THE

PAI5 AMA?!u'g
00
-

ANUAL 5

o t' 1

.
.7 s

,í" .

PRINCIPLES
D
i. rus,
ji.
41 a
-'
t
a

PRACTICE .-_ o
n r

FOR- THE %
F L-J}
WORLD ASOVE
s W.
s?1
50 MHz 1;' 4)

t ti

rrt

w
a , r . 4
a
. ,..;*1L
:
1

dt
. '

`.'
--7. A ''r1 'f ¡( .i
_y,.
''2`J,r
.---

- _ .. rf-_
 COPYRIGHT 1972 BY

THE AMERICAN RADIO RELAY LEAGUE, INC.

Copyright secured under the Pan-American Convention

International Copyright secured

This work is Publication No. 23 of The Radio

Amateur's Library, published by the League. All
rights reserved. No part of this work may be repro-
duced in any form except by written permission of
the publisher. All rights of translation are reserved.
Printed in U.S.A.

Quedan reservados todos los derechos

Third Edition
Second Printing
Library of Congress Catalog Card Number: 65-22343

$4.00 in U.S.A.
$4.50 elsewhere
Foreword
Probably no segment of amateur radio has moved so rapidly in
recent years as "The World Above 50 Mc" which is this book's
concern. Though The Radio Amateur's VHF Manual was introduced in
1965, and completely reworked in 1968, developments in the vhf
realm have come so fast, and in such bewildering variety, that this
third edition is already overdue.
Frequencies above 50 MHz have always been prime experi-
menter territory, but in the Seventies they are increasingly a com-
municator's world, as well. As such they are attracting an ever-higher
percentage of the world's users of the radio -frequency spectrum -
military, governmental, and commercial, as well as amateur. Just since
this book was first produced, we have seen amateur communication
via the moon become a worldwide reality on all amateur frequencies
from 144 to 2450 MHz. Amateur satellites have shown ever more
clearly that the days of "DX" (if the term means "distance" of the
conventional variety) are numbered. Who could listen to Apollo
conversations (frequently using channels close to our 2300 -MHz band)
and not realize that we are in the midst of nothing less than a
communications revolution?
The more conventional aspects of vhf and uhf hamming have
been moving with equal speed. MI-solid-state equipment of remarkable
utility has made the vest-pocket ham station a reality. Coming soon,
the "wrist radio"? Vhf and uhf repeaters make commonplace a degree
of 'round-the-clock reliability undreamed of even a few years ago.
Extensive use of ssb techniques has extended the range of voice
Communication far beyond the best we could manage heretofore. More
occupancy, more effective equipment, bigger antennas, and growing
propagation knowledge are showing new potentials of all our bands,
up through at least 10,000 MHz.
Obviously, it's time for a new VHF Manual. Edition Three is
full of new material, but it retains the down-to-earth readability of its
widely acclaimed predecessors. Like them, it is mainly the work of
QST's long-time VHF Editor, Edward P. Tilton, W1HDQ. Ed's exper-
ience span has seen the bands above 50 MHz all the way, from their
early status as get -lost room for a special breed of ham, to their
now -recognized stature as a major asset for all amateurs.
Assisting Ed with much interesting and practical information on
fm and repeaters is Douglas A. Blakeslee, WI KLK, for some years now
a fixture on the QST masthead, as Assistant Technical Editor. We
trust that you'll find them a good team.

JOHN HUNTOON, W1RW

General Manager, ARRL
Newington, Conn.
Ross A. Hull, vhf pioneer, and QST Associate Editor, 1931-1938.
Ross saw the potential of the then -uncharted world above 50 MHz
perhaps more clearly than any other man of his time. The technical
excellence of his equipment designs and his enthusiasm in print and in
person fired the imagination of a whole new generation of radio amateurs,
among them the author of this book. Hull's discovery and eventual
explanation of tropospheric bending of vhf waves has been called "one of
the truly outstanding examples of scientific achievement by an amateur in
any field of human endeavor."
CONTENTS
Chapter 1 How It All Started 7

2 A Vast Resource 14

3 Reception Above 50 MHz 30

4 Vhf Receivers, Converters and Preamplifiers 44

5 Vhf Transmitter Design 67

6 Vhf Exciters and Amplifiers 80

7 Vhf Stations 123

8 Antennas and Feed Systems 151

9 Building and Using Uhf Antennas 176

10 Fm - Theory and Techniques 218

11 Fm Transmitters, Receivers and Accessories . 235
12 Repeaters - Theory and Practice 263
13 Uhf and Microwaves 275
14 Test Equipment for the Vhf Station 306
15 Interference Causes and Cures 328
16 Bíts and Pieces 338
Index 350
 Abbreviations used in Text and Drawings
(Older form shown in parentheses)

A - ampere (amp.) muf - maximum usable frequency

ac - alternating current mV - millivolt
of - audio frequency mW - milliwatt
afc - auto matic frequency control nbfm, nfm - narrow -band frequency modulation
afsk - audio frequency -shift keying NC - normally closed
age - automatic gain control NO - normally open
a -m - amplitude modulation npn - negative -positive-negative
ATV - amateur television OD -
outside diameter
avc - automatic volume control osc -
oscillator
be - broadcast oz -
ounce
bci - broadcast interference PA -
power amplifier
BFO - Beat-frequency oscillator PEP -
peak -envelope power
ccw - counterclockwise pF -
picofarad (1..y1F - micromicrofarad)
CFM - cubic feet per minute PIV -
peak-inverse voltage
coax - coaxial cable pm -
phase modulation
COR - carrier-operated relay pnp -
positive -negative -positive
CRT - cathode-ray tube pot - potentiometer
et - center tap PRV - peak -reverse voltage
cw - continuous wave PTT - push -to-talk
dB - decibel rcvr - receiver
do - direct current rf - radio frequency
dpdt - double -pole double -throw rfc - radio -frequency choke
dpst - double -pole single -throw RFI - radio -frequency interference
dsb - double sideband rms - root -mean -square
DX -
long distance s.a.s.e. - stamped self-addressed envelope
EME -
earth -moon -earth, moonbounce SNR - signal-to-noise ratio
FCC -
Federal Communications Commission spdt - single -pole double -throw
FET -
field-effect transistor spst - single -pole single-throw
ft3/min -
cubic feet per minute (CFM) ssb - single sideband
fm -frequency modulation SSTV - slow -scan TV
GDO -grid-dip oscillator SWR - standing -wave ratio
GMT -Greenwich Mean Time TE - transequatorial (propagation)
H -Henry (hy.) tpi - turns per inch
hf - high frequency T-R - transmit-receive
Hz - hertz (cycle) TV - television
IC -integrated circuit TVI - television interference
ID -inside diameter usb - upper sideband
i -f -
intermediate frequency uhf - ultra-high frequency
JFET - junction field-effect transistor V - volt
k - kilo VCO - voltage -controlled oscillator
kHz - -
kilohertz (kc, kilocycle) VCXO - voltage -controlled crystal oscillator
kW - kilowatt VFO - Variable -frequency oscillator
If - low frequency vhf - very -high frequency
LO - local oscillator VOM - volt-ohmmeter
lsb - lower sideband VOX - voice -operated break-in
luf - lowest usable frequency W - watt
mA - milliampere WAC -
Worked All Continents
MARS - Millitary Affiliate Radio System WAS -
Worked All States
mH - millihenry ww -
wire wound
MHz - megahertz (Mc. - megacycle) wv -
working voltage
mic - microphone xtal -
crystal
mix - mixer p -
micro (10-6)
MOSFET - metal -oxide semiconductor field-effect transistor pF -
microfarad (uf.)
ms - millisecond µH -
microhenry (uh.)
m.s. - meteor scatter
 Chapter 1

How It All Started

Those of us who make the frequencies were the order of the day. The wonder was
above 50 Mc. our principal stamping ground that hams of the middle Twenties made gear
tend to think of v.h.f. as the "new frontier" of work on 7 or 14 Mc., let alone 5612.3,4
amateur radio. Actually it is as old as the art of This was c.w., remember. You had to chase
radio communication itself. While universal use a wandering signal with a receiver that was
of the upper reaches of the radio spectrum is a touchy beyond belief, even with stable signals.
fairly modem phenomenon, some of the earliest But 5 -meter gear was made, and it wórked after
work with electromagnetic radiation, and per- a fashion. When the first pioneers heard one
haps the first actual communication by radio, another across town they were ready for a shot
were on wavelengths near our present 2 -meter at Australia, or Europe. This technique had
band. brought results before on lower frequencies,
The resonator of Heinrich Hertz, and the why not on 56 Mc.?3 Then came several years
practical applications of it by Marconi, operated of largely fruitless effort. There were scattered
around 150 Mc. And if you think that the "heard" reports, some rather dubious in the
beam you are using is a recent development, light of present knowledge of v.h.f. propagation,
consider the fact that Hertz used a rudimentary but rarely was there two-way communication
form of Yagi in 1888, and Marconi employed a over more than a few miles.
parabolic reflector to extend the range of his By 1928, interest lagged. There were rumb-
first equipment before the turn of the century. lings of DX on our new band at 28 Mc. The
But Marconi and a generation of radio pioneers DX drive tended to move lower in frequency,
to follow him moved to the longer waves to and for about two years the world above 50
achieve greater coverage. The ultra -high frequen- Mc. was inhabited mainly by experimenters,
cies lay dormant for 20 years thereafter. rather than communicators.
The Drive for DX
ARRL and QST had been in existence for Working at 5 Meters
nearly ten years before frequencies higher than By S. Kruse. Technical Editor
15 Mc. were discussed in any detail. Trans- LAST month I aaid that ordinary cult I, shown because It la almple. 'Saris
methods worked perfectly well down
mitters using vacuum tubes had replaced spark to 20 meter. but apeeial can war
feed is used because this la to be a loose-
coupled net and therefore no harm will
rigs in the early '20s, and the unfolding possi- needed below that. Since that time come from aerie. feeding. Shunt feed can
hundreds of stations have been work- be used but there will be more trouble in
bilities of DX on wavelengths below 200 meters ing at 40 and 80 meten, not very many at making the chokes work wail.
20 meters-and very few indeed at 5 meten. Now we are down to a very mall helix,
caused adventurous amateurs to probe ever high- Most of the 20 and 5 meter work has failed and no capacity except that of the tube-
beeauae of an unsteady wave which could can we make this thing oscillate!
er in frequency. Each move upward produced not he read, altho very strong at the n -
The moral is to make a 5 -watt tube Tb. Circuit
new miracles of DX, culminating in worldwide work steadily rather than to make a 250-
watt tube work unsteadily.
The complete circuit is shown in Ftig. 2.
A little etody will show this to be the same
communication with low power - in daylight- Getting Dew.
circuit a, in Fig. 1, with the addibon of
the radio frequency chokes needed to make
when gear was finally made to work in the I aleo advised the use of one tube only.
This was correct at 40 meters, in still more
the tube osillate.
To tell when the net is oscillating the
14 -Mc. region. The next band then open to
amateurs was at 56 Mc. It was widely assumed
that if workable 5 -meter gear could be built,
this band would be even better for DX than 80,
40, and 20 meters had progressively turned out
to be.
Just getting there was thought to be the
principal problem. Technical Editor R. S. Kruse
a
pointed the way in the October, 1924, issue of
QST, with "Working At 5 Meters," perhaps the r MET=
A a
At M. lull la tbm
OSCILLATOR SET COMPLETE
b.W ud aaat I. It Me W.4.. CJ.r tub. .ida Its cana..
.oed.aa.r.. Buyout Wt I. Me Acme alum. traaalwew, 1b W«t.a
first v.h.f. constructional article ever published.' a..L n.
Dd.C. 1u.t,,.m.nt e..t to :47.0.12-e1. wt. wt. t -.w la .,, u.. u.
In the next few years much QST space would Slart tr.aalrr.4
.'"u. .é
cL<ult el tb.
the .4bi . of aaW Tí. ca.cW
trb«tawt Í.
Ta. circuit V *hewn la F4. r.
la ü.WMe.rr

be devoted to 5 -meter gear, but trying to use it important at 20 meten, and it is almost out :simplest teat is to touch the piste coil with a
was a frustrating business. Transmitters were of the question to make several American
mists work in parallel at 5 meters. The
wood handled screwdriver. Be careful-
the buns from even a 5-watt oscillator
simple oscillators; stabilization of any kind was meson for this is the insistence of our tube
makes in bringing all terminals out in a
are pretty painful.
If the ncrewdriver does not apark try
all but unknown, and not even considered for bunch at one end of the tube. If they raining the plate and filament voltage a bit,
would only bring the plate out somewhere then try putting the plate and gruff tarn.
5 -meter rigs. Receivers were regenerative detec- else would little t q. The t closer to they. g the screwrver
tors - hard to get going at all, and then in- nyr a l

credibly cranky to tune. Oscillators using de-

based tubes, mounted bottom-up to reduce lead In probably the first v.h.f. constructional article
inductance, and receivers with foot-long insula- ever published, Technical Editor Kruse described a
5 -meter oscillator in QST for October, 1924. The
ted tuning shafts to hold down hand capacity,
oscillator tube, barely visible just to the right of
The footnotes refer to the bibliography at the the tank coil, was a debased C-302 resting bottom -
end of the chapter. up on its glass envelope.

7
8 HOW IT ALL STARTED
Short -Range Phone Does It and it used up more than its share of high -
Up to this time, most amateurs were code
priority kilocycles. But there was plenty of
men. Phone was coming in, but it was frowned room on 5, and the unstable signal didn't sound
bad at all when received on the broad -tuning
on as wasteful of frequency and was often
treated as an unwanted stepchild. Then a few "rush -box."
u.h.f. experimenters began modulating their rigs, Soon it was found that these two castoffs
had something else in their favor: If stations
were not too close together in frequency, their

, ...-. .19:511 ia. s_,wa

- a orr m é .n
ák `_gi+3
e w
-
transmitters and receivers could be operated sim-
ultaneously. This was "duplex phone," a wholly
new concept. Two hams could converse as easily
by radio as over the telephone, or face to face.
--w áooes N. Duplex was an overnight sensation, with obvious
advantages over c.w., and the monologue voice
technique then used on lower bands.
The ill wind of economic depression blowing
across the land had made thousands of hams idle.
With much time and little money, they were ripe
for a kind of hamming that could be carried on
with makeshift gear, largely made by hand or with
parts robbed from discarded radio receivers. Not
only from junk sets, either; as 5 -meter interest
boomed, more than one family's radio listening
time was rationed, while the ham of the house
A utodyne receiver used by W8AZL in pioneering reddened the plates of Type 45 or 71A tubes,
5 -meter work with W8PK and W8ABX. V.h.f. lifted from the broadcast set for service in a
adaptation of the superregenerative detector, 5 -meter oscillator. Simple low-cost gear; duplex
which was to popularize v.h.f. operation in a big phone; the thrill of something new, yet within the
way, was still a year away. From September, 1930, reach of nearly everyone - these were magnets
QST. that drew countless newcomers, including the
author of these lines, into amateur radio in the
and unwittingly triggered off a boom that was early Thirties.
to establish the 5 -meter band as desirable com- As all through the history of the hobby, QST
munications territory in the minds of a whole struck the spark. The July, 1931, issue was fat with
new generation. v.h.f. lures. Technical Editor Lamb had 11 pages
The experience of early 5 -meter phone ex- on u.h.f. oscillators,7 some working as high as 400
perimenters John Long, W8ABX, and E. O. Mc., where some farsighted administrator had set
Seiler, W8PK, was typical. They were working aside a narrow band for amateur experimenters.
on 80 one night in the summer of 1930, when Associate Editor Ross A. Hull, who would become
a thunderstorm not far away made communica- one of the v.h.f. man's legendary heroes, fanned
tion difficult. W8ABX was running his 5 -meter the flame with down-to-earth 5 -meter receivers,8
rig simultaneously, and when W8PK listened on adding reports on mobile receiving tests, for good
5 he found, to his amazement, that signals were measure. Hull's "Duplex Phone on 56 Mc." in
far clearer than on 80. Here, for the first time, Augusts described more simple gear, and set forth
the 5 -meter band was seen to have real worth: 4
it would work over short paths with voice,
when noise levels, high activity, or other adverse
factors were present on lower frequencies.5
Receivers were still a bottleneck, however.
The regenerative receiver, critical enough on any
frequency, was an operator's nightmare at 56 -
Mc. Enter here the superregenerative detector.
Invented years before, the superregen had not
found much favor with amateurs.6 It was use- .
less for c.w., and its broad frequency response
and raucous audio quality gave it a bad name in
voice work. But on 56 Mc., where there was
band width to bum, broad tuning was almost a
.
blessing.
In retrospect we can see several factors
combining to accelerate 5 -meter interest as ham
radio moved into the Thirties. The wedding of
Cover picture from August, 1931, QST shows the
the modulated oscillator and the superregen- modulated -oscillator transmitter that helped trigger
erative detector would start things rolling again. the 5 -meter boom of that eventful summer. Two
The modulated oscillator sounded awful on sel- 71As in a push-pull oscillator were modulated by
ective receivers now in use on lower frequencies, parallel 47s.
Short -Range Phone Does It 9
the operating concept that was to build the fire to This sort of path had never been bridged on 5
conflagration proportions. meters. The best DX previously worked by Ross
The urge to work around the world was and his associate, W1ANA, and members of the
forgotten; the aim now was to work across town, Headquarters Staff who worked with them at
reliably, on voice. If 5 -meter waves traveled only in Selden Hill, had been a portable on a hilltop 35
straight lines, then we would get up on the hills to miles away. All hands were justifiably excited,
extend our horizons. "Working portable" became a therefore, when use of the new directive array
great weekend outdoor sport wherever there were produced two-way contacts with many stations in
hams and mountains. Duplex suggested voice relay- Eastern New England, 75 to 125 miles distant.
ing, and hilltop stations worked up haywire patch- Operating on a 24 -hour basis, the gang soon
ing systems, to hook valley dwellers up with found that although signals of some sort were
stations they could not hear direct. Message relay- nearly always receivable over at least 100 miles,
ing also boomed, and on weekends, points as far there was a tremendous variation in level. A signal
apart as New York and Boston were connected by would all but block the receiver at times, and then
snappy relay circuits, thanks to three or four a day, an hour, or even minutes later it would drop
hilltop portables along the way.10 almost into the noise.
Why? At this juncture what had been little
more than a lark to the enthusiastic Hull took on
? } the aspects of a long-term challenge. He set up
-1. J.
schedules around the clock, and kept them relig-
£ iously. He scanned reams of data on all kinds of
natural phenomena. Eventually he developed a
photographic recording technique, especially to
t compile a running record of the signal variations of
yr experimental station W1XW, at the Blue Hill
r}
.r
Observatory, near Boston. Records were plotted
against lunar cycles, weather data, temperature and
barometric pressure curves - anything that might
affect 5 -meter propagation.12
We won't attempt to retell the story here, for
you can find it all in Ross's own words in QST.
(See bibliography at the end of this chapter.) Read
Hull's account for yourself, and see why his
discovery of air -mass boundary bending of v.h.f.
waves, and eventual development of a theory to
' 4,
explain it,13 has been called "one of the outstand-
ing achievements by an amateur in any field of
Another QST cover shot shows Managing Editor scientific endeavor."
W1SZ hauling up the 8 -element beam that was to DX of a more remarkable variety was noted in
make 5 -meter history at Selden Hill in the fall of the spring of 1935.14 One day in May, 5 -meter
1934. men in eastern cities were amazed to hear strangers
in their midst; fellows signing W8 and W9 calls.
Michigan, Ohio, Illinois - on 5 meters? Impossible
DX Again
- must be bootleggers trying to pull a fast one!
But contacts were made and QSLs exchanged; the
Almost from the first widespread 5 -meter signals were genuine. It happened more often in
activity in the summer of 1931, it was seen that 1936, and soon everyone was looking for 5 -meter
signals did not always travel strictly in straight DX, as word spread like wildfire whenever the
lines. Stations not within pure line -of -sight range band opened up. Most contacts were between 500
could communicate at times, and the degree to and 900 miles, with some out to 1200. A pattern
which this was possible was seen to vary. But most for the propagation was evolved by Harvard Uni-
of us were too busy trying to improve our versity's J. A. Pierce, W1JFO, in a scholarly
equipment to pay much attention to propagation. analysis of 5 -meter DX in September, 1938,
Then, in the summer of 1934, Ross Hull began QST.15
experimenting with beam antennas.11 Hanging a In the same issue was the biggest DX news yet:
stick -and -wire system of 4 half waves in phase, details of a transcontinental QSO between
with reflectors, over the porch roof at Selden Hill, W1EYM, Fairfield, Conn., and W6DNS, San Diego,
West Hartford, Ross fired up a 200 -watt oscillator Cal., a record that would stand as long as we had
and blazed away in the direction of Boston. Selden the 5 -meter band.16 Dr. Pierce charged this all up
Hill was a fine location, with a clear sweep across to sporadic ionization in the E region of the
the Connecticut Valley, but it was less than 300 ionosphere, and pointed out some possible causes.
feet above sea level. The horizon, 12 to 20 miles His writings and those of Ross Hull are real
distant, was solid with hills 1000 feet and more in milestones in the amateur's contribution to radio
elevation, for 50 miles or so, before the country- propagation knowledge. We can be proud that they
side sloped down to the coastal plain of Eastern came years in advance of similar work by scientific
Massachusetts and Rhode Island. atid governmental agencies.
10

.
c,
n>n
i .,.f`
.. i.....s
HOW IT

:i;-
-
ALL STARTED

Nathaniel Bishop, W1EYM, Fairfield, Conn., and Harold Hasenbeck, W6DNS, San Diego, Cal., made
the first transcontinental 56 -Mc. QSO, July 22, 1938.

Meanwhile, v.h.f. was growing up. Power was A few gave up ham radio, rather than convert to
increasing; receivers were getting better; beam the more complex stabilized transmitters. Others
antennas were becoming common. QRM was get- dropped to lower bands. But a sizeable number
ting worse steadily, and it became obvious that moved higher, and picked up on 112 Mc. where
something must be done. That "something" was they had left off on 56. Using largely the same
the elimination of the broad signals radiated by techniques that had served so well on 5, these
unstable transmitters, so that selective receivers fellows and many newcomers went through a cycle
could be used effectively. The day of the simple rig of activity and development reminiscent of early
was drawing to a close, and effective December, 5 -meter days. Almost anything that had been done
1938, FCC required amateurs on the 56 -Mc. band on 5 was repeated on 2 1/2, except for the working
to meet the transmitter stability requirements of long distances via ionospheric propagation.
imposed -on lower frequencies. This brought a There was even considerable experimentation on
wonderful era to an end, but it started another. 1 1/4, and interest in both bands was still rising at
the time of the World War II close-down, Decem-
Simple Gear Moves to 112 and 224 Mc ber, 1941. Freed of their severe QRM problem
Some 5 -meter men saw in the stabilization after 1938, 5 -meter workers concentrated on im-
regulations the death knell of their favorite band. proving receivers, transmitters, antennas, and op-
erating techniques. Reliable operating ranges
stretched out to 200 miles and more, for the better
stations.
One more DX propagation mode would be
brought to light before the wartime close -down.
Who did it first, where, or when, is not precisely
known, but the first rumblings were heard in 1937.
"Rumblings" is the right word, for these weird
signals were characterized by terrific distortion,
rendering modulation of any kind almost unread-
able. Often the signal was little more than a rumble
or roar. At first nobody knew what was up, but
before long it was found that distortion and signal
strength peaked when antennas were aimed north.
Eventually this happened on a clear night, and all
across the northern sky was seen the eerie glow of
aurora borealis! Not much was known about the
nature of the aurora then, but amateur 5 -meter
observations were put to good use in studying it.
Until the wartime cessation of activity, and again
after 1945, amateur reports gathered by ARRL
contributed in a significant way to increased
understanding of auroral phenomena.17
The Modern Era
Except for limited use of the 112 -Mc. band in
the War Emergency Radio Service, all amateur
Two of the famous "Gil" cartoons in QST for activity came to an abrupt end on December 7,
December, 1938, summed up the v.h.f. aspects of 1941. Military communications and radar, mean-
new FCC regulations then becoming effective. while, expanded all through the spectrum, employ-
Worldwide V.H.F. DX at Last! 11
ing frequencies and techniques hardly dreamed of ing LU6DO, Temperly, Argentina, and other LUs
by most hams. Especially in v.h.f. and microwaves on 50 Mc. These contacts were made later in the
great strides were made during the war period, and day than F2 -layer predictions called for. Some-
many among us doubted that amateur radio could times propagation lasted well into the evening
ever catch up. But hams were in the thick of it, in hours, an unheard-of thing on frequencies this
laboratories and in the field, and they learned their high. Eventually labeled transequatorial scatter,
lessons well enough to be off and running when the this mode of propagation rates as one of the
war was over in the fall of 1945. A complete outstanding discoveries in amateur v.h.f. history.28
reshuffling of our allocations had been made, and
after a temporary start on the old bands at 56 and The New Bands Prove Their Worth
112 Mc., we moved to new assignments at 50, 144, Progress on 144 Mc. and higher frequencies
and 220 Mc. We also now had bands at 420 Mc. was also notable in the early postwar years. Aided
and at intervals all through the assigned portion of by the availability of the SCR -522 and other
the microwave spectrum. There was work to do! military communications units on the surplus
The change from 56 to 50 Mc. was especially market, 2 -meter men converted largely to stabil-
intriguing, in view of the rising solar activity curve. ized equipment and selective receivers, and opera-
Would the new band "open up" when we reached ting ranges expanded rapidly. Tropospheric propa-
the top of the sunspot cycle a year or two hence? gation was found to be more favorable on 144 than
By now, scientists were making predictions as to on 50 Mc., and the record for two-way work was
the maximum usable frequency for F2 -layer propa- extended kradually, reaching 1400 miles by 1951.
gation, but they were not overly optimistic. The Auroral communication was found to be possible
best guess was that 50 Mc. was a bit too high. on 144, and this mode provided much exciting
144 -Mc. DX. Exploitation of the reflecting proper-
Worldwide V.H.F. DX at Last! ties of ionized trails of meteors opened the way to
Fortunately, most v.h.f. men did not know more 2 -meter DX. Two leaders in this field were
about these predictions. Noting that British TV W4HHK, Collierville, Tenn., and W2UK, New
signals on 45 Mc. could be heard now and then,
and hearing from keen observers across the Atlan-
tic that American signals and harmonics were
filtering through in Europe on frequencies as high
as 47 Mc. on occasion, amateurs set up test
schedules in the fall of 1946. On mornings when
conditions appeared favorable, American 50 -Mc.
men transmitted toward Europe, listening for I 120 Pi
replies on 28 Mc., there being no 50 -Mc. band in 110
Europe. 4p.
Just before noon on November 24, 1946, a
test transmission by W1HDQ, West Hartford,
Conn., brought a frantic 'I'm hearing you on 50
megacycles!" from G6DH, Clacton-On -Sea, Essex,
England, and the first v.h.f. communication across
the Atlantic was on. G5BY, near Plymouth, heard
the test at the same time, joining G6DH in the
transatlantic cross -band QSO a few minutes later. {
Shortly after noon the same day, W4GJO, Orlando,
Fla., worked W6QG, Santa ANA, Cal., for the first
transcontinental 50 -Mc. F2 -layer QSO. Pacific DX ,
came in January, 1947, when KH6DD worked
J9AAK, Okinawa, extending the 50 -Mc. DX record
-ay
to 4600 miles.
In March, 1947, W4IUJ, West Palm Beach,
Fla., worked OA4AE, Lima, Peru, thereby winning
the Milwaukee Radio Club trophy for the first
two-way intercontinental v.h.f. QSO, the cup
having been resting at ARRL Headquarters for
nearly ten years. Crossband DX of phenomenal 1

proportions came at about this same time, resulting

from checks made by PAWN and PAOUM with
ZS1P and ZS1T, 6000 miles to the south. August, imp
1947, brought a new two-way record, 5300 miles,
between W7ACS/KH6 and VKSKL. Eileen and Denis Heightman of G6DH, British end
of the first transatlantic v.h.f. OSO, November 24,
Though not credited as such at the time, this 1946. Receiving W1HDO on 50 Mc., Denis replied
probably was the first DX QSO via a propagation on 28 Mc. The following year, operating with
medium that was to be exploited later on the special temporary authorization, G6DH was the
Mexico-to-Argentina path. Around the end of first British station to work two-way across the
August, 1947, XEIKE, Mexico City, began work - Atlantic on 50 Mc.
12 HOW IT ALL STARTED
4+- -ONE CYCLE, 6 SECONDS -L- SECOND CYCLE--
hi- -PULSE--y=4-N0ISE --ECHO--1.4+--NOISE---1.-1
KEYING I
¡TNANSIENT
1

- s ., 7..
-
,

-
Visual record of an historic achievement the first amateur signals sent to the moon and back. After
three years of work, Ross Bateman, W4AO, and William L. Smith,W3GKP, shown here checking
alignment of the huge stacked -rhombic array at W4AO, finally received echoes of their 144 -Mc. signa I

reflected from the moon. The date: January 27, 1953.

Brunswick, N.J., who received the ARRL Merit successful use of the moon for the reflection of
Award of 1955, for their outstanding meteor - amateur signals was accomplished by W4AO and
scatter work of 1953 and later. W3GKP in January, 1953.19 These two used
The 220 -Mc. and 420 -Mc. bands had appeal for advanced techniques on 144 Mc. to demonstrate
the experimentally inclined, and were soon shown that lunar communication was at least a possibility
to have great value for practical communications for amateurs. Two-way communication via the
purposes as well. Development of efficient equip- moon was a long time coming, and was finally
ment and high -gain antennas showed that these achieved first on 1296 Mc. The work of W1BU and
bands were capable of reliable coverage nearly W6HB in communicating over 2500 miles by way
approximating that of 50 and 144 Mc. of the moon in July, 1960, is a notable mile-
Making use of tubes and components largely stone.20
salvaged from war -surplus radar and navigational Propagation know-how paid off markedly for
equipment, amateurs developed workable com- W6NLZ, Palos Verdes Estates, Cal., and KH6UK
munications gear for all our microwave bands (also W2UK, mentioned above), Kahuku, Hawaii,
before the end of 1946, and in later years were when they were able to work across 2500 miles of
able to extend communications distances out to the Pacific on 144 Mc. in July, 1957.21 This was
several hundred miles on nearly all our u.h.f. and the longest path ever covered by tropospheric
s.h.f. bands. Development along these lines con- means by any communications service, and as such
tinues to this day. it achieved worldwide acclaim. Not satisfied,
Intrigued by the possibilities of weak -signal Chambers and Thomas went on in subsequent
work, amateurs made notable strides in utilizing years to bridge the path on 220 and 432 Mc. '!heir
various marginal modes of propagation such as superb work won for them the Edison Award for
moon reflection, ionospheric and tropospheric 1960, the only instance in which this award was
scatter, and even satellite communication. The first given for scientific accomplishment.
The New Bands Prove Their Worth 13
Thus we nave touched lightly on some high-
lights of amateur radio's long history of pioneering
the use of frequencies once thought to he useless ,
.r.;1
for any practical purpose. It is well for all of us,
hams of the present and future, that we have this
record of achievement behind us. In the years to 2z1,
come, the pressure on all frequencies above 50 Mc. r
is certain to rise, as if it were not already high
l)
,r
enough. Every kilocycle, even to frequencies only '. '^ - e3
dreamed of a few years ago, is eyed eagerly by "11;s;:
many users 'of the radio spectrum. We have shown,
from the earliest times, that it is good for everyone 't a`rwii! 4Li». i3, le: 17 I,.

that amateurs have access to samplings of the

radio -frequency spectrum, from bottom to top,
whatever that may be. To continue to merit the
confidence and support of the people and agencies
who will decide future allocation of frequencies
should forever remain one of our highest aims.
_

Historical QST References

1 Working at 5 Meters . . . Kruse, October,
1924. ~
Jones, 6AJF, a,
2 Pioneer Short Wave Work . ..
'

May, 1925.
3 Experimenter's Section . . . 1925 to 1928.
4 Gear for wavelengths down to 3/4 meter . The first amateur microwave station. A. E. Har-
. January, 1926; August, 1927.
.
rison, W6BMS/2, and Reuben Merchant, W2LGF,
5 Making Practical Use of the 56 -Mc. Band .
built two of these stations and had them ready for
. Long, W8ABX, September, 1930.
communication on November 15, 1945, the day
6 Superregeneration ,. July through Oct- that our microwave bands were opened to amateur
ober, 1922. use. Frequency: 5600 Mc.
7 Developments in U.H.F. Oscillators
.

. Lamb and Hull, July, 1931.

8 Five -Meter Receiving Progress . . . Hull, An All-Purpose 56 -Mc. Station . . . December
July, 1931. 1932. Summaries of activity appear throughout
9 Duplex Phone on 56 Mc. . . . Hull, August, 1932 issues.
1931. Behavior of U.H.F. Waves . . . Jones, March,
10 Progress reports and tests . . . January, 1933.
May, July, September, October, November; 1931. Graduating to Oscillator -Amplifier Transmitters
Fundamental Crystal Control . . . April 1932. for 56 Mc. . . Griffin, W2AOE, May, 1933.
.

Fun on 5 Meters . . . June 1932. Firing Up on the Newly-Opened Ultra -High

Frequencies . . . Hull, September and Nov-
ember, 1934.
11 Extending the Range of U.H.F. Stations .
. Hull, October and December, 1934.
12 Air-Mass Conditions and the Bending of
U.K.F. Waves . . . Hull, June, 1935.
13 Air -Wave Bending of U.H.F. Waves .
Hull, May 1937.
14 Five -Meter Signals Do the Impossible .
. August, 1935, was first published report of
authenticated 5 -meter skip. July issues of 1936 and
1937 contain summaries of reported DX.
15 Interpreting 56 -Mc. DX . . . Pierce, Sep-
tember, 1938 (E -layer theory)
"Further Reports of 50 -Mc. DX . . Sep- .

tember, 1938.
17 Moore, `Aurora and Magnetic Storms,"
June, 1951.
18 Cracknell, "Transequatorial Propagation of
V.H.F. Signals," December, 1959.
19 Tilton, "Lunar DX on 144 Mc.," March,
1953.
20 September, 1960.
John T. Chambers, W6NLZ, center, and Ralph
Thomas, KH6UK, right, receive Edison Award 21 September, 1957, p. 62.
trophies from General Electric vice-president L. Regular coverage of the v.h.f. and higher bands,
Berkley Davis, in Washington ceremony, February On the Ultra-Highs , began in December, 1939.
23, 1961. Award was in recognition of the trans- Later called The World Above 50 Mc., it has told
pacific communication by these outstanding ama- the month -by -month story of amateur v.h.f. pro-
teurs on 144, 220, and 432 Mc. gress ever since.
 Chapter 2

A Vast Resource
AMATEUR BANDS ABOVE 50 MHz
Thee true extent and worth of the frequencies by modulation methods. There are also restrictions
above 50 MHz that are available to amateurs are as to who may use some frequencies, depending on
rarely appreciated, even by those who spend most the class of license held by the operator. This
of their operating time there. The various vhf, uhf, information is subject to change. Though the table
and microwave bands are listed below, with the is accurate for the publication date, early 1972,
emissions that may be used in each. It will be seen readers should consult the latest edition of the Radio
that there is considerable subdivision of the bands Amateur's License Manual for current restrictions.
50 to 54 MHz 420 to 450 MHz*
50.0 to 50.1 MHz - Al (cw telegraphy) only. All above modes, plus A5 (television, slow or
Advanced and Extra Class licensees only. fast scan).
50.1 to 54 MHz - Al, A2 (tone -modulated
telegraphy), A3 ,(amplitude modulation and 1215 to 1300 MHz
narrow -band fm), A4 (facsimile). All classes All above modes.
except Novice.
51.0 to 54 MHz - AO (unmodulated carrier; 2300 to 2450 MHz
duplex operation), plus above. All above modes, plus pulse.
52.5 to 54 MHz - Wide -band fm, plus above.
3300 to 3500 MHz
144 to 148 MHz
All modes.
144.0 to 144.1 MHz -
Al only. All classes
except Novice and Technician. 5650 to 5925 MHz
144.1 to 148 MHz - AQ,= Al, A2, A3, A4.
All modes.
Wide-band or narrow-band fm. All classes
except Novice and Technician. 10,000 to 10,500 MHz
145.0 to 147.0 - All license classes. Novice
All modes except pulse.
must use code only, no more than 75 watts
input, and crystal control. No frequency 21,000 to 22,000 MHz**
above 147 MHz may be used by Novices.
All modes.
220 to 225 MHz*
40,000 MHz and all higher
All above modes. All classes except Novice.
All modes.
* All amateur frequencies above 220 MHz are
shared with Government Radio -Location Service ** Subject to change. World Conference on
which has priority. Operation in the 220-MHz band Space Communication Summer, 1971 assigned
is restricted in parts of Texas and New Mexico. 24,000 to 24,250 MHz in place of 21,000 to
Final -stage input in the 420 -MHz band is limited to 22,000 MHz. 'frequencies were also assigned for
50 watts in Florida Arizona, and parts of Ala- amateur transmitting satellites, as follows: 7.0 to
bama, Texas, New Mexico, Nevada and California, 7.1 14.0 to 14.25 21.0 to 21.45, 28.0 to 29.7
as set forth in Part 97.61 of the LIS Regulations. 144 to 146, 24,000 to 24,050, and (on shared
Permission to use 1000 watts may be obtained by basis) 435 to 438 MHz. Effective date of these
amateurs in restricted areas by individual appli- assignments not known at publication date. Watch
cation to FCC. QST for announcement.

WHAT CAN WE DO HERE?

In terms of band widths and potential occu- Our historical review, Chapter 1, emphasized
pancy by amateurs, this is a world almost beyond the amateur's role in uncovering the true worth of
comprehension. As seen in Fig. 2-1, amateur bands the vhf bands. Our potential in this field is far from
from 80 through 10 meters, which carry most of exhausted. Though great scientific strides have
the occupancy load, total 3300 kilohertz -less been made, by no means all is known of the ways
than the width of any vhf band. Vhf assignments by which signals in the vhf and higher frequency
include 13 MHz, or almost four times the frequen- ranges are propagated to distant points. Nature still
cy spread of all lower amateur bands combined. surprises even the best -informed amateur, and
The 420 -MHz band is wider than the entire admittedly this is a factor in the appeal of the
spectrum from do to the top of the 10 -meter band. world above 50 MHz. Knowing something of
Each band above 1000 MHz is wider still. Our propagation media we can, however, take advan-
inability to show these figures in scale is worth tage of the opportunities nature affords us, and we
remembering when we worry over congestion in will enjoy our work more and do it better than if
the amateur bands between 3.5 and 30 MHz! we merely take what comes our way, without

14
What Can We Do Here?
15
question or observation. Here are some propaga-
tion tips, band by band. '58
'48 '; i=r-r-91
50 MHz '37
,1
i;
i
¡I¡
68
Perhaps no band is more interestingly placed in
the radio spectrum than this, from the standpoint
;_
ll soWAS
Mc
of propagation vagaries. Working in borderline
territory between the "DX bands" and those
normally considered useful mainly for local com-
munication, the 50-MHz enthusiast samples both
33 '44 '54 '64

Z
75?

_:
;
I
Y
FP;1

WAC
so me
worlds. Though DX is not his daily lot, he will see,
at one time or another, nearly every known form
of long-distance propagation. His reliable range
with moderate power and relatively simple equip- 1 nw:
T3 rir
ment is likely to equal anyone's, for the 50 -MHz
region is less susceptible than lower frequencies to
adverse effects that tend to break up or impair
V,
communication. Consistent coverage over a radius
of 100 miles or more is not unusual, and this can
be extended considerably by use of optimum reflections from meteor trails afford the proficient
equipment and communications techniques. operator chances for work over 600 to 1200 miles
Variety is frequently afforded by tropospheric on a regular basis on 50 MHz. Transequatorial
bending, which extends local coverage by two to propagation is good for several thousand miles, in
three times the normal. Sporadic -E skip offers DX low latitudes and in periods of high solar activity.
in the range of 400 to 1200 miles or so, and More about each of these modes later.
multiple -hop effects may extend this up to 2500
miles or more on occasion. Auroral propagation to 144 MHz
all distances up to 1000 miles is fairly common in Except that it lacks some of the long-distance
the high latitudes. DX via the F2 Layer may be ionospheric possibilities, the 144 -MHz band is not
possible during the peak years of the sunspot cycle, unlike 50 MHz. Tropospheric propagation tends to
providing contacts at distances of 2000 miles and improve with frequency, so 144 MHz is superior to
more. F2 -layer backscatter fills in the shorter 50 in this respect. Whereas tropospherically propa-
distances at these times. Ionospheric scatter and gated 50 -MHz signals are seldom heard beyond 300
miles, 144-MHz work out to 500 miles or more by
3.5-29.7MHz TOTAL:3.3MHz this mode is fairly common. Up to 1400 miles over
land and 2500 miles over water have been covered
by tropospheric bending on 144 MHz.
50-54 MHz 4 MHz Sporadic-E skip is rare on 144 MHz, though
lack of alert observers in the more favorable areas
may have caused us to miss some 144-MHz DX
144-148 MHz 4 MHz opportunities of this kind in the past. Auroral
propagation is quite similar to 50 MHz, except that
borderline conditions may show on the lower
220-225 MHz 5 MHz frequency and not on 144 MHz. Distances up to
Total, all VHF Bands:13 MHz 1300 miles have been covered, but 200 to 700
miles is most common. Use of cw is almost a
necessity because of the high degree of distortion
33 NHz All HF Bands produced by the auroral reflection.
Of the rare modes, meteor scatter and tropo-
spheric scatter have been most exploited by
13 MHz All VHF Bands 144 -MHz operators. Each requires fairly high trans-
mitter power, skill in the use of cw, and the best
possible receivers and antennas. Communication by
420 -MHz BAND 30 MHz way of the moon is just possible on 144 MHz, and
considerable progress has been made in EME work
in the last decade by W6DNG, K6MYC, VK3ATN,
I BANDS ABOVE 1000 MHz \TOTAL: 2220 MHz KOMQS, W2NFA, ZL1AZR, F8DO, SM7BAE, and
(70 times width of42o -MHz Bardj others.

Fig. 2-1 -
Amateur bands at the upper end of the rf spectrum defy portrayal in scale. At the top are
the three
our hf bands, which total 3300 kHz (3.3 MHz) in width. Next below, on the same scale, are
amateur vhf bands, each wider than all hf bands combined. and this in turn fails
A new scale is needed to show these bands in relation to our 420 -MHz band,
to indicate the scope of amateur assignments above 1000 MHz. These would require 70 strips the size
from dc
of the one shown for the 420-MHz band, which is itself wider than the whole rf spectrum
through 30 MHz!
 16 A VAST RESOURCE
220 MHz
This band is similar to 144 MHz in its tropo-
spheric propagation possibilities. The overland
record is about 900 miles, and the 2500 -mile path
from the West Coast to Hawaii has been bridged
with good signals. No ionospheric propagation has
been observed. Auroral conditions are less favor-
able than on 144, but some DX of this kind has
been worked, mostly under 700 miles. Communi-
cation by way of the moon was not accomplished
until 1970, but 220-MHz possibilities appear at
least equal to those on 144 MHz. More universal
activity is needed on this and all higher amateur
bands to assess their real worth for long distances.
n
420 MHz
Exploitation of this band suffered because of
the power restrictions imposed until a few years amateur radio on lower frequencies. Our micro-
ago, but it is now known that tropospheric wave assignments have tremendous potential for
possibilities are excellent. The terrestrial two-way point-to-point communication, and they might
record, 1210 miles from Kansas to Connecticut, is well supplement lower frequencies for scheduled
an example. The West -Coast-to -Hawaii path has work.
Pulse modulation is one means by which the
been covered one way on 432 MHz with strong
microwaves can be put to practical use by ama-
signals. Lunar and satellite possibilities appear
teurs, this being usable on all our frequencies above
better than on lower frequencies, and distances as 2300 MHz except in the 10,000 -MHz band.
great as New England to Hawaii and California to
The existing DX records for each of our bands
Europe have been covered, by way of the moon.
above 50 MHz are listed below:
1215 MHz and Higher
Terrestrial Two -Way Records
Though not fully exploited by amateurs thus
far, the frequencies above 1000 MHz offer vast 50MHz: LU3EX JA6FR -
opportunities for interesting work. Here is the true -
12,000 Miles March 24, 1956
"frontier" - a wofld we must explore if the 144 MHz: W6NLZ -KH6UK
traditions of amateur pioneering are to be main-
tained. Distances up to 400 miles have been
-
2540 Miles July 8, 1957
220 MHz: W6NLZ -KH6UK
worked on 1215 MHz with low power under
conditions of tropospheric bending, and much -
2540 Miles June 22, 1959
greater distances certainly are possible. Reflection 420 MHz: WODRL KIPXE -
from the moon shows great promise. No two -
1210 Miles August 16, 1971
points on the earth's surface where the moon can 1215 MHz: W6DQJ/6 K6AXN/6 -
be seen simultaneously are beyond the possible -
400 Miles June 14, 1959
range of thIs mode of communication on these 2300 MHz: W4HHK -WA4HGN/4
frequencies.
Amateur experience in our microwave bands is
-
249 Miles July 11, 1970
too meager to permit us to assess their true 3300 MHz: W6IFE/6 K6HIJ/6 -
potential. Distances beyond line of sight have been -
214 Miles June 18, 1970
covered on all amateur frequencies up through 5650 MHz: K6HIJ/6 - W6OYJ/6
10,000 MHz, indicating the presence of tropo- -
214 Miles June 18, 1970
spheric bending. Necessity for use of high antenna 10,000 MHz: W7JIP/7 W7LHL/7 -
gains with resultant sharp beam patterns, almost -
265 Miles July 31, 1960
rules out the random operation that characterizes 21,000 MHz: W2UKL/2 WA2VWI/2 -
-
27 Miles Oct. 24, 1964

EME Two-Way Records

144 MHz: SM7BAE ZLIAZR -

-
11,055 Miles March 4, 1969
220 MHz: WB6NMT K2CBA -
MORE ACTIVITY
NEEDED
-
2650 Miles March 16, 1970
420 MHz: WA6LET G3LTF -
y -
5730 Miles Sept. 25, 1965
09;
1215 MHz: WB6IOM G3LTF -
""e, -
5492 Miles April 27, 1969
2300 MHz: W3GKP - W4HHK
840 Miles - Oct. 19, 1970
3.1
;,n -"k
Propagation Beyond the Horizon
17
PROPAGATION BEYOND THE HORIZON

Radio waves travel in straight lines unless

forced to do otherwise. In some respects, vhf
waves are less easily reflected than waves of lower
frequency, so consistent vhf communication with
low power tends to be essentially local in charac-
ter, covering only slightly more than line -of-sight
distances. There are many ways by which the wave
energy may be reflected, refracted, or scattered,
however, and the vhf man will do well to become
familiar with the principal ones at least. Some are
shown in Fig. 2-2.
Tropospheric Bending is the most common
form of vhf DX.1 Though observable on all radio
frequencies, it is most pronounced in the vhf range
and higher. It is the result of change in refractive
index of the atmosphere at the boundary between
air masses of differing temperature and humidity
characteristics. These boundaries occur in the first
few thousand feet above the earth, so their effect is
most prevalent at distances under about 150 miles,
though it may extend much farther.
Air masses often move on a very large scale,
retaining their original character over considerable Fig. 2-3 -A readily available guide to tropospheric
periods of time. A large mass of cold air of polar propagation conditions is a weather lines. map, showing
pressure distribution and frontal On the
origin may be overrun by warm air from the south. map appeared in eastern
When this happens, an inversion is said to exist, the October day that this
newspapers, the 2 -meter band was open from Nova
normal state of affairs being a 3 -degree drop in Scotia to at least North Carolina for several hours.
temperature for each 1000 feet of altitude. Such a
boundary may prevail for 1000 miles or more far inland. Though the exper-
along a more -or-less stationary weather front, mountainous terrain,
amazing DX in the vhf and uhf ranges. ienced meteorologist would call the following
producing of a complex picture,
There is an easily observed tie-in between visible advice an over -simplification
vhf coverage.2 Daily the vhf man should watch for slow -moving areas of
weather conditions and
concentrate on the
weather maps published in many newspapers and high barometric pressure and
often shown in rudimentary form in television trailing edges of such areas. See Fig. 2-3. Such
weather broadcasts may help the vhf enthusiast to favorable conditions are most
often observed in the
anticipate favorable propagation.3 Detailed early fall months.
weather maps may be obtained from the U. S. What is known as the U. S. Standard Atmos-
Weather Bureau on a subscription basis. phere curve4 is shown in Fíg. 2-4, left. The solid
with
Tropospheric bending is most common in fair, line is the normal decrease in temperature
the warmer months, though it can height. The broken line shows a relative humidity
calm weather of
feet. Figures
occur at any season. Atmospheric convection in of 70 percent, from ground to 12,000 vapor
coastal areas, over the Great Laeks Basin, or in the in parentheses are the ratio of grams of water
valleys of major rivers, produces the required to kilograms of air, called the mixing ratio. No
under
stratification of air, making these regions some- tropospheric bending would be observed
more desirable vhf territory than irregular these conditions. At the right are upper -air readings
what
Fig. 2-2 -The principal means by which
vhf signals may be returned to earth,
showing the approximate distances over
which they are effective. The F2 layer,
highest of the reflecting layers, may
provide 50 -MHz DX at the peak of the
11 -year sunspot cycle. Such communi-
cation may be world wide in scope.
Sporadic ionization of the E region
produces the familiar "short skip" on 28
and 50 MHz. It is most common in early
summer and late December, but may
occur at any time, and regardless of the
sunspot cycle. Refraction of vhf waves
also takes place at air -mass boundaries,
making possible communication over dis-
tances of several hundred miles on all vhf
bands. Normally it exhibits no skip zone.
18 A VAST RESOURCE
of an inversion over Toledo, Ohio, on a September E -layer vhf propagation is most common in the
evening some years ago. On this occasion 2 -meter months of May, June, and July. There is a shorter
signals were traversing the 750 -mile path from season in December and January, and the effect
Northern New Jersey to the Chicago area. may occur at random times throughout the year.
Particularly over water in the lower latitudes, The long and short seasons are reversed in the
and less often over land areas, something approx- southern hemisphere. Duration and extent of Es
imating a duct may form, in which vhf waves are openings tend to be greater in the long season.
propagated over very long distances, following the June is the peak month ordinarily, with country-
curvature of the earth in the manner of uhf energy wide openings lasting for many hours at a time at
within a waveguide. This ducting has accounted for this season. The early evening and before-noon
much of our extreme vhf and uhf DX. Notable hours are most productive.
examples are the spanning of the Pacific from The upper-frequency limit for sporadic E is not
Southern California to Hawaii on 144, 220, and known. It is observed fairly often up to about 100
420 MHz by W6NLZ and KH6UK,5 achievements MHz, and scattered instances occur in the 144 -MHz
which must always rank among the most signifi- band. Increased activity on 144 MHz has enabled
cant in amateur vhf annals. amateurs to observe the effect in this band more
Tropospheric communication on 144 MHz often in recent years. Ionization develops rapidly,
along the Atlantic Seaboard over distances as great with effects showing first on lower frequencies.
as New England to Florida, and overland contacts Observation of the 28 -MHz band, or commercial
up to 1400 miles, are also products of tropospheric frequencies between 30 and 50 MHz7 will usually
ducting. Scientific investigations over the South give the 50 -MHz enthusiast some advance notice of
Atlantic have shown ducts capable of propagating an impending opening. Similarly, the condition of
signals on frequencies even below 50 MHz, but true the 50 -MHz band or the vhf fm or television
ducting is rare in amateur experience below 144 frequencies may give clues as to the possibility of
MHz. It occurs most often in the uhf region, and 144 -MHz propagation.8
extensive occupancy of the amateur bands above As ionization density increases and the maxi-
400 MHz should enable us to exploit its possibil- mum usable frequency rises, the skip distance on a
ities more fully in years to come. given frequency shortens. Thus, very short skip on
Sporadic -E Skip results from reflection of vhf 50 MHz may portend a 144 -MHz opening. The
waves by dense patches of ionization in the E alert observer should watch for short 50-MHz
region of the ionosphere, roughly 50 miles above skip near the midpoint of a potential 1200 -to -
the earth. Causes are still not completely under- 1400 -mile path, as the best indication of a
stood and its occurrence is predictable only in a 144 -MHz DX chance. Hearing a 50 -MHz station in
general way, but its effects are well known to Cincinnati working another in St. Louis, for
generations of vhf enthusiasts.6 Layer height and example, would be a good omen for a 144 -MHz
electron density determine the skip distance, but operator in Washington, D.C., indicating the pos-
50 -MHz propagation is most common over dis- sibility of 2 -meter propagation to Oklahoma City
tances of 400 to 1200 miles. Often signals are very or Wichita. Plotting observed skip on a map of the
strong, though they may vary rapidly over quite United States will help one to grasp the signifi-
wide ranges. Ionization may develop simul- cance of what he hears.9
taneously in several areas, making multiple -hop Like most other vhf DX modes, sporadic -E skip
propagation possible and extending the working was discovered by amateurs (see Chapter 1) and it
range to as much as 2500 miles. Signals are usually quickly became a popular sport among 5 -meter
heard from intermediate distances at such times. men of the 1930s. To commercial users of the vhf

120o0 (23),(
.000 (Me)
w, (2.7) `,(
10,000 (MB).
W
6, 8,000 (3.4)\ 15,000

Q 6,000 (4.1) 1<` 6,000

,. 4.000 (1e).á
ti (5.0) 1
tiy
4,000 (1.9). .._ o
WATER VAPOR '. TEMPERATURE 2)k.
TEMPERATURE
ZÁ00 (6.3)\ 2.000 WATER VAPOR ,'s
.
o
- 5 0 5 10 15 20 0 5 10 15 20
TEM PERATURE (°C) TEMPERATURE CC)

Fig. 2-4 -
Upper -air conditions that produce extended -range communication on the vhf bands. At the
left is shown the U.S. Standard Atmosphere temperature curve. The humidity curve (dotted) is that
which would result if the relative humidity were 70 percent from the ground level to 12,000 feet
elevation. There is only slight refraction under this standard condition. At the right is shown a
sounding that is typical of marked refraction of vhf waves. Figures in parentheses are the "mixing
-
ratio" grams of water vapor per kilogram of dry air. Note the sharp break in both curves at about
4000 feet.
Propagation Beyond the Horizon 19
spectrum it is known mainly for its nuisance value, auroral disturbance may be getting under way. Its
but to 50 -MHz men it is the DX mode supreme. effects will not be long in showing on the bands
Though the number and quality of openings vary higher in frequency, if the disturbance is pro-
somewhat from year to year, Es propagation does nounced. Distortion of voice on 28 or 50 MHz,
not appear to be closely related to sunspot activity. when the array is aimed north, is evidence that the
Auroral Propagation involves reflection of vhf effect has reached these bands, and it is time for
waves from the auroral curtain in the northern the vhf man to go to work on cw.
skies,10 usually at acute angles. It is most common On 50 or 144 MHz the buzzing sound charac-
at 50 and 144 MHz, the number and duration of teristic of an auroral return may be heard even on
openings decreasing markedly at higher frequen- local signals, when both antennas are aimed north.
cies. Some auroral work has been done on 220 and The great -circle distance workable via the aurora
420 MHz. It may eventually become feasible above
1000 MHz, if very large arrays are used. Scientific
investigations with very high power and large
antenna arrays have shown auroral returns at
frequencies of several thousand Megahertz.
The reflecting properties of the aurora vary
rapidly, with the result that the returned vhf signal
is badly distorted by multipath effects. Voice
modulation is often unintelligible on 50 -MHz
signals, and nearly always so at 144 MHz. Keyed LAMEN INooueT-
,DJM NORTit
cw is, therefore, the most effective mode of
operation for auroral work. Suppressed -carrier ssb
is a poor second, followed by a -m, nfm, and
wide -band fm, in that order.
The number of auroras seen each year, and the
opportunities for vhf communication via the aur-
ora, vary with geomagnetic latitude. Since the
geomagnetic pole is quite near Thule, Greenland, extends from local out to more than 1000 miles,
geomagnetic latitude lines slope upward with re- but hops of a few hundred miles are most
spect to geographical latitude as we look to the common. Range depends to some extent on
west. Bangor, Maine, sees many more auroras than transmitter power, antenna gain, and receiver
does Seattle, though the latter city is farther north. sensitivity, but patience and operating skill are
New York, Philadelphia, and Washington far outdo important.
Reno and Northern California, which lie along the There is much to be learned about auroral
same geographical latitude. Aurora DX has been propagation. On 50 MHz, for example, an occas-
worked on 144 MHz as far south as 30 degrees in ional aurora will produce clear voice signals from
Southeastern U.S.A., but seldom or never in El distances out to 1200 miles or more, not unlike
Paso, Phoenix, or Los Angeles, all of which are well those encountered in sporadic -E skip propagation.
north of latitude 30. These may be accompanied by the distorted signals
Auroras follow seasonal patterns, being nest from shorter distances, the degree of distortion
common around the equinoxes (March and Sept- decreasing with frequency. On rare occasions, a
ember). They may occur at any time, however, and long -haul east -west skip may be observed, per-
summer and midwinter auroras are not uncommon mitting work over distances up to 2000 miles or
in the more northerly states. Aurora effects are more, such as between the first and seventh call
observed most often in the late afternoon or early areas.
evening, lasting for a few minutes to many hours. A somewhat similar type of propagation is
The southerly extent also varies greatly. Strong and observed more often by the few vhf operators of
widespread disturbances may peak in the early the far north. They have found 50 -MHz communi-
evening, drop off for about two hours before cation possible occasionally with stations in the
midnight, and then return, lasting until dawn or northern tier of states and adjacent Canadian areas,
after. apparently by something approximating an iono-
The optimum heading for a vhf antenna array spheric skip, using the auroral zone as a reflecting
varies with the position of the aurora and may medium.11
change rapidly, just as the visible display does. The number and geographical distribution of
Usually an eastern station will work the greatest auroras and auroral propagation on the vhf bands
distance to the west by aiming as far west of north vary with the solar activity cycle, the maximum
as possible, but this does not always follow. auroral incidence apparently lagging the sunspot
Constant probing with the antenna is recom- peak by approximately two years. The arctic
mended, especially if an array with a really sharp effects described immediately above have been
pattern is being used. observed on 50 MHz at the bottom of the solar
Developing auroral conditions may be observed cycle, so their relation to solar conditions is by no
by monitoring signals in the region from the means clear.
broadcast band up to about 5 MHz or so. If signals F2 -Layer DX may be possible in the peak years
in the 75 -meter amateur band, for example, begin of the 11 -year sunspot cycle. This ionospheric
to waver suddenly m the -afternoon or early mode, responsible for most DX on lower frequen-
evening hours, taking on a dribbling sound, an cies, opened the 50 -MHz band for world-wide
20 A VAST RESOURCE
communication during solar peaks of 1947 to ficult for amateurs in Eastern U.S.A. during peak
1950, 1956 to 1960, and 1968 to 1969. Partic- hours of Cycle 19, and hundreds of other Euro-
ularly in the late 1950s, the 50 -MHz band was pean signals and harmonics were audible whenever
excellent for distances of 2000 miles or more, for the band was open in that direction.
many hours at a time, almost daily during the Ionospheric prediction maps are available from
winter months. The first scattered F2 DX of solar the U.S. Government Printing Office, Washington,
cycle 20 came in 1967, but the peak was far below DC 20402. Ionospheric Predictions, OT-TRER 13,
those of the two previous cycles. is in four volumes: Vol. 1 explains the use of the
Frequencies near the maximum usable (muf) maps, 30 cents. Vols. 2, 3, and 4, 53.00 each,
produce the strongest F2 -layer signals, and multi- contain maps for predicted Zurich smoothed rela-
ple -hop effects and combinations with other forms tive sunspot numbers of 10, 110, and 160, respec-
of propagation may provide 50 -MHz communi- tively. Interpolation of data from two volumes
cation over very long paths. Africa to California, must be made for solar -activity periods of inter-
U.S.A. to Japan, Hawaii to Australia, and even mediate levels. Information on predicted sunspot
South America to Japan, -were covered frequently numbers in contained periodically in propagation
in the late 1950s, with signal strengths rivaling the forecast bulletins, transmitted by W 1 AW and many
best ever experienced on lower frequencies. Only Official Bulletin Stations.
lack of 50 -MHz privileges for amateurs of many Since the muf is related to the position of the
countries prevented 50 MHz from becoming the sun, it is highest at roughly noon at the midpoint
prime amateur DX band during this memorable of a given path. It tends to be highest in the low
period. latitudes, and lowest along paths traversing the
Whether like conditions will prevail during auroral zones. The highest recorded F2 muf was in
future peaks of solar activity is a matter of some the vicinity of 75 MHz.
conjecture. Sunspot records dating back to about Back -Scatter signals from amateur stations in-
1750, Fig. 2-5, show long-term trends indicating side the skip zone indicate high muf and also show
that we may be near the end of an era of generally the direction in which conditions are most favor-
high activity. Thus it is possible that peaks of the able for long-distance work. The F2 layer has
magnitude of recorded cycles 18 and 19 may not almost mirror qualities near the muf. Signals
recur within the lifetimes of readers of these pages. reflected from it come down at a distant point on
Cycle 20, peaking in 1968-69, showed a lower the earth's surface, from which they scatter in all
trend. directions. Some of the energy comes back to the
The muf can be checked readily with a general - ionosphere and is reflected back to earth. Thus a
coverage receiver.12 Since propagation near the station in Virginia, for example, will be heard by a
muf is very good, signals will be heard from station in Ohio, when both have their antennas
somewhere on any frequency that is alive. The aimed at Europe and that path is open for both of
10-meter band provides good clues, for its skip them.
distance shortens markedly as the muf rises toward Signals scattered back to or near to their point
50 MHz. If 10 is open for long periods daily, and of origin are weak and have a high degree of
the skip shortens to 1200 miles or less during the multipath distortion, somewhat like those reflected
peak hours, the muf is approaching 50 MHz. This is from the aurora. Voice may be only partly
the time to watch the frequencies just below 50 readable, and cw is highly effective under such
MHz, making note of the highest frequency at conditions. Back -scatter is usually strongest for
which DX signals can be heard, and the time of day stations no more than a few hundred miles apart,
when they appear. A daily record of these obser- but back-scatter QSOs have been made between
vations will show if the muf is rising. Enough use points as widely separated as New England and
of the vhf range is made, almost everywhere ín the Mexico City. When this occurred, both stations
world, so that there will be plenty of evidence of were aiming at a common `open" point in the
an imminent 50 -MHz opening. Usually there are South Atlantic. Alaska and California, with a
many signals, both in the band and near it. common opening to Japan, have had similar
European television made 50 -MHz DX work dif- experiences. Often the direct path between the two

RELATIVE SUNSPOT NUMBERS. 1749-1972

200 1957-58
180
16

110
120
10

Fig. 2-5 -Relative sunspot number records dating back to before 1750 show that the last two solar
peaks, known as cycles 18 and 19, were the highest in all of man's observation of the sun. Looking at
the long-term curve indicates that we may be moving into an era of generally low solar activity.
 -
Propagation Beyond the Horizon 21

..
stations will produce no signals, the circuit being
open only via the longer back-scatter route.

.-. /..'
A similar condition may be observed during
periods of sporadic -E propagation (see page 18) ~BM,/
though the shorter skip may make E-layer back - //rT-y' 1
1M...1
scatter hard to distinguish from other modes of
propagation. ':.../.
4,7,
Transequatorial Propagation during the after-
noon and evening hours is possible for 50 -MHz
stations situated at optimum distances from the
geomagnetic equator, roughly 1500 to 2500 miles
above and below. See Fig. 2-6. This mode is
GEOGRAPHICAL
associated with high sunspot periods, but because EQUATOR
it is effective at frequencies up to at least 1.5 times
the observed daytime muf for the F2 layer, it runs
for a longer portion of the solar cycle than does
the normal F-layer propagation described above. It MAIN ZONE
has been observed over paths that cross the equator
at angles as low as 45 degrees, and at greater
EXTENSIONS
distances than given above, but long paths and
large deviations from the north -south route show GMT +1 +P +3
lower muf and shorter openings.
Typical TE paths of high reliability are Puerto Fig. 2-6 - Main and occasional zones of trans -
Rico to Argentina, Japan to Australia, and the equatorial 50 -MHz propagation, as described by
Mediterranean area to Southern Rhodesia.13 These ZE2JV, show Limassol, Cyprus, and Salisbury,
circuits continue to show TE propagation in the
Rhodesia, to be almost ideally positioned with
respect to the curving geomagnetic equator. Wind-
50 -MHz band long after F2 -layer DX has expired. hoek, Southwest Africa, is also in a favorable spot.
At the peak of Cycle 19 the TE mode was observed Johannesburg somewhat less so.
over most of the United States, and in Europe as
far north as the British Isles on isolated occasions.
The spring and fall months show this mode to the for 144 -MHz work. Experience has shown that
best advantage. Signals may have a high degree of with optimum equipment, signals can be ex-
flutter, but voice readability is seldom seriously changed on 144 MHz consistently at distances out
impaired. to 450 to 500 miles15 regardless of propagation
Ionospheric Scatter is usable for marginal conditions. Such signals are very weak and difficult
50 -MHz communication over distances comparable to copy, and as in all weak -signal work, cw is the
to those encountered in single -hop E, mainly 600 only mode by which communication can be carried
to 1200 miles. Because only a very small part of on effectively in tropospheric scatter over the
the energy scattered in the E region of the extreme distances.
ionosphere returns to earth, such signals are ex- Meteor Scatter is one of the more esoteric
tremely weak. Large antennas, fairly high trans- forms of vhf DX currently worked by amateurs.
mitter power, and good receivers are essential, and Meteors entering the earth's atmosphere from
even with all these only cw emission can be outer space burn up rapidly in the E region of the
expected to produce consistent results. Though ionosphere due to friction. In the process a
ionospheric scatter is now widely used for military cylinder of dense ionization is formed as a trail
communication over long distances its adaptation behind the meteor. A 50 -MHz signal reflected from
to amateur use is limited.14 Operation on carefully this trail may appear as a few words of readable
kept schedules, with the precise frequencies to be voice, from a station up to 1200 miles away. If you
used known in advance, is almost a necessity, but make a habit of tuning the, 50 -MHz band carefully
amateur experience on many 50 -MHz scatter cir- for considerable periods, you may have heard these
cuits has shown that the bare essentials of com- bursts from time to time. On 144 MHz, the same
munication can be exchanged, if good equipment meteor would provide a much shorter burst,
and operating skill are available at both ends of the perhaps no more than a "ping" heard when the
path. receiver beat oscillator is on.
When there are aiding factors, such as devel- Meteors are constantly entering the earth's
oping sporadic -E ionization, or more than the atmosphere, so if a vhf receiver is left tuned to a
average number of meteors near the midpoint, distant station its signal pings will be heard at
ionospheric scatter circuits improve markedly, even random intervals at all times. The number and
though the path may not be considered "open" in duration of bursts increase greatly during major
the normal sense. showers listed in Table I, and they may come often
Tropospheric Scatter is similar to the iono- enough to permit communication of sorts between
spheric form, except that it occurs nearer to the cooperating stations.16 Meteor bursts are heard
earth's surface, and consequently shows mainly as frequently on any 50 -MHz ionospheric-scatter cir-
an extension of the normal working range of vhf cuit, rising suddenly far above the weak residual
stations. It is effective at all frequencies, 50 MHz signal that is characteristic of such communication.
and higher, but has been used by amateurs mainly On 144 MHz there is normally no residual signal
22 A VAST RESOURCE

beyond a few hundred miles, so only the meteor 420-, 1215-, and 2300-MHz bands, but only with
bursts are heard. Occasionally during major meteor very sophisticated equipment. Communication by
showers, bursts of up to a minute or more of reflection from the moon remains a challenge to
continuous 144 -MHz signal may be received. the more advanced amateur, and it is a fine project
Such fortunate breaks are by no means neces- for clubs having the facilities, ability, and deter-
sary for communication between dedicated en- mination to achieve a measure of success.
thusiasts. Using high keying speeds and precisely The requirements are fairly well established.
timed transmissions on accurately known frequen- They include the maximum legal power, the
cies, they often achieve exchanges of information ultimate in receiver performance, very large an-
on a series of bursts of no more than a few tennas capable of being aimed and controlled
seconds' duration each. accurately,21 and a willingness to work with very
The usual arrangement is for one station to weak cw signals. Because very high receiver selec-
transmit for exactly one minute, 30 seconds, or tivity is often used, to realize the best possible
even 15 seconds, following which the other takes signal-to-noise ratio, stability requirements in both
over for a like period. Detailed procedure is agreed transmitter and receiver are critical.
on in advance. Typically, a signal report (Si, S2, or Antenna problems are similarly staggering. At
S3, indicating duration of bursts, rather than signal 144 MHz, an antenna gain well in excess of 20 dB
strength) Is sent immediately upon identification is required. Polarization rotation along the path,
of the other station. When the signal report Is out and back, makes it appear that circular
received, an R is transmitted repeatedly, signifying polarization offers the best chance of success.
receipt. Because of the reflection of the signal, the polar-
Daily schedules kept through a major shower ization sense should be reversed between trans-
peak will usually yield enough bursts on at least mission and reception, in listening for one's own
one day to complete a contact within an hour or echoes.
less of such cooperative effort. A long loud burst is The polarization rotation problem is largely
usually the signal to abandon timed transmissions eliminated above 1000 MHz, but the difficulties of
abruptly and try for a complete exchange before generating sufficient power, and building a good
signals disappear. Very short sequences, such as 15 enough receiver are multiplied. Detailed discussion
seconds each way, have the advantage of making it of moonbounce problems is beyond the scope of
likely that any appreciable burst will be usable in this book. Study of the references at the end of
this way. this chapter is recommended to the serious would-
Information in the meteor shower table should be worker in this field.
be used as a guide, and not relied on completely. Progress in amateur communication via the
Showers may not peak at exactly the same time moon has been steady, though largely unspect-
from year to year, and those having widely spaced acular, in recent years. Because the ultimate in
periodic peaks may be deflected and not appear at equipment of every kind is required, it is probably
all. The times given are local standard. unrealistic to expect that any great "break-
After putting on the biggest show in amateur through" in this field will suddenly make reliable
radio history up to that time, October, 1946,17 communication by moon reflection practical with-
the Giacobinids were deflected, and were hardly in the framework of the amateur regulations. The
discernible 13 years later. On the other hand, the prospect of a lunar repeater is, however, quite
surprising resurgence of the November Leonids in another matter. Man's progress in the mastery of
1965 and 1966 provided 144 -MHz meteor DX of space travel being what it is, repeaters on the moon
unprecedented proportions.18 The August Perseids could revolutionize our whole approach to amateur
and December Geminids are dependable year after radio communication within a decade.
year, and are highly favored for vhf scheduling on Passive and Active Satellites offer possibilities
that account. for worldwide communication in the vhf range and
Minor showers listed in the table may offer higher frequencies. The former consists of a reflec-
little more opportunity for communication than tor of some sort, as in the Echo series, presenting
do the random meteors always entering the earth's problems even more severe than those involved in
atmosphere. Scheduled work over the right dis- lunar communication. Very good equipment is
tances, especially between 2200 and 0900 local required, and antennas designed for tracking must
time, should yield some results practically every be used, if this mode of communication is to be
day. used by amateurs.
Voice contacts are made occasionally during Active satellites pick up the transmitted signal
meteor showers, and success in this has increased as and relay it, usually on another frequency. While
voice-controlled ssb has become more widely used. such a device is relatively easy for the amateur to
Lunar Communication has been the dream of use, the design and operational problems are
vhf men for a generation, but successful use of the formidable. The number of signals it can accom-
earth-moon -earth (EME) path is still very difficult. modate at any one time is severely limited. The
Pioneering 2 -meter men received their own signals active satellite is expected to play a large role in
reflected from the moon in 1953,19 and long amateur radio of the future, as are other com-
distances have been covered two-way in recent munications ramifications of the space age. Prob-
years.20 Marginal communication via the moon has lems, possibilities, and results with active satellites
been carried out over long distances on the 220-, have been discussed extensively in QST. 22
Meteor Showers 23
Table I-Meteor Shower Data for V.H.F. Use
Time Visible Optimum Paths and Times Hourly Rate Velocity, Period, Next
Shower
and Dote Rise Set N -S NW-SE E -W SW -NE Visual Radio km/sec. Years Maximum

* January 3-5 2300 1800 - 0300-0800 0800-0900 0900-14C0 35 45 45 7 Note 1

Quadrantids
January 17 0230 2130 - SW S
0600-1100 1100-1300 1300-1800
SE
- - - - -
Cygnids
February 5-10 1200 0330 - SW
1400-1730 -
S SE
2130-0100 - - - - -
Aurigids
March 10-12
SW
2200 0830 2330-0030 W 0330-0530 0230-0330 0030-0230
SE
- - - - -
Bo6tids
March20
0530-0630 E' NE
1800 0630 2130-2300 W 2000-2130
N NW
0300-0430 - - - - - -
Coma Berenice:
April 19-23 2100 1100
0100-0300 E
0230 W
SW
2330-0100
SE
0700-0830 - 8 12 51 415 Note1
Lyrids
*May 1-6 0300 1200 -
0530 E SW
0830-1000 0630-0830 0500-0630
SE
12 12 66 76 Note I
Aquarids
May 11-24 1800 2130-2300 W 2000-2130
0630
NE
-
N NW
0300-0430 - - - - -
Herculids
May 30
0100-0300 E SW
2300 1200 0300-0430 W 0130-0300 - SE
0800-0930 - - - - -
Pegasids
June 2-17 2000 0300 -
0630-0800 E SW
0100 2300-24002200
SE
- - - - -
Scorpiids
June 27-30 Does not set; -
NE N NW
1500-1830 1830-2330 2330-0300 - - - - -
Pons Winnecke
July 14
min. at 0900
1800 1000 - SW
2100-2330 0130
S SE
0330-06C0 - - - - -
Cygnids
July 18-30 2030 0400 - SW S
0100-0200 2300-0100 2200-2300
SE

NW
- - - - -
-
Capricornids NE N
July 26-31 2200 0600 0300-0500 0100-0300 0000-0100 10 22 50 3.6 Note 1

NW
Aquarids
July 27 -August 14 Does not set; - NE N
2330-0300 0300-0800 0800-1130 50 50 61 120 Note 1

Perseids
August 10-20
min. at 1730
1200 0700 - SW
1700-1930 2130
5 SE
2330-0200 - - - - -
Cygnids
August 21-23 Does not set; - SW S
1500-1830 1830-2330 2330-0300
SE
- - - - -
Draconids
August 21-31
min. at 0900
Does not set; - SW S
1300-1630 1630-2130 2130-0100
SE
- - - - -
Draconids
September7-15
min. at 0700
2130 1200 -
SW
0030-0200 -
S SE
0700-0830 - - - - -
Perseids
September22 2100 1230 - SW
0030-0200 - SE
0700-0830 - - - - -
Aurigids
October 2
Quadrantids
0500 0000 - SW
0900-1400 1400-1500 1500-2000
SW
SE

SE
- - - - -
-
S

October 9 0600 0300 1 100-1600 1600-1700 1700-2200 Note 2 20 6.6 1972

Giocobinids
October 12-23
Arietids
1900 2130-2330 W
0700
0230-0430 E
-
SW
-
5
-
SE
- - -
October 18-23 0930 0000-0200 W 0430-0600 0330-0430 0200-0330
2230 15 30 68 76 Note 1

Orionids 0600-0800 E NE N NW
Oct. 26 -Nov. 16 0630 2100-2300 W 0130-0300 0030-0130 2300-0030
1900 10 16 27 3.3 Note 1

Taurids
November 14-18 0000
0300-0500 E
1230 0300-0500 W
NE N NW
- - - 12 Note 3 72 33.2 1999
Leonids
November 22-30 1300 0600
0800-1000 E

-
1600-2000 2300-0300 - Note 4 22 6J 1977
Andromedids
December 10-14 1900 0900 0030 W
SW
2130-2300 - SE
0500-0630 60 70 35 1.6 Note 1

Geminids
December 22 Does not set; -
0330 E
-
SW
0130-1530 -
SE
13 13 38 13.5 1972,
Ursids
May 19-21
min. at 2030
0530 1430 - 5
1100-1230 0900-1100 0730-0900 - -
Note 5
20 37 -
1985

Cetids
June 4-6 1730 0800-1000 W
0500 -
NE
-N
-
NW
40 29 -
Perseids
June 8
1300-1500 E
0330 1530 0600-0800 W - - - Note 6 70 38 -
Arietids
June 30 -July 2
Tourids
1100-1300 E
0500 1700 0700-0900 W 1130-1300 1030-1130 0900-1030
1300-1500 E NE N NW
- - 30 31 -
Major showers-Last four are daylight showers.
Times given are local standard at path midpoint
NOTES
1. These streams are evenly distributed and little year to year variation is to be expected.
2. Very concentrated stream. Peak years give up to 400 meteors per minute, but with duration of only 6 hours. 1946 peak was most
concentrated shower in amateur radio experience up to that time (see December, 1946, OST, page 431 but 1959 recurrence was de-
flected and was hardly observable.
3. Peak years give 60/hour visual. In the peak years of the 1800s, prior to being deflected by Jupiter and Saturn, this shower gave
1200 per minute. Spectacular results in 1965 and 1966 are reported in Jan. 1966 OST, page 80, and Jan. 1967, page 83.
4. Before being deflected by Jupiter this stream gave peok year rates of 100/minute. No notable rotes have been observed since,
though the stream could return.
5. Short duration shower. Peak years the radio rate is I65/hour.
6. This intense daylight shower begins June 2 and runs to June 14 with radio rates from 25 to 70/hour.
24 A VAST RESOURCE

RELIABLE VHF COVERAGE

In preceding pages we discussed means by bilities, two basic numbers must be determined:
which our bands above 50 MHz may be used station gain, and path loss. Station gain is made up
intermittently for communication far beyond the of eight factors: receiver sensitivity, transmitted
visual horizon. In emphasizing DX we should not power, receiving antenna gain, receiving antenna
neglect a prime asset of the vhf bands: reliable height gain, transmitting antenna gain, transmitting
communication over relatively short distances. The antenna height gain, and required signal-to-noise
vhf region is far less subject to disruption of local ratio. This looks complicated but it really boils
communication than are frequencies below about down to an easily made evaluation of receiver,
30 MHz. Since much amateur communication is transmitter, and antenna performance. The other
essentially local in nature, our vhf assignments number, path loss, is readily determined from the
could carry a much greater load then they pres- nomogram, Fig. 2-7. This gives path loss over
ently do, and this would help solve interference smooth earth, for 99 percent reliability.
problems on lower frequencies. For 50 MHz, lay a straightedge from the
Possibly some amateur unwillingness to migrate distance between stations (left side) to the appro-
to the vhf bands is due to misconceptions about priate distance at the right side. For 1296 MHz, use
the coverage obtainable. This reflects the age-old the full scale, right center. For 144, 220, and 432,
idea that vhf waves travel only in straight lines, use the dot in the circle, square, or triangle,
except when the DX modes enumerated above respectively. Example: at 300 miles the path loss
happen to be present. Let us survey the picture in for 144 MHz is 215 dB.
the light of modern wave -propagation knowledge
and see what the bands above 50 MHz are good for Station Gain
on a day-to-day basis, ignoring the anomalies that
may result in extensions of normal coverage. The largest of the eight factors involved in
It is possible to predict with fair accuracy how station gain is receiver sensitivity. This is obtain-
far you should be able to work consistently on any able from Fig. 2-8, if you know the approximate
vhf or uhf band, provided a few simple facts are receiver noise figure and transmission -line loss. If
known. The factors affecting operating range can you can't measure noise figure, assume 3 dB for.50
be reduced to graph form, as described by D. W. MHz, 5 for 144 or 220, 8 for 432, and 10 for
Bray, K2LMG.23 To estimate your station's capa- 1296, if you know that your equipment is working

- 5

120 30x30

- 10
130 50 mc
20
40
50Ñ
aE

- 140 5 90
lo - 144mc
101,

1 0 220mc
20 150 12= (Point only,
All distances)
14- 432mc
160
16-
30 170
18-
20-
40 30-
180 40-
50
50-
60 190 60-
70-
70
= 200 80-
80
- 90 90
E ¡p 210 100
120
140 ;
- 200
240 200'
- 260 :7, 220

280 300-
= 230
- 300 400
500
L 240 1296 mc

E250
- 400
260 Fig. 2-7- Nomogram for finding the capabilities of stations on amateur
bands from 50 to 1300 MHz. Either the path loss for a given distance or
- 500 vice versa may be found if one of the two factors is known.
Reliable VH F Coverage 25
IOOKC
17 -
-140 16-
15-
14-
-
- 10 KC
-150 13-
12
I
-- 160 10-

- I KC

8-
- 170 7-

6-
5-
= I00cps -- 180 4^
3-
2-
-190 I-
- 10 cps 0-
Fig. 2-8 - Nomogram for finding effective receiver sensitivity.
moderately well. Line loss can be taken from Table Transmitter power output must be stated in dB
8 -III for the line In use, if the antenna system is fed 1 watt. If you have 500 watts output, add
above
properly. Lay a straightedge between the appro- 500/1, or 27 dB, to your station gain. The
priate points on either side of Fig. 2-8, to find transmission line loss must be subtracted from the
effective receiver sensitivity in dB below one watt station gain. So must the required signal-to-noise
(dbW). Use the narrowest bandwidth that is prac- ratio. The information is based on cw work, so the
tical for the emission intended, with the receiver additional signal needed for other modes must be
you will be using. For cw, an average value for subtracted. Use 3 dB for ssb and 7 dB for a-m. Loss
effective work is about 500 cycles. Phone band- due to fading must be accounted for. It has been
width can be taken from the receiver's instruction shown that for distances beyond 100 miles the
manual. signal will vary plus or minus about 7 dB from the
Antenna gain is next in importance. Gains of average level, so 7 dB must be subtracted from the
amateur antennas are often exaggerated. For well - station gain for high reliability. For distances under
designed Yagis they run close to 10 times the 100 miles, fading diminishes almost linearly with
boom length in wavelengths. (Example: a 24 -foot distance. For 50 miles, use minus 3.5 dB for
Yagi on 144 MHz is 3.6 wavelengths long. fading.
3.6 X 10 = 36, or about 15 1/2 dB.) Add 3 dB for What It All Means
stacking, where used properly. Add 4 dB more for
ground -reflection gain. This varies in amateur After adding all the plus-and -minus factors to
work, but averages out near this figure. We have get the station gain, use it to find the distance over
one more plus factor: antenna height gain, obtain- which you can expect to work reliably, from the
able from Fig. 2-9. Note that this is greatest for nomogram, Fig. 2-7. Or work it the other way
short distances. The left edge of the horizontal around: find the path loss for the distance you
center scale is for 0 to 10 miles, the right edge for want to cover from the nomogram and then figure
100 to 500 miles. Height gain for 10 to 30 feet is out what station changes will be needed to
assumed to be zero. It will be seen that for 50 feet overcome it.
the height gain is 4 dB at 10 miles, 3 dB at 50 The significance of all this becomes more
miles, and 2 dB at 100 miles. At 80 feet the height obvious when we see path loss plotted against
gains are roughly 8, 6, and 4 dB for these distances. frequency for the various bands, as in Fig. 2-10. At
Beyond 100 miles the height gain is nearly uniform the left this is done for 50 percent reliability. At
for a given height, regardless of distance. the right is the same information for 99 percent
26 A VAST RESOURCE
10-
-lo
8-

6-

_ 20 4-
FOR HEIGHTS 2-
10 to 30 ft

-30 0-
0 50 100
10 500

- 40 DISTANCE (miles)
-2-
FOR HEIGHTS 30 to1OOft

- 50

- 60

- 70
- 80
-90
-100
Fig. 2-9 - Nomogram for determination of antenna -height gain.

Working out a few typical amateur vhf station coverage. Achieve all of them, and you will have
setups with these curves will show why an under- very likely tripled your sphere of influence, thanks
standing of these factors is important to any user to that hump in the path-loss curves. This goes a
of the vhf spectrum. Note that path loss rises very long way toward explaining why using a 10 -watt
steeply in the first 100 miles or so. This is no news packaged station and a small antenna, fun though
to vhf men; locals are very strong, but stations 50 it may be, does not begin to show what the vhf
or 75 miles away are much weaker. But what bands are really good for.
happens beyond 100 miles is not so well known to About Terrain
some of us.
reliability. For near -perfect reliability, a path loss The coverage figures derived from the above are
of 195 dB (easily countered at 50 or 144 MHz) is for average terrain. What of stations in moun-
involved in 100 -mile communication. But look at tainous country? Though an open horizon is
the 50 percent reliability curve: the same path loss generally desirable for the vhf man, mountain
takes us out to well over 250 miles. Few amateurs country should not be given up as hopeless until it
demand near -perfect reliability. By choosing our has been proven so. Help for the valley dweller
times, and accepting the necessity for some repeats often líes in the optical phenomenon known as
or occasional loss of signal, we can maintain knife-edge refraction.24 A flashlight beam pointed
communication out to distances far beyond those
usually worked by vhf men.
From the curves of Fig. 2-10, we see that path
loss levels off markedly at what is the approximate
limit of working range for average vhf stations
using voice. Work out the station gain for a 50 -watt
station with an average receiver and moderate -sized
antenna, and you'll find that it will come out
around 180 dB. This means about a 100-mile
working radius in average terrain, for good but not
perfect reliability. Another 10 dB may extend the
range to as much as 250 miles. Just changing from
a -m phone to cw can thus do wonders for you. A
bigger antenna, a higher one if your present beam
is not at least 50 feet up, an increase in power to
500 watts from 50, an improvement in receiver
noise figure if it is presently poor - any of these
things can make a big improvement in reliable
Operating Modes 27
COY 50

240 I

220

/'''.--
/ 220

296 Yt
l
200

3e0
552

12.0 MC
300

AlMil
:Fa-
160

1,10 l wl 1
1-
120n
I 20e 315/1 WO 50
:1 i
I

01ST44CE 15,411 DISTIYCE 1111

Fig. 2-10 -
Path loss vs. distance for amateur frequenceis above 50 MHz. Curves at the left are for 50
percent of the time; those at the right for 99 percent. The former is the more representative of
amateur radio requirements.

at the edge of a partition does not cut off sharply paths are common in high -mountain country. Mt.
at the partition edge, but is refracted around it, Rainier, Mt. Hood, and other majestic peaks of the
partially illuminating the shadow area. A similar Northwest are examples, and the mountains of
effect is observed with vhf waves passing over California provide many others. Mt. McKinley in
ridges; there is a shadow effect, but not a complete Alaska has demonstrated remarkable capabilities
blackout. If the signal is strong where it strikes the for both knife-edge and reflector service.
mountain range, it will be heard well in the bottom Rolling terrain, where obstructions are not
of a valley on the far side. sharp enough to produce knife-edge refraction, still
This is familiar to all users of vhf communica- does not exhibit complete shadow effect. There is
tions equipment who operate in hilly terrain. no complete barrier to vhf propagation; only
Where only one ridge lies in the way, signals on the attenuation, which varies widely as the result of
far side may be almost as good as on the near. many factors. Thus, even valley locations are
Under ideal conditions (a very high and sharp - usable for vhf communication. Good antenna
edged obstruction near the midpoint of a path long systems, preferably as high as possible, the best
enough so that signals would be weak over average available equipment, and above all, the willingness
terrain), knife-edge refraction may yield signals and ability to work with weak signals may make
even stronger than would be possible with an open possible outstanding vhf work, even in sites that
path. show little promise by casual inspection.
The obstruction must project into the radiation by the radiation from the array. Bulldozing them
patterns of the antennas used. Often mountains out of the way wouldhave substantially no effect
that look formidable to the viewer are not high on vhf signal strength in such cases.
enough to have an appreciable effect, one way or Mountains that are really high but not situated
the other. Since the normal radiation from a vhf or shaped so that they exhibit knife-edge effects
array is several degrees above the horizontal, may be useful in another way: as a reflector
mountains that are less than three degrees above visually common to two stations that do not have a
the horizon, as seen from the antenna, are missed direct open path between them.

OPERATING MODES
Almost every amateur has one mode of opera- inhabitant of the world above 50 MHz is wider
tion that he prefers over all others. Once this was a than for any other amateur.
simple choice between phone and code, but to- Though many of us tend to concentrate on one
day's picture is more complex. The voice operator mode, there is much to be said for versatility. With
on the vhf bands can use amplitude modulation, foresight in planning his station, the vhf enthusiast
suppressed -carrier single-sideband or double- can incorporate several modes of operation with
sideband, or frequency modulation, either wide or little difficulty. Though a-m phone is heavily
narrow -band. The code man can employ conven- entrenched in vhf work, other modes have much to
tional cw, keyed tone modulation, or frequency - offer. The most effective system for long-distance
shift keying. Other modes include television (slow- vhf work is keyed cw, and it is also the simplest of
scan or wide -band), radio teletype (with either all communications systems in its transmitter
audio- or radio -frequency shift), facsimile and requirements. There is no valid reason why cw
pulse. Because someof these are wide -band modes, capability should not be built into every vhf
taking up more space than would be permissible in station, yet a surprising number of vhf men make
crowded lower bands, the choice open to the no use of it. This is lamentable, for as shown in
28 A VAST RESOURCE
Fig. 2-10 the improvement gained through intel- requires a very wide band of frequencies, so it is
ligent use of cw can double or triple the effective confined to bands from 420 MHz up. Nearly all
operating range of any vhf station. Probably by no amateur TV is currently on the 420 -MHz band,
other means can vhf coverage be extended so between 436 and 450 MHz. -

easily. Pulse is a very wide mode inherently and is

An appreciable improvement in consistent permitted only on certain bands above 2300 MHz.
range with voice can be achieved by going to Though employed only to a limited extent in
single-sideband, as compared with a -m phone. Both amateur communication, it has interesting possi-
cw and sideband are more effective than other bilities on frequencies where its wide -band nature
modes mainly because they occupy a narrower can be accommodated25
band of frequencies. Cw requires essentially no Another emission is usable on all amateur
space at all, and thus it permits almost infinite frequencies above 51 MHz: unmodulated carrier,
receiver selectivity when suitable techniques are AO. It has many uses, not the least being remote
employed. Sideband requires less than half the control systems for model aircraft, boats, and the
spectrum space of other voice modes. No power is like. Much neglected by vhf men, an interesting
wasted in transmitting a carrier, this being sup- adaptation of AO is duplex phone. In using AO the
plied, in effect, by the receiver's beat oscillator. operator need identify only once every ten minutes
The bfo is also an important factor in the or less, and he is not required to have any
effectiveness of cw. intelligence on the signal in between identifi-
Frequency modulation, all but Ignored by a cations. This permits running the transmitter while
generation of vhf men, has great potential worth. tuning for other signals, either on the same band as
With suitable receiving techniques, fm provides the transmitter, or on another band. The 50- and
almost totally noise -free communication within its 144 -MHz bands tend to become congested at their
service area. Either wide -band or narrow-band fm Is low ends, so ít is well to use duplex technique only
easily incorporated in a vhf transmitter, and in the less -occupied upper portions. The U.S.
because it adds nothing to the power that must be Regulations allow AO only above 51 MHz for this
dissipated by the tubes, or to the voltage that reason, and it is well to check the frequencies to be
components must be capable of withstanding, the used for AO to see that they are unoccupied,
full cw ratings of all parts of the transmitter apply before embarking on duplex work.
to fm as well. Perhaps most important of all, fm Use of low power, separate antennas for trans-
eliminates practically all chance of audio -type mission and reception, and earphones in place of
interference in TV receivers, broadcast sets, hearing speakers are aids to effective duplex operation.
aids, and other audio amplifiers, and so is inval- Especially where signals are reasonably strong, as in
uable in solving interference problems for the local work, duplex provides a ready give-and-take
amateur in densely populated areas. Not widely exchange that can be far more pleasant and
appreciated is the fact that, with proper receiving efficient than the one-way monologue that conven-
techniques, fm is at least the equivalent of a -m in tional methods entail.
reliable vhf coverage. Only cw and ssb are superior.
Tone modulation (A2) is a simple means of
sending code on the vhf bands with any voice QST VHF PROPAGATION BIBLIOGRAPHY
transmitter. It is only slightly better than a-m
phone for weak -signal work, but it is fine for code 1 Hull, "Air-Mass Conditions and the Bending
practice, and for use on frequencies where either of UHF Waves," June, 1935, and May, 1937.
the receiver or the transmitter may not be stable 2 "On the Very Highs," July, 1944.
enough for cw, or where the receiver does not have 3 Hoisington, "Painless Prediction of Two -
Meter Openings," October, 1949.
a bfo.
4 Collier, "Upper-Air Conditions for 2 -Meter
Use of frequency -shift keying (fsk) is confined DX," September, 1955.
mainly to radio -teletype work, but it could be 5 "World Above 50 Mc.," September, 1957, P.
employed for cw. With today's voltage -variable 68, August, 1959, p. 68, and September, 1960, p.
capacitors, it is a simple matter to shift the 78.
frequency for fsk. For a report on conditions during typical
overland ducting, see Botts, "A Night to Remem-
Slow -scan television has no special vhf con- ber," January, 1970, p. 88.
notation. Conventional high-quality TV (fast scan) 6 Pierce, "Interpreting 56 -Mc. DX," September,
1938.
7 Helton, "Sporadic -E Warning Service for the
6 -Meter Man," July, 1961,
8 Ennis, "Working 2 -Meter E-Layer DX," June,
1967.
9 Wilson, "Midlatitude Intense Sporadic -E Pro-
pagation," December, 1970, and March, 1971.
10 Moore, "Aurora and Magnetic Storms,"
June, 1951. Dyce, "More About Auroral Propaga-
tion," January, 1955.
11 Mellen, Milner, and Williams, "Hams on
Ice," January, 1960.
12 Heightman, "Any DX Today?" January,
1948.
13 Cracknell, "Transequatorial Propagation of
VHF Propagation Bibliography 29
VHF Signals," December, 1959. "More on Trans - 20 A bibliography of moonbounce references is
equatorial Propagation," August, 1960, p. 47. available from ARRL, no charge. Send stamped
Whiting, "How TE Works," April, 1963. self-addressed envelope.
14 Moynihan, "VHF Scatter Propagation," 21 Michaels, "Tracking the Moon," January,
March, 1956. Taylor, "Working Ionospheric Scat- 1965.
ter on 50 Mc.," December, 1958. 22 OSCAR-Amsat Bibhography available from
15 Moore, "Over the Hills and Far Away,"
ARRL, no charge. Send stamped self-addressed
February, 1951. envelope.
16 Bain, "VHF Meteor Scatter Propagation," 23 Bray, "A Method for Determining VHF
April, 1957. Station Capabilities," November, 1961.
17 "World Above 50 Mc.," December, 1946, p. 24 Craig, "Obstacle Gain at 50 Mc. and High-
43. er," March, 1958.
18 Leonids summaries: January, 1966, p. 81,
and January, 1967, p. 83.
25 Guba and Zimmer, "Pulse - A Practical
Technique for Amateur Microwave work," Feb-
19 Tilton, "Lunar DX on 144 Mc.," March, ruary through May, 1963. Also see Chapter 13 of
1953. this manual.
 Chapter 3

Reception Above 50 MHz

RECEIVER CHARACTERISTICS
Nowhere in amateur radio is optimum recep- Sensitivity and Noise Figure
tion so important as on the frequencies above 50 Here we reach the parting of the ways with our
MHz. Selectivity, stability, sensitivity, smooth tun- hf brethren. Reception of weak vhf signals is
ing, ability to reject unwanted signals - important limited by factors quite different from those
attributes in any amateur receiver - are even more affecting bands below 30 MHz. On frequencies up
desirable in vhf work. Receiver problems are to the lower portion of the vhf spectrum, reception
similar on all amateur frequencies, but ideal solu- is limited almost entirely by noise picked up by the
tions vary greatly for different parts of the antenna. This may be man-made (ignition, power -
spectrum and modes of operation. Whether your line noise, electrical noise from motors, neon signs
station is a transceiver or a receiver -transmitter and the like) or natural, such as galactic or solar
combination, an understanding of basic receiver noise and atmospherics. With modern tubes, tran-
principles and consideration of your primary op- sistors, and circuits the noise contribution by the
erating objectives are vital to your future as a vhf hf receiver Itself is inconsequential. To prove this
enthusiast. We will examine each of the above for yourself, tune an hf receiver to some spot
qualities as it affects vhf communication. below 30 MHz where no signals are heard. Turn off
the agc and advance the gain controls until noise is
Selectivity heard. Now remove the antenna. If your receiver is
On lower frequencies selectivity is desirable stable and well shielded it should go dead quiet, or
mainly to prevent interference between stations in nearly so. The noise was practically all coming in
our crowded bands. It serves this end for vhf men, on the antenna.
too, but it is also important in achieving the best Now try the test with a vhf receiver, on 144
possible signal-to-noise ratio. Receiver noise out- MHz or higher frequencies. Chances are that the
put, for a given input signal level, is related to noise will change hardly at all, antenna on or off, if
bandwidth; the wider the bandwidth the greater
the noise. This is not important in reception of
strong signals, but for best readability of weak
signals no more bandwidth should be used than is
necessary to pass the intelligence on a signal. This
kind of selectivity is mainly determined by the i -f
system in a superheterodyne receiver, and the
means for achieving it do not change much with
signal frequency.
Stability
As selectivity is increased, the need for good
stability rises, and as we go higher in frequency the your location is a quiet one. This demonstrates
difficulty of achieving satisfactory stability also where most vhf noise comes from: the receiver
rises. Thus it is almost standard procedure to use makes it. The amount of noise it makes, over that
crystal-controlled frequency converters in vhf re- of a theoretically perfect receiver whose noise is
ception, doing the tuning at some lower or entirely external, is called the noise figure. Re-
intermediate frequency where stability is more ducing the noise generated within the hf receiver to
readily attained. Like selectivity, stability is then the ultimate would avail you little or nothing in
mainly a matter of the design of the lower -frequen- weak -signal reception on bands up through 30
cy components of the vhf receiving system. MHz, but you can improve reception in the vhf
range markedly if you can amplify incoming signals
faster than you build up the noise genetated
internally. Hence the emphasis on low -noise ampli-
fiers in vhf receiver design.
The noise output of a receiver, by itself, is of
no importance at any frequency. How much a
signal stands out above the receiver noise (called
signal-to-noise ratio) is important on any frequen-
cy, and it is compounded of many factors, inclu-
ding selectivity as well as noise figure. Noise figure,
on the other hand, is almost entirely a matter of
the design and adjustment of the first stages of a
Types of VHF Receivers 31
vhf receiving system, and it is independent of
bandwidth. We'll look into this complex business
of low -noise reception in more detail later on, but
ít is important to keep the above facts in mind.

Mechanical Considerations
The best selectivity, stability, and sensitivity are
of little use if you cannot tune the receiver
effectively. Nothing is more disconcerting than a
receiver that tunes too rapidly, or in sloppy
fashion. Backlash in tuning mechanisms is very
annoying, and the higher the selectivity the more
troublesome it becomes. These are mechanical
problems, but don't under-estimate their impor-
tance in building or choosing equipment for vhf
work. Few receivers are entirely satisfactory in
these respects, and many low-priced ones are all
but useless. The would-be vhf enthusiast will do
well to check the mechanical qualities of a re-
ceiving system with great care.
frequency, ability of a receiver to reject unwanted
Rejection of Unwanted Signals
signals may have to take precedence over other
A vhf receiver could score high in all the above desirable characteristics, particularly low -noise
categories and still be unsatisfactory if it responds figure. We may have to give the receiver help in the
to signals of other services near our bands, or form of a filter of some sort. See Chapter 15.
overloads readily when near neighbors come on the It should be obvious from what has been said
air in or near the band we're trying to use. No thus far that there is no one "perfect receiver."
receiver is completely free of spurious responses, so Even with unlimited resources and design skill at
we may have a difficult choice here. In areas of our disposal, we still must examine our own
high population density, where there may be vhf particular set of operating circumstances and objec-
men in every block, and TV, police, aircraft, fm, tives, and select equipment or techniques that offer
and other vhf services on almost every available the best overall hope for success.

TYPES OF VHF RECEIVERS

From here on we use the term "vhf" loosely. Reception on 50 through 450 MHz can be
Calling only those frequencies between 30 and 300 accomplished in many ways, but vhf receivers are
MHz by this name is a grouping more semantic of two principal types: the superregenerator and
than technical, and we will include the 420 -MHz the superheterodyne. The first may be very simple
band and even higher frequencies more often than - one tube or transistor and little else other than
not. In the light of present techniques, the logical an antenna, a tuned circuit, and headphones.
dividing line between vhf and uhf methods lies Common additions are one or more audio stages to
somewhat above our 420 -MHz band, rather than at operate a speaker, and an rf amplifier stage, to
300 MHz. improve performance and reduce detector radia-
tion. The superheterodyne may be complete in
itself, usually with four or more tubes or transis-
i tors, or it may be a combination of a vhf converter
and a communications receiver intended for use on
lower frequencies.

The Superregenerative Detector

11 Today's newcomer may not be too familiar
with this wonderful device (and probably it's just
as well!) but it was almost standard equipment in
early vhf work. To give the Devil his due, the
"rushbox" was a potent factor in popularization of
the vhf bands - and for good reason. Nothing of
comparable simplicity has been found to equal Its
ability to detect weak signals, but like all simple
devices the superregen has serious limitations. "
provides little selectivity, has a high and rather
unpleasant background noise level, and radiates a
broad interfering signal around its receiving fre-
quency. While various refinements may minimize
32 RECEPTION ABOVE 50 MHZ

(B)

Fig. 3-1 - Circuits of typical superregenerative detectors using field-effect transistor, A, and tetrode
a a

controlled by varying the drain voltage on the detector in the transistor circuit,
tube, B. Regeneration is
and the screen voltage in the tetrode or pentode. Values of L1 and Cl should be adjusted for the
frequency involved, as should the size of the rf choke, RFC1.
C2, C3 - 0.001-µF disk ceramic. Try different R1 -2 to 10 megohms.
values up to 0.005 for desired audio quality. RFC1 -
Single -layer rf choke, to suit frequency.
L2 - Small audio or filter choke; not critical. RFC2 -
85-mH rf choke.

these objectionable features, the superregenerative signal to 455 kHz, where gain and selectivity come
receiver is used today mainly where its small size, much easier. Also added here is an rf amplifier
light weight, and low power drain are all impor- stage, for improved noise figure and better sensi-
tant, as in short-range portable work. tivity.
Most tubes and transistors that do well in other You could go right to 455 kHz in the first
vhf receiving applications make good superregen- conversion, by suitable choice of frequency range
erative detectors. With tubes, tetrode or pentode for the tunable oscillator, but this gives rise to a
types are favored at 50 or 144 MHz, as variation of serious image problem. Suppose we want to listen
their screen voltage offers smooth control of to someone on 51 MHz. With an i -f of 455 kHz,
regeneration. Triodes are better for higher frequen- the oscillator would then be on 50.545 MHz, 455
cies. Typical circuits are shown in Fig. 3-1. If no rf kHz away from the signal frequency. The mixer
amplifier is used, operating conditions and the responds to signals 455 kHz on either side of the
coupling to the antenna should be adjusted so as to oscillator frequency, so someone on 50.090 MHz
permit superregeneration with the lowest satis- would interfere with the desired signal on 51 MHz.
factory power input, to hold down detector All superhets have this problem, but when a high
radiation. An rf amplifier stage ahead of the i -f is used (roughly 10 percent of the signal
detector 4v1í1 reduce or eliminate radiation, and by frequency is common) the selectivity of the first
isolating the detector from the antenna will make tuned circuits of the receiver is sufficient to reject
control of regeneration less critical. It will also add the unwanted image signal. Thus, for good image
some gain and selectivity. rejection, most vhf receivers employ double- or
triple -conversion circuits.
Superheterodyne Receivers So far we have used a tunable oscillator and
Because amplification is more efficient at low fixed intermediate-frequency amplifiers. In the vhf
frequencies than high, it is standard vhf practice to range, however, tunable oscillators may not be
use only as much rf amplification as may be stable enough for narrow-band reception, so we
needed for good noise figure, and then convert the more often use a crystal -controlled source for the
signal to a lower or intermediate frequency, to be first conversion, as in Fig. 3-2C. Here the first
amplified and detected. This is the basic super- intermediate frequéncy is variable, so the i-f,
heterodyne principle, used in nearly all radio detector and audio system (portion to the right of
reception today. the vertical broken line) may take the form of a
The simplest superhet receiver for 50 MHz is communications receiver that tunes the desired
shown in Fig. 3-2A. The antenna feeds a mixer frequency range, in this example 14 to 18 MHz.
stage operating at the signal frequency, in this case Our rf amplifier, first mixer, and crystal oscillator
50 to 54 MHz. A tunable oscillator, usually at are usually built in a single unit called a crystal -
some lower frequency (though it can be higher) controlled converter. This converter-receiver com-
supplies energy to beat with the signal and bination is the most common approach to amateur
produce an intermediate frequency (i -f) which is vhf reception in use today.
then amplified and detected. In our example the There are uses for the single -conversion recei-
oscillator is 5 MHz below the signal frequency, and ver, however, especially where simplicity, low cost,
the i -f is, of course, 5 MHz. You could build the and small size are more important than high
simple receiver with as few as two dual-purpose selectivity. Also, by use of advanced i -f design
tubes, but its gain would be low and its selectivity techniques, particularly involving the crystal -lattice
poor. You would not be happy with it for long. filter, it is possible to develop excellent selectivity
Gain can be increased with more i -f amplifier at high intermediate frequencies. Though fairly
stages, but a simple 5 -MHz amplifier, as in 2A, is expensive, the high -frequency crystal filter has
not sufficiently selective, so we go to what is much to recommend it. Requiring only a single
known as double conversion, Fig. 3-2B. Here a conversion for good selectivity, it reduces the
second oscillator and mixer convert the 5 -MHz possibility of unwanted signals being heard, and It
Transistors or Tubes? 33
makes possible optimum performance with fewer crystal -controlled, with 50 to 54 MHz as the
stages and circuits than multiple -conversion sys- tunable first intermediate frequency.
tems. It should be emphasized that the frequencies
A promising overall receiving system for the vhf given in the above discussion are examples only.
man is the use of such a single -conversion setup for Almost any combination of oscillator and inter-
the 50 -MHz band as the basic receiving unit of the mediate frequencies can be used and many factors
station. This would have a tunable oscillator; govern the choice. These will be taken up later.
converters for 144 MHz and higher bands would be

SIGNALS,
50-54MM:

AUDIO
SPAN: se n
I AMP DETECTOR

45-49MM[
Fig. 3-2 -
Development of the vhf superhet-
erodyne. The simplest receiver of this type, A,
(A)
would lack most desirable qualities. B shows a
TUNABLE double -conversion system, with rf and a second
osc
conversion added. A double conversion system
with crystal -controlled converter, or "front end" is
shown at C. Portion at the right of the broken line
can be a communications receiver capable of
tuning the desired frequency range.

(B)
TUNABLE
OSC

TRANSISTORS OR TUBES?
The proliferation of vhf and uhf transistors in reliable tubes do an entirely adequate job in
recent years has tended to give the impression that receivers for 50 and 144 MHz. Circuits for their use
anyone still using vacuum tubes in receiving equip- will be discussed in following pages.
ment is hopelessly out of date. Is there any Interest in portable or mobile work may tip the
justification today for going other than the solid- scales in favor of transistors. They have no com-
state route? There may be, especially for the fellow petition where power consumption is a factor.
who has been in the game long enough to have Also, at 220 MHz, and even more so at 420 MHz,
accumulated a considerable inventory of equip- the receiving performance of the better transistors
ment designed around tubes. Many inexpensive and is unequalled by tubes, at any price.
34 RECEPTION ABOVE 50 MHZ
The newcomer to the vhf scene probably will 220 MHz Paying a bit more fór transistors
start with solid-state receiving gear, but choosing rated for use up to 400 MHz or so may be
the right transistors for the various jobs may not worthwhile in the rf amplifier, at least, though
by easy. Some helpful pointers can be given, and inexpensive FETS still work very well at this
since the frequency of operation is important in frequency. Expensive uhf bipolar types are capable
making the choice, the merits and weaknesses of of lower noise figure than FETs and may have a
various transistor types are given by bands. slight edge in rf amplifier stages.
Transistor Selection 420 MHz Most FETs presently available offer
noise figures around 5 dB above about 250 MHz or
50 MHz Any small-signal vhf amplifier transis- so, whip the better bipolars yield as low as 2 dB,
tor should give more than adequate gain and noise up to ''S00 MHz. In a 2 -stage rf amplifier, use
figure at 50 MHz. Bipolar types are generally rather the best transistor available for the first stage,
poor as to cross -modulation and other overloading which will then set the noise figure of the system.
problems, so the insulated -gate FET (IGFET OR A lower-specification transistor will do well enough
MOSFET) is a logical choice for rf amplifier and in the second stage.
mixer service. The junction FET (JFET) is nearly Uhf mixers will be discussed in detail later. The
as good, and circuits for its use are somewhat oscillator in a vhf or uhf converter nearly always
simpler than for the MOSFET, though the latter is operates at a frequency in the lower vhf range, so
preferable for mixer use, at least. transistor selection Is of little importance in this
144 MHz Similar to 50 MHz, in general, except stage. Any inexpensive vhf transistor will do.
that low -noise types rated for use up to at least Preferred tube types will be discussed stage by
200 MHz should be used. stage, in the section to follow.

RECEIVER "FRONT-END" DESIGN

Whether the receiving portion of the vhf station commonly asked question: "If one stage is that
is a complete receiver, part of a transceiver or good, wouldn't two be better?" As far as ability to
transverter, or a converter ahead of a communica- detect weak signals in concerned, nothing would be
tions receiver for lower frequencies, the part of it gained by adding another rf amplifier stage.
that we'll be concerned with here is described in There are legitimate reasons for having more
connection with Fig. 3-2B or C. Only that portion than one stage. Some types of amplifiers that are
on the left side of the broken line in Fig. 3-2C is desirable in other ways may not have enough gain
important to us at this time, as this front end is a with one stage. And on frequencies above 200 MHz
major factor in the effectiveness of the receiving or so, adequate single -stage gain is increasingly
system. The hf, i -f, detector, and audio portions difficult to obtain. If more than one stage is
are covered thoroughly in any modern edition of required, only the first one need have the very best
the Radio Amateur's Handbook. We concentrate low-noise performance. In the example above, if
here on the vhf front-end circuits: the rf amplifier, the first -stage gain were only 10 dB (as might well
mixer, and oscillator. Combined in a single unit to be the case in a 432 -MHz converter) a second -stage
be used as a communications -receiver accessory, noise figure of 4 to 5 dB would be satisfactory, if
these comprise a vhf or uhf converter. (The mixer the total gain of the two stages is 15 to 20 dB.
stage of this unit may be called a "converter" in Typically this could mean a ten -dollar transistor in
some literature. We will call it a mixer here, and the first stage and a two -dollar one in the second.
use the term converter only in connection with the Much the same situation pertains in vacuum -tube
complete front-end circuitry.) amplifiers.
If more gain is required in the overall receiving
RF Amplifiers system than that needed to establish the noise
Each of the three stages of the converter plays figure, with a little to spare, it can. be made up in
an important role in achieving the objectives the i -f stages more readily and at less expense than
described in the first paragraphs of this chapter. in the front end. An i -f amplifier built into the
The main job of the rf amplifier is to establish the converter is a useful accessory, since it permits the
noise figure of the system. To do this it must gain of the converter to be adjusted to the
amplify incoming signals, while generating as little optimum for the receiver in use. Examples of
noise as possible on its own. A typical vhf amplifier converters with i -f amplifiers built in are given in
using a good transistor may have a gain of 15 dB, the following chapter. A total rf-stage gain of more
while generating only 2 dB of noise. If the gain than 20 dB ahead of the first mixer is seldom
minus the amplifier noise (15 - 2 = 13, in this required.
case) is more than the noise figure of the rest of
the receiving system, the small noise level of the rf Low -Noise Transistors and Tubes
stage will mask the noise generated in all following Transistor data sheets often list a confusing
stages. array of specifications, most of which need not
The average good vhf mixer will have a noise concern us here. Usually the noise figure and gain
figure of 7 to 10 dB, so the amplifier in the above that can be expected at various frequencies are
example is adequate for the job. This answers a given for transistors designed for if amplifier
Receiver "Front -End" Design 35
service. It is well to select types that are rated to though there are inexpensive transistors that work
well above the frequency you're going to use. well at 220. The best obtainable may be worth
There is no point in using an expensive 500 -MHz their cost at 420, and two -stage amplifiers are more
transistor on 50 MHz, but don't use a cheap one often used.
rated for up to 30 MHz, either. Tubes may not have noise -figure ratings given in
The ultimate in low -noise figure is not vital in a their characteristics sheets. If not, look for the
50-MHz stage, as external noise is still a limiting transconductance, which should be 5000 or more
factor at this frequency. Noise from the antenna is for a pentode, and 10,000 or more for a triode, for
amplified along with the 50-MHz signals. A system best results. Pentodes such as the 6AK5 and 6CB6
noise figure of 5 or 6 dB is plenty good enough, as work well at 50 MHz. A pentode rf stage would be
it will permit reception of several dB of "antenna very similar to the mixer circuit, Fig. 3-6C. Triodes
noise," even in the quietest location. Optimum are preferred for 144 and higher frequencies.
50-MHz reception is thus possible with dozens of Several dual triodes (6BQ7, 6BZ7, 6BC8, 6BS8,
inexpensive tubes or transistors, and a single and others) will do for 50 or 144 MHz. Currently
amplifier stage is adequate in most 50 -MHz front the most -used single triode is the Nuvistor (6CW4,
ends. 6DS4). Small in size and well adapted to circuit -
Good rf amplifier performance is still possible board use, it is probably the best low-cost triode
with a single stage at 144, though two may be available for vhf amplifier service.
needed In grounded -gate (grounded -grid) applica-
tions. Transistor selection becomes somewhat more
Amplifier Circuits
critical, and low noise figure is more important. Most transistor and triode rf amplifiers require
Still, the ultimate is not required, and a system neutralization. This is done inductively in the
noise figure of 3 dB is adequate. Transistors rated examples of Fig. 3-3, but capacitive methods are
to 200 or 250 MHz should suffice. usable, and examples appear in the following
At 220 MHz or higher, the better low -noise chapter. The JFET (A) MOSFET (B) and the
types rated to 500 MHz or so begin to pay off, triode (C) will be seen to be quite similar, except
Ln
2N4416
MPFI07
2N5245 CI

100
fI M FE 3008,
01, EC Ln1. 40673 etc.

CR OUTPUT

I N PtÑ

V
47K
X012
- Neutralized vhf amplifier stages. The
Fig. 3-3
JFET amplifier, A, and the triode tube version, C,
will be seen to be very similar. The MOSFET, B,
requires a few more components, but has good
resistance of overloading.

EXCEPT AS INDICATED, DECIMAL VAWES OF

CAPACITANCE ARE IN MICROFARADS (RF) ;
OTHERS ARE IN PICOFARADS (DF OR PLF1;
RESISTANCES ARE IN OHMS;
k 1000, M 1000 000.
36 RECEPTION ABOVE 50 MHZ
LI

Fig. 3-4 - Grounded -gate FET pream-

plifier and its grounded -grid triode equiv-
alent tend to have lower gain and broad-
er frequency response than other ampli-
fiers described. Use of ferrite-bead rf
chokes in the heater circuit is shown in
B. Select values of R1 and R2 to suit
tube type. The tuned input circuit,
L1C1, can be eliminated and the antenna
coupled directly to the input element,
for extreme simplicity, as in remote
amplifiers.

RFC RFC .001

6.3V

for the heater circuit in the tube version. All may be used wherein a tuned input circuit is eliminated,
be used effectively from 50 through 225 MHz. Coil and the low -impedance input is direct to the
and capacitor values depend on frequency. Opera- cathode or corresponding transistor element. If
ting conditions and adjustment of neutralization selectivity is needed in the input circuit, it is
should be set up for minimum noise figure, rather customary to tap the input element down on the
than maximum gain. tuned circuit, as in 3-4B. Both the cathode tap and
Interstage coupling in 3-3A is intended for the point of connection of the antenna line should
band-pass effect. The value of Cl should be be adjusted for minimum noise figure, though
adjusted for the desired bandwidth. Similar output - neither will be particularly critical.
coupling circuitry could be used with B and C. The The virtues of both types of amplifiers are
low -impedance coupling shown for these circuits is combined in the popular cascode circuit, in which
for preamplifier service, where a coaxial line is run a neutralized first stage works into a grounded -grid
between the amplifier and the receiver with which second stage. The latter loads the former heavily,
it is to be used. resulting in a broad -band amplifier that may be
Protection of the amplifier, shown ín B, can be stable, even without neutralization, though the
applied to any first -stage circuit. The diodes, CR, noise figure of the system is better when the first
are connected In opposite polarity between the stage is neutralized. The cascode is most often used
antenna line and ground, to conduct on either with dual triodes, but separate triodes and transis-
cycle of rf charges built up on the antenna. In tors work equally well. Two versions are shown in
circuit C, the rf choke shown in the 6.3 -volt line is Fig. 3-5, the series-cascode, B, being somewhat the
a ferrite bead, placed over the heater lead, close to simpler of the two, and most often used with dual
the socket. triodes. Optimum use of the cascode may require
An alternative to neutralization lies in the heater chokes, as in 3-5A. These must carry the
grounded -gate (grounded -base or grounded -grid) heater current without voltage drop. Ferrite beads
circuit of Fig. 3-4. The "grounded" element may are ideal, though handwound chokes also are
be connected directly to the chassis, or bypassed usable.
thereto, so long as it is maintained at ground Instability may arise in a cascode circuit as a
potential for rf. Bypassing may be critical, and result of ineffective bypassing of the grid, base, or
only the best rf bypass capacitors (button-mica or gate in the second stage. Such oscillation cannot be
ceramic feedthrough types) should be used. Cer- corrected by adjustment of the neutralizing coil,
amics of the disk variety are seldom effective above Ln. Series -resonant bypassing may be helpful when
about 100 MHz or so. Their use may encourage this is encountered.
oscillation, as the bypassed element must isolate
the input and output circuits. Adjusting RF Stages
When a vacuum tube is used, the heater circuit
should be kept above ground potential, along with The first step in adjusting an rf amplifier is to
the cathode. This is achieved through use of set the tuned circuits at the approximate frequen-
low -resistance heater chokes, which may be ferrite cy. This can be done with the aid of a dip meter, or
beads, as in 3-4B. by peaking the circuits for maximum response
The input impedance'of this type of amplifier is while actually receiving. In circuit A, the only
low, and it is broad -band and relatively low -gain, additional step is to adjust the tap position on Ll
by nature. Two stages may be needed to mask for best noise figure or signal-to-noise ratio. Noise
mixer noise, on this account. Where only a small figure adjustments are best made with a noise
amount of gain is needed, a simplified version may generator, as detailed in the test equipment chap-
 Using Rf Preamplifiers 37
ter. The work can be done on a weak steady signal, reasonably flat response across the desired band-
if care is used to get the greatest margin of signal width. Better bandpass characteristics are obtain-
over noise, rather than merely maximum gain. able with the double -tuned circuit in amplifier A of
In adjusting antenna coupling, the best noise Fig. 3-3. Its coupling capacitor, Cl, can be varied
figure may be found as the coupling is increased a to provide different degrees of coupling and
bit beyond the point where maximum signal is selectivity. Usually values of 1 to 3 pF are used.
obtained. The gain and signal level may drop These are readily obtained by twisting insulated
slightly, but the noise falls off faster at this point, leads of hookup wire together for a half inch or so.
so signal-to-noise ratio actually Improves at the
start of over-coupling. Gain can be made up Using Rf Preamplifiers
anywhere in the receiver, but noise figure is set by It is important to design the front-end stages of
the first stage, and predominantly in the first tuned a vhf receiver for optimum performance, but we
circuit. This cardinal principle of vhf receiver often want to improve reception with equipment
design should be borne in mind at all times. already built. Thousands of fm receivers formerly
In neutralizing rf stages other than the series in commercial service, now revamped for amateur
cascode, the operating voltages can be removed work in the 50-, 144-, and 420 -MHz bands, were
from the stage to be neutralized, and the coil Ln built before modern low -noise tubes and transistors
adjusted for minimum response on a strong signal. were available. Though otherwise useful, these
All other circuits are peaked for maximum re- receivers have excessively high noise figure. Many
sponse. There is interlocking of adjustments, so other commercial and home -built vhf converters
repeat these operations several times. Now put the and receivers are not as sensitive as they might be.
neutralized stage back into normal operation, and Though it would be better to replace the rf
recheck all adjustments for optimum signal-to- stages of such equipment with more modern
noise ratio, or lowest noise figure. Merely tuning devices, the simpler approach is usually to add an
for maximum gain can degrade the noise figure by outboard rf amplifier using a low -noise tube or
several dB, even when the amplifier is not actually transistor. In the fm example, the quieting level of
oscillating. some receivers can be improved by as much as 10
In any amplifier that is working properly, dB by addition of a simple transistor amplifier.
adjustments other than on the input circuit and Similar improvement in noise figure of some
neutralization of the first stage affect only gain. receivers for other modes is also possible; particu-
For this reason it is usually practical to stagger - larly band -switching communications receivers that
tune any following circuits, if necessary, to obtain have vhf coverage.

.>T r
so
1-10

.00>,00
(a)
Fig. 3-5 -Cascode amplifier circuit combines grounded -cathode and grounded -grid
and low -noise figure. Though tubes are shown, the cascode principle is usable withstages, for high -gain
transistors as well.
Examples are given in later constructional information. Rf chokes must be able to carry
current. the tube heater
38 RECEPTION ABOVE 50 MHZ
Any of the basic amplifier circuits of Fig. 3-3, Front -End Protection
4, and 5 can be adapted to preamplifier service.
Examples of amplifier construction are given in the The first amplifier of a receiver is susceptible tp
following chapter. Circuits shown in the vhf damage or complete burnout through application
converters described can also be used in preampli- of excessive voltage to its input element by way of
fiers. the antenna. This can be the result of lightning
Preamplifiers are useful mainly with older discharges (not necessarily in the immediate vicin-
equipment which is deficient in front-end gain or ity), rf leakage from the station transmitter
noise figure. Most vhf gear built in recent years through a faulty send-receive relay or switch, or rf
should be satisfactory in these respects, if it is power from a nearby transmitter and antenna
operating properly. Checks with a preamplifier system. Bipolar transistors often used in low -noise
should be made to determine if an improvement in uhf amplifiers are particularly sensitive to this
reception of very weak signals is possible. It is the trouble. The degradation may be gradual, going
margin of signal over noise, with and without the unnoticed until the receiving sensitivity has be-
amplifier, that counts. Any amplifier added to any come very poor.
receiver will result in higher S -meter readings, but No equipment is likely to survive a direct hit
these are meaningless unless weak -signal reception from lightning, but casual damage can be prevented
actually improves. by connecting diodes across the input circuit as
When a preamplifier is tried with the receiving shown in Fig. 3-3B. Note that these are in opposite
portion of a transceiver or transverter, be sure that polarity, to protect against damage during either
it is connected in the line to the receiving portion half of the cycle. Either germanium or silicon vhf
only. An external amplifier connected in the main diodes can be used. Both have thresholds of
line to the antenna will be damaged or ruined if the conduction well above any normal signal level,
transmitter portion of such equipment is accident- about 0.2 volt for germanium and 0.6 volt for
ally turned on. This has happened to many silicon. A check on weak -signal reception should be
transceiver owners who should have known better! made before and after their connection.

MIXERS
Conversion of the received energy to a lower some conversion loss. This must be added to the
frequency, where it can be amplified more effic- noise figure of the i -f amplifier following, to
iently than at the signal frequency, is a basic determine the overall system noise figure. Low -
principle of the superheterodyne receiver. The noise design in the first i -f stage is thus mandatory,
stage in which this is done may be called a for good weak -signal reception with a diode mixer
"converter," or "frequency converter," but we will having no rf amplifier preceding it. Purity of the
use the more common term, mixer, to avoid heterodyning energy and the level of injection to
confusion with converter, as applied to a complete the mixer are other factors in the performance of
vhf receiving accessory. Mixers perform similar diode mixers.
functions in both transmitting and receiving cir- Balanced mixers using hot -carrier diodes are
cuits. Transmitting applications will be found in capable of noise figures 1 to 2 dB tower than the
other portions of this book. best point -contact diodes. Hot-carrier diodes are
A receiver for 50 MHz or higher usually has at normally quite uniform, so tedious selection of
Ieast two such stages; one in the vhf or uhf matched pairs (necessary with other types of
converter, and usually two or more in the com- diodes) is eliminated. They are also rugged, and
munications receiver that follows it. We are con- superior in the matter of overloading.
cerned with the first mixer. Whether it works into The i -f impedance of a balanced hot-carrier
a communications receiver or the i -f stages of a diode mixer (Fig. 3-6B) is on the order of 90 ohms,
complete vhf or uhf receiving unit is not important when the oscillator injection is about one milli-
here. watt. Thus the mixer and a transistorized i -f
amplifier can be separated physically, and con-
The Diode Mixer nected by means of 93 -ohm coax, without an
There are many types of mixers, the simplest output transformer.
being merely a diode with the signal and energy on Conversion. loss, around 7 dB, must be added to
the heterodyning frequency fed into it, somewhat the noise figure of the i -f system to determine the
in the manner of the 1296 -MHz example, Fig. overall system noise figure. Unless a low -noise
3-6A. The mixer output includes both the sum and preamplifier is used ahead of it, a communications
the difference frequencies. Either can be used, but receiver may have a noise figure of about 10 dB,
in this application it is the difference, since we are resulting in an overall noise figure of 17 dB or
interested in going lower in frequency. worse for a vhf system with any diode mixer. A
With a good uhf diode in a suitable circuit, a good i -f preamplifier could bring the receiver noise
diode mixer can have a fairly low noise figure, and figure down to 2 dB or even less, but the system
this is almost independent of frequency, well into noise figure would still be about 9 dB; too high for
the microwave region. The effectiveness of most rf good reception.
amplifiers falls off rapidly above 400 MHz, so the An amplifier at the signal frequency ís thus seen
diode mixer is almost standard practice in amateur to be required, regardless of mixer design, for
microwave communication. All diode mixers have optimum reception above 50 MHz. The rf gain, to
 Mixers 39

100
1296 -MHz 30 MHz

TO 30 -MHz
I-FAMP

1266 M Hz
O5C INPUT

METERING
(A)
40673
MFP 3006
etc.

6AK5, 6CB6
RF (E)
AM
EXCEPT AS INDICATED, DECIMAL VAWES OF
CAPACITANCE ARE IN MICROFARADS I yF 1 ;
OTHERS ARE IN PICOFARADS I PF OR yyFI;
INJECTION
RESISTANCES ARE IN OHMS;
111000, M1000 000,

150V

Fig. 3-6 - Vhf and uhf mixer circuits. A diode mixer for 1296 MHz, with a coaxial circuit for the signal
frequency, is shown in A. CR1 is a uhf diode, such as the 1N21 series. A balanced mixer, as in B, gives
improved rejection of the signal and injection frequencies. If hot -carrier diodes are used for CR2, sorting
for matched characteristics is eliminated. The pentode mixer, C, requires low injection and works well
below aboút 200 MHz. The triode mixer, D, has a simplified 1-f output circuit, and bandpass coupling to
the rf stage. Cl and C2 can be insulated wires twisted together for about 1/2 inch, as needed.
Insulated -gate FET mixer, E, is ideal for transistor receivers.
override noise in the rest of the receiver, should be more current, the more noise. But it resists
greater than the sum of noise figures of the mixer overloading better than one whose current drain is
and the i -f system. Since the noise figure of the held low, in the interest of good noise figure. Most
better rf amplifiers will be around 3 dB, the gain vhf and uhf converters have rf amplifiers, so their
should be at least 20 dB for the first example in mixers are normally adjusted to reduce overload-
the previous paragraph, and 12 dB for the second. ing, whether the device used is a transistor or a
These facts show clearly the worth of building a vacuum tube.
low -noise i -f amplifier into any uhf converter. It In amateur vhf reception, overloading may
makes the job of the rf amplifier, a critical stage at come from near neighbors operating in the same
best, considerably easier. band, or from strong commercial signals on the
frequencies adjacent to the amateur band. In -band
Transistor and Tube Mixers signals can be prevented from overloading the
Any mixer is prone to overloading and spurious mixer only by adjustment of the mixer operating
responses, so a prime design objective should be to conditions, and by keeping the rf amplifier gain to
minimize these problems. Several factors enter into the minimum needed for good noise figure. Out-
this. Choice of the transistor or tube to j1e used, of-band signals can be held down by use of
the selectivity of the circuits preceding it, and the selective circuits in the antenna and if amplifier.
purity and level of the injection energy are Double -tuned coupling circuits for this purpose are
important. Usually a mixer can be set up for good shown in some of our rf amplifier and mixer
noise figure or good resistance to overloading, but examples, herewith.
not both. In a very simple receiver having no rf The injection level from the oscillator affects
amplifier, the mixer would be set up for low noise mixer performance, though it is not critical when a
figure. Where an effective if amplifier is used ahead good rf amplifier precedes the mixer. Until it
of it, the mixer can be tailored for best overloading affects the mixer adversely in other ways, raising
and cross-modulation characteristics. A mixer that the injection level raises the mixer conversion gain.
draws appreciable current tends to be noisy; the A simple check is made by observing the effect on
40 RECEPTION ABOVE 50 MHZ
signal-to-noise ratio as the injection is varied. At taking off the i-f output signal are shown in Fig.
preferred injection levels, the gain will vary but the 3-6 mainly as examples, rather than because of any
signal-to-noise ratio will not change. The injection advantage of one method or the other.
should then be set for conversion gain a few Oscillation, most commonly found in triode
decibels above that at which lower injection causes mixers, may result from inadvertent resonances
a drop in signal-to-noise ratio. near the signal frequency in the output circuit. The
Pentode tubes make simple and effective mixers 10-pF capacitor shown connected from the output
for 50, 144, or even 220 MHz. Triodes work well element to ground in some of our circuits is
on any frequency up to about 500 MHz or so, connected close to the mixer plate, collector, or
though diode mixers are commonly used in drain and to ground, with the shortest possible
420 -MHz converters and for all higher frequencies. leads, to break up this tendency to resonance, and
The pentode mixer of Fig. 3-6C gives good prevent oscillation. When such oscillation does
conversion gain and is readily adjusted for low occur it is often mistaken for rf-stage instability,
noise figure or resistance to overloading, by chan- and much time may be wasted trying to neutralize
ging the value of Rl. It may be as high as one -
it out in the wrong stage.
megohm, for low current drain and resultant The insulated -gate FET (IGFET or MOSFET,
low -noise figure, or as low as a few thousand ohms, 3-6E) is superior to other transistors in resistance
for best overload characteristics. Bias and supply to overloading, though the JFET is also good. An
voltage variations have similar effects on triode and objection to the former, the ease with which it can
transistor mixers. be damaged by inadvertent application of small
The triode mixer, D, is similar to the pentode, voltages in handling, has been taken care of by
except for the screen circuit in the latter. Various building in small protective diodes. Such transistors
methods of feeding ín the injection voltage, and of work as well as their unprotected predecessors.

INJECTIO N STAGES
Oscillator and multiplier stages that supply connected in the antenna line to absorb Channel -2
heterodyning energy to the mixer should be as TV signals that could otherwise beat with the
stable and free of unwanted frequencies as pos- second harmonic of a 36 -MHz oscillator in a
sible. They are similar to exciter stages of vhf 50 -MHz converter that works into a 14 -MHz i -f
transmitters, and their basic principles are dis- (36 X 2 - 14 = 58).
cussed in more detail in Chapters 5 and 6. Stability Unwanted frequencies also increase the noise
is no great problem in crystal -controlled conver- output of the mixer. This degrades performance in
ters, if the oscillator is run at low Input and its a receiver having no rf amplifier, and makes the job
supply voltage is regulated. Simple Zener regula- of an amplifier, if used, more difficult.
tion, as in Fig. 3-7A is adequate for a transistor Frequency multipliers in vhf receivers generally
overtone oscillator. A higher order of regulation is follow transmitting practice, except for their low
desirable for tunable oscillators. power level. The simple diode multiplier of Fig.
Unwanted frequencies generated in the injec- 3-7B will often suffice. Its parallel -tuned 130-MHz
tion stages can beat with signals outside the circuit emphasizes the desired third harmonic,
intended tuning range. In a typical example, Fig. while the series circuit suppresses the unwanted
3-7B, an FET overtone oscillator on 43.333 MHz second harmonic. The trap is tuned by listening to
feeds a diode tripler to 130 MHz. This frequency a spurious fm broadcast signal and tuning the series
beats with signals between 144 and 148 MHz, to capacitor for minimum interference. The 130 -MHz
give desired responses at 14 to 18 MHz. The tripler circuit should be peaked for maximum
multiplier stage also has some output at twice the response to a 2 -meter signal. Do not detune this
crystal frequency, 86.666 MHz. If allowed to reach circuit to lower injection level. This should be
the mixer, this can beat with fm broadcast signals controlled by the voltage on the oscillator, the
in the 100 -MHz region that leak through the rf coupling between the oscillator and multiplier, or
circuits of the converter. There are many such by the coupling to the mixer from the 130 -MHz
annoying possibilities, as any amateur living near circuit.
high-powered fm and TV stations will know.
Spurious frequencies can be kept down by Tunable Oscillators
using the highest practical oscillator frequency, no
multiplier in a S0 -MHz converter, and as few as Any tunable vhf receiver must have a variable
possible for higher bands. Some unwanted harmon- oscillator somewhere along the line. (See Fig. 3-2.)
ics are unavoidable, so circuit precautions are often At this point the intermediate frequency is fixed,
needed to prevent both these harmonics and the and the oscillator tunes a range higher or lower
unwanted signals from reaching the mixer. Selec- than the signal frequency by the amount of the i-f.
tive coaxial or trough -line circuits are practical aids In the interest of stability, it is usually lower. In
in uhf receivers. Trap circuits of various kinds may Fig. 3-7C a simple JFET oscillator tunes 36 to 40
be needed to "suck out" energy on troublesome MHz, for reception of the 50 -MHz band with a
frequencies. fixed 14 -MHz i-f. Its stability should be adequate
The series trap in Fig. 3-7B reduces the level of for a-m or fm reception, but it is unlikely to meet
the 86 -MHz second harmonic of the crystal fre- the requirements for ssb or cw reception fully,
quency. A 58 -MHz parallel-tuned trap can be even at 50 MHz, and certainly not higher bands.
Communications Receiver Problems 41
40080

(A)
TO pMIXER
MPF 102 43.3 MHz
o spf1-
130 M PFi02
4,3.3
MHz Q
MHZ
36-40 .1191
-MHz

N 0 r 9.1ÓO

_!
100 +
12V
DC

86.67- MHz 9.1 -VOLT

TRAP ZEN ER
-0 TO
BUFFE R
STAGE
Small
(B) +12V
(REG)
(C)
Fig. 3-7 -
Typical oscillator and multiplier circuits for vhf converters. A bipolar transistor with
Zener-diode regulation of the collector voltage is shown at A. The indicated frequency is for 144 -MHz
converters with 14 -MHz output. The FET circuit, B, includes a diode tripler to 130 MHz, with a series
trap to absorb unwanted second -harmonic energy at 86 MHz. A triode tube would require essentially the
same basic circuit. The tunable oscillator, C, would be suitable for a 50-MHz converter with fixed 14 -MHz
i -f.

Using a tunable oscillator and a fixed inter- Most vhf reception with high selectivity is with
mediate frequency (Fig. 3-2A or B) does have double -conversion setups, with the tunable oscil-
merit. It can be used with a crystal -controlled hf lator serving the second conversion, as in Fig. 3-2C.
receiver, or with any communications receiver that Such hf oscillators are treated in detail in The
has inadequate tuning facilities. Covering entire vhf Radio Amateur's Handbook, and in Chapter 6 of
bands with communications receivers having limi- this book. The oscillator should run at the lowest
ted tuning range is an example. With the i -f fixed, practical input level, to minimize drift due to
the precise frequency to be used can be selected heating. The supply should be well -regulated pure
carefully to prevent interference from signals riding dc. Mechanically rugged components and construc-
through in the hf range. Better selectivity at the tion are mandatory. The circuits should be shielded
intended receiving frequency is also possible, if the from the rest of the receiver, and coupling to the
rf and mixer -tuned circuits are gang-tuned with the mixer should be as light as practical. Drift cycling
oscillator. The front-end stages can then be de- due to heating can be minimized if the oscillator is
signed for optimum selectivity across the band, and kept running during transmitting periods. Leaving
no broad -banding of these circuits is required. the entire converter running is even better.

COMMUNICATIONS R ECEIVER PROBLEMS

Most of the information in this chapter has to Several degrees of selectivity are desirable: 500 Hz
do with crystal -controlled converters, since this is a or less for cw, 2 to 3 kHz for ssb, and 6 to 8 kHz
common approach to vhf reception. Unless the hf for a-m and fm phone are useful.
receiver with which the converter is to be used is Stability: Tuning with converters imposes no
satisfactory in the qualities discussed in the first special stability problems. Some receivers, particu-
paragraphs of this chapter, the best vhf converter larly older or inexpensive ones, may have pro-
design will be largely wasted. Let us consider the gressively poorer stability on each higher-frequency
communications receiver from the point of view of band, so this may be a factor in choosing the
the vhf converter user. converter output frequency range for use with a
Selectivity: If a receiver is satisfactory for the given receiver.
hf amateur bands, it will be selective enough for Sensitivity: The ability of a vhf receiving setup
vhf service. Most communications receivers will to detect weak signals is determined almost en-
satisfy all but the most critical users in this respect. tirely by the first stage of the vhf converter. Any
42 RECEPTION ABOVE 50 MHZ
communications receiver worthy of the name will
have more gain and sensitivity than you'll ever
need in using a converter with it.
Mechanical Qualities: These probably rate first
in choosing a receiver for vhf converter use. In
general -coverage receivers the tuning. rate of the
dial system is very important to the vhf man,
whereas it may be only a minor consideration to
other users. Such receivers ordinarily cover from
the broadcast band through 30 MHz, usually in
four to six ranges. On inexpensive receivers having
four bandswitch positions, each range covers such a
wide frequency spread that the number of kilo-
hertz per rotation of the tuning knob is almost
certain to be excessive. Receivers having five bands
may be better, but usually the tuning rate will be
Fig. 3-8 - Typical mounting for a vernier dial to be
used on a general -coverage receiver. Mechanical
too high on frequencies above about 10 MHz or so. arrangements can be worked out to fit most
This may be a factor in selecting the most desirable receivers without the necessity for drilling holes or
converter output frequency range. Some of the otherwise permanently disfiguring the receiver.
better general -coverage receivers have six or more Adjacent control shafts and nuts provide conven-
ient anchorages.
bandswitch positions. Only these give adequate
tuning smoothness fur the vhf man, ordinarily.
There are ways around this problem. One is to than at the low. Up through about 10 MHz, most
mount a vernier mechanism in place of the knob receivers are selective enough in their rf circuits so
on the general-coverage dial. Usually a mounting that the image is not bothersome, so a converter
arrangement can be worked out that will not output frequency range beginning at 7 MHz, for
disfigure the receiver in any way. Two examples example, may be preferable to one starting at 14
are shown in Fig.s 3-8 and 9. The first utilizes a MHz.
5 -to-1 vernier drive, the other a two -speed drive. Most receivers made prior to the late 1940s
An alternative approach with two -dial receivers were single -conversion models. Except for the
is to do the tuning with the bandspread dial, image problem, many such older receivers are quite
resetting the general -coverage dial as each addi- satisfactory for vhf converter service, so this is an
tional swing of the bandspread mechanism is made. important factor in selecting the converter output
This gives good tuning rate, but after the first frequency. An old but good receiver may be a
swing the calibration is lost. Because the two are better value for the money than an inexpensive
connected in parallel, changing the setting of the newer model, since the vhf man is primarily
general -coverage capacitor changes the number of concerned with receiver qualities that have not
kilohertz a given number of revolutions of the changed greatly in many years of receiver develop-
other capacitor will cover. A better solution is to ment. Some receivers dating back even to before
use separate crystals in the converter, for each World War II still serve the vhf man's needs quite
desired frequency range. well, if the converter output frequency is kept low.
Some receivers may not be completely shielded,
Rejection of Unwanted Signals with the result that stations operating in the
Image rejection is the principal concern here. converter output frequency range may be heard
Single-conversion receivers with 455 -kHz i -f sys- along with the desired vhf signals. This is a
tems have relatively poor image rejection, so there converter design problem as well, but corrective
is a tendency for strong signals to be repeated 910 measures may have to be applied at the receiver.
KHz away from the desired response. This trouble The three -terminal antenna connection plate used
is worse at the high end of the receiver's coverage on many communications receivers is shown in Fig.
3-10. Usually Terminal 1 is connected to the
chassis inside the receiver. This connection may act
as a coupling loop for i -f signals. Removing the
internal connection and grounding terminal 1 on
the outside of the chassis may correct this.
A better solution is to install a coaxial fitting
,d.º° on the rear wall of the chassis. When this is done,
the lead normally connected to Terminal 2 inside
the receiver should be permanently grounded, and
.
E the lead that went to Terminal 3 should be
connected to the inner conductor of the coaxial
fitting.
Receivers may pick up signals through the ac
line to some extent. Bypassing the ac lead inside
Fig, 3-9- Example of use of a two-speed planetary
mechanism to slow down the tuning rate on the
the chassis should take care of this source of i -f
leak-through. Some receivers, such as early models
general -coverage dial of an inexpensive communica- of the HRO, have separate power supplies. Power
tions receiver. cable leads may pick up signals, and also radiate
Receiving Frequency Modulation 43
harmonics of the receiver oscillator that may show
up in the tuning range of the vhf converter.
Filtering and bypassing power supply leads is
required in these instances, much as would be done
in TVI-prevention work on a transmitter. See
Chapter 15 for examples.
To tell whether signals at the intermediate
rFrt, r14,
frequency are coming through because of receiver
deficiencies or directly through the vhf converter, Rivet Rear Wall
or of Chassis
try running the receiver alone with its antenna Screw
terminals shorted. If no signals are heard in this
condition it can be assumed that the trouble lies in
the converter. The problem may be one of im-
Fig. 3-10 - The 3-terminal antenna connection
plate used on many communications receivers may
proper bonding between the receiver and converter be a source of interference from signals picked up
chassis. The outer conductor of the coax inter- at the converter output frequency. Remove the
connecting the units may not be enough. Place the loop from Terminal 1 to ground (inside the
converter close to the receiver, and then bond the receiver) and connect 1 to ground on the outside
two together with a heavy braid or strap of copper. of the chassis. Ground the -f coax to Terminals 1
i

Often this will eliminate the leak -through trouble. and 2, and connect the inner conductor to Ter-
minal 3. Better still, install a coaxial fitting on the
Effective bypassing of the converter power circuits rear wall of the chassis, to replace the terminal
is another step If the bonding fails. Use 0.01-11F board entirely.
disk capacitors for this purpose. Lower values may
not provide complete bypassing at frequencies those that can be covered readily on ranges where
under about 10 MHz. their receivers work best. This might mean using a
receiver's 400 -kHz coverage at 14 MHz with two
General -Coverage is. Amateur -Band Receivers converter crystals, for example, to give 144.0 to
Except for a few models that have a special 144.4 MHz and 145.0 to 145.4 MHz, or any other
tuning range for vhf converters, receivers designed segments that may happen to fit a local activity
for amateur -band service exclusively do not meet picture.
the vhf man's tuning needs fully. The obvious Some communications receivers themselves use
solution to this problem is to use more than one crystal -controlled oscillators with a tunable i -f
crystal in the vhf converter. Two crystals will give system. Crystals are usually supplied to cover the
the required 4 -megahertz spread with receivers that amateur bands from 3.5 to 30 MHz with such
tune 28 to 30 MHz, for example. This may not be receivers. In some instances (as for example the
entirely satisfactory, however, as the performance Collins S -Line), crystals can be obtained to extend
of some receivers is poor in the 28 -MHz range the continuous coverage. Continuous coverage
compared to lower frequencies. from 14 to 15.8 MHz is possible with a 75S-1 or
Vhf men who want the best reception with a S-3, for example, by substituting suitable crystals
minimum of crystal changing may decide to forfeit In the 21- and 28-MHz positions for those supplied
coverage of some parts of the vhf bands in favor of with the receiver.

RECEIVING FREQUE NCY MODULATION

Effective transmission and reception of fm The bandwidth used in fm broadcast reception
require techniques very different from those used is ideal for bands where there is room for it,
with other voice modes. Frequency modulating a particularly 220 MHz and higher. Suitable fm
transmitter is quite simple, but the i -f and detec- tuners for use with 220- or 420 -MHz converters are
tion circuits for fm are not readily incorporated everywhere, in the form of home -entertainment
into most communications receivers. The fm re- units, from pocket portables to expensive hi-fi
ceiver is specifically designed to reject all forms of tuners. Several communications receivers still
amplitude modulation, and the receiver bandwidth found on the used-equipment market are usable.
and the transmitter deviation must be matched The Hallicrafters SX-42, SX-62, and SX43 are
closely. Unless both qualities are provided, fm will ideal, as they offer both wide -band fm and
be markedly inferior to any other voice method. narrow -band a-m and cw in the 27 -MHz range, at
With well -coordinated design, vhf fm systems are the flip of the mode switch. Converters described
unmatched in their ability to provide reliable in Chapter 4 can be used with such receivers, for
noise -free communication over a local service area. multipurpose work. The S-27 and S-36 cover
Techniques that resulted in the current boom in similar frequency ranges, but do not have narrow-
fm and repeaters are described in detail elsewhere band cw or a-m capability. All these receivers are
in this book. We call attention here to a system fine fm broadcast sets, as well. The- S-240 is a
that has gone all but unused in amateur work, but modern fm a -m receiver that should work well if
which has great potential for reliable high-quality the amateur -band converter is modified for
voice communication on the higher bands. This is 88 -MHz output.
true wide -band fm, using equipment that is well Very simple transmitting equipment for true
within the capabilities of the relative newcomer to wide -band fm is discussed in Chapter 5.
home-building of ham gear.
 Chapter 4

V h f Receivers, Converters and

Preamplifiers

This chapter will present practical examples of currently heard in the 50- and 144 -MHz bands, and
principles discussed in detail in Chapter 3. To they are not compatible to any great extent.
conserve space and show a wide variety of projects The superregenerative tuner covering 14 to 18
for the home-builder of vhf gear, we will lean or perhaps 26 to 30 MHz, once an accepted
heavily on the previous chapter for explanatory simple -receiver approach for the beginner,1 is
material. It should be stressed here that building ineffective for these modes. It will work reasonably
equipment from this book, like working from well for a-m phone reception, but it is no more
nearly all material published by ARRL, bears little than an annoying makeshift for cw, ssb, or fm. The
resemblance to building from kit -type instructions. direct -conversion receiver, which has had quite a
Our purpose is not merely to provide "plans" for play in the simple -receiver field,2 does pretty well
vhf equipment, but rather to give the builder an on ssb or cw, but it is almost useless for fm, and
opportunity to learn something about design, not much better for a-m, especially if the frequen-
adjustment, and operation of the equipment, be- cy stability of the a-m signal is anything but the
yond the usual kit-builder's skills in the use of a best. A complete solid-state receiving system for all
soldering iron and simple hand tools. 1 For this and other numbered references, see
It is possible to build equipment described in bibliography at the end of this chapter.
this book without first acquiring some knowledge
of how the circuits work and why they are
designed the way they are, but the better method TABLE 4-I
is to get your money's worth from the principles Crystal and injection frequencies for use with
chapters. If you have not already done it, we common intermediate frequencies in vhf converters.
strongly recommend reading Chapter 3 thoroughly,
particularly those portions relevant to any project
you are about to embark on in the vhf receiving Band, Crystal and multiplier frequencies for i -f
field. This may well save you time and trouble, in MHz beginning at:
the end. An understanding of principles is basic to 7 MHz 14 MHz 28 MHz
success with projects to be described here. Multi Multi
Osc Multi Osc Osc
Though this is termed a "vhf" chapter, some
equipment for the 420 -MHz band is included,
where circuit features are not unlike those of
50 43.0 - 36.0 - 22.0 -
144 45.667 137 43.333 86.67 38.667 136
equipment for lower bands. The 420 -MHz band 220 53.25 213 51.5 206 48.0 192
tends to be borderline in nature, so other items for 432 Undesirable 46.444 418 44.889 404
this band will be found in the UHF and Micro-
waves chapter, where techniques employed are
more characteristic of higher bands.
RECEIVING SYSTEMS TABLE 4 -II
Nominal inductance of coils required for typical
A means of listening is the first requirement of converter circuits of Chapters 3 and 4. Resonant
the vhf newcomer. If he already has an adequate frequency is varied by movement of coil slug, or
receiver for the lower frequencies, a vhf converter (in air -wound coils) by changing capacitor value.
is his logical first step into the world above 50
MHz. The stringent requirements for good vhf
Freq, MHz Nom. L, pH Freq, MHz Nom. L, µN
reception do not lend themselves to simple sol-
utions otherwise, mainly because of the variety of 7-11 27 36-40 1
modes open to the vhf communicator. Fairly 19 4.5 40-45 0.8
simple receivers for a -m phone can be built, but 22 - 26 3 45 - 50 0.7
they will not work well for fm reception, and 26-28 2.5 50-55 0.6
probably not at all for ssb. The latter two modes 30.5 1.5 136 - 139 0.15
account for a considerable portion of the activity

44
Simple JFET Converter for 50 MHz 45
amateur bands up through 148 MHz is described in handled, rather than because of any marked
general terms in QST for January, 1971, and more advantage of one i-f with a particular converter.
completely in recent editions of the Handbook3 - The choice is more likely to be dictated by the
but it is far from being a beginner's project, and it nature of the receiver the prospective converter -
makes no provision for fm. A tunable converter for builder owns.
14 MHz, to work into car-radio or home receivers The 50-MHz converter immediately following
for the broadcast band4 could be used as a this section shows use of 7 MHz as the low end of
limited-range tuner for vhf converters having the tuning range. This might be fine with some
14 -MHz output. It works well with ssb, a-m, and receivers, but 14 MHz could be a better choice
cw, and not too badly as a slope -detection method with others, or 28 MHz with still others. Our
for fm. Probably the closest thing to a universal practical constructional examples show converters
system is discussed briefly in a QST article on with various intermediate frequencies. Tables I and
integrated circuits.5 It does well on all four modes, II present the basic information needed to build -in
and it is relatively simple to use, considering all the the injection and output circuits to fit your needs.
things it is capable of doing, but it hardly fits the
"A Simple . . . ." mold, so popular in beginner Receiving with Vacuum Tubes
articles of a bygone era.
Receiving systems and adaptations primarily for It will be apparent to the reader that receiving
fm appear in our chapters devoted to that mode, equipment described in this book is almost entirely
later in this book. What follows in this chapter is of solid-state design. This is not to imply that vhf
presented in terms of a separate converter used converters using tubes are no longer worth con-
ahead of a communications receiver, but the same sidering. Rather, it was felt that with an excellent
basic circuits apply to the receiver front-end selection of vacuum -tube designs readily available
portions of transceivers and transverters. from QST, the Handbook, and previous editions of
this Manual, available space could be used to better
Choosing the Intermediate Frequency
advantage by concentrating on equipment built
A portion of the preceding chapter goes into around transistors.
the problems encountered with various types of The bibliography at the end of this chapter lists
communications receivers, when they are used with several vacuum -tube receiver and converter articles
vhf converters. There is no one tuning range that is that are still in demands Where the publication in
ideal for all converters or for all receivers. The question is not available to the reader, ARRL will
converters to follow have different intermediate supply photocopies of the original articles for 25
frequencies, mainly to show how these can be cents per page.

SIMPLE JFET CONVERTER FOR 50 MHz

Field-effect transistors provide freedom from Motorola, and similar types available from other
overloading comparable to the better vacuum -tube sources, rated for rf amplifier service to 100 MHz
rf amplifiers. Inexpensive junction FETs are now or higher should be more than adequate. The rf
available that give excellent noise figure and gain in transistor is operated grounded -gate, which is the
the vhf range. The FET is quite similar in charac- equivalent of grounded -grid use of a vacuum tube.
teristics to a tube amplifier, and it will be seen Such a stage is relatively low -gain, but it is stable
from a comparison of the schematic diagrams that
the simple 50-MHz converter of Fig. 4-1 is not
unlike the triode converters previously described.
It should be equal to a tube version in perfor-
mance, and it has a considerable edge in overall
simplicity, since it can operate directly from a 9- or
12 -volt battery.

Circuit Details
/ 41,w

c». 4
In the interest of bandpass response and rejec-
tion of out -of-band signals, double -tuned circuits
are used for the input and interstage coupling.
Protective diodes, CR1 and CR2, between the r Pt

antenna connection and ground, prevent damage to

the first stage from rf leakage from the transmitter,
or transient voltages such as might result from
nearby lightning discharges.
The rf and mixer transistors may be any of
several vhf FETs. The inexpensive MPF-102 from

Fig. 4-1 - 50-MHz converter with field effect

transistors in the rf and mixer stages.
46 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS

Fig, 4-2 -
Looking into the under side of the 6-meter converter the mixer is in the center, with the rf
stage at the right and the oscillator at the left.

R.F. AMP. MIXER

50-54 Mc. 50-54 MC
5 7 -II Mc.
PF -102
0
LS
J2
L6
ANT.
TO REC.

.003

J3
O+9OR12V.
.0 .14,1;D.C.
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
IN MICROFARADS (Nf. ); OTHERS
ARE IN PICOPARADS(pf.ORppf.);
RESISTANCES ARE IN OHMS; LB
K 1000 3300 D S G

2N706A 3311
.005 r O

BOTTOM E
.00,1 BOTTOM
VIEW
VIEW 1000 MPF-102

Fig. 4-3 -
Schematic of the 6 -meter FET converter. All resistors are 1/2 -watt composition. All
capacitors are disk or tubular ceramic.
L6 --
5 turns insulated wire over cold end of L5.
CR1, CR2 -
Germanium diode (1N34A suitable). L7 0.33 µH slug -tuned (Millen 69054-0.331.
-
J1, J2 --
Phono connector.
Insulated banana jack one red, one black.
L8 1 turn small -gauge insulated wire over cold
end of L7.
J3, J4
L1 -L4, incl. -
0.68 pH, slug -tuned (Millen Y1 -
43.0 -MHz third -overtone crystal (Inter-
national Crystal Co. Type F-605).
69054-0.681. L1 has tap added at 2nd turn
from ground end. * Available directly from James Millen Mfg.
L5 -11 to 24 µH slug -tuned (Miller 4507). Co.. 150 Exchange Street, Malden, MA.
Advanced JFET Converters for 144 and 220 MHz 47
and does not require neutralization. At 50 MHz, at Adjustment
least, the gain is adequate for good reception, when
the stage is used with a grounded -source mixer, as Check the wiring, to be sure that it is correct
shown. and complete, then connect the converter to the
The oscillator uses a bipolar transistor with a communications receiver input, and apply dc volt-
43 -MHz crystal. Any vhf transistor will do here. age through J3 and J4. The receiver noise level will
The mixer output is 7 to 11 MHz for coverage of increase markedly, if the converter oscillator is
50 to 54 MHz, but the 14 -MHz range could be used working. Turning the slug in L7 should bring this
equally well. In that case the coil and crystal about, if it does not start immediately. Set the slug
information could be taken from Tables I and II. at a point where oscillation will occur each time
the voltage is applied.
Construction It should now be possible to hear any reason-
ably strong signal, actually on-theair or from a vhf
The converter case is a 3 X 5-1/4 X 2 -1/4 -inch signal generator. Peak all core studs in the rf and
Minibox, with the parts mounted on the cover mixer circuits for maximum signal strength, and
portion. Shields are mounted across the chassis to the adjustment should be nearly complete. It may
keep down unwanted interstage coupling. Parts be helpful to stagger-tune L3, L4, and L5 for
layout is not particularly critical, though the uniform response across the desired frequency
approximate relative positions of the principal rf range. This results in somewhat lower than maxi-
components should be followed, for best results. mum gain, but does not affect the noise figure
Phono connectors were used for the input and adversely, as this is determined mainly by the first
output fittings, J1 and J2. These work well tuned circuit. L1 and L2 should be adjusted
enough, though you may prefer the better quality carefully for best signal-to-noise ratio on a weak
of BNC or other coaxial fittings. signal, rather than for maximum gain, if there is a
Colored tip jacks J3 and J4 on the rear wall are difference discernible.
used to bring in the operating voltage. Small In adjusting the coil slugs, be sure that the
feedthrough bushings (Johnson Rib-Loc) are circuits actually peak. Occasionally there will
mounted in the interstage shields, for leads be- appear to be a peak which is actually the centering
tween sections. In Fig. 4-2 the rf amplifier stage is of the slug in the winding. If this happens, you
at the right side of the picture, the mixer at the need more turns in the coil or more capacitance
center, and the oscillator at the left. across it.

ADVANCED JFET CONVERTERS FOR 144 AND 220 MHz

The converters of Figs 4-4 through 4-11 were vers, gives enough overall gain for use of the
designed to provide optimum performance and converter with receivers whose performance at 28
flexibility. They show examples of several tech- MHz may be questionable. More important, it
niques that can be used to advantage in vhf permits setting the converter output level at the
converter design generally, as well. Two junction optimum point for any receiver. Using an inte-
FETs are used in a transistor version of the familiar grated circuit simplifies this stage, as most of the
cascode circuit. The JFET mixer has 28 -MHz parts are in the IC itself.
output, permitting coverage of at least 2 MHz of The oscillator and multiplier stages use inexpen-
the vhf band on most receivers, with a single sive bipolar transistors, with collector-voltage regu-
converter crystal. lation on the oscillator. These stages are isolated
The 28 -MHz i -f amplifier, while not absolutely from the rf portion with copper shielding. The
necessary with the better communications recei- supply voltage is fed in through a "polarity

A A 1
/ á

Fig. 4-4 -A rook at

the completed 2 -
meter etched circuit -
FET/IC converter
and its 12 -volt ac-
operated power sup-.
ply. Top appearance
/so°
of the 220 -MHz ver-
sion is almost iden-
tical.
48 Vhf RECEIVERS, CONVERTERS, AND PREAMPLI FIERS

r,

aJ

11
0.9^ J G

Fig. 4-5 -
I
Top surface of the 2 -meter converter. The -f gain control knob is at the upper right. The
i

input jack for the 12 -volt supply is just to the left of the gain control. The i -f output jack is at the
lower right, and the rf input jack is at the lower left on the board. The IC is located at the far right,
just above the i -f output connector.

insurance" diode, CR1. You can't hurt the conver- the various sections of the converter. Where these
ter by getting supply polarity wrong. If you're sure are soldered or bolted to the circuit board it is
you'll never do this, the diode can be omitted. necessary to trim away portions to prevent short-
ing out the circuits. The shields do not show
Options clearly 'in Fig. 4-8, so their approximate location
Several JFETs will do for Ql, Q2, and Q3, can be checked out by the dashed lines in Fig. 4-6.
some of them more expensive than the MPF-102s Locations of the key components can be deter-
and 107s used. The 2N4416, a top -rated JFET, for mined from the layout drawing, Fig. 4-9, and from
example, showed no measurable advantage in the top and bottom photographs.
either the 144- or the 220 -MHz converter. A power The chassis is a matching Vector assembly made
supply is shown, for use where operation from ac of two of their Fram-Loc rails 2 X 6-5/8 inches
power is a convenience, but all -battery power is (Vector SR2-6.6.062), two 2 X 4-1/2 inches
attractive for portable work, particularly in con- (SR2-4.6/062), and a bottom cover (PL4566). A
nection with an all -transistor communications re- standard chassis could be used, if the cover plate is
ceiver. The i -f amplifier, the voltage regulation of cut out to fit the circuit board.
the oscillator, and the polarity -insurance diode can
all be considered optional. You can build a good Power Supply
converter without them, but optimum perfor-
mance, versatility and safety are assured through The converter requires about 12 volts dc at 45
their inclusion. mA. The ac -operated 12 -volt do supply for fixed -
Construction station use, Figs. 4-4 and 4-7, is built in a
The converters described above are assembled on 4 X 5 X 2 -inch aluminum box with bottom plate.
4-1/2 X 6 -1/2 -inch etched circuit boards (Vector For portable work the converter might be operated
CU65/45-1). A layout drawing, Fig. 4-9, will help from the same source as a transistorized communi-
you to make your own layout, or a template can cations receiver with which it is to be used, or from
be obtained if you wish.* Ready-made boards can a car battery. A bank of 8 D -cells will provide
be purchased.* Shields of flashing copper isolate many hours of intermittent use.
* Full-size template similar to Fig. 4-9 sent If mobile operation is planned, it would be
upon receipt of 25 cents and stamped self-add- prudent to connect an 18 -volt Zener diode across
ressed envelope. Address ARRL Technical Dept.,
Newington, CT 06111, and mention figure num- J3, to protect the transistors from transient peaks
ber publication, and edition number. which occur in automotive electrical systems.
íteadymade boards may be obtained from Under normal conditions the Zener would not
individual suppliers listed from time to time in
QST. conduct.
 Advanced JFET Converters for 144 and 220 MHz 49
Checkout The inductance of Ll and L4, and the position
of the tap on L1, should be adjusted for best
Before applying operating voltage, make a (lowest) noise figure, if a noise generator is
thorough check of the soldering operations on the available. These adjustments can be made on a
circuit board, to be sure that the job is complete weak signal, if careful observation of the margin of
and correct, and that there are no incidental shorts. the signal over noise is maintained. There is
With a test signal (generator, or on -the -air, interaction between these adjustments, so several
starting with a high level) on about 145 MHz, resettings of each may have to be made for
adjust Ll, L2, L3, L4, L6, and L8 for maximum optimum reception. Reasonably flat response
output. If the test signal cannot be heard, it is across the desired tuning range can be achieved by
likely that the oscillator, Q4, has not started. In stagger-tuning the rest of the circuits, both rf and
this case, adjust L10 until an increase in noise i -f, as only the input and neutralizing coils will
occurs, indicating the start of oscillation. Detune affect noise figure measurably.
the slug out of the coil slightly, until the oscillator The gain control, R1, should be set so that the
will start whenever voltage is applied. noise level, with no signal, just shows on the

R F. AMP R.F AMP MIXER I. F. AMP

144-I48 Mc .
I

1x TT
NEUT 114

1
Sµ I

I
MDF102
144-148MC
- CI
MDF102 24 -32 Mc .005
JI Í MPFIO2 II

T001
I /22I
LI, '[
1118117 I
p1
144-
148 kW Q2
, I z
.005
51 28-3204c
J2
I
9 1<
OUTPUT
.005
330
5 4, '>
39K
livr
500K
I F
GAIN

05C DBLR.
Q4
40237 R2 5844c.
Q6
e
40237

12v. j+,CRt
J3

(45MA)
L12
14700
.r.
rtF '
68K LACERS AS IRDICAT(0, DCCIYAL
22..s. VALUES OF CAPACITANCE ARE

[
m+
RFC2
IN
ARE
YICROFARAOS
IR
(YI. ); 0T11CRS
mcorme.11 pl.OR/IFIS1;
4E5157'4114E5 ARE 0. DRYS:
AR1 R 1000
Q6 05
Q4, Flat IQI,Q2,Q3
Area.
C 5 M = SILVER MICA

SNORT LEAD
ECe D5G
(Cat off)
BOTTOM VIEWS

Fig. 4-6 - Schematic of the 144-MHz converter. Fixed -value resistors are 1/2 -watt composition.
Fixed-value capacitors are disk or tubular ceramic unless stated otherwise.

AR1 - Motorola MC -1550 integrated circuit. Take L5 -

2 turns insulated hookup wire over ground
end of L4.

C1
output from L7, if amplifier stage not used.
- Gimmick capacitor: two -inch lengths of is L6, L8
4503).
- Slug-tuned, 1.6 to 2.8 µH (Miller No.

L7, L9 - Three -turn link over cold ends of L6 and

1
insulated hookup wire, twisted 6 times. A 2-pF
fixed -value ceramic capacitor can be substi- L8. Use small -diameter insulated hookup wire.
tuted. L10 - 5 turns No. 24 enam. wire to occupy 3/8
C2 -
10-pF piston -type trimmer (Centralab inch on Miller 4500-4 slug -tuned form.
829-10). L11 -2 -turn link of small -diameter wire over cold
CR1 -
Silicon diode, 50 PRV or greater, at 200 end of L10.

J1
mA.
-
BNC-style chassis connector.
L12 -
5 turns No. 20 tinned copper wire (or
enam.), 5/16 -inch diameter, 3/8 inch long.
J2, J3 - Phono jack. L13 -
2 turns small-dia insulated hookup wire
L1 -6 turns No. 24 enam., wire to occupy 3/8
inch on slug-tuned form, 1/4 in. dia; (Miller L14
inserted in cold £4d of L12, 1/4-inch dia.
-9
turns No. 24 enam. wire, close wound on
4500-4) tap 1-1/4 turns above ground end. same style form as L10.
L2 - 4 turns No. 24 enam. wire to occupy 3/8
inch on same type form as L1.
Q1 -Q6, incl.
- -
For text reference purposes.
R1 500,000-ohm control, linear taper.
L-3 -5 turns No. 24 enam.to occupy 3/8 inch on
same type form as L1.
RFC1 --
50-µH rf choke (Millen J-300-50).
RFC2 22-µH rf choke (Millen J-300-22).
L4 -4 turns No. 24 enam. to occupy 3/8 inch on
same style form as L1.
Y1 -
58 -MHz third -overtone crystal (International
Crystal type F-605).
50 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
Ti Fig. 4-7 - Schematic of the con-
verter power supply. The 2000-1.1F
J4
limI1 CR111 capacitor is electrolytic, others
.aUTPUT are disk ceramic, 1000 -volt units.
1 2Y. The 56-ohm resistor was selected
,.005,£
15V. to give the proper power-supply
voltage when used with the circuit

1 OFF
CR1-CR4, incl.
1 ampere.
-Silicon rectifier, 50 PRV, of Fig. 4-6 (12 volts dc).

T1- 12 -volt, 1 -ampere filament transformer.

- 115 -volt ac neon lamp assembly.
-
11
115V. A.C. J4 Phono jack.
S1 - Spst toggle.

receiver S -meter. In this way the signal readings transistors may outperform one using anything but
will then be more useful than is often the case with the most expensive and hard -to -get vacuum tubes.
converter -receiver combinations where no i -f gain The 220 -MHz converter of Fig. 4-10 is almost a
control is included. The setting for various recei- duplicate of the 144-MHz model shown earlier in
vers may vary markedly, but it should be remem- this chapter. Its weak -signal sensitivity should be
bered that the position of this control has no better than has been possible heretofore at this
bearing on the ability of the system to respond to frequency, for anything of comparable simplicity
weak signals, if it is set high enough so that the and moderate cost. It was built by Tom McMullen,
noise output of the converter can be heard, or seen W1SL.
on the meter. To save space and avoid duplication, only those
portions of the converter that are different from
THE 220 -MHz MODEL the 144 -MHz version are discussed here. An iden-
tical circuit board is used. The circuit, Fig. 4-11, is
The superiority of transistors over tubes be- similar, but not identical to that of the 144 -MHz
comes more marked as the upper frequency limit converter. The same parts designations are used
of the tubes concerned is approached. Thus a insofar as possible. Self-supporting coils and cylin-
well -designed 220 -MHz converter using the better drical ceramic trimmers are used in the rf circuits.

. .
evoak
I

r Ir

Fig. 4-8 -Bottom of the circuit board. The i -f gain control and 12 -volt power jack are at the lower
right. The input circuit and rf stages are at the upper left. The mixer is at the upper center, and the IC
i -f amplifier is at the upper right. The oscillator chain extends along the lower portion of
the board.
The interstage shields are in place, but are difficult to see in this photo.
The 220 -MHz Model 51
Fig. 4-9 - Layout of the DRILL SIZES Q1
Qz Q3
G S D
etched circuit board. The lines A = 53 s
show where the key compon- 8 ='%32
ents are mounted and indicate C=Vá"
the way the semiconductor D=1/2'
leads are indexed. This is a E= No 33
bottom view of the board A//other PIN 1

(copper side). The white areas holes use

ARt
indicate the copper that re- No.43 drill
mains after etching. PIN 4
* Break and bridge with LI
100 ohms, 220 MHz only.
+12 VOLT
LN BUS
9 -VOLT R1
BUS
Y1

The first rf stage has capacitive neutralization.

Injection at 192 MHz (for 28 -MHz i -f) is provided
by a 48 -MHz crystal oscillator and a quadrupler.
Oscillator voltage is Zener-regulated at 9 volts.
Almost any silicon vhf transistor will work in
the oscillator and quadruples stages. The rf and
mixer are FETs. Judging from experience with the
preamplifiers described elsewhere in this chapter, i-f amplifier stage is not shown, as it is identical to
most vhf junction FETs should work well here. A that in the 144-MHz converter. It is definitely
noise figure of 3 dB or better should be obtainable recommended, not only to assure adequate gain for
with several different types, in addition to the some of the less -effective communications recei-
Motorola MPF series shown here. Some MPF types vers, but also to permit setting the desired con-
are now available only with 2N numbers. The verter output level to match the particular receiver
MPF-107 is now sold as the 2N5486. The 28 -MHz in use.

c
<4

«Pr
.

Fig. 4-10 - Interior of the 220-MHz FET converter. Minor differences from the 144 -MHz model, Fig.
4-6 are discussed in the text. The rf mixer and i -f amplifier circuits, left to right, occupy the upper half
of the circuit board. Board layout is similar to that of Fig. 4-9, except as described.
 52 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
R. F. AMP. R. F. AMP MIXER
MPF 107 MPF 107 MPF 102
1 220 Mc.
{
220 Mc, 220 Mc, 28 Mc.
rIE-
1.2 2
1.7

) TO I.f.AMP. I

JI

.330
C4 i

L_ too
OSC.
40237 15 48 Mc. 192 Mc.
c
1_12

48 CR1
Mt.= o+
12V.
5100

100

-+ 10o

-L
CR2

Fig. 4-11 -
Schematic diagram and parts information for the 220-MHz converter. Only those portions
wherein there are differences from the 144 -MHz circuit, Fig. 4-6, are shown. Parts are labelled similar
to those of the 144 -MHz converter, wherever possible. The MPF-107 is now sold as 2N5486.

L2 - 6-1/2 turns No. 22, 1/4 -inch dia, 1/2 inch

long. Tap at 2 turns from top end.
-
--3
C1 2 pF or 2 1 -inch lengths of insulated wire, L2A turns insulated wire between turns of L2.
twisted six times. L5 1 turn insulated wire between bottom turns
C2, C3, C6, C7 -
8-pF cylindrical ceramic trimmer of L4.
(Centralab 829-7).
- -
L6, L7 Same as in 144 -MHz model.
C4 9-pF subminiature trimmer (Johnson
189-503-41.
L10 -7 turns No. 22 5/16 inch long, on 1/4-Inch
iron -slug form (Miller 4500-4 form).
CR1 - Silicon diode, 50 PRV or greater, 200 mA L11 -2 turns insulated wire over bottom turns of
or more. L10.
CR2 -9 -volt Zener diode. L13 -
1 turn
J1 - BNC coaxial fitting.
L1, L3, L4, L12 -
turns of L12.
insulated wire between first two
3 turns No. 22, 1/4 -inch dia,
1/4 inch long. Tap L1 at one turn from ground
R2 - See text.-
RFC1, RFC2 25-µH rf choke (Millen J-300-25).
end. Y1 -
48 -MHz third -overtone crystal.

tor, C4, appears about where L14 was. The

One difference between this converter and the
cylindrical trimmers, C3, C5, C6, and C7, are
one for 144 MHz might not be readily apparent,
mounted where the slug coils are seen in the
but it is important. Note the resistor, R2, in the
144 -MHz model. Note the mounting positions of
line to the mixer drain circuit. This is not in the
the rf coils. LI, L3, and L4 are similar: their axes
2 -meter version. It was put into the 220 -MHz
model when a signal -frequency resonance devel-
parallel to the chassis. L2 is perpendicular to it.
oped in the circuit board, causing an oscillation Adjustment
problem that took some chasing down! Looking at The first step should be to get the oscillator and
the layout drawing of the circuit board, Fig. 4-9, multiplier running. It may be advisible to keep
pick out the 12 -volt bus running from near the voltage off the stages other than the ones being
middle of the board horizontally to the right, checked, at this point. Make sure that the oscillator
before dropping vertically into the lbwer half. This is on 48 MHz, and no other frequency. (In this
should be severed below the letter "A" on the type of circuit It is possible to get oscillation on
sketch. Tie 100-ohm R2 is bridged across the gap. the crystal fundamental, in this case 16 MHz, if the
Other minor mechanical differences resulting collector circuit does not resonate at 48 MHz.)
from the slightly modified circuitry in the rf Now fire up the quadrupler and peak C2 for
portion are apparent from the photographs. The maximum energy at 192 MHz.
small shield between Ll and L14 in the 2 -meter With the converter connected to the receiver,
model is not needed here. The neutralizing capaci- there should be a marked increase in noise when
FET Preamplifiers for 50, 144, and 220 MHz 53
voltage is applied to the rf, mixer, and i -f amplifier circuits affect only the gain and frequency re-
stages. The i -f can be peaked for maximum noise at sponse characteristics, so they can be adjusted for
28 MHz. It is helpful at this point to have a signal flat response across the desired frequency range,
on 220. A dipper signal will do. It is also desirable and there will be no sacrifice in the ability of the
to have a properly matched antenna connected to system to respond to weak signals.
J1, unless a good signal generator with 50 -ohm Most realistic operation of the receiver's S -
termination is available for alignment purposes. If a meter will be obtained if the meter adjustment is
random antenna must be used, put a 50 -ohm set so that there is an appreciable reading on noise
resistor across J1 to simulate the eventual load, for only, with no signal. The converter i -f gain control
neutralization purposes. is then set so that the meter reads S-0 or S-1, with
There may be no oscillation in the rf stages, the antenna on. In this way the relative strength of
regardless of tuning, if the converter is operated signals will be indicated on the meter, within the
with a proper load. If this is the case it is merely usual variations encountered with these none -too -
necessary to adjust the neutralizing capacitor, C4, reliable devices. The receiver's antenna trimmer, if
and the tuning of the input circuit, L1C3, for best there is one, can also be used as an auxilliary gain
signal-to-noise ratio on a weak signal. All other control, and it will have no effect whatever on the
sensitivity of the system.

FET PREAMPLIFIERS FOR 50, 144, AND 220 MHz

r,

Fig. 4-12 - Transistor preamplifiers for 50, 144, and 220 MHz, left to right. Appearance
except for the type of tuned circuit used.
is similar,

Where a vhf receiver lacks gain, or has a poor Many inexpensive transistors will work well in
noise figure, an external preamplifier can improve these amplifiers. Motorola MPF-102, 104, and
its ability to detect weak signals. Some multiband 106,* all low-priced molded -plastic units and the
receivers that include the 50-MHz band are not as more expensive metal -case 2N4416 were tried, and
good as they might be on 6. Converters for 144 all were more than adequate. The MPF-102 is the
MHz haing pentode rf stages, or using some of the least expensive, and surprisingly, it was as good as
earlier dual triodes, may also need some help. any, even on 220 MHz. Careful readjustment is
Some fm transceivers currently in use were built required when changing transistors, so the builder
before low -noise amplifiers were readily available. should not jump to conclusions about the relative
Most 220-MHz converters are marginal performers, merit of different types.
at best. The field-effect transistor preamplifiers of
Fig. 4-12 should improve results with these, and Construction
with any other receivers for these bands that may
The amplifiers were built in small handmade
not be in optimum working condition.
The circuits of the amplifiers are similar, boxes, aluminum for the 50- and 144 -MHz models,
though iron-core coils are used in the 50 -MHz and flashing copper for the 220 -MHz one, but any
model, and air -wound coils in the other two. The small metal box should do. Those shown are
1-1/2 X 2 X 3 inches in size. The transistor socket
grounded -source circuit requires neutralization.
This is done with capacitive feedback, rather than is in the middle of the top surface, and the BNC
with the inductive circuit commonly used. A input and output fittings are centered on the ends.
tapped input circuit is used in the 50 -MHz ampli- The tuned circuits are roughly 3/4 inch either side
fier, and capacitive input is shown for the other of the transistor socket, but this should be adjusted
two, though this was done mainly to show alter- for good layout with the parts available. Flat
native circuits. The output circuit is matched to * Only the MPF-102 is still available. Others
now have 2N numbers. The MPF-107 is now
the receiver input by means of C2. 2N5486.
54 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
ceramic trimmers are used for tuning the 144 -MHz on the low -frequency side of the gain peak. In
amplifier, and the cylindrical type in the 220 -MHz listening to a weak modulated signal, the fact that
one. Sockets were used mainly to permit trying the noise drops off faster than the signal with a
various transistors; they could be wired directly in slight detuning is quite obvious. Typically 'the
place equally well. Printed-circuit construction meter reading may drop about one full S -unit,
would be fine, if you like this method. while the noise level drops two S -units. The exact
Adjustment setting depends on the neutralization, and on the
loading, both input and output, and can only be
The preamplifier should be connected to the
determined by experiment, with a noise generator
receiver or converter with which it is to be used, or a weak signal.
with any length of coaxial cable, or by hooking J2
directly to the converter input jack with a suitable Results
adapter. If you have a noise generator or signal Because external noise is more of a limiting
generator, connect it to J1. If not, use a test signal factor in 50 -MHz reception than on the higher
from a grid-dip oscillator, or some other signal bands, tuning for best reception is not critical on
source known to be in the band for which the this band. Very likely you can set the neutraliza-
amplifier was designed. Preferably a matched an- tion to prevent oscillation, peak the input and
tenna for the band in question should be hooked output circuits roughly, and you'll be all set. On
to J1, if a signal generator is not used. A 50 -ohm 144 the job is fussier if the amplifier is to effect a
resistor across J1 may be helpful if a random real improvement, particularly if your receiver is a
antenna is used for the adjustment work. fairly good one. This preamplifier should get you
Set the neutralizing capacitor near half capaci- down to the point where external noise limits your
tance; then, with no voltage yet applied, tune the reception, for sure, if you were not there before.
input and output circuits roughly for maximum On 220 the preamp is almost certain to help, unless
signal. (The level may be only slightly lower than it you already have an exceptional receiving setup,
would be with the converter or receiver alone.) and optimum performance is worth the trouble
Now apply voltage, and check current drain. It you take to get it. With all three, you should be
should be 4 to 7 mA, depending on the voltage. certain that, if a given signal can be heard in your
Probably there will be an increase in noise and location, on your antenna, you will now be able to
signal when voltage is turned on. If not, the stage hear it.
may be oscillating. This will be evident from erratic Warning: if the preamp is to be used with a
tuning and bursts of noise when adjustments are transceiver, be sure to connect it in the line to the
attempted. receiver only, not in the main line from the
If there is oscillation (and it is likely) move Cl transceiver to the antenna. It is best to do this
in small increments, retuning the input and output before any work is done on the amplifier; other-
circuits each time, until a setting of Cl is found wise you're sure to throw the send -receive switch
where oscillation ceases, and the signal is amplified. inadvertently and finish off the transistor.
All adjustments interlock, so this is a see -saw If you're in doubt about the possibility of rf
procedure at first. Increasing the capacitance of C2 coming down the antenna line, connect protective
tends to stabilize the amplifier through increased diodes across the input, as shown with CR1 and
loading, but if carried too far will have an adverse CR2 in one of the circuits. Install these after the
effect on gain. The best setting is one where the preamplifier tuneup, and check weak -signal recep-
input and output circuits do not tune too criti- tion with and without them, to be sure that they
cally, but the gain is adequate. are not causing signal loss. Junction-type field-
The input circuit is first peaked for maximum effect transistors are capable of withstanding much
signal, but final adjustment should be for best more rf voltage than bipolar transistors, so this
signal-to-noise ratio. This process is very similar to kind of protection may not be needed in situations
that with tube amplifiers, and the best point will where it would have been mandatory with earlier
probably be found with the input circuit detuned types of transistor front ends.

17)
N!
+
t
Z
r '

Fig. 4-13 - Interiors of the FET preamplifiers, in the same order as in Fig. 4-12. The input end
toward,the right in each unit.
is
All-FET Converters for 50 and 144 MHz 55

Fig. 4-14 - Circuit diagrams and

parts information for the FET
preamplifiers. Values of capaci-
tors not described are in pico -
farads (pF or µµF).

JI

Cl -1.3- to 6.7-pF subminiature variable (John- J1, J2 - Coaxial fiiting. BNC type shown.
son 189-502-5). L1 - 50 MHz; turns No. 24 enamel on 1/4 -inch
7
C2
C3
--3- to 30-pF miniature mica trimmer.
0.001 -/IF feedthrough (Centrelab MFT-1000;
from
iron -slug ceramic form, tapped at 3 turns
ground end (Form is Miller 4500). 144 MHz: 3
FT -1000 in 220 -MHz amplifier). turns No. 22, 1/4 -inch dia, 3/8 inch long. 220
C4, C5 - 3- to 12-pF ceramic trimmer in 144 -MHz MHz: same, but with 2 turns 1/8 inch long.
amplifier;1- to 6-pF cylindrical ceramic in 220. L2 -50 MHz: 10 turns like L1, but center -tapped.
C6- 0.001-1/F 50 -volt mylar. Omitted in 220-MHz 144 MHz: 5 turns No. 22, 1/4 -inch dia, 1/2
model. inch long, center -tapped. 220 MHz: Same but 4
CR1, CR2 - 1N34A or similar germanium diode. turns.

were 6-pF miniature variables like C1 in the other

432 -MHz Version units. The coils were No. 20 tinned 3/16 -inch
Results with these preamplifiers were so gratify- diameter, 1-3/4 turns in Ll and 3 turns, center -
ing that a 432 -MHz model was tried. This was quite tapped for L2.
similar in layout, except that the metal case After some considerable juggling of adjust-
2N4416 was used, and it was wired directly in ments, this stage was stabilized, and then tested
place instead of using a socket. The transistor was with a poor crystal-mixer converter that serves as a
suspended in a small notch in the bottom edge of a trial horse in such work. The preamplifier gave
shield, which was mounted across the middle of about 10 dB gain, and this was all improvement in
the assembly. The case and source leads were the converter noise figure. But it was not enough;
soldered to the shield, with the gate lead projecting this test setup requires about 18 dB gain for
into the front compartment and the drain lead into optimum performance and complete over-riding of
the rear. The trimmers, C4 and C5, were 0.5 to 3 the mixer and i-f noise. For setups needing only a
pF and the input and output loading capacitors few dB gain, such a preamp should do very well.

ALL-FET CONVERTER S FOR 50 AND 144 MHz

The converters of Figs. 4-15 through 4-19 were
designed to be part of a complete receiving setup
for all amateur bands from 1.8 to 148 MHz.3 They
are included here to show matching converters for
50 and 144 MHz, of relatively simple design and
construction. Because the drawings are those used
in the complete Handbook project,3 the compo-
nent numbering does not start at 1 for each type of
part, in this condensed version. The sensitivity
resulting when these converters are used with any
28-MHz receiver of good performance should be
more than adequate, and the overall noise figure
should be around 2.5 dB. Their MOSFET mixers
provide high conversion gain, and resistance to
overloading and cross-modulation effects. Rejec-
tion of out-of-band signals is improved by the use _ Fig. 4-15 - The twin vhf converters are housed
in a homemade aluminum box which has
of double -tuned coupling circuits in several places removable top and bottom covers for easy
in each converter. access to the circuit boards. Each converter has
All essential circuit features and adjustment its own input and output jacks so that simul-
procedure have been discussed at length in con- taneous operation is possible.
 56 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
Fig. 4-16 -
The darker of the two circuit boards
is the 6-meter converter, shown at the bottom of
*7 the photo in this inside view. Since the circuit
boards are identical in pattern, some of the holes

'>
.
411.
1" 1
R
are left blank on the 6 -meter model, as there is
one less stage in its oscillator section. The protec-
tive diodes at the antenna jacks were not installed
E
(1. 1'
.
when these photos were taken.
E
s :OR 4

:1
. (1 CLW
! -- --_
hi
-,

CFm
c -' 1
. The converters are built on identical circuit
cl :
boards, though different board materials were used
115 here to make the separate units stand out pTioto-
719
0 e.
graphically. Full-scale templates for laying out the
boards are available from ARRL.*
The input circuits have protective diodes. Coup-
ling into the first tuned circuit is inductive, in the
50-MHz converter, as adjustment of coupling for
nection with earlier projects in this book, and in
lowest noise figure is not required at this frequen-
the preceeding chapter, so they will not be given in
* Converters originally described in October,
full here. Probably the only "different" circuit in 1969, QST, p. 37. Mention in ordering templates
these two converters is the capacitive divider used and send 25 cents and stamped self-addressed
to couple the i -f output (at 28 to 30 MHz) into the envelope. Circuit boards made for ARRL construc-
tion projects are available from individual sup-
associated communications receiver. pliers, listed from time to time in QST.
R F AMP L7 -
28-30 MNt
50-52 MNf

L .II 50-52
r1
MHz

CIB
y'evT L(f-" S ool
G2 D
_
C30

D I

.0 LS

AN

I -F
OUTPUT
CR5 CR4

T001
0}12V
C38
1005
S M!SLVER MICA_

C54
OSC

R19
BOTTOM VIER
22MHiI
IDOR
G2 GI
000
G S D

04,06
8.2 pH
EXCEPT AS INDICATED, DECIMAL VALUES OF '001 05
CAPACITANCE ARE IN MICROFARADS (pf.); C36 C37
OTHERS ARE IN PICOFAR A DS( pf.ORput); fIV I001
RESISTANCES ARE IN OHMS; K . 1000. 3

m
Fig. 4-17 -
Circuit diagram of the 6-meter converter. Resistors are 1/2 -watt composition.
Capacitors are disk ceramic unless specified differently. Numbered components not appearing in
the parts list were so identified for circuit-board layout purposes.
CR3 -
9.1 -volt, 1 -watt Zener diode (Motorola HEP-
104 or equiv.). close -wound, on J. W. Miller 4500-2 iron -slug
CR4, CR5 -
Small signal silicon switching diodes form.
J2
(1N914).
--BNC or SO -239 -type chassis connector.
04, 06 -
Junction FET, Motorola MPF-102
(HEP-802 or 2N4416 suitable).
J3 Phono connector. 05 -Dual -gate MOSFET, Motorola MFE3008
L5 -3 turns of small insulated wire wound over
the ground end of L6.
(RCA 3N141 also suitable. Gate -protected ver-
sion: RCA 40673).
L6, L8, L9 -
10 turns No. 24 enam. wire, close - RFC3 -
8.2-µH miniature rf choke (James Millen
wound, on J. W. Miller 4500-4 iron -slug form, 34300-8.2).
L7 -
25 turns No. 30 enam. on 4500-2 form.
L10 -L13, incl. -
12 turns No. 24 enam.,
Y1 -3rd -overtone crystal (International Crystal
Co. type EX).
 All-FET Converters for 50 and 144 MHz 57
R F AMP MIXER
L22 28-30 MHz
C41
144-146 MHz .001 NEUT I 144-148MHz C50
SM 680 G2
L14 C40 SM 5
D 08
39 07 C461\008 L17 LI8 Y
J4 LIS L,6
27
SM R2 IB
JS
ANT 42
C49, I 1f
.001 C4 C45 C53 1 OUTPUT
270 48
R26 ISO
005
CR6 .00S R2'
C47 270
C5I
_L001 5
R23
270
C
4124
C54
GNO
S M SILVER MICA .005
--r-

OSC DOUBLER
SBMHz 116 MHz

09 ^y
L19 I

010 _L20 L21

100,.
T2 R28 BOTTOM VIEW
58 MHz 1008 I
R29 I C6
SM C59
G2

G S D
RFC4
0 C60 D
001 07, 09,010

.I
EXCEPT AS INDICATED, DECIMAL C., R30 8.201 08
VALUES OF CAPACITANCE ARE
.001 .001
IN (9E1; OTHERS
MICROFARAOS
C58
ARE IN PICOFARADS IpF OR yyF); V
RESISTANCES ARE IN OHMS; CRB
1000.

Fig. 4-18 -
Circuit of the 2 -meter converter. Resistors are 1/2 -watt composition. Capacitors, unless
otherwise noted, are disk ceramic. See Fig. 4-17 and text for explanation of component numbering.
CR6, CR7 -
1 N914 or equivalent. L22 -9turns No. 30 enam., close -wound, on J. W.
CR8 -
9.1 -volt, 1 -watt Zener diode (Motorola
HEP-104 or equiv.).
Miller 4500-2 iron -slug form (J. W. Miller Co.,
19070 Reyes Ave., Compton, CA 90221; write
J4 - BNC or SO -239 -type chassis connector. for catalog and prices).
-
J5
L14
- -4
Phono connector.
turns No. 24 enam. to occupy 3/8 inch on
07, 09, 010 Junction FET, Motorola MPF-102
(2N4416 suitable).
J. W. Miller 4500-4 iron -slug form. Tap 1 turn Q8 -Dual -gate MOSFET, Motorola MFE3008
from ground end. (RCA 3N141 also suitable. Gate -protected
L15, L16, L19 -
5 turns No. 24 enam. to occupy version: RCA 40673).
-
8.2 -pH miniature rf choke (James Millen
3/8 inch on same -type Miller form as L14. RFC4
L17, L18 -
15 turns No. 24 enam. wire, close --
Y2
34300-8.2).
-
58-MHz 3rd -overtone crystal (International
wound, on J. W. Miller 4500-2 iron -slug form.
-
L20, L21 Same as L14, but no tap. Crystal Co. type EX).

cy. The point at which the antenna line is tapped stagger-tuned for uniform response across the
on the input coil in the 144 -MHz converter can be desired frequency range. This may drop the overall
adjusted for lowest noise figure, with the aid of a gain somewhat, but it should have no effect on the
noise generator or by listening to a weak test signal noise figure, if the first stage is working properly.
and making comparisons of the margin of signal
over noise.
Construction
The injection stages are similar, except for the Scale templates for the etched-circuit board are
addition of a doubler in the 144 -MHz model. If available from ARRL.* The semiconductors are
any difficulty arises from unwanted harmonics of available from most of the larger mail-order houses,
the oscillator or injection frequencies, a suitable or from any Motorola distributor. See October,
trap can be connected in the lead to Gate 2 of the 1970, QST, p. 14, for ordering restrictions. The
mixer. This can be a parallel or a series circuit. slug-tuned coil forms are made by J. W. Miller and
Examples of each will be found in Chapter 3. The should be only those numbers specified. It will be
double -tuned circuits in the injection stages should noted that some cod -form numbers have a numeral
be tuned carefully for maximum desired -frequency 2 at the end (4500-2) while others have a 4 at the
output, in the process of adjusting any trap circuit end of the number (4500-4). These numbers relate
for rejection of spurious signals. to the core material used, which is designed for a
Only the first -stage tuning and the neutraliza- particular frequency of operation. The core mater-
tion are important in obtaining best noise figure. ial has a significant effect on the tuning range of
Other circuits in the rf and mixer stages can be the inductors, and can seriously affect the coil Q if
58 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
Fig. 4-19 - Looking into the bottom of the
converter box, the 6 -meter unit is at the top of
A the photo. Each converter has four 1 -inch
standoff posts which secure the circuit boards
to the bottom plate of the cabinet. Rf shields
of flashing copper are soldered to the ground
foil on the circuit boards. They are notched out
wherever they come in close proximity to the
non -ground elements of the circuit.

of the wrong type. If substitute coil forms are 3008s to minimize the possibility of transistor
used, be sure that they're designed for the frequen- damage when soldering. The sockets are Elco
cy range over which they will be used. 05-3308. The binding posts used for connecting
These converters can be packaged in any style the +12 volts to the converters are E. F. Johnson
of box the builder prefers. In this instance, both 111-102s.
units are housed in a single homemade enclosure HEP-56 (Motorola) rectifier diodes are con-
which measures 6-3/4 X 5 X 2-1/2 inches. The top nected from the 12 -volt input terminals on the box
and bottom covers are held in place by means of to the 12 -volt terminals on the circuit boards, their
No. 6 spade bolts which are attached to the side anodes toward the Johnson binding posts. These
walls of the box. This style of construction can be diodes prevent damage to transistors should the
handled with ordinary hand tools, and only four operator mistakenly connect the power supply
90 -degree bends are required. This box was made leads for the wrong polarity. Positive voltage will
from a large aluminum cookie sheet purchased at a pass through the diodes, but negative voltage will
hardware store. The dull finish results from a be opposed.
lye -bath treatment given the aluminum after it was It is strongly recommended that the converters
formed. be housed in some type of metal enclosure, as was
The converters are mounted on the bottom done here, to prevent oscillator radiation, and to
plate of the box by means of 1 -inch metal standoff insure against random pickup of interfering com-
posts. Self-adhesive rubber feet are attached to the mercial signals by the mixer circuit. This pre-
bottom of the box. Black decals are used to caution is especially important in areas where
identify the terminals on the outside of the box. commercial fm and TV transmitters are nearby.
A 4 -terminal transistor socket is used for the Adjustment procedure for these converters is
6 -meter mixer MOSFET. At the time the 2 -meter essentially the same as for units described earlier in
converter was built a socket was not on hand, but this chapter, and given in general terms in Chapter
both converters should use sockets for the MFE- 3.

A LOW -NOISE 432 -MHz CONVERTER

There is usually great emphasis on the need for

a good low -noise uhf transistor in at least the first
stage of a 432 -MHz converter, but the rf stage or
stages cannot do the whole job. Good front-end
design is often hampered by deficiencies elsewhere
in the converter. These can include inadequate
injection to the mixer, and poor noise figure in the
stage immediately following the mixer. The conver-
ter of Figs. 4-20 through 27 does well with a single

Fig. 4-20 -
The 432 -MHz converter is built on the
cover plate of a standard aluminum chassis. The
"cigar-box effect" results from application of
contact paper with simulated wood -grain finish to
all exposed surfaces.
A Low -Noise 432 -MHz Converter
Fig. 4-21 - Interior of the 432 -MHz low -noise
converter, showing the rf and -f amplifier
i

assemblies. Strip-lines in the rf assembly are, left to

right, L1, L3, and L4. The wire loop coupled to L3
is the neutralizing device, L2, with a capacitor tab
coupling to L1. The small mixer diode is just
visible, close to the right side of the rf assembly.
The 28 -MHz -f amplifier is the square circuit board
i

at the lower right.

rf amplifier stage, because the mixer has plenty of

injection, and it is followed by a low -noise i -f
amplifier. The latter is especially important when
the intermediate frequency is in the 28 -MHz range,
as some communications receivers have rather poor
noise figure and gain in this region. 1-7/8 inches high and 8-1/2 inches long, supports
the oscillator -multiplier assembly. -The injection
Circuit circuit board is mounted with 3-48 screws and
The rf amplifier (see Fig. 4-23) uses a 2N5032 short metal spacers, to provide lead clearance.
uhf bipolar transistor, with grounded emitter and Output from the tuned -line filter, C14L17, is taken
neutralization. This gives more gain than the off through L18 and small coax to the mixer.
commonly used grounded -base circuit, so only one Small screws and spacers are also used to mount
rf stage is needed to override mixer noise and the i -f amplifier to the cover plate. A small pc -type
establish the noise figure of the system. Tuned control, R2, reached through a hole in the side of
lines are used in the input, collector, and mixer - the chassis, is used for gain adjustment. If the
input circuits, for some selectivity and rejection of converter is to be used with several receivers, a
unwanted frequencies. The segment of the band shaft -type control could be substituted, and
near 432 MHz used for weak -signal work is so mounted on the cover or chassis to facilitate
narrow that rf selectivity presents no problems. readjustment as needed. Lead lengths to R2 are not
The amplifier collector current is adjusted for critical.
optimum signal-to-noise ratio by varying the base The rf amplifier enclosure and tuned lines are
bias, by means of Rl. The bias source is Zener made of flashing copper. Dimensions are shown in
regulated by CR1. Fig. 4-25. All holes should be drilled before cutting
The mixer is a Schottky -barrier diode, CR2. It the metal and bending it to shape. Corners are
works into a 28 -MHz i -f amplifier stage using an soldered for rigidity and rf shielding. The lines, Ll,
integrated circuit. The amplifier may not be L3, and L4, should be laid out and drilled before
needed with communications receivers that work cutting the strips apart, as it is difficult to drill a
well in the 10 -meter range, but it adds little to the clean hole in a narrow strip of flashing copper. The
cost and complexity of the converter, and its gain ends of the strips are bent up and soldered to the
control, R2, is handy for setting the converter gain inside of the box.
to the optimum level for any receiver. A shield partition isolates the input line from
The injection system (Fig. 4-24) uses more the collector line, and also supports the transistor,
transistors and components than most converters Ql, which is mounted in a small hole near one end
have for this purpose, but the string of doublers is of the shield. The emitter lead is connected to the
easy and inexpensive to build, and it produces input side of the shield and the case lead to the
injection voltage to spare. Use of a starting collector side. Both leads are grounded by a 2-56
frequency of 50.5 MHz makes checking the oscil- screw and nut, with no soldering needed. The
lator a simple matter if the builder has a 50 -MHz entire assembly (the shield, Ql, R1, C2, and L2)
receiver. A tuned -line filter, loosely coupled to the can be completed before it is fastened in place with
last doubler, helps to suppress unwanted products small sheetmetal screws. Note the "sense" of the
of the multiplier system. neutralizing loop, L2. The grounded end is toward
the transistor, and the hot end goes through a hole
Construction in the shield, toward the grounded end of the
The converter is built in three principal sub- tuned lines. Cover the wire with insulating sleeving
assemblies, mounted on an aluminum plate that where it passes through the hole.
fits the top of a 5 X 9-1/2 X 2 -inch chassis, thereby
providing complete shielding. An L-shaped bracket

Fig. 4-22 -
The oscillator-multiplier assembly is on
a circuit board mounted on a full-length aluminum ,
bracket that provides both support and shielding.
The crystal oscillator is at the right end. The lie, ' -
!
strip -line at the left end is L17, the tuned circuit of
a 404 -MHz filter in the output of the last doubler
stage.
60 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS

.01 01

28MHz T Ul . -
28MHz
I -F
OuTPUT
JI L8

INPUT

Fig. 4-23 - Schematic diagram and parts information for the rf and i -f portions of the 432 -MHz
converter.
L2 -No. 24 enamel 1-11/16 inches long, bent as
Cl - 7-pF cylindrical trimmer (Centralab 829-7). shown in Figs. 4-21 and 25.
C2 -
Copper tab about 1/8 inch
1/4 inch square, L3, L4 - Copper strip 3/8 x 2-9/16 inches long
over L1, 1-5/8 inches from ground end. Adjust after bending. Tap L4 3/4 inch from ground
spacing for neutralization. end.
C3, C5
829-10).
-
10-pF cylindrical trimmer (Centralab L5, L7 - 20 turns No. 30 enamel on .162 -inch
slug -tuned form (Miller 27A013-7). Tap L5 10
C4 -
5-pF piston trimmer. turns from ground end.
C6, C7, C8 -
.001 -/IF feedthrough (Erie L6, L8 - 3 turns small enameled wire over cold
662-003-102K). C7 and a lug on the tie -strip serve ends of L5 and L7.
as test points, and should be joined by a jumper
when no meter is connected.
Q1 - Low -noise uhf npn transistor, Motorola
2N5032.
CR1 -6.2 -volt Zener diode. - Allied).
25K control, pc -type mounting (CTS R253B
CR2 - Hot -carrier diode (Hewlett-Packard 2811).
R1
-
CR3 - 1-A, 100-PRV diode (Int. Rect. 5A1). R2 - 100K control, pc -type mounting (CTS
J1, J2 - BNC coaxial jack. R104B - Allied).
J3 -2-pin polarized power connector. RFC1 - rf choke (Miller 4584).
.22-11H
L1 - Copper strip 3/8 X 2-1/4 inches after RFC2 - 6 turns No. 26 enamel 1/8 -inch dia, 3/8
bending. Tap at 7/8 and 1-3/4 inches from inch long.
ground end. U1 - Motorola MC1550G integrated circuit.

The coupling capacitor, C4, is supported by its oscillator should be checked to be sure that it is
own leads, between L3 and L4. The mixer diode, working only on the marked frequency. Improper
CR2, is tapped on L4, through a hole in the side of adjustment of the coil L9 may allow oscillation on
the enclosure, and is supported on a small tie-strip, the crystal fundamental, about 16.83 MHz, and
outside the assembly. The 3-pF injection-coupling output at 50.5 MHz will be much too low to drive
capacitor, one end of RFC2, and the end of the the following stage properly. There should be no
small coax from the injection board are also evidence of rf in the 16 -MHz region, or on twice
supported by this strip. the fundamental frequency. Be sure that the core
tunes L9 through the crystal frequency.
Adjustment Peak each doubler stage for maximum output
Alignment of the oscillator -multiplier chain at the desired frequency. If no indicator for 404
requires a dip meter or calibrated absorption MHz is available, a simple Lecher-Wire setup (see
wavemeter. To protect the amplifier transistor, Ql, Chapter 14) can be used, with a diode and
do not apply voltage to it, or the bias -adjusting milliammeter, to be sure that this circuit peaks
network, until all other stages are checked out. The properly. Injection on the proper frequency should .
 OSC. DB LR DB LR ()SLR

50.5 MHz
101 MHz . -1 ° 202 MHz L13

04 TO MIXER
505 MHz

16800

10

4-7

EXCEPT AS INDICATED, DECIMAL VALUES OF

CAPACITANCE ARE IN MICROFARADS ( pF l ;
OTHERS ARE IN PICOFAR4OS 1 pF OR ppri;
RESISTANCES ARE IN OHMS;
I, IODO, M1000 000,

Fig. 4-24 - Schematic diagram and parts information for the oscillator -multiplier stages of the 432 -MHz converter.

C10 35-pF ceramic trimmer.

- L12, L14 turns No. 22, 3/16 -inch dia, 3/8 inch
-4 long. Mount on small tie -strip at upper left
C11, C12, C14 5-pF miniature trimmer (Johnson
- long. corner, as seen in Fig. 4-22.
189-5048). L13 -4turns No. 22, 1/4 -inch dia, 3/8 inch long. Q2 -Q5, incl. - Vhf npn transistor (Motorola
C13 - 1/4x 3/8-inch copper tab. See Fig. 4-25 L15 5 turns No. 28 enamel, on 1/2 -watt resistor.
- MPS3563).
and text L16 -4turns No. 22, 1/2 -inch dia, 1/2 inch long. RFC3, RFC4, RFC5, RFC7, RFC8 - Small rf
CR4 -9.1 -volt Zener diode. L17 Copper strip, 5/16 x 2-1/2 inches. Bend at
- chokes, values as indicated on drawing.
L9 - 6 turns No. 26 enamel on .162-inch ground end, so that when run through the RFC6 2 turns No. 26 enamel through ferrite
-
slug -tuned form (Miller 27A013-7). circuit board it will lie parallel to it, at the bead.
L10 -6 turns No. 22, 3/16 -inch dia, 1/2 inch long. height of the stator of C14. See Fig. 4-22. Y1 - Third -overtone crystal, 50.5 MHz (Inter-
L11 - 7 turns No. 24 enamel, 3/16 -inch dia, L18 Hairpin loop, No. 24 enamel. Portion
- national Crystal Type EX).
closewound. coupling to L17 is 3/8 inch wide and 5/8 inch

Q1
 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
bias-adjusting control, Rl, should be set at maxi-
mum resistance, and a 10-mA meter connected in
place of the jumper across the test points in the
lead to RFC1. Be sure that solid connections are
made, as accidental application of base bias, with
no collector voltage, can ruin the transistor. A safer
arrangement might be to use a lower range meter,
with a suitable permanent shunt connected be-
tween the test points. The meter can then be
removed from the circuit safely.
The collector current should be set to around 2
mA by adjustment of R1. Tune Cl, C3, C4, and CS
HOLE SIZES for maximum response to the test signal, watching
A=fie" the collector current for any sign of change. Such
B= %4"
C = 3/K fluctuation, or bursts of noise or rough -sounding
D = Yo" notes in the receiver, indicates oscillation in the rf
E=3/31
stage. The positions of the coupling loop, L2, and
E Holes
positioned to the capacitor tab, C2, should be adjusted to stop
match D holes oscillation, if any is encountered.
/n box After the converter has been stabilized, adjust
the collector current for best signal-to-noise ratio,
24"_
using a weak test signal. The data sheet for the

,z
o 2N5032 indicates that best noise figure should be
/Beard up obtained with about 1 mA. Other good uhf
6 3 sides transistors may require different values of collector
E5/y current. Usually it will be found that highest gain
will occur with somewhat more collector current
than that giving the best signal-to-noise ratio, so it
is important to make all tuning adjustments and
current setting for the latter quality. Gain can be
made up in the i -f stage.
An additional check can be made to be sure
L2 & C2 that the oscillator -multiplier string is giving the
C13 ON INJECTION
No.24 ENAMEL&
BOARD -COPPER
mixer adequate injection voltage. If it is, and if the
'k"SQ. COPPER TAB rf stage is working properly, the injection can be
reduced enough to cause the converter gain to drop
Fig. 4-25 - Details of the copper parts of the rf several dB, before the signal-to-noise ratio is
amplifier and mixer assembly. It is suggested that adversely affected. If adequate gain is available
layout work and drilling be done before the sheet elsewhere in the converter,
of flashing copper is cut, as drilling of small pieces the best overall perfor-
of thin metal is difficult. mance will generally be obtained with the injection
level toward the lowest that will give good signal-
to-noise ratio.

also cause a noise increase in the communications

receiver, at 28 MHz. The copper tab, C13, provides
loose coupling to the stator of C12. Adjustment is
not critical, and a spacing of 1/8 inch should be
about right. There is some coupling between the
tuned line and the chassis, causing detuning and
loss of injection, when the converter is assembled
in its case. The circuit can be repeaked through a
small hole drilled in the bottom of the chassis.
Moving the i -f gain control through its range
should cause a smooth change in noise level at 28
MHz. LS and L7 can be peaked roughly on noise,
or on a 28 -MHz signal.
The mixer tuning should result in a small noise
increase as CS is peaked. This adjustment can be
made by using a harmonic of the dip oscillator, or
the third harmonic of a 2 -meter transmitter, for a
test signal.
The rf amplifier can now be adjusted. Before
voltage is applied to this stage, a 50 -ohm resistor
should be connected across 11, or a signal genera-
tor with a built-in termination should be used. The
Fig. 4-26 - Half -scale drawing of the injection
board, foil side.
A Blanker for Pulse -Type Interference 63
C C
Q2 Q3 C Q4 Q5 COPPER STRAP
B Bó B B L17
o
E
E C14
L9 C11 LI6 L15
10PF 1 27pF L1, 4.7j4H L12 L13

1_
4-:1, yqr
.l.
IN

.r `: _o. 171
Y1 ¡_ ',

1,
"
LWW
1
10011

12fcN
!ellriY
``_
fy. '4.i I

¡L11.01
o

ia
1

1'11.5rH1`
+

12
Cr.r? 1

Ó5ruF 12V
o

470í1 // 10 001
22K11
9 VOLT
4.7/4/-/
47011 .0f 10011 L14
RfC6
ZENER
Fig. 4-27 - Half -scale drawing

A BLANKER FOR PULSE -TYPE INTERFERENCE

Our frequencies from 220 MHz up are blessed eliminated before they reach the high -gain stages.
with a minimum of static, but cursed with a One advantage of this approach is that no modi-
maximum of pulse -type noise, thanks to the shared fication of existing equipment is required. The
nature of our assignments above 200 MHz. The noise blanker is connected in the cable between the
"Government Radio Positioning Service" is in converter and receiver, and requires no other
there, too. This means radar interference, and radar connections to either.
looks like about the worst interference you can
have. The situation could be worse, however, for Circuit Description
elimination of static has yet to be accomplished The blanker uses two 6AG5s as amplifiers at
while short pulses of the radar variety can be the converter output frequency, in this instance 28
blanked out very effectively with suitable equip- to 30 MHz or 14 to 18 MHz. It could be any other
ment. converter i -f, with suitable modification of the
Radar pulses, ignition, and most other noise tuned circuits. The 6AGSs were used mainly
encountered in the world above 200 MHz have fast because of their ready ayailability; other pentodes
rise time and thus can be dealt with if eliminated such as the 6BH6, 6BA6, 6AKS, and the like
from the receiver before they get to circuits that should work equally well. The input circuit of the
are selective enough to lengthen the pulse. Ideally a
noise blanker should be applied at the antenna
terminals of the receiver. This is not necessarily
impossible, but it is difficult, so customary practice
is to employ a noise blanker in the circuits after
the first mixer, before the selectivity -determining
stages. This type of circuit as described in all
modern editions of the ARRL Handbook for use at H
455 kHz is adequate for eliminating auto ignition
and the like, but radar interference is several orders
of magnitude more severe. A noise blanker used in i

' ~,
this manner is too late ín the circuit of a vhf
converter -receiver combination to prevent over- C

loading of earlier stages. " ,1

The blanker of Fig. 4-28 is installed between e oip
the vhf or uhf converter and the receiver used as ej..o
the i -f system, allowing the noise pulses to be

Fig. 4-28 -A noise blanker for insertion between a

vhf or uhf converter and the following communica-
tions receiver. Model shown is set up for 28 to 30
MHz, but other converter output frequencies can
be used by altering circuit constants.
64 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS

EXCEPT AS INDICATED, DECIMAL

VALUES OF CAPACITANCE ARE
IN MICROPARAOS (/lfI; OTHERS
ARE IN PICOFARADS (pt OR/Jpf);
RESISTANCES ARE IN OHMS;
N 1000 .001 + 0

j
--
I.001

-150V.
6.3V.

Fig. 4-29,- Schematic diagram and parts information for the noise blanker.

Cl - -1000-pF feedthrough capacitor. 3 turns. 14 MHz: 30 turns No. 30 enamel,

CR1
CR2 - - N64,
1 1 N60, or similar diode.
N920 or N3730 switching diode. L2-
tapped at 10.
Like L1, but no tap.
-
1 1

J1, J2 BNC connector. J2 has extension wire to L3 13 turns like L2.

be trimmed in length
with respect to L3. See text.
and adjusted in position R1- 5000 -ohm miniature control.
RFC1-RFC3 - 27-/1H rf choke.
J3 -4 -pin male power fitting. RFC4 - 500 -PH rf choke.
L1 - 28 MHz: 10 turns No. 30 enamel, close -
wound on 3/8 -inch slug -tuned form, tapped at

first stage is tapped to match the low -impedance Layout

line from the converter. Single-tuned circuits are
The noise blanker is built in a standard
used. A gain control is connected in the cathode
2 X 3 X 5 -inch Minibox, with all rf components in
leads of both tubes, and maximum gain is limited
a line down one side. The three coils are 2 inches
by 150- and 100-ohm series resistors.
apart, and the sockets are centered between them.
The output of the second stage is coupled into
The sockets are mounted so that the grid and plate
a pair of back-to-back diodes, CR1 and CR2. The
first is an ordinary second -detector diode such as pins, 1 and 5 respectively, are in a straight line, and
the 1N64 or 1N60, the other a selected computer nearest to their respective coil terminals.
A grounding lug is placed under each socket
diode such as the 1N920 or 1N3730. Output
brought out through Cl can be used to monitor mounting nut. A wire is run from the one adjacent
visually the noise pulses which are being blanked. to Pin 7 across the socket to the center shield ring
With a high -gain audio system connected to Cl all and then to Pin 3. Disk ceramic capacitors are
connected with the shortest possible leads in the
the signals in the bandwidth of the converter can
be monitored. This is not practical for hearing
following manner: Pin 2 to Pin 3, Pin 7 to ground,
weak signals, but it does provide continuous Pin 4 to adjacent ground lug, and Pin 6 to this
same lug. The bypasses at the bottom of the coils
monitoring on a lightly occupied band.
The i-f output coupling for the following
L2 and L3 are also returned to these lugs, which
are adjacent to the side of the unit. Spare lugs on
receiver is merely the very small capacitance of a
wire connected to the center conductor of J2, and the coil form for L3 are used for tiepoints.
placed near the output coil, L3. No direct connec-
tion is made at this point. The stray capacitive
coupling should be adjusted so that the gain
through the noise blanker is the same as it was Adjustment and Use
before the blanker was connected. After the wiring has been completed and
A word of warning: this is a high -gain lightly checked for errors the tubes should be installed
loaded amplifier. The rf chokes decoupling the and the coils L1, L2, and L3 adjusted to the
heaters and plate leads are essential if stable desired intermediate frequency with a grid -dip
operation is to be obtained. Physical layout like meter.
the original is also necessary, unless the builder is The unit should now be connected between an
experienced in such matters. operating converter and its i-f receiver. The noise
A Blanker for Pulse -Type Interference 65

Fig. 4-30 - Interior of the

noise blanker. The input
and power connector fit-
r
'
tings are at the left end in
this view. +-
_
a

°l. M, á
t o

.s WI
. o ,

level in the receiver should be adjusted by means of less of their amplitude when the blanker is out. If
the stray coupling between L3 and J2, to be the the blanker provides a substantial reduction in
same as before the blanker was installed. noise pulses, but does not completely eliminate
The gain control in the blanker is not for them, it is not working properly. There are only
purposes of adjusting output level. Its function is three reasons for this. The first is feedthrough
to adjust the input to the blanking diodes for around the blanker. The test for this was per-
optimum efficiency. Normally it will be at or near formed in a previous step and it is assumed that
full gain. While listening to the noise it should be there was no feedthrough. There may be insuf-
possible to remove the input cable from the ficient gain in the two amplifying stages. If you are
blanker and observe that the noise drops to using a normal vhf or uhf converter having 25 to
approximately the same level that it would if the 30 dB or more of gain, and tubes similar in
blanker was not turned on. characteristics to a 6AG5, the likelihood of low
It should be possible with a converter con- gain is quite small. The third and most likely
nected and the blanker gain control turned full on, reason for poor operation lies in the selection of
to peak up L1, L2, and L3 for maximum noise the proper diodes for CR1 and CR2. Some
without any evidence of regeneration. The 3 -dB experimentation with the polarity of the diodes
bandwidth of the blanking amplifier with the coils and the size of the diode load is usually required to
all peaked to the same frequency is approximately obtain optimum performance. The particular con-
500 kHz. Operation over 1500 kHz or more can be stants given in the circuit were successful in three
obtained with the coils adjusted in this manner. different models tested and no difficulty should be
In order to make the final adjustments on the encountered in obtaining optimum operation.
blanker it is necessary to have a source of noise and
a method of switching a blanker in and out of the Results
circuit. First, tune in a strong pulse -type signal The performance of the noise blanker in on -the -
such as radar or very strong automobile -ignition air operation leaves little to be desired from the
noise. With the blanker in the circuit, the power to standpoint of external noise elimination in uhf
the blanker should be turned off and the rf gain on work. With or without noise, the insertion of the
the receiver opened wide. At this time there should noise blanker in the circuit has no discernible
be no noise coming from the converter. Turning effect on the readability of a weak signal. In the
the converter power on and off should make no presence of pulse -type noise the signal continues to
change in the output noise of the receiver.
Now the blanker should be turned back on and
after warm-up the amount of noise reduction
should be observed. When the blanker is properly
operating there should be no noise pulses, regard-

Fig. 4-31 - Schematic diagram of the blanker

power supply. Filter capacitors are electrolytic.
J1 -3-pin female power connector.
T1 -
Small replacement transformer, 125 volts at
30 mA, 6.3 volts at 1 A. 6.3V.
CR1 -
400 -volt PRV silicon diode, 500 mA.
66 Vhf RECEIVERS, CONVERTERS, AND PREAMPLIFIERS
Fig. 4-32 -Power supply for use with the
blanker or other equipment items requiring
similar current and voltage.

Bibliography
1 Tilton, "Two -Band Station for the VHF
Beginner," QST, July through October, 1961.
Complete reprint of this widely used 4-part series is
available for 50 cents, from ARRL, Newington, CT
06111. All components of this station were des-
cribed in previous editions of this Manual: 14-MHz
tuner, 50- and 144 -MHz transmitters, modulator
and power supply, SWR bridge, and converters for
50 and 144 MHz.
2 DeMaw, "The D.C. 80-10 Receiver," QST,
May, 1969 and 48th and 49th Editions, The Radio
Amateur's Handbook.
3 DeMaw, "More Thoughts on Solid -State Re-
ceiver Design," QST, January, 1971. Also 1972
Handbook, "A Weceiving Package for 1.8 to 144
MHz,"
be perfectly readable and the noise is not evident 4 Hall and Wilson, "A Single -Band Converter,"
at all in the output of the receiver. QST, May, 1971.
Like all good things there are some drawbacks grated 5 Blakeslee, "A Second Look at Linear Inte-
Circuits," QST, July, 1971,
to the use of a noise blanker. The worst of these is 6 Constructional information on vhf receiving
that a very strong local signal will overload the equipment using vacuum tubes:
noise blanker and cross -modulate other signals on Nuvistor Converters for 50 and 144 MHz Part 4
of a 4-part series, see Reference 1. Matching
the band. This is an inherent trait of noise -blanker converters for 220 and 432 MHz were added later
circuits for which no solution has been found. In in the Handbook, and in previous editions of this
order to obtain sufficiently strong blanking pulses, Manual. Photocopies of information not appearing
in the QST series: add $1.00 to the reprint price of
high -gain amplifiers are required and high-gain 50 centss.
W2AZL Two -Meter Converter Excellent con-
amplifiers necessarily overload. This disadvantage is verter design, originally in December, 1959, QST.
far outweighed by the ability to copy weak signals (Photocopies $1.25) Also in Edition 1 of this
in the presence of strong pulse -type interference. Manual.
50- and 144 -MHz Reception at Low Cost
Even when the blanker is overloaded, signals which Simple converters and modification of BC -455.
Photocopies from Edition 1 of this Manual,
could not be heard through radar interference Bandswitchingg converter, 50 through 220 $1.25.
MHz,
without it are readable. from Standard Coil TV tuner. Photocopies from
previous editions, 11.00.
Power Supply Superregenerative Receivers for 50 and 144
Mhz Photocopies from previous editions, $1.50.
Details of a simple power supply for the tors50
-MHz mobile converter with 12 -volt Nuvis-
Photocopy from Edition 1, 50 cents.
blanker are given in Figs. 4-31 and 4-32. This type
of supply is useful for many purposes around a Copies of QST still in stock are available for the
station where the owner likes to build small units current cover price. Photocopies of out -of -print
items will be made for 25 cents per page, on
such as preamplifiers, test equipment, and other request. Glossy 8 by 10 prints of ARRL construc-
items requiring only a small amount of current. tion project photographs are available for $2.00
per print.
 Chapter 5

V H F Transmitter Design

Before we discuss transmitting techniques for main emphasis will be on transmitters that employ
the amateur bands above 50 MHz in detail it will crystal control or its equivalent in stability and
be well for us to see what standards are set for us freedom from spurious emissions. That such con-
in the U.S. Regulations. The two numbered para- cern for clean signals is required by law only on the
graphs below are not exact quotes, but they 50 -MHz band will be largely ignored.
summarize pertinent regulations. The frequencies above 50 MHz were once a
97.71 -
Amateur stations operating below 144 world apart from the rest of amateur radio, in
MHz must employ adequately filtered dc plate equipment required, in modes of operation and in
power fdr transmitting equipment, to minimize results obtained. Today these worlds blend increas-
modulation from this source. ingly. Thus, if the reader does not find what he
97.73 - Spurious radiations from an amateur needs in these pages to solve a transmitter problem,
station below 144 MHz shall be reduced or elimina- it may be covered in the hf transmitting portions
ted in accordance with good engineering practice of the ARRL Handbook. This chapter deals mainly
. . . In case of A3 emission, the transmitter shall
. with aspects of transmitter design and operation
not be modulated to the extent that spurious that call for different techniques in equipment for
radiation occurs, and in no case shall the carrier 50 MHz and up.
wave be modulated In excess of 100 percent. . .

Simultaneous amplitude and frequency modulation DESIGNING FOR VHF SSB AND CW
is not permitted. .. . The frequency of the
emitted carrier wave shall be as constant as the Almost universal use of ssb for voice work in
state of the art permits. the hf range has had a major impact on equipment
It will be seen that stability and quality design for the vhf and even uhf bands. Many
requirements imposed on all lower amateur fre- amateurs have a considerable investment in hf
quencies apply equally on the 50 -MHz band, but sideband gear. This equipment provides accurate
not to 144 MHz and higher. This is not to say that frequency calibration and good mechanical and
we should not strive for excellence on the higher electrical stability. It is effective in cw as well as
bands, as well as on 50 MHz, but it is important to ssb communication. These qualities being attractive
remember that we may be cited by FCC for failing to the vhf man, it is natural for him to look for
to meet the required standards in 50 -MHz work. ways to use his hf gear on frequencies above 50
A sideband signal having excessive bandwidth, MHz.
an a -m signal whose frequency jumps when modu- Thus increasing use is being made of vhf
lation is applied, an fm signal that is also ampli- accessory devices, both ready-made and homebuilt.
tude -modulated, a cw signal with excessive keying This started years ago with the vhf converter, for
chirp or objectionable key clicks - any of these is receiving. Rather similar conversion equipment for
undesirable on any band, but they are all illegal on transmitting has been widely used since ssb began
50 MHz. Any of them could earn the operator an taking over the hf bands. Today the hf trend is to
FCC citation in 50-MHz work. And misinter- one -package stations, called transceivers. The ob-
pretation of these points in an FCC examination vious move for many vhf men is a companion box
could cost the would-be amateur his first ticket. to perform both transmitting and receiving con-
From the stand-point of the law, there is a vast version functions. Known as transverters, these are
relaxation in the technical standards we must meet offered by several manufacturers. They are rela-
above the 50-MHz band. Looking at the question tively simple to build, and are thus likely projects
from the amateur point of view, however, there is for the home -builder of vhf gear.
little or none. The desirability of radiating the best
signal that is technically feasible is the same Transverter vs. Separate Units
throughout the vhf region, if the user is interested It does not necessarily follow that what is
in worthwhile results, and in causing a minimum of popular in hf work is ideal for vhf use. Our bands
trouble for his fellow users of the vhf bands. are wide, and piling -up in a narrow segment of a
It should be remembered that the use of band, which the transceiver encourages, is less than
unstable equipment is legal above 144 MHz, so ideal use of a major asset of the vhf bands -
long as the radiation from the transmitter remains spectrum space. Separate ssb exciters and receivers,
entirely within the assigned frequency band. There with separate vhf conversion units for transmitting
are some circumstances where very simple and and receiving, tend to suit our purposes better than
therefore unstable gear may serve useful ends, and the transceiver-transverter combination, at least in
these will be touched on later in this book, but our home -station service.

67
 68 VHF TRANSMITTER DESIGN
Future of Other Modes power level. A 144 -MHz transmitter may start at 8
to 8.222 MHz, which Is then tripled, doubled and
It should not be assumed that ssb will mono- tripled to 144 to 148 MHz. A 220 -MHz lineup
polize voice work in the world above 50 MHz in would use 8.149 to 8.333 MHz, tripled three times
the way that it has the amateur voice frequencies to 220 to 225 MHz. The portions of the 420- and
below 29 MHz. Sideband is unquestionably far 1215 -MHz bands where stabilized equipment is
superior to other voice modes for weak -signal DX usually employed are in third and ninth harmonic
work, but where there is plenty of room, as there is relation to the low end of the 144 -MHz band.
ín all vhf and higher bands, both amplitude and From these figures and from the chart, Fig. 5-1, it
frequency modulation have merit. A low -powered may be seen that coverage of 8 to 9 MHz in the
a-m transmitter Is a fine construction project for a oscillator stage could take care of most of our
vhf beginner, and fm has been gaining in popularity requirements through 1300 MHz.
rapidly in recent years. A reprint of a very popular Many other starting frequencies and orders of
4 -part QST series describing a complete two -band multiplication are usable. The oscillator -multiplier
-a-m and cw station for the vhf beginner is available approach is convenient in multiband designs, and it
from ARRL for 50 cents.' is widely used in equipment other than the
The decline in use of amplitude modulation has single-sideband variety. One weak point is that any
been mainly in high-powered stations. The heavy - instability in the controlling oscillator is multi-
iron modulator seems destined to become a thing plied: 6 times in the 50 -MHz example of the
of the past, but this should not rule out use of a-m. previous paragraph, 18 times for 144 -MHz opera-
Many ssb transceivers are capable of producing tion, 27 times for the 220MHz band, and so on. An
high -quality a-m, and one linear amplifier stage can oscillator that seems quite stable at 8 MHz may
build as little as 2 watts a-m output up to 200 suffer from drift, hum modulation, mechanical
watts or so, with excellent voice quality, if the instability or frequency modulation to the extent
equipment is adjusted with care. It should be that the signal at the 54th harmonic, 432 MHz,
remembered that the transmitting converter (or may be unacceptable to the critical worker. This is
heterodyne unit as it is often called) is not a particularly true of the conventional vfo. The
sideband device only. It will serve equally well with crystal oscillator is much superior for frequency
a -m, fm, or cw drive. control on 144 MHz and higher bands.
The possibility that harmonics other than the
THE OSCILLATOR -MULTIPLIER desired ones will appear in the output should be
APPROACH considered in designing a vhf exciter. Such un-
wanted frequencies may be a source of interference
A basic difference between ssb transmitters and to TV, fm, and other vhf reception in the vicinity
those for other modes is that the ssb signal, once of the amateur station. They can be reduced by
generated, cannot be run through frequency multi- taking suitable circuit precautions, but their inter-
plier stages to get to a higher frequency or band. ference potential should not be ignored.
Where modes other than ssb are used, generation of The number of frequencies that could cause
the signal usually involves an oscillator in the hf trouble can be reduced by using a high starting
range, followed by one or more frequency multi- frequency in the vhf exciter. A 24 -MHz oscillator
plier stages, and at least one amplifier stage instead of one at 6 or 8 MHz eliminates most of
working at the final frequency. A signal of any the harmonics that are potential sources of TVI in
kind at any frequency can be heterodyned to the low TV channels. Starting at 48 or 50 MHz is
another frequency. but only amplitude-modulated still better. There are good reasons for using 6 or 8
and sideband signals must be. MHz, however, if precautions are taken to prevent
The starting frequency of the oscillator -multi- radiation of unwanted harmonics. A stable oscil-
plier transmitter is usually in the 8 -MHz range, lator is more readily built in this frequency range
though 6, 12, 18 MHz and other frequencies are than for 24 MHz or higher. Crystals for 6 to 9 MHz
also used. Typically, a 50-MHz transmitter is are inexpensive, reliable and easy to use, while
controlled at 8.334 to 9 MHz, the frequency then those for 12 MHz and higher cost more initially
being tripled to 25 to 27 MHz, and doubled to 50 and require more care in application. Vhf crystals
to 54 MHz, before amplifier stages build up the are used mainly where economy in number of
1 For this and other numbered references, see stages and over-all current drain is an important
bibliography at the end of this chapter. consideration.

8000 KC 8100KC. 8200 KC, 8300 KC. 30 8400KC 9000 KC.

20 40 60 8O 20 40 BO 20 40 60 BO 20 40 60 80 I

114-148 i0 54
Mc
432 Yc 8000- 222 Kc 220 225
8000 KC X54 X 18 Mc. 8334 -9000KC.
819-8333 KC X6
1296 Mc. X 27
8000 KC. X162

Fig. 5-1 -
Our bands above 50 MHz are nearly harmonically related. The possibility of using a single
frequency -control system for all bands from 50 through 1300 MHz is illustrated in this chart. The
example is for oscillators in the 8 -MHz region, but other frequency ranges such as 6 or 12 MHz may be
used.
The Oscillator -Multiplier Approach 69

8-9Mc

Fig. 5-2 - Block diagram of a

34.33 Mc
(UNWANTED)

typical heterodyning process

for producing stable V FO - 50 51 Mc, TO AMPLIFIER
controlled signals at 50 and
144 MHz. 144-145 OR
42 Mc, 145 146 Mc
TO AMPLIFIER

Heterodyning 47 OR
47.5 Mc
Any two frequencies may be fed into a mixer
stage to produce resultant frequencies equal to the
sum and difference of the two. This process is example we use the sum of 42 and 8 to 9 MHz and
inherent in the superheterodyne receiver, but it 50 and 94 MHz but the differences are also
was not widely used in transmitters until the produced. The selectivity of the tuned circuits may
advent of single sideband. If a keyed or modulated be sufficient to reject the unwanted products but
signal is heterodyned properly, the product is an this should not be assumed. The output of any
exact replica of the original signal, with no more transmitter employing heterodyning should be
frequency instability or bandwidth than was pre- checked carefully to be sure that frequencies other
sent in the two components mixed. Heterodyning than the intended one are not being radiated. A
is thus a good way of obtaining variable frequency mixer stage requires only a very small amount of
control, since it is relatively easy to build a variable energy on the mixing frequencies to produce
oscillator for the hf range that is adequately stable output, so harmonics and other components of the
at its fundamental frequency. signals being mixed may beat with each other and
The process is shown in block diagram form in produce all manner of unwanted frequencies.
Fig. 5-2. The control signal is generated at some Mixing at low level, careful examination of the
frequency below about 10 MHz, 8 to 9 MHz in this spectrum for spurious products, and use of highly
example. A 42 -MHz signal from a stable source selective circuits for passing on the desired product
beats with the control signal in the mixer. Two and rejecting others are musts for the builder of a
main products, one at 34 to 33 MHz, and the other heterodyne exciter.
at 50 to 51 MHz, result. The unwanted difference
Crystal Oscillators
product is rejected by the filter, while the desired
sum at 50 to 51 MHz is passed on to succeeding Quartz crystals of many kinds and cuts are used
amplifier stages. for frequency control but all have one character-
The vhf man who works 50 MHz and higher istic in common: when a voltage is applied across
bands may employ heterodyning again to repro- it, the crystal is distorted mechanically. The
duce the 50 -MHz signal on another band. In the converse is also true: mechanical distortion of the
example the 50 -MHz signal is mixed with energy at crystal develops a voltage across it. This is the basic
94 or 95 MHz, to give coverage of the lower half of piezoelectric effect, discovered many years ago and
the 144 -MHz band. Other crystals can be used in applied to crystal control of oscillators as far back
either crystal oscillator to extend the coverage to as the 1920s.
any one -megahertz segment desired. The greatly magnified edge views of crystal
The chief problem in heterodyning is to prevent plates in Fig. 5-3 show, in simplified form, what
unwanted products from being radiated. In our happens to an oscillating crystal. The quiescent

FUNDAMENTAL OVERTONE

Fig. 5-3 - Greatly magnified edge views of quartz crystals, showing the mechanical distortion effect
when voltage is applied across the crystal. At the left is a fundamental crystal, and at the right is one
oscillating on its third overtone. Frequency of oscillation depends on crystal thickness the thinner -
the crystal the higher the frequency.
 70 VHF TRANSMITTER DESIGN
state is at the far left. The next two sketches show important as one goes to higher frequencies. The
the distortion at the positive and negative peaks of crystal oscillator should always be regarded as a
the oscillation cycle. The crystal is a very high-Q device for controlling frequency, not as a source
device. It will oscillate on one frequency only, of rf power. Control of feedback is also important.
determined principally by the thickness of the However control is achieved, feedback should be at
crystal. (The thinner the plate, the higher the a level that will allow the oscillator to start readily,
frequency.) Connected properly in an oscillatory but not enough to cause heating or frequency
circuit, the crystal will control the frequency of jumping.
oscillation within very narrow limits.
Crystals and circuits for their use in vhf Crystal Oscillator Circuits
transmitters are of two principal types: fundamen-
tal and overtone. It is important for the vhf worker An almost infinite variety of crystal oscillator
to understand the basic difference between them. circuits may be employed in vhf transmitters. Only
The fundamental crystal, whose mode of operation a few will be described here, to demonstrate basic
has just been described, is usually supplied for principles. These will satisfy most requirements,
frequencies up to about 18 MHz. Though funda- and though the literature contains special claims
mental crystals can be made for frequencies up to for innumerable variations, proper adjustment and
about 30 MHz, they are very thin and difficult to operating conditions are the principal factors in
handle and process above the normal commercial achieving the desired results.
limit of 18 MHz. (Overtone crystals can be A simple circuit useful for both fundamental
supplied for lower frequencies, and 12 MHz was and overtone oscillators is shown in Fig. 5-4A. It is
the dividing line for many years.) essentially the same, whether the device used is a
At frequencies above 12 to 18 MHz it is field-effect transistor, as shown, or a triode tube.
customary to go to overtone oscillators. Almost An overtone circuit for bipolar transistors is shown
any crystal can be made to oscillate on its third at B. Feedback to sustain oscillation is mainly
overtone, which is roughly three times the frequen- through the device capacitance in 5-4A, and the
cy for which the crystal was ground. In overtone feedback frequency is determined by the tuned
operation the crystal in effect breaks up into an circuit, L1C1. With a fundamental crystal in place,
odd number of layers, as shown in the right half of the circuit may oscillate on the crystal frequency,
Fig. 5-3. The oscillation cycle is given in the two regardless of the setting of Lid , but output will
sketches at the right. Because of mechanical rise sharply as the circuit is tuned through the
considerations, the overtone may not be an exact crystal frequency.
multiple of the fundamental frequency, though it An overtone crystal is merely one that has been
is always close to it. Only the odd multiples are processed and mounted in such a way as to
available as overtones; there is no such thing as a encourage overtone oscillation. With an overtone
second, fourth, or sixth overtone. The fifth, crystal in place, if feedback is not concentrated on
seventh, and ninth overtones can be obtained quite the desired overtone frequency by proper design
readily with suitably processed crystals. and adjustment of L1C1, the circuit may oscillate
Overtone operation with crystals processed for only on the fundamental frequency, or perhaps not
fundamental service depends on several factors, oscillate at all. The tuned circuit should have fairly
principally the flatness of the crystal and the high Q, and it should be adjusted with care, to be
method of mounting ín the holder. Because the sure that oscillation is on the desired overtone
layers for third-overtone oscillation in an 8 -MHz frequency, and that there is no energy at the
crystal are less than 0.004 inch thick, and for fundamental or second-harmonic frequencies. With
higher-order overtones progressively thinner, it can a 24 -MHz crystal, for example, there should be
be seen that minor variations in flatness or surface energy only on the marked frequency of the
imperfections quickly inhibit overtone oscillation. crystal, and none on 8 MHz (the fundamental) or
Crystals clamped between metal plates, as in the 16 MHz (the second harmonic). If a signal is heard
common FT-243 holder, seldom work well above at 8 or 16 MHz, the oscillator may be loaded too
the third overtone. heavily, or possibly the tuned circuit does not
Crystals processed for overtone operation usu- quite reach the overtone frequency. The latter is
ally can be made to oscillate on higher -order often the case with slug -tuned coils that may
overtones than the intended frequency.2 A crystal appear to tune, but which are actually only
marked for 24 MHz, normally an 8 -MHz funda- approaching resonance as the core centers in the
mental, will often work well on 40 MHz, 56 MHz, winding.
or even 72 MHz in suitable circuits. Unless the If overtone oscillation is attempted with crys-
purchaser specifies otherwise, crystal companies tals made for fundamental use, more feedback is
customarily supply third -overtone crystals for fre- usually needed at the overtone frequency than is
quencies from about 18 to 54 MHz, fifth -overtone provided by the circuit just described. Two meth-
for 54 to 70 MHz, and seventh -overtone for ods of supplying this feedback are shown in Fig.
frequencies up to around 100 MHz. Overtone 5-4C and D. In D the crystal is effectively tapped
crystals for frequencies as high as 150 MHz can be up the output circuit by the capacitive divider, of
made, but in amateur service frequencies above which C3 is the adjustable component. The lower
about 72 MHz are seldom used for direct control. the capacitance the greater the feedback. Usually a
For best stability any crystal oscillator should fixed 50-pF capacitor will do for C3, though lack
be run at low power input, and this is increasingly of an adjustment for feedback may allow this
 Crystal Oscillator Circuits 71
40-bo MC.

67054 MHz

(A) (B)

C2
24-27Mc.

470

(C) (D) E)

Fig. 5-4 -
Typical crystal oscillator circuits. The FET circuit, A, works with either fundamental or
overtone crystals, up to at least 54 MHz, though circuit Q is more critical for overtone operation. A
comparable bipolar transistor circuit is shown at B. Circuits C and D can be used to make fundamental
crystals oscillate on their third overtone, through control of feedback. Value of R1 depends on the
tube type; 47,000 ohms to 100,000 ohms is common. Values of Cl and Li, L2 depend on frequency.
A pentode oscillator -multiplier with double -tuned inductive coupling is shown in E. C2 should be the
lowest usable value; about 2 pF in E, 10 pF in others.

circuit to oscillate on the fundamental frequency In the tetrode oscillator the tube cathode is
of the crystal. In C the crystal is part of an above rf ground by virtue of the rf choke and
inductive feedback loop, L2, coupled to the tuned capacitor combination. Feedback is controlled by
circuit, LIC1. Though adjustment of the feedback the variable grid -cathode capacitor, C3, which can
is not convenient this way, the circuit has the be replaced with a fixed type once the approxi-
advantage of oscillating only on the overtone mate value needed is found. In its most common
frequency. In all such circuits it is possible to have application in vhf transmitters, the oscillator plate
too much feedback, which can result in over- circuit tunes roughly 24 to 27 MHz, permitting
heating of the crystal, multiple oscillation frequen- frequencies in this range to be taken off while
cies, or even free -running oscillation. The use of using fundamental crystals at 6 to 6.75, 8 to 9, or
fundamental crystals for overtone service is an 12 to 13.5 MHz. The stage is then an oscillator-
economy and a convenience that exacts its price in quadrupler, tripler or doubler, respectively. With a
careful adjustment for the desired results. wide -range variable capacitor for Cl, adjacent
Multiples of the crystal frequency (not over- harmonics can be tuned if this serves useful design
tones) can be obtained with the oscillator of Fig. ends.
5-4E. Here the screen of the tetrode or pentode This type of oscillator always works on the
tube simulates the plate of a triode oscillator. The crystal fundamental frequency. The tuned plate
plate of the tube serves as a take -off element for circuit merely emphasizes the desired harmonic; it
harmonics of the crystal frequency, usually the does not completely eliminate the other harmon-
second or third, though higher -order harmonics are ics. The circuit is useful with pentode -triode tubes,
present and can be used. The values of L1 and C1 with the triode portion doubling to 48 to 54 MHz.
determine which harmonic will be emphasized. The bipolar transistor oscillator, Fig. 5-4B,
Double -tuned inductive coupling, through the cir- should have a high value of capacitance across the
cuit L2C4, helps to reject unwanted harmonics. collector circuit, compared with other oscillators.
For best selectivity the capacitor C2 should be the Circuits for bipolar transistors must cope with the
lowest usable value, and the coupling between L1 low impedances characteristic of this type of
and L2 should be the least that will pass the transistor. It may be necessary to tap the collector
desired amount of energy at the intended harmonic down on its tuned circuit, in order to develop
frequency. Tighter coupling serves no useful pur- sufficient Q for the circuit to work properly.
pose, and may raise the level of unwanted- For best stability with any oscillator, crystal or
harmonic energy passed on to succeeding stages. other, the designer must follow certain rules:
 72 Power Amplifiers
1. Operating voltages must be constant, and mounting methods, both in the crystal holder and
except for tube heater voltage, pure dc. 2. The the crystal socket. A rule-of-thumb for crystal
oscillator load should be unvarying, and the stage pulling allows a 750-Hz change, maximum, for each
should not have to deliver power. 3. The oscillator megahertz of crystal frequency, without the series
should run at very low input, to avoid drift, both coil . Thus, the 4.5 kHz to be expected with a
long-term and cyclical, due to heating and cooling. 6 -MHz crystal, and 6 kHz with an 8 -MHz one
The ideal arrangement Is to build the best oscillator translate to roughly 35 kHz per crystal at 50 MHz
that is practicable, and follow it with a class-A and 110 kHz at 144 MHz, without the series coil.
amplifier. This is called a buffer, implying effective This can be increased to 100 kHz at 50 MHz and
isolation from load variations resulting from adjust- 300 at 144 MHz with the coil in the circuit, before
ment, keying or modulation of succeeding stages. the stability of the oscillator is degraded to that of
the average good VFO.
VARIABLE OSCILLATORS - The VXO principle is widely used in fm
VXO AND VFO transmitters, where precise "netting" of trans-
mitter frequencies is required in repeater work.
A variable capacitance or inductance, usually Obviously it also offers a simple approach to
the former as a matter of convenience, can be frequency modulation of the transmitter, by con-
made to vary the frequency of oscillation in a necting a varactor diode in place of or in parallel
crystal oscillator circuit to some extent. In free - with Cl. The bias on the varactor is varied at an
running oscillators a variable capacitor is usually audio rate, with simple speech circuitry.
the principal tuning device. Both types of oscilla-
tors will be discussed here, and practical examples
The VFO
are given in Chapter 6. The potential user should
have a clear idea of the strengths and weaknesses of Variable -frequency oscillators of the free-
both these means of variable frequency control. running type (VFO) are in great demand for vhf
transmitter frequency control, but except where
The VXO heterodyning to a higher frequency is used, as
If the inherent capacitance (crystal holder plus opposed to frequency multiplication, the VFO is
circuit strays) is kept low, a crystal oscillator can generally unsatisfactory. Small instabilities, hardly
be "pulled" over a small frequency range by means noticeable in hf work, are multiplied to unaccept-
of a variable capacitor across the crystal. A crystal able proportions in the oscillator -multiplier type of
oscillator In which this is done intentionally is transmitter. The fact that many such unstable VFO
called a VXO. A simple form merely varies the rigs are on the air, particularly on 6 meters, does
pressure (and thus the capacitance) on the metal not make them desirable, or even legal. Only
clamping plates in the pressure -mounted type of careful attention to all the fine points of VFO
design and use can result in satisfactory stability in
crystal holder. See Chapter 16 for an example.
vhf transmitters of the oscillator -multiplier type.
A better method employs a variable capacitor
across the crystal circuit, which may include a
variable inductor in series with the crystal, as FREQUENCY MULTIPLIERS
shown in Fig. 5-5. The coil L1 "rubberizes" the
crystal to a degree depending on the inductance Circuits for frequency multipliers are quite
value. Flexibility is obtained at the expense of similar to those in straight -through amplifiers dis-
stability, and the crystal oscillator is no better than cussed in the following section, except that higher
the free-running type, if the series-coil idea is driving power and bias are usually needed in
carried too far. With the variable capacitor alone multiplier stages. It is important to keep in mind
the frequency pulling is slight, but the oscillator that in multipliers for the vhf bands the probability
stability can be excellent if the numbered rules a is that frequencies other than the desired har-
few paragraphs back are followed. monics will be present in the output. These can be
With or without the series coil, most of the sources of TVI in vhf transmitters. Examples are
frequency change occurs with the first few pF of the 9th harmonic of 6 MHz and the 7th harmonic
capacitance variation, so keeping stray capacitance of 8 MHz, both falling in TV Channel 2. The 10th
to a minimum pays off in increased variable- harmonic of 8 -MHz oscillators falling in Channel 6
frequency coverage. Usable frequency change also is a similar problem. These unwanted multiples can
depends on the quality of the crystal and on the be held down by the use of the highest practical

SO

60R8 MHz

Fig. 5-5
uses
- Typical variable crystal oscillator (VXO)
the basic Pierce circuit. The variable capacitor,
4101C
C1, may be connected directly across the crystal
(high stability, small tuning range) or the series
coil, Li, may be added for more frequency
+I50v variation. Stability of the oscillator is degraded as
REG. the value of Li is increased.
VHF TRANSMITTER DESIGN 73
degree of selectivity in interstage coupling circuits filter is recommended, whether the multiplier
in the vhf transmitter, and by proper shielding and drives an amplifier or works into the antenna
interstage impedance matching. This last is par- directly.
ticularly important in transistor frequency multi- Frequency multipliers are usually single -ended
pliers and amplifiers. More on avoiding TVI will be though other combinations are possible. The
found later in this chapter, and in the chapter on "push-push doubler" (input push-pull, output para-
interference problems. llel) gives good efficiency and rejection of odd
The varactor multiplier (see UHF and Micro- multiples of the driving frequency. A push-pull
waves) is much used for developing power in the tripler, often used with dual tubes, both tetrodes
420 -MHz band. Requiring no power supply, it uses and triodes, is fairly efficient and tends to cancel
only driving power from a previous stage, yet quite even multiples of the driving frequency. There is
high orders of efficiency are possible. A varactor little tendency to oscillation in frequency multi-
triplet to 220 MHz is shown in Chapter 6. A pliers, and neutralization, screen -circuit bypassing
220 -MHz exciter tuned down to 216 MHz makes a and other stabilization devices of amplifier cir-
good driver for a 432-MHz varactor doubler. More cuitry are seldom needed. Some points discussed in
commonly used is a tripler with 144 -MHz drive. more detail in connection with transistor amplifiers
The output of a varactor multiplier tends to have may also apply to solid-state frequency multipliers
appreciable amounts of power at other frequencies to some extent, but instability problems are not
than the desired, so use of a strip-line or coaxial common in multiplier stages.

POWER AMPLIFIERS

Principles of transmitter design and operation available from the ssb exciter, it is also useful with
tend to be similar regardless of frequency. Basic the Class AB 1 linear amplifier, for only a watt or
amplifier theory is covered thoroughly in the two of driver output is required. When used with
ARRL Handbook, so modes of operation and an a -m phone transmitter, the linear amplifier is
circuit design of power amplifiers are discussed inherently a low -efficiency device, but it has its
here only insofar as the special aspects of the vhf virtues, particularly when other modes of opera-
field are concerned. We look at the principal tion are planned.
problems the vhf operator faces in thinking about Various routes to Increased power are shown in
increasing transmitter power. Fig. 5-6. Our basic a -m and cw transmitter, A,
requires only the linear amplifier, B, to run up to
Transistors or Tubes? full legal power of 1 kilowatt. This can be a
high -efficiency system on cw, delivering up to 750
In receiving, this question has been answered on watts output. As an a-m phone linear, its maximum
a strongly solid-state note. The problem is not
power output is 350 watts, and it is likely to be
quite the same in transmitting, except where much less. If the original transmitter includes
low -power mobile and portable operation is the provision for ssb, as well as cw and a -m, the linear
main concern. For more than a few watts of
becomes more attractive. Setup C can give full
power, transmitting with transistors may have little
power on cw and ssb, and medium power on a -m,
to recommend ít. The portable exciter using without auxiliary audio equipment.
transistors, that can serve as a complete rig for
Maximum power output on a-m phone requires
portable work and as a driver for the home station,
a high-level modulator, as in D. A 500 -watt audio
has some appeal, but tubes of some sort are almost
inevitable in running appreciable power at the system, needed to modulate a kilowatt amplifier, is
an expensive and bulky proposition, and with the
home base. Special transistor problems in ampli-
fiers will be discussed as they arise in individual current trend to ssb in amateur voice communica-
units, leaving tube techniques in amplifiers as our tion, more and more vhf men are thinking twice
principal subject here. before making the considerable investment in
terms of money, space and weight that a kilowatt
a-m phone station entails. There will undoubtedly
Linear or Class C? be considerable use of a -m in vhf work for many
Amplifiers in vhf transmitters all once ran Class years to come, despite the inroads of ssb, so the
C, or as near thereto as available drive levels would relative merits of linear and high-level modulated
permit. This was mainly for high -efficiency cw, and amplifiers deserve careful thought.
quality high-level amplitude modulation. Class C is If one is to concentrate on a-m, to the
now used mostly for cw or fm, and in either of exclusion of other modes, a plate -modulated power
these modes the drive level is uncritical, except as amplifier of no more than 200 to 500 watts input
it affects the operating efficiency. The influence of may be desirable. The cost of both rf and audio
ssb techniques is seen clearly in current amplifier components rises very rapidly above the 500 -watt
trends. Today Class AB 1 is popular and most level, and it may well be that the extra cost could
amplifiers are set up for linear amplification, for be better spent in other ways.
ssb and - to a lesser extent - a -m. The latter is For the all -mode operator, the linear approach
is very attractive, since minor modification of the
often used in connection with small amplitude -
modulated vhf transmitters, having their own operating conditions will permit high -efficiency
built-in audio equipment. Where a-m output is operation on cw and ssb, while retaining a -m
 74 VHF TRANSMITTER DESIGN
PHONE-CW
HIGH -POWER CW
MEDIUM - POWER PHONE
Fig. 5-6 -Some ways to
increase power in a vhf
(A) (B) station. Transmitter A is a
MIC.
typical packaged unit,
10- WATT complete with modulator
TRANSMITTER TRANSMITTER LINEAR
D WITH A AMPLIFIER and power supply. Adding
MODULATOR a linear amplifier, B, can
give up to 300-watts out-
KEY
put on a -m phone, or
HIGH - POWER
750 -watts output on cw.
CW OR SS B., The side-band exciter, C,
MEDIUM -POWER usually also makes pro-
A M
10 -WATT vision for cw and a -m, so it
EXCITER LINEAR
CW AM AMPLIFIER combines well with a linear
OR S S B amplifier for high power
on ssb or cw, and medium
power on a -m phone. A
HIGH - POWER small rf unit, D, Is used to
C W OR A.M drive a Class -C amplifier
(D) PHONE for high -efficiency cw.
10-WATT Addition of a modulator is
CLASS - C
EXCITER
AMPLIFIER required for high -efficiency
CW
a-m phone.

500 -WATT was the day of the 100 -watt transmitter, usually an
MODULATOR 829B, a pair of 6146s, or some similar setup, with
a companion 50 -watt modulator.
When vhf sideband became popular the usual
ssb exciter was also a low -powered device, usually a
capability at a somewhat lower power level - all low-level mixer and a Class -A amplifier, with no
with a station that can be built compactly and at a more than a few watts output on the vhf band in
moderate cost for the high -power portions. question. The logical amplifier for getting into the
A distinct advantage of the linear approach is medium- or high -power brackets with this type of
the matter of driving power. With a Class AB1 exciter is the grounded -cathode type, using tet-
linear (most commonly used with vhf tetrodes), no rodes. This puts the station capability up to several
driving power is required; only voltage. Kilowatt hundred watts, with a single stage having very low
amplifiers of the grounded -cathode type can de- driving power requirements.
liver over 300 watts output with nothing more Meanwhile, hf sideband was growing up around
than a 3 -watt a-m rig as a driver. A little more drive the 100 -watt exciter and the grounded -grid kilo-
will push the cw output to as much as 600 watts. watt amplifier. It was only a question of time
With a driver output of 7 to 10 watts, the before vhf stations would follow the same route.
amplifiers will give up to 750 watts Class C output Today there is an increasing trend to 100 -watt
on cw, ssb or fm. transmitters and transceivers in vhf work, espec-
Operating conditions for linear service are ially on the 50-MHz band. Thus, in the following
critical. The amplifier must be heavily loaded. If it chapter we show 50 -MHz examples of both the
uses tetrodes, the screen voltage, and preferably low -drive grounded-cathode amplifier and the
the bias as well, should be regulated. The drive grounded -grid type requiring an exciter capable of
level must be watched closely, to be certain that delivering 25 watts output or more.
the amplifier is never driven into the grid -current There is merit in both methods. The grounded -
region, if it is operated Class AB1. An oscilloscope cathode amplifier using external -anode tetrodes is
is practically a necessity, if true linear conditions capable of very high efficiency, even with only a
are to be achieved and maintained. In all these few watts of driving power. It can be shifted
respects the linear is more demanding than Class-C readily from one operating mode to another, to
cw or plate -modulated a-m service would be. suit the type of drive to be used, and it lends itself
nicely table -top style, requiring only a very small
Grounded -Cathode or Grounded -Grid? exciter. In return, it exacts a price in the form of
Nearly all vhf power amplifiers in amateur rather critical adjustment for optimum results, and
service have been of the grounded -cathode type, the need foi neutralization. Its power-supply and
mainly because of the high drive requirements of metering requirements are fairly complex.
the grounded -grid amplifier. In the heyday of a-m The grounded -grid amplifier can be quite simple
communication the typical rig was a transmitter or in regard to circuit, construction and operation.
transceiver of no more than a few watts output. It Triodes work well, simplifying power-supply prob-
had its own modulator, and the owner was not lems through elimination of the screen supply
ordinarily interested in building additional high- needed with tetrodes. Efficiency tends to be lower,
powered audio equipment. His choices were then but some of the driver power appears in the
either a high-powered linear amplifier, or a amplifier output. Neutralization is not ordinarily
medium -powered Class -C stage, the latter requiring required. If the driving power is available, the
a modulator of modest power to go with it. This grounded -grid amplifier is logical for increasing
Single -Ended, Parallel or Push -Pull? 75

__
T_C D

(A) rts (B)

Fig. 5-7 - Loading effect of input and output capacitances in single -ended circuits limits their use at
higher frequencies in the vhf range. In the push-pull circuit, right, these strays and also the tuning
capacitances are in series across the tuned circuits, permitting use of a given tube type at much higher
frequencies.
power, at least on 50 MHz. At 144 MHz and the minimum of C2 and circuit stray capacitance,
higher, drive is lidely to be more of a problem, and across L2. Obviously it will not be possible to
tubes designed for grounded -grid service are mostly resonate conventional tuned grid and plate circuits
types that do not perform well much above 100 at 144 MHz and higher, with tubes connected in
MHz. parallel, even when they are types designed for vhf
The tetrode grounded -cathode amplifier is cur- service.
rently almost standard procedure for 144, 220, and In the push-pull circuit B, the input capaci-
420 MHz. Above 1000 MHz, the trend reverts to tances are in series across the tuned circuit. So are
grounded -grid, mainly because tubes available are the two halves of the split -stator tuning capacitor,
coaxial uhf types designed especially for this Cl. The effective total capacitance across the tuned
application. The 2C39A, and its later version, the circuit will be about one fourth that of the parallel
3CX100A5, are widely used in medium -power connection. The same is true in the plate circuit.
grounded -grid service in the 1215 -MHz band. It can be seen that our chances for reasonably good
vhf circuit efficiency are vastly better with push-
Single-ended, Parallel or Push -Pull? pull than with parallel.
On lower bands use of two or more tubes in With single -tube amplifiers the parallel effect of
parallel is almost standard practice. Often it is less the tube and circuit capacitance still prevails, but it
expensive to use several small tubes in parallel than is not nearly so bad as with two or more tubes in
one larger one of the same total power capability. parallel. Most single -ended amplifiers for the higher
Parallel is preferred to push-pull in hf transmitters bands employ tank circuits which permit direct
mainly because of its simpler circuitry and ready connection to the tube element or socket tab, with
adaptability to bandswitching. Where tube and no leads in the usual sense. Coaxial lines or
stray circuit capacitances do not represent a large flat -strip tank circuits are preferred, especially for
percentage of the total, parallel connection of higher-power amplifiers. Even with the lowest
tubes is entirely satisfactory. possible capacitance, rf circulating current will run
Looking at Fig. 5-7A, we can see readily why very high in a vhf amplifier, so low dc and rf
parallel operation is not practical for the higher vhf resistance is of utmost importancee. Large con-
bands, using conventional tubes and circuits. The ductors have the added advantage of helping to
tube input and output capacitances, Cg and Cp, dissipate heat developed in the tube elements.
shown in broken lines, are in parallel with the Because of their compact construction and
tuning capacitors, Cl and C2, across the tuned short leads, power transistors work well in parallel
circuits. Suppose we select a pair of good vhf tubes up through 150 MHz, at least. It is also possible to
like the 4CX250B. This tube's input capacitance is use certain vhf and uhf tubes in parallel, with
16 pF. Thus, in circuit A we have 32 pF, plus the properly -designed strip -line circuits. This requires a
minimum of Cl, plus unavoidable circuit capaci- complete break with conventional coil -and-
tances, all in parallel across L1. Output capacitance capacitor concepts, as practical examples seen in
is 4.4 pF, so the plate circuit has 8.8 pF in Cp, plus later chapters will demonstrate.

(A) (B)

Fig. 5-8 -Basic functions of, tuning and impedance matching are performed equally well in the
pi -network, A, and inductively coupled output circuit, B. Choice in single -band vhf amplifiers is
mainly a matter of convenience in a particular design.
76 VHF TRANSMITTER DESIGN
Pi -Network or Inductive Coupling? 5-9A still work well if tubes and components
designed for vhf service are used. Conventional
The pi-network tank circuit, Fig. 5-8A, is circuitry may serve at 144 and even 220 MHz with
popular for transmitter use, largely because of its suitable tubes, but in general the usefulness of
adaptability to band -switching amplifiers. In single - coil -and -capacitor circuits is limited above 100
band vhf designs there may be little choice MHz.
between It and the inductively coupled circuit, Transmission -line adaptations of conventional
5-8B. The output circuit of an amplifier has two tuned circuits, 5-9B, extend the range and improve
basic functions: to tune the stage to the desired performance as we reach frequencies where we
frequency, and to act as a matching device between "run out of coil" with the circuits at the left. In
the stage's high output impedance and the low - the push-pull version, the inductance LI may take
impedance load. In the pi-network the tuning and the form of a U-shaped loop, or it can be a pair of
loading capacitors, Cl and C2, serve these pur- copper pipes, 1/8 to 1 inch or more in diameter,
poses. With inductively coupled circuits, either the with an adjustable shorting device at the end away
single -ended, 5-8B, or push-pull, 5-7B, the coil and from the tubes to adjust the total inductance in the
the output coupling loop comprise the matching circuit. The single -ended version below it can be
transformer. The two circuits work equally well, grounded at the left end, if a blocking capacitor is
and choice between them can be dictated by used at the grid, and the resistor R1 is connected
adaptability to the particular amplifier being built. from grid to ground. The effective electrical length
of Ll can be made variable by use of a sliding
Coils or Lines? contact.
On lower frequencies the fact that any capaci- With either circuit of B the upper limit of
tor has some inductance and any coil some frequency is reached when Cl is removed and the
capacitance can be neglected in most circuit -design ground point is moved up to the grid terminal. In
work, for these "strays" are too small to have any practice, the limit is reached when there is no
significant effect. At frequencies in the upper vhf longer enough exposed circuit to permit effective
range they become all-important. Connecting leads, coupling. We then can go to the circuits of C. The
which at lower frequencies merely join coils and rf gnd voltage Eg, is shown by the curve above
capacitors, may, in a vhf circuit, have more each set of circuits. In A and B the zero -voltage
inductance than the "coil" itself. Similarly, leads point is at the center tap or bypassed end of Ll, or
within tubes and sockets may become appreciable at the left end of the line. If minimum rf voltage
portions of a wavelength. Unavoidable capacitance occurs close to the tube, the line can be extended a
in rf circuits also severely restricts the upper limit quarter wavelength to the left, and the tuning
of frequency for satisfactory vhf amplifier per- capacitance connected across the left end. The
formance. whole circuit, including tube and tuning capaci-
At 50 MHz these factors are not insurmount- tance, now becomes an electrical half wavelength
able, if care is used in laying out amplifier stages. of line, loaded capacitively by the tube at one end
Single -ended or push-pull circuits such as Fig. and Cl at the other.

MOO. E9
Mox.
Molt

Mtn }Min.

(A) (B) (C)

1%4
a2

Fig. 5-9 - Evolution of the tuned circuit in vhf amplifiers. Conventional coil -and -capacitor tuning,
A,
becomes a circuit in B. A half -wave circuit is shown at C. Each has a progressively
quarter-wave line
higher upper useful frequency limit for a given type of tube, whether single -ended or push-pull design,
is used.
Multi band Amplifier Circuits 77

(B)

Fig. 5-10 -
Simple tricks for achieving multiband capability in vhf circuits. The 144 -MHz line, L1,
becomes merely a lead or pair of leads when the 50 -MHz circuit, L2, is plugged in at the point of
lowest rf voltage.

The bias resistors R1 and R2 should be con- that the connection on a half-wave line is made at
nected at the point of lowest rf voltage in C. This the point of lowest rf voltage, where it has no
can be determined by feeding rf power into the effect on the operation of the line, we realize that
circuit and touching L1 with a lead pencil or the 50 -MHz circuit of B can even be permanently
insulated metal object, until the point is found connected. This has been done many tines in both
where there is no reaction on the circuit.
The half -wave line circuit will extend the useful AMP.
frequency range well above the maximum obtain-
able with quarter -wave or coil -and-capacitor cir-
cuits. Though a grid circuit is shown in the
examples of Fig. 5-9, the principle is equally
applicable to plate circuits. The next steps after OUTPUT

this, coaxial and cavity circuits, will be discussed in

our uhf chapter.
Multiband Amplifier Circuits
Though conventional bandswitching and plug-in
coil arrangements are ineffective at 144 Mhz and
higher frequencies, it is possible to build multiband
tank circuits for vhf transmitters. Simple adapt-
ations of the plug-in coil idea are shown in Fig.
5-10A. Here a 144-MHz circuit, L1, is completed
by plugging in a shorting bar at the end of the line.
To use the circuit on 50 MHz or even lower
frequencies, we plug in a suitable coil, L2. This
general idea was used effectively in the plate circuit
of a 4-65A amplifier for 28, 50, and 144 MHz
described some years ago in QST3 and the Hand-
book. 4
A similar principle is applied to half-wave lines
in B. Again the 144 -MHz half -wave line, L1, 4400V.
BIAS
becomes merely the "leads" between the 50 -MHz
coil, L2, and the tuning capacitor. Remembering AMP.

Fig. 5-11 -Typical neutralization circuits for vhf

amplifiers. The circuit at A is used with triodes and
most other amplifiers, with Cl coupling energy
from the plate circuit back to the grid in the
proper phase to cancel feedback through the tube.
Circuits B and C may be needed with tetrodes,
when the frequency is above that at which the tube
is inherently neutralized. The variable capacitor Cl
in circuit C is adjusted to provide a low -impedance
path from screen to ground at the operating
frequency.
 78 VHF TRANSMITTER DESIGN
AMP. 21 Fig. 5-12 - Parasitic
suppression meth-
ods for vhf ampli-
fiers. In circuit A,
for 6 -meter opera-
PARASITIC
CHOKE
tion, Z1 is 3 or 4
turns of No. 14
(A) wire, wound on a
100 -ohm 2 -watt
noninductive resis-

AMP. Z2
f B+
PARASITIC NETWORK

1
tor. Z1 tends to
overheat in other
than low -power cir-
cuits, so circuit B is
CH
100 OHMS more practical for
SO Mc. 2W. high power. The
windings of Z2 are
INPUT determined by ex-
S0 Mc.
OUTPUT periment, usually 2
(B) RFC turns of No. 14, side
RFC by side on a resistor
0 z.001,U.f like that in A, tuned
BIAS 8+ to the parasitic fre-
quency by capacitor
AMP. C. Parasitics in cir-
cuit C are damped
144-148 Mc out with 56-ohm re-

(c) C
- OUTPUT
sistors bridged a-
cross portions of the
plate leads, to give
loading at the para-
/T' JYYY'1-0 a+
sitic frequency. The
.00luí-Z transistor amplifier,
D, is bypassed for
DRIVER both If and vhf, to
220 Mc.
discourage feedback
in both parts of the
RFC spectrum.

.00t,uf.
FT.

+24V.

manufactured and home -built gear for 50 and 144 completely removable plate circuit also brought
MHz, and it can be adapted to line tank circuits for 432 MHz into the picture, permitting use of the
still higher frequencies. stage as a doubler or tripler. This design by
Some critical problems are involved in turning W1VLH appears in Edition 1 of this Manuals
this trick for 50 and 144 MHz, especially when
both the grid and plate circuits of an amplifier are Stabilization
operated in this way. Because of the nearly Most vhf amplifiers, other than the grounded -
third-harmonic relationship, considerable care must grid variety, require neutralization if they are to be
be exercised in proportioning the tank circuits to
satisfactorily stable. This is particularly true of
prevent radiation of energy on unwanted frequen- ABl amplifiers, which are characterized by very
cies, or oscillation troubles due to unwanted
high power sensitivity. An example of conventional
resonances in the grid and plate circuits. An
neutralization is shown in Fig. 5-11A.
example of a design in which these potential
A tetrode tube has some frequency where it is
troubles were avoided was shown ín an amplifier inherently
by WOIC in QST. 5 neutralized. This is likely to be in the
lower part of the vhf region,
By thinking in terms of the job to be done, hf service. Neutralization for tubes designed for
rather than of the way such tasks have been may be required in of the opposite sense
such amplifiers, as in the
handled in the past, it is often possible to come up example shown in Fig.
5-11B.
with solutions that are unique to the vhf field. A Conventional screen bypassing methods may be
grid circuit tuning both 144 and 220 MHz made ineffective in
the vhf
possible an efficient transmitter for these bands, in screen to ground, as in range. Series -tuning the
5-11C, may
which only the plate circuit was changed. Ai situation. A critical combination be useful in this
of fixed capaci-
Wide -Band FM with Simple Gear 79
tance and lead length may accomplish the same text. Bandwidth comparable to that employed in
result. Neutralization of transistorized amplifiers is entertainment -type fm broadcasting relaxes trans-
not generally practical, at least where bipolar mitter stability problems. Even a simple modu-
transistors are used. lated -oscillator transmitter can be made to deliver
Parasitic oscillation can occur in vhf amplifiers, good -quality fm, if the power supply is well -
and, as with hf circuits, the oscillation is usually at filtered and some provision is made to keep the
a frequency considerably higher than the operating deviation within the limits a 100-MHz fm
frequency, and it cannot be neutralized out. broadcast receiver will accept. The stability of the
Usually it is damped out by methods illustrated in modulated oscillator is not sufficient for use in
Fig. 5-12. Circuits A and B are commonly used in heavily-occupied bands, but the simple approach is
6 -meter transmitters. Circuit A may absorb suffi- logical for the 420-MHz experimenter.
cient fundamental energy to burn up in all but Equipment of elementary simplicity for use in
low -power transmitters. A better approach is to use the 220 -MHz band described and demonstrated by
the selective circuit illustrated at B. The circuit is W1CTW years ago7 is still potentially useful.
coupled to the plate tank circuit and tuned to the The author of this book demonstrated an even
parasitic frequency. Since a minimum amount of simpler arrangement for 420 -MHz work at many
the fundamental energy will be absorbed by the radio clubs and conventions. A little 6J6 oscillator
trap, heating should no longer be a problem. was modulated by a 6AQ5 audio stage.8 With
At 144 MHz and higher, it is difficult to speech input held so low that the modulation
construct a parasitic choke that will not be percentage was only about 5 per cent, the signal
resonant at or near the operating frequency. could be received with quite satisfactory quality
Should uhf parasitics occur, an effective cure can with a simple tunable converter and an fm broad-
often be realized by shunting a 56-ohm 1 -watt cast receiver. The signals sound more like a
resistor across a small section of the plate end of buzz -saw than speech, when picked up on a
the tuned circuit as shown in Fig. 5-12, at C. The selective communications receiver, but with wide -
resistor should be attached as near the plate band fm detection they can be above reproach. We
connector as practical. Such a trap can often be have plenty of room for them above 220.5 MHz,
constructed by bridging the resistor across a and in the low part of our 420 -MHz band.
portion of the flexible strap -connector that is used
in some transmitters to join the anode fitting to
the plate -tank inductor.
Instability in solid-state vhf and uhf amplifiers Bibliography
can often be traced to oscillations in the If and hf
1 Tilton, "A Two -Band Station for the VHF
regions. Because the gain of the transistors is very Beginner," July through October, 1961, QST.
high at the lower frequencies, instability is almost Includes simple 14 -to-18 -MHz tuner; 10-watt trans-
certain to occur unless proper bypassing and mitters for 50 and 144 MHz; receiving converters
decoupling of stages is carried out. Low-frequency for 50 and 144 MHz; power supply, modulator and
control unit; and vhf SWR bridge. Complete
oscillation can usually be cured by selecting a reprint of QST series, with templates for drilling
bypass-capacitor value that is effective at the principal surfaces, 50 cents per copy, from ARRL,
frequency of oscillation and connecting it in Newington, Ct. 06111.
parallel with the vhf bypass capacitor in the same 2 Tilton, "Overtone Crystals
to Use Them," March, 1955, QST.
- How and Where
part of the circuit. It is not unusual, for example, 3 Chambers, "450 Watts on VHF," September,
to employ a 0.1-µF disk ceramic in parallel with'a 1949, QST.
.001-µF disk capacitor in such circuits as the 4 ARRL Handbook, 27th, 28th, and 29th
emitter, base, or collector return. The actual values Editions, Chapter 17.
used will depend upon the frequencies involved. 5 Meer, "The Perseids Powerhouse;' October,
This technique is shown in Fig. 5-12D. 1959, QST.
6 Southworth, "Using the 4X250B on 144,
Wide -Band FM with Simple Gear 220, and 432 Mc.," February, 1957, QST.
7 Hadlock, "Wide-Band F. M. Gear for
Wide -band fm is very easily achieved with any 220Mc.," March, 1961, QST.
transmitter that is VFO-controlled, using methods 8 Tilton, "Simple Gear for the 420 -Mc. Begin-
described in Chapter 6, Fig. 6-13 and associated ner," May, 1949, QST.
 Chapter 6
Vhf Exciters and Amplifiers
As was done with the
subject of receiving in companion items will be pointed
two previous chapters, we are out. Power
covering
design, adjustment and operation in transmitter supplies and modulation
equipment are seldom
chapter, and practical examples in detail in one included, as these usually follow
descriptive items to follow,
another. In the changes only slightly over periods practice that
of many years.
and adjustment procedure explanatory material The reader is referred to
minimum necessary for adequate
will be held to the the ARRL Handbook for appropriate sections of
details of accessories that
unit described. The reader is coverage of each are needed in these fields.
urged to examine Some mention of the 420 -MHz
Chapter 5 thoroughly before
embarking
construction of equipment to be described on the found herein, despite the designation band will be
here. MHz semantically as "vhf."
of 30 to 300
This section will deal mainly Where the design
portions of transmitters of the with the rf techniques involved are truly uhf in
oscillator -multiplier for the 420 -MHz band will be nature, items
type. Where items are coordinated found in
units that appear elsewhere in in design with dealing with uhf and microwaves, later the chapter
the book, the in the book.

FREQUENCY CONTROL
Most transmitters of the
oscillator -multiplier critical ear. By the time we multiply
type in this book are shown with 54 times, to
Though being able to move crystal control. 432 MHz, even average crystal
around control is not good
becoming almost as important in vhf at will is enough for narrow -band work.
lower bands, the fact remains that work as on There are two solutions:
heterodyning, which
frequency -control systems presently many variable duplicates the fundamental -frequency
stability on
50 MHz are far from in use above a higher frequency, and very
satisfactory, except when special attention to
heterodyning methods replace frequency the stability problem in oscillators
multipli- that are to be
cation. In the case of 50 MHz followed by one or more frequency
downright illegal. From 144 MHz up
operation, some are example of the latter approach is multipliers. An
we are not detailed below.
required by law to transmit stable
signals, but self
respect and consideration for others A VXO FOR 50 THROUGH
we keep our signals above dictate that 450 MHz
reproach, regardless of Crystal control has many advantages.
frequency. By the
This is not easy when very nature of the quartz
crystal, the frequency of
continously-variable
frequency control is used, especially a crystal oscillator is
at 144 MHz maintained very close to the
and higher. A VFO that sounds desired spot. The effects of heating
8 -MHz region may be
good enough in the
only fair at its 6th harmonic contraction of the oscillating device(expansion and
and its circuit
in the 50 -MHz band. At elements), mechanical vibration and
the 18th harmonic, 144 supply voltages are greatly variations in
MHz, it very likely will be reduced, in comparison
unacceptable to the with these effects in any
self-controlled oscillator.

=-

5
Fig. 6-1 -A VXO espec-
ially for vhf use. Calibra-
tion on the front panel is
for a favorite crystal used
for cw work on 144 and
432 MHz. Crystal sockets

\,
; t-
a
F =-
,,a
at the lower left
mounted on insulating
are
material, to reduce circuit
' - . capacitance to the lowest
s+ possible value. Frequency
variation per crystal de-
pends on which socket is
used. Pointer knobs are for
the output plate circuit
and the spotting and power
switches.
The vernier dial is a
National type AM.

80
Frequency Control 81
osc CLASS A AMP MULTIPLIER
6A K5 6U8

Fig. 6-2 -
Schematic diagram and parts information for the VXO and power supply. Unless specified,
resistors are 1/2 watt. Decimal values of capacitance are in µF; others in pF. Capacitors with polarity
marked are electrolytic. Terminal strips J6 and J7 may be omitted and connections made directly
where the power supply is built in. Pin 4 of J7 permits use of the supply for other purposes.

Cl - 100-pF per-section split -stator variable L2 - 24-35-1.1H,

4508).
iron slug, ceramic form (Miller
(Hammarlund HFD100). 50 pF per section also
usable. L3 -3.541H, 21 turns No. 24 tinned, 1/2-inch dia,
Cl C3 - 50-pF miniature variable (Hammarlund
L4
32 t.p.e.
-3 turns like L3, spaced 1 turn from it. Make
H F-50). Higher maximum capacitance
(HF-100) may be used. Grounded-rotor type both from single piece of B&W Miniductor No.
preferred. 3004.
J1 - Crystal socket for 0.05 -inch pins, spaced L5 - Same as L3, but tapped at 3 turns. Coax
0.487 inch. from L5 to P2 may be any convenient length.
-
Crystal socket for 0.095 -inch pins, spaced P1 - 300 -ohm line plug.
P2 - Coaxial cable fitting.
J2, J3
RFC1 - 750-µH rf choke.
0.487 inch.
J4 -2 -terminal barrier strip. Omit if fm is not to
RFC2 - 1.0 mH rf choke.
be used. Remove jumper when fm is connected.
- Coaxial S1 - Spst switch.
S2 - Spst switch. (See text).
J5 receptacle.
J6 -3 -terminal barrier strip. T1 - Power transformer capable of delivering 200
J7 -4 -terminal barrier strip.
Lt - 16-24-µH, iron slug, ceramic form (Miller to 250 volts dc at 50 mA through filter, 6.3
volts ac at 1 A, and 5 volts ac at 3 A.
4507).

But even with crystal control, the fundamental limitedby the amount of instability you are willing
requirements must be met if we are to have highly to accept.
stable control of frequency. These become more The variable crystal oscillator (VXO) shown in
stringent as the order of frequency multiplication Figs. 6-1 through 6-4 allows the operator a choice
is increased. of variable capacitance alone, or in conjunction
It is possible to "pull" the frequency of a with a series coil. Furthermore, the amount of
crystal oscillator a small amount in several ways. A inductance in series with the crystal, and con-
mechanical method is described in a later chapter, sequently the frequency shift obtained by rotating
but it is adapted to use only with pressure- the variable capacitor, can be adjusted to suit the
mounted crystals. Controlled voltage variation builder's desires. Since temperature variation is the
causes some shift, but is usually associated with principal cause of drift in crystal oscillators, this
large changes in output. Adding capacitance across one is run at low input, and drift is held to a very
the crystal works well with some crystals, and the small amount, even from a cold start. The oscilla-
swing with a given amount of capacity change can tor runs continously, so there is no heating and
be increased by adding inductance in series with cooling cycle effect in transmitting.
the crystal. (See previous chapter.) The frequency With just variable capacitance (no series coil)
change with these methods (as with any other) is the maximum usable swing is roughly 4.5 kHz at 6
 82 VHF EXCITERS AND AMPLIFIERS
MHz, or 6 kHz at 8 MHz, for crystals in the small The frequency is pulled by the split -stator capaci-
metal hermetically -sealed holders. FT-243s and tor, C1, connected between plate and screen. The
other pressure-mounted crystals having high holder oscillator plate voltage is regulated 150. Input is
capacitance may swing quite a bit less. There is a held to about 3 mA, combined plate and screen, so
certain total capacitance at which each crystal goes this oscillator is not going to move much unless
out of oscillation, and it varies markedly from one you move it with Cl. An rf choke is used in the
to another, depending on crystal activity and plate circuit, instead of a resonant coil, as tuning
mounting methods. here would pull the frequency.
The 6000 -kHz crystal shown plugged into the To build up the low oscillator output to a
VXO in Fig. 6-1 covers 432.24 to 432.0 MHz and usable level, and to provide isolation, a buffer
144.08 to 144.0, without use of the series coil. amplifier follows, using the pentode section of a
This gives all the coverage usually needed for 6U8, V2A. This tube was selected because it has
weak -signal cw work in these two bands, but the the lowest grid -plate capacitance of any dual tube
oscillation frequency goes lower than 6000 kHz, of the pentode-triode class. The triode portion
and if the crystal had been a shade higher V2B, is a multiplier, the output frequency depend-
frequency it would have been more useful for ing on the crystals used. Provision is made for
144 -MHz service. About 6001 or 6002 would have covering 12 to 26 MHz with C2L3. The plate
been ideal. Available swing is mainly on the low circuit of the pentode amplifier is broadly tuned,
side of the marked frequency. and an intermediate setting of the slug in L2 can be
With the series coil, Ll in Fig. 6-2, about three found that will permit use of either 6- or 8 -MHz
times as much variation is possible without serious crystals in the oscillator. The plate circuit of the
degrading of the stability. A range of 100 kHz at multiplier may then be tuned to the second, third,
50 MHz and 300 kHz at 144 is average for 6- or or fourth harmonic of 6 MHz or to the second or
8 -MHz crystals. Rubberiness varies considerably third of 8 MHz. Which output frequency you use
from one crystal to the next, with the series coil may depend on the type of circuit into which the
and without it. One ordinary FT -243 surplus VXO works. More on this later.
crystal at 8.38 MHz was adjustable over an
operating -frequency range of 50.34 to nearly 50.0
MHz, but this is exceptional. The need for variable
Construction
control is confined mainly to narrow high -activity Mechanical layout of the oscillator portion was
segments of the vhf bands, so a few selected dictated by the need to keep circuit capacitance to
crystals will do the job for most vhf operators. a minimum. The lower the total capacitance in the
Random crystals for other parts of the bands need circuit, the higher the frequency will go with Cl at
not be swingers, ordinarily. mínimum, and the wider swing you'll achieve per
Operation on 220 or 420 is almost always crystal. This rules out crystal switching, though if
channeled to one narrow segment per band, and convenience outranks crystal economy in your
this is handled easily with one crystal per band. objectives, switching can be used. Crystal sockets
Usually one crystal will serve for both 144 and are mounted on a Plexiglas insert in the front
432, for use ín the parts of these bands where high panel, instead of directly on the metal. The tuning
stability is most desirable. Use of the series coil is capacitor ís shimmed up an extra quarter inch
not recommended above the 144 -MHz band. above the chassis, to hold down its minimum
capacitance, and rf leads through the chassis have
Circuit Details half-inch clearance holes. Any one of these steps
yields little, but combined they net quite a bit
The oscillator, V1, is a 6AK5, but almost any more coverage at 432 MHz. This dividend is at the
small receiving rf pentode will do. The basic VXO low-C end of the range of Cl, where oscillator
idea can also be used with transistor oscillators.
stability is at its best.

1.1

'

Fig. 6-3- Interior view of

the VXO. The oscillator
tube is at the right. The
power supply, shown here
as a separate assembly,
could be built On the same
chassis with the rf circuits,
if the constructor wishes.
A VXO for 50 Through 450 MHz 83
Fig. 6-4 -
Bottom of
the VXO.
The oscillator
components 11 r," ""C.er t >Q <7 r7 .
are at the
right, the am-
plifier
multiplier
stages near
and

-
WIt---
the center, r'
_it
1
,,,
y
and the pow- -
i ft.
the left. -1..4 ;I
`!1
s
_;,

S
-.

Three crystal sockets, J1, J2, and J3, are wired cuit, L5C3, permits use of any convenient separa-
so that a crystal may be plugged into the circuit tion between the VXO and the transmitter. Rf
either with or without the series coil. Two differ- from the plug, P1, can be fed into the transmitter
ent types of sockets in parallel, J1 and J2, permit in several ways. Some experimenting may be
small -pin or large-pin crystals to be plugged into needed with your setup, but typical circuits are
the high -stability low-swing portion of the circuit. shown in Fig. 6-6.
Use a wider variety of sockets if your crystal stock Triode Overtone Oscillators. Don't try to plug
requires it, though each one adds a little capaci- directly into the crystal socket without modifying
tance. the circuit. Mounting an extra socket, J2 in Fig.
Any crystal you plug into this circuit will 6-6A, allows you to return to direct crystal control
oscillate on its fundamental, including those in- at will, yet gives optimum transfer of power from
tended for overtone operation. Most crystals above the VXO. Remove the regular crystal from J1
12 MHz are overtone types, the third overtone when the VXO is used, of course. With the
being used up to about 54 MHz. A crystal marked capacitive feedback circuit, Fig. 6-6B, the 50-pF
for 24 MHz will oscillate near 8 MHz, but not capacitor should be shorted out, and the VXO
necessarily at exactly one-third the marked fre- output fed to J1.
quency. If you're ordering crystals especially for Pentode Oscillators. The pentode crystal oscilla-
this purpose, we recommend 6 to 6.5 MHz, which tor circuit used in many vhf transmitters should
will cover 50 to 52 MHz, 144 to 148, 220 to 225, have its cathode rf choke shorted by means of a
and 432 to 450 MHz. The output frequency would switch. Plugging into the crystal socket may work
then be, preferably, 12 to 13 MHz, as this will with such circuits, unmodified, but more reliable
allow the crystal oscillator stage of most vhf
transmitters to work as a frequency multiplier
when driven by the VXO. Use of 8 -MHz crystals
and 24 -MHz VXO output is usually satisfactory
where the first stage in the transmitter proper is a
pentode, but triodes may self-oscillate, unless
operated as multipliers.

Coupling to the Transmitter

The coupling system shown in Figs. 6-5 and 6-6
is not the simplest way of hooking the VXO to a
transmitter, but it has certain advantages. Low-
impedance coupling terminated in the tuned cir-

Fig. 6-5 - Coupling assembly to be used for f--".

plugging into the exciter driven by the VXO.
Components are L5, C3, and P1, of Fig. 6-2. The
tuned circuit covers 12 to 26 MHz. A larger
variable capacitor may be used to make the value
of L5 less critical, if desired.
 84 VHF EXCITERS AND AMPLIFIERS

(A)

Fig. 6-6- Modifications of various crystal oscillator circuits for VXO drive. J1 is the original crystal
socket. J2, where required, is an additional socket, for VXO Input. A and B are typical triode
overtone
oscillators. C is a popular pentode oscillator. Two options are shown for C. To convert the
a multiplier stage, close Si and S2 and feed drive into J1. The
oscillator to
oscillator may be disabled by opening
S2, in which case drive is fed to J2.

operation is likely when the cathode is grounded Refinements in the spotting technique can be
for rf, as with Si in Fig. 6-6C. made to suit the operator's preference, though the
Another possibility in working into an existing circuit is useful as shown. With power applied to
transmitter is to disable the transmitter crystal the amplifier and multiplier through S2, the signal
oscillator, and couple into the grid of the second is just plainly audible on 432 MHz, when the
stage from the VXO. Opening the screen or B -plus heaters are on in the rest of the transmitter. It is
lead of the first stage, as is done with S2 in Fig. stronger progressively on each lower band, but the
6-6C, is handy for this, and a crystal socket may be signal from the oscillator alone is inaudible, even
connected to the grid of the second stage, as shown on 50 MHz. If you make a practice of zeroing the
by 12 in Fig. 6-6C. Here again, reversion to other fellow's frequency most of the time you may
standard crystal control is easy. want to install a small relay, actuated by your main
Simplification is possible when the VXO is built transmitter control, in parallel with S2. Then leave
directly into a transmitter designed for it. Here, the the switch in the open position normally, closing it
output of the isolation amplifier will be sufficient only for spotting purposes. A spring -return substi-
to drive a frequency multiplier to 24 MHz, so one tute for S2 may be desirable in this case.
stage is saved compared with the system wherein The series coil, Li, is adjusted by the core stud
the VXO is used to work into the crystal oscillator seen on the front panel, just to the right of the
stage of a transmitter designed for 6-, 8-, or crystal sockets. Moving the core into the coil raises
24 -MHz crystals. But don't skip the buffer ampli- its inductance and increases the swing per crystal.
fier; its functions are vital. Some practice with various crystals will be needed
before you know just what to expect from each
Operation one. The coil comes into play only when the
For maximum stability, particularly in crystal is plugged into J3. Instability increases with
432 -MHz cw work, it is well to leave the VXO on inductance, and also with increasing capacitance in
continously during an operation period, and prefer- Cl. Listen to the note critically, and check for
ably warm it up a few minutes before going on the mechanical effects when the unit is jarred. Don't
air. This way there is almost no frequency change, push your luck, or expect to swish all over the
except those deliberately made by moving Cl. band with one crystal, even though you'll find one
now and then that will make this possible.

Generating FM
Frequency modulation of the VXO is easily
done. A small audio voltage applied to the screen
at J4 will give good-quality narrow-band fm on 220
MHz and higher, even with the high -stability
oscillator arrangement. For 50- and 144 -MHz work
it may be necessary to use the series-coil circuit to
get enough deviation for good audio recovery at
the receiving end. All that is needed in the way of
audio equipment is a microphone transformer, a
flashlight cell and a carbon microphone. Remove

0 the jumper shown across J1 in Fig. 6-2, of course.

A topnotch nfm signal can be generated with a
very simple audio amplifier having a limiter, and a

Fig. 6-7 - Transistor VFO for vhf use. Tuning

range is about 8 to 8.42 MHz as described.
Transistor V FO for the Vhf Bands 85
Fig. 6-8 -
Interior of the transistor VFO. The
oscillator portion is built on a circuit board. The
amplifier stage is close to the rear wall.

good microphone. Swinging the frequency with a

varactor diode across Cl offers interesting possibil-
ities, and somewhat more deviation than is possible
with the screen-modulation method. Varactor-
-diode modulation is shown in the 220 -MHz
transmitter later in this chapter. For more on fm,
see Chapters 10-12.

TRANSISTOR VFO FOR THE VHF BANDS

r
Sources of hum in variable-frequency oscillators
using tubes (ac used on tube heater circuits, and
inadequate filtering of the plate supply) can be
eliminated with a transistor oscillator. Some other
instability problems are more readily solved with
transistors than with tubes. With a battery supply,
frequency changes due to voltage fluctuation are
nil. Drift cycles resulting from heating and cooling s
of the oscillator and its circuit components are v
greatly reduced in good transistor VFO design.
This is not to say that the transistor is a cure-all trift, and this can be corrected by adjusting the
for VFO problems, but with proper attention to temperature-compensating padder across the tuned
basic principles of oscillator design a reasonably circuit. A buffer -amplifier provides some isolation,
good VFO using transistors can be made at and builds up the output up to permit operation
moderate cost. An etched circuit board takes care with most vhf transmitters having crystal -con-
of most vibration problems. The transistor should trolled oscillators that multiply into the
run far below its rated dissipation, to hold down 24 -25 -MHz range.
heat cycling, and preferably run continuously
during an operating period. The VFO should be in Construction
a position as free of external heat changes as The VFO is built in a box 3 inches square and 5
possible. The oscillator should be isolated from the 1/4 inches long. This was made to fit the job, and
rest of the transmitter with suitable buffer and is not difficult to duplicate, but standard cases of
multiplier stages. something like this shape and size could be
These points were considered in the design of substituted, as there is nothing sacred about the
the VFO unit shown in Fig. 6-7. Its transistor layout. Tuning is by means of a small imported
oscillator tunes 8.0 to about 8.42 MHz, or enough vernier dial, the knob of which was replaced with a
to cover the entire 144- and 220 -MHz bands, and larger one, in the interest of smooth control. The
the 50 -MHz band from the low end to about 50.5 front panel is 3 1/2 by 3 1/4 inches in size,
MHz. More coverage can be included with a larger fastened to the case with bolts passing through
capacitor for C6. The oscillator input is less than 11/16-inch metal sleeves. These were filed down
40 milliwatts, so there is little heating and warmup from their original 3/4 -inch length.

Fig. 6-9 - Schematic diagram and parts infor-

mation for the 8 -MHz VFO unit. Parts not des-
cribed below are numbered for identification in
text. Where not otherwise indicated, capacitor
values are in picofarads (pF or µµF). Those across
tuned circuits are dipped -mica.
C6 - 15-pF variable (Millen 22015). For more
tuning range use larger value.
C7 - 8- to 50-pF neg. temp. coef. ceramic trimmer
(Centralab N-650).
J1, J2 -phono jack.
L1 - 3-µH slug -tuned coil: 13 turns No. 22 on
3/8 -inch iron -slug form (Miller 42A336CB1,
with 2 turns removed).
L2 - 1.3-µH slug -tuned coil: 14 turns No. 26 on
1/4 -inch iron -slug form (Miller 4502).
L3 - 2 turns No. 26, wound over low end of L2.
CR1 -9 -volt Zener diode.
P1 - Phono plug. Short out if no spotting switch is
used.
S1 - Remote spotting switch, any type.
 86
VHF EXCITERS AND
AMPLIFIERS
Fig. 6-10 -Bottom
bly. The oscillator of the transistor VFO assem-
over a 2-3/4 -inch circuit board assembly mounts
. round
bushings are in upper righthole. Power feedthrough
t
,5, t . corner.
Almost any hf or vhf
The circuit board is transistors may be
Ç. ` t n-p -n can be designed so that either p -nused.
-p or
accommodated, and there is no
advantage either way.
case p -n-p types are Motorola 2N4125 plastic-
shown. RCA 2N1177s were
also tried. A zener diode,
CR1,
diagram, does not appear in shown in the circuit
is used to regulate the photographs. This
the supply voltage if other than
a battery source is
used. It would be
the VFO were to be important if
used for mobile work,
car battery. In on the
operation from a small
battery the Zener can be separate
voltage applied directly toomitted, and the positive
and R4. the junction of R2, R3,

I<. Adjustment and Use

Make certain that all
ly, then apply voltage. circuits are wired correct-
the current drawn, It would be well to
which will be about 10 check
12 volts and 7 to 8
mA mA at
two thirds is drawn with 9 volts. Of this, about
by the amplifier. If
oscillator current is the
low-range meter, a measured separately
slight flicker in current on a
seen when the oscillator will be
coil or tuning capacitor is
touched, if the circuit is
can be checked oscillating.The frequency
All oscillator roughly with a wavemeter,
signal can be monitored or the
capacitor, C6, arecomponents except the tuning
calibrated receiver tuning
on a receiver. A well-
circuit board designedmounted on a multipurpose
for service in ARRL handy, but not necessary. the 8 -MHz range is
where tunable projectsin your 50- or 144 The signal can be
oscillators are required. A layout -MHz receiver, if heard
given in Fíg. 6-11, if is connected the VFO is
you
If not, prepared boards want to make your own. used. to the transmitter
with
can be 'purchased Methods of connection to which it is to be
suppliers mentioned in QST from various
The drawing and from time to time. crystal oscillator circuits are discussed types of
photographs tion with the in connec-
where the various parts should make clear VXO. See Fig. 6-6 and
are mounted. text. A shielded 8 -MHz associated
The amplifier stage is Fig. 6-12. coupling unit is shown in
The transistor is at the back of the chassis. To calibrate the
mounted
through bushings (Johnson on three small. feed- all -in, and oscillator, turn capacitor C6 to
place when pressed Rib-Loc). These stay ín adjust the core slug in LI
so that
other Rib-Loc terminals into a 0.136 -inch hole. Two frequency
be
is approximately 8000 kHz. This the
heard at the low end of the 144 will
of the power source. provide for the connection you are -MHz band, if
on the back wall of the These and the output jack are a link from listening there. It may be
the back, is chassis. On a side J1 to the 144 -MHz necessary to run
another phono jack for wall, near you are not driving the receiver input, if
connection as a spotting external transmitter at this time.
the amplifier stage on or switch or relay, to turn
off.
Small front and back
edges. The cover is a plates have folded -over
these, and to the sidesU-shaped piece, fastening to
of the chassis. The
capacitor mounts on the
the rotor shaft. The front panel with a tuning
type that grounds the nut on
this way is preferred to rotor in
ing studs, as the those having small
electrical and mechanical mount-
ing is much better. ground-
If more frequency
desired, the tuning coverage is
larger size than the capacitor should be the next
then be removed if the 15-pF type shown. Plates
coverage is too great. can
Fig. 6-11 -
the transistorGeneral purpose circuit board used
VFO is for
areas are etched. Board2-7/8 inches square. Dark
hole. mounts over a 2-314 -inch
Frequency Modulating the Transistor V FO 87
Fig. 6-12 -Coupling assembly, for plugging into a
transmitter crystal socket. Tuned circuit is similar
to that shown in Fig. 6-2, except L5 has 24 turns.
Interior appearance is similar to Fig. 6-5.

A drift check should next be made. Be sure that

the receiver is not drifting, then turn the VFO on
and note the drift for the next two minutes or so.
If it is appreciable, it can be reduced by proper
setting of the compensating trimmer, C7. Move the
trimmer a few degrees either way, and reset the
core slug so that the signal is again heard at 8 MHz.
Check the drift cycle again. If there is less drift this
time, from a cold start, you moved the trimmer in
the right direction. If there is more, move the
trimmer about as far in the opposite direction,
reset the slug, and try again. A combination will be
found eventually where the temperature -compen-
sating qualities of the trimmer will almost exactly
nullify the effect of the slight transistor warm-up e

drift. The temperature compensation, is really just

a refinement. the stability of the oscillator is quite
good without compensation, If you don't mind
letting it run for 10 minutes or so before using it.
It's not bad, even from a cold start.
A dummy load for the amplifier can be made Fig. 6-12, was plugged into the crystal socket. In
with a 2 -volt 60-mA pilot lamp, wired to a phono each instance the note quality was acceptable, and
plug and inserted in J1. It will just glow when the there is substantially no frequency modulation.
amplifier output circuit is peaked, indicating an Drift is no more than some crystal-controlled
output of 30 to 40 milliwatts. This is enough transmitters show at comparable frequencies. Sta-
to drive any of the common crystal oscillator bility does not approach that of the VXO just
circuits, if the coupling system of Fig. 6-2 is used. described, but it is better than many VFO units
For an 8 -MHz circuit, L5 in Fig. 6-2 should have currently heard on the vhf bands.
24 turns. Frequency Modulating the Transistor VFO
The simple bipolar transistor buffer stage does
not afford a high degree of isolation. There will be Noting that only about 12 pF of capacitance
an appreciable frequency change as the coupling change is required to tune the VFO just described
circuit is adjusted, but the stability in actual use is over an operating-frequency range of nearly 500
fairly good. This is not a device for working close kHz at 8 MHz, it can be seen that frequency
to the band edge, however. Play safe, and be sure modulating it enough for present-day vhf fm
that you know where you are. communication would require only a tiny capaci-
This VFO was tested on 50 and 144 MHz with tance change across the tuned circuit. It was found
the "Two -Band Station" transmitters described in by experiment that changing the capacitance across
earlier editions of this Manual.1 These were modi- the buffer circuit, L2, with a varicap diode could
fied by installing a switch that shorts out the rf pull the oscillator frequency adequately, and with
choke in the cathode circuit. See Fig. 6-6. The reasonable stability and linearity.
VFO works well with the 220-MHz transmitter A speech amplifier and limiter similar to that
described latter in this chapter, if the oscillator shown with the 220 -MHz fm-cw transmitter of Fig.
circuit of this unit is modified in a similar manner. 6-26 was used, with the varicap diode connected
It drove a Clegg 22-er nicely when the coupling through a small capacitor to. the VFO amplifier
1 "Two -Band Station for the VHF Beginner," collector circuit. An RCA Integrated Circuit Kit
QST, July through October, 1961. Complete audio amplifier and tone oscillator, KC4003, was
reprint available from ARRL, Newington, CT
06111, price 50 cents. All components were also used as an audio voltage source.
described in previous editions of this Manual. As a matter of convenience, the jack J2 in Fig.
6-9 was used for the audio voltage input. The diode
lay

RI 181(
.I,uF
AUDIO 4700 10 TO LI OR L2
IN ) F196-9

Fig. 6-13 -Basic circuit of a varactor diode

modulator for generating fm with the VFO of Fig.
VARICAP

6-9. Values of R1 and R2 can be adjusted to give

the bias required for the varicap diode in use.
 88 VHF EXCITERS AND AMPLIFIERS
D- SPEECH
AMP
VF0 DIODE
MODULATOR
7.5-8 MHz
Fig. 6-14- Typical example
of a heterodyne -type exciter,
MOSFET CLASS -A TO POWER

MIXER AMP. p--». -535 giving 500 -kHz tuning ranges

XrAL 055 100 mW ORR 51.5.5 -53 in two bands. Stability and
MHz
42.5 OR
45.0 MHZ
L -
KEY
d FM OR CW deviation in the fm signal are
established ín the VFO, tuning
7.5 to 8 MHz. The first ampli-
fier can be keyed for cw.
50.0-50.5 OR

MOSFET
MIXER
j

-
52.5-53 MHZ

94 MHZ
CLASS -A
AMP.
100 HnW
TO POWER AMP
144-144.5 OR 146.5-147 MHZ
FM OR CW

XTAL OSC-
DOUBLER
47- 94

and associated components were mounted

under tor fed into Gate 1 and a crystal oscillator
the chassis in the open area visible in Fig. 6-10. The of
added circuitry is shown in Fig. 6-13. R1 and R2 suitable frequency fed into Gate 2. The sum
can be adjusted to give the bias value frequency is a 500-kHz tunable range, depending
recom- on the crystal selected, 50.0 to 50.5 or 52.5
mended for the varactor diode being used. This is to
not critical, and anything from 1 to 3 volts 53.0 MHz. A Class -A amplifier using any small
-
worked signal vhf amplifier transistor builds the
satisfactorily with several low -capacitance varicap mixer
diodes tried in this application. Only a fraction of a output up to usable level.
volt of audio is needed, when the VFO The 500 -kHz ranges in the 50 -MHz band thus
output produced can be heterodyned to similar ranges in
frequency is being multiplied. the 144 -MHz band, using another mixer
If operation is to be attempted in the parts and a
the 6- and 2 -meter bands where fixed channels of crystal oscillator -multiplier with 94 -MHz
output.
are Additional crystals can be used in either crystal
in vogue, it Is recommended that the
oscillator be oscillator for more ranges. Adding crystals is
run continously, so that when the desired
cy is zeroed the VFO frequency will frequen- preferable to Increasing the VFO tuning range, for
remain obvious mechanical and stability reasons.
constant. The stability should be more than ade-
quate to meet the fairly stringent requirements of The voltage applied to the first mixer, or to
this kind of communication, if this is done. either Class -A amplifier, can be keyed for
cw
(See operation. Amplifier keying is probably the better,
Chapter 10.) The deviation should be adjusted so as the mixer has some small output,
that the signal is of good quality when slope - even with
voltage removed, if both oscillators are left
detected on a communications receiver having run-
average a -m selectivity. (Do not use a ning. Keying of the oscillators results in
broad inevitable
chirp problems. Amplifier keying is
receiver such as is common in small a -m excellent; there
trans- is no frequency pulling and no
ceivers for this check.) backwave.
The upper poruon of the system has been
used
on the air, as shown, for local work, and
as a driver
Heterodyne FM-CW Exciter for the tetrode amplifier of Fig.
6-38, which
The upper portion of the system shown in Fig. resulted in about 50 watts output. Results in
narrow- and wide-band fm work were very
both
6-14 has been tried in mock-up form, working good,
with and the cw signal is excellent. The setup
both fm and a -m stations. The VFO circuit was
tuned down by means of the core in Ll so that the to 100 mW on 50 MHz, with about 100delivers up
mA drain
at 12 volts. Many contacts have been
coverage is roughly 7.5 to 8 MHz. The
deviation made in the
can be adjusted by listening in this portion of the 50-MHz band where a -m phone
range, as there is
is no increase in deviation or still widely used. Zeroing any
instability when the frequency, for
signal is heterodyned to the operating communication with a -m or fm stations, is easily
The mixer is a MOSFET, with the variable frequency. done if a communications receiver is
used.
oscilla- key-up condition in the Class-A amplifier The
leaves
10011. 150K
VFO
IN
T.5-BMHZ

50-50.5 OR
525-53 MHZ
XTAL
IN AMP
42.5 OR
45 MHZ

100
Fig. 6-15 -
Circuit of the MOS-
FET mixer used in the
dyne exciter shown in hetero- block-
* GATE PROTECTED
diagram form in Fig. 6-14. A
-L"-°12V Class -A amplifier is added to bring
output up to usable level.
Vhf Transmitter Design and Packaging 89
just about enough signal for spotting purposes, and should be encouraged for mobile and repeater
ordinarily. work, a tunable approach to fm communication is
The all -transistor 2 -meter portion has not been surely a "must," if we are to make the most of the
tried, but the 50 -MHz stages have been used with home -station potential of this mode.
the heterodyne exciter described in Chapter 7, Further steps in the direction of versatility
with equally good signal quality and zeroing ability immediately suggest themselves to operators fam-
in the 2 -meter band. If we are to have tunable fm iliar with ssb transverters. The output of a sideband
exciters that meet the stability requirements of exciter in the 7 -MHz range could be substituted for
channelized fm on the vhf bands, there is little the fm input to the mixer, to develop vhf ssb with
doubt that the heterodyne approach is the logical the mixer and crystal oscillators as shown. The
way to do the job. It is good enough for ssb, which VFO accessory could be eliminated entirely, if a
is much more critical as to stability than is fm. varactor fm system were added to the tunable
Though fixed -channel fm has many advantages, oscillator in the ssb rig.

VHF TRANSMITTER DESIGN AND PACKAGING

The amateur about to begin assembling a vhf powered rf units for 144 and 220 MHz are laid out
station is frequently confused by the choice of so that the exciter and amplifier can be built and
transistors, tubes, circuits, operating modes and operated separately, if desired, though they go
equipment packaging available to him. Having been together in standard 17 -inch -wide packages.
exposed to the "station in a box" approach now so Use of circuit boards facilitates design by
common on lower amateur frequencies, he may subassembly, permitting a compact end result
feel that he wants a similar one -package station for without sacrificing easy modification. But metal -
his vhf work. This is certainly convenient, espec- plate and chassis methods need not be considered
ially if he plans to take ham radio with him on his obsolete, as the continuing popularity of many
travels, but it is by no means the most versatile "old but good" designs demonstrates. A classic
way to set up a home station. example of the long -life potential of sound sub-
Many vhf operators do not build stations for assembly design is seen in the still -extensive use of
the sole purpose of talking to people. We like to components of the "Two -Band Station for the
try different circuits, methods, and modes. We VHF Beginner," featured in QST more than a
look forward to gradual building -up and refine- decade ago, but still going strong in reprint form,
ment of our stations, as time, experience and despite its all-tube format,1 Its transmitter units
financial resources permit. With some planning, are still good for the man who wants an effective
today's low -powered transmitter becomes to- a-m and cw rig of simple design and moderate cost.
morrow's exciter. If it is a separate unit, it can still Another all -tube transmitter set for 50 and 144
serve for portable work. MHz is available from QST and recent editions of
There is much to be said for subassembly the ARRL Handbook.2 It provides somewhat more
design, even if the eventual objective is the power than the 10-watters of Reference 1, and is
containment of the whole station in one unit. Most complete with power and audio equipment.
equipment described in this book is worked out 2 "50-Watt Transmitters for 6 and 2," Radio
along these lines. Even the two complete medium - Amateur's Handbook, Editions 44 through 48.

A MEDIUM -POWERED 144 -MHz TRANSMITTER

pie transmitter of Figs. 6-16 through 6=23 was

built in the ARRL Laboratory, specifically for the
Headquarters Station, W1AW. Some aspects of its
design may be slightly different from those the
average vhf man would build into his station, but
anyone wanting an efficient and reliable 144 -MHz
transmitter should find it of interest. The exciter
and final amplifier are built on separate standard -
sized chassis, and either may be used with other
suitable equipment. The exciter makes a fine
low-powered rf unit by itself. and the amplifier will

Fit]. 6-16 - The 500 -watt 144 -MHz transmitter is

built in separate assemblies using standard chassis :
sizes, yet it can be mounted on a standard 19 -inch
rack panel. The exciter, left, may also be used as a
low -powered transmitter, capable of up to 10 watts
output. Amplifier, right, has built-in bias supply.
Its simple strip -line plate circuit is enclosed in a
removable cover.
 90 VHF EXCITERS AND AMPLIFIERS
work well with any exciter capable of delivering for 8 -MHz
crystals, though fundamental crystals at
over 5 watts output. Combined, the two units 6 or 12
MHz will also work. The switch S1 selects
mount on a standard rack panel, making a compact one of five
rf section capable of delivering 300 watts output VFO input crystals. One position may be used for
of the switch S2 (not in the original
on cw and 200 watts a -m phone.
unit) is incorporated to short out the cathode rf
There is much to be said for a 500 -watt level as choke, RFC1.
the maximum power input for a vhf station. The oscillator plate circuit multiplies to 24
Overall cost and complexity are far lower than MHz. Its
plate coil, L1, is inductively coupled to
when the full legal power is used, and the the
following grid circuit, L2, and the coupling is
difference in signal at a distant point is no more increased by
than barely noticeable. The economic advantage VI the link, L3. The second pentode of
doubles to 48 MHz. Its plate circuit and the
gained, ín staying below S00 watts input is particu- grid
circuit of the push-pull tripler, V2, are also
larly marked when a -m phone is the most -used link
-coupled. The tripler plate circuit and the grid
mode of operation. circuit of the 144 -MHz amplifier, V3, are induct-
ively coupled. Amplifier output is taken off
Exciter Design through a series -tuned link to a coaxial fitting on
The exciter is designed to permit shifting the rear wall of the exciter chassis.
frequency over most of the band without extensive Grid current in the tripler or amplifier stages
retuning. If the stages are peaked near the middle can be measured by connecting a low -range milli -
of a two -megahertz range normally used, very little ammeter between the exposed terminal of C3 or
readjustment of the exciter will be required. Even CS and the chassis. This is helpful in initial
in moving from one end of the band to the other, adjustment, and for trouble shooting, íf needed.
only repeaking of the tripler and amplifier tuning
capacitors will be needed. The tubes run at a The Final Amplifier
conservative level, to assure trouble -free operation It is desirable that a transmitter be capable of
in the continuous nightly service encountered in running at
bulletin and code -practice transmissions from the work, where moderate power, particularly in vhf
about 50 watts
Headquarters station. Double -tuned inductively - most communication. The output is adequate for
coupled circuits throughout give the desired band- vhf tetrode fits this external -anode type of
need admirably, as it will run
pass response, but with selectivity sufficient to efficiently at
inputs of 100
attentuate unwanted multiples of the oscillator be pushed up to 500 watts watts or less, yet it can
with complete safety.
frequency that might go through to the amplifier The 4CX250R shown can be replaced with
and be radiated if simpler circuits were used. any tube of this family. Many builders will want to
There are three dual tubes in the exciter. The use the 4X150A,
which can be found on the
oscillator and multiplier stages are 6AR11 dual surplus market at
pentodes, and the output stage is a 6360 dual equally well, except for attractive prices. It will work
tetrode. A 6CX8 or similar pentode -triode could be The socket specified takesslightly reduced ratings.
substituted for the 6ÁR11 oscillator-tripler- and 4CX250B and R. Withthe 4X150A, 4X250B,
a suitable socket, and
doubler with only minor circuit changes. A 6360 is possible modification
usable for the push-pull tripler, if the Compactrons care of differing of the grid circuit to take
are hard to come by. The oscillator, VIA, is set up many
input capacitances, any of the
tubes of this general type can be used.

-4:11
w a

n
0

A
w
a

-4111, .,) MI
/rig
OM.<
+r
Fig. 6-17 -Rear
view of the com-
plete transmitter,
with amplifier shield
removed.
A Medium -Powered 144 -MHz Transmitter 91

Oh L-1 la

' .

v/'i- '
111.
l

J
Fig. 6-18
stages are
- Interior view of the 144 -MHz exciter. The oscillator portion is at the left. Note that all
inductively coupled, for maximum protection against spurious frequencies in the output.

The amplifier tank circuits are made from for cw, or if the ultimate in oscillator stability is
flashing copper, a readily -available material that wanted. Otherwise Pins I and 3 of the power plug,
can be cut and bent without special tools. Details P2, can be connected together.
of the grid inductance, L12, and the plate line, The 4 -pin fittings on the exciter and amplifier,
L13, are given in Fig. 6-23. Many plate circuits J5 and J6, are wired so that meters for the final
were made and tested to derive the shape of L13. grid and screen current can be connected exter-
These included copper pipes from 1/8 to 1 inch in nally. The meters are not in the photographs, as
diameter, and copper strips of several widths and they are mounted on a separate panel in the W1AW
configurations. The plate circuit was operated as a setup.
pi -network, as well as in the inductively coupled The filament transformers Ti and T2 cón-
form shown here. When optimum L/C ratio was nected back-to-back give isolation from the ac line
achieved (tuning to the desired frequency range for the bias supply, and take care of the heaters of
with the lowest usable capacitance) there was no the transmitter. The blower motor, B1, comes on
essential difference in results, so the convenient whenever the primary of Ti is energized. The
and safe grounded tank shown was adopted. switch 54 is connected externally and does not
appear in the photographs.
Power Circuits The exciter heaters should be operated at 6.3
volts, but the amplifier tube should run at. 6.0, plus
Only two external power supplies are needed or minus 5 percent. With today's line voltages
for the rf portion of the transmitter: one delivering
250 volts dc at about 150 mA, and a high-voltage
supply giving anything up to 2000 volts at 300
mA. Control of the ac voltage input to the 1

i
final -stage plate transformer by means of a Variac
or Powerstat is an excellent way of adjusting the
transmitter power level to the needs of the
moment. A bias supply for the amplifier is built in.
The single 250 -volt source handles the exciter
stages and final amplifier screen. It should have
good regulation. The oscillator screen voltage
4á
should be regulated if the transmitter is to be used

Fig. 6-19 -
The amplifier plate circuit is mainly a
piece of flashing copper. It is grounded for dc,
making for safety in operation and ease of con-
struction. Tuning is by means of a disk capacitor
on a brass lead -screw, right. Plate voltage is
shunt-fed through the rf choke, upper right.
 OSC-TRIPLER
SARI! DOUBLER
TRIPLER
VIA 2 AMPLIFIER
V2 2 6ARII
IA4MHt V3 e
4
144 ARM

CIO

RFC2

TO PIN I,
J5
EXCEPT AS INDICATED,
DECIMAL
VALUES OF CAPACITANCE
ARE J6 SCREEN
IN MICROFARA DS
(NI ); OTHERS + P2
ARE 114 PICOFARADS
I01 OR NNI); TO OSC.
RESISTANCES ARE IN L15
K .
OHMS; SCREEN +150V REG
1000

p+ 250V
AMPLIFIER
4CX25OR
V4
L12 RFC4
T1"TZ
-0 TO 150V 4

CRI 100
2,4,6,8 CI8
100
C13
CIT
100 350%7 1I5V AC

PLATE
JB

er +1000 TO 2000V 300 , MA

0-500

Fig, 6-20 - Schematic diagram and

parts information for the 500
-watt 144 -MHz transmitter.
A Medium -Powered 144 -MHz Transmitter 93
C ° O °m ` O C
d
-aE
. o yL-°
m a
E
ó,.-daóÑ
a c U

c a,o g
óNEó«á `-'có
áL ) U
w L 9P L
cm
H
7o
O
ú pN
a ta)

m a a d Á p
C0 N Ñ o- ym c
;, a .0
5 áy ºrQl E o-o.Y
m L -Ñ Óa
Ocmr°)La)
« O N
.°).. U w O d

m=12 H0.)54)

°i O U w V a)
a)
2 O a>
..

L M C
rn°'aanm°)a)
"r 111-mi..OÑm

`O
c
UyaiL
a)

ó..E
ÚC
N
p, 0

.
¡<0
-mm~
Z'-
oá
y
dL E
«.
º'
<0

op..óC.3-

° J^ -'`wE
>112.5C-1'aioo
NóQCóOóa
Z L
m

Z
YO

fL0

c
M'X M L
a
p._V
a¡<;,YO

gil
- t
Ú
c

_C
L.2 c

'61-É
L U

d
ú
cp

0.

maOMm.EcSZox~«á)y-ma)
..,.d. -,_z
pNO c io
Ó.

O._ó)
a)LL
a?
(S

=UEc
'
m °YXi¡
tir a>cp

-iv
-0i.;
mÚ

Qcv
Ñ Ñ OL_J¿-I.--6, (O-
N
.

N
L
..

C)
pm

Cw
__
U
O

fl.

12 xc0«,á pUc
M a

Q.
O
cº
c
m
2t
c
(dd
.Y
cLi

cco
L!_
á)'
p.

L a« Ót° vvóJó-° vI.

° m
aciv pa_v a:.= d 3
H°moQ<°L`>>>-.oca.. .
7
O

0
.

,
Ñ
-
?
^
ri
c7

Ñ
Ó

Ñ
.
CO

.c
U
c

C)

v3`m
?`
p0
°
3m
0)

d7N>aó«; 3d-
ONCO

m
O

ctn
z
m
Ó
C O

N
Ó.Ñ.m-
V d

<0r

e
c2
o
p

m
LS
3
«

w-.

_ o o
á ó E H
~O
z
(.i
LL

Ú
YO

71
.

Y
c`

C
á)
Eó
O C
m
12
,d
aci
E
mw
wQ
.d

E
m

d
'

á M Ó Ú OaL w S.'aÓ ÓC N L º'T ° ó

CO

i.3 E y«Úr Ó
C« V Oa-áNU C JLL,.cOInMÓZ CQQ\ ó H-CMM
óÓ E
«
d
a a°'
t ó m
ó,M
° ó c
S..ac._
Uco
É
vu
MÑ
ó
ó O
C i
I^ E.' F. ÑcocD
3

mN ÁM
1----,»

^
Ó)

á$á
1 I

l
ÚÚMÚÚ
I
roi
ó c m I.Eci
I

11
m
'_
I

1 1 1 1

I
LLLL LL LL ,--N
GJo
J J J J J Jd(LOC S v)f/) coM 1-H

Ó p, Ó
a
j
d
=á -ó U"
N

i ó E
11111
ó n d C,---
cLi

'5u_ Ó - Ú d w >
O
c rn
C
'-

'
m <Ó Vi fl. O
:_ó<a,vE óu-1'5 H
.
E o a; =á Mcm
o
L aá
coQ

c E u E E Eci
mJ E
m U `o °:° a
13 -E.0
m o E E '5,-
15,-'- 1 > á.Y 2 _d M
L E c u .-,-,_o LU
C O _co N
c°'
O` N m a)
m
E
C
7 VC É 69,3 E C-N
.,-.31,-.5
m
..;13-5,81279 C Ñ(V `i
2=,,,"¡I
' ' C Ú
.É 7006
E1.12,101
2 0
m
° Y<C0
,
1.11Z0
E,-75,-0
,:=45.
o
d.pc`
« J c01i °
a
m v
> m
cya. d L
E'
~m

VoY ca)
>
a)d
= ó°-J.:' °i d ó.z d
~
Ó
O_Sa)
>o _- N>w
Nz d É ".11-1
RN "2, -0-

wó=co.°.
yLLoám
m aalo c
Em pom,.,ciu°om
;'c N. oL-
ó 3 0
a., a,EawO'
°
=zo
N- o aa m

waól`J._
o0 7 E E aN °ó>thmº>'.mmn¢ov
<Óo ÑS'7'X f0'
oÑ
d?ó;N:°

E5810-
-Li_
ÜO]
y2 Z Ú
LL Ó
mO áLLW >Ñ Oa U0
y nLL u. 0 y
1.M0

E
(fl-:
rníoi,óo0` rn c2
E O C
cOiZ á)
á,
'a a a U cSnMI Ocm
101. ó mót-ó' d
I y-p c c
3covñ2o
UU OLíMLÑ
.
áM .

QL C000
óQ1
1 n M a
d dLLÓ
rnÑM Y 0 m
u -
vr--
<E`OLLJ mL
U'111%Iplóc-
I I
1
1

m
IUUÚÚavi1=1=UW11Cj111
. No é_-<nco-n_o_orn_¢ _M N J J
U 55 U 55 UUUU-)l J J J
I
LO (O
J
 94
Fig. 6-22 - VHF EXCITERS AND
Hole
the exciter chassis. layout for AMPLIFIERS
Hole sizes 24i
A-1/8 inch,
are i 134'-
C-3/4 inch, and B-1/4
- 2/t 2'-
inch, }--IVId-1 I'
D-1
Because of variations in
inch.
parts I'---t'-I=-i-I.
,A lA A
-7----4- 1 -A -3/i-{----¡
I

- A
sizes, the builder
should check - Br m e
with his components
before
drilling to these dimensions.
v4S
)K
A
!
2/Z---,=-2/2
Aó
I7/s I ISrl n'Vé-I
"\8 Ibé
BY -AT
- iA
4
B

---
7'"
21/4é--?
i, iv_ .
1-541 A

-f-
Q
i

S I -+A -A
1
I
14i--t-I/is-2Sps
lo'

often running over 120, a

rated at 6,3 volts may filament transformer
doubler plate
much too high for the give as much as 7. This is between Vl and tripler grid coils, L4 and L5, are
was introduced to 250 series tubes, so RFC4 and V2. All are slug
144 -MHz coils are -tuned.
perform a dual role.
drops the heater voltage Mainly, it leads. The supported mainly on their The
to the desired tripler and own
the rf isolation of the level, but tors, C8 amplifier plate
heater circuit it also provides and C9, are alongside -tuning capaci-
certainly does no harm. tubes. They and the their respective
tion is not critical, the Because the choke func- extension crystal switch are driven by
wire size and/or number shafts and couplings.
turns can be varied to give of Liberal
tube terminal, with your 6.0 volts at the 4CX250 mounting use is made of tie-point strips for
average line It nected small parts, where they
is well to measure
this with an ac metervoltage.
of known support directly to other components that will
are not con-
accuracy; vacuum-tube them.
voltmeters are notoriously Disk -ceramic capacitors
inaccurate on ac readings. The choke was not in bypassing in circuits up through 48 are used for
place when the pictures 144 -MHz MHz. The
were made. It is circuits
alongside the tube socket, mounted use button -mica (except the heater circuit of V3)
from the cooling fan, so directly in the air stream as disk or ceramic feedthrough capacitors,
ceramics are unreliable
wire sizes as small as No.heating is no problem with frequencies. at this
28. During initial testing of and higher
Modulation of the transmitter instability the
(the latter is used in for voice or tone in the output stage was traced exciter
to the
presence of rf in
practice transmissions) W1AW bulletin and code- bypass on Pins 4 the heater circuits. The heater
unit of at least 150 watts requires an external audio the and 5 of V3, a 100-pF ceramic of
"dogbone"
L15, connected in the output. The audio choke, series -resonant in variety, with 1/4-inch
screen lead, is shorted by the 144-MHz region. leads, is
when cw of fm is used. S3 out to be a simple and This turned
2000 for cw and 1500 for Maximum plate voltage is effective way of getting the
heater down to ground potential for rf,
works well with voltages a-m, but the amplifier stabilizing the 6360 and thus
voltage is brought in on as low as 750. High stage.
a separate fitting, Heater leads are made with
similar fitting, J7, inverted J8. A power leads can be shielded wire. All
tube, is used to terminate alongside the amplifier mandatory. Heater made this way, though it is not
the high -voltage feed to voltage
the shunt plate choke,
RFC3. brought through the side for the exciter is
former, Ti, in the wall from the trans-
feed -through capacitor, amplifier compartment on
Building the Exciter C20. Plate power comes a
Layout of parts in the exciter via the 4-pin in
connectors JS and P2, which
clear from the should be fairly separate terminals for have
the oscillator
sions for drillingphotographs. The
the top surface ofprincipal dimen- this element can be supplied with screen, so that
3-inch chassis are the 5 by 10 by if desired. regulated voltage
given in Fig. 6-22. The
do well to check builder will
his components for
minor
Adjustment
variations before going ahead The exciter should
with
Controls on the front wall of the drilling. 250 volts from the be tested with no more than
the
left are the crystal chassis at the the builder can play supply. Less c.an be used, and
6360 plate capacitor, selector, Sl, lower,
C9, above it. In the and the 10-watt resistor in
safe by inserting a 5000
the loading capacitor, middle is series with the voltage -ohm
C10, temporarily. This will prevent source
tripler plate capacitor, C8. and at the right is the of malfunction, and it tube damage in case
6-17, we see the crystal On the rear wall, Fig. will protect the supply in
case of a dc short.
cathode jack, J1, in the sockets at- the right, the Start with voltage only
nector, J5, at the left, and center, the power con- on the
and screen, leaving
output connector, J2. below it the coaxial other power leadsoscillator plate
temporarily. Listen for the disconnected
The tubes, V1, V2, on 24, 48, or 144 oscillator on 8 MHz, or
to front in that order. and V3, are lined up back MHz, whichever of
frequencies is available for these
The
doubler grid coils, Ll and L2,oscillator plate and should be a pure crystal receiving. The note
are beside Vl. The
should vary little or none tone, and the frequency
as Ll is tuned. The
value-.y
A Medium -Powered 144 -MHz Transmitter 95
of Cl may require change for some inactive they are adjacent to the plate terminals, 6 and 8,
crystals, though ordinarily the 10 pF specified will respectively, adjusting their position until the
be satisfactory. change in grid current as the plate circuit is tuned
An rf indicator is now needed. This can take is eliminated. Normally the 6360 does not require
many forms. A 2 -volt 60-mA pilot lamp with a neutralization, and the exciter shown here was
1/2 -inch diameter loop soldered across it can be completely stable without it, after the heater
hung over the end of L1, and the core position bypass capacitor, C19, was installed.
adjusted for maximum lamp brilliance. A grid -dip Now apply plate and screen voltage to the
meter in the output -indicating position may be 6360, and connect a dummy load to J2, preferably
coupled to L1 for this test. The latter is preferred, through a power -indicating SWR bridge. Adjust C9
since it provides a check on the frequency of the and C10 for maximum power output, which
output, which should be in the 24 -MHz region, should be 6 to 8 watts, with a 250 -volt supply.
three times the crystal frequency. Now, using a crystal frequency near the middle of
A high -resistance voltmeter or vtvm can be used the range over which you will normally operate,
to measure relative grid voltage at the cold end of tune all adjustments through C8 for maximum
L2. A simple wavemeter (see Fig. 14-3) may be amplifier grid current, and C9 and C10 for best
used to check the approximate frequency, as the output.
grid voltage will dip sharply as the wavemeter is It will be seen that input to the 6AR11
tuned through the oscillator output frequency. pentodes runs only about the rated plate dissi-
Tune the cores in Ll and L2 for maximum drive to pation for the tubes. This makes for long tube life
V1B. and trouble -free operation. The 6360 also operates
Now apply plate and screen voltage to V1B, conservatively, yet its output is adequate to drive
and check similarly for 48 -MHz output. Any of the the final amplifier. The exciter may also be used as
above methods may be used, and in addition we a low -powered transmitter, and it is well -adapted
have provision for measuring tripler grid current to portable work, since its total drain is only a
built into the exciter. Connect a low -range milli - little over 100 mA at 250 volts. The output stage
ammeter from the exposed terminal of C3 to can be modulated with 6 to 10 watts of audio, or
ground, and tune all core studs for maximum grid keyed in J 1 for cw work.
current. This should be about 1 mA, though more
is fine if you can get it. Check with a wavemeter to
Amplifier Construction
be sure that the energy in L4 and LS is on the 6th The final stage is built on a 10 by 12 by 3 -inch
harmonic of the 8 -MHz crystal, and on no other aluminum chassis, which when fastened to the 5 by
frequency. 10 -inch exciter makes a complete 10 by 17 -inch
Next connect the meter from the exposed assembly that can be rack mounted. Our photo-
terminal of C5 and ground, and apply plate and graphs were made before the panel was added, in
screen voltage to V2. Tune C8 for maximum the interest of clarity. The construction is ex-
amplifier grid current, which should be around 2 tremely simple, and with the drawings of the grid
mA. Check with a wavemeter to be sure that the and plate inductances, Fig. 6-23, the builder should
drive is on the 18th harmonic of the 8 -MHz crystal have little trouble in duplicating the original.
frequency. A check on the need for neutralization, Arrangement of parts, other than in the rf circuits,
if any, should now be made. Tune the plate circuit is not important.
of V3 slowly through resonance while watching the The Eimac SK -620A socket has a shield ring
amplifier grid current. There should be no drop in enclosing the screen contacts, a feature that may
grid current as this is done. A downward flicker contribute to the exceptional stability of this
would indicate feedback, which would require amplifier. Other air-system sockets leave the screen
neutralization. This is easily done with a 6360 by ring of the tube exposed, and this has been a factor
soldering 1/2 -inch pieces of insulated wire to the in neutralizing problems encountered with various
grid terminals, Pins 1 and 3. Bend the ends until external -anode tubes of the 150-250 series in the

3j 1-i
38 I/S DRILL

-- ]- 13 -F-
t-
m
u
Í i

SEND 341

IS/¡ 21/2

93,4

Fig. 6-23 -
Details of the grid and plate induc-
tances. The material is flashing copper, though
brass or copper of heavier gauge may be used.
Silver plating is optional. Approximate positions of
coupling loops are shown in broken lines.
 96
VHF EXCITERS AND AMPLIFIERS
past. The push-pull amplifier
for 144 MHz des- along the
cribed elsewhere in this chapter surface that
plates alongside the tube required shield cover. Good electrical makes contact with the
contact is
achieve complete stability.
sockets, in order to and also at the grounded end of theimportant here,
plate
Neutralization of this
amplifier, if needed, is described at the folded -over end of L13 is clamped to line. The
end of this the chassis
section. with a metal strip and two screws and nuts.
The grid circuit is a short strap For effective cooling with a small
blower it
its main portion about one inch of copper, with is important that there be very low air leakage out
away from the of the main chassis, except up
chassis. One end is supported on a through the tube
capacitor, C13, and the other on the grid button-mica socket. To this end, the
holes in the corners of the
of the tube socket. The input coupling terminal chassis, the overlaps at the corners, and all holes
is supported on a tie-point loop, L11, made in mounting the various parts
strip adjacent to the with plastic cement. A screened were sealed
grid line, the loop extending hole in the top
copper strip. underneath the cover allows the warm air to flow out of the
The main portion of the plate compartment directly over the tube. A tight plate
stationary plate of C15 are a single
line and the bottom cover is important for good -fitting
piece of perhaps more than for shielding. cooling,
flashing copper. This is fastened
chassis at the left' end, as viewed in directly to the The built-in bias supply, the audio
choke in the
Fig.
voltage is shunt-fed through RFC3 6-19. Plate screen lead, and the various components
other than
to the tube those in the rf circuits can be
anode. A copper strap wrapped placed almost
supports the blocking capacitor,around C16,
the anode anywhere that suits
the builder's fancy.
bolted to both the plate strap and which is
inductance, L13. At the grounded end of
the plate Firing Up
be seen the series capacitor, C17, L13 may The first step in placing the
and the output is to check the grid amplifier in service
coupling loop, L14. The nature of circuit. Input coupling is best
tant. It must be a transmitting-type C16 is impor- adjusted with a standing-wave bridge
connected in
capable of withstanding heat, high rf capacitor, the line between J2 and J3. A
high voltage. The TV-type current, and be between terminals 3 and milliammeter 4 of
should
often used for this purpose on "doorknob" capacitor amplifier grid current. The object nowPl, to read
is to obtain
lower
definitely not recommended for 144 frequencies is optimum coupling into the amplifier.
-MHz service. Apply power to the exciter,
When the amplifier was placed in
W1AW it was found that vibration service at vates the amplifier bias supply. which also acti-
of the plate line meter disconnected Leave the screen
caused by the blower motor was a for the present, so that there
source of will be no voltage on the amplifier
operational difficulties, so a ceramic
mounted near the middle of L13, to standoff was the exciter tuning and loading screen. Adjust
support it for maximum
more rigidly on the chassis. We amplifier grid current. Now adjust C11 and
plate line components, but silver-plated all the minimum reflected power on the SWR C12 for
measurements made this is not zero, try various bridge. If
carefully before and after show only a perceptible positions of L11 with
improvement from the plating. respect to L12, readjusting their
The movable plate of CIS is a 2 time for lowest reflected power. capacitors each
1/4 The best power
aluminum disk mounted on a 1/4-20 -inch transfer between exciter and amplifier will occur
brass lead this point. at
screw. A matching nut soldered to a
bolted to an aluminum bracket copper to
plate is Adjust the bias control so that the
amplifier
bearing and electrical ground. When the provide a grid current is 10 mA or less,
and apply plate and
place a tension spring can be added panel is in screen voltage to the amplifier. Be sure
slipping it over the brass screw. externally, by amplifier is loaded at all times, to prevent that the
The shunt-feed rf choke, RFC3, screen current. Satisfactory operation excessive
may be seen in possible with plate voltages as low as should be
a horizontal position beside the 700, with 250
tube, level with the volts on the screen. If
top of the anode. Its back end is lower plate voltage is used
highvoltage feedthrough, J7. Underconnected to a for initial testing, the screen
voltage should be
the chassis the dropped also, to keep screen current below
matching portion of J7 is connected to a about
fitting, J8, on the back wall of the similar 30 mA. Keep a 50ohm dummy load
chassis, by J4 at all times, and be sure that connected to
means of coax used in lieu of high C15 and C17 are
-voltage shielded adjusted so that power is being
wire. Another run of coax connects delivered to the
fitting, J4, with the hot end of the the output load. Tube damage is more likely to develop from
loop, L14. output coupling excessive screen dissipation than from
that can happen to the plate in anything
The shield cover for the normal service, so
amplifier is a standard keep a close watch on the
7 by 12-inch chassis, screen meter, and be
notched to pass the shafts of sure that
C15 and C17, and held in place
by wing nuts atop dissipation is kept below 10 watts.
six 3 1/2-inch 6-32 threaded Adjust the position of L14 with
fastened at the corners, and at the
brass rods. These are for maximum output, readjusting respect to L13
the tuning and
each long side, with hex nuts above midpoint of loading capacitors, C15 and C17, with each
change
main chassis surface. If you do your and below the in coupling. The tuning and the position of
work you may be able to make a own metal- coupling loop will change with the
plate line than this; the better shielded voltages, so final adjustment should various plate
dimensions of ours were the plate voltage at the point be made with
dictated by available chassis sizes. where maximum
was polished with emery The main chassis efficiency is desired. If an
paper and steel wool accurate
wattmeter is available, it should indicate bridge or
operating
A 500 -Watt Fm and Cw Transmitter for 220 MHz 97
efficiency in excess of 65 percent. Power output in using voltages above 250, so long as the screen
well over 300 watts was measured at 2000 volts, dissipation is kept low.
and 200 watts at 1500 volts, with inputs of 500 Neutralization can be added, if necessary, as
and 300 watts, respectively. follows: A feed -through bushing (National TPB) is
The amplifier can be run under a wide range of mounted under L13, so that it projects through the
plate and screen voltages, bias and driving power, chassis under L12. A loop of wire about 1/2 inch
so long as none of the maximum ratings for the on a side is connected from the bushing rod to the
various elements is exceeded. With fixed screen chassis, under L12. A brass capacitor plate about
supply, best efficiency will be obtained by juggling 1/2 x 1 inch is soldered to the top, under L13.
the grid bias, checking output meanwhile. Keep the Vary the position of the loop with respect to L12,
final plate current below 250 mA and the screen and the plate with respect to L13, to achieve
current under 30. Screen current will be progres- minimum rf feedthrough, with the exciter running
sively lower as the plate voltage is raised, and may and the amplifier having only heater voltage
even go negative at plate voltages above 1000 or so, applied. Check with a sensitive rf indicator coupled
particularly with low drive. If a separate variable to J3.
screen supply is used, there may be some advantage

A 500 -WATT FM AND CW TRANSMITTER FOR 220 MHZ

The 220 -MHz transmitter of Fig. 6-24 was
designed and built by R. B. Stevens, W1QWJ, and
was first described in May 1969 QST. It is capable
."541 511
of 300 watts output, cw or fm, or the exciter
portion can be used alone to deliver approximately $
8 watts output.
-w
The RF Circuits
Looking at the schematic diagram, Fig. 6-26, it
will be seen that the first three stages of the
transmitter look very much like any vhf trans-
mitter using vacuum tubes. A conventional 6CL6
crystal oscillator, V1, uses 6-, 8-, or 12 -MHz
crystals, multiplying in its plate circuit to 24 MHz
(12-MHz crystals should be the fundamental type.) "
A 6BQ5, V2, triples to 73 MHz, and drives a 2E26
amplifier, V3, straight -through on this frequency.
A variable capacitor, C6, across the crystal, permits
Fig. 6-24 - The 220 -MHz transmitter is set up for
rack mounting on 8-3/4 -inch panel. Meters at the
a small adjustment of the frequency. left can be switched to read driver plate, amplifier
A varactor tripler, driven by the 2E26, is used screen and amplifier plate currents, and amplifier
to get up to 220. Requiring no power supply of its plate voltage.
own, it is capable of more than enough power
output at 220 to drive the 500 -watt amplifier. through some kind of rf bypassing device. Thls and
The output of a varactor multiplier contains the other mechanical features of a good coaxial
harmonics other than the desired one, so a strip -line tank are not readily made with the simpler tools.
filter is connected between the varactor output and Details of the assembly are given in Fig. 6-30.
the final amplifier grid circuit. The filter is a The final grid circuit, visible in Fig. 6-32 along
separate assembly mounted on the end of the with the varactor multiplier and the strip -line filter,
chassis, visible in two of the photographs. Full is a half -wave strip-line. The fan blows cooling air
details of the filter may be found in any edition of into the grid compartment, up through the
this Manual. 4CX250 socket, and out through the end of the
The final amplifier is a 4CX250 series external -
anode tube, with a coaxial tank circuit. The B
version is used here, but the R and F types have
the same mechanical design.
The coaxial plate circuit follows a standard
design. Such a tank has extremely high Q, and the
heavy copper (or brass) construction offers con-
siderable heat sinking. Probably its only disadvan-
tage is the necessity for feeding the high voltage in

Fig. 6-25 - Rear view of the 220 -MHz transmitter.

The exciter stages are on a circuit board in the
foreground. Chassis at the right side houses the
varactor tripler and the amplifier grid circuit. Air
blows into this compartment and out through the
center conductor of the coaxial plate -circuit assem-
bly. r' 3
 osc

TO
v RACTOR
TRIPLER

TO
METER
SWITCH

CRI

MODULATOR

EXCEPT AS INDICATED, DECIMAL VALUES OF

CAPACITANCE ARE IN MICROFARADS 1 yF ) ;
6.3V AC OTHERS ARE IN PICOFARADS (PF OR yyF);
RESISTANCES ARE IN OHMS; k 1000.
LIMITER
ÁEMP .1 ATA
zNss

YIC

Fig. 6-26 -
Schematic diagram and parts information for the W1 QWJ 220 -MHz exciter and frequency
modulator. Capacitors with polarity marked are electrolytic. Components not specified below are
marked for text reference purposes. Cl through C5 are dipped -mica or silver -mica.
C6 - 30-pF miniature trimmer (Johnson 160-130). CR2, CR3 - Any silicon diode (Motorola MV L3 - 7 turns No. 22, 1/2 -inch dia, 3/8 -inch long.
C7, C8 - 20-pF miniature trimmer (Johnson 2105 or similar). Tap 4 turns from grid end.
160-110). J1 - Closed-circuit jack. L4 -5turns No. 16, 1/2 -inch dia, 1 inch long.
C9 - 15-pF variable, double-spaced (Hammarlund J2 - BNC chassis fitting. Y1 -
8150 -kHz crystal, HC -6/U holder preferred.
HF-15-X). L1 - 10 turns No. 22 enamel, closewound on 6112 -kHz or 12223 -kHz fundamental crystal
C10 - 140-pF variable (Hammarlund HF-140). 1/4 -inch slug-tuned form. also usable. Frequencies given are for low-
CR1 - Varicap diode. L2 -4 turns No. 22, 1/2 -inch dia, 7/16 -inch long. frequency end of the band. Use C6 for slight
frequency adjustment.
 A 500 -Watt Fm and Cw Transmitter for 220 MHz 99
T3.SS irl! 2206.N,
J3 IS
LS ..LT CIS

INPUT LB
S
L6 C1146 C
J
146.7-61111
tOL ER
7, TO P
CRS

m GRID

1I Is
x777

Fig. 6-27 - Circuit of the varactor multiplier, 73 to 220 MHz.

C11, C13, C14, C16 -
15-pF miniature variable L6 - 4 turns No. 16, 1/2 -inch dia, 1/2 -inch long.
(Johnson 160-107). Rotor of C11 must be L7 - 3 turns No. 16, 3/8-inch dia, 3/8 -inch long.
C12
insulated from chassis.
- 20-pF miniature variable (Johnson
L8 - 3 turns No. 16, 3/8 -inch dia, 3/8 -inch long,
tapped at 1 turn from grounded end.
160-110).
- CR8 -
Varactor diode (Amperex H4A/1 N4885).
C15 5-pF ceramic.
L5 - 8 turns No. 16, 1/2 -inch dia, 7/8 -inch long.
J3, J4 -
BNC fitting.

tank assembly, by way of the hollow inner it,and this changing capacitance is used to "pull"
conductor, L10. The coaxial output fitting, J6, the the frequency of the crystal oscillator slightly. A
coupling loop, L11, and its series capacitor, C21, good 8 -MHz crystal can be pulled up to 600 Hz in
are mounted on a small detachable plate bent to fit this way, depending on the values of Cl and C6 in
the curvature of the coaxial assembly, and Fig. 6-26. With 27 times frequency multiplication
mounted near the outer end. The varactor triplet is this gives a maximum deviation in excess of 16 kHz
built into the top of the amplifier grid assembly, as at the operating frequency, close to the optimum
seen in Fig. 6-32, above the final grid circuit and for some of the fm receivers currently in use in
the strip -line filter. fixed -frequency service on 6 and 2. Lesser devia-
tion, for work with communications receivers,
Generating the Frequency Modulation most of them having about a 3 -kHz bandwidth
today, is merely a matter of applying less audio.
Where only a small swing at the control
frequency is needed, as in a vhf or uhf transmitter Adjustment and Operation
having a high order of frequency multiplication, This is not intended to be a beginner's project,
the modulation can be applied very easily. A so detailed discussion of the mechanical layout will
voltage -variable capacitor, CR1, changes capaci- be omitted. The mechanical arrangement of the
tance in relation to the audio voltage applied across components could be altered to suit one's own
AMP

+1000 TO
2000V

-
Fig. 6-28 Schematic diagram and parts informa- C22 - Built-in bypass capacitor; see Fig. 6-30.
tion for the 220 -MHz final amplifier. Decimal val-
ues of capacitance are in microfarads
C23 - 500-pF 5-kV or more.
in pF.
(µF); others J6 -N -type fitting.
C17 - 20-pF miniature variable
L9 - Brass strip, 1/16 X 3/8 X 6-1/2 inches. Bolts
(Johnson to grid terminal on socket. Tap C18 7/8 inch
160-110). Stator supports end of L9. from grid.
C18 - 15-pF silver -mica. L10 -
Coaxial line inner conductor; see Fig. 6-30.
C19 - Capacitor built into socket assembly (John- L11 -
Output coupling loop made from 3-1/4
son 124-109-1 socket, with 124-113-1 bypass inches No. 16. Cover with insulating sleeving
ring and 124-111-1 chimney).
C20 -- Disk -type tuning capacitor; see Fig. 6-30.
and bend to 3/4 inch high and 1-3/4 inch long.
See Fig. 6-30.
C21 15-pF miniature variable (Johnson
160-107).
RFC4, RFC5
J5 - -
0.84-µH rf choke (Ohmite Z-235).
BNC fitting.
100 VHF EXCITERS AND AMPLIFIERS
requirements, since the complete transmitter is
made up of many subassemblies. Adjustment for
best results may be strange to anyone who has not
had experience with varactor multipliers.
The first step is to get a good 52 -ohm load. For
the present, it will have to handle a maximum of
about 10 watts. A good SWR bridge is also needed
CR7
for the tests. The first step is to adjust the exciter. -SOS BIAS
Procedure here is like that for any similar lineup of TO P
Dev STAGE
AC 5000
tubes, but the 2E26 must be adjusted for optimum ION
results when working into a 52 -ohm load. Once an
output of 10 to 12 watts is obtained in this way,

Fig. 6-29 - Circuit details of the built-in power

supplies for amplifier bias (lower) and speech
amplifier -modulator (upper) for the 220 -MHz
transmitter. Capacitors with polarity marked are
electrolytic. All diodes are 200 -volt PRV, 1 A. R1
and R2 are approximate values. Select for 12 and
minus 50 volts output, respectively. Capacitance is
in microfarads.

s' leave the tuning of the 2E26 and preceding stages

alone thereafter.
Now connect the SWR bridge output to J3 of
the varactor multiplier, and tune C11 and C12 for
lowest SWR indication. Leave the 2E26 adjust-
ments alone.
Now connect a coaxial cable from J2 to J3, and
connect the bridge or wattmeter in a line from J4
to the dummy load. Adjust C13, C14, and C16 for

L
'
?.--.
e i maximum output at 220 MHz. Adjustments in the
multiplier interlock, and several passes through all
siZ
4' adjustments may be needed for best output. But
. {p4, remember that the 2E26 is set for a 52-ohm load.
Leave it alone, and make the multiplier adjust -
SOLDER ENTIRE
CIRCUMFERENCE

muss

COPPER ql
BRASS PIPE

r.
I.

638

Fig. 6-30 - Details of the coaxial -line plate circuit

of the 220 -MHz transmitter.
L. `DA.

C20

r-1,BRASS
A 500 -Watt Fm and Cw Transmitter for 220 MHz 101
Fig. 6-31 - Looking
underneath the
chassis of the
220-MHz transmit-
ter, we see the qr.

speech amplifier -1
clipper at the upper
right, the exciter cir- -.41114.
cuits across the bot-
tom, power supply
components at the
lower right, and
meter switching,
upper left
r-_ /a

t
F
, L, 1
tit

ments do the job. An indication of some 8 watts or Adjustment of the coupling loop, L11, and the
so of output should result in maximum grid loading capacitor, C21, will be fairly critical when
current in the final amplifier. striving for the absolute maximum output. Follow-
It is likely that getting enough grid current for ing the manufacturer's recommendations as to
the 4CX250B will not be difficult, as the lineup maximum plate voltage and current, 2000 volts at
described gives more than ample drive. Up to 20 250 mA, resulted in about 320 watts output.
mA grid current has been obtained, but not this Raising the plate current to 300 mA, by increasing
much is needed. In fact, with fm or cw operation, the screen voltage, netted 400 watts output. Even
only a slight increase in efficiency is noted after at this input the tube seemed to be operating well
the drive is raised beyond the point where grid cur- and the tank circuit did not indicate excessive
rent begins to flow. heating.

o u

Fig. 6-32 - Looking

into the amplifier
grid compartment.
The varactor tripler
is in the upper left
portion. Below the '
compartment is the
220 -MHz strip -line
filter.
720 Kc r1iTt11
102 VHF EXCITERS AND AMPLIFIERS
AN 829B AMPLIFIER FOR 144 MHz

The dual tetrode known variously as the 829, 3/8 -inch wide strips of copper or brass, to reduce
829B and 3E29 has been a fixture on the vhf scene lead inductance. The .001-µF. capacitors at Pins 3
for many years. Commonly available on the surplus and 5 are returned to Pin 4, using the shortest
market since the end of World War II, it is still one possible leads. The grid coil, L2, is mounted
of the better vhf amplifier tubes in the 100-watt directly on the socket terminals, with the link, Ll,
class. At surplus prices, it is also the cheapest. inserted between turns at the center. A 3 -lug
Inclusion of a rather old 2 -meter amplifier in the terminal strip attached to the rear wall supports Cl
first edition of this manual showed that there is and Ll. A 5 -terminal barrier strip on the outside
still a considerable interest in this tube, so this rear wall is used for power supply connections.
modern version by W10ER, Figs. 6-33 to 6-37, is Coaxial connectors for input and output are on
presented here. It features complete shielding, a opposite sides of the rear of the chassis. A
recessed socket with shield ring, for isolation of the UG106/U shield hood covers the back of J2, to
grid and plate circuits, and a metal strap plate isolate it from J1 and prevent stray coupling
circuit, for improved efficiency. between the input and output. The lead from J2 to
This amplifier was designed specifically for the the feed -through terminal and the high-voltage lead
829 -series tubes, but there are several other types from the barrier strip to its terminal up front are
that could be used, with minor modification of the made with coaxial cable.
design. The 5894 is a more efficient dual tetrode, Details of the plate circuit assembly and top
capable of somewhat more power than the 829, enclosure are given in Fig. 6-37. The top edges of
and requiring less drive. Because of lower input and the plate line, L3, are soldered the full length of
output capacitances, it will require more induc- the stator posts of C2, for minimum stray induc-
tance in L2 and L3. The 832A, a smaller version of tance at this point. The tuning capacitor is sup-
the 829 taking lower power and less drive, is also ported on a plastic mounting block, which has
usable. narrow slots for L3. These can be cut in the plastic
with a keyhole saw, after drilling starting holes at
Construction the top. See detail B. If Teflon of suitable
The amplifier is built on a 3 by 5 by 10 -inch thickness is available, it would be ideal for this
aluminum chassis, with an aluminum cage on top, support, as it is impervious to heat of the order
9 1/2 inches long, 4 inches wide and 4 1/2 inches encountered here. Plexiglas and other clear plastics
high. Holes in the sides and rear of the top are usable.
compartment, at the tube end, allow for air Teflon shafting would also be best for the rod
circulation. The cover is perforated aluminum, that is to run from C2 out through the front panel.
permitting the heat to rise from the tube, as cool Wood dowelling is also suitable. Do not use metal
air moves in from the side holes. stock, as it would be closely -coupled to L3. The
The 829B socket is an E. F. Johnson Type rotor of C2 must be isolated from ground.
122-101, designed for recessed mounting. Leads The low-impedance end of L3 is supported on a
1 -inch ceramic pillar. Mount a No. 6 spade bolt at
from the socket terminals 1, 4, and 7 to ground are
the exact center of the U bend in L3, thread the
standoff onto this, and then bolt the bottom of the
insulator in place. The coupling loop, L4, is
supported on the stator post of C3 and the
feed -through bushing to which the coax to J2 is
connected, on the underside of the chassis. C3 is
on the front wall of the shield enclosure, so L4 is
soldered to it after the cover is in place.
The plate line was made of sheet brass, and
then silver plated. Flashing copper will work
equally well. If not plated, it should be polished
thoroughly, and then coated with clear lacquer to
reduce tarnishing. The lacquering should be done
only after the assembly job is complete. It will be
seen from detail C, Fig. 6-37, that there are two
strips of thinner stock bolted to the ends of the
4 stiff material of L3. Holes for these bolts should be
larger than needed, so that the line, the straps, and
the Fahnstock clips for the plate connections can
be assembled loosely at first, then tightened in a

Fig. 6-33 - The 829-B amplifier, with its shield

cover in place. Air circulation is provided by the
screened holes and cover.
An 829B Amplifier for 144 MHz 103

4011

ia

Fig. 6-34 - Top of the amplifier chassis, as seen from the rear with the shield cage removed. The
output link with its black spaghetti tubing is just below the U-shaped plate tank inductor. The loading
control, C3, is mounted on the shield cage and is not shown here.

i L
(r_ :.,
9'' 1
r.
N't

o
Fig. 6-35 - Looking into the bottom of the chassis. The feedthrough bushings for plate power and rf
output are at the left. Coax cable is used for the high voltage dc lead. Wide copper straps ground the
filament and cathode pins of the tube socket. A hood over the back of J2, lower right, helps isolate
the input from the output.
104 VHF EXCITERS AND AMPLIFIERS
position such that no strain in placed on the tube Screen voltage should be regulated, in linear
plate pins. Be sure that the tube is seated properly service, either a-m or ssb. For cw, fm, or high-level
in the socket before the final tightening of the line a-m, the screen can be supplied through a dropping
assembly. resistor from the plate voltage source. The value
will depend on many factors, but should be about
Adjustment and Use 10,000 ohms at low plate voltages, rising to 35,000
The amplifier may be driven in Class -C service at the high end of the range. Grid bias may also
with any exciter delivering 3 to 10 watts output. vary, and it may be obtained from a bias supply, or
Operating conditions and maximum plate voltages from a grid resistor (connected between RFC1 and
for cw, a -m, and ssb service are given below. The Terminal 3 of the barrier strip) or both. In ssb or
82913 works well at lower plate voltages, and is a -m linear service, it preferably should be regulated
often operated at about 400 to 450 volts in vhf and adjustable.
applications. The maximum plate current at 450 Any tetrode amplifier can be run under widely
volts is 200 mA, and this amplifier delivers about varying conditions, so it can be adjusted to give
55 watts output this way. A suitable supply for optimum results with the power supplies you may
this voltage level can be made with a TV receiver have available, for modes of emission you are most
power transformer. interested in. The "typical operating conditions"
Many amateurs look for a linear amplifier that listed in tube tables are guidelines, not laws, But
can be used with the small a -m transmitters when the tables say "Maximum Ratings," they mean
commonly used in vhf communication. This ampli- it.
fier will operate as a linear, but unless the exciter is To adjust the 829B amplfier, apply heater
very low -powered the step-up may not be attrac- voltage, and then connect the exciter to J1.
tive. Output in a -m linear service is no more than Connect a milliammeter between Terminals 3 and
half the maximum safe plate dissipation for the 5, and turn on the exciter, noting the grid current.
tube used. This means that an 82913 linear is Adjust the position of Ll with respect to L2, and
limited to about 15 watts output on a -m which the turn spacing of L2, for maximum grid current.
may be good enough for use with a 1 -watt Now tune the plate circuit slowly through its
transistor rig, but not very attractive at higher range, watching the grid current. There may be a
levels of exciter power. For more on linear slight rise at resonance, but no downward dip. The
amplifiers, their uses and limitations, see the latter would indicate need for neutralization,
preceding chapter, and "Tips on Linears" in this which was not required in this version. Grid
one. current should run 7 to 12 mA for Class -C service.

AMPLIFIER
144-148 Mc. Ly 144 -148 M.
L2
J2
JI
LI
OUTPUT
INPUT

RFC,
TB,
PLATE
SCREEN
GRID
EXCEPT AS INDICATED, DECIMAL VALUES OF 829-8
CAPACITANCE ARE IN MICROFARADS HEATER
OTHERS ARE IN PICOFARAOS (PI. ORpuplk GROUND
RESISTANCES ARE IN OHMS, K 1000.

Fig. 6-36 - Schematic diagram of the 2 -meter amplifier

L1 -2 turns No. 22 insulated hookup wire in

Cl - 27-pF silver mica. center of L2.
-
C2 - 18 pF per section, butterfly variable (E. F.
Johnson 167-22 with 3 stator plates removed
L2 5 turns No. 20 tinned wire, 5/16 -inch
diameter 1/2 inch long (see text).
from each side. Also, two rotor plates are L3 -
Plate inductor. See Fig. 6-37 for dimensions.
removed). L4 -6-inch length of No. 12 enam. wire bent into
C3 --50-pF variable (Millen 200501. a U with -1/4 -inch spacing between sides
1

C4 0.001-µF transmitting ceramic (Centralab (cover with spaghetti tubing).

858S). RFC1, RFC2 -
2.7-41H choke (Millen 34300-2.71.
C5, C6
--
0.001-1IF 1000-volt disk.
J1, J2 SO -239 connector. TB1
-
RFC3 0.841H rf choke (Millen 34300-.821.
-5 -terminal barrier strip (Millen 37305).
An 829B Amplifier for 144 MHz 105
ALUM. ANGLE
TO HOLD COVER.

O Di DIA. (ALIGN: WITH

f"-\
ALUM.
41/2 STRIP C2 SHAFT I.
INSIDE L5
s/16 DIA.

NO.6 1-1yi--1
21/2' DIA. `. 1O.6 SPADE /s HOLES
(3) BOLTS.
ALUM. SCREEN
(USE FOR TOP COVER TOO).

NO.6
HOLE
3/e DIA.
THIN BRASS
PLEXIGLASS OR COPPER NO.10 GAUGE
('/pTHICK OR STRIPS. (OR THICKER)
GREATER). COPPER OR BRASS.
(C)
(B)

L ALUM.
BRACKET.

Fig. 6-37 -A -
General layout of the shield box is shown at A. The box is made from No. 16 gauge
aluminum stock. B -
Details of the mounting block which supports C2 and L3. C
L3 and its connecting strips (see text).
Dimensions for -

It may be more in the static condition, as It will A lamp load may be connected across J2 for a
chop some when the amplifier is actually running, rough indication of power output, though a good
and loaded. dummy load and a power-indicating watt-meter or
If neutralization is needed, run wires from the SWR bridge is much to be preferred. Apply plate
grid terminals of the socket up to the top of the and screen power, tune C2 and C3 for maximum
chassis on feed -through bushings, and then bring indication, and then adjust the position of L4 with
wires up alongside the tube envelope adjacent to respect to L3 carefully, retuning each time the
each plate. The wires are crossed over under the loop is moved. Coupling should be the loosest that
chassis, and the desired feedback is obtained by will give satisfactory power transfer. The lamp load
varying the position of the top wires with respect will be of no value in the adjustment, as it
to the tube plates. represents a load of far different impedance than
will be used ultimately with the transmitter.

829-B Operating Conditions

Service Esc jsc Eg Ig

EP Ip
(Max.) (minus)
Class C cw- 750 V 160 mA 200 V 17 mA 50V 7mA
Class C -a
-m 600 V 150 mA 200 V 16 mA 60V 7mA
Class AB1
no sig.
ssb - 600 V 110 mA 200 V 26 mA 18 V 0
40 mA (reg.) 4 mA

Maximum plate input for a -m linear: about 40 watts, Class AB1

 106 VHF EXCITERS AND AMPLIFIERS
KILOWATT AMPLIFIERS FOR 50 AND 144 MHZ
The amplifiers shown in Fig. 6-38 were de- 4CX300A, 8122 and others, but differences in
signed for versatility. Though capable of running at tube capacitance might require modification of the
the maximum legal power for amateur stations, circuit elements described here. The air-system
they operate efficiently at much lower levels. They sockets (required for all external -anode tubes
work well as linears, for use with a -m or ssb, or mentioned) may be the same for all types in the
they can be modulated or keyed in high -efficiency second paragraph, but those just above require
Class-C service. Though the tube type shown is different sockets.
expensive when purchased new, an effective substi- Both amplifiers take a kilowatt on cw or ssb
tute is commonly available on the surplus market with ease. The 144 -MHz model must be held to
at much lower cost. Operated as a rack -mounted 600 watts input for plate -modulated service to stay
pair, as pictured, the amplifiers offer convenient within the manufacturer's ratings. On SO MHz the
band -changing from 50 to 144 MHz, merely by three tubes in parallel loaf along at 1000 watts in
snapping on the appropriate heater voltage switch, the low -duty -cycle modes. The permissible input
and changing the air connection from one to the on a -m phone is 900 watts. Class C efficiency is on
other. the order of 75 per cent, over a wide range of plate
The external -anode type of transmitting tube voltages. It is possible to run all the way from 800
has many variations. The family originated with to 2000 volts on the amplifier plates without
the 4X1SOA many years ago, and tubes of the altering screen voltage or drive levels appreciably.
early type are still available, and widely used. A
later version, with improved cooling, is the
4X250B, capable of higher power but otherwise Mechanical Layout
very similar to the 4X150A. More recently the The amplifiers are similar packages, to mount
insulation was changed from glass to ceramic, and together harmoniously, though this is of only
the prefix became 4CX. All the general types thus incidental interest to the fellow concerned with
far mentioned were made with variations in basing one band or the other. They are built in standard 4
and heater voltage that will be apparent to any by 10 by 17 -inch aluminum chassis, mounted open
reader of tube catalogs. The 4CX250R used here is side up and fitted with shield covers. In the
a special rugged version, otherwise very similar to author's station a single blower is used for all
the 4CX250B, and interchangeable with it for transmitters. This explains the air-intake sleeve
amateur purposes. Similar types are supplied by seen on the back of each amplifier. An air hose
other makers as the 7034/4X150A 7203/4CX250B from the remote blower is pushed into the ampli-
and 7580. There is another version for linear- fier being used.
amplifier service only, called the 4CX350A. The transmitters are all hooked up together, to
If one then goes to other basing arrangements meters, power circuits, audio equipment and power
similar power capabilities may be found in the supplies common to all. Changing bands involves

Fig. 6-38 - The kilowatt amplifiers

for 50 and 144 MHz in a rack made
from aluminum angle stock. At the
bottom is a meter panel with controls
for meter and mode switching.

sit

s':
Kilowatt Amplifiers for 50 and 144 MHz 107
4Cx25OR, J2

t23456 781J3

-'T6.0 VOLTS

1-
51/1 I15 VOLTS

t^ l i.

(1 2 3 4 5 6 7 8 IP2
TO GRID METER 1 I

AND BIAS t t I 115V, A C.

TO SCREEN
METE R, SW ITCH, AND
REGULATED 250 OR 350 VOLTS

Fig. 6-39 - Schematic diagram and parts information for the 50 -MHz amplifier.
C1- 100-pF miniature trimmer (Hammarlund Strip jacket and braid back about 4 inches.
Insert between center turns of L2.
MAPC-1001.
C2 - 35-pF per section split -stator (Hammarlund L2 -
8 turns No. 14, 5/8-inch dia, 1-1/4 inches
HF D -35X). long, center tapped.
C3 - -
Neutralizing capacitance see text. L3 -3 turns 2 inches dia, 3 inches long, 1/4 -inch
C4, C5, C11 - 500-pF 5000 -volt transmitting capa-
P1
copper tubing.
-
High -voltage power connector, male (half of
citor (Centralab 858S-500).
C6 - Tuning capacitor made from 3 -inch alum-
-
inum disks see text and Fig. 6-40. P2
Millen 37501).
-8
-pin cable connector to match J3, female.
C7 - 200-pF variable, .03 -inch spacing (Johnson R1 -20 -ohm 10 -watt slider -type resistor. Set so
167-12 or 200L15). that heater voltage is 6.0 at socket.
C8, C9, C10 - -.001-µF disk ceramic. R2, R3, R4 -150 -ohm 1/2-watt resistor. Connect
C12, C13, C14 Bypass built into special air - at socket screen terminal.
system socket. - RFC1 -No. 32 enamel wire closewound full
- Green -jewel pilot lamp holder.
J1, J2 - Coaxial chassis receptacle.
11 length of 1 -watt resistor, 10,000 ohms or
higher.
J3 -8 -pin male power fitting. RFC2 -
No. 28 dsc or enamel,wound 1-3/4 inch
J4 - High -voltage power connector female (half of on 1/2 -inch Teflon rod. Space turns 1 wire dia
Millen 37501). for 8.3 µH. For winding information see Chap-
L1 - turn insulated wire about -inch dia. Make
1 1

from inner conductor of coax running to J1. S1

ter 16.
--
Spst toggle.
T1 6.3-volt 8-A. Adjust R1 to give 6.0 volts.

mainly the switching on of the desired heater locations are slightly different. No attempt was
circuits, and the insertion of the air hose in the made to achieve symmetry through mechanical
proper intake sleeve. Separate antenna relays are gadgetry, since the unbalance of the front panels is
provided for each final stage, and power switching not unpleasing. The rack shown in Fig. 6-38 was
and plugging and unplugging are largely eliminated. made up from aluminum angle stock to fit the job.
Tube sockets are the air -system type, mounted Several screen and bias control arrangements were
on 4 -inch high partitions with folded-over edges tried before the circuit shown in Fig. 6-43 was
that are drawn up tightly to the top, bottom, front settled upon. Meters read driver plate current, and
and back of the chassis with self-tapping screws. amplifier grid, screen and plate currents. Switches
Air is fed into the grid compartments at the left enable the operator to check the grid and screen
side, as viewed from the front. Its only path is currents to each tube in the 144 -MHz amplifier
through the sockets and tube anodes, and out separately, and the screen currents in the 50 -MHz
through screened holes in the right side of the amplifier likewise. A mode switch provides proper
chassis. Panels are standard 5 1/2 -inch aluminum. screen operating conditions for a -m, linear, or cw
Controls for the amplifiers are similar, though their service.
 111 VHF EXCITERS AND AMPLIFIERS
Fig. 6-40 -
Interior of
the 50-MHz amplifier.
Note method of para-
lleling grid and plate
N connections.
AIN

1,1

shielded wire. Where the high voltage comes into

The 50 -MHz Amplifier the plate compartment it is bypassed at the
feed -through fitting.
The use of three tubes in parallel in the 50 -MHz 5) The plate circuit is made entirely of copper
amplifier was an experiment, tried with the ex- strap and tubing, for highest possible Q and low
pectation that parasitics, unbalance, excessive tank resistance losses. It may be of interest that the
circuit heating and all manner of troubles would entire tank circuit was silver-plated after the
develop. These problems never materialized; use of photographs were made. Efficiency measurements
paralleled tubes seemed to introduce no problems made carefully before and after plating showed
on its own, and extensive experience with the identical results.
amplifier has confirmed the worth of the idea. This Looking at the interior view, Fig. 6-40, we see
happy state of affairs Involves a few basic consid- the grid compartment at the left. The coaxial input
erations that should be stated here. fitting, J1 in Fig. 6-39, is in the upper left corner
1) Paralleling straps in the grid and plate of the picture. Coax runs from this, out of sight on
circuits were made "three of a kind." The two the left wall, terminating in a loop, L1, made from
going to the outer grids were bent identically, and its inner conductor. This is inserted between turns
then the one for the middle tube was bent back on at the center of the grid coil, L2. The series
itself as necessary to use the same total length of capacitor, Cl, is just visible on the left chassis wall.
strap. The same was done in the plate circuit. It is not particularly critical in adjustment, so no
2) The grid circuit was split -stator tuned, to inconvenience results from its location away from
get a reasonably -sized grid coil, even with the the front panel.
combined input capacitance of the three tubes plus Screen voltage, bias, and 115 volts ac come
circuit capacitance - some 60 pF or more. This through an 8 -pin fitting, J3, mounted between the
also provided a means for easy neutralization. air intake and the heater transformer, T1. On the
3) The pi-network plate circuit is tuned with a front panel are the heater switch, Si, and the
handmade disk capacitor. This has a far lower pilot -lamp holder.
minimum C than the more conventional tuning The three air -system sockets (Eimac SK -600,
capacitor, and it is devoid of the side bars and SK-6620, SK-630, Johnson 124-110-1 or 124-115-1,
multiple ground paths that are so often the cause with chimneys) are centered on the partition,
of parasitics in vhf amplifiers. No parasitic reson- spaced so that there is about 1/4 inch between
ances were found in this amplifier, other than one their flanges. The small angle brackets that come
around 100 MHz introduced apparently by the rf with the sockets should be tightened down with
choke. This caused a blowup when grid -plate their inner ends bearing against the ceramic chim-
feedback developed with a similar choke In the gnd neys, to hold them in place. Note that the
circuit. The problem was solved easily by use of a 150 -ohm isolating resistors R2, R3, and R4 are
low -Q choke of different inductance in the grid connected right at the screen terminals.
circuit. Do not use a high -quality rf choke for Both grid and plate straps are cut from flashing
RFC 1 ! copper 5/8-inch wide. Lengths are not critical,
4) All power leads except the high-voltage one except that all grid straps should be the same
are in the grid compartment, and made with length, and all plate straps identical. The plate
The 50 -MHz Amplifier 109
straps are made in two pieces soldered together in the past. Maybe one would have handled the job,
T shape, to wrap around the anode and join at the but two do for sure.
coupling capacitors, C4 and C5. These T-shaped
Some Possible Variations
connections could be cut from a sheet of copper in
one piece, with a little planning. It is always risky to suggest variations on a
The copper -tubing plate coil, L3, is mounted on design unless they have been checked out in use, as
stand-off insulators not visible in the picture. bugs may develop in unforeseen ways. The follow-
Connections to the coupling capacitors, the tuning ing are ideas only, to be used at the builder's risk,
capacitor, C6, and the loading capacitor, C7, are since they have not been tested by the designer.
made with copper strap. It will be seen that these You might not care for three tubes in parallel.
various pieces are bolted together, but the straps Two should work well, handling a kilowatt except
were also soldered. The connection from C7 to the in a -m linear or plate -modulated service. Many
output fitting, J2, is a single strap of copper, builders report success with 2 tubes.
bolted and soldered to L3. For those who can afford it, a vacuum variable
The disk tuning capacitor can be made in capacitor should be ideal for C6. One with about
several ways. Flashing copper is easy to work, and 10 pF maximum capacitance should do nicely.
the 144 -MHz capacitor was made of this material. For lower tube cost, 4X150As from surplus
A more sturdy disk can be made from 1/8 -inch should work without mechanical changes. Use
aluminum. Those shown in Fig. 6-40 were 3 -inch plenty of air, if you intend to push the ratings of
meter cutouts from an aluminum panel. Disk -type the 150As. A 100-cfm blower is not too much. The
neutralizing capacitors (if you can find them; ability of the anode structure to withstand heat is
they're not common catalog items these days) the main difference between the 150A and later
provide ready-made disks and lead screw for versions of this tube, and some people have gotten
tuning. For the latter we used 3 -inch 1/4-20 brass away with 250 ratings with 150-type tubes. In this
screws from a neighborhood hardware store. A connection, the 50 -MHz amplifier will take a
panel bushing with brass nuts soldered to it kilowatt at 1200 to 1500 volts, if your power
provided the lead-screw sleeve. The stationary disk supply will handle the current. This approach plus
is supported on 1/2 -inch-diameter Teflon rod, a plenty of air, is preferable to using plate voltages
material also used for the rf choke form. Teflon much in excess of the 4X150A ratings.
works easily and is unexcelled for insulating
applications where high temperatures are encoun- The 144 -MHz Plumber's Special
tered. We found it reasonably priced, in various
diameters, at a local plastics house. Use of 1 5/8 -inch copper tubing for a 2 -meter
The plate rf choke, RFC2, is important. You'll tank circuit is by no means new.* We simply went
probably have to make it to get one of sufficiently one step further and made the entire circuit from
good quality. For more on this see information standard plumbing components. All the heavy
under Fig. 6-39 and "RF Chokes for the VHF metal you see in the plate compartment of Fig.
Bands," Chapter 16. Two coupling capacitors were 6-41 came from the plumbing counter of the local
paralleled because we've experienced trouble with * "High-Efficiency 2 -Meter Kilowatt," QST,
Feb. 1960, p. 30. "Top Efficiency at 144 Mc. with
exploding capacitors in pi -network plate circuits in 4X250Bs," Breyfogle, QST, Dec., 1961, p. 44.

Fig. 6-41 -Interior of the

144 -MHz amplifier, showing
the plate circuit made from
standard plumbing compon-
ents. Brass pipe junctions
make connection to the an-
odes, and T fittings are modi-
fied to form the short at the
end of the line.
'.

e
;
 110 VHF EXCITERS AND AMPLIFIERS
4CX 250 R's

RFC,

J4

IP
TO HIGH VOLTAGE

TO GRID METER, ( i 2 3 4 5 6 7 8 1 P2
SWITCH AND 1, .i-,
( 4 4
TO SCREEN
1 1
TO 115v -A.0
BIAS
M ErER,SWITCN, AND
REGULATED 250 OR 350 VOLTS

Fig. 6-42 - Schematic diagram and parts information for the 144-MHz amplifier.

Cl - 5-pF differential trimmer (Johnson 160-303

or 6MA11).
about 3/4 -inch dia. Remove jacket and braid
about 3 inches. Adjust position with respect to
C2 - 15-pF per section split -stator (Hammarlund L1, L2 for maximum grid current.
C3
HFD-15X). Leave rotor ungrounded.
- 30-pF miniature trimmer (Hammarlund
L4 -
Plate line 1 -5/8 -inch copper pipe, with
junctions and T fittings. Exposed portion of
MAC-30).
C4 - Tuning capacitor made with 3-inch disks. See
text and Fig. 6-41.
pipe is 8 inches long. Cut right end of T fittings
to 1/4 -inch shoulder, and joined ends to 3/8 -
C5 - 3 -inch disk movable with respect to L4. See L5
inch shoulders.
-
1/2 -inch strap of flashing copper, U portion 4
text and Fig. 6-41.
C6
C7
-- 50-pF variable (Hammarlund MC -50).
500-pF 5000 -volt (Centralab 858S-500).
inches long and 1-1/4 inch wide. Make loop and
connections from single piece. Support L4 and
C8, C9 -
socket.
Bypass capacitor built into air -system P1
L5 on standoffs of ceramic or Teflon.
-
High -voltage connector, male (half of Millen
11 - Green -jewel pilot lamp holder.
37501).
-8
J1, J2
J3 -8
- Coaxial chassis receptacle.
-pin male chassis connector.
P2
R1
-pin female cable connector to match J3.
-
20 -ohm 10 -watt slider -type. Adjust for 6.0
-Millen volts at socket.

- Spst toggle.-
J4 High -voltage power connector, female (half of R2, R3, R4, R5 150 -ohm 1/2 -watt resistor.
37501).
L1, L2 - 3-1/2 turns No. 14, 5/8 -inch dia, turns
S1
T1 -
6.3 volt 8 A. Adjust R1 for 6.0 volts.
spaced 1/2 -inch. R2 and R3 tap on about
turn in from grid end. text.
1 RFC1 - 2.15 µH rf choke. No. 22 enamel
L3 - -turn inner conductor of coax from J1,
See closewound 1-3/16 inch on 1/4 -inch Teflon
1
rod.

Sears store. The picture and Fig. 6-41 should be open ends of the T fittings are cut down to about
largely self-explanatory. 1/4 -inch in length.
At the tube end of the plate line, L4 in Fig. The last instruction and the information about
6-42, we have brass castings normally used to join the plate line given under Fig. 6-42 apply only if
sections of the copper pipe. They make a nice the fittings are identical to those obtained by the
sliding fit over the tube anodes. For tighter fit, cut builder. Since there are several types of fittings
thin brass shim stock and insert as much as needed available from plumbing supply houses, the follow-
between the anode and the sleeve. The end of the ing overall dimensions should be heeded: tube end
fitting can be slotted and then clamped firm on the of the plate line to center -line of short - 10 3/8
anode with a hose clamp, as an alternative. The
short at the B -plus end of the line is made with two
inches; spacing of pipes center to center 3 1/2 -
inches.
T fittings, with their flanges cut down to 1/2 inch In using tube types other than those specified,
and slipped over a short section of the pipe that is it may be that some change in plate circuit
not visible. Joints throughout the assembly were inductance will be needed. A simple check will
silver -soldered with a torch, but conventional show if this is needed. Slip the castings and pipe
soldering should do equally well. The flanges at the together without soldering, and assemble the plate
The 144 -MHz Plumber's Special 111
circuit temporarily. Check the tuning range by C3 on the side wall, just under Cl, to make it more
means of a grid-dip meter. No plate or heater readily adjustable. Note that the rotor of C2 is
voltage is needed for this rough check, but it is well ungrounded.
to have the coupling loop in place, and a 50 -ohm
resistor connected across J2. About Neutralization
The coupling loop, L5, is cut from a single These amplifiers were tested without neutraliza-
piece of flashing copper 1/2 inch wide. This tion and we almost got away with it, but use of all
delivered slightly more output to the load than was modes, particularly a -m linear and ssb, imposes
obtained with loops of wire of various lengths strict requirements on stability. Conventional
tried. The loop should be positioned so that the cross-over neutralization employed in the 144 -MHz
bottom edge is approximately flush with the amplifier is omitted from Fig. 6-42 in the interests
bottom of the pipes. Optimum coupling to a of clarity. The schematic representation, C3 in Fig.
50 -ohm load is achieved when the closed end of 6-39, is not very informative either.
the "U" is about 1/4 inch lower than the open end. In the 50-MHz amplifier the lead visible in Fig.
Looking down at the plate -line assembly, the 6-40, attached to the rear stator terminal of C2,
coupling loop is centered between the pipes. runs to a polystyrene feedthrough bushing (Nat-
The loop and plate line are supported on Teflon ional TPB) mounted in the partition between the
rod insulators. The rf choke is also wound on rear and middle sockets. Even, this bushing's wire
Teflon. Note its position outside the U of the plate stub projecting into the plate compartment turned
line. First mounted inside the loop, it went up in a out to be too much "C3" and it was trimmed off
furious burst of smoke when high power was 1/16th inch at a time, until minimum feedthrough
applied to the amplifier. was indicated on a wavemeter coupled to L3 and
Our tuning disks are 3 -inch sheets of flashing tuned to the driving frequency.
copper. For nicer appearance and better mechan- Similar feedthrough bushings are used in the
ical stability, use 1/8 -inch aluminum as In the 144 -MHz amplifier, but here a small wire had to be
50 -MHz model. Three-inch brass 1/4-20 screws are added to each one. The wire connected to the grid
threaded through the pipe fittings. The rear one is of the front tube is aimed toward the anode of the
held in place with a lock nut, and the other Is rear tube, and vice versa. Small sheets of this brass
rotated by the tuning knob, a bakelite shaft or copper should be fastened under the adjacent
coupling, and a length of 1/4 -inch Teflon rod edges of the sockets, and bent up at right angles to
running in a panel bushing. the partition. These 3/4 -inch high barriers act to
A third disk is mounted adjacent to the rear shield the screen rings of the tubes from the
portion of the tank circuit. Its position is adjusted feedback "capacitors" and assure that the coupling
to achieve perfect balance in the tank circuit, but is from grid to opposite plate, and not to the
in practice this turned out to have no measurable screen.* Length and position of the feedback wires
effect. It is felt that a really good choke at RFC1, are adjusted for minimum feedthrough of driver
and careful adjustment of Cl, can practically energy to the plate circuit, as described above.
eliminate the effect of any slight unbalance if the About a half inch of wire was needed in addition
point of connection of RFC1 to the tank circuit is to the terminal stub in this case.
not bypassed to ground. When used as linear amplifiers the tubes must
The 144 -MHz grid circuit, L1L2, looks like two be biased to permit them to draw considerable
coils, but actually is a coiled -up half-wave line. This plate current with no drive, so perfect neutraliza-
is somewhat more compact than a half-wave line tion is a "must." Properly neutralized, the ampli-
with its conductors out straight, and it seems fiers will be stable when run at or near maximum
equally effective. The grids are connected to the safe plate dissipation with no drive, even when the
outer ends and the tuning capacitor to the inner. grid and plate circuits are swung through their
The point of connection of the bias -feed resistors entire ranges. If they will not pass this test the
should be 'determined in the same way as with the amplifiers are not ready to be used for linear
usual half-wave line: by coupling in 144-MHz service.
energy and touching a pencil lead along the Controls and Metering
inductance while watching the grid current. The everyone who builds his own equip-
correct point for final connection of the resistors is Almost
reaction on grid current is ment has a favored way of controlling it, so the
that at which no
schematically in Fig. 6-43 may not
observed. Isolating resistors here, and for feeding system shown
suit everyone. It is for use in a station where power
screen voltage to the sockets, are preferable to rf
The inner conductor of the coaxial line is supplies are actuated by closing the primary
chokes.
circuits to all that the operator wants to have come
used to make the coupling loop, L3, which is
on for transmitting purposes. They are mounted
placed between the inner ends of the grid circuit.
away from the transmitting position, and a cable
Balanced drive is maintained by adjustment of carries the various voltages to the rf position. At
the differential capacitor, Cl, connected in parallel the left, JI, J3, J4, and J5 are terminals carrying all
with C2, and mounted on the side of the chassis voltages from the power -supply position. These are
adjacent to it. The series capacitor, C3, is out of distributed through meters, controls and output
sight under the tuning capacitor, which is mounted fittings, J6, J7, and J8, to various transmitters.
on standoff insulators. It is adjusted by inserting a * Alrsystem sockets are now available with
small screwdriver in a hole in the side of the built-in shielding of the screen ring. The Eimac
chassis, but if we were doing it again we'd mount numbers are SK -620 and 630.
 112 VHF EXCITERS AND ANi?LIFIERS
J.
PI JI 0 f00 5014c 144 Me OTHER
100

1
-- JA J7 Je

_,w+51 LI

-
-C, 50 TO Jy
TO

H
1
100
90 Y. 2 2 0-SO 10 SOMA 2 2

TO 100 LI 3 3 3
3 3

-h
+500V. 2000 A I
4 4 4
DRIVER
RI

+ 25W 100 100 4S3a O'IO 5iO 5 5 5

PLATE (---- 4
-003 -`o--. 6 6 6
SUPPLY IOO
C 7 ___- 7
DRIVER 5 5 LIN
PLATE 8 8
CIRCUIT
6 6 l

.8
AM
7
+C 53A
-H V -.8 o
-IN Fig. 6-43 - Schematic diagram
15- TO PINS
6 AND 7, JI
and parts information for the
A C J3
J.,JzJe control unit used with the vhf
amplifiers. Resistors are 1/2-watt
J4 composition, unless specified;
values in ohms.
JS

J1 -8 -pin male power connector. R1 - 2000 -ohm 25 -watt resistor. Value may be
-
J2, J9, J10, J11 Tip jack. reduced to as low as 1000 ohms if regulation at
-
J3 Ac connector, male. high values of screen current is desired, pro-
-
J4, J5 High -voltage feedthrough connector (Mil- vided current measured in J10 and J11 does not
len 37501). exceed 40 mA under low -screen -current con-
J6, J7, J8 -8
-pin female power connector. ditions.
L1 - 10-H 50-mA choke. Must be shorted out for S1 -
Single -pole 2 -position switch.
other than plate -modulated service.
-8 -
S2 Single -pole 3 -position switch.
P1 -pin female cable plug. -
S3 Double -pole 3 -position switch.

Circuit breakers at the supply position are used to the regulator tube strings can be measured between
turn everything off when the station is closed J10 or .111 and ground.
down.
Adjustable bias, 50 to 90 volts negative, is Operation
brought in through Pin 2 to a 50-mA meter and
Because a variety of tubes may be used, with a
appropriate shunts that keep the circuit that is not
wide range of conditions as to plate voltage and
being metered closed. The switch SI enables the
drive, we're not going to be too specific here. If
operator to read the grid currents separately in the
144 -MHz amplifier. Grid voltage may be read when
you follow the tube manufacturer's recommen-
dations for the plate voltage you intend to use you
required, at J2.
Similarly, a 500 -volt positive source is con-
won't be far wrong. All tubes of this class are quite
versatile as to drive level and plate voltage; unless
nected through Pin 3, a voltage -regulating system,
an audio choke, a 100-mA meter and a 3 -position you are running close the maximum plate -input
switch, S2, to the screens. Currents can be read ratings the principal factor to watch is screen
separately here, too, and this facility is important dissipation, as far as safety of the tubes is
in determining that all tubes are running within concerned. Set up your amplifier with a dummy
load and then try the various conditions given in
ratings. The VR system is switched by S3A to
tube data sheets, observing the operation on all
provide regulated 250 or 350 volts to the screens.
Ganged to it is S3B, which shorts the audio choke
meters. In this way you'll soon learn your way
for all modes except plate-modulated a-m. This around. A few words of preliminary advice may,
must be done, as the choke will cause trouble on however, be ín order.
the other modes. The series -parallel VR-tube bank First, don't feel that you have to run a kilowatt
is by no means an ideal regulating system, but it
right off the bat. Put a Variac in your final plate
prevents soaring of the screen voltage under con- supply primary and run the voltage down for initial
ditions of low or negátive screen current. These testing, or use a lower-voltage supply until you
become familiar with the way the rig works. Watch
occur only in linear operation, and on cw when the
the screen current closely, particularly at low plate
key is up. It is not particularly important that
voltage or with high grid drive or light loading. The
screen voltage be held constant for high screen
current, as in plate -modulated a-m and key -down provisions for checking individual screen currents is
cw conditions with low plate voltage. The screen important, otherwise you may learn too late that
voltage will be kept down by the heavy load on the one tube has been taking all or most of what you
supply at such times. Actually a single string of have seen on a meter that ret Is total screen current
three regulator tubes will do the job quite well, and only. In the push-pull amplifier it may be advan-
both amplifiers have been worked successfully with tageous to balance screen currents by Cl, rather
this simpler screen arrangement. Current through than grid currents, if balance of both screen and
grid currents does not occur at one setting.
Tips on A -M Linear Amplifiers 113
Tune up for Class C and get the feel of the get 750 watts out of it in Class C. For the 144 -MHz
amplifiers before trying linear operation. Then, if amplifier, 200 watts out with 700 in is about the
linears are unfamiliar to you, read up on them safe maximum for a-m linear service. These are
below, and in chapter S before jumping in. Use a optimum figures; you may get less, but you can't
scope; there is no sure way to set up and operate a get more and be linear.
linear without one. The Heath Monitor Scope, For higher plate efficiencies go to ssb, cw, or
HO -10 or SB610, is ideal for this job because of its plate -modulated a-m. In any of these modes these
built-in tone oscillator and in-the -transmission -line amplifiers will give you the biggest legal signal
features. Running a linear, either sideband or a-m, around, if that's what you want. Or they'll throttle
without a scope check is inviting trouble. down nicely to 300 watts input or less, merely by
Finally, is you must use an a-m linear, don't lowering the plate voltage. They'll work efficiently
expect 70 percent efficiency from it. Don't expect at much lower inputs if the screen voltage is
50. Expect and see that you get, no more than 35 dropped appropriately. Chances are that you'll still
percent from a Class AB1 linear, or no more than have a signal that will stand out in most neighbor-
about half the rated plate dissipation for the tubes hoods, on either 6 or 2, and you'll have no worries
used. This means 350 watts out of our 50 -MHz about over heating.
amplifier with a kilowatt in, even though you can

TIPS ON A -M LINEAR AMPLIFIERS

It is no small wonder that the a-m linear fed to a linear results in more of the same; lots
amplifier appears attractive to the neophyte look- more! The exciter should be stable and its output
ing for his first step up the vhf power ladder. At stage as perfectly modulated as we can make it.
first glance it seems almost too good to be true. A Since the driver operates at very low level, this is
Class AB1 linear, the type most often used, not hard to do. If an exciter is being built
requires no driving power at all. Class AB1 is especially to drive a linear, it might be well to go
operation without the amplifier drawing grid cur- with a neutralized -triode output stage, with no
rent at any time. With the amplifier consuming no more than about 5 watts input. A Class -A modu-
power from the driver stage, only a mere handful lator employing inverse feedback and some form of
of exciter is needed. You could use a one -watt output limiting would be good. Peak limiting is
transistor rig, and have output to spare. important, to keep the average modulation percen-
This applies whether the amplifier runs 100 tage high and prevent overmodulation.
watts input or 1000, so it can be seen that the Most vhf transmitters will have a lot more
linear is most attractive in the high -power bracket. output than is needed, so the drive applied to the
The inevitable price to be paid is low efficiency. amplifier must be reduced in some way. Detuning
Thus, there is hardly any point in building a linear the driver output circuit or the amplifier grid
for less than about 200 to 300 watts input; you circuit will not do, as it may leave the driver
won't get enough step-up in power to make the without a proper load, and impair its modulation
project worthwhile. And since any amplifier is a quality. A simple solution is to connect a 50 -ohm
fairly expensive undertaking, it may be well to dummy load parallel with the driver output. A
build it for kilowatt capability, even if you don't coaxial T fitting is connected to the driver output
expect to push it that far right away. The receptacle. The dummy load is connected to one
amplifiers of Fig. 6-38 through 6-43 can be run as side of the T, and the amplifier grid input to the
low as about 300 watts input if you wish. At this other. The amplifier grid circuit still may have to
level they deliver about 100 watts to the antenna - be detuned slightly, if the exciter output is more
no mean signal on a vhf band. There is plenty in than 2 or 3 watts, but this will not be harmful for
reserve when you need it, and the final tubes only a small reduction in drive. Driver output may
hardly know they're working. also be reduced by lowering its plate or plate -and-
As its name implies, a linear amplifier is one screen voltage, though it is well to check the
which reproduces the wave form of its driver stage quality to be sure that linear modulation character-
exactly, but at higher power level. This requires istics are being obtained in the driver.
considerable attention to details. Everything has to
be right, or the signal quality suffers, and it will Checking Signal Quality
occupy far more space in the band than a signal The Heath Monitor Scope, Model HO -10 or
should. Grid bias, drive level and antenna loading SB610, is ideal for use with a vhf linear, as it may
are all critical. Regular use of an oscilloscope is a be left connected to the transmission line for
must. Meters alone are not enough, if you want to continuous monitoring. Some modification may be
be sure that your signal is above reproach. necessary for effective use of this scope on 144
MHz, though it works nicely on 50 MHz and lower
About Driver Stages bands as is. Two coaxial receptacles of the SO -239
Obviously the driver stage is important in the type are mounted on the back of the scope, with
linear picture. If we are going to amplify it in their inner terminals joined by a wire about 1 1/2
exactly its original form, the signal had better be inches long. The transmitter is connected to one
good to start with. A distorted splattering signal receptacle and the antenna coax to the other. The
 114 VHF EXCITERS AND AMPLIFIERS

(A) (B) (c)

Fig. 6-44 - Typical oscilloscope envelope patterns for amplitude modulation. Unmodulated carrier is
shown at A. The single -tone pattern for 100 percent modulation is shown at B. Peaks should rise to
twice the envelope -pattern height, and valleys should just reach the center line. The effects of
excessive drive or too -light loading, or both, on a linear amplifier are shown at C. Note the flat -topping
and small increase in amplitude over the unmodulated envelope.

unshielded wire inside the scope causes an appre- tune for maximum output.) Run up the drive now
ciable impedance bump in a 144 -MHz line. This to the point where grid current just starts to show,
may be corrected by connecting a coaxial T fitting and then back it off slightly. Readjust the plate
to one of the terminals, and using its two arms to and loading controls for maximum output. Be sure
make the above connections from transmitter to that you're putting every watt you can into the
antenna line. Internal scope connections and func- transmission line for this amount of grid drive.
tions remain intact, and the impedance bump is Maximum loading is a must for linear operation.
held to manageable proportions. Try modulating the driver, while watching the
The scope, milliammeters in the grid, screen scope pattern. It should look like the patterns
and plate circuits of the amplifier, and a power- shown in Fig. 6-44, A and B. These are envelope
indicating device in the coaxial line are useful in patterns, which are most readily obtained with the
setting up the linear for maximum effectiveness. Monitor Scope. Unmodulated carrier is shown at
The power meter will tell you if you are getting all A. The Heath scope has a built-in tone oscillator.
you should from the amplifier. If you're getting Using this or a steady whistle into the microphone
too much, the scope will tell you. The meters are should produce a pattern like the one at B, when
necessary to assure operation at both safe and the modulation level is 100 percent. The peaks and
optimum conditions. valleys are sharp, and the valleys (negative peaks)
The tube manufacturers' data sheets give typi- just reach the zero line. Positive peaks are just
cal operating conditions for various classes of twice the total height of the unmodulated envel-
service, usually including a-m linear. These are the ope, Pattern C shows effects of excessive grid drive
best guides available and you'll do well to follow or too -light loading, or both. Note the flat-topping,
them closely, especially when just learning your and the lower height of the positive peaks. If you
way around with a linear. They do not tell the don't have some form of negative-peak limiting,
whole story, however. They are merely "typical"; watch out for excessive modulation in that direc-
there may be other combinations that will work tion. That's where the splatter comes from first if
well, if you know how to read the indications your audio and rf operation is clean otherwise. In
meters and scope provide. Conversely, it may be watching your voice modulation beware of the
possible to radiate a less -than -admirable signal, bright flashes at the zero line of the modulation
when meter indications alone seem to be in order. pattern that indicate over-modulation on negative
You'll need that scope! voice peaks.
In using the 6- and 2 -meter linears of Fig. 6-38 Practice the adjustment routine with a dummy
the plate voltage can be almost anything, provided load connected to the transmitter, and you'll soon
that the amplifier is adjusted carefully whenever get the hang of it. Deliberately over -drive the
the plate voltage is changed. From 800 to 2000 amplifier and see how quickly you can detect the
volts has been used on 4CX250Rs and Bs. Screen results (pattern 6-44C) on the scope pattern.
voltage should be what the sheet calls for; in this Observe the meter action, too. You'll see that you
case 250 volts for Class C and 350 volts for Class can't draw any grid current without spoiling the
ÁB1. Bias should be variable and adjusted so that picture. You'll also see that when the scope picture
the tube or tubes will draw the recommended is right the plate current stands still on all
no-drive plate current. In this instance it's about modulation peaks. The screen current will prob-
100 mA per tube. It is well to start with bias on ably be just a bit negative. Output will absolutely
the high side (no-drive plate current low) to be on not exceed 35 percent of the input. If it does,
the safe side until set up correctly. you've got some meter inaccuracies, or you're
With the amplifier running in this fashion, feed cheating on the interpretation of the scope pattern.
in enough drive to make the plate current rise and The scope is the final authority; you have to
output start to appear. Tune the final plate circuit believe it.
and adjust the loading control for maximum Now, once over lightly again. Loading is all-
output, as indicated by the height of the scope important. Keep it at the maximum output you
pattern or by the power-indicating meter in the can get for a given value of grid drive. Recheck it
transmission line. Disregard the final plate current, for every frequency change or change in plate
so long as it is at a safe value. (Do not tune for dip; voltage. Grid current will always be zero. Grid
Grounded -Grid 50 -MHz Amplifier 115
drive can be lower than optimum as regards making any adjustments, if you want to switch
output, but never more than optimum. (You can instantly. A good linear is more versatile than this,
read grid vvoltage for a reference on amount of however. It's possible to do a lot better than the
grid drive, if you like.) The scope will tell you very a-m conditions on sideband, and still stay in the
clearly the minute you go too high. So will the AB1 mode. Efficiency on cw will shoot up
sound of the signal, but this may be hard to markedly with just a slight increase in grid drive,
determine, if your receiver overloads on your own with no other changes. Same for fm, which is
signal. Most receivers will. Final plate current will identical to cw, as far as the tubes in the final are
rise with increasing grid drive, but it must stand concerned. If you want the ultimate in cw or fm
still during modulation. If it kicks on modulation output, switch to 250 volts on the screen, and run
peaks, you've got distortion, and very likely up the grid drive some more. Drive level is very
splatter. uncritical, so about all you have to watch for is to
All adjustments react on one another to some keep the final input below the kilowatt level, and
extent, and each time you change any operating avoid swinging the plate current on fin. Readjust-
condition you have to go through the routine ment of the plate tuning and loading will be
completely again. This sounds as if you'd spend the needed for top efficiency. Plate -modulated voice
rest of your life tuning the rig, but once you get service is quite similar to the cw conditions, except
the hang of it you can make the necessary that the maximum plate voltage permissible is
corrections in seconds. lower with most tubes. Grid drive requirements are
usually slightly higher for good plate modulation
Using SSB and Other Modes conditions than for cw or fm, and the bias should
Since a -m linear is the most critical of all, it is be juggled for best modulation characteristics.
in order to switch to any other mode without Scope indication should be like Fig. 6-44B.

GROUNDED -GRID 50-MHZ AMPLIFIER

Increasing use of 50 -MHz transceivers and antenna line. Most of the lower part of the
transmitters having outputs of 25 watts or more schematic diagram has to do with control and
has created a demand for amplifiers to be used metering, and is largely self-explanatory. The ex-
with such equipment as the driver. The grounded- citer voice -control relay shorts out R1, allowing
grid amplifier of Fig. 6-45 is designed for this use. grid current to flow, and making the amplifier
With 30 watts or more of driving power it will operative, if the filament and primary -control
deliver 600 watts cw output. As a Class-B linear, switches, Si and S2, have been closed. Feeding ac
single -tone conditions, its rated PEP output is 750 voltage to the plate-supply relay through J4, J5,
watts. and PI makes application of plate voltage without
the filament and blower being on impossible.
Circuit
The Eimac 3-500Z triode ís designed for
grounded -grid service. As may be seen from Fig.
6-48 driving power is applied to the filament
circuit, which must be kept above rf ground by
means of high -current bifilar rf chokes, RFC1, and
RFC2. These are a central feature of the bottom
view, Fig. 6-47. The input impedance is low, so the
input circuit, LIC1, tunes broadly, and the 50 -ohm
line from the exciter is tapped well up on L1. The
plate circuit is merely a coil of copper tubing, L2,
inductively tuned by means of a "shorted turn" of
copper strip, rotated inside its cold end. See Fig.
6-46. Tuning is smooth and the rotating loop
avoids many problems commonly encountered in
tuning high-powered amplifiers by conventional
methods. Plate voltage is shunt -fed to the tube, to
prevent the high dc voltage from accidentally C
appearing on the output coupling loop or on tht

c 1

Fig. 6-45 - Table-top 50 -MHz amplifier of

grounded-gridklesign, only 10 X 12 inches in size. e
Grid and plate current are monitored simultan-
eously. Knobs at the right are for input tuning,
bottom, amplifier loading, center, and plate tuning,
top.
116 VHF EXCITERS AND AMPLIFIERS
Fig. 6-46 - Interior view of the 50 -MHz amplifier
shows the shorted -turn tuning system, plate coil
and output coupling, upper right. The tuning and
loading controls are mounted on a bracket to the
right of the 3-500Z tube and chimney. Meter
shielding is partially visible in the left front corner.
I:Jn

Connections to the grid terminals (on opposite

sidesof the socket) are made with short 1/4 -inch
copper straps soldered to the pins and bolted to
the chassis with No. 6 screws, nuts, and lock -
washers. Be sure that a clean tight rf ground
results.t
In Fig. 6-46 it will be seen that the hot end of
L2 is supported on the top of the two blocking
capacitors, C3 and C4, which, in turn, are mounted
on the Teflon rod that serves as the form for
RFC3. The ground end of L2 is supported on a
.vertical post made of 3/8 -inch copper tubing, 1-3/8
inches high. The end of the coil can be fitted with
a heavy copper lug, or pounded flat. A hole is
Construction drilled in the flat portion and a 2 -inch brass bolt
The amplifier chassis is aluminum, 10 by 12 by runs through it and the post and chassis. Be sure
3 inches in size, with the tube socket centered 3 that there is a permanent solid rf ground at this
1/8 inches from the front edge. The sheet- point.
aluminum panel is 10 inches high. The decorative The shunt -feed rf choke is effectively across the
edging is "cove molding," used by cabinet makers tuned circuit, so it must be a good one. Hand-
for counter tops. Sides and back are also sheet -winding as described below ís strongly recom-
aluminum. Where they need not be removable, mended, as no ready-made choke is likely to be as
parts are fastened together by pop -riveting. Tools good. Teflon is slippery, so a light thread cut in the
and rivets for this work can be found in most form will help keep the winding in place. If this
hardware stores. Perforated aluminum (cane metal) cannot be done, prepare and wind two wires, as for
is used for the top, and for covering the panel the filament chokes. Feed the wire ends through
viewing hole. one hole in the form, and wind a bifilar coil. Pull
Stretch the wire for the bifilar rf chokes, before the other ends through the finish hole, bending one
winding. Then, with the wires side by side, under back tightly at the hole edge. Remove the other
tension, wind them on a form of wood or metal. winding, which should leave a tight evenly spaced
This is left in until the choke ends are soldered in coil that makes an excellent vhf choke.
position. Then remove the form and coat the The blocking capacitors, C3 and C4, are
windings with coil cement, to help maintain turn mounted between brass plates, one of which is
alignment. fastened to the top of the rf choke form with a
sheet-metal screw. The other plate is connected to
the hot end of L2 by means of a wrap-around clip
of flashing copper. The lead to the tube plate cap is
made with braid removed from a scrap of coax. A
strip of flashing copper about 1/4 inch wide is also
good for this. Use a good heat -dissipating con-
nector, such as the Eimac HR6.
The shorted -turn tuning ring is centered be-
tween the first two turns of L2. The ring is
71Ps
attached to a ceramic pillar, and that to a 1/4 -Inch
shaft, the end of which is tapped for 8/32 thread.
...v This shaft runs through a bearing mounted in a
, a
bracket 4 inches high and 2-3/4 inches wide,
fastened to the chassis and the side of the
enclosure. The output loading capacitor, C6, is also
mounted on this bracket. It is one inch above the
chassis, and the tuning -ring shaft is 3-1/4 inches

Fig. 6-47 -
With the bottom cover removed, a look
into the chassis from the rear shows the input
circuit, Li C1, right, the bifilar filament chokes,
foreground, filament transformer and control
switches. Opening in the rear wall is for air intake.
Grounded -Grid 50 -MHz Amplifier 117
50º
C3 5KV
AMPLIFIER 1

3-500 2

2.3 C
50 MHz 50 MHz
C2
6000 RFC 3

INPUT Ji 75
J J3 OUTPUT

500
CS 6.01 C6
RFC 1 RFC 2 50

01 J2)\ Kv
6000 ro
3000 V DC
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
IN MICROFARADS ( yF I ; OTHERS
ARE IN PICOFARADS (PF OR yyF);
RESISTANCES ARE IN OHMS;
M
t 1000.
0-150

TO RELAY
IN MV
OFF SUPPLY
OFF PRIMARY
52
H v C ONppYROL

2w 1 6 7 6 J4

-MV VOY 317V ONO

RELAY AC

Fig. 6-48 - Schematic diagram and parts information for the 50 -MHz grounded -grid amplifier.

B1 - Blower, 15 ft3/min or more. ring is closed loop of 1/2-inch copper strip,

Cl - 75-pF variable (Johnson 167-4). 2 -5/8 -inch dia.
-
C2 - 1000-pF dipped mica. L3 1 turn, 3 -inch dia, and leads, made from one

C3, C4 - 500-pF 5 -kV transmitting ceramic (Cen- piece of 1/8-inch copper tubing or No. 8 wire.
tralab 858S-500). M1 - Dc meter, 0-1 A (Simpson Wide-Vue, Model
C5 - 500-pF, 10 -kV or more, TV "Doorknob." 1327).
C6 - BNC
50-pF variable(Johnson 167-3). -
M2 0-300 mA, like M1.
-
J1 - coaxial receptacle.
J2 - High -voltage connector (Millen 37001).
P1
R1 -
Ac plug, on cable to power supply.
47,000 -ohm 2 -watt resistor.
J3 - Type coaxial receptacle.
N RFC1, RFC2 -
21 turns each, No. 12 enam.,
J4 - 8-pin male power connector, chassis -
RFC3
1/2 -inch dia, bifilar.
-
30 turns No. 20 enam., spaced wire dia,
mounting.
J5 - Ac receptacle, chassis -mounting. on 3/4 -inch Teflon rod, 3-3/4 inches long. Drill
-4 turns No. 12 enam., inch long, -inch dia. end holes 1/2 and 2-3/4 inches from top.
L1
Tap 2-1/2 turns from ground end.
1 1
Si, S2 - with snap -in bracket).
Spst, rocker -type, neon -lighted (Carling
L2 -
3-1/2 turns 1/4 -inch copper tubing, 3 -1/2 -
T1
LT1 L,
- Filament transformer, 5 V, 15 A, (Stancor
inch dia, 5-1/4 Inches long. Diameter is finished
P6433; check any electrical equivalent for fit
dimension, not that of form used for winding.
See text and photo for turn spacing. Tuning
under 3-inch chassis).

above the chassis. The input tuning capacitor, CI, bypassed immediately inside the compartment
is mounted under the chassis, with equal spacing with a TV "doorknob" high -voltage capacitor, C5.
between the three, for symmetrical appearance. The blower assembly in the left rear corner of
The output coupling loop, L3, is just inside the the chassis draws air in through a hole in the back
cold end of L2. It can be adjusted for optimum of the compartment, and forces it down into the
coupling by "leaning" it slightly into or out of L2. enclosed chassis. The only air path is then back up
Be sure that it clears the shorted turn throughout through the socket and chimney (Eimac parts
movement of the latter. SK -410 and SK-406 recommended) and out
The coaxial output jack, J3, is on the rear wall through the top of the enclosure. The data sheet
of the enclosure. A small bracket of aluminum for the 3-500Z specifies an air flow of at least 13
grounds it to the chassis, independent of the cubic feet per minute, when the tube is operated at
bonding between the chassis and the enclosure. 500 watts plate dissipation. The ac leads for the
Plate voltage enters through a Millen 37001 high- blower motor come into the enclosure on feed -
voltage connector, J2, on the rear wall, and is through capacitors.
118 VHF EXCITERS AND AMPLIFIERS
The meters are enclosed in a shield fastened to plane. Dimensions that affect tuning range are as
the front and side panels. Meter terminals are follows: grounded support for L2 - 1-1/8 inches
bypassed for rf inside the shield, and leads come from right side of chassis, and 3-1/4 inches from
through the chassis on feedthrough capacitors. The rear. RFC3 mounting position -4 inches from rear
rocker-type switches just below the meters have and 5-1/2 inches from left. Shorted turn approxi-
built-in illumination. The high -voltage switch is not mately centered between turns 1 and 2 of L2. The
meant to control the plate supply directly, but start of L3 bends from the stator of C6 to near the
rather through a relay, as in the 3000 -volt supply start of L2. The end toward J2 passes between the
shown in the Handbook. The plate meter is in the first two turns of L2, clearing the tuning ring in
negative lead, so be sure that your supply is any position of the latter.
compatible with this arrangement. Do not use this Once the amplifier seems to work normally at
system where a potential difference exists between moderate plate voltages, apply higher, up to the
the amplifier and power supply chassis. All power maximum of 3000. Plate current, with no drive,
leads are made with shielded wire (Belden 8862) should be about 160 mA. It can be lowered by
and all exposed points are bypassed to ground. inserting 0.1 to 0.4 ohm in series with R1 and the
filament center -tap. A Zener diode, 2 to 9 volts, 10
Adjustment and Use watts, could do this job, as well.
Do not apply drive to the 3-500Z without the Keep the amplifier tuned for maximum output.
plate voltage being on. Also, it is recommended Do not decouple to reduce output; cut down drive
that initial testing be done with low drive, and with and/or plate voltage instead. Adjustment for linear
a plate voltage of 1500 or less. With a 50 -ohm load operation requires a scope. Maximum output,
connected to J3, apply 1000 to 1500 volts through minimum plate current, and maximum grid current
J2, and turn on the driver. Adjust the tuning ring should all occur at the same setting of the plate
inside L2 for a dip in plate current. Tune Cl for tuning. If they do not, the output loading is
maximum grid current. Tune C6 and adjust the over -coupled, or there is regeneration in the ampli-
position of L3 with respect to L2 for maximum fier. The plate -current dip at resonance is notice-
output. If the amplifier seems to be running able and smooth, but not of great magnitude.
properly, connect an SWR bridge between the Typical operating conditions given by the man-
driver and J1, and check reflected power. It should ufacturer, and in the tube-data section of the
be close to zero. If otherwise, adjust the tap Handbook, are guides to good practice. The ampli-
position on Ll. fier works well with as little as 1000 volts on the
Tuning range of the plate circuit can be tube plate, so varying the ac voltage to the
checked with a grid -dip meter, with the power off plate -supply transformer is a convenient way to
the amplifier. The range is affected by turn spacing control power level. It is seldom necessary to run
overall, and at the cold end. The closer the first the maximum legal power in vhf communication,
two turns are together the greater the effect of the so some provision for this voltage control is
tuning ring. No other tuning device is used, so recommended. With just one high-voltage supply
some experimentation with diameter and length of needed, and no critical tuning adjustments, power
L2 may be needed if you want other than the 49.8 variations from 100 to 600 watts output are
to 52.7 MHz obtained with the graduated turn quickly and easily made. This amplifier was built
spacing visible in the interior view. The highest by Tom McMullen, W1SL, and first described in
frequency is reached with the ring in a vertical QST for November, 1970.

KILOWATT AMPLIFIER FOR 144 MHz

The 144 -MHz amplifier of Fig. 6-49 can be run
in Class AB1, for a-m or ssb linear service; or Class
C, for high -efficiency a-m, cw, or fm. Driver power
output should be 2 to 3 watts for AB1, and 10
watts or more for Class C. For more on operating
conditions, see information on linear amplifiers in
this chapter, the tube manufacturer's data sheets,
or the tube data section of the Handbook.
a
7
Construction
The principal difference between this amplifier J
and its many predecessors using similar tubes lies in
the plate -circuit design. The inductor is cut from
flat sheet brass, in the form of a U. The circuit is Fig. 6-49 - The 144 -MHz amplifier is built in
conventional rack and panel style, with the entire
tuned by a simple handmade variable capacitor
that avoids problems commonly encountered in top of cane metal, to provide free air flow.
Controls are grid -circuit tuning, C2, lower left;
this part of a high-powered vhf amplifier. The output loading capacitor, C5, center; and plate -
circuit is practically identical to several previously circuit tuning, C4, with vernier dial, right. The
described in QST, this Manual, and recent editions slotted end of the Teflon shaft on Cl is visible as a
of the Handbook. white spot just below the loading control.
Kilowatt Amplifier for 144 MHz 119
Fig. 6-50 - Interior of the 2 -meter amplifier,
showing the brass plate -inductor and vane tuning
system. Note the position of RFC1, at the far left,
out of the main rf field. The output coupling loop,
L4, just below the plate line and barely visible
here, is connected to the output jack, J4, on the
rear wall with a short section of coax; and to the
loading capacitor, C5, on the front panel by means
of copper strip.
.w
The amplifier is built on a 17 X 8 X 3-inch
aluminum chassis, fitted with a bottom cover
which completes the shielding and directs the flow
of cooling air. The top portion of the enclosure is
of similar size, except that it is 3-3/4 inches high,
and it has a cane -metal top. It was made by Leads to the neutralizing tabs, C9 and C10, are
bending up the necessary sheet aluminum, but tapped on the grid lines at a point 1-3/4 inches
angle stock and flat sheets could be used equally from the grid end. Feedthrough bushings (not
well. Angle stock along the back of the front panel visible in the photographs) are under the lines. The
completes the enclosure. The gray -wrinkle alum- crossover is made by copper strips from the lines to
inum panel is 7 inches high. the bushings. Variable capacitance to the plate line
The tube sockets are mounted 2 inches in from is provided by copper tabs 1/4 X 5/8 inch in size,
the right side, as seen in the photographs, and soldered to the top ends of the bushings, just
2-5/8 inches apart, center to center. The Eimac below the plate line, L3. Adjusting their position
SK620A sockets, with their integral screen-ring with respect to L3 provides the required neutral-
shielding, are recommended. Other sockets may izing capacitance.
require slightly greater spacing, and some modifica- Connections to the grid ends of L2 are wrap-
tion of the plate -circuit dimensions. The raised around copper clips slipped over the tubing ends
screen -ring shield is also a great aid in neutralizing and fastened to the grid posts of the tube sockets
the amplifier. Some form of shield should be added with screws. They are soldered to the line ends, for
if early flat sockets are used. This need is particu- permanence. The connections to C2 are made in
larly acute if the amplifier is to be operated in the somewhat the same way, except that the tabs are
Class AB1 mode, which is characterized by very soldered to the stator lugs. Note that the rotor of
high power sensitivity. C2 is not grounded. It is supported on ceramic
The halfwave-line grid circuit, L2, is tuned at standoffs 5/8 inch high.
the end away from the tubes by the split -stator The grid-circuit isolating resistors, R1 and R2,
variable, C2, and balanced to ground by means of are connected to L2 by means of spring clips which
C3, a differential capacitor. This is supported on its are slid over the line before assembly. These can be
stator tabs, which are soldered directly to L2, tube grid clips, if available. They are moved along
immediately adjacent to C2. A strap of 1/4 -inch the line to the point of minimum rf voltage, using
copper connects the rotor of C3 to the chassis, in the familiar lead-pencil test.
the shortest practical manner. The slotted shaft of The shaft of C2 is rotated through an insulating
C3 is reached through a hole In the bottom cover shaft, fitted with an insulating flexible coupling, to
of the chassis. This hole is sealed with black plastic minimize any tendency to unbalance in the grid
tape after the adjustment is completed, in order to circuit. The shaft from Cl is also insulating
avoid air leakage. material, and it has a flexible coupling. The
The input coupling loop, Ll, is mounted capacitor is not adjusted often, so the shaft end is
between and just below the grid lines, with its slotted, and is allowed to protrude through the
closed end near the midpoint of the lines. The end front panel. It is just visible in the front view,
toward the panel is soldered directly to its tuning below the output -loading control.
capacitor, Cl, and the other to an insulating
tiepoint, which also has the center conductor of
the RG58/U coax to J1 connected to it. The
position of Ll with respect to L2 can be adjusted
by means of an insulating rod, through a hole in
the bottom plate near the closed end of the loop.
This hole is also taped over to prevent air leakage.
9
Fig. 6-51 - The principal feature of the bottom
view is the halt -wave grid circuit. Its split -stator
capacitor, C2, is at the left end of the line, L2. The
differential balancing capacitor, C3, is also across
the line, just to the right of C2. Isolating resistors
in the grid circuit, R1 and R2, are near the middle
of the picture. The screen isolating resistors, R3
and R4, run to tiepoints on the right wall of the
chassis.
 120 VHF EXCITERS AND AMPLIFIERS
AMPLIFIER

14A-1ABMHZ .001
144-118 MNZ

>rG I `OUTPUT
I J2

RFC 1

INPUT C6
JS
O
500
4.01V
J4
41000 TO
2000 V

NTRS.
EXCEPT AS INDICATED, DECIMAL
Y
VALUES OF CAPACITANCE ARE /47
IN MICROFARADS(pF); OTHERS
ARE IN PICOFARA DS IpF OR yyr);
RESISTANCES ARE IN OHMS;
BIAS 11...SCR.V-0 I 1 VAC
L 1000, M.1000 000
W1 1041

Fig. 6-52 -
Schematic diagram and parts infor-
mation for the 144 -MHz amplifier. Capacitors not J3 -8 -pin power connector, male.
described are disk ceramic. J4 - High -voltage power connector (Millen
Cl -25-pF miniature variable (Hammarlund
MAPC-25B).
37501).
L1 - Copper strip 1/4 X 4 inches. See Fig. 6-54.
C2 -
25-pF per section split -stator (Hammarlund L2 - 1/4 -inch copper tubing 10-1/2 inches long,
HFD-25). 15/16 inch center to center. Bend to Y shape 2
C3 -1.5- to 5-pF differential (Hammarlund
MAC -5-5). L3
inches from tube end.
--
.065-inch sheet brass; see text and Fig. 6-53.
C4 -
Vane -type tuning capacitor; see text and
photos.
L4 Copper strip 15/16 X 7-1/2 inches, bent to
roughly elliptical shape. See text and Fig. 6-54.
C5 -
35-pF variable (Hammarlund HF-35). --
R1, R2 150 -ohm composition, 1/2 watt.
C6 -
500-pF 10 -kV TV "doorknob." R3, R4 150-ohm composition, 1 watt.
C7, C8 -
Screen bypass; part of Eimac SK -620A -
R5 20 -ohm 10-watt, slider type.
socket.
C9, C10 -
Neutralizing tabs 1/4 X 5/8 -inch sheet
RFC1 -
32 turns No. 24 enamel, closewound on
1/4 -inch Teflon rod. See mounting position in
copper, soldered to top of National FTB bushing. interior photo.
I1
J1
--
115 -volt neon pilot lamp.
BNC coaxial jack.
S1
T1
--Spst6.3-V
toggle switch.
6-A filament transformer (Merit
-
J2 Type N coaxial jack. P-2947) .

All power leads are made with shielded wire,

bonded together by frequent spot -soldering, and to assembly in the interest of clarity, but their
the chassis by means of grounding lugs. Exposed position is clearly visible in the photographs. These
terminals are bypassed wherever necessary, to tabs are curved slightly after bending, to provide
prevent rf pickup. more contact surface to the anode. Clamping rings
Each cathode pin on the socket is grounded made of flashing copper wrap around the anode.
through a separate lug, and nothing else uses these structure and hold the tabs tightly to it. This is a
lugs for a ground path. Minimum cathode -lead point of low rf current, so a large contact area is
inductance is important. Even the shortest lead not vital.
shared with another circuit can cause unwanted The plate line was made flat originally, but
coupling in a vhf amplifier. when the amplifier was tested it was found that
The plate inductor. L3. is made of sheet brass. this did not allow room to adjust the output
in the form of a U. Principal dimensions are given coupling loop, L4, to the optimum position. The
in Fig. 6-53. The stator plates of the tuning half-inch offset shown in Fig. 6-53 (but not in the
capacitor, C4, part A, are soldered to the plate line photographs) netted a marked improvement in
with their right edges 5/8 inch from the tube efficiency. The entire plate circuit was silver-plated
anodes. Connection to the latter is made with two after the photography. Careful checks on perfor-
brass tabs, part B, at the tube ends of the line. mance indicated no difference, before and after
These were omitted from the drawing of the plating. Plating may be desirable on a long-term
Kilowatt Amplifier for 144 MHz 121
ii"2

uP.-B2A55
BEND

A
0 r
2 REOUEIED
2

1n 2

-.1:11

ii`11
BEND UP

C .030-BRASS
.I 4 REQUIRED

BEFORE BENDING

Fig. 6-53 - Principal dimensions of the brass parts of the amplifier plate circuit. The U-shaped
inductor is shown in both top and side views, with the stator plates of C4 in place. These plates (A) are
shown before bending, at the upper right. The small brackets (B) make contact with the tube anodes.
Slight curvature, to fit tube anode, can be imparted by tapping with a small hammer, against a
1 -1/2 -inch pipe or rod, used as an anvil.

basis, as silver oxide is a good conductor and other stops the rotor in the horizontal position. The
oxides are not. rotor is prevented from "going through the roof'
The "stators" and the tabs for the anode by a 1 -inch setscrew in the vernier-drive hub, and a
connection were silver-soldered to L3. Ordinary longer -than -normal screw for the lower left mount-
soldering will be adequate, but it might be well to ing screw for the drive assembly.
use screws to hold the tabs onto L3, as a The rotor in its horizontal position Is approx-
precautionary measure. The stator plates have imately 1/4 inch above L3, and the spacing at the
flat-head screws running through them and L3, ends of the rotor is also 1/4 inch. The tubes are
into the insulating supports for the latter. These fitted with Eimac SK626 chimneys. The under
are 1 -inch ceramic pillars. The closed end of the surface of L3 should just clear these. If it does not,
loop is supported on a 1 -1/2 -inch pillar. raise it by putting washers on the screws that run
The holes for these supports can be made into the 1 -inch pillars.
slightly oval, to position the assembly so that no The output loop, L4, is supported under L3 by
strain on tubes or sockets is caused when the anode two 1/2 -inch ceramic insulators. If the threaded
rings are tightened. The mounting hole in the holes go the whole length, be sure that the
closed end of L3 is also elongated. The screw that mounting screws do not ground the loop, or come
holds the line on its support has Teflon washers close enough to allow arcing to ground. Connec-
above and below L3, to permit the line to move on tion to the coaxial output jack, J2, is made with a
its support, if expansion and contraction with short piece of RG-8/U coax, using a shielding cone
heating and cooling of the line should be appre- at the J2 end. The coax shield is grounded to
ciable. chassis with a copper strap at the L4 end also, to
The rotor of C4 is in the form of a shallow box make the rf path to ground independent of the
made of flashing copper. It is shown in flat form in chassis bonding. The rotor of CS is also grounded
Fig. 6-54, along with other copper parts of the independently. A copper strap connects the stator
plate circuit. Its ends, 1 inch high, provide the of CS to the end of L4. After the final form and
variable capacitance to the stator plates on L3. size of L4 have been determined, the connection to
After the box is bent to the desired form, its the strap should be soldered, to maintain a good rf
adjoining surfaces are soldered for additional bond. These circuits carry high rf currents, and
strength and rigidity. The edge away from the tube permanent low-resistance connections are impor-
anodes is supported on a fiber glass rod with 4.40 tant. The performance of many amplifiers falls off
screws, the rod surface having been filed flat in this with aging, because factors like this were over-
area previously. Reducing couplers at each end of looked.
the rod permit use of a 1/4 -inch shaft bearing at An adequate supply of cooling air must be
the rear, and a National Velvet-Vernier dial mech- provided. The manufacturer stipulates 4.6 cubic
anism at the front. Do not use heat-sensitive rod feet per minute, per tube, minimum, but much
such as Lucite or Plexiglas. Nylon and some types more should be available. The blower used here has
of Bakelite are unstable in strong rf fields, and are a 3 -inch diameter wheel, turning at 3300 rpm. It is
also unsatisfactory. Teflon is probably good, but connected to the rear of the chassis by way of an
the fiber glass rod is stronger and easy to work. It automotive defroster hose 2-1/8 inches In dia-
is 6-3/8 inches long and may be 1/2- or 3/8 -inch meter.
diameter.
Mechanical stops for the rotor are provided at Adjustment
both ends of its normal travel. A 3/8-inch Teflon Heater voltage (at the socket) should be 6.0
rod 1-3/8 inches high, fastened to the chassis volts. This is adjusted by means of the slider on
between the neutralizing feedthrough bushings, RS. Set the sliding clips on L2 at the approximate
 122 VHF EXCITERS AND AMPLIFIERS
6"

4_s" 1500E CLAMP

12 FEGI
2

Li

_f"
(

Fig. 6-54 - Flashing-copper parts used in the 2 -meter amplifier. Broken lines indicate 90 -degree
bending required. The surfaces of the C4 rotor are soldered together after bending, for ridigity. The
anode clamps, upper right, wrap around the tube cooling ring, and hold the brass tabs (Fig. 6-53)
firmly in place. L1 and L4 are shown in the approximate shape, after bending, at the right.

midpoint. Now apply 1 to 2 watts drive to the grid insulated screwdriver or other safe shorting device.
circuit, adjusting the position of L1 and the tuning Do this every time before touching anything inside
of Cl and C2 for minimum reflected power, the compartment in any other way. Play it safe!
indicated on an SWR bridge connected between Apply plate and screen voltage, in that order.
the exciter and J1. Adjust bias so that the plate current is about 150
With enough drive so that grid current will be mA. Apply drive, and tune C4 and C5 for
measurable, meter each grid separately, and adjust maximum power output. With enough drive for
the balancing capacitor, C3, for as near to the same about 5 mA grid current per tube, the plate
value for each grid as possible. Readjust C2 for efficiency should approach 70 percent, after the
each change. When the currents are approximately position of IA with respect to L3 is adjusted with
equal, the neutralization should be adjusted. With a some care. Loop position and all tuning adjust-
50 -ohm load connected to J2, and with the screen ments change with plate voltage and drive level, so
and plate circuits having some dc path to ground, in linear service all adjustments should be made
such as through power supply bleeders, couple a under the conditions for which you want best
sensitive rf indicator to L3. Still with no plate or linearity.
screen voltage applied, tune C2 and C4 for maxi- The shape and position of L4 are quite critical.
mum indication, then adjust the positions of the Best efficiency was obtained with the loop roughly
neutralizing tabs, C9 and C10, carefully for mini- elliptical in shape, and about 3/8 inch below L3.
mum rf feedthrough. Recheck the grid circuit Best results show at plate voltages between 1200
balance and tuning each time a tab setting is and 1800. The tube maker's typical operating
changed. conditions are the best guide to efficient operation,
The points of connection of the resistors R1 but they are only typical. If safe levels of grid,
and R2 on the lines comprising L2 are not critical, screen, and plate dissipation are not exceeded,
unless the exciter is low on output, but they many variations are possible. See "Tips on Linears"
should be near the points of lowest rf voltage on earlier in this chapter.
the line. Check by running a pencil lead along the This amplifier was built by W1SL, and des-
line and watching the grid current. The point at cribed in February, 1971, QST.
which there is no change in the meter indication is
where the clip should be. Recheck all adjustments.
The approximate tuning range of the plate References
circuit can be checked with a grid -dip meter, with The 50- and 144-MHz amplifiers
described
no power on the amplifier. It should tune more incorporate features from many previous QST,
than the width of the 2 -meter band. Now, with an Handbook, and VHF Manual projects.
output indicator and a good 50 -ohm load con- Maer, "Perseids Powerhouse," QST, Oct., 1959
nected to J2, the amplifier is ready for power. (dual -band amplifier for 50 and 144 MHz).
For initial tests, the plate voltage should be 800 "High -Efficiency 2 -Meter Kilowatt," QST,
to 1000 volts. Screen voltage should be no snore Feb., 1960 (PP 4CX300As).
than 250, preferably regulated. There will be little Breyfogle, "Top Efficiency at 144 Mc. with
difference in tuning or output with the cover on or 4X250Bs," QST, Dec., 1961.
off, so, with due regard for safety, leave it off, at "Kilowatt Amplifiers for 50 and 144 Mc.,"
QST, Feb., 1964. Basic information also in this
first. Never reach inside the plate compartment Manual, all editions, and in the Handbook, 1966
when high voltage is applied. To be sure that it is through 1970. Metering and control information
off, short the plate inductor to ground with an applies to the 144 -MHz amplifier Described here.
 Chapter 7

Vhf Stations -Transceivers,

Transverters, and Transmitting
Converters

Buy or build? This question faces every new still essentially a technical avocation. The fellow
amateur, and it is likely to remain with him as he who learns his way around through building his
advances in the art. Though there is nothing wrong own equipment and making it work at topnotch
with the all -commercial station, if its owner is well efficiency is sure to get more out of it than the
informed technically and he operates intelligently mere purchaser of boxes. Finally, there is the
and with consideration for others, there are still personal pride and satisfaction in operating some-
sound arguments for building one's own equip- thing you built yourself. It Is your station, in a
ment, or at least some of it. Many hams new and sense that no manufactured package can ever be,
old still play the game that way. and you will be a better ham for having done the
Parts cost money and may be hard to come by, job yourself!
but building your own gear can save you money on "Nobody builds ham gear any more?" Don't
the total station cost. This results from the choices you believe it! Constructional articles, particularly
open to the builder. Unlike the buyer of commer- those dealing with vhf, bring in thousands of letters
cial gear, he can have the station he wants, which from readers of QST, the ARRL Handbook, and
will do best the things he wants most to do. A this Manual. Prospective builders order bales of
station covering 80 through 6 meters, for example, templates for constructional projects, whenever
is a waste of money for the vhf enthusiast having they are offered. Fm items described in Chapter 11
no interest in the hf bands. The vhf-only approach are current best-sellers. The better projects have an
is sure to deliver more watts -per-dollar and better amazingly long life. Components of the "Two -
reception for the money than the multiband Band Station for the VHF Beginner," pictured
station. Perhaps more important, there may be herewith, are still being built from the original QST
nothing commercially available that really fits his series published more than a decade ago, and from
requirements. The continued popularity of high- descriptions of parts of the station that have
efficiency one -band kilowatt amplifier designs such appeared in the Handbook and this Manual, in past
as those described in Chapter 6 is proof of this. editions. A reprint of the QST articles is still a
But most important is the nature of the hobby brisk seller, even though it describes all-tube
itself. Despite all easy approaches, vhf hamming is designs, in what is admittedly the age of the

.` ZI,r
Il.

1.7-
(. n
.
111
AI%

A
@
.9 4111 ¡'e
7-1 -A complete two-band station, shown here set up for 50 MHz. The transmitter, left, and
converter, right, have companion plug-in units for 144 MHz. The control unit, left center, contains the
power supply and modulator, and all units of the station draw their power from it. The simple tuner,
right, may be omitted if the builder has a communications receiver. At the far right is an SWR bridge
that doubles as a test meter.

123
124 VHF STATIONS
transistor and the integrated circuit. Today the together, so completely new elements of the
series stands as the most -used vhf material ever station can be tried without abandoning the entire
published, and one of the classic successes in the project. Such a station can be made the basis for a
history of QST.1 more advanced setup quite readily, since the
transmitters can serve as drivers for higher powered
Station Planning amplifiers, and the converters can be used with a
Too often, amateur stations "Just grow," rather communications receiver, replacing the simple
than developing along planned lines to make the tuner shown.
best use of the considerable financial outlay they If the would-be vhf operator wishes to build all
usually represent. This applies to equipment pur- or part of his station unit -style, he is advised to go
chased ready made, as well as to that built at over the basic information given in other chapters.
home. We accumulate transmitters, receivers, con- Receiving basics are discussed in Chapter 3, with
verters, modulators, and so on down the line, with the special aspects of fm reception added in
little thought as to their integration into a working Chapter 10. Practical receiving units are shown in
unit for vhf communication. Some commercial Chapters 4 and 11. Transmitter design is mainly in
gear leans to the opposite extreme -
the one -box Chapter 5, with fm factors in Chapter 10. Trans-
station that may be neat and unobtrusive, but is mitting equipment you can build is In Chapters 6
often lacking in versatility. and 11.
Amateurs are individualists. We like our stations This chapter is devoted mainly to one -package
to be unique, tailored to our special needs. With applications of information in the above chapters.
some advance planning we should be able to Most of it is the result of the trend to ssb operation
assemble a station that is both effective and in the vhf bands, which has had an Incidental
versatile, without its necessarily becoming elabor- dividend In the form of good cw capability, almost
ate or tremendously expensive. With these objec- inherent in the qualities needed for good ssb.
tives in mind, most of the equipment we describe Examples of simple a-m portable transceivers are
here is built unit -style, with few built-in heavy also included.
items like power supplies and modulators. These It should be emphasized that, as far as their rf
tend to be static in design; long-term investments circuitry is concerned, there is little difference
that can be used with a succession of rf units we between a-m and fm transmitters. The prospective
may wish to build and try. Subassembly design has builder of vhf transmitters for other than fm use
much to recommend it, and cost is by no means should not, therefore, overlook the transmitting
the only consideration. The ability to try different units for fm described in Chapter 11.
circuits without becoming involved in the kind of
rat's nest all too often seen in amateur stations
should rate high in our planning. The Transverter
The beginner's two -band station, shown set up Heterodyning, which is basic to the modern
for 50-MHz operation in Fig. 7-1, demonstrates communications receiver and to most ssb trans-
that a home -built setup can be neat, yet versatile. mitters, is being used increasingly in vhf stations.
Though some components may be considered For review, two frequencies combined in any way
obsolete by present-day transistor standards, the produce two other frequencies, equal to the sum
transmitter and converter units of this station are and difference of the original two. Either product
still representative of good vacuum -tube technique. of the mixing process can be used, or both may be.
The block diagram of the station, Fig. 7-2, helps to See "Heterodyning," Chapter 5.
show how a low -powered a -m and cw station for Typical vhf applications of heterodyning are
50 and 144 MHz can be set up. All units plug shown in Fig. 7-3. The receiving use, A, involves
1 For a reprint of the 4 -part OST series vhf signals, in this instance at 50 MHz and higher, a
describing equipment shown in Fig. 7=1 send 50 36 -MHz energy source, a mixer, and a communica-
cents (no stamps) to ARRL, Newington, CT
06111. tions receiver which acts as the detector system for

50 OR 144 MC.
ANTENNA
Fig. 7-2 - Block diagram of an
a -m and cw unit -style station. A
SWR central unit contains the speech
BRIDGE equipment, power supply, and
TRANSMITTER OUTPUT INCOMING SIGNALS
control circuits. The antenna con-
SO -54 OR 144- 14$ Mc. 50-54 OR 144-14$N,. nects to a send -receive relay on
RELAY
SEND RECEIVE SPEAKER the back of this unit through a
OR PHONES standing -wave bridge. Transmitter
KEY FOR c( 14 TO IS Me, rf assemblies for 50 or 144 MHz
C.W
TRANS- MODULATOR, plug into the left side of the
MITTER
POWER SUPPLY control unit, and a tuner for 14 to
POWER AND
R FUNIT t CONTROL POWER TUNER 18 MHz into the right side. Con-
87010 AND
CIRCUITS
14 TO POWER CONVERTER verters for 50- or 144 -MHz recep-
WATTS AUDIO 18 Mc AT SIGNAL
OUTPUT FREQUENCY
tion plug into the right side of the
tuner. The various units may be
interconnected with cables, in-
MIC. stead of being plugged together, if
operating convenience so dictates.
A Transmitting Converter for 50 to 144 MHz 125
Fig. 7-3- Typical examples of heterodyning in vhf
V 50-36=14
communication. Use of a 50 -MHz converter with a
14 -MHz receiver is shown at A. The same frequen- 50 -MHz RECEIVE 14 -MHz
cies are used in transmitting, B. A single oscillator SIGNALS
MIXER RECEIVER
serves both transmitting and receiving functions in
C, permitting operation on a vhf band with a
transceiver designed for lower frequencies. 36 -MHz
CRYSTAL (A)
OSCILLATOR

14 + 36 50
the heterodyned signals. The same frequency com- 14 -MHz TRANSMIT 50 -MHz
binations are used in transmitting, 7-3B. Here a XMTR -e - MIXER AMPLIFIER
50
14 -MHz transmitter and the 36 -MHz energy source
feed a mixer, and the output is a 50 -MHz signal
having (hopefully) the same characteristics as the 36-MHZ
14 -MHz driving signal. The devices described are
CRYSTAL
OSCILLATOR
(B)
usually known as receiving and transmitting con-
verters. As far as the functions we're discussing are
concerned, the transmitted or received signals can
be of any mode: cw, ssb, a -m, fm, or whatever. ¿ANT. RELAY
Since the same frequencies are involved in both 50 -MHz SIGNALS 50 POWER

converters, It follows that a single energy source Y

Í
can be used for both heterodyning jobs. This is RECEIVING 36 -MHz TRANSMIT 50 -MHz
done in the transverter, Fig. 7-3C, a device increas- CONVERTER OSCILLATOR CONVERTER AMPLIFIER
ingly familiar on the vhf scene.

The Transmitting Converter 14- MHZ

Receiving converters discussed extensively In I -F 14 MHz TRANSCEIVER 14 MHZ OUTPUT

(C)
Chapters 3 and 4 are essentially unchanged when
they perform the receiving function in a transver- SPKR (-(1 MIC

ter. Our concern in this chapter will be with

complete transverters and transmitting converters.
Remember, in considering the heterodyning ap- as free of harmonic content as possible, and that
proach to vhf transmitting equipment, that it has they be chosen to keep unwanted mixing products
two basic advantages: The low-level signal can be far away from the desired frequency range, so that
modulated or keyed in any manner before mixing, they can be rejected with the inherent circuit
and the resultant signal on the vhf band has the selectivity of the system, or by suitable filters
same stability and modulation characteristics as the added for this purpose.
original, if the process is handled properly. Another major consideration in mixer design is
There are problems, too, the most important that the mixing frequencies be applied in such
being that the heterodyning process can very easily ways and at such power levels that the mixer does
result in radiation of unwanted products, which not have excessive distortion products that may
can cause interference if not properly suppressed. cause poor signal quality or excessive transmitted
It is very important that the mixing frequencies be bandwidth, or both.

A TRANSMITTING CONVERTER FOR 50 TO 144 MHz

If you have a 50 -MHz transmitter that operates transceiver. Ideas for doing this can be picked up
to your satisfaction, you can have the same kind of from a transverter described later in this chapter.
signal on the 144-MHz band quite readily, through
use of the transmitting converter of Fig. 7-4.
Though this 3 -tube rf unit was built primarily to Circuit and Layout
develop ssb and cw drive on 144 MHz with a The schematic diagram, Fig. 7-5, and the
Heathkit HX-30 50 -MHz sideband exciter, it works bottom view, Fig. 7-6, may be "read" from left to
well with other modes used on these bands. If the right. First we have a simple triode crystal
50 -MHz rig is VFO-controlled, the 144 -MHz signal oscillator, VIA, on 47.0 or 47.5 MHz, depending
will have the same stability as the 50-MHz one, and on the crystal, Yl. The 47 -MHz plate coil, L1, and
the dial calibration will be the same for both Its tuning capacitor, Cl, are in the upper left
bands. corner of the picture. The second triode of VI is a
Though only the transmitting circuits are doubler to 94 or 95 MHz. Its tuned circuit, L2C2,
included, the first stages could also be used to is seen adjacent to the oscillator, but with its axis
generate the injection voltage for receiving, if one perpendicular to L1. Inductively coupled loosely
is interested in 144-MHz operation with a 50 -MHz to L2 is L3, the grid circuit of a 94-MHz amplifier,
126 VHF STATIONS
Fig. 7-4 - Transmitting converter for duplicating a
50 -MHz signal on 144 MHz. Two tubes at the left
and center comprise the oscillator, multiplier,
es amplifier, and mixer stages. At the right is the
144 -MHz output amplifier. Provision is made for
metering all stages by means of tip jacks and test
points. Note crystals taped together to prevent loss
of the one not in use.
m

or anything but Class -C conditions adequate drive is

very easily achieved. Here again, be sure that L7
actually tunes through the desired frequency, in
order that rejection of unwanted frequencies will
be at a maximum.
Construction is on a standard 5x 10 -inch
V1C. On the right side of the first tube is the
aluminum plate and 3 -inch chassis. A layout
amplifier plate circuit, L5C3, straddled by L6, the drawing, Fig. 7-7, is given for those who wish to
split grid coil of the mixer, V2. Below the mixer make an exact duplicate. To check every circuit
tube is the 50-MHz input circuit connected to the during the adjustment phase of the project, an
mixer cathodes. unusual combination of feedthrough bypasses and
From here on the layout and circuit look like
tip jacks is used. Oscillator plate current is
any other low-powered 144 -MHz transmitter. The measured by plugging a meter into J1 and clipping
amplifier grid coil, L8, is purposely made too small to the exposed terminal of C9. All other plate
to resonate in the 144- to 146 -MHz region with the currents may be read by plugging one side of the
input capacitance of the 6360, V3. Being on the meter into J2 and connecting to C10, for doubler
high -frequency side of resonance, it offers little plate current, C11, for 94 -MHz amplifier plate
feedback coupling to the output circuit, even current, C12, mixer plate and screen current, or
though there is no shielding between the two. The C13, amplifier plate and screen current. Amplifier
amplifier plate circuit, L9C7, is at the far right. grid current, if any, is checked at J3 and J4. A
Output is taken off through a series -tuned link, table of operating conditions is given later.
L1 008. Bias for the mixer and output amplifier is
Positioning of the various coils is important. obtained from a small 22 -1/2 -volt battery. Builders
Note that coils are placed so that unwanted may prefer some other bias source, but the battery
coupling between circuits is kept down, even with does the trick simply and inexpensively. There is
a fairly compact layout. It is suggested this no current drain, and it may even be charged a bit
principle be followed unless the builder is willing when the amplifier runs into Class-C conditions, so
to cope with a new set of neutralization problems. life should be long and voltage constant. Just be
The oscillator and doubler circuits are standard careful not to short out the battery when working
practice. In the grid circuit of the 94 -MHz on the unit. Information on suitable bias supplies
amplifier, the input capacitance of the 6M11 may be found in any edition of the ARRL
pentode was too high to permit resonating L3 at Handbook.
94 MHz in the usual way. Some checks with a The 6360 amplifier operated satisfactorily
variable series capacitor showed that a coil the
without external neutralization, but a small
same size as in the previous plate circuit could be amount was added when a slight reaction on
resonated at 94 MHz with about 10 pF in series, so amplifier grid current was noted as the plate circuit
the fixed capacitor shown in Fig. 7-5 was used. was tuned through resonance with voltage off. The
Only a small amount of energy is needed for the
grid and plate leads are crossed over inside the
mixer grids, so neither the tuning nor the coupling
6360, providing inherent neutralization in the vhf
between circuits is at all critical.
range, so only a tiny amount of additional
Getting the 94 -MHz amplifier to operate in a capacitance is needed. A half-inch wire is soldered
stable manner is mainly a matter of achieving
to each grid terminal and bent over toward the
ground potential for rf at the screen. This is done
adjacent plate terminal. The position is adjusted
with the series circuit, L4C4, the setting of which
until reaction on amplifier grid current is
is not particularly fussy. Coupling between L5 and
eliminated. For circuit simplicity, this neutraliza-
L6 should be adjusted to the minimum that will
tion is not shown in Fig. 7-5.
provide satisfactory output from the mixer. Make
sure that both circuits actually tune, as it is
possible to get enough output with one or the Adjustment and Use
other not actually peaking. Best rejection of It should not be taken for granted that the
unwanted frequencies will not be assured unless heterodyning approach is for the sideband operator
the circuits are tuned to the desired frequencies. alone. Given any of the popular small 50 -MHz
Coupling between L7 and L8 should also be as transmitters, homebuilt or commercial, this
loose as it can be and still provide adequate drive heterodyne unit will duplicate its signal on the
for the 6360. Drive requirements depend on the 144 -MHz band at a comparable power level. You'll
class of operation of the output amplifier. For need no big batch of crystals or two separate VFO
 OSC, ()SLR. AMP. MIXER AMP.
470 OR
475Mc. 61v111 e4 OR GAR11 ó36O
i0 94 OR 44 TJ
95 MC \j 144 TO
vI L5
Vlc 95 M ,
74b M"
4 b Mc ,
VIA
C CZ s 14 Ls L4
470 OR 8 L
47.5 ('.C3
5 W
Mc 1,
4 1009 100 oá
33K
Y, sou TkJ
500
00 100 fS-
4700 .8100 3 30 RF0j
IW. C10 10K 71M
C9 C47
c11
J4
2
2 B, 100
J 104
00 : =
T+
+150V. KEG. % .F+ TO -22V ,B,
2
CI$
20 J3
V1 V2 V3
82 7
I J2 O
RFC2 RFC3
.r4-::53 +250v. D.C.
6.3V

Fig. 7-5 -Schematic diagram and parts informa- with 3/8 -inch lengths of spaghetti when not in L8 - Like L6, except 1/2-inch space at center.
tion for the heterodyne unit. Decimal values of use for metering. L9 - Like L6, except 3/16 -inch space at center.
capacitance are in microfarads (µF1; others in C14 500-pF button -mica.
- L10 -One turn insulated hookup wire inserted
picofarads (pF). Unless otherwise indicated, fixed C15 20-pF miniature trimmer (Hammarlund
- halfway into center of L9. Leads are 1-1/2 and
capacitors are ceramic. Resistors are composition, MAC-20). 2 inches long.
1/2 watt, unless specified. J1, J2, J3, J4 Insulated tip jack.
- L11 5 turns No. 24, 1/2 -inch dia, 32 tpi, tapped
-
B1 - 22 -1/2 -volt miniature battery (Eveready No. J5 Coaxial receptacle, SO -239.
- 2 turns from ground end (B&W 3004).
412). J6 Phono jack or coaxial receptacle. L12 -2 turns made from same piece as L11,
L1 6-1/4 turns No. 24, 1/2 -inch dia, 32 tpi spaced 4 turns away; cold ends adjacent.
--
C1 - 15-pF miniature trimmer (Hammarlund
(B&W 3004). RFC1 No. 28 enam. close-wound full length of
-
MAC -15).
L2, L3 3-3/4 turns, like L1. .
1/2 -watt 1-megohm resistor. (Any ready-made
C2, C3 -10-pF miniature trimmer (Hammarlund -
MAC -10). L4 -6 turns No. 20, 1/4 -inch dia, one inch long. 144 -MHz rf choke is suitable.)
L5 -3 turns, like L1. RFC2, RFC3 Ferrite -bead rf choke, 1.3 /tH.
-
C4 - 3-30-pF mica trimmer.
C5, C6, C7 - 8-pF per section miniature butterfly L6. - 2 turns each side, No. 20, 1/2 -inch dia, 16 (National R-45-251) No. 24 enamel wound full
tpi, center tapped. Make from single piece of length of 1 -watt resistor also usable.
(Hammarlund MACBF-8). Y1 - Third -overtone crystals, 47.0 and 47.5 MHz
C8 - 30-pF miniature trimmer (Hammarlund B&W 3003 cutting all but one plastic strip.
MAC -30). Leave 3/8 -inch space at center. (International Crystal Mfg. Co. F-605). Tape
C9, C10, C11, C12, C13 500-pF feedthrough
- L7 - 4-1/4 turns No. 20, 1/2 -inch dia, 16 tpi, together, opposite ends up, to prevent loss of
bypass (Centralab FT -500). Cover exposed ends center tapped (B&W 3003). one not in use.
128 VHF STATIONS
units to give coverage of both bands. If you're a The various operating voltages are brought to a
Technician or Novice at present, use only the terminal strip visible in the upper center portion of
47.5 -MHz crystal in the oscillator -multiplier. Fig. 7-6. In firing up the unit apply plate power to
Heterodyning from 50 -MHz frequencies will start one stage at a time, beginning with the oscillator.
your coverage at 145 MHz. A crystal that gives This stage works simply, showing the usual sudden
operation on 50.2 MHz will put you on 145.2 downward kick in plate current from about 12 to 5
MHz, and so on. A VFO that covers 50 to 51 MHz mA when the crystal starts oscillating. Set Cl so
(not for Novice use, of course) will give you that oscillation starts every time voltage is applied.
coverage of 145 to 146 MHz, which can be If you have a grid -dip meter you can set all
extended to 144 MHz with the insertion of the following circuits close enough without applying
47 -MHz crystal at a suitable time. power to the unit. The dip meter can also be used
The output stage of the conversion unit can be to indicate power output relatively from the
run as a linear amplifier for sideband, cw, or a -m, various stages, and to determine that output is on
or it can be driven into Class-C conditions for the desired frequencies.
higher efficiency on cw. Plate modulation may be The pentode amplifier should be checked for
applied in the usual way for high -efficiency a-m stability by removing power from the preceding
service. The linear way will probably be the more stages briefly and watching the amplifier plate
attractive to most users, however, as it eliminates current while tuning C3. Should any fluctuation
the heavy and power -consuming audio equipment. appear, adjust C4 to stop it.
If your 50 -MHz rig is plate modulated, you can We are now ready to "mix" and to obtain
make provision for switching the audio power over output on 144 MHz. Feed 50 -MHz power into J6.
from its final stage to that of the conversion unit. With power on V1 and V2, check for output on
144 MHz at L7. A pilot lamp connected to a loop
of insulated wire wrapped around L7 may be used
Cathode Injection temporarily as an output indicator. When output
Initially we ran the 50-MHz energy into the has been obtained, connect a one-mA meter to J3
mixer grids and applied the 94 -MHz injection to and J4, and look for amplifier grid current. Leave
the cathodes, but it was easy to saturate the grids plate and screen voltage off the 6360 for the
with the swinging drive from the 50 -MHz sideband moment.
rig. With the circuits swapped around as shown, the The lead from J3 can be removed from the
mixer takes the full output of an HX-30 (about 2 negative terminal of the bias battery and connected
watts a -m or 6 watts ssb) without flat-topping. to the chassis, to make it easier to obtain grid
Output is several times what it was with the other current for purposes of adjustment, if necessary.
arrangement, and linearity is extremely good. Peak all adjustments for maximum grid current,
Every circuit tunes uncritically, and it is possible to making sure that this drive is on the desired
set up almost on-the -button merely by peaking the frequency. You'll need something larger than a
circuits to approximate frequencies with a grid -dip one-mA meter if everything is working correctly,
meter. or you can reconnect the bias battery once you

'
w

Fig. 7-6- Interior of the 144-MHz heterodyne exciter. 47 -and 94 -MHz circuits are at the left, the mixer
in the center, and 144 -MHz amplifier at the rioht.
A 50 -MHz Transverter 129
Fig. 7-7 - Layout drawing showing
principal hole locations and sizes, for
those wishing to make a duplicate
unit. Hole sizes: A-1 inch, B-3/4
inch, C-1/4 inch; others 1/8 inch.
Chassis and plate are 5 X 10 inches.

have obtained a reasonable current reading. Opera- characteristics similar to those of the Compactrons
tion of the amplifier from here on is exactly like it can be used. The functions of the double -triode -
would be in a conventional transmitter. pentode 6M11 can be performed by a 12AT7 dual
When the conversion unit is used for sideband triode for VIA and B, and a 6AK5 in place of
or a -m, the 6360 operates as a Class AB1 linear V1C. Or a single triode for VIA and a 6U8 or
amplifier. Thus the drive must be kept below the other triode -pentode for V1B and C will do nicely.
level at which grid current starts to flow. In driving Some differences In layout will be needed either
an amplifier like the 144 -MHz 4CX250 amplifiers way.
described in Chapter 6, it is not necessary to drive The 6AR11 was used for the mixer mainly as
the 6360 into grid current for any class of service. an economy, compared with the cost of a 6360.
On cw, for example, it is possible to develop 600 The latter dual tetrode makes an excellent push-
watts output from the 4CX250s with the 6360 pull mixer, as well as amplifier, and little or no
stage running Class AB1 (no grid current). If a change in layout is required for this substitution.
harder-to -drive final stage is used it may be A higher amplifier output level can be obtained
necessary to push the 6360 into Class -C conditions with other dual tetrodes for V3 than the 6360.
for full -power cw work. This will also be necessary This tube's larger brother, the 8458, is an excellent
if the 6360 is to be plate modulated. candidate, delivering up to 25 watts output at 400
In practice, it is convenient to use the output volts on the plates.
control on the 50 -MHz exciter as the sole means of
controlling the operation of the conversion unit, Operating Conditions
whether the mode of operation be sideband, cw, or
a-m. Keying for cw is done in the 50 -MHz exciter, With the exciter as described, any power supply
and modulation of the signal is also done there. We capable of delivering 250 to 300 volts at 100 mA
have encountered no linearity problems in the and 150 volts, regulated, should be satisfactory.
mixer or its following amplifier at any level of Some typical operating conditions are:
operation needed with the 4CX250 push-pull Oscillator plate current: 12 mA without crystal os-
amplifier running at power output levels from 50 cillating; 5 mA with.
to 600 watts. Doubler plate current: 8 mA.
The conversion unit is plugged into a power Amplifier plate current: 10 mA.
supply of conventional design. Power is left on the Mixer plate and screen current: 15 mA with no
setup during all operating time, as the current drain 50-MHz drive; up to 20 mA with maximum drive.
without 50 -MHz drive is well below the rated 6360 Amplifier plate and screen current: 25 mA
dissipation of all tube elements. with no 50 -MHz drive; 48 mA for operation as
linear amplifier; 70 mA max. for Class-C cw.
Tube and Circuit Alternatives Amplifier grid current: None, except for Class-C
Experience with this exciter has turned up operation; about 1.5 mA max.
several desirable alternatives. Perhaps you prefer Output: 6 watts cw, sideband, or plate -modulated
other tube types. If so, individual or dual tubes of a-m; 2 watts a-m linear.

A 50 -MHz TRANSVERTER

With the increase in use of ssb on the vhf bands, quate gain and noise figure, and freedom from
there is much interest in adapting hf ssb gear to use overloading problems.
on higher frequencies. The transverter of Fig. 7-8 Circuit Details
will provide transceiver-style operation on 50 MHz,
when used with a low-powered 28 -MHz transceiver. The receiving front end uses a grounded -gate
The output of the transmitter portion is about 40 JFET rf amplifier, Q1 in Fig. 7-10, followed by a
watts, adequate for much interesting work. It can dual -gate MOSFET mixer, Q2. Its 22 -MHz injec-
be used to drive an amplifier such as the grounded - tion voltage is taken from the oscillator and buffer
grid 3-500Z unit described later in this chapter. stages that also supply injection for transmitter
The receiving converter combines simplicity, ade- mixing. The difference frequency is 28 MHz, so the
130 VHF STATIONS
transceiver dial reading bears a direct 28-50 rela-
tionship to the 50 -MHz signal being received. For
more detail on the converter construction and
adjustment, see Fig. 4-15 and associated text. The
transverter uses the grounded -gate rf amplifier
circuit, while the converter referred to above has a
grounded-source, but they are quite similar other-
wise.
The triode portion of a 6LN8, VIA, is a
22 -MHz crystal oscillator. The pentode, V1B, is a
buffer, for isolation of the oscillator, and increased
stability. Injection voltage for the receiving mixer
f is taken from the buffer output circuit, L8,
through a two-turn link, L9, and small -diameter
coax, to gate 2 of the mixer, through a 10-pF
blocking capacitor.
The grid circuit of the 6EJ7 transmitting mixer,
V2, is tuned to 22 MHz and is inductively coupled
to the buffer plate circuit. The 28 -MHz input is
1 applied to the grid circuit through a link around
L11, and small -diameter coax. The mixer output,
L12, is tuned to the sum frequency, 50 MHz, and
coupled to a 6GK6 amplifier, V3, by a bandpass
circuit, L12 and L13. The 6GK6 is bandpass-
coupled to the grid of a 6146 output stage, V4.
This amplifier employs a pi-network output stage.
7-8 - The 6 -meter transverter, with shield The 6146 plate dissipation is held down during
cover in place. Large knobs are for amplifier the receiving periods by fixed bias that is switched
tuning and loading. Small knob, lower right, is
for a meter sensitivity control. The meter switch in by relay K1. The mixer and driver tubes have
is just above it. their screen voltage removed during receiving, by
the same relay, which also switches the antenna
Fig. 7-9 - Top view of the transverter. The and 28 -MHz input circuits for transmitting and
receiving converter is inside the shield at the left. receiving. The relay is energized by grounding pin 7
The 22 -MHz crystal oscillator and buffer are in of PI through an external switch, or by the VOX
the lower left portion of the chassis. In the right relay in the transceiver.
corner is the transmitting mixer. Above it is the
first amplifier. The 6146 output amplifier is in
the shielded compartment at the upper right. Construction
A 7 x x 2 -inch aluminum chassis is used for
9
the transverter, with a front panel 6 inches high,
made of sheet aluminum. The top and sides are
enclosed by a one-piece cover of perforated alum-
inum. The output -stage tuning control, C5, is on
the upper left of the panel, 2 inches above the

r chassis. The loading control, C6, is immediately

below, under the chassis. The meter, upper right,
monitors either 6146 plate current or relative
output, as selected by the switch, Si, immediately
below it. A sensitivity control for calibrating the
output -metering circuit completes the front -panel
controls.
The output connector, J2, is centered on the
rear apron of the chassis, which also has the input
jack, J1, the 8 -pin connector, PI, and the bias -
adjusting control mounted on it.
The meter is a 1-mA movement, with multiplier
resistors to give a full-scale reading on a current of
200 mA. The front cover snaps off easily, to allow
calibration marks to be put on as desired.
An enclosure of perforated aluminum, 3-1/4
inches high, 4 inches wide and 4-3/4 inches long
shields the 6146 and its plate circuit. There is also
an L-shaped shield around the 6146 socket, under
the chassis.
The receiving converter is built on a
2-1/2 x 4 -1/4 -inch etched board, and mounted
vertically in a three-sided shield of sheet aluminum.
 A 50 -MHz Transverter 131
RF AMP MIXER
.001
50 MHz MPFI02 50 MHZ 0420673
12 01
$
- D 27
0
28MHz

27
La
15
270
í
.005
12 1.6
15 1.4 10 I 150 4 J1

.001 .005
100K 1°1K1B TRANSCEIVER
270 270 270
10
m 47

v
TO K 1 F
22 MHZ .001 2w TO +12V,
Pl

OSC. BUFFER
/26LN8 22MHZ Y261.148 22MHz MIXER
V2 50 MHZ
VIA I VIB r
22 22
a ;;:1
22 VI i L7

,r
L9 1.10 L11 001
5
I TO L1

MHz 47K
8
1000
.005
22K
3 .001
22K
L12 >I'L13
e
TO
/1s
.005 IW
47K
470
.001
.001 ¡ 2200
IW
o-o-o

RFC1
m
i7s
001

W
IW
NC
560
U 12
1
KIE 11 K1F ANT.
AMP. .01 6aim TO V3 4© J2
SCREEN
66K6 RFC3 AMP XRFC4 20 23
©RFC2
TO
6I46A SOMHZ
RFCS 22 KID
C4
,YYYY'j
V3 47 100 TO
1

1
3` V4 AL to0LI61
50 MHz 6 1W K1E 50 MHz I 0011W C5 C6
RFC RFC6
t i. 22
2

.T..001 Cl_.¿10
U_ 7
.00r 1
.001

1.12:
. *'L14
CR2

--Y
---
.SOK
.001 330 PLATE tILAT SENS.
22K .001 OUTPUT ó oautpuT
47K -220 TO
RFC1 rJ,
(
12V

RFC2 OSIA SIB /p\DNv

001 I W0.001 .001 Cot, K1A
.001 9
BIAS 10K
{ 600V .W1 .oOtT-t- /{, 1
/F2
VI V2 V3 9 50K
2W
500
1-;/'
I '} r 7 1

6 5W IOW +250v(112113i415161718)P1
4 S 4 S 2 7 K1C tNC , 1-6.3VAC 1

-IOOV 1
ris CONTROL
.001'

ACEPT A5 INDICATED, DECIMAL SAI.UES 0f

CAPACITANCE ARE IN MICROFARADS (pi).
OTRCRS ARE IN PICOFARAOS (pi O11/1M11,
RESISTANCES ARE IN OWNS, 0 1000.

Fig. 7-10 - Schematic diagram and parts information for the 50 -MHz transverter.

Cl -10-pF subminiature variable (Hammarlund L7, L8, L11 -

Iron -slug coils adjusted for 4.1,
5.5, and 5.5 µH, respectively (Miller 4405).
C2
MAC -10).
-5-pF subminiature variable (Hammarlund L9, L10 2
ground ends
- turns small insulated wire over
of L8 and L11.
C3
MAC -5).
-2 -1/2 -inch length No. 14 wire, parallel to L12, L13 - 1-12H iron -slug coil J. W. Miller 4403,
and 1/4 inch away from tube envelope. Cover 3 turns removed.
with insulating sleeve. L14 -
7 turns No. 20, 1/2 -inch dia, 1/2 inch
C4 -
500-pF 3000 -volt disk ceramic. long (B&W 3003).
-
C5 -10-pF variable (Johnson 149-3, with one L15
L16
Like L14, but 6 turns.
-
6 turns No. 20, 5/8 -inch dia, 3/4 inch
stator and one rotor plate removed).
C6 --
140-pF variable (Millen 22140).
1N128 diode. P1
long (B&W 3006).
-8
-pin power connector.
CR1
-
--1N83A diode. RFC1 68-pH rf choke (Millen 34300).
CR2
J1 phono jack. RFC2 -- 8.2-pH rf choke (Millen J-300).
J2 --
Coaxial jack, SO -239. RFC3
resistor.
5 turns No. 22 on 47 -ohm 1/2 -watt
6 -pole double -throw relay, 12 -volt dc coil.
RFC4 - 4
K1
L1 -2 turns small insulated wire over ground
resistor.
turns No. 15 on 47 -ohm 1 -watt
end of L2.
-
10 turns No. 24 enamel closewound RFC5, RFC6 -
8.2-µH rf choke (Millen 34300).
L2, L3, L4
on J. W. Miller 4500-4 iron -slug form. 51 -
Dpdt toggle.
L5, L6 -
12 turns No. 24 enamel on J. W. Miller Y1 -
22 -MHz overtone crystal (International
4500-2 iron -slug form. Crystal Co., Type EX).
 132 VHF STATIONS
Fig. 7-11 - Bottom of the transverter, with the
6146 socket inside the shield compartment at the
right. Three sets of inductively coupled circuits
are visible in the upper -right corner. The first
nib two, near the top of the picture are on 22 MHz.
Next to the right and down, are the mixer plate
, - and first -amplifier grid circuits. The self-
supporting 6GK6 plate and 6146 grid coils are
just outside the amplifier shield compartment.
The large variable capacitor is the loading control.

MHz and L12 and L13 are tuned to 50 MHz. The

driver and output circuits should also be tuned to
50 MHz. Check to be sure that slug -tuned coils
really tune through the desired frequency. Quite
often troubles are eventually traced to coils where

t
the circuit is only approaching resonance as the
core centers in the winding. Such a circuit will
i appear to work, but drive will be low, and spurious
outputs will tend to be high. This is a common
-FL trouble in overtone oscillators, with slug -tuned
coils.
Once the circuits have been set approximately,
apply heater and plate voltage to the oscillator, and
tune L7 for best oscillation, as checked with a
wavemeter or a receiver tuned to 22 MHz. Connect
a 28 -MHz receiver to the input, J1, and apply dc to
the converter. It should be possible to hear a strong
local station or test signal immediately. Peak all
coils for best reception, then stagger -tune L5 and
Before mounting the converter shield, be sure to L6 for good response across the first 500 kHz of
check for clearance with the terminals on the the band.
meter. Remember, the meter has full plate voltage Before applying plate voltage to the 6146, it is
on it when the switch is set to read plate current, advisable to protect the tube during tuneup by
even when the transverter is in the receiving mode. inserting a 1500- or 2000-ohm 25 -watt resistor in
Testing of the transverter was done with the series with the plate supply. Connect a 50-ohm
General -Purpose Supply for Transceivers, described load to the output jack, and energize Kl. Adjust
in the power supply chapter of the ARRL Hand- the bias control for 25 to 30 mA plate current.
book. Separate provision must be made for 12 Apply a small amount of 28 -MHz drive. A fraction
volts dc for the receiving converter. of a watt, enough to produce a dim glow in a No.
Injection voltage, signal input, and i-f output 47 pilot lamp load, will do. Some output should be
connections to the converter are made with small - indicated on the meter, with the sensitivity control
diameter coax. These and the 12 -volt wiring are fully clockwise. Adjust the amplifier tuning and
brought up through small holes in the chassis, loading for maximum output, and readjust all of
under the converter. As seen in Fig. 4-19, the input the 50-MHz circuits likewise.
JFET, Q1, is on the left. The mixer is near the After the circuits have been peaked up, adjust
center. The 28 -MHz output coils, L5 and L6, are the bandpass circuits by applying first a 28.1 -MHz
just to the right of Q2. input and then a 28.4 -MHz input, and peaking
Note that there are two sets of relay contacts, alternate coils until good operation is obtained
KM and K1F, in series in the receiver line. This over the range of 50.0 to 50.5 MHz. Most ssb
guarantees high isolation of the receiver input, to operation currently is close to 50.1 MHz, so
protect the rf amplifier transistor. Another protec- uniform response across a 500-kHz range is not too
tive device is the diode, CR1, across the coil of the important, if only this mode is used. If the
relay. If there are other relays external to this unit 10 -meter transceiver is capable of a-m operation,
that use the same 12 -volt supply, it is advisable to and you want to use this mode, coverage up to
pet diodes across their coils also. Spikes of several 50.5 with uniform output may be more desirable.
volts can be induced with making and breaking of Adjust the position of the neutralizing wire, C3,
the coil circuits. for minimum rf in L16, with drive on, but no
screen or plate voltage on the 6146.
Adjustment
Now apply full plate voltage, with no drive, set
A dip meter is very useful in the preliminary the bias adjustment for a 6146 plate current of 25
tuning. Be sure that L7 and L8 are tuned to 22 to 30 mA. With the dummy load connected,
A 144 -MHz Transverter 133
experiment with the amount of drive needed to uous monitoring with a scope is the best preventive
reach maximum plate current. Preferably, use a measure.
scope to check for flat -topping as the drive is Because of the frequencies mixed, and the
increased. An output of 40 watts, cw, should be bandpass coupling between stages, the output of
obtainable. The quality of the ssb signal is deter- the transverter is reasonably clean. Still, use of an
mined first by the equipment generating it, but it antenna coupler or filter between the transverter
can be ruined by improper operation. Over -driving and antenna is good insurance. The same treatment
the mixer or the 6146, and improper loading of the of the transverter output is desirable when driving
amplifier will cause distortion and splatter. Contin- a linear amplifier.

A 144 -MHz TRANSVERTER

The transverter of Fig. 7-12 is designed to be

used with any 14- or 28 -MHz ssb exciter capable of
delivering approximately 20 watts peak output. It
can provide up to 20 watts PEP output at 144 MHz
- sufficient for driving a pair of 4CX250 tubes in
Class C for cw operation, or the same pair of tubes
can be operated AB1 to provide 1200 watts PEP
input with this unit as a driver.
It is not recommended that beginners attempt ``0
this project since vhf ssb circuits require special .F =

care in their construction and operation, a require-

ment that may be a bit beyond the inexperienced
builder.

How It Operates
Starting with VIA, the oscillator, Fig. 7-13, a
43.333 -MHz or overtone crystal is used at Y1 to Fig. 7-12 - Panel view of the 2 -meter transverter.
provide the local -oscillator signal for the exciter. This version is patterned after a transmitting
Output from VIA is amplified by V1B to a converter design by K9UI F. The on -off switches
suitable level for driving the tripler, V2. 130 -MHz for ac and dc sections of the power supply are
mounted on the front panel of the unit as are the
or 116 -MHz energy is fed to the grids of V3, a pilot lamps and plate meter for the PA stage. The
6360 mixer, by means of a bandpass tuned circuit, tuning controls for the various stages are ac-
L3C1, and L4C2. The selectivity of this circuit is cessible from the top of the chassis.
high, thus reducing unwanted spurious energy at
the mixer grids.
Output from the exciter is supplied through an
attenuator pad at J1 and is injected to the mixer,
V3, at its cathode circuit, across a 270 -ohm The receiving section uses a low -noise uhf
resistor. The attenuator pad can be eliminated if a MOSFET as the rf amplifier and a second dual -gate
very low -power exciter is to be used. The values MOSFET as the mixer. See Fig. 7-15. The gate 1
shown in Fig. 7-13 were chosen for operation with and drain connections of the rf amplifier are
a Central Electronics 20-A exciter operating at full tapped down on the tuned circuits so that uncon-
input, or nearly so. The driving power needed at ditional stability is achieved without neutraliza-
the cathode of V3 is approximately 4 or 5 watts tion. Oscillator energy is sampled with a two -turn
PEP. link wound over L3. A short length of RG-58A/U
After the 130 -MHz and 14 -MHz signals are carries the injection energy to Q2. The converter is
mixed at V3, the sum frequency, 144 MHz, is built in a 5 x 2-1/4 x 2 -1/-inch box constructed
coupled to the grids of V4, the PA stage, by means from four pieces of double -sided circuit board that
of another bandpass tuned circuit - further re- have been soldered on all abutting edges. The unit
ducing spurious output from the exciter. PA stage is mounted on the transverter front panel.
V4 operates in the AB1 mode. Its idling plate
current is approximately 25 mA. The plate current Construction Notes
rises to approximately 100 mA at full input.
If ew operation Is desired, the grid -block keying The photographs show the construction tech-
circuit in the mixer stage (J3) can be included. If niques that should be followed for duplicating this
ssb operation is all that is contemplated, the minus equipment. The more seasoned builder should have
100 -volt bins line can be eliminated along with J3, no difficulty changing the prescribed layout to fit
R1, and the shaping network at J3. In that case the his particular needs, but the shielding and by-
15,000 -ohm grid resistor from the center tap of L4 passing methods used here should be adhered to
would be grounded to the chassis. even if changes are made.
 r T- r
OSC. BUFFER 43.33 TRIPLER AMP.
1S 38.666MHz 116 00. Mlx'ER
6EA8 IS 6CL6 130 MHz. 1
636C 6360
144-148 MHz. 144-148 MHz.
SM I
43.33 OR VIA V2 I 1-1 V3 V4
38.666MHz
L9.i L7 .J2
3 IC
ANT

20
C6
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
M MICROFARADS Ij,FI¡OTHERS
ARE IN PICOFARADS (PF OR OFF);
RESISTANCES ARE IN OHM.Y,
k 1000.
1000
T -J
12
220K

(REG)
i-- +250V.
1
33
2W.~ 4µf 2 + ISO V (REG.)
t KEY 150V.
14 MHZ. 100
F.T . FEED THROUGH 3 -100V. (BIAS)
S.5.8. S .SILVER MICA -i-. IOK
INPUT B M.. BUTTON MICA RI BTI 15V. (REG.)
SOK iyy, 4 +150V.
BIAS ADJ SW 6.3V.
5
6 GN D.
270 Fig. 7-13
Schematic diagram of the transmitting converter
-
"Iw portion of the transverter. Fixed -value capacitors are disk ceramic 7
1 TO TI
unless noted differently. The polarized capacitor is electrolytic. 8 PRIMARY
Fixed -value resistors are 1/2 -watt carbon unless otherwise noted. 9
TRANSMIT
B1 -Small 15 -volt battery. L5 5 turns No. 18 wire, 1/2 -inch dia, 5/8-inch
- 10
TO Ti SEC.
Cl - 20-pF mniature variable (E. F. Johnson long, center -tapped. +300V
160-110 suitable). L6 3 turns No. 18 wire, 1/2 -inch dia, 5/8-inch
-
C2, C3, C5 REMOTE
10-pF per section miniature butter-
- long, center -tapped. CONTROL 10K
fly (E. J. Johnson 16721 suitable). L7 4 turns No. 18 wire, 1/2-inch dia, 1/2 -inch
-
C4 -5-pF per section miniature butterfly (E. F. long, center -tapped.
Johnson 160-205 suitable). L8 1 -turn link of insulated hookup wire,
-
C6 -20-pF miniature variable (same as C1). 1/2 -inch dia, inserted in center of L7.
11,12 117 -Vac neon panel lamp assembly.
- L9 2 turns of insulated hookup wire over L3.
- VI V2 V3 V4
J1,J2 SO -239 -style coax connector
- M1 0- to 200-mA dc meter.
J3 - Closed-circuit phone jack. P1 11 -pin chassis -mount male plug (Amphenol
--
J7 - Phono jack. 86PM11).
L1 - 15 turns No. 28 enam. wire, close -wound, R1 50,000 -ohm linear -taper, 5 -watt control.
-
on 1/4 -inch dia slug -tuned form (Millen 69058 RFC1-RFC3, incl. 2.7 -pH rf choke (Millen
-
form suitable). 34300-2.7).
L2 - 12turns No. 28 enam. wire, close -wound, on S1, S2 Spst rocker-type switch (Carling
-
same type form as L1. TIGK60).
L3 - 5 turns No. 18 wire space -wound to Y1 43.333 -MHz third -overtone crystal for
-
7/8 -inch length, 1/2 -inch dia, center -tapped. 14 -MHz input- If a 28-MHz transceiver will be
L4 -3 turns No. 18 wire, 1/2-inch dia, 3/8-inch used, a 38.667 -MHz crystal is required.
long, center -tapped.
A 144 -MHz Transverter 135
Fig. 7-14 -
Inside view of the converter. Shields
are used between the rf amplifier input and
output circuits, and between the latter and the
mixer input circuit. The cable entering the
bottom side of the enclosure carries the oscillator
injection energy. Output to the associated re-
ceiver or transceiver is taken through the jack to I1
the left.

116 OR 130 MHz

J6

144 MHz
5

144 MHz 15 40673

2 3300 02
#c7 3N200 15

'T- IS 5M
QI 11tD <
L10
L12
1 150 1

J4
i1
500
f T5 )J5
' I
28MHZ
OR 14 MHz
-04 T0 TRANSCEIVER
ANT
47K 47 117V
"s.7 AC

Q1, Q2
150K
.001 C10 T
1000

is +12V
62
EXCEPT AS INDICATED, DECIMAL VALUES OF
al CAPACITANCE ARE IN MICROFARADS 1 .1+F I ;
OTHERS ARE IN PICOFARADS 1 or OR AAF1;
RESISTANCES ARE IN OHMS;
I000, 111000 000.
5 M=SILVER MICA
*= GATE PROTECTED

Fig. 7-15 -
Diagram of the converter section. ID. Tap at 3 turns up from the cold end for
the 01 drain connection.
Resistors are 1/4 -watt composition and capacitors
are disk ceramic, except as noted otherwise. L11 -5 turns No. 18 tinned wire, 1/4 -inch ID.
C7-C9, incl. -
Air variable, pc mount (Johnton L12 - 1.99 -2.42 -pH slug -tuned coil, pc mount,
189-505-5) for 28 -MHz output (J. W. Miller 46A226CPC);
C10 -
Feedthrough type. or, for 14 -MHz output, 7.3-8.9-µH (J. W.
L9 -
4-1/2 turns, No. 18 tinned wire, 1/4 -inch
ID. Tap at 1-1/2 turns up from the ground
Miller 46A826CPC).
J4 -J6, incl. -
Phono type.
end for the antenna connection, and at 3 -
01,02 RCA dual -gate MOSFET.
turns for the 01 gate. Z1 - 12-V miniature power supply, transistor
L10 -
4-1/2 turns, No. 18 tinned wire, 1/4 -inch radio type.

Fig.
Looking
7-16
into the
-
bottom of the g
chassis, the rf
section is enclosed M
in a shield compart-
ment made from
flashing copper.
Additional divider
sections isolate the
r
input and output S ' a
tuned circuits of 11

the last three stages 4

of the exciter.
Feedthrough ca- _

pacitors are
mounted on one
wall of the copper
compartment to
provide decoupling
of the power leads.
136 VHF STATIONS
An 8 x 12 x 3-inch aluminum chassis is used off-the-air weak signal, peak L9, L10, and L11 for
for this equipment. An internal chassis, 5 inches best signal-to-noise ratio.
wide, 3 inches deep, and 12 inches long, is made The transmitter section can be powered by the
from flashing copper and installed along one edge circuit of Fig. 7-17, or the builder can design a
of the main chassis. This method makes it possible supply of his own choice. Regulated voltages are
to solder directly to the chassis for making positive recommended for best operation.
ground connections rather than rely on mechanical With a dummy load connected to J2, apply
joints. Shield partitions are made of copper and are operating voltage. Couple a wavemeter to L1 and
soldered in place as indicated on the schematic tune the oscillator plate for maximum output.
diagram and in the photo. An aluminum bottom Then, detune the slug of Ll slightly (toward
plate is used to enclose the underside of the chassis minimum inductance) to assure reliable oscillator
for confining the rf. starting. Couple the wavemeter to L2 and tune for
Feedthrough capacitors are used to bring power peak output. With the wavemeter applied to L4,
leads into the copper compartment. Though this adjust Cl and C2 for maximum indicated output.
adds somewhat to the overall cost of the project, it The next step is to connect the transceiver to
provides excellent bypassing and decoupling, thus J1 and supply just enough drive to cause a rise in
reducing unwanted interstage coupling. It also PA plate current of a few milliamperes. Tune C3
contributes to TVI reduction. Most surplus houses and C4 for maximum indicated plate current at
stock feedthrough capacitors, and offer them at M1, then adjust C5 and C6 for maximum power
reasonable cost. output to the dummy load. Cl, C2, C3, and C4
should be readjusted at this point for maximum
Tune Up plate current of the PA stage. Use only enough
An antenna-changeover relay and a set of drive to bring the PA plate current up to 100 mA
normally open relay contacts, both operated by at maximum dc input power.
the exciter, must be provided. The remote control A closed-circuit keying jack is used at J3 so that
leads, from P2, should be connected to the relay the mixer stage is not biased to cutoff during voice
contacts. With power applied to the converter, L12 operation. Inserting the key permits full bias to be
should be set for maximum noise input to the applied, thus cutting off V3. Rl should be adjusted
transceiver. Then, using a signal generator or for complete cutoff of V3 when the key is open.

S
3pp
yyyy
00,4.F
H., 450V
51
CR1
11

TI 100,ÚF +
1 100K 800 3500 c.r.
2W IOW 101V 10

CR2 + 250V
1

6.3V 0821
+ 150V
+150V
2
TO
4
6.3V EXCITER
5
3
-10ov

T2
0A2 0A2
.!T 6
6N0.

9
P.M"'
PRI.

-100V BIAS
CR3
+T 150V

II
3A
FUSE

117V AC

Fig. 7-17 - Schematic of the power supply section. On -off switches for the ac
and dc circuits are mounted in the rf deck along with the pilot lamps. Polarized
capacitors are electrolytic, others are disk ceramic CR1 and CR2 are 1000 -volt,
1 -ampere silicon diodes. CR3 Is a 200-PRV 600-mA silicon diode. T1 is a
power transformer with a 540 -volt ct secondary at 120 mA. Filament windings
are 5 volts at 3 A, and 6.3 volts at 3.5 A. T2 is a 6.3 -volt, 1 -ampere filament
transformer connected back to back with the 5 -volt winding of Ti. S1 is an
11 -pin socket (female). A 10,000 -ohm resistor and a .01-/.lF disk capacitor are
connected in series between the center tap of T1's secondary and ground for
transient suppression when S2 is switched to on. The suppressor is mounted at
S2, in the rf deck.
The A -m Transceiver -Is It Obsolete? 137
ADJUSTING DRIVE LEVEL TO VHF TRANSVERTERS

In using hf exciters to drive vhf transmitting COAXIAL

T
converters and transverters, the problem of main- H F XMTG
EXCITER CONVERTER
taining optimum drive level is often troublesome.
Many hf sideband rigs are in the 100 -watt category,
yet the more desirable mixers require no more than SHIELDED (A)
a few watts of drive. This problem is complicated LOAD '513

by the necessity for running the ssb exciter at full

load, if the quality and carrier suppression of the
signal are to be maintained. Reduction of the voice
level (to keep down the ssb peak output) is not
recommended, as this tends to destroy the effec-
tive carrier suppression in the driving signal.
Several solutions are possible. In some hf
exciters drive can be taken from a stage preceding
the output amplifier, though this is not popular
with owners of commercially built rigs, since it
requires modification of the circuitry. In some (B)
output amplifiers it may be possible to remove the Fig. 7-18 - Methods of attenuating the signal from
screen voltage, or greatly reduce it, and still get an hf exciter used with a transmitting converter
enough output to drive the vhf mixer. The most requiring low driving power. A conventional 50 -
popular method is to run the hf exciter more -or - ohm dummy load is connected across the driver
less normally, and swamp out most of its output In line, in A. An attenuator capable of dissipating up
to 15 watts ssb or cw power is shown in B. This is
a dummy load. inserted in the driver line. All resistors are 2 watts,
This approach ís shown in Fig. 7-18A. A coaxial R1 33 ohms, R2 270 ohms.
T fitting is Inserted in the line between the exciter
and the mixer, and a shielded dummy load is
connected in parallel. The input tuning of the the drive to the desired level. Where the hf exciter
transmitting mixer can then be used as a level output level is moderate, as with the Central
control to some extent, if the power is still too Electronics 10B and 20A, the attenuator of Fig.
high, or an attenuator can be inserted between the 7-1813 alone may serve the purpose, without use of
dummy load and the transmitting mixer, to adjust the dummy load.

THE A -M TRANSCEIVER - IS IT OBSOLETE?

Rapid increases in use of fm and ssb in vhf Meaning no disparagement of other modes,
communication have tended to create the impres- each of which has qualities that make it unique in
sion that amplitude modulation is obsolete. It may its field, we express here the hope that enthusiasm
be almost gone from the busy hf bands, where the for them will not cause an untimely demise of the
narrower spectrum of ssb is an overpowering a-m transceiver. In what other way could we match
advantage, but where there is plenty of room for it, the simplicity, versatility, and low cost of the two
a -m is by no means dead. And it should not be. If portable stations that follow these lines? They do
we examine all the qualities of all known forms of some things well that are difficult, expensive, or
modulation, we may well conclude that a -m has impossible with competing modes. Most important,
legitimate roles in the vhf realm that other modes they are great fun!
handle poorly, or not at all. These two portables were chosen for use here
The vhf bands have always been great territory not because they are either new or optimum
for the do-it-yourself beginner, but if vhf operation designs, but rather because they are basically
goes to more complex modes completely, at the simple, yet effective. Many other items appearing
expense of a -m, much of this quality could be lost. in QST in recent years will suggest methods you
Amplitude modulation is the only mode with may prefer in building your own "mountain -
which the ultimate in simple receivers, the superre- topper." We suggest a check on references below,
generative detector, can be used effectively. An before you start.
a-m transmitter, at least in the low-power brackets,
is easy and inexpensive to build. No mode can
match A2 (keyed -tone modulation) for on -the-air Tilton "Featherweight Portable 5O-MHz Sta-
tion," QST, November, 1964.
code practice; it is easy to generate, and easy to Utz, "Mlniwatt 2 -Meter Transceiver," QST,
copy on any receiver. There are no critical stability October, 1967.
Campbell, "A 1969 -Model 50 -MHz Transistor
problems, transmitting or receiving, in an a -m Transceiver," QST, January and April, 1969.
system. Finally, good a -m can be very good indeed, Becker "More Power on 144 MHz with Tran-
sistors, QST, August, 1969.
even with home -built equipment that is within the Preiss, A 2 -Meter QRP Mountain -Topper,"
capabilities of the average tinkerer-type ham. QST, May, 1970.
138 VHF STATIONS
TRANSISTOR PORTABLE STATION FOR 50 MHz

Though "working portable" from the high These items, a Minilog and miscellaneous small
spots has been an integral part of the vhf game tools and spare parts are carried in a zippered
since the earliest times, it remained for the age of plastic "gym bag" 5 x 9 x 12 inches in size. A
transistors to bring lightweight portable vhf sta- lightweight portable 3 -element beam that makes
tions into the realm of full practicality. The this little station "really talk" is described In
50 -MHz station in Fig. 7-19 is complete, including Chapter 9.
even the antenna system and microphone, yet it is
a mere 5 -pound handful. It delivers a good-quality Transmitter Rf Unit
voice signal of up to one watt, and its receiver will
The transmitter is shown in Figs. 7-21 and 7-23,
pick up any a -m signal that you could hear on the
with its circuit diagram in Fig. 7-22. Parts are
best home -station setup, yet the station will
mounted on an aluminum plate made from a sheet
operate for many hours on íts self-contained pack 3 x 7 inches, with 3/8 inch folded up at the
of D cells. You can run the rig from the car battery bottom. This fastens to the transceiver bottom
when it is convenient to do so, but you can also plate with self-tapping screws. The oscillator and
take it to any spot you can reach on foot and be
buffer stages use 2N706 transistors, or similar vhf
ready to operate in minutes. npn types. These are at the left side of Fig. 7-21
and the right of Fig. 7-23. Aluminum shields
What's Inside 2 x 2-1/2 inches are mounted on spade lugs at 2
The handmade sheet aluminum box is and 4 inches in from the left side, as seen in Fig.
4-3/4 x 6 x 9 inches in size. Inside are separate 7-21. Leads from L2 and L4 run through these
units for transmitting, receiving, modulation, and shields and are insulated from them by sleeves of
power, any of which can be changed without polyethylene made by removing the conductors
dismantling the others. The receiver is a small from small pieces of RG-58 or 59 coax.
imported pocket broadcast set, with a crystal - The output stage has two silicon vhf power
controlled converter ahead of it. The converter is transistors in parallel. Several types are usable, and
seen in the left foreground of Fig. 7-20. Its bargains may be found occasionally in surplus
adjustments are reached through holes in the right flyers. RCA's 2N3553, 2N3866, and 2N4427 also
side of the case, as viewed from the front. The work well and can be run at higher input than
broadcast receiver, attached to the front panel, quoted here, if drive for them is available. 2N706s
shows in the left rear corner of Fig. 7-20. At the work well in the final, but will not stand amplitude
right rear, in back of the battery pack, is the modulation voltage peaks encountered with 12
transmitter rf unit. Just above this assembly is a volts on the collectors. Nearly all silicon vhf
readymade 1 -watt audio amplifier, modified for transistors will do for the oscillator and buffer, but
modulator service. Each of these units will be the 2N706 has a higher dissipation rating than
described in detail. most. They also can be obtained for as low as five
With the transceiver, in Fig. 7-19, are the for a dollar from surplus sources.
microphone, a 35 -foot "long-wire" antenna, a The safe dissipation rating for transistors can be
dummy load for testing the transmitter, and a raised by even the simplest of heat sinks. A strip of
small antenna coupler built into a plastic parts box. thin brass or flashing copper 3/16 inch wide can be

.
1
d
1
m
...

Fig. 7-19 - The 50 -MHz

portable station, complete
with all necessary opera-
ting accessories -total
weight: under 5 pounds.
Accessories grouped
around the transceiver: the
microphone, miniature an-
tenna coupler, pilot -lamp
dummy load, and a 35 -foot
"long-wire" antenna.
Transistor Portable Station for 50 MHz 139

-
. --_
ti

01.

Fig. 7-20 - Interior of the 50 -MHz transistor transceiver, with top plate, right side and rear panel
removed. Parts of the switching circuits and the small broadcast receiver used for the i -f system are
seen on the front panel. The C -shaped subassembly at the left is one of several converters tested in the
transceiver. The transmitter rf assembly is seen in back of the package of 7 "D" cells. At the upper
part of the left -side panel is the modulator.

Fig. 7-21 - Transmitter portion of the 50 -MHz transistor rig. At the left side are the crystal oscillator
of two
and buffer stages. The two transistors in the output stage are connected in parallel by means
brass plates, which also serve as a heat sink. The amplifier collector circuit is tuned by means of
the
knob at the lower right, the surface of which is slotted to permit adjustment with a screwdriver,
through a hole in the left side of the transceiver case. The crystal and the two tuning slugs are also
provided with access holes.

t,4

(
 140 VHF STATIONS
XTAL OSC. BUFFER AMPLIFIER.

Fig. 7-22 - Schematic diagram and parts information for the transmitter portion of the 50 -MHz
transceiver. Resistors are all 1/4 watt. Decimal values of capacitance are in µF; others in pF. Suffix F
indicates feedthrough type. All others not described are Mylar or dipped -mica, 50 -volt rating or more.
The modulator is shown in outline form only, since it is a ready-made unit. Lead colors given are for
Radio Shack audio amplifier, type 277-038, having a 1 -watt rating.
C1 - 35-pF subminiature variables (Millen 25035E). Q3, 04 -
C2 -- 180-pF mica trimmer (Arco 463).
Silicon vhf power transistor, 1 -watt or
higher dissipation. See text.
C3
J1
5-µF 25 -volt electrolytic.
- Phono jack. RFC1-RFC4, incl. - 8.2-µH iron -core rf choke
(Millen J300-8.2).
L1, L3 - 5 turns No. 24 enamel, closewound on T1 - Input transformer, high -impedance micro-
1/4-inch iron -slug form. (Miller No. 4501, with phone to amplifier input, 200k to 1000 ohms
3 turns removed or wind on No. 4500 form.) (Archer 27-1376).
L2, L4 - 2 turns insulated wire wound near
bottom end of L1 and L3, respectively.
T2 - Output transformer; 45 to 50 -ohm primary,
L5 -10 turns No. 20, 16 tpi, 1/2-inch dia, center
tapped (B&W No. 3003 Air-Dux 416T, PIC
3.5 or 8-ohm secondary. Connect low -imped-
ance winding to amplifier output, and run
final -stage collector current through 50 -ohm -
1730). winding (Knight 5404147).
L6 -2 turns insulated wire around center of L5. - Third -overtone crystal, 50.11 to 54 MHz
Q1, 02 -
2N706 or equiv. See text.
Y1
(International Crystal Mfg. Co. Type F-605).

bent into keyhole shape and slipped over the adjusted for optimum transfer to a 50 -ohm load
2N706 case for this purpose. The brass plates and left set thereafter, since adjustment is not
holding the final stage transistors together (Fig. critical.
7-21) serve the dual purpose of heat sink and
parallel collector connectors. Dimensions are not Modulation
critical, but ours are 0.041 x 5/8 -inch brass, about The audio amplifier used for the modulator
1-3/4 inches long. Aluminum would be equally (Radio Shack 277-038, 1 -watt rating) has an
good. Be sure that these do not touch the output transformer with a low-impedance secon-
mounting plate or the socket-mounting screws at dary. This must be replaced with one suitable for
any point, as the collector and case are connected modulation purposes, or a step-up transformer can
together in power transistors, and thus the case has be added. We chose the latter, as it was easier to
the supply voltage on it. Center-to-center spacing find than one designed specifically for modulator
of the holes should be the same as that of the service. An input transformer to match the high -
transistor sockets, one inch in this case. impedance microphone must also be added. The
Venous output circuits were triea, with the extra transformers, Ti and T2, in Fig. 7-22, are
series-tuned center -tapped arrangement shown in visible in Fig. 7-20, mounted at opposite ends of
Fig. 7-22 working out best for this setup. Output is the amplifier. The modulation transformer is con-
taken off through a series -tuned loop, L6, wound nected "back to back" with the output trans-
around the midpoint of L5. The series capacitor, former of the amplifier, and has the collector
C2, is a high -minimum mica trimmer, visible current of the final stage of the transmitter running
directly over the tank coil in Fig. 7-23. It can be through its 50 -ohm winding.
Transistor Portable Station for 50 MHz 141

j
141
I

1
)-F
1f .

Fig. 7-23 -
Back view of the transmitter, with the output stage at the left. Partitions isolate the
three stages; crystal oscillator at the right, buffer at the middle.

The amplifier has pnp transistors, so it is set up good modulation. About 300 milliwatts output is
for positive ground, as is the broadcast receiver. In possible, with good modulation, at 6 volts.
using the transceiver in negative -ground cars (U.S.
standard) the "ground" side of the amplifier and Transmitter Adjustment
broadcast receiver must be isolated from the
transceiver case. The amplifier is mounted on an Tuneup is very simple. Checking individual
aluminum bracket, making sure that the mounting stages for current drain is desirable, and adjust-
screws do not come in contact with the positive- ments can be made at lower than rated voltages
voltage circuits of the module. Parts of the initially. Operation at 6 volts is similar to that at
amplifier circuit that connect to the positive lead higher voltages, and it may be safer in the
(brown lead in the unit used here) are bypassed to check-out phase. Apply voltage through the oscilla-
the transceiver case with an electrolytic capacitor, tor feedthrough capacitor only, at first, and check
C3, in Fig. 7-22. the current drain. As the slug in L1 is moved there
The amplifier unit is intended for 12 -volt will be a downward dip in collector current as the
service, but it works well at lower voltages. Its crystal begins oscillating, to around 10 mA at 6
output tracks with the input to the final stage of volts. At 9 volts the oscillator current is 15 to 20
the transmitter as the supply voltage is changed, so mA. Output is enough to light a 2 -volt 60-mA pilot
the modulation percentage remains about the same
regardless of the power source used.
Signal quality and modulation percentage de-
pend on many factors. With our operating condi-
tions the best modulation is obtained with audio
applied only to the collectors of the amplifier
stage, and with the final collector circuit detuned l
slightly on the high -frequency (low capacitance)
side of resonance. When tuned for maximum
output the stage- shows little upward modulation, #1,4k
and when Cl is detuned to the high-capacitance
side the quality is poor and the modulation
distinctly downward. The amount of detuning
needed depends on the collector voltage, increasing
with voltage level.
_

Output capability is about one watt at 9 volts

and two watts at 13 volts, but the stage must be r
detuned to one-half and one watt, respectively, for

Fig. 7-24 -
Front view of the converter portion of
the 50 -MHz transceiver. Core studs at the right side
are for adjusting the rf amplifier collector circuit,
r
the mixer base circuit, and the oscillator collector
circuit. The rf stage input circuit is at the lower
left.
 142 VHF STATIONS

Fig. 7-25 -Schematic diagram and parts informa tion for the transistor converter. Decimal values of
capacitors are in µF; others in pF. All are Mylar or dipped mica, 50 -volt rating or more. Resistors are
1/4 -watt composition. Parts are numbered serially following those of the transmitter.

C4 - Leads of insulated hookup wire twisted

L12 - About 8 turns No. 24 enamel, wound over
turns of built-in loopstick of broadcast receiver.
together 4 turns. See text.
J2, J3
-
-Insulated tip jack (Johnson 105-800).
Position and number of turns not critical.
330-pF capacitor also uncritical.
L7 2 turns of the inner conductor of the lead to
S1 , wound over bottom turns of L8. See text
P1, P2 -
Insulated solderless tip plug (Johnson
105-300).
and Fig. 7-26.
L8, L9, L10, L13 - 6 turns No. 24 enamel, on
Q5, 06, Q7 -
Silicon vhf transistor (RCA 40235
used; 40236 through 40240 also tried/.
1/4 -inch iron -slug ceramic form (Miller 4500, or
4501 coil with 3 turns removed). L8 is tapped
R1 - 680 and 68 -ohm 1/4-watt in series. Check
different values for optimum amplifier perfor-
at 2 turns from ground end. If made from mance.
prepared coil, unwind, clean insulation at tap -
RFC5 8.2-µH rf choke (Millen J300-8.2).
point, solder on tap, and rewind. Space out S2 -Two -pole two-position slide switch.
turns on any coils if needed to obtain resonance
within core range.
Y2 -
Third -overtone crystal, 49.5 MHz Inter-
national Crystal Mfg. Co. Type F-605.
L11 - 2 turns No. 24 enamel, wound over bottom
turns of L10.
Y3 - Same as Y2, but 51.5 MHz, or as desired; see
text.

lamp dimly, if a loop of wire is soldered to its lamp mounted in a BNC fitting is a desirable
terminals and slipped over Ll. Set the slug in Ll accessory.
for the highest output at which the oscillator starts Tune Cl and C2 for maximum lamp brilliance,
readily each time voltage is applied. at first. Recheck the settings of Ll and L3 also.
Now apply voltage to the buffer through RFC3, The lamp will light very brightly at 9 to 10 volts,
and check current drawn by Q2. It will rise as the indicating about one watt output. Peak C2 for
oscillator is tuned toward maximum output, and maximum output, and leave it that way. When
the pilot lamp load should glow fairly brightly modulation is to be applied, detune Cl on the
when coupled to L3. Adjust the stud in L3 for low-capacitance side while talking into the micro-
maximum output. Current drain will be 20 to 30 phone, detuning only enough to get a good upward
mA with the stage working correctly. modulation indication in the lamp. Note the fmal
Check the amplifier similarly, applying voltage collector current under these conditions. At full
through RFC4. The current to this stage will be output it will be 150 to 200 mA at 9 volts, with
practically nil until drive is applied, after which it detuning to 125 to 150 mA for best modulation.
is proportional to the drive level. A 6 -volt 150-mA At 12 volts the best setting will be around 150 to
pilot lamp (brown bead, No. 40, 40A, or 47) 175 mA.
makes a good dummy load when the rig is intended If you have several 2N706s, try various ones in
to work into 50 ohms. Other lamps will light up, the oscillator and buffer stages, selecting the ones
but they do not approximate 50 ohms at normal that drive the final collector current to the highest
brilliance. Solder short wires to the base and plug value at the maximum -output tuning condition.
these into the BNC fitting, or temporarily solder When the detuning procedure outlined is fol-
the lamp across the coax lead connected to L6. A lowed the resulting modulation characteristics are
The Receiver 143
at least as good as those of any small pentode or signáls, this produces Intermediate frequencies be-
tetrode tube transmitter for the vhf bands. Voice tween 500 and 1500 kHz for a signal range of 50
quality is good and "talk power" is high, as there is to 51 MHz. The broadcast receiver may not go
some inherent clipping effect that tends to prevent down to 500 kHz unless its oscillator padder is
excessive modulation and splatter. fudged a bit, but the lowest frequency usable for
voice in this country, 50.1 MHz, comes in at 600
THE RECEIVER kHz. If you don't care about tuning as high as 51.1
MHz the crystal frequency for Y2 can be modified
Use of a simple crystal -controlled converter to suit your desires.
working into a pocket broadcast receiver for the i -f Use of a crystal on 51.5 MHz for Y3 permits
and audio system gives more than adequate sensi- tuning of the first megahertz of the band ín the
tivity, and the selectivity is good. A friction -drive reverse direction on the broadcast dial. The low
vernier, to be described later, provides smooth end appears at 1500 kHz and 51 MHz is at the 500
tuning. There are weaknesses however, as in any kHz end. This provides a quick solution to image
very -simple approach. The main problem is spur- problems that may crop up locally, since image
ious responses. Image rejection is inherently low, rejection is much better at the 1500 -kHz end of
with such a low Intermediate frequency, but this is the receiver's tuning range. Mobile services around
turned to an advantage by setting up the converter 48 MHz ride through strongly as images when Y2 is
injection so that it can be on either the high or the used, but disappear when Y3 is switched in. A local
low side of the signal frequency. MARS net just below the band edge takes over the
receiver when Y2 is used, but gives no trouble with
Converter Circuit Features Y3. On the other hand, Y3 puts most of the band
The schematic diagram, Fig. 7-25, makes most occupancy in the part of the dial where tuning rate
circuit details self-evident. Most silicon vhf trans- ís least favorable. Signals in the upper half of the
istors work well in these stages. The rf amplifier, band (if there are any) appear as images in the
Q5, is a common -base stage. Its collector circuit is tuning range when Y3 is used. So it boils down to
band-pass coupled to the mixer, Q6. The mixer using whichever crystal does the best job under
collector circuit is a few turns of wire wound over conditions of the moment.
the built-in antenna (loopstick) of the broadcast Plugging other crystals in for Y3 provides
receiver. The oscillator, Q7, has one crystal (Y2) coverage of any one -megahertz segment in or near
wired to a selector switch, S2A. The other side of the 50 -MHz band. For ranges other than 50 to 51
the switch, S2B, is connected to a crystal socket on MHz the rf circuits must be repeaked for optimum
the front panel, so that crystals may be plugged in reception, but this is done readily enough by
for Y3, to do any of several jobs. The crystal moving the core studs in L8, L9, and L10.
socket is omitted from Fíg. 7-25 for simplification. Repeaking these lower in frequency gives 48 to 49
When crystal Y2 is selected by S2 the injection MHz with Y2. Running them out and switching in
frequency is 49.5 MHz. Beating with incoming Y3 gives 52 to 53 MHz.

Fig. 7-26 - Rear

view of the 50 -MHz
converter. The rf r
amplifier transistor
socket and the input 1"'
coil are isolated
from the rest of the a
converter by an
L-shaped shield, f.,.
lower right. Leads at
the top run to the
crystal switch.
Those with tips ºt-
tached plug into
jacks connected to
the mixer collector
winding on the
loopstick. Coax at
the lower right goes
to Si A.
 144 VHF STATIONS
Receiver Construction the three coils. Q6 is midway between the center
From Figs. 7-19 and 7-20 it will be seen that lines of L1 and L13 (3/8 inch up from L10), and
the broadcast receiver is mounted on the front Q7 is the same distance above the level of L13. The
panel of the transceiver, with the back of its case rf amplifier is isolated from the rest of the
removed and the speaker facing forward. No converter by means of an L-shaped shield mounted
specific dimensions can be given as there is an on spade lugs. The amplifier collector lead runs
almost unlimited variety of small receivers avail- through this shield to L9. The converter assembly
able. We recommend that one of the better types is held on the bottom plate by two self-tapping
be used; agc action and audio quality are consider- screws.
ably better in most 8 -transistor models than in the The antenna coupling winding, L7, is made
very cheap 6 -transistor ones. from the inner conductor of the RG-174/U coax
Most pocket sets use pnp transistors, and so used for the lead to S1A. Strip the braid back
have opposite battery polarity to that required for about two inches and leave the polyethylene intact
the npn transistors in the transmitter and conver- except for about 1/8 inch at the end, for soldering
ter. This poses no real problem, as the receiver to the series capacitor. Wrap the insulated conduc-
cases are plastic and there is no "ground" as such. tor around the winding in the same direction as the
We drilled holes near the four corners of the case bottom turns of L8, and solder the braid and one
for mounting. With some sets it may be necessary side of the series capacitor to a ground lug under
to install wire screening inside the speaker hole to the coil mounting. Leave some surplus length in
prevent pickup of broadcast stations, but this was the coax, so that the converter can be removed
not needed with the receiver used here. with the connection to SI A left intact for minor
A vernier drive for the broadcast receiver dial adjustments.
can be made quite simply. A 1/4 -inch panel bearing The common positive supply lead and the
(E. F. Johnson 115-255) is used with a drive shaft mixer collector lead are fitted with solderless tip
of 1/4 -inch tubing or rod. A small rubber grommet plugs (E. F. Johnson 105-300) which fit into
with a 1/8 -inch hole is stretched to slip over the matching jacks (105-800), to permit easy discon-
shaft in a position to bear against the edge of the necting for converter removal. (This maker's tiny
small circular dial of the receiver. The mounting Rib-Loc plugs and jacks would be fine here.) The
hole for the bearing can be filed slightly oval in jacks are soldered to a tie -point strip visible in Fig.
shape, to permit adjusting the pressure of the 7-20, just adjacent to the top of the broadcast
grommet on the dial. One grommet will stand up receiver. The oscillator base and collector leads
for months of operation. running to S2A and S2B are made just long enough
The converter chassis is a C -shaped piece of to reach the terminals of the switch, and must be
aluminum, cut to 2-1/2 X 5 inches and then bent unsoldered to remove the converter.
over one inch top and bottom. The physical layout PACKAGING AND POWER
is not particularly critical, except that the holes for
the three coils (left side of Fig. 7-26) should be 3/4 Presumably the transceiver could be fitted into
inch center to center. They are on a vertical line some standard -size case, but the metal work
3/8 inch in from the side of the plate, with L9 1/2 involved in making your own is not extensive. The
inch up from the bottom. Next above it is LIO, front and back panels are 4-3/4 X 6 inches, with
with the oscillator coil, L13, at the top. The rf 3/8 inch folded over on all sides. Metal size before
input coil, L8, is 1/2 inch in from the other side, bending: 5-1/2 X 6-3/4 inches. Sheets for the sides
and the socket for QS is centered approximately are cut 5-1/2 X 9-3/4 inches, and bent up to
between L8 and L9. The sockets for Q6 and Q7 are 4-3/4 X 9. Top and bottom plates are 6 X 9 inches.
along a vertical Iine 1-1/8 inches over from that of Self-tapping screws hold the case together. Access
holes for the transmitter and receiver adjustments,
and holes for the microphone jack, transmitter
CONVERTER crystal, and receiver audio gain control should be
7' INPUT, Fí9.7
located according to the parts used. Jacks for
metering in the negative lead, and for internal -
TRANSMITTER
L6. OUTPUT, F9.4

Fig. 7-27 -
Switching and power
circuits for the transceiver.
BT1 -7or 8 "D" cells in series.
J4, J5, J6 -Phono jack.
.F.^ 1 J,
Ps 0-300
(I
p6 J7 - Polarized power plug on re-
TO EXTERNAL + il ceiver battery lead (part of broad-
12 -VOLT _ cast receiver).
-
P
TO CONVERTER.
STEM W - F197 JS Coaxial socket, BNC type.
r
+.15
P R 1

.. I »I + RCVR.
P3, P4, P5, P6, P7
P8 - -
Phono plug.
Similar to J7, but polarity
P3 5i8OOFF JT
II

o .1 Coo reversed. Can be removed from

T top of 9 -volt transistor radio bat-
tery.
TO TRANSMITTER
AND MODULATOR, F(9.4
S1 -2 -pole 3 -position wafer switch,
miniature type.
Packaging and Power 145
external power selection (see Fig. 7-27) can be
mounted wherever convenient on the rear wall.
The send -receive switch is a wafer type with J4
horizontal lever action, though any small 3 -posi-
tion 2 -pole rf switch will do. The crystal switch is
an ordinary slide type. Anteníia leads are small -size
coax (RG-174/U) throughout.
The seven or eight "D" cells are wired in series STRONG TWINE
with strips of metal or stiff wire. They are in
LIGHT FLEXIBLE WIRE.
layers, wrapped with electrical tape to hold them ANY LENGTH"---..._
in place, and clamped in a wrap-around metal strip
ANY
that is screwed to the bottom plate. SUPPORT
The cells shown are inexpensive, and stand up
ANTENNA
very well. Transistors have a great advantage over COUPLER
tubes in overall efficiency, and even smaller bat- 50 -Mc.
teries can be used if light weight is the primary STATION
consideration. Mercury, alkaline, and nickle-
cadmium cells are more uniform and longer-lived Fig. 7-28 - Circuit of the antenna coupler and its
than ordinary "D" cells, but because of the application in feeding a long wire in portable work.
intermittent nature of the load, and their recupera- Tip jacks J1 and J2 may be used for a balanced -line
tive powers, the cheaper cells make a logical choice system. Any of the three jacks may be used for
for most users.* Another transistor "plus" is that, random -length long wires, merely by checking for
best reception. Peak C1 for maximum signal on
with no critical filament temperature to be main- receiving. Gain and directivity of the long wire will
tained, the efficiency of the transceiver remains depend on length and slope.
constant over a wide range of battery voltage. - 11-pF per section butterfly variable (Johnson
C1
Output drops off with fading voltage, of course, 160-211 or 11MB11).
but the quality of the signal holds up until the
batteries are almost dead.
C2 - Fixed ceramic capacitor, 39 to 68 pF. Check
with variable temporarily, if possible.
The transceiver may be run from a car battery L1 - 18 turns No. 24, 1/2 -inch diameter, 32 tpi.
or other external power source by removing the Tap at 5 turns from each end and 1-1/2 turns
jumper (P3 and P4, Fig. 7-27) and plugging P7 into from one end (B&W No. 3004).
J5. A cigarette lighter plug and cable to P7 is L3 - 2 turns insulated hookup wire around center
handy for operation from a car battery. The car's of L1.
electrical system must be negative -ground, the U.S. -
J1, J2, J3 Tip jack.
standard. In case you're worried about running 12
-
J4 BNC cable fitting. Connect J4 and rotor of
C1 with copper strip.
volts on a 9 -volt transistor radio, this has been tried
with several different types with no apparent
damage resulting. If you still want protection, it's a
simple matter to install a 9 -volt Zener regulator on
the receiver line.
cannot be used, various "long wires" are effective,
Adjustment and Use if properly tuned and matched to the transceiver
input. Wire antennas and the little plastic-case
Adjustment of the transmitter was described antenna coupler of Fig. 7-28 will be found very
earlier. Monitoring of the total drain can be done superior to the collapsible whip type of antenna so
with a milliammeter plugged into J6, Fig. 7-27. If often used with hand -carried equipment. Tilted
the meter is removed a phono plug with its wires respond to various polarizations, and they
contacts shorted (P6) is plugged into 16. A 150-mA have some gain and directivity.
lamp is a must for a dummy load, used as described Various wire lengths can be plugged into the tip
earlier. jacks connected to taps on L1. A balanced line, or
The tuned circuits of the transmitter and even an improvised V or rhombic, can be plugged
converter are broad enough so that repeaking is not into J1 and J2. Anything will work, but usually the
necessary in the course of normal use between 50 longer the better. Tune in a signal on the receiver
and 51 MHz, except for the retouching of C1 in and peak the coupler for maximum signal strength.
the transmitter. With the twisted -wire coupling The coupler can be connected directly to the
capacitor, C4, made as described, receiver response BNC fitting on the transceiver, or a length of coax
is nearly flat from 50 to 51 MHz. can be used. The support for the far end of the
As with most receivers using bipolar transistors, wire can be a fire tower, tree, building, or whatever
it is important to use a properly tuned and happens to be handy. If there is room to maneuver,
matched antenna system, to avoid overloading walk around (maypole fashion) until maximum
problems from out -of-band signals. A well -matched signal is found. Contacts have been made at
50 -MHz beam accomplishes this ordinarily, and distances up to 125 miles on several occasions
something like our 50 -MHz portable job (see employing this haywire but effective approach.
Chapter 9) is highly recommended. When a beam The above information is a condensation of a
* A review of the various types of batteries
suitable for use with transistor gear is given in QST two-part QST article, "50-Mc. Transistor Trans-
for September, 1967. ceiver, Mark II," February and March, 1967.
 146 VHF STATIONS
SELF-CONTAINED PORTABLE FOR 144 MHz
Originally described by DeMaw as "The Con-
necticut Bond Box," in August, 1968, QST, the
2 -meter transceiver of Fig. 7-29 represents the
simple approach to 2 -meter portable gear as well as
anything before or since. It can be operated from
its own D cells, from a car battery, or from an
ac -operated 12 -volt supply. It has a 3 -transistor
transmitter, a superregenerative receiver with iso-
lating rf amplifier stage, and an audio module that
serves both portions of the station. Construction is
the circuit -board type, and templates for board
"=---
layout work are available from ARRL, for 25 cents
and a stamped addressed envelope.

Receiver Section
A JFET grounded -gate rf amplifier provides
some isolation for the IGFET superregenerative
detector. The insulated -wire "gimmick" capacitor,
C19, in Fig. 7-31, provides light adjustable coup-
ling between the rf amplifier and the detector.
Quench -frequency voltage is provided by R14C26,
in the source lead of the detector. The output of
the detector passes through a quench filter con-
Fig. 7-29 - The 2 -meter transceiver is housed in a
legal -bond box. A dial-calibration chart for the
sisting of C24, C25, RFCS, and C27. L9 isolates receiver is pasted on the inside of the lid. Two
the of signal from the B+ line. Drain voltage is plastic cable clamps serve as holders for the
varied by R15, for regeneration control. R16 is the two -section 1/4 -wavelength whip antenna (inside
audio gain control. lid) when the unit is not in use. The antenna is held
When soldering the IGFET, Q5, into the circuit, together at the center by a 1/4 -inch -diameter
be sure to connect a clip lead between the tip of threaded coupling.
the soldering iron and a good earth ground. This

OSC 72 MHt 144 Wit

DBLR

J3
03 EXT
12V

BTI-BTB, INC
ESC ESC Q5 112V TOTAL)

EXCEPT AS INDICATED, DECIMAL

VALUES OF CAPACITANCE ARE

414
D S
0
CASE
ANT
W MICROFARAOStyr);
APE IN PICor*NAo,
RESISTANCES ARE IN
OTHERS
t9r ON
OHMS¡
yyr);
1000.
Fig. 7-31 -
Schematic of the 2 -meter transceiver. Fixer -value capacitors are disk ceramic except
with polarity marking, which are electrolytic. Resistors are 1/2-watt those
composition. Component
numbering is for identification of parts on the circuit-board templates. Significant
below in the usual manner. parts are listed
Self -Contained Portable for 144 MHz 147
AR1 -
200-milliwatt audio module (Round Hill L7 - 5 turns No. 22 enam. wire, closewound on
1/4 -inch dia ceramic slug -tuned form (J. W.
Associates Model AA0100').
BT1-BT8, incl. -Eight size -D flashlight cells, series Miller 4500-4).
connected and mounted inside box by means of L8 -4 turns No. 10 bare copper wire, 1 inch long,
four Keystone No. 176 dual -battery clips. 3/8 -inch ID. (The tap shown is not a physical
C10, C12 -
5 to 25-pF ceramic trimmer, Erie one; see description of C19.
822-CN or equiv. (Midget 3- to 30-pF mica L9 - Total primary winding of 500 -ohm ct
trimmer also suitable.) transistor output transformer. 8 -ohm secondary
C15 -
8- to 50-pF ceramic trimmer, Erie 822 -AN winding not used. (Argonne AR -164 or similar.)
or equiv. (Midget 8- to 60-pF mica trimmer also R15 -R17, incl. -
100,000 -ohm audio -taper carbon
suitable.) control.
C19 -
3 turns insulated wire, wrapped around low
end of L8.
RFC1 -
Miniature 50-µH choke (Millen
34300-50').
C20 -
15-pF subminiature variable (E. F. Johnson RFC2-RFC4, incl. -
Miniature 2.7-µH rf choke
160-107). (Millen 34300-2.7).
C22 -5-pF min. variable (Hammarlund
MAPC-15B, all but -one rotor and one stator
RFC5 -
Subminiature 10-mH rf choke (J. W.
Miller 73F102AF).

CR1
plate removed).
-
18 -volt 1 -watt Zener diode (used for
-
RFC5 Subminiature 10-mH rf choke (J. W. Miller
73F102AF).
J1
transient protection during mobile operation).
- SO -239 coax fitting (chassis mount). -
-
Si, S4 Spdt slide switch.
4 -pole 2 -position phenolic single -section
J2, J3 -
Two -terminal single -contact audio con-
S2
rotary wafer switch (Mallory 3142J).
nector (Amphenol 75PC1M or similar. S3 -Spst slide switch.
L1, L2 -
3 turns No. 22 enam. wire spaced to
occupy 1/2 inch on 1/4-inch-dia ceramic slug -
Y1, Y2 -
72 -MHz overtone crystal (International
Crystal Co. in HC -6/U holder.').
tuned form (J. W. Miller 4500-4'). * Round Hill Assoc., Inc., 434 Sixth Ave., New
L3 - 4 turns No. 20 bare wire, 1/2 inch long, York, N.Y. 10011
5/16 -inch inside diameter. *J. W. Miller CO., 19070 Reyes Ave., Compton,
L4 - 6 turns No. 20 bare wire, 1/2 inch long, CA 90221.
* International Crystal Co., 10 N. Lee St., Okla.
5/16 -inch ID.
-
L5 Same as L3. City, OK 73102.
L6 - 8 turns No. 20 bare wire, 1 inch long,
* James Millen Mfg. Co., 150 Exchange St.,
Malden, MA 02148.
5/16 -inch ID. Tap 5 turns from source lead of
Q4.

DET
PA 144 YNz R F AMP
C23
2K3312
02
03 L

L4. 05
3/1125

.0
C25
n
RFC5

oa
C27

aP
WO GAIN
TOOK
^"
10 K

$.MODIFICATION
(SEE TEXT)

M -2
NPUT S2D
1113754
412V WOO)
174

yK OOK J2
GAIN
R17
 148 VHF STATIONS
Fig. 7-30 - Top-chassis
layout of the transceiver.
The receiver section is at
the left. Controls for regen-
eration and modulation are
in the foreground near the
center of the chassis. The
audio module is at the low-
er right, and the trans-
mitter board is near the
panel, directly under the
loudspeaker. The home-
P, ' made heat sinks are visible
at the left end of the audio
- .: board.
n,r. _.
z ' i

will help prevent damage to the gate of the 3N128

assemblies of this general type are available, and
should static charges be present. Also, do not numerous modifications of them have been shown
handle the leads of Q5. The leads should be in QST articles of this type, including the 50 -MHz
removed from their shorting collar by means of a transceiver immediately preceding this text.
wooden toothpick. Once Q5 is soldered in place, it Because the module is designed for a positive -
should be quite safe from static-charge damage. ground bus (pnp transistors are used), it is neces-
The transmitter portion has a 72 -MHz crystal sary to "float" the entire assembly above chassis
oscillator, Ql, in which the Cl and the internal ground to prevent short-circuiting the power sup-
capacitance of the transistor control the feedback. ply. Information on the mounting techniques and
Bandpass coupling and a capacitive divider provide some modifications to the board is given later.
optimum coupling at the intended frequency and
impedance matching, for minimum harmonics in Building the Transceiver
the drive to the doubler, Q2. The oscillator and
doubler transistors are inexpensive Motorola types, The packaging of this circuit can be up to the
but other vhf transistors should be usable. The builder. In this instance a standard legal-bond box
RCA 2N3512 is a low-cost type that works well, was chosen. It measures 5 X 6 X 11-1/2 inches.
and may be easier to stabilize than higher beta The chassis and panel are made from 16 -gauge
transistors might be. A heat sink is recommended aluminum sheeting. The chassis measures
for Q3. (Wakefield Engineering NF205 used.) 11-1/4 X 4 X 1 inch. The panel is 11-1/4 X 4-3/4
Another capacitive divider, C1OC11, matches inches. After the panel holes are drilled, a coating
the output of the doubler to the base of the of zinc chromate should be sprayed on it. Then,
amplifier, Q3, with C10 tuning the circuit. The after thorough drying, a coat of spray -can enamel
amplifier is biased for Class AB conditions, for or lacquer can be added for the final touch.
maximum low -drive output. The combination L The receiver and transmitter are built on
and pi network in the amplifier collector circuit etched -circuit boards, but point-to-point wiring
provides optimum harmonic rejection. Tuning is could be used if done neatly and with short
done by C12, with C15 serving as the loading connections. AR1 is insulated from the main
capacitor. chassis to prevent short-circuiting the power sup-
Power leads are decoupled by means of C3, C9, ply. It has a plus -ground bus; the rest of
the
and C14, in combination with R4, R8, and R11. transceiver circuit uses a negative ground. A piece
The three 33 -ohm resistors also serve as current- of cardboard is mounted between the circuit board
limiting devices for the transistors. and the chassis to prevent accidental contact
between AR1 and the chassis. AR1 is bolted to the
Audio Section chassis at four points. The four mounting holes in
the main chassis contain small rubber grommets,
The audio unit, AR1, is a ready-made import
each serving as an insulator. Terminals 1 and 9 of
assembly with two input impedances, 50 ohms and the audio board arc common to its plus-ground
100k ohms, and two output Impedances, 8 and
bus. These terminals must be disconnected from
500 ohms, the latter used for modulation of the the ground bus by removing the thin copper
amplifier collector voltage. The audio assembly has connecting strip which joins the circuits.
a 200-mW rating at 9 volts, but this can be raised
to 300 or more by using 12 volts, and adding heat
To operate AR1 at 12 volts it is necessary to
add heat sinks to the two transistors nearest the
sinks to the two output transistors. Several audio
output transformer. The sinks can be made from
Self -Contained Portable for 144 MHz 149

,F
'LL19$11:-

q_.-_..'. 1
L.

7".
)

n
I
*el-'
l
.

'
_..

a
, Irs

Fig. 7-32 -
Bottom view of the chassis. The receiver board is at the right. The transmitter board is at
the upper left. A 2000-µF 15-volt electrolytic is mounted near the rear lip of the chassis.

pieces of thin brass, copper, or aluminum. They are tune in a weak signal on the air or from a signal
1-1/2 inches long, formed by warping the stock generator and tune L7 for a peak response.
around a drill bit which is slightly smaller in Chances are that when the peak is reached, the
diameter than the body of the transistor. detector will stop oscillating. If this happens,
All interconnecting rf leads are made with advance R15 until the hiss returns. If it does not,
subminiature coax cable, RG-174/U (Belden detune L7 slightly until a compromise is reached
8216). Shielded audio cable should be used for all (L7 usually loads the detector somewhat when it is
of wiring which is more than a couple of inches in tuned to the operating frequency). Alternatively, a
length. A bargain -house import is used for the 1000-ohm swamping resistor can be connected
receiver tuning dial. No slippage was noted with across L7 to reduce its effect on the detector.
the 2 -inch -diameter model used here. The next Trimmer C20 is used to set the tuning range of
smaller model is not recommended because it may C22. The turns of 18 can be spread or compressed
not handle the torque of the tuning capacitor. A for additional frequency adjustment.
2 -1/2 -inch -diameter loudspeaker is used. Its protec- A No. 49 pilot lamp makes a suitable dummy
tive grille can be made from perforated aluminum. load for visual tune-up of the transmitter, though
Two 3 -inch-long brass angle brackets, each with somewhat reactive at 144 MHz. First, determine
3/4 -inch sides, mount the panel -chassis assembly
inside the box. Two 6-32 hex nuts are soldered to
the bottom side of each bracket, directly under
No. 10 access holes. Four 6-32 X 3/8 -inch screws
hold the transceiver in place. The brackets are
attached to the sides of the box with 4-40
hardware.
Tune Up and Use
The receiver should be tested first. With an
antenna connected to J1, apply operating voltage
and adjust R15 until a rushing noise is heard in the
speaker. Do not advance R15 beyond this point as
the sensitivity of the receiver will decrease. Next,

-.
Fig. 7-33 - Eight size -D cells are series connected
to provide 12 volts. They are mounted in Keystone
holders on the back wall of the bond box. The
1/4 -inch -diameter hole in the front of the cabinet
(upper right of photo) permits final calibration of
the receiver (C20) after the installation is com-
pleted. The hole ís opposite the shaft of C20.
150 VHF STATIONS
that the oscillator, Ql, is operating by coupling a foregoing steps should be repeated a few times to
wavemeter (or grid -dip meter in the diode -detector assure maximum output. Final adjustments should
position) to Ll and look for an indication of be made with the antenna connected, and with an
output. Adjust the slug in L1 for maximum SWR indicator in the line.
output, then turn the transmitter on and off a few A sensitive SWR indicator ís needed at this
times to make sure the crystal always kicks in. If power level. One of the Monimatch indicators with
not, detune L1 slightly toward the high -frequency a 4 -inch -or -longer line (air -dielectric element type)
side of resonance until the oscillator does start can provide full-scale readings if a 100-4A meter is
each time. Next, peak L2, C10, C12, and C15 for installed. Alternatively, see QST, August, 1967, for
maximum indication on the bulb. There will be a low-power bridge. Also, see the "Monimatch
some interaction between the circuits, so the Mark II," QST, February, 1957.
 Chapter 8

Antennas and Feed Systems

Every radio station since the first one has had cautioned not to look for simple all-purpose
an antenna, but what a bewildering array these answers. There is no one antenna or feed system
skywires represent! Antenna experimentation has that "has everything." Nowhere are better an-
long been a favorite pursuit for the amateur and tennas more needed than on the vhf scene, but
for good reason. Probably by no other effort can making an intelligent choice involves more than a
the average ham so improve his results, at so little perusal of performance figures found in antenna
cost, as by putting up a better antenna. But what is manufacturers' catalogs.
a better antenna? Though numerous books have As in our equipment chapters, we will discuss
been published on every aspect of antenna design principles first, presenting information that will be
and talk about beams is heard wherever hams helpful whether the reader intends to build an
gather, there is probably more misinformation antenna or buy one ready made. This will be
about antennas than about any other subject in followed by practical examples, for the fellow who
amateur radio. wants to build his own arrays.
In the following pages we will attempt to sort
the wheat from the chaff, but the reader is

OBJECTIVES

Choice of a vhf antenna system begins with strength in the favored direction with that of an
some decisions about the type of work we want to antenna that would radiate equally in all direc-
do best, since there Is no antenna that does all tions. Such an antenna exists in theory only, since
things well. Does highest possible gain overrule it would have to be a point source. Called an
other considerations? Is broad frequency response isotropic radiator, it has a special appeal for the
important? Can we live with a sharp beam pattern, man who would have his antenna look good on
or would something broader and less critical in paper. If it has a gain of 1, a half-wave dipole has a
aiming serve our needs better? How about omni- gain of 2.14 dB. "Gain over isotropic" is a handy
directional coverage? Can we go all-out for size, or and legitimate way of stating antenna performance,
must there be some compromise with what simple and it is coming into more general use in antenna
mounts and inexpensive rotators will handle? What literature. But remember that figures so quoted are
is the nature of nearby terrain? Are there trees, more than 2 dB higher than those for the more
wires, and buildings to be cleared? Is there a familiar half -wave dipole comparison. Unless other-
neighbor problem? Let's think about these and wise stated, gains mentioned in this text are with
other points a bit before we get down to how -to - respect to a half -wave dipole, a comparison we can
build -it details. measure.

About Gain Frequency Response

Sad to relate though it may be, we should Antenna gain is often achieved at the expense
recognize the fact that antenna gain figures are of frequency coverage, especially in arrays having
often on the optimistic side. Even when given
accurately, and with the best intentions, gain
information may be in terms confusing to the
average reader. True measurement of gain is diffi-
cult; few amateurs can do it accurately. On -the -air
1.1 TO I ACROSS
TME 138UC"
/
. .

evaluation is also far from simple. It is entirely

possible to get several contradictory observations
in as many tries, and usually only long-term
comparisons will show if a new antenna is doing
what we wanted it to do. What really counts is
whether or not an antenna provides a stronger and
more consistent signal at distant points than was 'GÁ161
I
Overt
possible before. If a new installation does not yield ISOTROPIC' I
this result, impressive numbers have little meaning.
There is no magic about antenna gain. It is
achieved only by taking radiation from some
portions of the antenna's field where it may serve
no useful purpose, and putting it into areas where
it will do some good. A logical way of expressing
antenna gain, therefore, is to compare the field

151
152 ANTENNAS AND FEED SYSTEMS
usually not obtained with the same antenna adjust-
ment that gives maximum forward gain, however.
It is another of those factors that you obtain in
trade for something else. Usually the choice is
between highest gain and optimum front -to -back
ratio.
Omnidirectional Coverage
There are several ways of building up gain
without losing a circular pattern in the horizontal
plane. Compression of the vertical angle Is involved
here. Since radiation at angles other than those

TRANSMISSION-
LINE LOIS MAY NULLIFY
: close to the horizon ordinarily does the vhf man
little good, trading off the high -angle power loss
for gain along the ground looks almost like
}{EIGNTGeIN something for nothing. Omnidirectional antennas
have many uses, and we'll be describing practical
many parasitic elements. As will be shown later, systems later on, but they, too, have their draw-
the gain of even a small antenna of this kind is backs. Gain achieved without modifying both
available over only a narrow frequency range, horizontal and vertical patterns is limited to a few
whereas an array of the same stated gain built with decibels, even when high and somewhat cumber-
more driven elements may work well over all or some stacks are involved. More important to the
most of a vhf band. This point is of little concern man in the midst of high vhf activity concentra-
to the 2 -meter DX enthusiast who is interested tion, interference problems multiply as omnidirec-
only in the first few hundred kHz at the low end of tional gain is built up. Noise problems increase, and
the band, but to the fellow who wants to work heterodynes and cross -talk seem to grow as if by
effectively over a wide frequency range it could be magic.
the deciding factor in a choice between two very
different antenna systems having the same adver- Height Gain vs. Line Loss
tised gain. The broad -band array will be larger and In nearly every instance, the higher the vhf
perhaps somewhat more difficult to handle and antenna the better. Clearing obstructions in the
install, but with several times the bandwidth it may immediate vicinity is of utmost importance. Wires,
be a better investment, depending on one's opera- trees, and buildings in the line of fire can ruin the
ting habits. antenna pattern, absorb power, and aggravate TVI
and other interference problems. Putting the vhf
Pattern Shape antenna up high enough so that Its main radiation
An antenna with very high gain inevitably has a pattern is completely above nearby TV antennas
main lobe shaped like a cigar or a baseball bat. This may be one of the best TVI-prevention measures
can be both good and bad. In areas of high vhf that can be taken.
activity, a narrow antenna pattern Is fine, if it helps As may be seen from the height-gain informa-
to hold down the level of some of the local signals tion of Chapter 2, increased height may do as
you have to work through. It may be helpful in much for you as putting up a much larger antenna
nulling out man-made noise, if the sources of such at rooftop height. But how you get the added
noise lie mainly in one direction. On the other side height is important. Particularly at 144 MHz and
of the ledger, a very narrow main lobe imposes stiff higher, the added transmission -line loss may be
requirements on the rotator and direction - considerable. An antenna installed at the top of a
indicating devices, and it may keep you from hill nearby may be a useful approach, but it
hearing some choice DX that pops up a little off probably will entail some special attention to feed
your line of fire. The local rag -chewer won't get his methods.
money's worth from really sharp beams. And so it goes. Working with antennas is always
Front -to-back ratio is allied to sharpness of interesting and often rewarding, but the important
pattern, but antennas can be built and adjusted to thing is to remember that there is more to it than a
have high rejection off the back and still retain a choice that promises the most decibels for the least
broad frontal lobe. High front -to -back ratio is dollars.

TYPES OF VHF ANTENNAS

The simplest antenna commonly used in vhf you a very large sphere of influence, unless
work consists of a single driven element. It may be perchance you live on the top of the highest
called a dipole, a whip, a halo, or some more fancy mountain in your state - and maybe not even
name. It may be horizontal, vertical, or something then. We'll go into dipole design later.
in between. It always has some gain over that To build up antenna gain we do various things
theoretical isotropic antenna we spoke of earlier, with dipoles. We may make several, hang them in a
but never very much. It is handy for getting on the curtain arrangement, and feed them all in phase.
air quickly and unobtrusively, but it will never give This is called a collinear array. We'll have many
Yagi Antenna Design 153
practical examples in Chapter 9. Or we may line up a Yagi is more complex, Involving length as well as
one or more elements in the same plane with the number of elements, but a significant improvement
dipole, in front or in back of it, but not connected always means a much larger antenna. To double
to the feedline. These are called parasitic elements. the Yagi gain we must nearly double the number of
They may be directors (one or more, shorter than elements, and more than double the length. Ap-
the driven element, and placed ahead of it) or proximately 3 dB gain can be achieved by stacking
reflectors (usually only one, longer) placed in back. two similar antennas side by side or one above the
Such an antenna is called a Yagi, in honor of its other, and feeding them in phase. The spacing for
Japanese co -developer. optimum gain increases with the length of the Yagl
Both systems have their uses in vhf communica- antenna. Pairs of antennas can be stacked, as can
tion. In either, gain is related to size. To double the be pairs of pairs, and so on, but the size -doubling
gain of a collinear array (increase it by 3 dB) we requirement for increasing the gain by 3 dB makes
must use twice as many elements, in a frontal area it obvious that Increasing -gain projects reach the
roughly twice as large. The size-gain relationship in point of diminishing returns quite quickly.

YAGI ANTENNA DESIGN

Though it is possible to make a 2-element workers. There are many ways to make a good
parasitic array bidirectional by adjustment of its Yagi, particularly a long one. Numerous combina-
element lengths or spacings, the objective in tions of element lengths and spacings give almost
making a Yagi is usually a pattern that is essentially identical results. You can have an interesting time
unidirectional. Whether the parasitic element op- of it proving this for yourself, if you are experi-
erates as a reflector or a director Is determined by mentally inclined. If you aren't you can make a vhf
the relative phase of the currents in the driven and array by following the tables given herewith, and
parasitic elements. With element spacings com- be assured of good re.sults, provided you have a fcss
monly used (1/4 wavelength or less) the current in simple instruments to check system performance.
the parasitic element will be the right phase to In practical terms, the element lengths and
make it act as a reflector when it is tuned to the spacings in a Yagi array, even a long one, are by no
inductive or low -frequency side of resonance. In means so critical as some harried workers would
other words, it will be longer than the driven have you believe. If you can match the antenna to
element. To act as a director, the element must be its transmission line and transfer power to it
tuned to the high -frequency side of resonance, by efficiently, it will perform well, even if element
making it somewhat shorter than the driven ele- lengths vary by as much as 1 percent and element
ment. spacings by 5 percent from the recommended
values. The reflector length is more tolerant on the
Element Lengths and Spacings long side and the director lengths on the short side
The maximum gain that is theoretically possible of the table values. A small error in these directions
with a 2 -element parasitic array is shown in Fig. increases the antenna's bandwidth with only a very
8-1. The gain obtainable at various director and minor effect on gain. Error In the opposite
reflector spacings is also given. It is assumed that directions can ruin Yagi performance much more
the length of the parasitic element is adjusted for quickly. This is the same as saying that frequency
each change of spacing, as it must be if it is to response of a Yagi is broader on the low side than
deliver maximum gain. The curves of Fig. 8-1 are on the high side of resonance, a fact that is well
for 2 -element antennas only, and results in practice to bear in mind in deciding on an optimum
may not work out exactly as shown. To see why frequency for the array.
we look at Fig. 8-2, which shows the radiation
resistance at the center of the driven element when
the parasitic element is adjusted for the conditions
of Fig. 8-1. Note that the radiation resistance goes
very low at close element spacings.
With the low feed impedance of the close-
spaced array, the rf current is very high, so ohmic
losses go up. The bandwidth of the system goes
down, and the difficulty of feeding the antenna
properly rises. The result is that, for practical
purposes, and especially with more than one
parasitic element, some modification of these
theoretically optimum spacings is necessary for
best overall results. Thus we find the "wide - 005 01 0 5 02 025 03 035 04
spaced" array in common use where three or more ELEMENT SPACING -WAVELENGTH
elements are involved.
Exact analysis of what happens in a Yagi
antenna is difficult if not impossible, and it would
Fig. 8-1 -The maximum possible gain obtainable
with aparasitic element over a half -wave antenna
serve little use here since optimum arrangements alone, assuming that the parasitic element tuning is
have been worked out experimentally by many adjusted for greatest gain at each spacing.
154 ANTENNAS AND FEED SYSTEMS
A perceptible increase in gain can be had by
spacing the second director 0.25 wavelength from
70 the first, though the difference may not be worth
the extra boom length in some instances.
Uniform element spacing can be continued for
60
more elements, but some improvement in gain can
With PeÍ/e[ior be achieved by going to graduated spacing. Here
ó
each successive director is spaced somewhat greater
1 50
/ than the previous one, until a spacing of about 0.4
x
/
/ wavelength is reached, after which all directors are
40
gh
W
r spaced this much. A good rule of thumb for
1".--With Olre[for medium and long Yagis is:

Dipole to DI - 0.15 wavelength (X)

W 30

0
20 ;Pr Í Dl to D2 -0.18X
D2 to D3
D3 to D4
0.25 ñ -
-
0.35 to 0.42. All directors beyond D4,
i0
..- to be the same spacing.
Dipole to Reflector 0.2 - (2)
o
0 005 01 0 5 02 025 03, 035 The 5 -percent -longer, 5 -percent-shorter rule ap-
ELEMENT SPACING -WAVELENGTHS plies. For slightly more bandwidth, start with D1
at 4 percent shorter than the driven element and
Fig. 8-2 -
Radiation resistance at center of driven take off 0.5 percent for each succeeding director.
element as a function of element spacing, when the Lengths of individual elements are sufficiently
parasitic element is adjusted for the gains given in tolerant that you can change by 0.5 percent one
Fig. 8-1.
way or the other and find only a barely measurable
difference in gain. A listener at the other end of a
communications circuit would not know that
Recommended dimensions given later take this
anything had happened. Even at 432 MHz, where
into account, but remember, if you want more
half -wave elements are only about a foot long, it is
bandwidth make the reflector longer and the
possible to change all elements by as much as 1/16
directors shorter. If for some reason you want to
inch without much effect, if the matching system
try close parasitic-element spacing, make the direc-
is readjusted after the change is made.
tor nearest the driven element longer and the
reflector shorter than the recommended values, by Element Diameter
1 to 2 percent, depending on the closeness of the
spacing. This is in line with the point made above Information given in (1) applies with tubing
that frequency response sharpens with closer spa- sizes commonly used in vhf antennas. Elements 1/2
cing. to 1 inch in diameter are customarily used at 50
There is little point in pursuing the matter MHz. For 144 MHz, 1/8 to 3/8 inch is common.
further, points are covered in interesting
as these
detail in all modern editions of the ARRL Antenna
Book. A study of that manual, and some of the
references it cites, is recommended to the experi- W a,000
mentally inclined amateur. 7p0p
R
Yagis - Short and Long O
s0
400oo0

Ó3000
Element spacing of 0.15 to 0.2 wavelength is
2000
recommended for small Yagis (up to 4 elements)
commonly used in 50-MHz work. With convenient
U 1000
tubing and boom sizes the following figures apply
700
for all vhf parasitic arrays:
500
? 400
5600 W 300
Driven Element Length, Inches -
Freq (MHz) 200
(1)
Reflector - 5 percent longer é 100
Director -5 percent shorter 70
Second Director -6
percent shorter 50
40
30
QC 20
Fig. 8-3 -
Length factor for the range of con-
dúctor diameters used in practice. This curve Oti 10
applies to either quarter-wave (grounded or 0 92 0.93 0.94 0.95 0.96 0.97 0.98
ground -plane antennas) or half -wave antennas. LENGTH FACTOR (KA)
Yagi Antenna Design
Fig. 8-4

can be
-
Design information
for Yagi antennas. Curve A
shows the optimum boom
length in wavelengths for any
number of elements. Curve B
shows the maximum gain that
expected when the
design information of Curve A
is used.
co
LS

6.0

5.5

5.0

4.5

4.0

15

3.0

2s
2.0

1.5

IA
r

A
¡ `
155

20

le

14

12

10

0.5

02 3 444444 10 11 I!
NO. OF ELEMENTS
13 14 IS te 1T IS IS 20 21

For 220 and 420 MHz, we may go as small as 1/16 The elements in this array had to be extended 6
inch, and anything larger than 1/4 inch is seldom inches beyond the usual lengths before this array
used. There may be exceptions however, and worked properly. This is an increase even greater
appreciable changes in element lengths must be than that indicated by the graph of Fig. 8-3.
made for large variations from the above practice.
The larger the element diameter the broader the Gain and Size in Yagi Arrays
frequency response, so variations on the large side
may not be too critical, but when elements much As mentioned earlier, there is an optimum
smaller than standard practice are involved it is boom length for a given number of elements in a
well to check the antenna performance carefully. Yagi array. This is obtained from Fig. 8-4, curve A.
Some idea of the practical extent to which small Note that a 4 -element array is about a half
element diameter can affect element length can be wavelength long. A 6 -element beam of optimum
seen in a 50 -MHz portable beam described in length is twice as long, and an 8-element array
Chapter 9. In planning construction of antennas should be four times as long as one with 4
having unusual length/diameter ratios the informa- elements.
tion of Fig. 8-3 can be used to good advantage. What these combinations should yield in gain is
Here the vertical scale is the free -space half given by curve B. Our 4 -element Yagi is capable of
wavelength divided by the conductor diameter. just under 9 dB. The 6 -element goes over 10, and
The horizontal is the percentage of a free -space the 8 -element to about 12. There is no limit to the
half wavelength that should be used for a driven gain that can be achieved with longer booms of
element. practical length, as has been demonstrated by
The standard -practice range of element sizes experimenters willing to build Yagis 50 feet or
mentioned above represents only about 1 percent more in length for 144 MHz, but after the first 13
change in element length, from lowest to highest K to 15 elements it becomes a rather dubious
factor. The 50-MHz portable array mentioned business.
above has a factor of nearly 2000, if the diameter Optimum element placement is given in Table
near the ends of the telescoping elements is used. 8-I. As may be seen from these figures, this is not

TABLE 8-1

Optimum Element Spacings for Multielement Yagi Arrays. DE - driven element; - R

-
reflector; D director.
No.
Elements R -DE DE -DI Dl -D2 D2 -D3 D3 -D4 D4 -D5 D5 -D6
2 0.15X-0.2X
2 0.07X-0.11X
3 0.16 -0.23 0.16 -0.19
4 0.18 -0.22 0.13 -0.17 0.14X-0.18X
5 0.18 -0.22 0.14 -0.17 0.17X\ 0.23X
6 0.16 -0.20 0.14 -0.17 0.16 -0.25 0.22 -0.30 0.25X-0.32X
8 0.16 -0.20 0.14 -0.16 0.18 -0.25 0.25 -0.35 0.27 -0.32 0.27X-0.33X 0.30X-0.40X
8 to N 0.16 -0.20 0.14 -0.16 0.18 -0.25 0.25 -0.35 0.27 -0.32 0.27 -0.33 0.35 -0.42
N = any number; director spacings beyond D6 should be 0.35-0.42X.
 156 ANTENNAS AND FEED SYSTEMS
the collinear is useful mainly where ham haywire is
the order of the day. Care in design and construc-
.95A
tion pays off in performance, but the collinear is
95A
2 tolerant of amateur methods; considerably more so
than the Yagi.
1 J2 Nondirectional Collinears
Simple vertical collinears like those of Fig. 8-5
provide some gain over a vertical half-wave dipole,
without introducing directivity in the horizontal

12
.95A plane. At A we have two half -waves in phase, fed
by means of a folded half wavelength at the center.
A balanced transmission line is shown, but coax
and a balun could be used equally well. Slightly
Fig. 8-5 - Vertical collinear antennas for vhf use. increased gain and lowered radiation angle result
from lengthening the radiating portions and short-
Antennas A and B use the same total length of
wire, A being arranged as two half-waves in phase, ening the stub, as in the extended double-Zepp, B.
and B as an "extended double-Zepp." Antenna C is The total wire length is the same: 3 half -wave-
three half-wave elements in phase. All give some lengths. Three or more half -wave vertical elements,
gain over a single half -wave radiator, without kept in phase by means of quarter -wave stubs, C, is
directivity in the horizontal plane. another common omnidirectional vertical antenna.

3.5
particularly critical. These graphs and table and the
information of (1) are all one needs to design
effective Yagi arrays.
Collinear Arrays
The collinear (elements along a common line) is cc 2.5 YAGI, 9De. GAIN

one of the oldest forms of directional arrays. The

"Two Half-Waves in Phase," the "Extended 2

Double-Zepp," and other simple collinears are

almost as old as short-wave radio. They and their
larger multielement relatives that grow on some of 1.5

our best vhf antenna farms are still among the most COLLINEAR, 14 De, GAIN
useful. Because the collinear is made up of many
driven elements, with only reflectors for parasitic 429 430 431 432 433 434 435 436
elements, it is much more frequency -tolerant than FREQUENCY
a Yagi of the same gain. A 2 -meter collinear can be
Fig.
cut and matched for the middle of the band and it of a 8-7 - Comparison of the frequency responses
small Yagi antenna and a large collinear array.
will work over the entire four megahertz with A Yagi of comparable gain would have
a still
only moderate variation in gain. sharper frequency response.
Properly designed, a collinear system is easy to
feed with common types of transmission line. It
can be strung together with sticks and wire, hauled
up into a tree, and rotated by pulling on ropes - Directional Arrays
and it will work. One of the first and most Larger arrays with 4, 6, 8, or more half -wave
renowned vhf beams ever built was handled in just elements stacked side -by -side and one above the
that way. See Chapter 1. This is not to infer that other are what most vhf men think of when they
hear the term, "collinear." These may be driven
.95AJ 2 1
elements only, as in Fig. 8-6, wherein a bidirec-
tional pattern is obtained, or reflectors may be
added for unidirectional characteristics, as in sever-
al examples shown in Chapter 9. Directors can be
added, but this is seldom done. Large arrays with
directors are better arranged and fed as combina-
tions of Yagi bays, rather than as collinears.
Reflectors in a collinear array are usually
parasitic in nature, but a reflecting metal plane can
be used. This can be of sheet metal, though more
often wire mesh is used, in the interest of
decreased wind resistance. Spines of wire or small
Fig. 8-6 - Bidirectional collinear array using 6 tubing are also usable, so long as the spacing of the
half -wave elements in phase. Parasitic reflectors or spines is well under 0.1 wavelength. To be fully
a screen may be placed in back of the driven effective, the plane reflector should be at least a
elements for increased gain and unidirectional quarter wavelength larger in both dimensions than
pattern. the curtain of driven elements it backs up.
Polarization 157
An interesting comparison between the band- the Yagi exceeds this mismatch in less than 3
width of a 6 -element Yagi and that of a collinear megahertz. Had the Yagi in question been a long
array having 8 half -wave elements in phase, backed one, with a gain similar to that of the collinear, its
by a screen reflector is shown in Fig. 8-7. The Yagi, useful frequency range would have been very much
with a gain of about 9 dB, has a much sharper sharper still. A typical 2 -meter long Yagi may be
frequency response than the collinear, with a gain expected to work well over about 1 megahertz,
of 14 dB. Both antennas were matched carefully while a collinear of large size for this band will
between 432 and 433 MHz. The collinear shows an work nicely over at least 3 times as much frequen-
SWR under 1.8 over a range of 7 megahertz, while cy range.

POLARIZATION
The wave emitted by an antenna perpendicular ratios than vertical ones. This has led to adoption
to the earth is said to be "vertically polarized." of halos, turnstiles, and other horizontally polar-
Radiation from an antenna parallel to the earth's ized mobile antennas, despite the concern of some
surface is termed "horizontal." In the space age family passengers who may not be sold on the
these terms may mean nothing. Once we lose the esthetic virtues of these devices.
reference of ground there is no longer any "hori- In other vhf work not Involving mobiles noise is
zontal" or "vertical," but merely what is more still a factor, but it may or may not be predom-
accurately called plane polarization. The radiation inantly vertical in nature. In general polarization is
from any straight wire or rod is mainly plane - not an important consideration, as far as signal-to-
polarized, but it can be horizontal, vertical, or noise ratio is concerned, other than with mobiles.
anything in between. Long experience has shown that if there is any
Much of the time it is something in between so signal -strength advantage It usually lies with hori-
the horizontal -vertical argument that raged for zontal polarization. This is probably because of a
years without ever being entirely settled tends to combination of the vertical nature of some noise
be a specious one. There is no one "best" and the observed tendency of polarization to roll
polarization, and going along with what others in a over to horizontal in passing over hilly terrain. This
given area are using offers the best hope for good is hard to pin down, however, and some vhf men
vhf coverage. Because a vertical dipole or whip has with extensive experience in high -mountain coun-
an essentially omnidirectional pattern in a horizon- try insist that vertical is superior to horizontal in
tal plane, vertical antennas were employed for working with mobiles. This may well account for
most of the early vhf communication, before the the predominance of verticals in California and
days of high -gain arrays. When beams began to take preference for horizontals in most other areas,
over the burden of vhf work it was only natural to where the terrain is either open or rolling in
mount them in a horizontal position. Gradually character.
then vhf men went over to horizontal antennas, Near saturation of the country with television,
except in a few areas where mobile work was a which employs horizontal polarization, introduces
major factor. In mobile fm communication, with a factor not present when the move to horizontal
repeaters, vertical is the accepted standard In most standardization for amateur vhf work began in the
areas. late 1930s. Because polarization shift Is slight in
the immediate vicinity of the transmitting station,
Horizontal or Vertical? horizontal polarization for both home TV and
There is no consistent large difference in
coverage between horizontal and vertical, so long
as the same polarization is used at both ends of the
path. Reflections and the passing of the wave over
intervening hills modify polarization to a marked
extent. Probing with mobile antenna installations
will show the polarization shifting with a car
movement of a wavelength or less. Discrimination
between horizontal and vertical may amount to 20
dB or more, and at 144 MHz or higher it can be
found to reverse itself at times in a matter of
inches of travel. The results of this are familiar to
any vhf mobile worker, in the form of "mobile
flutter" that is so pronounced at certain car speeds,
in anything but the most wide-open terrain.
In vhf mobile communication our effective
working range is nearly always limited by noise,
mostly ignition racket from our own car and
others. Such noise tends to be vertically polarized,
so in areas of appreciable motor traffic horizontal
antennas yield considerably higher signal-to-noise
158 ANTENNAS AND FEED SYSTEMS
turned into a threaded hole. A circularly polarized
system will accept waves of any plane polarization,
as well as circularly polarized waves, so it is useful
COAX FEED,
Zo . 140
J in amateur vhf communication in areas where both
OHMS horizontal and vertical are in use. It suffers a 3-dB
penalty for its universality, however, so it is usually
not as good as matched plane polarization in such
1 circumstances.
Fig. 8-8 - Schematic drawing of a helical antenna. Two-way work involving circular polarization at
Circumference should be one wavelength, diameter both stations should be equal or superior to
0.32 wavelength, turn spacing 0.22 wavelength, matched plane polarization, and it may be used on
and reflector diameter 0.8 wavelength. paths where there is marked polarization shift, over
land as well as in space communication. There is a
polarization-matching problem with circular sys-
amateur work does increase the possibility of TVI tems also: the direction of rotation or "sense,"
of the front -end -overload variety. It should not be must be the same for both stations. A right -hand -
assumed that changing back to vertical would be a polarization wave encounters approximately the
cure-all for TVI problems however, for the causes same barrier in a left -hand -polarized antenna as
and cures of TVI are much more involved than does a horizontal wave at a vertical antenna, the
this. Furthermore, since interference that may discrimination amounting to 20 dB or more.
result from matched polarization is due to receiver A further complication is introduced in work
deficiencies and is not the fault of the amateur, he via reflection paths, as in the earth -moon-earth
should not be required to sacrifice communications route: the reflection produces a reversal of polar-
efficiency as a TVI expedient. ization sense. To receive our own signal reflected
We had interference problems when everyone from the moon, we must reverse the sense of
used vertical vhf antennas. We still have them in antenna polarization between transmitting and
areas where verticals are predominant today. TVI receiving. The problem is lessened for two stations
resulting from harmonic radiation and the all -too - communicating by way of the moon or a reflecting
common audio-rectification problem would be satellite, since one merely needs to use right-hand
largely unaffected by changing the polarization of and the other left.
the amateur antenna. Cross -polarization has Circular polarization is inherent in the helical
demonstrated no marked TVI cure-all properties in antenna, Fig. 8-8, in which the driven element is a
Great Britain, where television is vertical and coil or wire or tubing, fed at one end and usually
amateur vhf antennas are horizontal. backed up by a screen reflector or "ground plane."
Dimensions of the helix are not critical, so it is
Space-Age Polarization
useful over a very wide frequency range. Each turn
A third type of polarization is coming into of the helix is one wavelength at the midpoint of
widespread use in communication involving natural its useful frequency range. Combinations of hori-
and artificial satellites, because of the constantly zontal and vertical elements placed at right angles
varying polarization encountered in this work. to one another and fed in the proper phase also
Called helical or circular polarization, this is best produce circular polarization. The sense can be
symbolized by a screw thread, with the wave reversed by reversing the feed system in such an
boring through space in the manner of a bolt being array.

OFF THE BEATEN PATH

Collinear and Yagi antennas of conventional quad can be built with any number of elements, in
design are so universally used in vhf communica- the manner of a Yagi, or several driven elements
tion that most amateurs give little consideration to can be arranged in sets fed in phase, much like the
other systems. To a degree, antenna principles are collinear. The same basic stacking and feeding
the same regardless of frequency, so it may be to principles apply for quads as for Yagis.
the vhf man's advantage to try methods used on
other frequencies, both higher and lower than his Long -Wire Antennas
accustomed stamping ground. An amateur with a wire antenna for lower
bands never need wait for the erection of a vhf
The Quad beam before operating on 50 MHz or higher bands.
An antenna very popular among DX men on Antennas for 80, 40, or 20 are not often well
10, 15, and 20, but little used in vhf work, the adapted to vhf use, but they can always be made to
Quad has interesting possibilities. It can be built work, after a fashion. On rare occasions they may
from sticks and wire, if need be, so its cost can be be outstanding, for a long wire operated on its high
close to the ultimate low for beans. It is an harmonics has interesting properties. Two -meter
antenna that is readily adjusted, since all elements men the length of the Atlantic Seaboard fondly
can be stub -tuned. It has an appreciable frontal recall the booming signal put out for years by
area, and it is inherently a lower -angle radiator W4CLY, using a 75 -meter dipole that sloped down
than a Yagi of the same height above ground. A from the lighthouse at Cape Henry, Virginia. More
Off the Beaten Path 159
TABLE 8-I1
Dimensions of V and Rhombic Antennas for Vhf
Use. Columns 1 and 2 are for V designs. For Rhom-
bic. use 1, 3, and 4.
Freq Side Length V Over-all Length Width
(MHz) "A" in Feet Angle "B" in Feet "C" in Feet
50.5 58 60° 96.5 65.5
145 58 35' 109 39
28.7 68 70° 101.5 84
50.5 68 55° 106.5 70.6
145 68 35° 129 41
50.5 106 42' 192.5 91.5
145 106 35° 205 47.5
28.7 136 52° 237.5 133
50.5 136 37° 252.5 102

often, the low -band wire will show its best proper- corner angles 45 to 180 degrees (flat plane) is given
ties in the least -useful directions, but there's no in Fig. 8-9. Gain with the flat -plane reflector
harm In trying. remains nearly constant from 0.1 to 0.25 wave-
Principles that make the V and rhombic useful length, so it can be seen that varying the spacing
on lower frequencies still apply at vhf. If designed may be a convenient way of accomplishing an
for lower bands these antennas will not have impedance match.
dimensions that are optimum for 50 or 144 MHz, The parabolic reflector produces a very sharp
but they can be pressed into service in a pinch. and clean pattern, if it is large in terms of
With side lengths and angles adjusted for a vhf wavelength. A reflector diameter of about 10
band they may do very well. A rhombic large wavelengths is the minimum for appreciable focus-
enough for appreciable gain at 50 MHz may fit on ing effect, which is the basis of the system. This
a residential lot, and if it can be aimed to take care means about a 25 -foot "dish" for 432 MHz which
of major activity areas it may be worth a try. may look like the hard way to develop an
Unterminated, the rhombic is bidirectional, which outstanding signal at that frequency. Where the
may help in this. A main problem with wire arrays reflector can be set up at or near ground level, as
is getting them high enough to make them really for moonbounce work, a sizeable installation is
pay off in vhf work. Practical V and rhombic well within the capacities of the kind of workers
dimensions are given in Table 8 -II. who are apt to band together for a group project in
this field.
Plane and Parabolic Reflectors Because of constant improvement in reflector
Looking higher in frequency, the vhf man can design for military and scientific needs, some large
borrow techniques from uhf practice. Plane, cor- reflectors have become available to amateurs
ner, and even parabolic reflectors begin to be through surplus channels. Several of the larger
attractive at the upper end of the vhf range. Large amateur installations have used surplus dishes, but
nonresonant reflector systems offer broad frequen- other individuals and groups have demonstrated
cy response, clean pattern, and noncritical adjust-
ment, but from the standpoint of gain for a given

MEWL
size, they are not outstanding. A corner-reflector
array having a gain of 10 dB, for example, is larger
and more difficult to erect than a Yagi or collinear

AMR
of the same gain.
The flat -plane reflector backing up collinear
elements may have more potential. A light frame
covered with chicken wire, window screening, or
hardware cloth, with sets of elements for two
bands on opposite sides of the structure is a
convenient way of operating on 220 and 420, or
420 and 1215 MHz with one rotating array. Except
as it affects impedance, spacing of the corner or
flat -plane reflector from its driven elements is not
AIM"
r NW
"AMU
particularly critical. The impedance of the driven
element for various spacings (D) from the vertex of

Fig. 8-9- Feed impedance of the driven element

in a corner-reflector array, for various corner angles f
of 180 (flat sheet), 90, 60, and 45 degrees. "lf IN WAVELENGTHS
 160 ANTENNAS AND FEED SYSTEMS
that construction of a suitable parabola is not practical. In general therefore, the 1215 -MHz band
beyond the realm of possibility. is the dividing line above which the parabola
A reflector as small as 6 feet in diameter can be becomes a thoroughly practical approach. For
pressed into service at 420 MHz, but it will have 2300 MHz and higher it is practically standard
large minor lobes and low gain. A Yagi or collinear equipment, and even a 4 -footer works very well
array of equal or better performance is more from this point on up.1

STACKING PROBLEMS

In stacking horizontal Yagis one above the

other on a single support, certain considerations
apply whether the bays are for different bands or
for the same band. As a rule of thumb, the
minimum desirable spacing is one-half the boom
length for two bays on the same band, or half the
boom length of the higher frequency array where
two bands are involved.
In the stacked two-band array of Fig. 8-10, the
50 -MHz 4-element Yagi is going to "look like
ground" to the 7 -element 144 -MHz Yagi above it,
if it has any effect at all. It is well known that the
impedance of an antenna varies with height above
ground, passing through the free-space value at a
quarter wavelength and multiples thereof. At one-
quarter wavelength and at the odd multiples
thereof, ground also acts like a reflector, causing
considerable radiation straight up. This effect is
least at the half-wave points, where the impedance
also passes through the free-space value. Preferably, Fig. 8-11 - Approximate horizontal patterns of a
32 -element 2-meter collinear, showing the effect of
then, the spacing S should be a half wavelength, or
multiple thereof, at the frequency of the smaller increasing spacing between the inner element ends.
antenna. The half-the -boom-length rule gives about Pattern C (solid line) is with the element ends two
inches apart, the procedure normally used in such
the same answer in this example. For this length of arrays. Pattern B resulted when the spacing was
2 -meter antenna, 40 inches would be the minimum increased to 1/4 wavelength. Pattern A was taken
desirable spacing, but 80 inches would be better. with 5/8 -wavelength spacing between inner ele-
The effect of spacing on the larger array is ment ends. Note that the main lobe is longer and
usually negligible. If spacing closer than half the sharper with the wider spacings. Minor lobe con-
boom length or a half wavelength must be used, tent also increases, and this is a limiting factor in
the principal thing to watch for is variation in feed bay spacing in all types of arrays. A bay spacing of
impedance of the smaller antenna. If the smaller 5/8 wavelength is optimum for short Yagis, as well.

antenna has an adjustable matching device, closer

spacings can be used in a pinch, if the matching is
adjusted for minimum SWR. Very close spacing
and interlacing of elements should be avoided,
unless the builder is prepared to go through an
extensive program of adjustments of both element
lengths and matching.

Stacking for Gain

In stacking bays for the same band fed in phase,
the minimum spacing for appreciable gain is a half
wavelength for Yagis of up to four elements or so.
Fig. 8-10 - In stacking Yagi arrays one above the For such small Yagis, and for dipoles and omni-
directional systems such as the Big Wheel and the
other the minimum spacing between bays, S,
should be about half the boom length of the turnstile, a spacing of 5/8 wavelength will give
smaller array. Wider spacing is desirable, in which appreciably more gain. This is convenient in that
case it should be a half wavelength, or some an electrical full wavelength of coax may be used
multiple thereof, at the frequency of the smaller
array. If the beams shown are for 50 and 144 MHz, 1 A comprehensive discussion of parabolic an-
S should be 40 inches minimum, with 80 Inches tenna theory and practice, prepared by WA9HUV,
preferred. Similar conditions apply for stacking is available from ARRL for $1.00 per copy. A
bays for a single band.
condensation appeared in QST for June, 1971, p.
100.
 Stacking Problems 161
for phasing. We'll get into phasing and feed
problems later.
As bay spacing is increased in directional arrays
the main lobe becomes sharper, but minor lobe
content also increases. This becomes self-defeating
if carried too far. Small Yagis spaced a half
wavelength show a beautifully clean pattern, but
only moderate gain from stacking. For Yagis up to
two wavelengths long, a bay spacing of one
wavelength is good, though minor lobes are quite
rfl
pronounced when individual bays have 6 elements
or less.
For arrays of more than two wavelengths, keep
that half -the-boom -length minimum in mind, but
space them wider if you can. It can be seen from
this that stacking of long Yagis makes for large and
ungainly structures, but gain never comes easily
once you get into the upper brackets.
Stacking Yagis one above the other increases
gain without sharpening the horizontal pattern - Fig. 8-12 -Arrays of several driven elements
usually a desirable objective. In stacking another should be fed at the center of the system, so that
pair beside the first two, the optimum spacing currents will be balanced about the feed point.
depends on the length of the bays. One wavelength Array at the left was ineffective until the feed was
center -to-center is ordinarily used with booms less changed to the center connection, as in the
than one wavelength long. The half-the -boom- right-hand sketch.
length minimum applies with longer ones.
Pattern effects with stacking are illustrated in
Fig. 8-11, made with two 16 -element collinears function properly. The author learned this the hard
mounted side by side. Note that the pattern is way years ago with a curtain array of 8 vertical
markedly sharper with each wider spacing between half-wave elements in phase. This bidirectional
halves of the array. The gain is also higher, but system was first erected as shown at the left in Fig.
minor-lobe content increases rapidly at the wider 8-12. This was desirable mechanically, but the
spacings. Pattern A, made with 5/8 wavelength array worked very poorly. Changing the feed point
between the inner element ends, shows excellent to the center of the phasing system corrected the
gain, but the pattern is extremely sharp, and minor current unbalance, and turned this admittedly
lobes are larger than for B, which was made with rather haywire arrangement into an effective vhf
the bays spaced to leave 1/4 wavelength between array.
inner element ends (3/4 wavelength between bay The more driven elements there are in a phased
centers). system the more difficult it is to keep them in
Going beyond 5/8 wavelength would result in balance. Thus it is often desirable to break up a
no improvement in gain, for the minor lobes would large driven system into several sets of elements,
be much larger. These would grow from here on, at
interconnected with phasing lines. The 48 -element
the expense of the main lobe. A likely compromise 432 -MHz array mentioned above is an example.
between the maximum obtainable gain and the risk Ordinarily no more than 8 elements should be in a
of large minor lobes is between 1/4 and 1/2 single set, and breaking these up into two sets of 4
wavelength between inner element ends, depending. each may be better. See Fig. 8-13.
on what the builder wants most from his effort.
Also apparent from these patterns is the fact
that with large collinears, as with large Yagis, it is
not to our advantage to fill up an array with
elements. Spacing out the inner element ends to a
half wavelength probably nets as much as putting
another set of elements in the space between the
bays. The 48 -element collinear array for 432 MHz,
Fig. 9-64, is probably at least as effective as would
have been a 60-element collinear of the same
frontal area - and the former is much easier to
build and feed properly.

Phasing and Feedpoint

Fig. 8-13 - In phasing large arrays no more than 8
elements should be connected to one line terminal,
Arrangement of phasing lines and the point of as at A. Even with 8 half waves in phase, it may be
connection of the main transmission line are desirable to break the systems up into two parts, as
important factors in the performance of large at B, joining their midpoints with a phasing line.
The phasing harness so used should be a half
arrays. Balance of currents about the central wavelength or multiple thereof each side of the
feedpoint is the critical point here, as the driven main feed point. The universal stub, Fig. 8-180, is
elements must be in phase if the system is to very useful for feeding such a system.
 162 ANTENNAS AND FEED SYSTEMS
TRANSMISSION LINES

The best antenna is of little value if it cannot be per hundred feet at 144 MHz and 2.3 dB at 420
made to accept power from the transmitter or MHz, but losses go up markedly in wet weather,
transfer signals it intercepts to the receiver. Thus, and performance is very erratic. Flat ribbon gives
selection of the right transmission line and an the most trouble, but even the best tubular line
effective method of matching it to the antenna are will show fluctuating loading in heavy -rain con-
of utmost importance. These factors are more vital ditions. Cheap lines with small conductors and thin
to the vhf man than to the occupant of lower insulation should be avoided entirely, unless the
frequencies, for even with the best lines losses run line is to be indoors or no more than a few feet long.
higher in vhf installations than in the 80- Book figures make open-wire line look best of
through -10 station. It is easy to waste more than all. If a good open line has only 0.2 dB loss at 144
half our transmitter power in heating up the MHz, why doesn't everyone use it? Even at 420
transmission line, and still more can be lost in MHz, the loss per 100 feet can be under 1 dB. This
radiation from it that should have gone on to the picture has the biggest "ifs" of all, however. Such
antenna itself. Many 144 -MHz installations are at fine results are achieved, if ever, only under the
least this bad, and on higher bands power and most carefully controlled conditions. The con-
received-signal losses may run up to 90 percent, ductors must be large, yet spaced closely so that
with some lines that are fairly common in vhf radiation from the line will be negligible. Wire
circles! alignment must be kept constant, yet with a
minimum of insulating spreaders and supports.
Coax, Twin -Lead, or Open -Wire
There can be no sharp bends in the line, and it
There are three principal types of transmission must be positioned so that it is balanced to ground.
line commonly used in vhf installations today. These conditions definitely are not met in most
Each is obtainable in many styles and sizes, and amateur installations. We use TV -type lines, with
each has its strong and weak points. There is no too -small conductors and spacings generally too
one "best" line, or we would not still be using all wide, at least for 420 MHz. There are spreaders
three. Choice of the right one begins with the every few inches. The line is often run close to a
line -loss information, Table 8-IIl, but this is by no metal tower or eavetroughs, with little or no
means the whole story. These figures are for new consideration of balance to ground. Nearly always
lines, properly installed, and used in dry weather. there are bends of a sharpness that can be very
Under average amateur -station conditions losses harmful. One 220 -MHz line installed with reason-
will almost certainly be greater than the table able care and using half-inch spaced open TV line
indicates. showed a measured loss of 4 dB in a 125 -foot run.
Coax has relatively high loss in the tables. This represents a transmitter power loss of 60
RG-8, perhaps the most commonly used line, percent, yet it was probably a better-than -average
reputedly has a loss of about 2.5 dB per hundred amateur installation.
feet at 144 MHz -if the line is working perfectly. The potential low-loss qualities of open line can
At 420 MHz, the same line, in new condition and be realized in amateur work if sufficient care is
perfectly installed, will dissipate 70 percent of taken in the construction and use of the line. Large
your transmitter power and received -signal strength conductors are a must; never less than No. 14, and
in a 100 -foot run. Discouraging as these figures No. 12 or larger is better. Spacing must be close in
may seem, they are not the whole story. Trans- terms of wavelength; not more than 1 inch at 144
mitter power loss can be made up to some extent MHz and proportionately less at higher frequencies
by increasing power, at least up to the legal limit, if at all possible. Teflon is preferred for spreaders,
but in receiving the signal lost can never be and they should be several feet apart. If bends
recovered. must be made, keep them to very obtuse angles, or
Good coax, on the other hand, is tolerant of in a continuous arc of large radius.
installation. It is almost impervious to weather Baluns (about which more later) should be
changes, and it can be installed anywhere. Tape it made and used with care. A 100 -foot straight run
to a steel tower, or bury it; let it wrap around the of No. 12 enamelled wire, spaced 5/8 inch center -
tower and unwrap again as the beam is rotated - to -center with Teflon spreaders every 6 feet, fed
the loss will stay the same, almost regardless of with baluns at each end, was measured for loss,
conditions that adversely affect other types - of including baluns, at 144, 220, and 432 MHz. It
lines. A prime advantage of coax that is often showed 1.1, 1.35, and 1.56 dB, respectively, on
ignored is the fact that it permits measurement of these frequencies. By comparison a 1/2 -inch TV
the system performance readily, and with fairly line tested on 432 MHz under identical conditions
inexpensive equipment. You can measure your showed a loss of 2.3 dB. These losses are somewhat
SWR and line loss, and the effects of any adjust- higher than those of Table 8 -III, but they represent
ments are immediately apparent. This is not easy the best that can be expected in a practical
with other types of line. amateur installation. They also demonstrate the
Twin -Lead is inexpensive and convenient to worth of good open -wire line, when it is used
use. Its advertised losses look good on paper, properly. If the line must be long, a good open -
compared with coax. The best grade of tubular wire installation is probably the best way to do the
Twin -Lead, transmitting type, is quoted at 1.25 dB job at moderate cost.
 TABLE 8 -Ill CHARACTERISTICS OF COMMONLY USED TRANSMISSION LINES

Conductor Z0 Velocity Coax OD Attenuation in dB/100 ft Power Rating, Watts

Size Ohms Factor Inches MHz:50 144 220 420 MHz: 50 144 220 420
Type of Line
400-600 0.975 0.13 0.25 0.5 1 Over 1 kW
Open wire' 12

400 0.95 0.3 0.75 1 1.8 Over 1 kW

Open-Wire TV Líne, 1/2-inch2 18
450 0.95 0.3 0.75 Not recommended Over 1 kW
Open-Wire TV Line, 1-inch2 18

Parallel -Conductor Solid -Dielectric

Twin-Lead3
0.82 0.85 1.55 1.9 2.8 Over 1 kW when dry
Standard Flat (214-056) 7/28 300
0.82 0.85 1.55 1.9 2.8 Over 1 kW when dry
Tubular (214-271) 7/28 300
Tubular, Transmitting Type (214-076) 7/26 300 0.82 0.68 1.25 1.6 2.3

0.82 0.5 1 1.3 2 Over 1 kW

Extra -Heavy Flat (Federal K-200)2 7/22 200

Coax, Solid -Dielectric

20 53.5 65.9 0.195 3 6 7 15 350 175 125 90
RG-58/U4
RG-59/U 22 73 65.9 0.242 2.3 4.2 5 8 500 250 180 125
7/21 52 65.9 0.405 1.5 2.5 3.5 5 ' 1500 800 650 400
RG-8/U
7/26 75 65.9 0.405 1.5rá 2.8 3.7 5 1500 800 650 400
RG-11/U
RG-17/U 0.188 52 65.9 0.87 0.5 1 1.3 2.3 4500 2300 1900 1200

Foamed RG-8A/U 7/21 50 75 0.405 1.22 2 2.75 3.9 1500 800 650 400

Aluminum -Jacket Foamflex

0.117 50 75 0.435 0.85 1.5 2 3 2200 1200 900 600
3/8-inch5
0.162 50 75 0.60 0.65 1.2 1.5 2.3 3000 1600 1100 800
1/2-inch5
3/8-inch5 0.077 70 75 0.435 0.82 1.5 1.9 2.9 similar to
1/2-inch5 0.108 70 75 0.60 0.62 1.2 1.5 2.3 50 -ohm types

Loss figures neglect

'Spreaders at least 3 feet apart. Maximum spacing between conductors 1-1/2 inches for 50 MHz, 1 Inch for 144, 3/4 inch for 220, 1/2 inch for 420.
radiation.
2Estlmated loss, neglecting radiation.
3Numbers with 214 prefix are American Phenolic Corp.
4 Wlth all coax listed except RG-58; letter A, B, or C after number signifies noncontaminating jacket. With 58, only RG-58/U has this type jacket.
SNot Including vinyl Jacket.
 164 ANTENNAS AND FEED SYSTEMS
Tips on Selecting Coax
bought at moderate prices on the surplus market.
Coaxial line comes in two principal imped- It gives ,a constant impedance through the con-
ances: 52 and 72 ohms. There are small variations nection and can be had in all types required.
either side of these nominal figures, but they are of Propérly installed, it is weatherproof.
no significance for our purposes. Other impedances Series C fittings provide constant impedance
are available, but are seldom found in amateur and are weatherproof. In addition, they are quick-
installations. From the standpoint of overall effect- disconnect and very handy on that account.
iveness there is no preference between the above However, they are not on surplus and are quite
impedances, but practical factors tend to make the expensive.
52 -ohm types the more useful. Most test equip- The BNC Series is nice, but too small for the
ment is set up for 52 ohms, for example. On the RG-8-size line. The Type HN is a constant -
other hand, 72 -ohm coax and a balun of the same impedance series, for the larger sizes of coax.
material provides a good match to the 300 -ohm Whatever series you select, be sure that the
balanced load that some vhf antennas represent. installation job is done properly. Water leaking into
Losses in any line are related to conductor sizes fittings will ruin the best system in short order. A
and types of insulation. The small sizes of coax, sprayed coating of lacquer, and tape wrapping,
with inner conductors of No. 22 wire or smaller, help to prevent moisture absorption.
are bound to have high losses, regardless of quality
or price. An inner conductor of No. 14 wire or its G- Line
equivalent in stranded wire is about the minimum
that should be used, except for short runs. Coax Most uhf amateurs are aware that there is a
like RG-58 or 59 is convenient, but it should never single -conductor transmission line, invented by
be used for vhf applications where the run is more Goubau, and called "G -Line" in his honor. Papers
than a few wavelengths. There is no easier way to by the inventor appeared some years ago, in which
waste power and lose receiving effectiveness! seemingly fantastic claims for line loss were made;
Any coax costs money and good coax is quite under 1 dB per hundred feet in the microwave
expensive, but all things considered the best may region, for example. Especially attractive was the
turn out to be a good investment. Cheap coax is statement that the matching device was broad-band
likely to be old, and its measured loss may be in nature, making it appear that a single G -Line
higher than figures given in the table. More installation might be made to serve on both 420
important, older types of coax and some inexpen- and 1215 MHz.
sive new ones deteriorate quite rapidly when used When uhf TV first appeared on the scene, a
G -Line kit was put on the market. Mainly because
outside. Be sure to find out whether or not the
coax of your choice will stand up in outdoor of its high cost (about $30.00, plus installation) it
service. "Non -contaminating" is the word for it. never sold well for home TV use, but G -Line has
since come into íts own in cable TV systems. Here
Coax guaranteed for 15 years of use, underground
very long lines must be run, and losses must be
or otherwise exposed, is now available at moderate
cost. held to a minimum, so the G-Line principle looks
more attractive.
Coax is available in infinite variety. Worth
looking for is the "polyfoam" version of standard The basic idea is that a single conductor can be
an almost lossless transmission line at ultra -high
types. These cost slightly more than solid-dielectric
types, but losses are typically one-third less. Watch frequencies, if a suitable launching device is used.
A similar launcher is placed at the other end.
the velocity factor. An electrical half wavelength
will be a greater portion of a physical half Basically the launcher is a cone-shaped device
wavelength with foam or other low -density dielec- which is a flared extension of the coaxial feedline.
tric than with solid. In effect, the cone gets the energy accustomed to
Various lines are made with semiflexible sheath- travelling on the inner conductor, as the outer
ing, usually aluminum, and with spiral wrap or conductor is gradually removed. The inner conduc-
foam insulation. These axe fairly costly, but they tor should be large and heavily insulated. No. 14,
deliver excellent results and are fine for permanent vinyl covered, is supplied with the kit.
installation. Flexible sections for rotation are Since the kit was designed for home TV use,
needed with these, and a good way to handle a the small end of the horn launcher has a balun of
multiband installation is to put in a remotely sorts for conversion from unbalanced to balanced
operated coaxial switch to permit the use of one line. This can be removed for amateur purposes
line for all antennas. and the system fed directly with 72-ohm coax. The
G-Line is very sensitive to bends. If any must be
About Coaxial Fittings made, they should be in the form of an arc of large
radius, this being preferable to even an obtuse -
If you go to the expense of a good coaxial line, angle change in the direction of run. The line must
it is approaching the ridiculous to pinch pennies on be kept several inches away from any metal and
the fittings to be used with it, particularly on 220 should be supported with as few insulators as
MHz and higher. The so-called "UHF" fitting isn't possible.
to be trusted in the uhf range, especially if you A 100-foot run using direct 72-ohm feed to the
want to be able to measure antenna and feed-line launchers measured for loss at 432 MHz showed
performance with any degree of accuracy. 2.7 dB, which may have been mainly in the
Probably the best fitting, for most of us, is the launchers, since they were much too short to be
series N, a constant-impedance type that can be really effective at this frequency. Theory states
Transmission Lines 165
Driven element
Fig. 8-14 - Flexible
sections for rotatable
arrays. Coax may be
Driven element used, as at A. If the
coax section Is any mul-
tiple of a half-wave-
length, the antenna im-
pedance will be repeat-
'300- ohm
. ed at the bottom end.
Twin -Lead may be used
Twin'Lead Hq or odd multiple for either as a Q section or
'Q Section; k or ony as an impedance repeat-
multiple for repeating er, as shown in B.
impedance
or any multiple
Support
here-"-

450 -ohm fine

Coaxial
bolus

(B)
Support`
here

tin Coax can be taped to a tower, so long as there

is no abrasion to cut through the insulation. Sharp
Transmission bends are best avoided, but only for mechanical
line or
matching section
rather than electrical reasons. Where coax must
swing free, as in the portion that will rotate with
(A)
the antenna, be sure that enough slack is left to
assure free rotation without additional strain. An
extra turn or two around the tower, near the point
that the cones shou d be at least 3 wavelengths of attachment to the beam, is usually desirable.
long, and the kit type is less than one wavelength Make all supports extra strong, to take care of
at 432 MHz. Since loss in the line itself is
extra loads imposed by ice and wind.
presumably very low when properly installed, the Properly handled, coax makes the best available
G-Line idea should be useful where very long runs
rotating section for antennas that are fed with
are required in uhf and microwave work.
other types of line that may be more critical as to
Practical Line Installations proximity to metal. Open -wire lines are partic-
ularly susceptible to breakage or shorting out
It is one thing to quote losses for a straight unless special precautions are taken. Usually some
Twin -Lead or open -wire line, without bends or form of insulated flexible line is connected be-
insulating supports, and well away from metal or tween the antenna proper and a stationary support
semiconducting objects such as trees, roofs, and at the top of the tower or mast on which the
walls. It is quite another to put up a practical antenna is mounted.
installation for an amateur station, where a line Such a flexible section can take several forms,
must be run from inside the house, be fastened to a and it can be made to do double duty as a
tower part way up, and then allowed to swing free matching device. Probably the most satisfactory
as the antenna is rotated. The inevitable losses and method for arrays that are not to be fed directly
mechanical troubles that result from compromises with coax, is to use a flexible section of coax with
inherent in the amateur approach, particularly with baluns at each end, as shown in Fig. 8-14A. If the
rotatable arrays, make a strong case for coax. But flexible section is made any multiple of a half
with any line these problems must be dealt with, wavelength electrically the impedance of the array
and how we handle them can make a good many will be repeated at the bottom of the flexible
decibels difference in our signal reports, sending section.
and receiving. A similar method is to use Twin -Lead for the
If coax is used it is best to support it frequently rotating section, as shown in Fig. 8-14B. The
throughout the run and not depend on strain to 300 -ohm tubular transmitting-type line is recom-
keep it up out of harm's way. Burying coax is fine, mended. Here again, halfwave sections repeat the
provided that it is the noncontaminating variety. antenna impedance at the bottom end. Such a
Lines of this type have a letter following the rotating section can also be made any odd multiple
number. Example: RG8A/U is the noncontam- of a quarter wavelength, to act as a Q section,
inating version of RG8/U. The letter may be A, B, giving a step-down between a 450 -ohm open line
or C, depending on other characteristics. Most coax and a 200 -ohm antenna impedance. More on these
made today is noncontaminating, but the buyer applications will be found in the text relating to
should watch this point in picking up "bargains." matching devices that follows.
 166
IMPEDANCE MATCHING
We know, or can determine, the impedance of
the transmission line we want to use. If we knew
the impedance of the antenna with equal certainty,
matching one to the other would be a simple
matter and one of our major vhf antenna problems
would be solved forever. Unfortunately, the actual
impedance of an antenna is subject to so many
variations that it is seldom possible to put a precise
value on the impedance the transmission line must
work into. Some kind of adjustable matching
device is, therefore, a very useful tool.
Matching systems are many and varied, but all
perform one basic function: that of impedance
transformation, so that the feedline will "see" an
impedance similar to its own, regardless of the
actual antenna impedance. Matching may be com-
bined with other functions, such as conversion
from an unbalanced line (coax) to a balanced load
(center -fed antenna element). Matching may be
included in the phasing lines connecting the bays
b

of stacked arrays. The matching element may also Fig. 8-15

About Antenna Impedance

This was discussed briefly earlier, but to review,
a half-wave dipole in free space has an impedance
of about 72 ohms. When the dipole is close to
ground, or objects that simulate ground, its imped-
ance changes. In the first half wavelength from the
ground up, the impedance swings from a few ohms
near ground, through the free-space value near 0.25
wavelength to as much as 100 ohms at 0.3
wavelength, and then back to 72 ohms at the
half-wave point. Beyond here it drops off to 60
ohms and rises through 72 ohms again to nearly 85
ohms, then drops back to 72 ohms again at one
"n
'1.
NI
NI
r
ANTENNAS AND FEED SYSTEMS

IIv
...-.-.
,0,1
Ill '.1
I1/I1=

ob as e7s is tzs u 17s zn zn z

HEIGHT ADOvE GROUND-WAVELENGTHS

-
Variation in radiation resistance of a
be used to tune the system to resonance. We'll get horizontal half-wave antenna with height above
to examples of all these methods shortly, but first perfectly conducting ground.
a bit more about what they are going to be called
upon to do.
changes in the impedance of the driven element, to
the point where it is almost impossible to predict
what the feed impedance of a Yagi array will be.
The best course, then, is to make the antenna first,
determine its impedance by estimate or experi-
ment, and then make a matching device to fit the
requirements. If we can make a reasonable guess at
the impedance, we can make an adjustable match-
ing device of small range that will do the job.
If our antenna is just a half-wave dipole, Fig.
8-16A and B, we can assume 72 ohms, knowing
from the curve of Fig. 8-15 that it cannot vary
much more than 30 ohms either way. Adding a
-
reflector will bring the impedance down to 40 or
wavelength. The effect of ground on impedance 50 ohms, on the average. Putting on directors will
becomes relatively insignificant beyond two wave- lower it further, to something around 20 ohms. All
lengths, as shown in Fig. 8-15, but it can be seen these are for a feed point of the split dipole, A. At
that in situations most hams encounter in putting the center of a dipole that is unbroken, Fig. 8-16B,
up antennas the impedance of a dipole is anything the rf voltage between the element and ground is
but a sure thing. zero. This point can thus be grounded, as in
Ground is only one factor. Adding parasitic all -metal arrays, and the impedance matched by
elements drops the impedance, but how much is tapping the line out on the element in various
anyone's guess, especially in arrays with both ways.
reflector and director elements. Length, diameter, Rf voltage and impedance at the ends of
and spacing of these elements can effect great half-wave elements are very high. So is the feed

[ow Z +
(72 ~IS) Nigh Z
Nyl. z

Fig. 8-16 - The halfwave dipole, A, is fed at its center, the point
free space, and at certain heights above ground, this impedance isof lowest impedance. For a dipole in
72 ohms. Rf voltage on a halfwave
dipole is shown by the curved line in B. Since there is no voltage ground
dipole, this point can be grounded to the metal support. Rf voltagetoand at the center of an unbroken
two collinear dipoles in phase, as at C. impedance are high at the ends of
Impedance Matching 167
impedance of two dipoles fed in phase at their ing factors, including that of coupling between
inner element ends, Fig. 8-16C, the simplest elements, but 200 ohms is a good starting point for
collinear array. The feed impedance of an "H" setting up a matching system for this type of array.
array of four half-waves in phase is somewhere All these assumptions are valid approximations
around 600 ohms. The popular vhf collinear only for the frequency at which the system is
16 -element array (8 halfwaves in phase as in Fig. resonant. If the array is out of tune all bets are off.
8-13, but with reflectors) gets down to around 200 We then must have some means of tuning the
-
ohms maybe! Remember that there are modify - system before we can match it.

COMMON MATCHING METHODS

We will not describe all kinds of matching

systems, but will consider only those commonly
used in vhf work, or those that should get more
attention.
any
The Delta unPéda,

First there is the delta or Y-match, Fig. 8-17A.

Here the transmission line is fanned out and tapped
onto the driven element at points equidistant from
the center. The taps can be adjusted until an
impedance match is achieved, and then fastened
permanently in place. One of the first impedance -
matching devices ever employed, it still has its
merits, not the least of which is simplicity. Chief
Fig. 8-17
C

-
i e S
D

The transmission line and antenna

impedances may be matched by tapping the
Ba/anredlitre
any unpe.lr.

fault is the likelihood of radiation from the feedline out on the dipole in various ways. The
fanned -out portion, if it is not properly propor- delta or Y -match is shown at A. A variation for
tioned. It is also quite frequency sensitive. coaxial feed, using a balun, is given at B. The
The delta works well with a balun made of gamma match, C, is popular where coax feed is
coax, or an antenna coupler of some kind. A used. The T-match, D, may be fed with balanced
coaxial balun connected at the base of the delta is line, or through a balun as in the case of B.
shown at B. If this is made of 72-ohm coax there
could be a 300-ohm line of any convenient length
between the balun and the delta. Adjustment is arm. Without it the gamma system cannot be made
very easy when the delta is combined with coax to work perfectly, as a slight unbalance is always
feed. You merely insert an SWR bridge in the present. The gamma arm is usually made of tubing
coaxial line near the balun and adjust the delta side of about the size of the driven element, and a
length and spread for zero reflected power. If the sliding clip is used between the two, to facilitate
balun or balanced line is connected directly to the adjustment. The capacitor can be at either end of
delta as shown in Fig. 8-17A and B, the lines can the arm.
be of any impedances commonly available. More Once the proper value is found for Cl, it can be
on baluns below. removed and a fixed capacitor substituted. An
assumed value for your line can be taken, and only
Gamma and T -Match the point of connection of the arm made adjust-
Variations of the tapping-out idea are seen in able. Suitable fixed values for 50 ohms are as
the gamma and T-match, C and D of Fig. 8-17. The follows: 50 MHz -
65 pF, 144 MHz 20 pF, 220 -
gamma is fine for coaxial feed, while the T is most MHz - 15 pF, 432 MHz - 8 pF.
often used with balanced line. A balun and coaxial
feed could be used with the T, of course, just as
with the delta. The series capacitor, Cl, is used to
tune out the inductive reactance of the gamma
Soo-ak7n Any balance
A lute or balwr. !Liu wait mutable
of 7s-dun Lint Jspole nitro
Fig. 8-18 -A single conductor may be bent as at A
to form a folded dipole, giving an impedance four
times that of a simple split dipole. It may thus be
fed with 300 -ohm balanced line, or 72 -ohm coax 3.H) OILMS "low la.l
and a balun. Higher impedance step-up can be
achieved by making the unbroken portion of the
dipole of a larger conductor, as at B. A quarter - Q section.
wavelength matching transformer, or Q section, is C z= soo,3oa
= .375 oiunr
shown at C. A matching device that is useful for
any balanced load is the universal stub, D. The goo -alum
transmission line can be coax or balanced line, any lute,
impedance. length.
168 ANTENNAS AND FEED SYSTEMS

Jill-i
element can be made for 50 or 144 MHz by using
e

6_i A Y /44 /,' the nomogram, Fig. 8-19.

Where the spacing between the portions of this
type of dipole is an appreciable portion of a

\=
wavelength, as it must be at 220 or 420 MHz, the
information of Fig. 8-19 is no longer reliable. A
better method of matching arrays for these fre-

WE
.
quencies is to use the universal stub, Fig. 8-18D, or
the Q section, Fig. 8-18C. For more on matching
:p Yagis for 220 and 432 MHz, see practical examples
in Chapter 9.

"
Bu= 2
- -4, A problem with folded dipoles is that one must
know the impedance to be matched in order to
design one to do the job. Educated guesses may
come close enough for most practical purposes.

i.:_
For example, if we assume the feed impedance of a
ó =:i.- Yagi array to be 20 ohms, we can use a folded

as-
VIM
dipole with a 15 -to -1 step-up as the driven element,

~11"
and feed the array with 300 -ohm Twin -Lead. The
"`"11 mismatch will be slight, even if the dipole imped-

1/
04

,, ance turns out to be 15 ohms, or 25 ohms, instead

as

02

5_
`,,,' dz
j1 dl
of 20. The SWR will be only about 1.2 to 1 in
either case. We could use a 10 -to -1 dipole and a
50 -ohm balun equally well.
The folded dipole is easy to make, and it is
somewhat more frequency tolerant than some
other matching systems. It is very useful in
1

F«° `cw+i stacked-Yagi arrays having open -wire phasing lines.

0.1
1
1
ü
1

2 3
I
4
II
5 75 10 12315 2D A fairly high value of dipole impedance is desirable
S/d2
Fig. 8-19 -
Impedance step-up ratio for the
two -conductor folded dipole, as a function of
19

conductor diameters and spacing. Dimensions d1,

d2, and S are shown on the inset drawing. This
information is not reliable for use on amateur
bands above 148 MHz.

Strictly speaking, series capacitors should be

used with the T system too, but since omitting
them does not upset the balance of the dipole, as it
would with the one-sided gamma, they are not
always used.
140
Folded Dipole
One of the most commonly used matching 13
devices is the folded dipole, shown in various forms
in Fig. 8-18. When a single conductor is bent 12
4-
around as shown at A, the impedance seen by the 1/
N.
transmission line is quadrupled. Thus a folded o é .
00
dipole made from one size of conductor through-
out has an impedance of 4 X 72, or 288 ohms, and 0 4 b m lo
it can be fed with 300 -ohm line, or with a balun
and 72-ohm coax, without appreciable mismatch.
`b
po . \ QOtiPO

'
2'0 O
The dipole element can be made from a piece of 90 r'17a =o
Twin -Lead, with each outer end shorted and one
conductor broken at the midpoint, for connecting
eo
the transmission line. This is a convenient arrange-
ment for temporary or indoor use.
Additional impedance step-up can be obtained 70

by making the unbroken portion of the dipole of

larger cross-section than the fed portion, as shown 60
in Fig. 8-188. This is widely used in parasitic
arrays, where the feed impedance is nearly always 50
much lower than 72 ohms. Impedance step-up .15 .2 .3 .4 .5 .6 .7 .9 .9 LO
INSIDE DIAMETER (INCHES)
depends on the ratio of the conductor sizes, and on OE OUTER CONDUCTOR

the spacing between them. If the approximate

impedance of the antenna is known, a suitable
Fig. 8-20 - Characteristic impedance of typical
air -insulated coaxial lines.
 Common Matching Methods 169

Wire
1 size
700
%/.
mo,00,r0°11.r
Fig. 8-22 -
Clip for use in adjusting the point of
connection of a balun, or the adjustable short of
Fig. 8-18D, made from a piece of perforated
600 aluminum. Balun leads are soldered to the lug.
When the adjustment process is completed, the clip
may be removed and the connection soldered

500 41.,'' permanently to the line.

inches apart, or two 1/4-inch rods 2-3/4 inches

apart, to show two examples.
An adjustable Q section is a convenient way of
riØo matching two impedances that are known only
approximately. Two 1/4 -inch rods can be made to
provide impedances of about 210 to 400 ohms, by
varying their spacing from 3/4 to 3 inches. The
300 system can be used to step up as well as down, and
it works with coaxial or parallel conductors. We'll
have examples later.
too The Corrective Stub
Probably the most useful matching device of all
is the universal stub of Fig. 8-18D. Because the
too -
2 3 4 5 6 78910 12 matching stub must be a half wavelength or more
.8 15

SPACING (5) INCHES, CENTER TO CENTER to start, it is cumbersome at 50 MHz or lower, but
it is ideal for 144 MHz and higher bands. No
Fig. 8-21 -
Characteristic impedance vs. conductor
size and spacing for parallel -conductor lines.
impedances need be known to utilize it, and within
limits the system to be matched does not have to
be resonant. The short on the line is adjusted to
here, but the exact value is not important, as resonate the system to be fed, and then the
matching will be taken care of where the main transmission line is tapped onto the stub at the
transmission line connects to the phasing section. matching point. The load can be any impedance,
the stub can be any convenient wire or tubing size,
The Q Section and any spacing. The feedline can be coaxial or
Aquarter wavelength of transmission line has balanced, any impedance. A balun is used with
the property of acting as a matching transformer coax as indicated in the sketch. The shorting bar
between two different impedances. Such a trans- can be grounded, and the unused portion of the
former is called a Q section, and an example is stub cut off, once adjustment is completed.
given in Fig. 8-18C. Here a 300 -ohm dipole is Two variables are involved, which complicates
matched to a 500 -ohm line by using a Q section the adjustment procedure a bit, but with a stand-
whose impedance is equal to the square root of the ing -wave bridge in the line the job is quite simple.
product of the two impedances to be matched. A You merely move the position of the short and the
375 -ohm section is required here, but the principle point of connection of the transmission line until
may be applied to many vhf matching problems. zero reflected power is in dicated on the SWR
The impedance obtainable with various conductor bridge. Coupling at the transmitter is then adjusted
sizes and spacings is obtainable from Fig. 8-20 for for the desired loading.
coax, and 8-21 for balanced lines. Our 375 -ohm Q Where the point of connection of a balun or
section could be two No. 10 wires spaced 1-1/4 shorting bar must be made adjustable, a small .clip
of perforated aluminum, Fig. 8-22, is handy for a
Fig. 8-23 -
Circuit and parts information for the
vhf antenna couplers.
Cl -
100-pF variable for 50 MHz, 50-pF for 144
MHz (Hammarlund MC-100 and MC -50).
C2 -35-pF per -section split -stator variable,
0.07 -inch spacing (Hammarlund MCD-35SX).
Reduce to 4 stator and 4 rotor plates in each
section In 144 -MHz coupler for easier tuning; diameter, 1/8 -inch spacing. Slip over L2 before
see text. mounting.
-
---
J1 Coaxial fitting, female. L2 50 MHz: 7 turns No. 14 tinned, 1 -1/2 -inch
J2 Two -post terminal assembly. diameter, 1/4-inch spacing (Air Dux No. 1204).
L1 50 MHz: 4 turns No. 18 tinned, 1 -inch Tap 1-1/2 turns from each end. íV
diameter, 1/8-inch spacing (Air-Dux No. 808T) 144 MHz: 5 turns No. 12 tinned, 1/2-inch
inside L2. diameter, 7/8 inch long. Tap 1-1/2 turns from
144 MHz: 1-1/2 turns No. 14 enam., 1 -inch each end.
170 ANTENNAS AND FEED SYSTEMS
MHz - -
25 pF, 220 MHz 15 pF, and 420 MHz -
10 pF. A variable capacitor used for Cl should be
chosen so that these values can be reached with
some to spare. Often a fixed capacitor of approxi-
mately the above value will suffice, adjustment
then being made entirely with C2.
For adjustment of the coupler an SWR bridge
should be connected in the coaxial line between
the antenna changeover relay and J1. The two
Fig. 8-24 -A balun for working from coaxial to capacitors are then adjusted for zero reflected
balanced line is shown at A. Impedance at the power, as indicated on the bridge. If this results in
balanced end, top, is four times that of the coaxial unsatisfactory transfer of power from the trans-
line used. The loop is an electrical half wavelength. mitter, the loading control in the transmitter
Its resonant frequency may be checked with a dip should be readjusted for maximum forward power
meter as shown at B. on the bridge meter, Do not adjust the antenna
coupler for maximum forward power reading;
always set it for zero reflected. This applies in any
temporary connector. The holes are already made, matching adjustment.
and with some tension on the clip the edges of the Connected as described, the antenna coupler
aluminum bite into the conductor slightly, assuring will aid in reception, reducing the strength of any
good contact. Small Fahnstock clips are also out -of-band signals before they reach the receiver,
useful. When adjustment is completed, remove the where they might otherwise cause overloading and
clip and solder the connection permanently, using other spurious responses. The coupler is also an
the same overall lead length. effective filter, attenuating any unwanted frequen-
cies present in the transmitter output, before they
Making and Using Baluns reach the antenna.
As its composite name implies, a balun is a The coupler can be connected at any point
device for working between an unbalanced line between the transmitter and the antenna where the
(coax) and a balanced line or load. It can take conversion between the unbalanced and balanced
several forms, some of which also include the lines is desired. Because of the need to retune the
function of impedance matching along with the coupler for appreciable frequency excursions, it is
unbalanced -to -balanced conversion. usually mounted within easy reach of the operating
The Antenna Coupler Probably the most ver- position.
A versatile 2 -band antenna coupler (Trans-
satile of baluns, the antenna coupler, Fig. 8-23, can
be made to work from any impedance of coaxial match) of the type shown in Fig. 8-23 is described
in detail in Chapter 14. It is also capable of
line at Jl to any impedance balanced load at J2.
The low-impedance input circuit, L1C1, is series working between unbalanced lines of different
resonant at the operating frequency, and induc- impedance, and it includes a built-in SWR indica-
tively coupled to the balanced circuit, L2C2. The tor.
balanced output, connected to J2, is tapped down 4-to -I Baluns Broad-band baluns of several
on L2 an equal amount from each end. types are readily constructed. Bifilar-wound coils
Component values in the antenna coupler are can be used in the same manner as on lower
not critical, and it will handle a wide range of frequencies, but this method is seldom used above
impedance combinations merely by adjusting the 30 MHz. The most common balun for vhf service is
capacitors. Changing the tap positions on L2 made from an electrical half-wavelength of coax,
extends the range of impedances still further. The usually the same type as used for the main
values of Ll and Cl should be roughly those that transmission line, folded back on itself and con-
have inductive and capacitive reactance equal to nected to the main line and the antenna as shown
the value of the coaxial line impedance. Since the in Fig. 8-24. This balun provides an impedance
value of capacitance is the more readily estimated, step-up of 4 to 1, while handling the unbalanced -
it is customary toaim for this and adjust the size of to -balanced conversion.
Ll to resonate with it. Approximate values for the The physical length of the balun loop will vary
-
various bands are as follows: 50 MHz 65 pF, 144 with different types of coaxial lines. With solid -

BALANCED LOAD

-NO CONNECTION

Tv4 Fig. 8-25 - Two types of baluns, for

conversion from a coaxial line to a
(A) (B) balanced load. The coaxial sleeve, A,
is the preferred type, and it also can
\CONNECT be used as a matching device, as
COAXIAL
TOGETHER described in the text. Both serve the
LINE same main purpose: prevention of
current flow on the outer conductor
of the coaxial line.
Common Matching Methods 171

Fig. 8-26 - Coaxial -sleeve balun for 144 MHz, showing the parts that
make up the air -dielectric matching section.

The impedance at the end of the balun is 4

TABLE 8 -IV times that of the coax used. A balun with 52-ohm
coax will match a 200 -ohm load, and 72 -ohm coax
Lengths for the decoupling sleeve, A, and and a balun match a 300 -ohm load.
copper pipe outer conductor, B, for 144, 220, 1 -to -1 Baluns The unbalanced to balanced con-
and 432 MHz. version can also be made without an impedance
A B change, using the balun shown schematically in
144 MHz 19-3/4" 20-5/16" Fig. 8-25A. Here a split dipole or other balanced
220 MHz 12-1/2" 12-29/32" load is fed directly with coax. This would make for
432 MHz 6-1/4" 6-31/32" unbalance and rf flow on the outer conductor of
the line, but for the detuning sleeve (or bazooka)
that has been added to the last quarter -wavelength
dielectric coax the loop will be about 65 percent of of the line. Being open at the top and shorted to
the free-space value for a half wavelength. Less the outer conductor at the bottom, this sleevl
dense insulation such as foamed polyethylene may presents an infinite impedance to rf at the resonant
increase this to as much as 80 percent of the frequency, effectively choking off current flow and
free -space value, so it is well to check the loop for preventing radiation from the line.
resonance. Using the length of leads that will be A similar effect can be achieved with the
involved in the eventual connections, short the bazooka in B, wherein a quarter-wave line section
ends as shown in Fig. 8-24B, and couple the dip shorted at the bottom is formed by connecting an
meter to one end. additional piece of coax or tubing as shown. This is
The coaxial balun is cumbersome for use below less effective than the sleeve method however, and
50 MHz, and it ceases to be practical above 450 is seldom used above the 50 -MHz band. It is used
MHz. At the lower frequency the loop can be occasionally for vhf mobile antennas wherein it
rolled into a coil if desired. Usually it is taped to
the main transmission line in U shape. It is best to TABLE 8-V
run the balun perpendicular to the load, which
would mean dropping it vertically from the boom Inner conductor wire sizes to be used with
of a horizontal Yagi. The main line can be looped 9/16 -inch ID copper pipe outer conductors, for
back to the boom and taped in place, if this is various impedance -matching jobs commonly
desirable mechanically. A permanent position for encountered in vhf work. The impedance of the
the balun with respect to its load is particularly main coaxial transmission line, Zs, is given in
important at 220 or 420 MHz. The loop for the the left column. Next is the balanced load, Z1.,
latter frequency is only about 8 inches overall, so to be matched.
mechanical variations can throw the balance the Z5, ohms Z,, ohms Wire Size, A. W.G.
loop is supposed to provide quite a bit off. 50 72 4
Small coax such as RG-58 or 59 is not 50 200 10
recommended for baluns. Soldering weakens and 50 300 12
distorts the insulation, making shorts likely, and 50 450 18
the small conductors break very easily. Losses in 75 200 12
small -coax baluns often run prohibitively high. 75 300 18
RG-8 or similar sizes are much better, if the balun 75 450 24
is made and mounted with care.
172 ANTENNAS AND FEED SYSTEMS
1

911
Fig. 8-27 -
Characteristic impedance of coaxial
60 matching sections for various conductor diameter
ratios. The outside diameter of the inner conductor
I

70
60i..---, and the inside diameter of the outer conductor are
So used.

w Ab40 balun. Examples are the beer-can baluns of Figs.

8-26 and 28, made by K6HCP and WA6GYD.
20
These are assemblies to which the main coaxial line
run is attached by means of standard coaxial
fittings. By making the inner portion of the balun
of the right combination of conductor sizes it is
10 made to act as a coaxial Q section. Construction of
9
e
the balun is detailed in Fig. 8-28. Lengths of the
i sleeve and coaxial section for 144, 220, and 432
6 MHz are given in Table 8 -IV.
S I The inner coaxial portion of the balun can be
4
made'td continue the line impedance, or transform
it to other impedances. If standard wire sizes are
3 used for the inner conductor, load impedances
from 70 to 450 ohms can be matched by using
2 9/16 -inch ID copper pipe for the outer conductor.
The matching combinations for various wire sizes
are given in Table 8-V. You can choose your own
combinations for various Q -section impedances by
0 50 f00 ISO 200 230 3f
using the graph of Fig. 8-27. Remember, the
CHARACTERISTIC IMPEDANCE,'ZO
formula for finding the needed Q-section imped-
ance, Zo, is
may be convenient for feeding a split driven
element directly with coax. In this application the Zo = Zs Zr (3)
bazooka is usually a piece of small coax similar to where Zs is the impedance of the main trans-
the main line. Taping the two pieces together mission line, and Zr is the impedance to be
leaves the issue of true electrical length somewhat matched.
in doubt, and other feed methods are generally The load to be matched can be either balanced
preferable. or unbalanced, but if it is the latter the outer sleeve
Impedance -Matching Balun Since the sleeve is not needed. The diameter of the detuning sleeve
,assembly of Fig. 8-25A is a quarter -wavelength is not critical; it just happens that beer cans work
long, it is a simple matter to make it serve well as a out conveniently. Let's say we want to match 50
Q section for impedance matching, as well as a to 300 ohms, a frequently encountered situation
that cannot be handled with a flexible balun. From
formula (3) we find that we need a Q section with
BEER CAN
DECOUPLING SLEEVE an impedance of 122 ohms. Fig. 8-27 tells us that a
b/a ratio of 7.5 is needed. With standard 9/16 -inch
ID plumbing copper pipe and a No. 12 wire we can
take care of this job nicely. Many other usable
combinations can be worked out, using pipe and
TEFLON WASHER wire sizes that are readily available.
1/16 THICK

QN
:- .5<
COPPER PIPE
5/6 O.D.,9/16 I.O.

TEFLON WASHER

BRASS BOTTOM PLATE 1/1 BRASS

DRILL B TAP FOR 4-40 SCREWS

SOLID COPPER WIRE

Fig. 8-28 - The beer-can balun in exploded

N CONNECTOR

form. Sleeve and outer conductor lengths 7-4-40 SCREWS

are given in Table 8 -IV. The wire inner
conductor should be approximately 1
inch longer than B.
Feeding Stacked and Phased Arrays 173
FEEDING STACKED AND PHASED ARRAYS

If individual bays of a stacked array are A double Q section of coaxial line is illustrated
properly designed they will look like noninductive in Fig. 8-29C. This is useful for feeding stacked
resistors to the phasing system that connects them. bays which were originally set up for 52-ohm feed.
The impedances involved can thus be treated the A spacing of 5/8 wavelength is optimum for small
same as resistances in parallel, if the phasing lines Yagis, and this is the equivalent of an electrical full
are a half wavelength or a multiple thereof. The wavelength of solid-dielectric coax, such as
latter point is important because the impedance at RG-11/U. If our phasing line is made one quarter-
the end of a transmission line is repeated at every wavelength on one side of the feed and three
half wavelength along it. quarters on the other, one driven elemeht should
In Fig. 8-29 we have three sets of stacked be reversed with respect to the other to keep the rf
dipoles. Whether these are merely dipoles or the currents in phase. If the number of quarter-
driven elements of Yagi bays makes little differ- wavelengths is the same on either side of the
ence for the purpose of these examples. Two feedpoint, the two elements should be in the same
300 -ohm antennas at A are one wavelength apart, position, not reversed as shown in C. See "5 -over-
resulting in a feed impedance of approximately 5 for 50 MHz," Chapter 9.
150 ohms at the center. (It will be slightly less than One marked advantage of coaxial phasing lines
150 ohms, because of coupling between bays, but is that they can be wrapped around the vertical
we can neglect this for practical purposes.) This support, taped or grounded to it, or arranged in
value holds regardless of the impedance of the any way that is convenient mechanically. The
phasing line. Thus, we can use any convenient type spacing between bays can be set at the most
of line for phasing, so long as the electrical length desirable value, and the phasing line placed any-
is right. where necessary to use up the required electrical
The velocity factor of the line must be taken length.
into account. As with coax, this is subject to so Making Adjustments
much variation that it is well to make a resonance
check if there is any doubt. The method is the Wherever adjustable matching devices are used,
same as for coax, Fig. 8-24. A half wavelength of any really effective adjustment procedure must be
line is resonant both open and shorted, but the carried out either with the antenna in the position
shorted condition (both ends) is usually the more in which it will eventually be used, or under
readily checked. conditions simulating the eventual installation. The
The impedance -transforming quality of a quar- thought of making adjustments at the top of a
ter -wavelength of line can be employed in com- tower is often a bit staggering to the budding
bination matching and phasing lines, as -shown in B big -antenna enthusiast, but fortunately such a
and C of Fig. 8-29. In B, two bays spaced a high -wire act is not really necessary. There are right
half-wavelength are phased and matched by a and wrong ways to do the job at ground level,
400 -ohm line, acting as a double Q section, so that however. From preceding discussion of the effect
a 300 -ohm main transmission line is matched to of ground on antenna impedance it is easy to see
two 300-ohm bays. The two halves of this phasing that matching adjustments made with an array
line could each be 3 or 5 quarter-wavelengths long close to ground could be quite a bit off when the
equally well, if these lengths serve any useful array is hoisted to its eventual resting place 50 feet
purpose. An example would be the stacking of two or more in the air. Furthermore, even if there were
Yagis, where the desirable spacing is more than no impedance change from the effect of ground,
one-half wavelength. objects quite some distance out in front of the

300 OHMS

52 OHMS
300 OHMS
-ANY IMPEDANCE
LINE
72 OHM
.65%A COAX

IA
150
OHMS
5/8AI 52 OHM
IF COAX

.65A/4 72 OHM
(A) COAX

52 OHMS
x/300 OHMS (8) (C)

Fig. 8-29 - driven

Three methods of feeding stacked vhf arrays. A and B are for bays having balancedphasing
matching and
elements, where a balanced phasing line is desired. Array C has an all -coaxial
system. If the lower section is also 3/4 wavelength, no transposition of bays
is needed.
 174 ANTENNAS AND FEED SYSTEMS
array may reflect enough power back to it so that
than top performance at the low end. But if 2:1 is
an appreciable reflected -power reading is observed
as low as you can get, and it is around
from effects other than actual mismatch. The the
bigger and sharper the array, the more troublesome
frequency you work most often, you don't need to
worry too much if the transmitter loads satisfac-
these reflections become.
torily.
The solution to this problem is obvious, but not
With high power a high SWR runs you into the
too many antenna workers seem to think of it: aim
the beam straight up, with the reflectors close to danger of flash -over of the line, but this doesn't
happen very often in vhf circles, at least with any
ground. If the front -to -back ratio is 20 dB, the
coax worth using.
amount of power that will be radiated downward Exact matching is important in making meas-
with the beam in the straight up position is
negligible, and so is the effect of ground on the urements of antenna performance. If you would
learn anything from attempted gain measurements
antenna impedance. This lazy -man approach has you have to know exactly how much power you're
been used many times, on bands from 50 through
432 MHz, and on each occasion it has resulted in putting into the antenna, or at least you have to
very close to optimum matching when the array know that you're using the same power every time.
was finally installed in the tower position. Very Forward -power readings with the usual SWR bridge
much better results are possible in this way than are useless for antenna evaluation purposes, unless
with the array's line of fue parallel to and near the the system is perfectly matched. This means
ground. adjusting for zero reflected power, every time a
comparison or measurement is made.
How Important is Matching? Much of the conflicting evidence reported in
Due mainly to over-exposure to the term, a articles on antennas over the years has resulted
good many hams tend to worship perfect match- from a lack of understanding of the importance of
ing. To have a 1 -to -1 SWR is the ultimate achieve- this precaution. Just putting up a field -strength
ment, for them. But is it so very important? Not meter and then pruning the elements or adjusting
necessarily! It depends on what you're going to do. their spacing for maximum meter reading may
As may be determined from Fig. 8-30, a 100 result in your having a fairly good antenna, but it is
-foot a wholly unreliable way to make measurements. If
line of RG-8 coax at 144 MHz will have its loss
increased by less than 0.5 decibel with a 2:1 SWR you find the element lengths and spacings recom-
compared to a perfectly matched line. If the mended in much of the literature on antennas
loading on the transmitter is adjusted properly and confusing, failure to keep the radiated power
the line is trimmed for length, if necessary, a constant, or inability to determine it accurately,
listener at a distant point would not be able to tell may well be at the bottom of most of the
the difference. Note that this line trimming is to inconsistencies.
achieve a resonant condition and proper loading. It
does not affect the SWR! Using the SWR Bridge
Mismatch is important in some ways, and it can Coaxial feed is recommended, if only for the
tell you things about your antenna system. Make a reason that it permits easy monitoring of the
frequency run, measuring SWR at 144, 144.5, 145, matching process. You merely connect a standing-
145.5, 146, and so on. If your SWR dips to near

,'
wave bridge in the coaxial line and adjust
1:1 at 147 MHz and is 3:1 at 144, you need the
matching device for lowest possible reflected pow-

'i
some
work on your array. You're sure to be getting er. This should be zero, or very close to it. All that
less
is left then to make your antenna
radiate effec-

--
tively is to adjust the coupling at the
for maximum forward power on the bridge transmitter
° meter.
Note that you do not adjust the matching
i/r--
) {

for maximum forward power; you adjust fordevice

SWR= 20
.

't zero
reflected.
4
ID ---- Where the bridge is inserted in the line is
3
,/.4111110_~' SwR=
=7
important. Many hams are happy about their
antenna systems because a bridge connected in the
'Aroodir it= 1

to
a0
....~.~I~...~Im
as
01F... .
Fig. -
8-30 Increase in line loss because of
standing

I
OA SwR2 waves. To determine the total loss in
decibels in a
line having an SWR greater than 1, first
determine
/
aa
the loss for the particular type of line, length
frequency, on the assumption that the line and
a2 is
perfectly matched (Table 8-III). Locate this
SMRLS
on the horizontal axis and move up point
to the curve
corresponding to the actual SWR. The corres-
ai
az w a4asa u in ponding value on the vertical axis gives
2 3 4 s s e R the
additional loss in decibels caused by the standing
LINE 1.055 IN DB.WHEN MATCHED
waves.
Feeding Stacked and Phased Arrays 175
line at the transmitter output shows zero reflected be connected at or near the antenna, when making
power, but they may be in a fool's paradise. If the matching adjustments.
transmission line is long in terms of wavelength, There is no way to adjust an antenna properly
and lossy (all coaxial lines are lossy enough to without an SWR indicator. Don't try to do without
throw us off) the line may, in effect, bé self- one, for it is probably the most important instru-
terminating. That is to say you can have the ment you can own. It need not be fancy or
world's worst mismatch at the end of a 100 -foot "commercial." Very simple units are described in
run of RG-8 on 432 MHz and you'll never know it Chapter 14. The one built into the two -band Trans-
if the bridge is connected at the transmitter. Try a match should work well on all vhf bands. The line -
direct short on the end of your line, or disconnect sampler" rivals expensive in-line devices in versatility
the antenna entirely, and see how little difference and accuracy, and it can be made to work from at
it makes on your line. Remember these are the least 50 through 1300 MHz. Both types are inex-
ultimate extremes of mismatch! The bridge must pensive to build.

DANGER - HIGH RF FIELD STRENGTH!

For years we have been reminded to "Switch to exists possibility of disturbances to the human
Safety," and techniques for doing this are well nervous system, as yet unconfirmed in this coun-
established, at least as far as dangerous ac and dc try, which has caused some other authorities to set
voltages are concerned. But how about strong rf standards from 2.5 microwatts/cm2 to 1 mW/cm2
fields? At what level is there any real danger to as allowable limit's.
humans in exposure to uhf power? K2RIW feels we should be on the conservative
There is much difference of opinion, but rough side, and treat any radiation from high -power uhf
guidelines are available. Dick Knadle, K2RIW, transmitters with healthy respect. He recommends
discusses the problem in some detail in his QST staying at least 30 feet away from the front of a
treatment of the kilowatt amplifier for 432 MHz, 432 -MHz antenna carrying 700 watts of rf power,
described briefly in Chapter 14. He reports that the and he would not climb a tower carrying this
US Government has set a limit of 10 mW/cm2 antenna, while power was on. He probed the
averaged over a 6 -minute period as the safe kilowatt amplifier, running at full legal input, and
maximum for human exposure. More than this found no more than one milliwatt, and that only
causes thermal heating of skin tissue. Heating within 1/4 inch of the air -exhaust screening, on the
effect is especially dangerous to the eyes. There cover plate.
 Chapter 9

Building and Using V H F

Antennas

To some extent an antenna is an antenna, adapt inexpensive items to his purposes. Service-
regardless of frequency. Certain basic principles able beams have been made by coating wooden
apply all across the rf spectrum, but the wavelength dowels with conducting paint, or even by wrapping
factor makes for very large differences in practical them with aluminum foil. Neither of these tech-
problems encountered in building and erecting niques is recommended, but they are examples of
antennas, even within the vhf range. Mainly for this what can be done in a pinch. Salvage should not be
reason, the explanatory material of Chapter 8 may overlooked, if costs are really important. Lumber
not be enough for many vhf enthusiasts who would yards; electrical, welding, or plumbing -supply
like to try their hands at building their own beams. houses; metal -smelting companies; junkyards and
Arguments in favor of building rather than surplus lots - these are a few places in addition to
buying are not greatly different for antennas than the usual channels where we may find usable metal
for other equipment we need for communication, products. Hardware stores often have stocks of
except perhaps that fabrication of antennas may be aluminum and accessories useful in vhf antenna
more within the capabilities of the home craftsman construction.
than other equipment phases of the game. The Almost anything that is strong enough can be
hardest part of the job, the erection of the used for booms and supporting frames, whether it
antenna, has to be done by the amateur in any is insulating or conducting material, and there is no
case, so he is more likely to go the whole way and law requiring that elements be round in cross-
build the skyhook himself. Any able-bodied ham section, so long as they are of a metal that is a
with a few simple tools can build and erect his own reasonably good cordtuctor. And though electrical
antennas, and usually he will enjoy the work and rotation systems have become almost standard
learn much from it. Very likely he will stretch his equipment in amateur antenna practice, there is
dollars somewhat further too, for good antennas much to be said for simple "armstrong system"
come rather high these days. rotating devices.
Material for the construction of arrays may be Rotating provision of some kind is important,
costly, depending on where you do your shopping, however. In earlier days at least, much of the magic
but there are many ways for the ingenious ham to ascribed to a ham's first beam was actually the

TABLE 9-I

Dimensions for Vhf Arrays in Inches

Freq (MHz)* 50* 144* 220* 432* Dimensions are for most -
Driven Element 111 38-5/8 25-7/16 13 used sections of each band: 50 to
Change per MHz* 2 1/4 1/8 1/32 50.6 MHz, 144 to 145.5 MHz,
Reflector 116-1/2 40-1/2 26-3/4 13-1/2 220 to 222 MHz, and 432 to 434
1st Director 105-1/2 36-5/8 24-1/8 12-11/32 MHz. The element lengths should
2nd Director 103-1/2 36-3/8 24 12-9/32 be adjusted for each megahertz
3rd Director 101-1/2 36-1/8 23-7/8 12-7/32 difference in frequency by the
1.0 Wavelength 236 81-1/2 53-5/8 27-114 amount given in the third line of
0.625 Wavelength 149 51 33-1/2 17 the table. Example: if optimum
0.5 Wavelength 118 40-3/4 26-13/16 13-5/8 performance is wanted much
0.25 Wavelength 59 20-3/8 13-7/8 6-13/16 above 145 MHz, shorten all ele-
0.2 Wavelength 47-3/4 16-1/4 10-3/4 5-7/16 ments by about 1/4 inch. For
0.15 Wavelength 35-1/2 12-1/4 8 4 above 146 MHz, shorten by 1/2
inch. See text.
Element spacings are not critical, and table figures may be used, regardless of element lengths
chosen. Parasitic element lengths are optimum for collinear arrays and
small Yogis, having
0.2 -wavelength spacing.

176
Antennas for 50 MHz 177
result of its having been the first antenna he ever To aid those who like to work strictly on their
put up high and in the clear, and equipped with own, as far as materials and mechanical construc-
some form of rotator. Even a simple dipole with tion are concerned, Table 9-I gives the principal
these attributes is not a bad antenna, but the bigger dimensions needed in building antennas for 50
and better a vhf array is, the more it needs a through 450 MHz. Note that a most -used portion
rotator and some means of telling where the of each band is used for this information. Line 2 of
antenna is headed. If these requirements can be the table is a change factor to be applied to table
handled adequately by pulling on ropes and look- element lengths when other parts of a band are to
ing out the window, then there is no reason to be be emphasized. Only element lengths are ordinarily
ashamed of doing it that way. this critical. Element spacings and phasing -line
Because the band for which they are designed lengths can be left as given, or midband values
makes such a difference in the size and structural used.
details of vhf antennas, our practical constructional In the practical construction examples the
information is given by bands. It should be stressed dimensions of the original are given. Where the
that the following are examples; they by no means array is one that will be highly frequency-sensitive,
cover the range of possibilities. Nor should it be as in a long Yagi, the portion of the band where
inferred that, because a particular antenna is shown the antenna works best is stated. The change factor
for only one band, it cannot be used, in principle of Table 9-I can be applied if some other band
at least, for others. These are ideas, to be adapted segment is to be favored. When in doubt, check
as the reader may see fit. The true ham will "take back through Chapter 8, for basic information.
it from there."

ANTENNAS FOR 50 MHz

Nearly everyone, at one time or another, has and lower parts of dipole B are connected at the
need for something simple and/or inexpensive for outer ends by means of metal pillars or sleeves. If
an antenna. This means a half-wave dipole of some the tubing is sufficiently flexible it can be bent
sort, ordinarily. It can be any of the arrangements around as in dipole A, but this is usually done only
described in Chapter 8, so we will consider here in dipoles for the higher bands, where small
only those that are most commonly employed in conductors are stiff enough to be self-supporting.
50-MHz work. The feed line can be Twin -Lead or open -wire. If
the latter, the half-inch -spaced type is preferable to
The Folded Dipole the 1 -inch, as it is closer to 300 ohms impedance.
Coax and a balun can also be used. If this is done,
Probably the most universally useful 50 -MHz
72 -ohm coax will give a better match than 52 -ohm,
dipole, all things considered, is the folded variety. though even with the latter, the SWR will not be
It is broad in frequency response and not critical as more than about 1.5 to 1, which is not serious.
to construction or adjustment. It can be made of a
wide range of conductor sizes and materials, and it The "J" Antenna
is adaptable to various mounting arrangements. It
can be fed directly with 300-ohm balanced line Center -fed systems like the folded dipole are
(Twin-Lead or open -wire) or coax and a balun. well adapted to horizontal polarization, but the
A folded dipole can be suspended from rope or need for running the feed line perpendicular
to the
wires or supported on a mast, depending on how
the element is made. The center of the element can 110"
be grounded for lightning protection, or left Growuutrece ?
floating electrically. The dipole of Fig. 9-lA can be Tt"
made entirely of Twin -Lead, ready-made open -
wire, or of any wire you may have on hand. When
made of Twin -Lead, it is occasionally tacked or
taped to a wall, when a temporary and unobtrusive
axy
3oo-o/un l.ixe,

(A)
antenna is the principal requirement. .,--
Where it is to be mounted on a support such as no" +

Ground here
a rotating mast, the dipole of 9-1B is preferable.
The conductor size is not critical, except that both T
the broken and unbroken halves should be the Metal pillar Insulate wz to.'
or sleeve front support
same size. The unbroken portion can be attached
to or run through a metal pipe or tubing support, Soo-ohux lute.
in which case only a small cross -arm, or perhaps anY Leitlt (B)
none at all, will be needed. The wire dipole, A, can
be supported on a wooden "T," using vinyl- Fig. 9-1 -
Folded dipoles for 50 MHz. Either may
insula ted screw eyes or the type sold for TV be fed with 300 -ohm line, or 72 -ohm coax and a
installations. The inner ends of the broken portion balun. Dipole A is made of wire or Twin -Lead;
of either dipole, where the feed line connects, dipole B of any convenient size tubing.portion. Either can
should be insulated from the support. The upper be grounded at the center of unbroken
 178 BUILDING AND USING VHF ANTENNAS
conductor of the coax. The radials of the ground-
plane and the skirt or lower element of the coaxial
are connected to the outer conductor, and to the
support, if desired. The skirt of the coaxial antenna
should be so connected only at the top; the rest of
it must be insulated from the coax and the
support. Element and skirt lengths are not critical
in the coaxial antenna. About 54 inches should be
i"Appmx. suitable for work across the whole 50 -MHz band.
The ground -plane is perhaps the best all-
purpose vertical antenna for coaxial feed. As the
(A) 03) nllTnll name implies, the horizontal radials simulate
ground; consequently the impedance of the an-
tenna is little affected by variation in height above
Anti coax, actual ground and the nature of the supporting
wttfi balusc structure.
A simple and often-used version would be as
shown schematically in Fig. 9-4A, except fed
Fig. 9-2- Three versions of the "J" antenna, with entirely with 52 -ohm coax, without a matching
dimensions for 50 -MHz operation. Grounding for section. The feed impedance of a ground -plane is
lightning protection may be done as indicated by low; of the order of 30 ohms, so there will be some
the ground symbol. mismatch when it is fed with 52 -ohm line. The
SWR is under 2 to 1, however, and performance
should be satisfactory.
dipole for some distance makes them cumbersome Matching can be achieved in several ways. A
for most vertical applications. The "J" system of simple method is shown at A, but it requires that
Fig. 9-2 is more useful for vertical polarization. the main line be 72 -ohm coax. The quarter-wave Q
This is a vertical dipole with the matching arrange- section of 52 -ohm line makes an almost perfect
ment at the bottom end, for convenience. It may match, and it can be connected very simply. The
be fed in various ways. lower end of the Q section and the upper end of
Antenna A has a balanced-line feed. This can be the main line can be fitted with coaxial connectors,
any impedance, as the point of connection is and a coaxial junction used between them. The
moved along the stub portion until a match is length of 38 inches for the matching section is for
achieved. Antenna B is fed at the bottom end with solid-dielectric coax with a velocity factor of 0.65.
coaxial line. This is a good approximation, if the Foam and other low-density insulation will make
antenna is to be erected and used without atten- the matching section longer. Means for checking
tion to matching. Though it may not be a perfect resonant lengths are outlined in Chapter 8. See Fig.
match, it will be close enough for practical 8-24.
purposes. Antenna C is for use where adjustment Another method for matching with 52 -ohm line
for match is desired. As in A, the coaxial line and is to shorten the radiating element slightly, and
balun are moved along the stub until an SWR then tune out the reactance so introduced by
bridge in the line shows zero reflected power. The
bottom end of the system can be grounded for
lightning protection in either A or C. In B, the
bottom of the stub portion can be grounded.
The basic idea of the "J" is that the stub should
not radiate, but in actual practice it does, to some
extent. This radiation interferes with that from the
main portion of the antenna and may result in
raising the effective radiation angle. For this reason
other matching methods are preferable, though the
"J" has the virtue of simplicity. It is occasionally
used as the driven element about which parasitic
elements are rotated in a simple vertical Yagi.
Except in B; it is preferable to run the feed line as
nearly ,perpendicular to the stub as possible, for at
least a quarter wavelength.

Coaxial and Ground -Plane Antennas

Particularly where the supporting structure is
metal, the coaxial antenna, Fig. 9-3, and the
ground -plane, Fig. 9-4, are superior to the "J" for
omnidirectional vertical use. Both are intended
primarily for coaxial feed, and the line can be run
up inside the pipe mast, if one is used. The upper
Fig. 9-3 -
Coaxial vertical antenna for 50 MHz. A
supporting pipe mast can run up inside the skirt
element is, in effect, an extension of the inner portion.
 Omnidirectional Horizontals 179

Radiator 54" laity

4 R4,-tl.s
sa-1ocry

sº=olurc
coax 38 /049
72- Ohm coax,
any lenytk

(A) I (B)

Fig. 9-4 - The ground -plane antenna, shown with Q matching section, A. One method of making the
antenna, with a metal mounting bracket and ceramic insulator ís shown at B. Radials, omitted from 8
in the interest of clarity, are shown attached to the metal mounting bracket in C.

connecting a closed-end stub in parallel horizontal polarization is used. In 50 -MHz work

with the
antenna. If the antenna assembly includes a coaxial this is usually accomplished with a halo or turnstile
T fitting at the base of the driven element, the stub antenna. Basically, the halo is a half -wave dipole,
can be connected at will, and trimmed for length bent around into a circle or some other shape that
until a perfect match is achieved. With a 53 -inch will give it fairly uniform radiation in the horizon-
radiator the stub should be about 21 inches, if tal plane. Any of the common matching systems
made with 52-ohm coax, but it is well to start with described in Chapter 8 can be used with the halo.
one somewhat longer and trim for match. Remem- Where the halo is used for 50 -MHz mobile work
ber to short the fax end of the stub each time a the total length of the ring is usually reduced by
check is made. capacitive loading between the ends, as shown in
Another matching trick with the ground -plane Fig. 9-5. The circumference of the ring so formed
is to droop the radials downward, adjusting their is usually 60 to 70 inches.
angle below the horizontal until the antenna feed The gamma matching system is convenient for
impedance becomes 52 ohms. This usually occurs mobile halos, as it permits matching to coaxial line
at about a 45 -degree angle. The antenna ceases to and the use of an unbroken driven element. The
be a true ground -plane under these circumstances, step-up type of folded dipole is also used in halos,
but the method is often a satisfactory compromise. though the mechanical work involved has limited
There will be some radiation from the radials in the this type of feed mainly to manufactured antennas.
drooping position, but this is not necessarily bad. A three-ring model of this kind has long been a
Mixed polarization could be a "mixed blessing" fixture on the 6 -meter mobile scene.
under some propagation conditions. The capacitively loaded halo is a high-Q device,
Ground -planes can be made in many ways. One and it must be tuned with care or it will be all but
is shown in Fig. 9-4B. The vertical radiator is useless. Usually some provision is made for varying
1/4 -inch rod, threaded at the bottom end, held in the end-to -end capacitance, Cl. If a gamma -
the top of a ceramic standoff insulator with nuts matching system is used, the series capacitor, C2,
above and below the top of the cone. Before the and point of connection of the gamma arm to the
insulator is bolted to the angle bracket that serves element should be made variable, at least tempor-
as a mounting, a wire or flexible strip of copper is arily. Adjust these and the tuning capacitor, Cl,
fastened under the nut. It is left long enough so
that it can be soldered to the coaxial fitting
mounted on the angle bracket, before the insulator
is bolted in place.
Radials are omitted from B in the interest of
Fig. 9-5- The halo
antenna is a half -
clarity. They can be fastened to the angle bracket, wave radiator bent
as shown in 9-4C. The angle bracket can be into circular shape
fastened to any vertical support, of wood or metal. for nearly uniform
A metal support preferably should be grounded radiation pattern.
independently of the outer conductor of the coax, Capacity plates, Cl,
for lightning protection. permit use of a
small radiator for 50
MHz. Gamma match
Omnidirectional Horizontals and series capacitor,
Often it is desirable to maintain uniform field C2, are for coaxial
strength in all directions about the station when feed.
 180 BUILDING AND USING VHF ANTENNAS
for minimum reflected power in an SWR bridge The optimum spacing for stacked halos is 1/2 to
connected ín the transmission line. This is made 5/8 wavelength.
easier if the element is first resonated roughly at Two 50 -MHz halos adjusted individually for
the middle of the desired operating range with Cl, 50 -ohm feed can be stacked physically 5/8 wave-
checking resonance with a grid -dip meter. The best length apart by connecting them with half-wave
point for coupling the dip-meter coil is near the (77 -inch) sections of 52 -ohm coax, with a T fitting.
feed point, just to the left of the coaxial line A 38 -inch 52 -ohm Q section at this point will then
ground point in Fig. 9-5. Variation of the effective match a 72-ohm transmission line. Specific halo
capacitance of Cl is usually done by mounting a designs follow in the mobile section, later in this
small disk on an adjusting screw, equipped with chapter. For more stacking details see Fig. 9-44.
lock nuts, to one of the plates, and then adjusting The turnstile antenna, shown for 144 -MHz use
its position with respect to the other plate. in Fig. 9-39, can be adapted readily to 50 -MHz
Halos can be stacked one above the other in a service. It is larger physically than the halo, and it
vertical line, to lower radiation angle and build up should provide slightly more gain and considerably
gain. Thís is done occasionally for fixed stations broader frequency response. A stacked turnstile
where omnidirectional coverage and something system makes a very good omnidirectional antenna
better than a single halo are required. Matching and for home -station use. Stacking methods similar to
feeding can be done in the same ways as with any those outlined for halos and other coax-fed an-
two 50 -ohm antennas. See Chapter 8, and below. tennas can be employed with turnstiles.

YAGI ARRAYS FOR 50 MHz

in pairs without requiring a cumbersome support.

The elements are half -inch aluminum tubing of
1/16 -inch wall thickness, attached to the 1 -1/4 -
inch dural boom with aluminum castings. (Kirk
Electronics Yagi clamps.) The mounting method of
Fig. 9-14 is also usable. By limiting the element
spacing to 0.15 wavelength, the boom is only 6
feet long. Two booms for a stacked array can thus
be cut from a single 12 -foot length of tubing.
The folded -dipole driven element has No. 12
wire for the fed portion. The wire is mounted on
3/4 -inch cone standoff insulators and joined to the
outer ends of the main portion by means of metal
pillars and 6/32 screws and nuts. When the two
halves are pulled up tightly and wrapped around
the screws, solder should be sweated over the nuts
and screw ends to seal the whole against weather
Fig. 9-6 -A lightweight 3 -element 50 -MHz array. corrosion. The same treatment should be used at
Feeder is 52 -ohm coax, with a balun for connec- each standoff. Mount a soldering lug on the
tion to the folded dipole driven element. Balun ceramic cone and wrap the end of the lug around
loop may be coiled as shown, or taped to the the wire and solder the whole assembly together.
supporting pole. These joints and other portions of the array may

The Yagi antenna is almost ideally suited to

REE D.E.
50 -MHz operation. Usually only a relatively small
portion of the band is covered so the Yagis limited
frequency response presents no problems, and
arrays having gains of up to 10 dB are easily built
and erected. Except under the most severe weather
conditions, rotation of 50 -MHz Yagis of up to at ns=
M
least 6 elements can be handled with inexpensive
TV-type rotators, provided means are taken to
prevent the entire weight of the structure from
bearing on the rotator driving mechanism. Some
rotators have thrust bearings available as access-
ories, for this purpose.
3-Element Lightweight Array
The 3 -element 50 -MHz array of Fig. 9-6 weighs Fig. 9-7 -
Dimensions of the 3 -element array of
Fig. 9-6, for working in the lower portion of the
only 5 pounds. It uses the closest spacing that is 50 -MHz band. Driven elements are 1/2 -inch alum-
practical for vhf applications, in order to make an inum tubing. The folded dipole driven element uses
antenna that could be used individually or stacked No. 12 wire for the fed portion.
Yagí Arrays for 50 MHz 181
Fig. 9-8 - Closeup view of the boom
mounting for 50-MHz arrays. A plate of
aluminum about 6 inches square is
backed up by wood or Masonite.
TV -type U clamps hold the boom and
vertical support together at right angles.
The aluminum casting shown here for
mounting the beam elements is no longer
sold, but suitable substitutes are avail-
able.
-be . , -ralb
3,

be sprayed with clear lacquer as an additional variable capacitor can be used experimentally and
protection. replaced with the nearest-available fixed value,
The aluminum casting shown in Fig. 9-8 is no once the needed capacitance is found experimen-
longer sold but Kirk Yagi Clamps (Kirk Elec- tally. Coaxial variable capacitors, handmade or
tronics, 134 Westpark Road, Dayton, OH 45459) manufactured, are also usable. Some form of
serve equally well. The inner ends of the fed gamma match is especially desirable
if two bays are
portion of the dipole can be supported on ceramic to be stacked. The matching and phasing arrange-
cones, whose mounting screws go through the ment, shown later in connection with a 5 -over
-5
equidistant from the for 50 MHz, is applicable to smaller bays as well.
1/2 -inch element at points
midpoint, the actual separation being determined
by the element -mounting hardware used. Any High -Performance 4 -Element Array
inner -end separation from 1-1/4 to 3 inches should The 4 -element array of Fig. 9-9 was designed
be usable. The plane of the two portions of the for maximum forward gain, and for direct feed
with 300-ohm balanced transmission line. The
Director /oJ' parasitic elements may be any diameter from 1/2
to 1 inch, but the driven element should be made
s- as shown in the sketch. For a 1/2-inch driven
l:/rector ,o9' element use the information for the antenna of
t
Fig. 9-6. The spacing between driven element and
*c reflector, and between driven element and first
Radíato, r director, is 0.2 wavelength. Between the first and

`
is
u
e
300 Ohre (me
t
Reflector s,6 .t.,
second directors the spacing is 0.25 wavelength.
'Tubinq

Fig. 9-9 -Dimensions of a 4 -element 50-MHz .--40" I 45"

array having maximum forward gain in the lower
part of the band. The folded dipole details are for
300 -ohm balanced feed. _0 e0.
o
-J 11
Bakelcte. Woodo Boosss
2 needed
folded dipole is unimportant. It need not be
vertical, as in Fig. 9-6.
The element lengths in Fig. 9-7 are for best -1b Steel plate
operation between 50.0 and 51 MHz. The reflector To fit verttcal
sappoct
and director are approximately 4 percent longer
and shorter than the driven element. The close
T
spacing of the parasitic elements (0.15 wavelength) 11 Boom,
makes this deviation from the usual 5 percent
desirable. For higher parts of the band, shorten all
elements by 2 inches per megahertz higher in Steel pia le
frequency.
The folded dipole gives a single 3 -element array
Fig. 9-10 -
Suggested construction for an inexpen-
sive 4 -element array for 50 MHz using a wooden
a feed impedance of about 200 ohms. It may be boom. Dimensions are a slight modification of the
fed with a balun of the type shown in Fig. 8-25, optimum given in Fig. 9-9, to fit a 12-foot boom.
using 52 -ohm coax. A gamma -matched dipole may Mounting arrangement is for clamping to a pipe
also be used. See Fig. 8-17C and associated text. A mast.
 182 BUILDING AND USING VHF ANTENNAS
REF,
D.E.
is a method of attaching the boom to a pipe
DI D2 D, support, using wood blocks. A convenient way of
supporting metal-boom arrays on a pipe mast is
T shown in Fig. 9-8.

5 50 -MHz Elements on a 12 -Foot Boom

116 110 105' 103
As aluminum or dural tubing is often sold in
12 -foot lengths, this dimension may impose a
practical limitation on the construction of a
50-MHz beam. A 5 -element array that makes
optimum use of a 12 -foot boom may be built
according to Fig. 9-11. If the aluminum -clamp
method of mounting elements shown in Fig. 9-14
--r -35"-
1

+f«-35'
i

r- 35 - 35' is employed, the weight of a 5 -element beam can

be held to under 10 pounds.
Fig. 9-11 -
Five -element 50 -MHz Yagi for a
12 -foot boom. Dimensions are for working in the
The gamma match and coaxial line are recom-
mended for feeding such an array. A folded dipole
lower part of the 50 -MHz band. similar to that used in the 3 -element array will
provide an approximate match to 52 -ohm coax and
The same general arrangement may be used for a balun, but the gamma system is preferable, as it
a 3 -element array, except that the solid portion of permits adjustment for exact match at the favored
the dipole should be 3/4 -inch tubing instead of 1 frequency range.
inch. The boom length would then be about 8 feet. Gain and bandwidth of this compromise design
With the element lengths given, the array will are both slightly below optimum, but it represents
give nearly uniform response from 50 to 51.5 MHz, effective use of a 12 -foot boom. It will have a bit
and usable gain to about 52 MHz. more gain than a 12 -foot 4-element array, but will
If a shorter boom is desired, the reflector be somewhat more critical in frequency response.
spacing can be reduced to 0.15 wavelength and
both directors spaced 0.2 wavelength, with only a LONG YAGIS ON 50 MHz
slight reduction in forward gain and bandwidth. A
slight modification for mounting on a 12 -foot Once we go beyond the convenient and readily
wooden boom is shown in Fig. 9-10. Also included available 12 -foot boom appreciably, any structure
tends to become cumbersome, so building for
optimum performance is usually the main consider-
ation. A 5 -element 50-MHz of optimum design is
shown as part of the 5 -over -5 described later. Yagis
of 6 elements, or even longer, need not be heavy or
overly difficult to mount and rotate. Choice of
materials and judicious use of bracing are impor-
tant here.

6 -Element 50-MHz Yagi

The 6 -element arrays of Figs. 9-13 and 9-19 are
examples. The 20-foot boom can be made of light
aluminum TV masting. This comes in 10 -foot
lengths which telescope one inside the other for
about 6 inches. The joint can be held firm with
self-tapping screws. Any light boom more than
about 12 feet long should be braced to keep it in
alignment, preferably from above the boom, as
shown in Figs. 9-12 and 9-14. A method of splicing
tubing for long booms is shown in Fig. 9-18.
Dimensions in Fig. 9-13 are for operation in the
first megahertz of the band. Set up in this way, the
array can be adjusted for perfect match at 50.3 and
it will show an SWR under 1.7 to 1 from 50 to
50.6 MHz. Slightly more bandwidth can be ach-
ieved by making the directors one-half inch shorter
than shown. Additional directors should be spaced
70 inches apart. They may be similar to D4, or
each one inch shorter for greater bandwidth.
Elements are mounted in the same way as for
Fig. 9-12 -A
6-element Yagi for 50 MHz, with the 3 -element job, using aluminum castings. The
center support is the same as in Fig. 9-8. Elements
optimum spacing for forward gain given in Fig.
9-13. Boom is 20 feet long. Antenna at the top is can be run through the boom and clamped, as
16-element all -metal collinear array for 144 MHz. shown in Fig. 9-14, but the aluminum-casting
Long Yagis on 50 MHz 183
Fíg. 9-13 - Dimensions of the 70" 59- 42--.-36"-.i.-36-
6 -element 50 -MHz array. If the
boom length is limited to exactly
20 feet, reduce spacing of D3 and
Ad elements
D4 by 2 inches each. Dimensions
Dunn.
are for the low megahertz of the
band.

Boom './
Dave,

D4 D3 02 D1,.. DE REF.
IOIZ 102.1 104' 105y 110?. ,

method of mounting is stronger. It ís recom- two 6/32 machine screws for mounting the brack-
mended for light boom materials. Matching is by et.
means of a coaxial gamma arrangement shown in The sliding arm that is the movable element of
Fig. 9-15. Once adjustment is completed the open the coaxial capacitor is made of 1/4-inch tubing or
end of the gamma arm can be wrapped with plastic rod, about 14 inches long. It is maintained coaxial
tape, and it and the joint between the sliding and with the main arm by means of two polystyrene
stationary parts of the gamma capacitor sprayed bushings. One is force -fitted to the end of the rod
with clear lacquer. If anyone doubts the ability of that goes inside the main arm. The other is fitted
this arrangement to withstand weather, the original tightly into the far end of the main arm, but
was used for over seven years, and was working reamed out to permit the movable rod to slide
nicely when replaced by that in Fig. 9-19. freely in and out. These bushings can be made
The main gamma arm is cut from the same from 3/8 -inch polystyrene rod, or they can be
material as the elements. It is suspended parallel to fashioned easily from small polystyrene coil forms.
the driven element by means of two 1 -inch ceramic A clip of sheet aluminum makes contact be-
standoffs and 4 sheet aluminum clips, as shown in tween the driven element and the sliding rod. Be
the photograph. The 1/2 -inch tube is 15 inches sure that all surfaces at the points of contact are
long. Its inner end is connected to the inner completely clean, as solid low -resistance electrical
conductor of a coaxial fitting, which is mounted contact is of utmost importance here.
on a small bracket screwed to the boom casting. Proper adjustment of the gamma match re-
Holes are drilled and tapped in the casting to take quires an SWR bridge. If the work cannot be done
with the beam in the position in which it is to
operate, set it up temporarily with the boom
vertical and the reflector close to the ground.
Insert the bridge in the line near or at the antenna.
If more than a few watts of power is present in the
line, adjustment of the gamma will best be done
with gloves on the hands, to prevent rf bums. The
operation is twofold; both the point of connection
and the value of the series capacitor must be
adjusted. Start with the clip set about 16 inches
Fig. 9-14 -
Method of mounting elements through out from the boom, with enough tension on the
a metal boom, left, and of bracing the boom to the clip to insure a good electrical connection. Adjust
vertical support, right. Suspension bracing is
recommended for long booms. Shorter ones can be the capacitance and the point of connection for
braced below the boom. A one-piece wrap -around zero reflected power at the midpoint of the
clamp on the boom takes the angle brace. frequency range you want to work over effectively.

Fig. 9-15 - Gamma -matching sec-

tion for the 6 -element array, using
a coaxial variable capacitor. The
sheet aluminum clip at the right
and the length of the small rod
protruding from the arm are ad-
justed for minimum reflected
power. The movable rod element
of the capacitor is about 14
inches long, and is insulated from
the fixed portion, of 1/2-inch
tubing.
184 BUILDING AND USING VHF ANTENNAS
If you are interested only in the first 600 kHz or so open -wire line. Coaxial -line phasing and matching
of the band, use 50.3 for the adjustment frequen- is ideal for gamma -matched bays, and it has
cy. For good coverage of 50 to 51 MHz, use 50.5. obvious mechanical advantages. Coaxial phasing
In the latter case, the SWR should be below 2 to 1 tends to result in somewhat cleaner radiation
over the 1 -megahertz range. patterns, there being no pickup or radiation from
the phasing system. A properly designed balanced -
STACKED 50 -MHz YAGIS line system should also work well in this respect,
but chances for stray signal pickup and phasing -line
The lower radiation angle resulting from the radiation are greater.
stacking of two 50 -MHz Yagis one above the other Coaxial matching and phasing are shown in the
has a very beneficial effect on station effectiveness, 50-MHz 5 -over-5 that follows. A 4 -over-4 using
particularly on long paths. The gain may exceed balanced transmission line feed and phasing can be
the theoretical 3 dB obtainable from stacking, made very simply with two bays of the type shown
where radiation angle is a critical factor. (Low in Fig. 9-9. A 400 -ohm line of No. 12 wire spaced
radiation angle can be harmful, too, but the effect 1-1/8 inches connecting two such bays a half-
is discernible mainly on paths that are not too wavelength apart presents a 300 -ohm feed imped-
difficult to cover, ordinarily.) Gain from stacking is ance at its center. It can be fed with a main
achieved without appreciable sharpening of the transmission line of 300 ohms impedance, balan-
horizontal pattern, and with little effect on fre- ced, or with 72 -ohm coaxial line and a balun.
quency response, whereas gain from using more Half-wave spacing gives somewhat less than opti-
elements in a single plane tends to make both mum gain, but the lowered radiation angle still
factors more critical. makes such a 4 -over -4 very much worthwhile.
Choice between a single long Yagi and a stacked A bay spacing of 5/8 wavelength (12 feet) and
pair of the same theoretical gain thus becomes a all -coax matching and phasing is appreciably bet-
matter of operating objectives. Stacking a pair of ter. A full wavelength (20 feet) is optimum for
3 -element bays like that in Fig. 9-6 should give medium -length bays, but it may be more than most
about the same gain as a single optimum-spaced vhf-antenna builders want to tackle.
6 -element bay, Fig. 9-12. The 3 -over -3 will require
less rotating, but the 6 -elements -in -line will reject 5 -OVER -5 FOR 50 MHz
strong signals from off the desired line of fire more The information provided in Fig. 9-17 is useful
effectively. for a single 5 -element Yagi, or for the stacked pair
of Fig. 9-16, either to be fed with a 50 -ohm line.
Stacking Methods The phasing and matching arrangement may be
The functions of phasing and matching of two used for any pair of Yagis designed for 50 -ohm
Yagi bays can be handled with open -wire or coaxial feed individually. With slight modification it will
lines. Uses of both are discussed in Chapter 8. See serve with Yagis designed for 200 -ohm balanced
Fig. 8-29 and associated text. Antennas designed feed.
for balanced -line feed are more readily phased with
Mechanical Details
Construction of the single Yagi bay or a stacked
pair is simplified by use of components that should
be available to most builders. Element -to-boom
and boom -to -mast mounts are aluminum castings
designed for these applications by Kirk Electronics,
134 Westpark Road, Dayton, OH 45459. The
gamma matches shown schematically in Fig. 9-17
are of coaxial construction, waterproofed for long
life, available from the same supplier.
Booms are made of two 8 -foot lengths of
1 -1/4 -inch aluminum (Reynolds) found in many
hardware stores. Reynolds makes a special fitting
for joining sections of the tubing, but these are not
widely available from the usual hardware -store
stocks, so a handmade splice was substituted. A
piece of the same -diameter tubing as the booms,
12 inches or more in length, is slotted with a
hacksaw, and then compressed to fit inside the
ends of the two 8 -foot lengths, as seen in Fig. 9-18.
If the splice is held in the compressed position with
large pipe pliers or a hose clamp, the ends will slide
inside the boom sections readily. When the splice is
released from compression, the two tubes can be
driven together. Self-tapping screws should be run
through the tubes and the splice, to hold the
Fig. 9-16 - 5 -over -5 stacked-Yagi array for 50
MHz, with all -coax feed.
assembly firm. Use at least two on each side of the
splice.
5 -Over -5 for 50 MHz 185
Elements are 1/2-inch aluminum tubing, Alcoa DRIVEN ELEMENT
alloy 6061-T6. Almost any aluminum should be
suitable. Kirk Yagi clamps, one-piece aluminum so -Ohm Coaxial
castings designed for this job, are available for 3/8 - 0.6613/4.1 gamma match
1
as well as 1/2-inch elements, and 1 -1/4-inch boom. total length
The eyes through which the elements pass are zs-0/vx Pkasixy Line
drilled, but must be tapped for 10-32 setscrews to
tighten the elements firmly in place, two screws T Fitting
per element. The portion of the clamp that
surrounds the boom can be spread slightly to allow
the clamp to slide along the boom to the desired 12FT"
!` so -Ohm line,
tany length
point. The interior surface is slightly rough, so
1 \ Tluough
tightening the yoke with the screw provided with
the clamps makes the element set firmly on the connector
boom. The reflector, driven element, and first
Zs-Ohm
director are all in back of the boom splice.
The vertical member of the stacked array is
1 -1/4 -inch thick -wall anodyzed steel tubing, com-
monly used In large antenna Installations for home
TV. Do not use thin-wall aluminum or light
0.66
totall
Xe ,ñ.
th 8

DRIVEN ELEMENT
phasiny line
so -Ohm
9 h
galvanized steel masting. The aluminum is not
strong enough, and inexpensive steel masting rusts
yí'+j F16FT
inside, weakening the structure and inviting failure. 0 o
Spacing between the bays can be a half wave-
length (10 feet), 5/8 wavelength (12 feet), or a full 4erY4-4..- as,. 1- 46'h"
\ 02
wavelength (20 feet), though the wide spacing REF. DR. EL. DI M
1161/2"' 111' 105'f2 103'/í"
imposes mechanical problems that may not be MAST TO
worth the effort for most builders. The 5/8 -wave BOOM CLAMP

spacing is a good compromise between stacking Fig. 9-17 - Principal dimensions of the 50 -MHz
gain and severe support problems, and is recom- 5-over -5, with details of the 3/4-wavelength Q -sec-
mended with the materials used here. tion matching system. The propagation factor of
The 10 -foot lengths of steel masting could be 0.66 applies only with solid-dielectric coax. Gam-
used, with the bottom 8 feet running through the ma -matching assemblies are coaxial -capacitor units
(Kirk Electronics C6M).
tower bearing to the rotator. A heavier main
support is preferable, however, and it is "1 -inch
double Q -section of 75 -ohm coax, as shown in Fig.
water pipe" in this installation. This is iron, about
1 -3/8-inch outside diameter, extending about 8
9-17.
The Kirk gamma -match assembly has an SO -
feet out of the tower. The steel masting between
239 coaxial fitting built in, so the phasing lines are
the Yagi bays is fastened to the pipe with four
TV -type U-clamps, spaced evenly in the over-
fitted with PL-259 coaxial connectors at both
lapping area of the two supports. ends. The inner ends attach to a matching coaxial
T fitting. The main run of 50 -ohm line connects to
The booms are braced to the mast fore and aft,
the center of the T, with a coaxial through-
using the longest pieces of element stock left over
connector and a PL -259 fitting. When the antenna
when the forward directors are cut from 12 -foot
is installed all connectors should be wrapped
lengths. Ends of the braces are flattened about one
tightly with plastic tape, and sprayed with Krylon
inch, and bent to the proper angle. Outer ends or other protective spray. Dow -Coming Silastic
fasten to the booms with two self -tapping screws
RTV-732 sealant is also good for this use. If the
each. The mast ends are clamped to the support coaxial phasing sections are wrapped around the
with one TV U-clamp for each pair. This bracing is
booms and vertical support a few times, they will
good insurance against fluttering of the booms and just reach the T-fitting, when 12 -foot spacing is
elements, which can cause failures after long periods, used.
even though a structure appears adequately strong. The lines should be any odd multiple of a
quarter-wavelength. If both are the same length the
Phasing and Matching gamma arms should attach to the same side of the
A single 5 -element Yagi can be fed directly with driven elements. If there is a half-wavelength
50 -ohm coax, through the Kirk coaxial gamma - difference in the lines, the arms should connect to
match assembly (Type C6M). This has an adjust- opposite sides. The length given in Fig. 9-17 is
able coaxial capacitor, and an arm that connects to nominal for solid-dielectric coax. If foam -dielectric
the driven element with a sliding clip. Both the line is used, the propagation factor given by the
capacitor and the point of connection should be maker should be substituted for the 0.66 figure. It
adjusted for minimum reflected power, at the is best to grid -dip the line sections for resonant
center of the frequency range most used. Doing frequency, in any case. Cut the line three inches or
this between 50.2 and 50.4 MHz is suitable for more longer than the expected length. Solder a
most operators, other than those using fm above loop of wire between the center pin and the
52.5 MHz. Each bay of the stacked pair should be mounting flange of an SO -239 connector. Attach
set in this way. The pair can then be fed through a this to the PL -259 connector at one end of the
186 BUILDING AND USING VHF ANTENNAS
with 50 -ohm line, in this position, and the gamma
r 12-

8 -FT
TUBE
_
iI_ -
'1

--- Z= 1 8FT
TUBE (!
111
settings should be the same as obtained in the first
check, described above. Now connect the two
75 -ohm phasing lines, and insert the SWR bridge in
the 50 -ohm line to the T fitting. The SWR should
Fig. 9-18 -
Details of the boom splices used in the be the same as when the bays are fed separately
5 -element 50 -MHz Yagis. Two 8-foot lengths of 1
through the 50 -ohm line; close to 1:1. The array
1/4 -inch tubing are joined to make the 16-foot
booms. can be dismantled and reassembled atop the tower,
and matching should remain correct.
line, and couple it to the dip -meter coil. Trim the The matching-phasing system described is use-
line length until resonance at the midpoint of the ful for any two loads designed for 50-ohm feed.
intended frequency range is indicated. This will not The 5/8 -wave spacing is usable with up to at least
change appreciably when the other coaxial con- 6 -element bays, though wider bay spacing is
nector is attached. needed for maximum gain with long Yagis. Individ-
The line used in the model described is ual antennas intended for 200-ohm balanced feed
RG-59A/U, which is satisfactory for any amateur can be matched with 75 -ohm coax in the phasing
power level, so long as the SWR is kept low. Larger harness and baluns at each load.
coax, such as RG11A/U, is recommended for a Bay spacing is not critical. Close spacing gives
greater margin of safety. somewhat lower gain, but a very clean pattern. The
main lobe gets sharper and larger as spacing is
Adjustment and Testing increased, but minor lobes also increase. These take
over from the main lobe if spacing of bays is
An individual Yagi can be tested and matched carried too far. The effect of increasing bay spacing
properly by mounting it a half-wavelength above
is shown graphically in Fig. 8-11.
ground, in a large area that is clear of obstructions
for many wavelengths. The boom can also be tilted
up, until the ground-reflected wave is not a factor 144 -OVER -50
in the field -strength meter reading. The SWR
bridge should be connected at the gamma match, Stacking of arrays for 50 and 144 MHz, and a
or an electrical half-wavelength therefrom. Apply method of using a single low -loss transmission line,
low power (not over 10 watts) and adjust the are illustrated in the two-band system of Fig. 9-19.
gamma capacitor and the point of connection to The 4 -bay 144 -MHz array will be described in the
the driven element for zero reflected power, at the next section. The 6 -element 50 -MHz Yagi is very
desired frequency range. The model was flat from similar to the one already described, except for the
50.2 to 50.4 with just perceptible reflected power simplified gamma matching system and special
showing at 50.1 to 50.5. Adjusted in this way the attention to lightweight element design.
array should work well up to about 51 MHz. The 50 -MHz elements have center sections of
The best way to check operation of the stacked half-inch aluminum tubing making up about half
pair is to support the array with the reflectors their total length. Thin -walled fuel -line tubing
resting on the ground and the booms pointing inserts in each end keep the total element weight
straight up. A 6 -foot step -ladder can be used for a down, and provide a means of adjustment of
temporary support. The bays can be fed separately length, if the builder wishes to experiment with
tuning. The entire element can be made of the
heavier tubing, though the arrangement described
is lighter, and has a bit lower wind resistance. Some
shopping around in surplus houses, aluminum
smelting places, and hardware and plumbing supply
stores, as well as the usual aluminum tubing
sources, will turn up several tubing size combina-

Fig. 9-19 -All metal arrays for 50 and 144 MHz.

All parts of both beams can be assembled readily
with ordinary hand tools. In this installation the
two beams are fed from a single feed line, with a
Fig. 9-20- Model showing the method of mount-
ing 1/2 -inch elements on a 1 -1/4 -inch boom with-
waterproofed coaxial switch at the top of the out drilling holes through the latter. For stróngest
tower permitting selection of the desired array permanent assembly, self -tapping screws should
from the operating position. fasten the sheet -metal clamps in position.
144 -MHz Antennas
tions that can be used in this way. The ends of the i' -Z'/=_.1
larger tubing can be slitted with a hacksaw to a
depth of about three inches, and then tightened
onto the smaller material with a wrap -around
Y4
Io
I ,
--
0--Nc.22 acta o

clamp. The elements can be mounted through the

i ©
boom, as in Fig. 9-14, fastened with Kirk Yagi
Clamps, as in the 5 -over-5 just described, or held in
o
¡-Bbtáup o
1 Bewdup1
o o
place with sheet -aluminum clamps shown in Figs.
9-20 and 21.
The lips of the clamps should be bent upward
v4.
io
Bend
at right angles first. Forming the "U" ís started by
placing the tubing in a vise in a vertical position, Oup
B

and bending the clamp around it. The actual U No. 22 D.rLL .yb
shape is achieved by opening the vise to slightly
more than the width of tubing-plus-clamp, placing
the clamp U-down loosely in the vise with the
tubing lying in it, and then tapping the tubing
lightly with a hammer. Alignment of the holes in
Fig. 9-21 -
Dimensions of aluminum plates used to
make the assembly clamps in the vhf arrays. Sheet
the clamps is not fussy, and if they are drilled metal should be 3/64 inch or thicker. Two A -type
slightly larger than needed to pass the screws there clamps are needed for joining 3/4 -inch tubes at
will be no assembly problem. We used a No. 22 right angles for assembling the frame of the
144 -MHz array. One B and one C are needed to
drill and 6-32 screws. The nuts should be pulled up mount a 1/2 -inch element on a 1 -1/4 -inch boom, as
only tightly enough to hold the assembly firmly in Fig. 9-20. The "figure 8" clamp, D, made from a
together. Check the nuts after the array has been in 1/2 -inch -wide strip approximately 6 inches long, is
use for a few days and tighten as necessary. used to ground the coax to the 50 -MHz boom.
The gamma method is about as simple as you
can get: the coax is merely brought along the hermetically sealed coaxial switch was mounted on
boom to the driven element, bent at right angles, the vertical support between the two antennas for
and run out far enough to match the antenna this purpose. An ordinary coaxial relay will do, if
impedance when fed through a 100-pF fixed mounted in a waterproof box. The remote switch
capacitor. The point of connection was found operates on 115 volts ac, but to make for a safe
experimentally, though 20 inches and 100 pF installation a 6.3 -volt transformer was used in the
should do. Put an SWR bridge in the line and move station, connected to a similar transformer on the
the connection along the element, for minimum tower through rotator cable with pairs of wires in
reflected power. The outer conductor of the coax parallel. Low -voltage ac is used for the main run.
should be grounded to the boom at about 54 Antenna selection at the control position is done in
inches from the capacitor end. Strip a narrow band the primary circuit at the station end.
of the outer covering off, and fasten it to the boom The other 144-over -50 combination, Fig. 9-12,
with the "figure 8" clamp of Fig. 9-21D. Water- is set up for separate open -wire lines. The rotating
proof by wrapping with plastic tape and spraying section of line to the 144 -MHz 16 -element col-
with Krylon. Treat the capacitor similarly. It must linear is 300 -ohm transmitting Twin -Lead. The
stand high rf current. The Centralab 8505-100N is 50 -MHz rotating section is coax, terminated in a
adequate. A variable may be used if mounted in a balun and a Q section, to match the 400 -ohm line.
weatherproof box. Small mismatches in these lines are unimportant, as
The 144- and 50 -MHz arrays can be fed with both main runs are high -quality balanced trans-
separate 52 -ohm lines, but a single high -quality mission line. They show no increase in loss when
transmission line was used in this installation. A operated with a moderate SWR.

144 -MHz ANTENNAS

Though some information in the 50 -MHz sec- whether built. for 50 or 144 MHz. It is accepted as
tion may be useful to the builder of 144-MHz quite a good antenna on 6, but if everyone used
arrays, the roughly 3-to -1 difference in size tends nothing larger on 2 our results on that band would
to make construction ideas for the two bands be dismal indeed. Thus, we find long Yagis and
mutually exclusive. The 2 -meter, antenna nearly large collinear arrays in common use on 144 MHz
always has more elements than the one for 6, for and higher bands, but relatively rare on 50 MHz.
two good reasons: size permits it and performance Frequency response may be important on 144,
demands it. too. Most 50 -MHz activity is concentrated near the
Though gain over a dipole comes progressively low end of the, band, but we spread out more in
easier as we go to higher frequencies and smaller the 144 -MHz band. Its ability to work over a
elements, overall communications effíkiency does relatively wide frequency range makes the collinear
not. Regardless of frequency, the physical size of a good choice for many 2 -meter men, while those
the antenna is what really counts in determining who want optimum performance in one narrow
how well a station will work out. A 5 -element segment probably will go for the long Yagi.
Yagi, properly designed, will have the same gain
188 BUILDING AND USING VHF ANTENNAS
COLLINEAR ARRAYS

is higher, due to the lesser number of elements and

.,
connection of the main line at the inner ends of
z
/'
the middle pair instead of at the midpoint between
two pairs, so the reflectors are spaced 0.15
wavelength to bring the impedance down. There is
little gain difference with reflector spacings from
0.15 to 0.25 wavelength.

z coax-°d bal/ Anye A supporting frame may be made of wood or

metal, if elements are supported at their centers
and no insulating mounts will be needed. It is best
to keep the supporting structure in back of the
plane of the elements insofar as possible, and to
avoid use of insulating material near the element
ends. All -metal construction is illustrated in Figs.
Fig. 9-22 - Schematic drawing of a 12 -element 9-24 through 9-26. All -wood construction is used
in 432 -MHz arrays described later.
collinear array for 144 MHz that may be fed with
coaxial line and a balun, or 300 -ohm balanced line.
The supporting frame of Fig. 9-24 was intended for
use with this type of array.

Two collinear systems for 144 MHz and higher

bands are shown in Figs. 9-22 and 9-23. Either can
be fed directly with 300 -ohm balanced line, or
with coaxial line and a balun. The actual feed
impedance depends on many factors, but if no
means of matching is used the SWR on the main
line will never be very high. For precise matching,
use of the universal stub (Fig. 8-18D) is recom-
mended.
Feed impedance can be controlled to some
extent by varying the spacing between the driven
elements and the reflectors. In the 16 -element
array, Fig. 9-23, the impedance tends to be on the
low side of 300 ohms, so the reflectors are spaced
0.2 to 0.25 wavelength behind the driven elements.
In the 12 -element system, Fig. 9-22, the impedance

Fig. 9-24 - Supporting framework for a 12 -

element 144 -MHz array of all -metal design. Dimen-
sions are as follows: element supports (1) 3/4 X 16
inches; horizontal members (2) 3/4 X 46 inches;
vertical members (3) 3/4 X 86 inches; vertical
support (4) 1 -1/2 -inch diameter, length as required;
reflector -to -driven -element spacing 12 inches. Parts
not shown in sketch: driven elements 1/4 X 38
inches; reflectors 1/4 X 40 inches; phasing lines
No. 18 spaced 1 inch, 80 inches long, fanned out
to 3-1/2 inches at driven elements (transpose each
half -wave section). The elements and phasing lines
are arranged as shown in Fig. 9-22.

Elements should be rigid enough so that light

tension on the phasing lines will not bend them
appreciably. Aluminum or dural tubing 1/4 to 3/8
inch in diameter is commonly used, and 1/4 -inch
Fig. 9-23 -A 16 -element collinear array similar to
aluminum rod is good. Frequency response of the
the 12 -element of Fig. 9-22, except that the
feedpoint is midway between the middle pairs of collinear array is broad enough so that element
elements. Wider spacing of the reflectors results in lengths and other dimensions can be taken from
nearly the same feed impedance. An example of Figs. 9-22 or 23 or from Table 9-I.
this array is shown above a 50 -MHz Yagi in Fig. Large collinear arrays should be kept to a
9-12. maximum of 8 driven elements per set of phasing
Yagis for 144 MHz 189
Fig. 9-25 - Model showing method of assembling -
all -metal arrays, using clamps detailed in Fig. 9-26.

lines. See Fig. 8-13 and the 432 -MHz collinear of

Fig. 9-64 for recommended methods of feeding
arrays of more than 8 driven elements. Two
16 -element 2 -meter arrays can be mounted side by
side or one above the other, and fed in phase. A
side -by -side 32-element setup was used to collect
the data of Fig. 8-11, on the effects of spacing
"stween bays., Increasing the bay spacing makes a
large array that is hard to handle mechanically, but
a very worthwhile improvement in gain and pattern
sharpness results.
The feed impedance at the center of the
phasing line between two such bays is roughly half
that of one bay alone. A Q section with variable
spacing, or the universal stub, Fig. 8-18C and D,
will provide for matching.

All -Metal Construction

Collinear arrays of all -metal design can be very
light in weight, yet rugged enough to withstand
extreme weather hazards. The 16 -element array of
Fig. 9-12, built according to Figs. 9-23 through 26,
survived four severe winters, mounted 70 feet in
the air on a windswept New England hilltop, yet it
was in good condition when taken down. It
weighed less than 10 pounds and was relatively
easy to handle. The entire frame, except for the
1 -1/2 -inch vertical support, was made of 3/4 -inch
aluminum tubing. Elements are 1/4 -inch tubing,
center -mounted. Phasing is done with open -wire
TV line soldered to lugs bolted to the inner Bend on radus R

element ends. =d
The rotating portion of the transmission line B=é
was 300 -ohm tubular Twin -Lead, brought to an
insulating support just below the tower bearing.
Here it joined the main transmission line, which
was open -wire pulled up tight on strain insulators Fig. 9-26 - Clamps used for ambling all -metal
collinear arrays. A, B, and C are before bending
at each end. As pointed out in Chapter 8, straight Into U shape. Right-angle bends should be made
unsupported runs are recommended for low radia- first, along dotted lines, then the plates may be
tion loss with open line. This installation had 125 bent around pipe of the proper diameter. Sheet
feet, yet it performed in outstanding fashion. stock should be 1/16 -inch or thicker aluminum.

YAGIS FOR 144 MHz

Where a small antenna with appreciable gain is amateur vhf needs. It is not mandatory to scrimp
needed, a Yagi is the usual choice at 144 MHz and on vhf antenna costs, and there is a very useful
higher, as well as at 50. There is normally little source of top -grade materials and hardware in
point in building 2 -meter Yagis with less than five many stores in the form of Reynolds Aluminum
elements, as a 5 -element array of optimum design stocks. The methods already outlined for all -metal
fits handily on a 6 -foot boom. With some shopping construction are adaptable to Yagis as well as
and make -do methods, an effective 2 -meter Yagi collinears. In using wood for booms and frames the
can be assembled for almost nothing. range of usable methods is as wide as the builder's
Bamboo or wooden tool handles make excel- imagination.
lent booms. Dowelling of up to one-inch diameter Mounting Vertical Yagis
is available in most hardward stores and lumber-
yards. Welding rod and aluminum "picket wire," It is obvious to anyone who has tried to use
available almost anywhere, are good for beam available 2-meter Yagis that most are designed for
elements, which can be held in place with self- horizontal polarization. To use a single Yagi
tapping or wood screws. TV hardware is useful and effectively in a vertical position requires a non-
inexpensive. Even TV antennas themselves may be metallic support. Closet poles available in lumber-
of such design as to be readily converted to yards serve this purpose well for moderately sized
190 BUILDING AND USING VHF ANTENNAS
clamp like that of Fig. 9-14. The booms are 3/4- or
REF.
41"
D.E.
387411.
DI
381/8" 3r365"
D2

7%" Baum
D3
1 -inch diameter aluminum. Wood dowelling could
be used equally well.
A delta match is used in conjunction with a
coaxial-line balun to feed the single 5 -element
Yagi. Some experimentation with delta dimensions
17'.1.1.-13"4-16" -.-21" may be required to achieve the best match. (See
Fig. 8-17 and detailed description of the delta
match in Chapter 8.) Start with dimensions of Fig.
9-28A. This arrangement makes a fine small Yagi
Fig. 9-27 -
Optimum design for a 2 -meter Yagi,
using 5 elements on a 6 -foot boom. When used
that can be dismantled readily, for carrying about
in portable work.
singly, this antenna can be fed as shown in Fig.
8-17B, with 4 -inch delta arms connected 3 inches 5-Over -5
either side of center. The balun loop would be Use of two 5 -element Yagis with 1 -wavelength
about 27 inches long. With lengths shown, the spacing is shown in Fig. 9-28A. The phasing
antenna works well from 144 to above 146 MHz,
but gain drops sharply above 147 MHz. harness can be any open-wire line, preferably not
spaced more than one inch. Delta dimensions are
2 -meter antennas. Hardwood poles, not so widely not critical in this application, as the matching is
available, are recommended if the entire support, done with the universal stub at the center of the
through the tower bearing and down to the harness.
rotator, is to be of wood. A simpler solution is to 5 -Over -5s Stacked Side by Side
clamp the wood pole to a vertical support of strong
metal tubing, using wood only for the portion of The 4-bay 20 -element system in Figs. 9-19 and
the vertical member that is to be within a 9-28B uses two sets of 5 -over-5, connected be-
quarter-wavelength of the bottom end of the tween centers with another 1 -wavelength line. The
elements. universal stub is connected at the center of the
Side-by -side stacking of vertical Yagis gets away horizontal section. In each case, the stub length
from this problem and is highly recommended. If and line-connection point are adjusted for mini-
the array is to be relatively small in size, two mum reflected power in the main line.
5 -element bays side -by -side would be preferable to An interesting phasing method was used in the
a single Yagi of the same theoretical gain, from 4-bay array. Common electric zipcord, available in
both mechanical and electrical considerations. any hardware store, was split into its two parts.
The insulation was left on, and spreaders made of
5 -Element 144 -MHz Yagi ordinary 1/2-inch wood dowel were used to hold
An optimum design for 5 -element 2 -meter
the wires one inch apart. Holes were drilled in
these of such size that the zipcord could just be
Yagis, to be used singly or combined in stacked
systems, is shown in Fíg. 9-27. Dimensions given pulled through them. They are held in place with
work well from 144 to 146 MHz, if the matching is any good cement. If supported with TV -type
screweyes that grip the spreaders, such a low-cost
adjusted at 145. Lengths should be reduced 1/4
inch for each megahertz higher center frequency line is very durable. The array shown was taken
than 145 MHz. The original elements have center down after two years of use in a very exposed
sections of 1/4 -inch aluminum tubing with 5/32 - location, and no deterioration was apparent. There
inch rod inserts that slide into the center members. was no breakage, even under several heavy ice loads
One-piece elements of 1/8- to 1/4 -inch tubing or each winter. Using several supports on each harness
section is the key to this long life.
rod will work equally well. The larger size will
permit fastening in place with self-tapping screws The transmission line was switched between the
six- and two -meter arrays by means of a water-
bearing on the elements. For smaller sizes, use a
proofed antenna relay. To avoid the dangers of a
115 -volt line run, 6.3-volt transformers were used
6..F.
H at each end. This one -line hookup makes it possible
to use a good low -loss coaxial line to its fullest
.4ro1" potential on two bands.

(B) Wood -Boom 7 -Element Yagi

A Yagi with folded -dipole matching
7 -element
is shown in Fig. 9-29. A good material for booms
in medium -length Yagis is round wooden stock
90 -ohm line so -ohm line available in most lumberyards. Commonly called
a d d?/at 1240' ha/an rug or closet poles, they come in various lengths
Fig. 9-28 -Stacking details for the 5 -element
Yagis of Fig. 9-19 and 9-27. The short on the
and diameters. The boom shown is 10 feet long,
and 1-1/4 inches in diameter. It should hold
universal stub, and the point of connection of the alignment without bracing, if properly treated to
main transmission line, are adjusted for minimum prevent moisture absorption. Select stock that is
reflected power in the latter. Balanced line could thoroughly dry and free of knots. When holes have
be connected similarly for the main run. been drilled, spray or brush on clear lacquer, dry,
Yagis for 144 MHz 191
í

l
Clamp x 1i aluminum
(Yx /j for parasitic ele xement )
I" Tubing.
with
wood dowe
8 Element
Thread
é Rod.6-32
end (1" Wood boom.
I'StasrdofF !O' long
Bend oh ó'radius
i3
Element
Thread 6-32
Wraprousrd
P

Di Dz D3 D5

I
REF. D.E. 04
o o o o 0 0I
o
SPACING- 17" 21" 25' -11- 29

I 1
LENGTH -401" 38" 37" 364- 361" 361" 36..

Fig. 9-29 -Dimensions and structural details of a 10-foot wood -boom Yagi for the 144 -MHz band.
The end view of the folded pole driven element is shown at the upper left, and the method
of
a balun.
mounting it in the boom, at the right. Dimensions are for feeding with 52 -ohm coax ofandthe driven
Element lengths are for optimum performance between 145 and 146 MHz. Detail
element is shown for one side only, in the interest of clarity.

and then brush on outside paint. When all mechan- method of holding the elements in place is shown
ical and electrical work has been completed, the at the upper right of Fig. 9-29. This clamp
assembled antenna can be sprayed with clear arrangement works equally well with any round
lacquer to prevent corrosion of metal parts. boom, regardless of material used.
Parasitic elements can be hard -drawn aluminum
wire or welding rod, 1/8 inch or larger, or tubing
Driven -Element Construction
up to 3/8 -inch diameter. The smaller stock is Construction of the ratio -type dipole ís shown
preferable. The 10-foot boom is probably about in Fig. 9-29. The unbroken portion is 3/8-inch
the longest that will hold up well without bracing, tubing, the ends of which are plugged with wood
but if a longer beam is wanted it can be braced as dowels to permit tightening nuts against it to hold
shown in Fig. 9-14. Additional directors should be the fed portion of the dipole in position. The latter
spaced 29 inches apart, and each made progress- is 1/8 -inch wire or rod, with the ends threaded for
ively 1/4 inch shorter. Element spacings are not 6-32 nuts. An alternative method of making end
particularly critical. The reflector can be anywhere connections, if you do not have a threading die, is
from 12 to 20 inches in back of the driven to hammer the rod end flat and then drill it to pass
element, with only a slight effect on performance, a 6-32 screw. The outer end can be bolted to the
provided that the matching system is adjusted to 3/8 -inch portion, and the inner to soldering lugs
take care of varying feed impedance. Director attached to the ends of the coaxial line and balun.
spacing can be varied plus or minus an inch or so
with no noticeable change in characteristics. Ele-
ment lengths can vary plus or minus 1/8 inch
without any change that could be observed except
by the most careful check on frequency response.
With the ratio-type dipole shown in Fig. 9-29,
the antenna can be fed directly with 50 -ohm coax
and a half-wave balun. A small range of adjustment
can be had by bending the 1/8-inch portion of the
folded dipole nearer to or farther from the
3/8 -inch portion. The curve of Fig. 9-30 shows the
standing wave ratio of the array with the dimen-
sions given, using this method of feed. This is a
good setup for the fellow interested mainly in
operating above 145 MHz. If optimum perfor-
mance is wanted near the low end of the band the 145 146 147
144
elements can all be made 1/4 to 1/2 inch longer. REQUENCV
In mounting the elements the boom should be
drilled just large enough so that the elements fit
tightly into the holes. They can be held in place by
Fig. 9-30 -
Curve of standing -wave ratio taken
with the 10-foot Yagi of Fig. 9-29. For optimum
wood screws run into the boom and bearing firmly matching between 144 and 145 MHz, make all
against them. These screws can be bonded together elements 1/4 to 1/2 inch longer, leaving all other
with a wire running down the boom, and this can dimensions as shown. Note that mismatch rises
be grounded, for lightning- protection. A better more rapidly on the high -frequency side.
 192 BUILDING AND USING VHF ANTENNAS
phasing section, the SWR on this line will be lower
r_upprox than if a uniform-conductor dipole is used, but this
is not an important consideration, in view of the
short run of phasing line. Two of these 10 -foot
Yagis stacked should nearly equal the single 24 -
footer of Fig. 9-32 in gain. The stacked pair will
41"
or more have broader frequency response, and in some
Any /ure.orcoax circumstances may be easier to install.
and ba/ust

Long Yagis for 144 MHz

Though there is no theoretical limit to the
Fig. 9-31 - One-piece folded dipole and universal
amount of gain that can be achieved in making ever
longer arrays, the practical limit is reached in
matching stub for substitution in place of the
ratio -type dipole. Adjustment of the stub length 144 -MHz Yagis at somewhere around 24 to 30 feet
and point of connection of the feedline can be for most of us. If the limited frequency response of
made for optimum matching anywhere in the arrays this long is not a severe handicap, they are
lower half of the band, with the element lengths an attractive means of developing outstanding
given in Fig. 9-29. antenna performance.
Optimum element spacings and lengths for long
Yagis were worked out experimentally by W2NLY
The two portions of the dipole are held in
and W6QKI some years ago, and similar work has
alignment by means of 1 -inch ceramic or teflon
been done by many others since. These projects
standoffs, one on each side of the boom and about
have resulted in published figures that appear to be
one inch out from it. A 6-32 screw running
contradictory, but their seeming disparity merely
through the 3/8-inch upper portion and a wrap-
shows that there are many ways to arrange the
around clamp of thin metal, also with a 6-32 screw,
elements in a Yagi for roughly the same result in
hold this assembly together as shown in the sketch.
gain and bandwidth for a given length. This point
An alternative driven-element design that allows
explains the variations that will be found in
a wide range of adjustment, and use of any type of
comparing the Yagi systems shown in these pages.
transmission line, is shown in Fig. 9-31. This is One product of the W2NLY-W6QKI work was a
made from a single piece of stiff wire 160 inches
13 -element 144 -MHz Yagi that has since become
long, bent as shown to include both the dipole and an almost standard long-Yagi design. Lengths and
the universal matching stub. The sliding short on
spacings for one version are shown in Fig. 9-32.
the stub, and the point of connection of the line or This antenna is just under 24 feet long, but it can
balun, are adjusted for zero reflected power, using be made very light in weight and easy to handle.
a frequency in the middle of the range where
With the dimensions given in the drawing (all
optimum performance is desired. Readjusting for directors the same length) optimum performance is
various frequencies will extend the useful range of maintained essentially from 144 to 145 MHz. (The
this array beyond that in the SWR curve of Fig. SWR curve will show a rise at the low end, and a
9-30. The stub matching method can be used to steeper rise approaching 145 MHz, but gain and
extend the range of the ratio dipole, as well, if minor -lobe content do not change markedly.)
desired.
Polar plots of this antenna at various frequen-
Stacking cies are given in Fig. 9-33. At the left are runs at
144 and 145 MHz. Note that the broken line for
Where two of
these Yagis are to be stacked the 145 MHz shows a narrower main lobe, but some
type of dipole used is unimportant, for matching increase in size of the minor lobes. Plots at 145.5
will be taken care of at the central feed point, and 146 MHz show greatly reduced gain, and at the
preferably with the universal stub. The information higher frequency almost no difference between the
of Fig. 9-28 can be used. If a ratio -type dipole is major and minor lobes. Above 145 MHz appre-
used in each bay, with open -wire line for the ciably, the array has no practical value.

REF
02 D3 04 Do 07 D6 D9
s
Dio Du

e'-.r. 16" 32" 32" 32` 32" 32-

41" 393 131- Dp Narrow-bard version., all 36f'

Fig. 9-32 - Element lengths and spacings for a 24 -foot high-performance 144
bandwidth can be achieved, at some sacrifice in forward gain, by tapering the -MHz Yagi. Greater
described in the text. Design information is from W2NLY and W6QKI. element lengths as
Yagis for 144 MHz 193
Fig. 9-33 - Polar plots made with
the long Yagi of Fig. 9-32. At the
left are plots at 144 and 145
MHz. Note that at the higher
frequency the main lobe is sharp-
ening and minor lobes are larger.
At the right we see that the gain is
much lower and the minor lobes
are much larger in the 145.5 -MHz
curve, solid line. The array is of
little value at 146 MHz, broken
line, there being little difference
between major and minor lobes.

This situation is improved somewhat by taper- Hauling a long Yagi up to its position may be
ing the director lengths, as may be seen from Fig. the most difficult part of an antenna installation.
9-34, showing the relative gain level across the Many antenna workers solve this problem by
band with elements tapered 1/8 inch (solid line) designing the antenna so that it can be assembled
and 1/4 inch (broken line). The 1/8-inch taper atop the tower. The necessary components can be
extends the useful range to about 145.5 MHz, and run up on a pulley rope with ease. This is practical
the 1/4 -inch to above 146, but both involve if the boom is made in several sections that
appreciable sacrifices in peak gain. When tapered telescope together. The method has been used by
element lengths are employed, the first director, the author of these lines on many occasions to put
Dl, should be 37-3/4 inches, D2 37-5/8, D3 up large arrays single-handed.
37-1/2, and then each additional director 1/8 or Feeding the long Yagi presents no special
1/4 inch shorter, depending on the performance problems, except that one should be sure that it is
desired. matched at the frequency to be most often used.
These curves and pattern plots represent the Impedance of the array will depend on many
best that can be expected. In practice, the pattern factors, but if a basic impedance of about 25 ohms
of a long Yagi working above its upper useful is assumed, and the matching system worked out
frequency limit is little more than a mass of minor accordingly, the result will never be very far off. A
lobes and deep nulls. For work over all or most of step-up folded dipole with a half -inch solid por-
the band, a collinear array, or a smaller Yagi, is to tion, and the fed portion of 1/8 -inch wire spaced
be preferred. 1-1/4 inches between centers, will give an 8 -to -1
A convenient way to build a long Yagi is to impedance step-up, and something close to 200
obtain several telescoping aluminum mast sections, ohms feed impedance. Typical construction is
available from radio and TV distributors. These shown in Fig. 9-29. This can be matched with
should be pinned together with self-tapping screws. 52 -ohm coax and a balun, or a quarter-wave Q
Elements can run through the boom, if clamps section can be installed at the feedpoint for higher
such as shown in Fig. 9-29 are used to hold them in
place. Hard drawn aluminum wire or rod 1/8 inch
in diameter is a good element material. It is strong, 8.0
and springy enough so that ice does not form on it 7.0
readily. If an ice load does build up, the elements
droop and the ice slides off, after which they p 6.0
spring back to their original position. A long Yagi 1

5.0
of this construction, with suitable bracing for the %-I N
TAPER
boom, is practically indestructable. w 4.0
With very long booms it is best to hang the 113.0
antenna from its braces, though booms up to about
24 feet stand up well with bracing below the "I 2.0
boom. The boom bracing of Fig. 9-14 is suitable W
Q 1.0
345 -IN TAPER
for suspension. Where braces are below the boom,
greater strength can be achieved by leaving the
140 142 144 146 148
brace material round, and driving a tight-fitting
wood plug into the end several inches, so that it
FREQ -M C
will not compress when tightened in place. Light-
weight aluminum angle stock also makes good
Fig. 9-34 -
Relative response of the long Yagi with
director lengths tapered 1/8 inch, solid line, and
boom braces. 1/4 inch, broken line.
 194 BUILDING AND USING VHF ANTENNAS
impedances. If the spacing is made adjustable it the feed impedance of a stacked pair is about half
will be possible to adjust the impedance of the Q that of one bay, if the phasing line is any multiple
section for perfect match to any desired feedline. of one wavelength long, electrically. If made any
The Q section, if used, and the main feed line odd multiple of a half wavelength, the phasing
can be run along the boom toward the vertical system will act like a double Q section, when fed at
support. This is usually preferable to letting the its midpoint.
transmission line dangle, as the feedpoint is a long Because of the mechanical and electrical prob-
way from the center of the system, and it puts lems in handling stacked combinations of long
considerable unnecessary strain on the support. Yagis, more bays of shorter Yagis often may be a
better solution. As an example, four 24 -foot Yagis
Stacking in a box configuration take roughly 6000 cubic
Long Yagis have large "aperture"; consequently feet of space. It may give 20 dB gain, if the job is
they require very wide spacing when stacked, if a done properly. A set of 8 6 -element Yagis spaced
real improvement is to be made. There is little to one wavelength each way takes only about 1000
be gained from stacking two 24 -foot Yagis closes cubic feet of space, yet it should give almost as
than 12 feet apart, and 16 to 18 feet is better. In much gain, and probably over a wider frequency
matching stacked systems, remember the rule that range.

USING QUADS ON 144 MHz

Though it has not been used to any great extent at the bottom of the forward vertical support. The
in vhf work, the Quad antenna has interesting top portion of the element runs through the
possibilities. It can be built of very inexpensive support and is held firm by a screw running into
materials, yet its performance should be at least the wood and the bearing on the aluminum wire.
equal to other arrays of its size. Adjustment for Feed is by means of 52 -ohm coax, connected to
resonance and impedance matching can be accom- the driven element loop. For a perfectly non-
plished readily. Quads can be stacked horizontally radiating line the coax should be fitted with a
and vertically, to provide high gain, without detuning sleeve (see Fig. 8-25) but omission of this
sharply limiting the frequency response. precaution does not seriously affect the perfor-
mance of the Quad.
The 2 -Element Quad The reflector is a closed loop, its top and
The basic 2 -element Quad array for 144 MHz is bottom portions running through the rear vertical
shown in Fig. 9-35. The supporting frame is support. It is held in position with screws, top and
1 X 1 -inch wood, of any kind suitable for outdoor bottom. The loop can be closed by fitting a length
use. Elements are No. 8 aluminum wire. The -driven of tubing over the element ends, or by hammering
element is one wavelength (83 inches) long, and them flat and bolting them together, as shown in
the reflector 5 percent longer, or 87 inches. the sketch.
Dimensions are not particularly critical, as the The elements in this model are not adjustable,
Quad is relatively broad in frequency response. though this can easily be done by the use of stubs.
The driven element is open at the bottom, its It would then be desirable to make the loops
ends fastened to a plastic block, which is mounted slightly smaller, to compensate for the wire in the
7zr" wood frame
adjusting stubs. The driven element stub would be
trimmed for length and the point of connection for
Reflecfoc the coax would be adjustable for best match. The
reflector stub could be adjusted for maximum gain
F/ctt#a ekd or front -to -back ratio, whichever quality the build-
and bolt er wishes to optimize.
together In the model shown only the spacing is ad-
justed, and this is not particularly critical. If the
wooden supports are made as shown, the spacing
between the elements can be adjusted for best
match, as indicated in an SWR bridge connected in
the coaxial line. The spacing has little effect on the
gain, from 0.15 to 0.25 wavelength, so the varia-
Plastic Block tion in impedance with spacing can be utilized for
Coax
matching. This also permits use of either 52- or
72 -ohm coax for the transmission line.
Fig. 9-35 -
Mechanical details of a 2 -element Quad
for 144 MHz. Driven element, L1, is one wave- Stacking Quads
length long; reflector, L2, 5 percent longer. Sets of
elements of this type can be stacked horizontally Quads can be mounted side by side or one
and vertically for high gain with broad frequency above the other, or both, in the same general way
response. Bay spacing recommended is 1/2 wave- as described for other antennas. Sets of driven
length between adjacent element sides. Example elements can also be mounted in front of a screen
shown may be fed directly with 52 -ohm coax. reflector. The recommended spacing between ad-
The Skeleton -Slot Antenna 195
jacent element sides is a half wavelength. Phasing for conventional Yagis. In an experimental model
and feed methods can be similar to those employed built by W8HHS the reflector was spaced 0.25
with other antennas described in this chapter. wavelength and the director 0.15. A square array
using four 3 -element bays worked out extremely
well.
Adding Directors Workers using Quads on 144 MHz have re-
Parasitic elements ahead of the driven element ported reduced fading, compared with horizontal
work in a manner similar to those in a Yagi array. Yagis. Possibly this is due to the presence of some
Closed loops can be used for directors, by making vertical polarization with the Quad, making it less
them 5 percent shorter than the driven element, or affected by polarization changes that tend to occur
about 79 inches. Spacings can be similar to those on long or obstructed paths.

THE SKELETON -SLOT ANTENNA

A novel antenna that is very popular with replace the usual driven elements in a stacked-Yagi
British vhf enthusiasts is shown in Fig. 9-36. system, and parasitic elements are lined up with
Developed by B. Sykes, G2HCG, and sold by his them in the same way as in a conventional Yagi.
company, J-Beams, Ltd., on both sides of the Vertical spacing is 5/8 wavelength. Dimensions
Atlantic, this so-called "skeleton -slot" array gets its given are for 145 MHz, and broader frequency
name from the nature of its driven element, coverage is claimed than would be the case for a
derivation of which is as follows: Yagi of similar dimensions. Up to 7 parasitic
Start with two half-wave dipoles spaced 5/8 elements are commonly used in each half of the
wavelength, one above the other, as at A. Radia- array. Element spacing is similar to that employed
tion is mainly from the center portions of these, so in Yagi design.
the ends are bent toward each other, as in B. Then Two or more of these stacked slot-fed systems
they are joined with what is essentially a wide- can be placed one above the other or side by side,
spaced transmission line, C, and fed with a fanned- and fed in phase in the manner of stacked Yagis.
out Y section and coaxial or balanced line. Starting with about 72 ohms for the first set, a
Balanced -to -unbalanced conversion, for feeding stacked pair will have a feed impedance of about
with 75 -ohm coax, is accomplished with a coaxial 36 ohms, and so on. Spacing of the sets varies
sleeve as detailed in Chapter 8, Fig. 8-25. This is between 1 and 3 wavelengths, depending on the
not shown in the sketch. The name of the array length of each bay, following the rules for Yagis set
comes from the fact that this radiator behaves in forth in Chapter 8. Slot -fed arrays are common on
much the same manner as a slot in a plane of both 144 and 432 MHz throughout the United
metal, but in this case the plane is reduced to a Kingdom, and elsewhere in Europe. The foregoing
closed loop. information is published with the kind permission
Polarization is in the plane of the 15 -inch of The Radio Society of Great Britain.
portions, or horizontal in the example. These

Fig. 9-36 - Derivation of the skel-

eton -slot vhf array. Only the driven -r
element is shown, in the interest of
clarity. Parasitic elements are lined
up with top and bottom portions of
the driven element, giving the effect
45.
of a stacked Yagi with 5/8-wave-
length bay spacing.

75 -OHM
, COAX

OMNIDIRECTIONAL 2 -METER ANTENNAS

Net operation and the rapid growth of vhf directional systems of both polarizations. The
repeaters, especially in the 2 -meter band, have simple forms of vertical antennas described for
focussed interest on omnidirectional antenna sys- SO -MHz work in the first pages of this chapter can
tems for this band. Repeater work is almost wholly be duplicated for the 144-MHz band by using
done with vertical antennas, and the trend in dimensions from Table 9-I. Specific designs given
2 -meter mobile work has swung back toward here for 144 MHz will be confined to those for
vertical on that account. On the other hand, most which models have been built and tested on this
home-station operation other than above 146 MHz band. Mobile antennas are treated separately later.
is still horizontal, so there is need for omni.-
196 BUILDING AND USING VHF ANTENNAS
array that is almost impervious to the effects of
bad weather. Such an all-plumbing vertical two
A
2
half-waves in phase is shown in Fig. 9-38. Recom-
A
mended mounting points are shown. Preferably
4 those on the radiating elements would have cer-
amic stand-off insulators, mounting the antenna to
a wooden support.
14 ANY LINE Initially the matching section at the center was
OR MATCHING operated in the manner of the universal stub,
STUB
shown on several other arrays, as at the right. When
the optimum shorting and connection points were
ANY LINE determined experimentally, the assembly was com-
OR BALUN
pleted with plumbing fittings as shown. The
coaxial line and balun connection points shown
should provide a near-perfect match, if the antenna
construction follows the sketch. Practically no
reflected power could be measured from 145 to
147 MHz with a 52 -ohm line and balun connected
Fig. 9-37 -Basic details of nondirectional vertical at the points on the stub indicated. Match should
collinear antennas. If constructed of stiff rod or tu- be adequate across the entire 2 -meter band.
bing, they can be supported at points indicated by All parts were obtained at the plumbing coun-
dots. ter in a local Sears store. Soldering can be done
with a 100 -watt or larger iron, or with a small
Vertical Collinears torch.
Vertical elements can be stacked end-to -end Very similar construction can be used with the
and fed in phase, to build up gain without three -element stack shown in Fig. 9-37, except that
introducing horizontal directivity. Any number of the feed is at the center of the middle element of
elements could be used, but stacks of two or three three. This antenna can be fed at this point with
are common. Fig. 9-37 shows these types, with 300-ohm line with fair matching, or a universal
preferred mounting points indicated with black stub can be used. It should be as least a half-
dots. The midpoints of the stubs can be grounded wavelength long initially. When the proper point
for lightning protection, if desired. for the short is found experimentally, the unused
Either antenna can be made of wire, but a portion can be cut off. The center points of all
better arrangement involves use of copper pipe and stubs can be grounded.
plumbing fittings, which can result in a very rugged As with all verticals except the end -fed variety,
positioning of the transmission line presents a
problem, whether it is coax -or open line. Ideally, it
should run away from the antenna in a horizontal
19" plane, but this is rarely practical. If the supporting
structure has metal anywhere above the bottom of
the antenna it will probably affect the antenna
C'o.ieStandoff pattern more than the feedline will, so the position
of the latter is unimportant. If possible the
1911'
structure should be of nonconducting material,
where ít is near to any part of the antenna proper.
In this case, run the line down at least a quarter-
wavelength away from the radiator. For more on
vertical collinears, as used in repeater work, see
Chapter 12.
The two antennas of Fig. 9-37 are useful for
some horizontal -polarization situations, such as the
need for some gain and directivity in a limited
space. Made of wire, they can be strung up in an
L'okre Struadoff BALUN, RG-8/U attic, to give a measure of antenna performance in
OR BETTER
50 -OHM
places where a rotatable beam is impractical. For
COAX indoor use, a transmission line of 300 -ohm Twin-
Lead works very well.
OMNIDIRECTIONAL HORIZONTAL
ANTENNAS
Fig. 9-38
2 -meter
- Mechanical
collinear
and electrical details of the
antenna. Element and stub
Fairly uniform radiation in all directions with
lengths are the actual lengths of tubing required. If horizontal polarization is occasionally needed in
adjustment is to be attempted, the connection vhf stations. The control station of a local net,
points for the transmission line, and the position of most of whose members normally use horizontal
the short on the stub, should be adjusted for zero 2 -meter beams, is one example. An Official Bull-
reflected power in the transmission line. etin Station in the midst of a large population
Omnidirectional Horizontal Antennas 197
Fig. 9-39 - Schematic drawing of the turnstile
antenna. Crossed dipoles (dimensions from Table
9-1) are fed 90 degrees out of phase through a
quarter -wavelength loop. The Q section is not
ordinarily used in mobile installations where the
line is short.

concentration is another. The Headquarters Sta-

tion, W1AW, has demonstrated the worth of a
stacked omnidirectional array on 2 meters for the
latter application, for some years now.
Remember that gain is achieved only by taking
radiation from some parts of the antenna pattern
where it would be wasted for ordinary purposes,
and putting it somewhere else, where it will add to "Q°-SECTION
RG-58A/u
t4A RG-59/U
the signal in the receivers we want to reach. In an 4
A

omnidirectional antenna with horizontal polariza-

tion, this can be done only through compression of
the vertical pattern by stacking two or more
antennas and feeding them in phase.
RG-59/U
Very few simple horizontal antennas that are ANY LENGTH
called omnidirectional really are, but when they
are stacked in a vertical plane they can be oriented
dipole elements, each 19 inches long, threaded to
differently, so that their individual patterns com-fit in the holes in the round block. Lock washers,
plement one another, resulting in more nearly soldering lugs, and nuts hold the rods tightly in
uniform radiation in all directions. place and provide for connection of the line and
stub, in the manner shown schematically in Fig.
The Halo 9-39. Be sure that the dipole rods do not come in
Being merely a dipole bent into circular shape, contact with the center support.
the popular halo antenna for 2 meters is far from The Q -section arrangement shown is desirable if
omnidirectional. Most of the radiation from a exact matching is important. In a mobile instal-
half-wave element is from the middle of the lation, where only a short feed line is needed,
radiator, regardless of what is done with the ends. direct feed with any convenient small coax will
The principal virtue of the 2-meter halo is that it work about equally well.
makes possible a horizontal antenna of compact Though primarily for mobile use, this design
design and convenient shape. The pattern ap- may be adapted for neat home-station installation.
proaches omnidirectional character only when the The block can be fitted to the top of a pipe mast,
loop is made small, and the antenna is capacity and the coaxial line run down inside, if desired.
loaded, as is done in 6 -meter installations. This Turnstiles may be stacked for additional gain, in a
makes the antenna a high -Q device having a very manner similar to that shown for the Big Wheel.
low feed impedance and narrow frequency cover- Because of the light weight, up to four turnstiles
age. Halos of various types, for both 6 and 2 may be stacked vertically with only a relatively
meters, are described in the mobile section of this small support. In mobile work the turnstile is
chapter. somewhat superior to the halo, and is broader in
frequency response as well.
Turnstile for Two
An unobtrusive but effective omnidirectional
antenna for 2 -meter mobile or fixed -station use is Element Rod

the "turnstile," Figs. 9-39 and 40. This adaptation

by W1CUT is two half -wave dipoles crossed at right ®--6-32 nut
angles, fed with equal power but 90 degrees out of ..-solder lug
phase, as shown in Fig. 9-39. The quarter - C)...-lock washer
wavelength stub provides this phasing of the Plastic block
about 7" thick 4
second dipole. Note that this is not the usual
coaxial balun; it is a quarter wavelength long, not a
half wavelength. The pattern of the turnstile is 45.
I
Tap all holes
(except center.)
essentially circular in the horizontal plane. foró Sz thread
Mechanical details of the turnstile are shown in 1s6-32
et screw
Fíg. 9-40. The insulating support is a 1 -inch piece
of 1 -112 -inch diameter polystyrene rod, drilled in
the center to fit over a 1/4 -inch rod used for a
support. A setscrew keeps this mount tight on the ALL ELEMENTS
rod. Tapped holes 90 degrees apart take the four %ROD

Fig. 9-40 - Mechanical details of the lightweight %{ ROO OR TURING

turnstile.
198 BUILDING AND USING VHF ANTENNAS
Note from the schematic presentation, Fig.
942, that the elements are in parallel. Looking
down at the antenna, if the left side of one element
goes to the angle bracket, its right end goes to the
triangular plate. Moving around to the right, the
next element connects the same way, and so on, to
the third. Only the stub length is critical, and since
it is merely a strip of aluminum, several of various
lengths can be made and tried. Another method is
to slot the mounting hole in one end of the stub,
so that its electrical length can be adjusted.
Distorting the shape of the stub also will tune it to
some extent. The only objective in this is to get an
SWR bridge in the line to show zero reflected
power at 146 MHz. The SWR will then be just
detectable at opposite ends of the band. The size
and shape of the elements contribute to the
excellent broad -band characteristics.
A single Big Wheel is nearly as effective in all
directions as a horizontal half -wave dipole in its
favored directions, and it is some 2 to 3 dB better
than a single halo. A very marked improvement
comes with stacking a pair of Wheels, and this is
easily handled electrically and mechanically. A
Fig. 9-41 - The Big Wheel, an omnidirectional
stacked pair, and a pair of pairs, are shown In Fig.
horizontal antenna for the 144 -MHz band designed
by W1FVY and W1 1.1D. Radiating elements occupy 944. Physical spacing is 5/8 wavelength. This is
an area approximately 40 inches in diameter. not critical; it may be set to whatever the phasing
sections make convenient, in the vicinity of 50
inches.
The Big Wheel
The phasing lines are 1/4- and 3/4 -wavelength
A weakness of the halo is its small size, and sections of 75 -ohm coax. These act as a double Q
resultant low gain. It is considerably below a section, resulting in 50 ohms impedance at the T
halfwave dipole, compared with the latter's pattern fitting junction. An electrical wavelength of solid -
in its favored directions perpendicular to the dielectric coax works out to be about 5/8 wave-
dipole. The halo is also quite limited in frequency length physically, the optimum stacking dimen-
response, particularly when capacitively loaded. sion. The length of the phasing sections should be
The turnstile is somewhat better than the halo, and checked out with a grid-dip meter, as described in
is much broader in frequency response. Chapter 8.
One of the best horizontal omnidirectional The off -center connection for a pair of an-
arrays yet devised for amateur vhf work, in gain, tennas requires that one be inverted with respect to
frequency response, and uniformity of pattern is the other. Polarities for sets of 2 and 4 are shown
the "Big Wheel," developed by W1FVY and in Fig. 944. In the 4 -stack, the feed impedance is
W1IJD, and shown in Fig. 9-41. kept at 50 ohms through use of two 3/4 -wave-
Almost harder to picture and describe than to length Q sections of 75 -ohm coax, joined at the
build, the Big Wheel consists of three one -wave- midpoint of the array with a T fitting, and fed
length elements connected in parallel and arranged there with 52 -ohm coax for the main run.
in clover -leaf shape. The parallel connection results The slight irregularities in the horizontal pat-
in a very low impedance, which is raised to 50
tern of a single bay can be smoothed out in a
ohms with an inductive stub. Frequency response stacked pair by positioning the bays so that the
is very broad. With the stub adjusted to the proper
centers of the radiator elements of one line up with
length for perfect match at 146 MHz, the SWR is
the notches of the other in the vertical plane.
negligible from 144 to 148 MHz. The radiation
Coupling between bays of a stacked system re-
pattern is not perfectly circular, having slight dips
in line with the notches in the antenna.
Elements (A in Fig. 9-43) are 80 inches long, of
any convenient size tubing or rod stock. With
tubing, the strength and, stability of the antenna
are improved if wooden plugs are driven into the
element ends. One element end is fastened to a
grounded angle plate, B in Fig. 943. The other
connects to a floating triangular plate, C. The two
plates are kept in alignment by a ceramic or
bakelite insulator about 1-1/2 inches high. The
inner conductor of the coaxial line connects to the
triangular plate, and the sheath to the angle
Fig. 9-42 - Schematic representation of the Big
Wheel. Three one -wavelength elements are con-
bracket. The stub connects between these two nected in parallel. The resulting low feed imped-
plates. ance is raised to 52 ohms with an inductive stub.
Omnidirectional Horizontal Antennas 199
(INVERT)

son RD -8

son. RG-8
o NR.27 DRILL
O NR.17 DRILL
is-ss)
px 4i DRILL

5
(B)

101-174--10 Fig. 9-44 - Stacking arrangements for two and

four Big Wheels. Off -center feed in the two -bay
system requires that one bay be inverted with
respect to the other. In the 4 -bay stack the two
center bays are the same side up and the two outer
ones are inverted. Both systems are for 50 -ohm
I/! ALUMINUM feed. Dimensions given in Xc should include the
velocity factor of the coaxial line.

quires some modification of stub length for perfect

matching. The usual stub length for a single bay is
TV U CLAMP
S inches. For two bays stacked, the stubs will be 6
inches each. In a 4 -bay system, the top and bottom
stubs are 6 inches and the inner pair 7 inches. A
longer stub than 5 inches may be needed for a
single bay mounted near a car top or other large
metal body, if perfect matching is to be achieved.
(C) Gain from a stacked pair tends to average well
above the 3 dB that theory would indicate,
especially where the antennas are not at great
height above ground. This probably results from
the lowering of radiation angle that comes with
stacking. Greatest improvement is observed with a
pair, compared with one. With four, the overall
gain is more in line with what one would expect,
and the improvement over a stacked pair tends to
I/li
3/i
ALUMINIUM
be less than the theoretical 3 dB. The end result,
FLAT STRIP however, is omnidirectional gain roughly compar-
able to the gain of a small Yagi in its favored
direction.
NO.17 DRILL (o) Stacked omnidirectional systems are fine for
STUB control-station use in vhf nets, or for situations
where erection of a rotatable antenna is not
possible, but they are not ideal substitutes for
Fig. 9-43 -
Structural details of the Big Wheel. rotatable arrays. Interference problems may be-
One element is shown at A. For strength the ends come acute when gain is built up in an omni-
are plugged with wood. The grounded lower directional array. Likewise, the broad frequency
support is shown at B. It is fastened to the pipe response of 'the Big Wheel is not an unalloyed
support with a TV U clamp. One end of each
element is connected to this plate, and the other to blessing. The antenna may increase the trouble one
the triangular plate, C. The tuning stub is shown at has with spurious receiver responses, in a location
D. where other vhf services are operating close by.
200 BUILDING AND USING VHF ANTENNAS
VHF MOBILE ANTENNAS
The simpler antennas already described can be
adapted to mobile service, to the extent that other
users of a family car will permit its disfigurement.
A 2 -meter Big Wheel or a 3 -element array for
hidden -station hunting go pretty far in this direc-
tion, but they are used at times. On the other
hand, perhaps the most -used mobile vhf antenna
may be the broadcast receiver whip, or a reason-
able facsimile thereof.
Some vhf mobile enthusiasts, including the
author of this book, like to order new cars with no
radio or antenna, and then make their own
arrangements for both. Often vhf mobile operation
is a casual business, typically a Sunday -afternoon
drive with a rig that may also serve the home
station. Or operation may be and increasingly is on
fm, working through repeaters, which reduces the
requirement for mobile -antenna efficiency quite
markedly. In any event, with a little thought and
planning it is always possible to come up with a
pretty fair vhf mobile antenna, even if it also has to
serve the car broadcast receiver as well.

No-Holes Mounts
Increasing use of other than amateur com-
munications equipment in cars has led to the
development of many types of antenna mounts
that can be installed on cars without hole drilling Fig. 9-46 -Top and side views of a bumper mount
easily made from sheet aluminum. Clamps A, B,
or disfiguration of the vehicl e in other ways. It is
also quite easy to make your own.
and C, 1/8-inch stock, hold vertical member, D,
tightly to the bumper. Vertical support can be
Two detachable whip mounts are shown in Fig. tubing, for heavy antennas like a 50 -MHz halo, or
9-45. At the top is a sheet aluminum bracket with wood, as shown. The rod E is 1/2-inch aluminum,
drilled at the top end to take the turnstile support
of Fig. 9-40.

a coaxial fitting attached. This can be bent to fit

the car door, or frame. The temporary coaxial lead
can run over the top of the window, or in most
cars it can run through the door opening, as it will
not be subjected to long-term abrasion from
opening and closing the door. The weather strip-
111 ping around most car doors will pass RG-58 or 59
coaxial line easily. Similar arrangements can be
made for temporary mountings around the rear
deck openings of most cars.
The turret mount of Fig. 9-45 can be left in
place as long as desired, and the antenna and line
removed when not in use. Its construction from an
COPPER ordinary can top, soldered to a sheet of copper or
SHEET
brass, should be obvious from the drawing. Plastic
tape holds the mount to the car top, permitting
removal without damage to the finish. Electrical
SOLDERED ELECTRICAL
contact to the car roof is not needed for good
TAPE TO operation. The capacitance of the base plate to a
CAR TOP
metal car top is enough to simulate electrical
Fig. 9-45 - Two no -holes mounts for vhf mobile
whips. The clip, a temporary expedient, is merely a
grounding.
In many years of vhf mobile operation in a
coaxial fitting mounted on a light strip of alum- variety of cars, the author has always found it
inum, which can be bent over the top of the
window or around the edge of the door opening. possible to work out ways of mounting effective
The turret is cut from a can top, soldered to a thin antennas without drilling visible holes. Removable
copper plate. This is taped to the car top. Tape, or trim and the holes for mounting it have provided
a sheet of thin plastic, under the plate will permit means for fastening antenna mounts in place, and
easy removal of the assembly, leaving no damage to bringing in coaxial -line feed. Air vents offered
the surface. ready-made holes and access to the car interior in
Vhf Mobile Antennas 201
Fig. 9-47- Three -purpose whip serves for broad-
cast reception and for 50- and 144 -MHz mobile
work. Set to optimum 50 -MHz length, it works as a
3/4 -wave antenna on 144, though performance is
improved slightly if length is adjusted when chang-
ing bands. Mount is an aluminum bracket fastened
inside the well around the gas tank opening.

the case of the all-purpose mount that is just visible

in Figs. 9-51 and 53. This takes antennas mounted
in a PL -259 plug, or on a 1/4 -inch rod or tubing
support. A compartment around the gas tank fill
pipe was used for the whip bracket of Fig. 9-47.
Even the screws and holes already in the car were
used for mounting this one, and only a small hole
in the side of the compartment, inside the rear
deck, was needed for the coaxial line.
A detachable bumper mount for mobile an-
tennas such as the turnstile or halo is detailed in
Fig. 9-46. This is handy for supporting portable
ro
r
beams, as well. See Fig. 9-54. It requires only a few
pieces of sheet aluminum, a section of round
wo?den closet pole or broom handle, and an
aluminum rod or tubing for the vertical member.
---
Dimensions will vary with each Installation, so only
the basic ideas are given here. For a 6 -meter halo too much variation with changes of an inch or two
the support should be aluminum malting, and the either way, but the best length should be found
brackets should be made of 1/8 -inch stock. For experimentally, preferably with a field -strength
light 2 -meter antennas, brackets of 1/16- or 3/32 - meter, as it may not be the theoretical quarter -
inch aluminum should suffice. wavelength. The position of the antenna on the car
may affect both the optimum length and the feed
Mobile Whips impedance. Do not adjust length for best match, as
Where vertical polarization is in general use the the nominal impedance of a whip is below 50
whip is quite a satisfactory mobile vhf antenna. It ohms. Trim the coax line length, if necessary, for
is also used by casual mobiles in horizontal areas, optimum loading.
regardless of polarization disadvantage, because of A 50 -MHz whip of optimum length, usually 54
its convenience and unobtrusive appearance. Cross to 56 inches, works reasonably well in the 3/4 -
polarization may not pose too severe a problem in wave mode on 144 MHz. If you can extend the
mobile operation; the polarization of the received whip about 4 inches for 2 -meter work, resonance
signal ís likely to be mixed, as a result of multiple can be obtained for both bands. The radiation
reflections. The vertical whip serves well in work- angle is high for a 3/4 -wave whip, but it is a
ing ionospheric DX on 50 MHz, again because of convenient expedient for casual two -band work. If
the random nature of the polarization of a re- you work only the 2 -meter band, a 19 -inch car -top
flected signal. Additionally, a quarter-wave antenna whip will almost certainly be better.
working against ground (the car body) tends to be If the antenna mount is made with an SO-239
a good low-angle radiator. The chief weakness of coaxial jack and 50 -ohm -line feed, many different
the whip is that, like other verticals, it is more antennas can be pressed into service. The antenna
susceptible to ignition noise than horizontal sys- can be changed readily for different bands, and the
tems. socket can be used for connecting portable beams
Whips for 144 MHz or higher bands should having coaxial feed. Such a coaxial setup perman-
preferably be mounted near the center of the car ently mounted on the car is a very useful access-
roof, as this gives a large ground plane, nearly ory.
omnidirectional coverage, and a low radiation Most collapsible whips are small enough so that
angle. Roof mounting is good for 6 -meter whips, they can be mounted in a PL -259 coaxial plug,
too, but not everyone likes a 5 -foot vertical on a using some kind of insulating sleeve. The imped-
car top. Family circumstances may dictate use of ance is low at this point, so the quality of the
the broadcast whip, or reasonable facsimile. The insulation is not vital. Even a few layers of
whip of Fig. 9-47 serves three purposes. It is ordinary black plastic tape will do nicely. Insulating
connected to a coaxial switch on the dash, and rod of various kinds can be drilled or turned down,
leads from this run to the broadcast receiver and to as required. A good way to make a solid connec-
mobile rigs for 6 and 2. Other positions of the tion between whip and plug is to fit a 6-32 brass
switch can be used for testing other gear, without screw into the whip end, and tap the center
disrupting the "permanent" installations. connector of the PL -259 plug for 6-32 thread.
Length of a 50 -MHz whip is not particularly Soldering the screw into the whip base and into the
critical. A field-strength indicator may not show plug will result in a strong connection.
202 BUILDING AND USING VHF ANTENNAS
One versatile type of whip that should not be conductor of the coaxial line is tapped up one turn
overlooked is the many -sectioned variety made for from the bottom.
use with small portable radios. The whip shown at The coil is wound on 1/2 -inch poly or Teflon
the left in Fig. 9-48 has 10 sections, and it rod 2 inches long, turned down to 11/32 -inch
collapses to 8 inches. It cost two dollars at a local diameter for the first 3/8 inch of its length, and
radio bargain counter. Mounting it in a PL -259 3/8 -inch diameter for the second 3/8 inch. This
plug takes a little doing, but the result is worth leaves a 1 -1/4 -inch shoulder extending above the
some work. It will resonate on any band from 220 fitting, for the coil support. Drill the center of the
through 50 MHz, as it is adjustable from 8 to rod its entire length about 1/8-inch diameter. Drill
54-1/2 inches, and it will stay at any length you set two No. 35 holes perpendicular to this, one about
it. 1/4 inch from the top and the other one inch from
the top.
5/8 -Wave 2 -Meter Whip Drill out the top end of the insulator 1/4 -inch
Also in Fig. 9-48 is a close-up view of the lower diameter, to a depth of 1/2 inch. A 10 -inch piece
part of a 5/8-wavelength whip for 2 -meter mobile of 1/4 -inch aluminum tubing is inserted in this
work. Such an antenna has a low radiation angle hole, and drilled to match the No. 35 hole in the
and appreciable gain over any shorter vertical whip. insulating material. Tap the No. 35 hole for 6-32.
Manufactured versions of the 5/8 -wavelength whip A 1/2 -inch 6-32 screw runs through the side of the
are widely used in commercial and 2 -meter fm tubing to press against the inner wall on the other
mobile work, but there is no reason why you side, to assure good electrical contact for the top
shouldn't make your own and save some money. end of the coil. If a whip that extends to 45 inches
The basic ideas for this antenna were supplied or more is available, it can be fitted directly to the
by Vern Epp, VE7ABK, in October, 1965, QST. insulating coil mount.
The whip is a lightweight portable -radio type with A length of No. 22 wire is used for making the
a base section about 3/16 -inch diameter. It extends tap on the coil. Bend the end upward at 45
to a maximum of 37 inches, so a 10 -inch base degrees, about 1/4 inch, so that the point will
section is added to bring the total to over the come up into the lower No. 35 hole when the wire
optimum for this application, which is around 45 is threaded into the tip of the coaxial fitting and
inches. The actual length is not particularly critical, up into the drilled insulator. It can be bent around
if the capacitor across the matching circuit is the coil wire and soldered. The coil is No. 14
variable, as this adjustment can compensate for tinned, 4 turns wound on the insulating rod, with
length variations of several inches. The coil is turns spaced about 3/16 inch, center to center, and
connected to the sleeve of the coaxial plug at the another complete turn around the top of the
bottom, and to the whip at the top. The inner coaxial fitting, soldered in place. Wrap the top end
around the 6-32 screw inserted in the top horizon-
tal hole in the insulator, and solder.
The capacitor shown in the photograph was
initially a small 15-pF trimmer, soldered across the
coil. Obviously this is not a permanent arrange-
ment, it having been put on to determine the
optimum value experimentally. This was found to
be about 10 pF, for a total antenna length of 45
inches, for use in the upper half of the 2 -meter
band. Adjustment of the whip length will take care
of other frequencies in the band, with one value of
fixed capacitor. Adjustment can be made for
minimum reflected power in the coaxial line, or for
highest reading on a field-strength indicator. Whip
length or coil turn spacing can be adjusted, after
the fixed capacitor is installed, as shown in the
photograph.
The 5/8 -wave whip has only one weakness
when compared with other whips mentioned: it is
a one -band device, and it will not work for
broadcast reception. The grounded matching coil
effectively shorts the whip to the car frame at
broadcast frequencies. For 2-meter communication
it will be considerably better than either 1/4- or
3/4 -wave whips.
Fig. 9-48 -Detachable whips for vhf mobile use.
Both have PL -259 coaxial plugs fitted at the base.
Using the Broadcast Whip
At the left is a 10 -section portable -radio whip, Many whips installed on cars for broadcast
fastened to a Teflon rod fitted inside the plug reception can be extended long enough to work
sleeve. At the right is a 5/8 -wavelength 2 -meter well on 50 MHz. If you check on it, you may be
whip, with matching circuit mounted on a polysty- able to get a whip of the desired length installed
rene rod. when you buy a new car, and you might even get it
Horizontal Mobile Antennas 203
put on where you want it. There is likely to be a
transmission-line problem with broadcast whips, as
installed on cars. The line normally used is a
high -impedance type designed for low capacitance.
This will not work well in amateur installations set
up for 50 -ohm lines, unless it happens to be just
the right length. Conversely, 50-ohm coax may not
be too good for broadcast reception, especially if
the run is long. Some broadcast receivers have
tunable input circuits that may help to compensate
for loading effects of a low -impedance line, but it
is best to keep the run from the antenna to the
broadcast receiver short, if RG58 or similar
50 -ohm line is used.
The length of line to the vhf rig will not be
important, so the best way to use a bc whip is to
run a 50 -ohm line from the whip mount to a
coaxial switch nearby. The line to the bc receiver
can be made with the low -capacitance line that
came with the whip, and the 50 -ohm lines for the
ham gear can be any convenient length, as may be
needed to run them under carpets and out of the
way otherwise. With some thought a whip instal- Fig. 9-49 - This 144 -MHz halo appears to have no
lation can be made that will work well for both transmission line or matching device. Attached
purposes, and the car will show no evidence of its off -center to the whip of Fig. 9-47, it uses the whip
use as a hamshack on wheels. This is a real as a single -wire transmission line. Optimum whip

advantage at trade-in time. length is approximately 40 inches, so top section is

telescoped when the halo is used. Very little
HORIZONTAL MOBILE ANTENNAS vertical polarization is in evidence.

Where horizontal polarization is in use for 6 -meter whip is shown in Fig. 9-49. The brainchild
home stations, the mobile operator must use a of W3KDZ, it uses the ancient principle of single-
horizontal antenna if he is to achieve maximum wire feed. In this case the whip acts as the
local coverage on the vhf bands. Aside from the transmission line, and is connected to the halo
importance of matching polarization, the horizon- off -center, at the approximate matching point. The
tal antenna pays a considerable dividend in re- halo should preferably be about 40 inches up on
jection of ignition noise, especially on 50 MHz. In the whip, and if possible the whip should not
ssb communication it is particularly important to extend above it. Minimum vertical radiation is
keep down noise input to the receiver, as simple obtained in this way. The distance off center on
noise -limiting circuits are not effective with the ssb the halo should be adjustable, but 3 to 5 inches is a
mode. Several practical horizontal antennas are good starting point.
described below. The length of the halo element will depend to
some extent on the diameter of the circle. The
Novel 2 -Meter Halo smaller the circle the shorter the element, because
A quick-disconnect 2 -meter halo that can be of increasing capacitance end-to -end. The one
dropped onto the broadcast antenna or added to a shown is 40 inches long, and resonates at 145 MHz,

Fig. 9-50 - Halo for 50 or 144

MHz. As shown it is set up for the
50 -MHz band. Change to 144 is
made by moving connecting clip
closer to the center post, and
changing the spacing of the capa-
citor plates by swapping the cer-
amic insulators.
 204 BUILDING AND USING VHF ANTENNAS
. capacitor, and some adequate provision for keeping
. the halo rigid during driving. Any flopping of the
halo causes intermittent detuning, and severe mo-
bile flutter, in addition to that normally encoun-
tered.
Operation on two bands is achieved by chang-
ing the point of connection on the halo arm, the
setting of the variable series capacitor, and the
spacing of the square -plate capacitor at the element
ends. The plate spacing is changed by using a
3/4-inch ceramic standoff to hold the plates in
position for 50 MHz and changing to one 5/8 inch
long to bring resonance down to 48 MHz, so that it
will operate on its third harmonic.
Resonance must be achieved before the antenna
can be matched on either frequency. This can be
Fig. 9-51 -Shortened dipole for 50 -MHz mobile facilitated by use of a grid-dip meter, coupling to
service. Loading coils are inserted either side of a the halo near the vertical support. Resonance does
solid center section. Inserts are 19 inches long, not need to be exact for 144 -MHz work, as the
threaded into forms that support the loading coils.
See Fig. 9-52. antenna tunes more broadly on this band, but it is
critical in the 50 -MHz band. When it is resonated
near the desired frequency range, apply power and
when bent so that the opening is 9 inches. In the move the sliding clip and adjust the variable series
original by W3KDZ the element was 34 inches, and capacitor for lowest reflected power. These adjust-
the ends were fitted into a polystyrene insulator ments must be made with care for either band.
about 2 inches long. A grid-dip meter resonance For 50 -MHz work the clip connection is near
check is desirable, in any case. the outer end of the arm. For 144 it works at
Walking around the car with a field-strength about 3 to 4 inches out from the capacitor. With a
meter showed some interesting pattern and polari- 25-pF capacitor in series with the arm, tuning is
zation variations with the whip-halo combination. near the middle of the range for 50 MHz and close
With the 55 -inch whip alone the 144 -MHz polari- to minimum for 144. Properly adjusted the halo
zation was predominantly vertical, with a major works well over only about 50 to 50.5 MHz
lobe off the back of the car, at an angle of about without readjustment. In the 2 -meter band, satis-
15 degrees to the right of the line of travel. With factory operation is possible over about half the
the halo at 40 inches up the whip, left in its band without retuning. The range with the antenna
extended position, the polarization was mixed, set up for 50 -MHz service seems to be normal for a
with vertical strongest on the main lobe, but with halo on 6. On 144, the antenna seems to give
several horizontal and 45 -degree lobes elsewhere. somewhat better coverage than the conventional
With the whip run down to the halo (whip now 40 2 -meter halo, probably because of its larger size.
inches long, thus mismatched in the vertical mode)
horizontal was mainly evident, with some energy at A.Neat 50 -MHz Dipole
45 degrees, but almost no vertical. In other words,
the single -wire feed principle does work. The halo Where a horizontal antenna is needed for
is by no means omnidirectional, however; its main 50 -MHz mobile, and a halo is too much of an
lobe is forward and to the left, perpendicular to eyesore, the dipole of Fig. 9-51 works reasonably
that portion of the halo near its high -current well and is unobtrusive in appearance. If it can be
midpoint. A lesser lobe appears off the back and to mounted near the middle of the car the element
the right of the car heading. ends will not extend far enough to be dangerous to
passersby. Because radiation is largely from the
2 -Band Halo.
center of a dipole, this design works better than
one in which a single loading coil is used at the
A halo that can be set up to work on either 50 center of the element. Mounted just back of the
or 144 MHz is shown in Fig. 9-50. This antenna is rear window, as shown, the antenna has a major
customary 50-MHz size, 67 inches in circum- lobe to the rear and a lesser one forward. In
ference, with 2 -1/4 -inch square capacitor plates stationary operation the antenna can be rotated for
fitted to each end. The gamma matching arm is best signal by loosening the plug sleeve slightly.
14-1/2 inches long, of the same material as the The connection is maintained, so the received
halo, and separated from it by ceramic insulators. signal can be monitored as the antenna is turned.
A clip of sheet aluminum provides sliding contact As may be seen from Fig. 9-52, the horizontal
between the arm and the halo. As always with the dipole has loading coils at equal distances either
gamma match, be sure that this makes a clean tight side of center. The dimensions given are one of
contact. many possible arrangements. They were dictated
The halo shown was put together mainly to try by a desire to use 2 -meter turnstile elements for
out the two -band idea, so its mechanical details are the outer portions, and keep overall length to
not spelled out here. Important points in a about 50 inches.
permanent installation are the arm-to -halo contact The 13 -inch rod which is the center portion of
already mentioned, waterproofing of the series the dipole is drilled and tapped at each end for
 Horizontal Mobile Antennas 205
z

Ir i
Fig. 9-52 - Principal details of the loaded 50-MHz
RG-174/u Coax
mobile dipole.
Cl - -
Li, L2
15-pF dipped mica.
11 turns No. 20, 5/8 -inch dia, 16 tpi
4 Aluminum
30" Tubisry
(B&W No. 3007). L2 tapped 3/4 turn from
inner end, or as required for minimum SWR.
Coils are supported on 1/2 X 1 -inch ceramic
pillars (Millen 31001).

6-32 thread. The loading coils, Ll and L2, are

PL -259 with.
made in a manner similar to those used with the tic -776/u Adapter
two-band antenna. Prepared coil stock is slipped
over 1/2 X 1 -inch ceramic standoff insulators, and
the wire ends are soldered to lugs at each end of The antenna used by the writer was trimmed
t the insulators. The element ends are 1/8 -inch for resonance at about 50.25 MHz. The next step
aluminum welding rod, threaded 6-32 for about was to find a value of series capacitor, Cl, and a
one-half inch at their inner ends. A 6-32 nut is point of connection on the antenna or loading coil,
threaded onto the element, and this acts as a stop L2, that would provide a 50 -ohm termination for
when the element is screwed into the insulator. the coaxial line. This was done experimentally with
The 13 -inch center section is supported in a the antenna support clamped in a vise on the
1/2 -inch piece of solid aluminum rod about one workbench. A recheck of the SWR and operating
inch long, with a setscrew running in from the top frequency, when the antenna was installed on the
to hold the rod tightly in place. The lower portion car, showed little change.
of the block is drilled to take the vertical support, Frequency response is about the same as with a
which is 1/4 -inch aluminum tubing. This can be capacitively loaded halo. Resonated and matched
any length that will stand the strain; ours is 30 at 50.25, the dipole is usable from the low end to
inches long. 50.5 MHz before the SWR rises above 2 to 1, a
The diameter of the bottom end of the vertical mismatch that is tolerable in a mobile setup.
member is filed down just enough so that it can be
forced into the UG-176/U adapter, which, in turn, 2 -Meter Dipole with Built -In Balun
screws into the PL -259 coaxial plug. Small - A dipole makes a good horizontal 2 -meter
diameter coax was first used for the feedline, mobile antenna. Nearly all mobile antennas have
bringing it out through a hole in the vertical some directional characteristics, depending on their
support. This turned out to be fragile, so a piece of position on the car, so the natural bidirectional
zip cord (one conductor and its covering discarded) pattern of a simple dipole may not be too much
was substituted and found to work just as well. different from what usually results from the use of
a supposedly omnidirectional mobile antenna.
Adjustment The dipole of Fig. 9-53 is clean in appearance,
The top end of the line extending through the and it has an invisible built-in balun. Installed on a
hole as shown forms the arm of a gamma match. car top it will have very nearly a true bidirectional
The series capacitor, Cl, was first set up as a
variable, permitting the right combination of capa-
citance and tap position on L2 to be selected
experimentally.
First the antenna by itself must be resonated at
the center of the desired frequency range. This will
be a narrow frequency spread, a limitation not too
important, with most 6 -meter operation being in
the first 500 kHz of the band ordinarily. The
resonant frequency can be checked with a grid-dip 7r'
meter, putting the GDO coil adjacent to the
13 -inch center section of the antenna, close to the
center block. The trick now is to trim the lengths
of the outer elements, or the number of turns in
the loading coils, until you hit the desired fre- Í
r
quency. It will be a sharp indication; when you -
approach the desired frequency, do not trim
elements by more than one-half inch at a time, or
the loading coils by more than 1/4 turn. Whichever
you cut, be sure that the same change is made on
both halves of the antenna. When you're through, Fig. 9-53 - Horizontal dipole for 2-meter mobile.
The vertical support is 1/4 -inch tubing, with small
the coils should be identical, and the outer ends of
the element the same length. coax inside, thus serving as a 1:1 balun.
206 BUILDING AND USING VHF ANTENNAS
dipole pattern, without the distorted lobe that The screw is not needed if the support is
usually results from direct feed of a split dipole approximately a quarter wavelength high, as the
with coax. It is very easy to make, and is handy for ground will be made automatically by the con-
portable as well as mobile use. nection of the coax to the PL-259 plug. In fact,
The construction is similar to that for a several different lengths have been tried, and the
turnstile. A center insulator is mounted at the top sleeve support seems to work like a balun regard-
of a support made of 1/4 -inch aluminum tubing, of less of length. In the dipole pictured the grounding
any convenient length. The bottom of the support screw is about 6 inches above the fitting (support
is forced into a UG-176/U adaptor, which screws just over 2 feet high) and the radiation from the
into a PL-259 coaxial plug. The feedline is dipole is a good figure 8, whether the screw is
RG-58/U or RG-174/U coax, running inside the contacting the outer conductor or not. It might be
tubing. The coax should not make contact with the useful to be able to ground the outer conductor at
tubing at the top. The outer conductor is con- this proper point in longer supports, however, and
nected to one half of the dipole and the inner it is a simple thing to do. Be careful that the screw
conductor to the other. The support is made to act is not run in far enough to puncture the inner
like a 1 -to -1 balun by running a self-tapping screw insulation, and short out the coax.
in at a point about 18 inches down from the top,
to contact the outer conductor at that point.

PORTABLE ANTENNAS
From earliest times one of the great joys of vhf disappointing unless a good antenna is available.
hamming has been "working portable." Every Halos, whips, and the like are pretty poor stuff,
enthusiast dreams of someday having a station on a compared with a beam antenna of even moderate
mountaintop, with an unobstructed view for miles size. Fortunately, fixing up a vhf array so that it
in every direction. Few of us ever see this dream can be moved about readily is no great task.
become a permanent reality, but with the mobility Antennas can be built to encompass many
we enjoy today nearly everyone can bring it off for degrees of portability. Probably the simplest are
a few hours now and then. antennas and supports that come apart enough to
The catch is that even the finest vhf locations permit tying to a ski rack or other car-top carrier.
have a way of turning out to be somewhat Most manufactured beams are shipped knocked-
down, so dismantling for some portability is no
problem. Yagis made for the TV trade nearly
always have folding elements, to make life easier
for the serviceman -installer. These can be modified
for amateur bands quite readily. Occasionally the
element spacings can be left as they were in the
original, and the element lengths adjusted accord-
ing to Table 9-I.
The next step is to make your own boom and
supporting structure, using TV masting. This can
be cut to the maximum length that your car
storage space permits. Usually something around 4
feet maximum length is convenient. Element
mounting methods described elsewhere in this
chapter can be adapted to portable beams handily.
The principal problem then becomes how to feed
the array, since two popular matching systems, the
folded dipole and the gamma match, do not lend
themselves readily to quick dismantling and re-
assembly.
One thoroughly practical feed 'method is the
delta or Y match, Fig. 8-17A and B, and 9-56. If
the connection to the driven element is made with
removable clips the delta provides a connecting and
matching system that can be coiled up and carried
in your pocket. The bottom of the delta can be
terminated in a coaxial balun and coax of the
desired length, or Twin -Lead can be used for all or
any desired part of the main run to the equipment.
If the portable beam is to be used alongside the
car the support can be tied or clamped to the door
handle or bumper. An example of an effective
Fig. 9-54 - Portable beam set up for 144 MHz,
using the bumper bracket of Fig. 9-46.
bumper clamp is shown in Fig. 9-47, and in actual
use in Fig. 9-54. A screwdriver or small stake can
Portable Antennas 207
REF, DR.EL. DIR.
be driven into the ground to anchor the bottom of 124-

-12 -r --1e --t Ir --Ir

120-
-1 114-

the mast.
'
21"

io 0 o ..q."": o 0 01
Lightweights for 6 and 2 REF. DR EL 1 D2 CO

2 e r+ly. 343C
If you're satisfied to operate only in high spots 4oie MSERT

that are accessible by car, extreme light weight is woo SIDE VIEW
D
of no great interest, so long as the antenna can be
taken apart and packed away conveniently in the 2é
car. But you really need the performance of a
beam when you work a flea -powered transistor
ap
IUUye 2y1'
portable from some remote mountain top, miles S. N1.1

from the nearest road. If you go for this kind of

vhf hamming, light weight and easy carrying are of
primary importance. The arrays of Fig. 9-55 were
la VIEW

designed to be carried on foot. They sacrifice

nothing in performance for their lightness, so they Fig. 9-55 -
Details of the portable beams for 6 and
also serve nicely for general-purpose portable 2. One boom can serve for either array, or two sets
operating. of hardware can be made, permitting simultaneous
Provision is made for a 3 -element 50 -MHz beam operation on two bands. The hole just back of the
U -clamp holes is used for both 6- and 2 -meter
and a 5 -element 2 -meter one. The booms are elements.
duplicates, so that one can be used for either band,
or the two antennas can be set up together. All the
material for the 2 -band setup, with booms, 15 -foot 3/4 -inch hardware store aluminum tubing. A
vertical support and feedlines, adds up to less than TV -type U clamp holds the boom to the vertical
5 pounds. It stows in a canvas golf bag, as seen in support. Drilled as shown in Fig. 9-55 the boom
Fig. 9-57. In more -or -less this form they have supports the center sections of the elements for
served the author in vhf portable ventures from either beam. The 50 -MHz elements are 1/4 -inch
Maine to California and the Canadian Rockies to aluminum tubing, with their ends drilled one inch
Florida. They evolved gradually, being made ever deep with, a No. 5 drill, to take collapsible whips
lighter and easier to set up, until it is felt that they that make up the balance of each element. Various
are now near the ultimate in portability and sizes of these whips can be found in distributor
performance. catalogs. Ours are Lafayette Type 99-C-3005,
0.210 inch in diameter at the base, and 47 inches
The 3 -Element 50-MHz Yagi long, extended.
The booms are 3/4 -inch aluminum tubing, Because these whips are very thin at the outer
originally two 6 -foot pieces, now cut into 4 3-foot ends, the elements must be made several inches
sections. An insert at the center made of 5/8 -inch longer than would be the case for normal element
copper pipe joins two 3 -foot pieces to make a diameters. The driven element is 120 inches, the
6 -foot boom. This size is a nice sliding fit inside reflector 124 inches, and the director 114 inches.

22"-.1

is-\

iV RG-SB
9"/o/U
ny
92"
30o -ohm/
Twin Lead

(A)
so -ohm coax
ar y length C3

Fig. 9-56 - Two methods ofatfeeding the 50 -MHz portable array. A half -wave balun and delta match
A. The Twin-Lead delta and line, with adjustable antenna coupler, B,
a

of flexible wire are shown

permits use of the array over a wider frequency range. With readjustment,
it provides a constant load
for the transmitter, from 50 to 52 MHz.
C1 - 75-pF miniature variable (Hammarlund L1 -
15 turns No. 20, 1/2 -inch dia, 16 tpi. Tap at
MAPC-75B). 3-1/2 and 11-1/2 turns.
C2 - 11-pF miniature butterfly variable (Johnson L2 -3
turns insulated hookup wire, around center
160-211). of L1. Coupler is assembled in a
--
C3 30-pF miniature mica trimmer (ARCO)
J1, J2 Insulated tip jack.
1-5/8 X 2 X 3 -1/4 -inch Minibox, with the tip
jacks at one end and the coaxial connector at
J3 - BNC fitting. the other.
 208 BUILDING AND USING VHF ANTENNAS
range of the array can be extended considerably. A
small coupler unit can be connected at the bottom
of the delta, and coax run from there to the
equipment, or part of the transmission line can be
Twin -Lead or other balanced line. This can be any
length, though the electrical half wavelength shown
is a good system.

QI
e a With the short lines normally used in a portable
setup, this array works extremely well, and its gain
and directivity are very helpful in working out with
low -powered gear. When the coupler is used, and
:, WI+HD retuned for each frequency change, the system can
..
be matched perfectly over the first two megahertz
of the band. Gain is 6 dB or better from 50 to 51.5
MHz, and in excess of 7 dB in the most -used lower
part of the band.

The 5 -Element 2 -Meter Beam

- Beams for 6 and 2 with 15-foot vertical
Fig. 9-57
support and all hardware,
The boom for the 144 -MHz antenna can be the
ready for travel. one used for 6 -meter operation, or it can be
another made similarly. The center sections of the
2 -meter antenna are 1/4 -inch aluminum rods, with
With 47 -inch whips inserted to a depth of one inch, their ends drilled and tapped 6-32, to a depth of
center sections should be as follows: reflector 32 1/2 inch. Element extensions are 1/8 -inch alum-
inches, driven element 28 inches, and director 22 inum wire or welding rod, threaded 6-32 at the
inches. The elements can be held in place with clips inner end. To prevent confusion in assembling the
in the manner shown for several other antennas in antenna, the extensions are all the same, and the
this chapter, or self-tapping screws can be run center sections are made progressively shorter,
through the top of the boom, to bear on the back to front, to give the optimum element
element midpoint. The ends of the center sections lengths. The dimensions given in Fig. 9-55 are for
are slotted with a hacksaw, and a small wrap- best performance from the low end to about 145.5
around clamp is used to squeeze the ends tightly MHz.
around the whip. A delta and balun are used with this array also.
The array can be fed conveniently in two ways, The arms of the delta are 4 inches long, including
as shown in Fig. 9-56. The simpler is the delta and the alligator clips used for connecting to the driven
balun, A. This works fine at the design frequency, element. Pieces of zip cord make flexible long-
but the SWR rises quickly on either side. If an lasting delta arms. The balun loop, of RG-58/U
antenna coupler is used, at B, the useful frequency coax, is 27 inches long.

ANTENNAS FOR 220 AND 420 MHz

Since physical size must be maintained as we go fed properly. Anyone interested in working the full
to progressively higher frequencies, if communi- band, 420 to 450 MHz, would be foolhardy to use
cations effectiveness is to be maintained, the them, but where operation is confined to a narrow
number of elements in large arrays for 220 or 420 segment like 432 to 436 MHz, for example,
MHz tends to stagger the imagination of users of frequency response should be no problem. This is
lower bands. On the other hand, a half wavelength likely to hold for 220-MHz work as well, as activity
becomes of such proportions (roughly 2 feet on is frequently channelled to one part of that band,
220, 1 foot on 420) that high-performance beams by mutual local agreement.
are not difficult to build and adjust. Line losses are a problem at these frequencies.
Antenna work at these frequencies is an absorb- As pointed out in Chapter 8, the best line may be
ing field in itself. There is no better way to work none too good. Matching is important, particularly
out pet antenna ideas than with models built for if any real antenna evaluation work is to be
420 MHz. If reasonable care is used in scaling undertaken. Line radiation is troublesome, and
significant dimensions, models can supply answers must be countered with every reasonable pre-
in a few hours of interesting work that would take caution.
days (and dollars) to work out on some lower Basic antenna systems already described work
frequency - if, indeed, they could be developed well on both 220 and 420. Collinear arrays
may be
satisfactorily there at all, by amateur methods. built using the midband dimensions of Table 9-I
We formerly heard the statement, "Yagis don't and the mechanical arrangements suggested for
work on 420!" This completely false notion is 144 -MHz arrays, except that most parts can be of
mainly the result of the workers having neglected smaller cross-section and lighter in weight. In
the question of scaling. Yagis and most other kinds all -metal construction of collinears or Yagis, ele-
of antennas work just as well on 220 and 420 MHz ment and boom diameters should be kept in scale
as on any lower frequency, if they are made and with the 144 -MHz designs, insofar
as possible. A
Yagis for 220 and 432 MHz 209
12 -inch element running through a 1 -1/4 -inch element spacings and lengths that are close to
metal boom, for example, ís out of place in a optimum for gain. The antenna should be suppor-
420-MHz array. Hard-drawn aluminum wire or ted near its mechanical balance point, roughly 2
welding rod, preferably not larger than 1/8 inch in feet from the reflector end. If a TV -type U clamp
diameter, is good for elements. Booms no larger is used, it is well to bend up a U-shaped metal plate
than 1/2 inch are recommended, if made of metal. the width of the boom and about 3 inches long,
Matching methods follow the basic principles and slip it over the boom at the point where the
detailed in Chapter 8, but the small physical holes are to be drilled for the clamp. This protects
proportions of the adjustable Q section, the delta the boom from crushing when the U -clamp nuts
match, and the universal stub make these probably are tightened, and leaves it strong enough to stand
the most desirable matching devices. The ratio -type up well without bracing. Gusset plates of wood or
folded dipole is suspect at these frequencies. Masonite make stronger assemblies. See Fig. 9-63.
Plane and comer reflectors assume practical Parasitic elements in the 432 -MHz model are
proportions. Where broad frequency response and made of 3/32 -inch aluminum welding rod, and
high front -to-back ratio are desired, driven ele- 1/8 -inch is used for the 220 -MHz model. This
ments backed up by a wire grid or metal screening material can be purchased very reasonably at
make a fine antenna. This is particularly inviting welding supply houses, usually ín 3 -foot lengths.
where two bands must be taken care of with a Any stiff wire or rod up to 1/8 inch diameter will
single structure, since elements can be mounted on do. Drill the boom with a size that will just take
opposite sides of such a reflecting plane. The gain the elements with a force fit, then run a 1/2 -inch
of a screen reflector array is little affected by the brass or aluminum screw into the boom to bear on
spacing of the elements from the reflector, but the element and hold it in place. The screws can
there is a marked change of impedance when the then be bonded together and connected to ground
spacing is changed. (See Fig. 8-9, 180 -degree for lightning protection, If desired.
curve.) From this it is clear that matching to a The driven elements originally tried were step-
up folded dipoles similar to those used In the
transmission line can often be achieved by selecting
144 -MHz Yagis, but it was found that these did not
a suitable spacing by experiment, if the impedance
of the system is not known. work well at .220 and 432 MHz. This is probably
the result of the spacing between the two parts of
such a dipole being a considerable portion of a
YAGIS FOR 220 AND 432 MHz wavelength at these frequencies. The 432 -MHz
Moderate -size Yagis for the 220 and 420 -MHz Yagi was made with a driven element of the same
bands can be built at very low cost, and with only material as the parasitic elements, mounted as
simple tools, if the suggestions of Figs. 9-58 and 59 shown in Fig. 9-58 A and B. Blocks of insulating
are followed. Booms are 1 X 1 -inch wood, available material 1/4 inch thick and 1-1/2 inches square are
in any lumberyard. (Your dealer will call it "one fastened to the boom with two 1 -1/2 -inch brass
by one" but the actual size will be more like screws and nuts. The upper portion of the dipole
3/4 X 7/8 inch.) Be sure that it is straight, dry, and runs through the boom, just above the center, and
free of knots. Take the man's advice as to which the lower is held in place with 4-40 nuts on either
kind of wood will be best for out -door use, as side of the insulating plates, as shown in the end
available stocks vary around the country. Ours was view, B. The 3/32 -inch rod is easily threaded for
red cedar. Prime and paint it well, if you want long 4-40, if this is done before the element is bent. The
life. total length of the wire is about 25 inches. An
An 11 -element array is shown for 432 MHz, alternative to threading is to hammer the ends flat,
and a 7 -element one for 220 MHz, both using and drill for 4-40 screws.

/zI wood
r
m Poly Blocks
Boom \ Foldeádipole
we
4'3z

for 4-4o nuts

l
Details of a
Bend acozord
Fig. 9-58 1 "cyluedec
6-foot 11 -element Yagi for
432 MHz. The square (B)
(A)
boom and one polystyrene
mounting block for the
driven element are shown
at A. The blocks, element,
and boom are shown in
C\^
i drell
`.
`Cl
B,
; i
Cut off
ttltsF/endS
k,=-
Í
detail in the end view, B.
Matching stub, C, fastens PolyBlocEs
to ends of the driven ele- 15"
ment, and is mounted un-
der the boom between two
D4 Dr De Dg
poly blocks. Element REF.
1-1:1
D.E. 0,
0
D2 D3
o
D5
o o o 0
lengths and spacings for o

the middle of the 420 -MHz SP4cING ,bYi.f« 7Ys7-«

Io
9%í 9%á 9%i+ 9%í+1
I0l

band are shown in the side

view of the complete array. LENGTH -19'li2Yí 12" 11V11Vua,"
I.1_"iTxw.{. n3/,- n{6 11%- 111:%i
 210 BUILDING AND USING VHF ANTENNAS
t x i"wood Poly Block
8 x25"tuÓLH9
Fig. 9-59 Element -
6 ft. long
T 220 -MHz Yagi on a 6-foot
wood boom. Poly blocks
each side of the boom sup-
port the modified folded-
di pole driven element. Lat-
POLY BLOCK
ter has sloping lower por-
tion, for matching 52 -ohm
(A) o coax and balun, connected
to lugs at the bottom of
sketch B. With element
lengths and spacings given
D.E
in the side view of the
REF. D2
DI D3 04 D5 array, optimum perfor-
0 o 0 0 o) mance is obtained over the
first 3 to 4 megahertz of
SPACING - 10"-.j_72 82 - 13' 1s the band.
LENGTH - I

261 " 25" 23r 231" 231 231 231"

The antenna is matched by means of a universal The bridge should be at a point in the line a
stub, C, made of the same material as the elements. multiple of a half-wavelength from the antenna, for
It should be cut about 15 inches long, and the greatest ease of adjustment.
suspended under the boom. An adjustable short The 220 -MHz Yagi can be made in the manner
and two sliding clips for connecting the trans- just described, using a dipole made of a single piece
mission line or balun are provided for adjusting the of wire. The universal stub for matching should be
matching. The ends of the stub that connect to the about 28 inches long, to assure an adequate range
dipole are pounded flat with a hammer, and then of adjustment. A variation of the ratio -type folded
drilled to pass the threaded ends of the dipole. dipole was made for the 220 -MHz antenna as
These are held in place by the 4-40 nuts shown in shown at B in Fig. 9-59. Here a flat strip of
B. A ceramic cone standoff insulator (not shown in aluminum comprised the fed portion of the dipole,
the drawing) is fastened to the underside of the and a 3/8 -inch tube the unbroken portion. The
boom. Two pieces of polystyrene similar to that aluminum strip is bolted to the underside of the
used for the dipole mounting blocks, one above tubing at the outer ends. The slope down to the
and one below the matching stub, are fastened to feed point at the polystyrene blocks determines
this cone, clamping the stub in place. the impedance. With the dimensions shown the
The short and the point of connection of the array can be fed with 52 -ohm coax and a balun,
balun are adjusted for zero reflected power, as connected to the lugs at the insulating plates. The
indicated in an SWR bridge connected in the line. SWR is under 1.5 to 1 from 220 to 224 MHz, with
optimum match at about 221.5 MHz.
I B 1'
16

I-
II
Is2

I I I I I I
Fig. 9-60 - Details of the 15-element "WOEYE
15-etement Yagi" for 432 MHz. Director spacing is
50-
Owl
I I I r I' uniform throughout. Five forward directors, all the
same length, are omitted from (A) for clarity.
uNE
, t. t1., t t .. Driven -element mounting and feed arrangement
113..
13 2 123,. 123 123 " 113 u2
B 16 16 32 4 uS uts 1116 are shown in (B). Gusset -plate mounting for the
(A)
LAST 5 OIRECTORS, boom is shown in (C). Note method of holding
ALL 1116 , NOT SHOWN elements in place. Mounting plate dimensions are
uncritical.

X
- WHOLES FOR
U -BOLTS

TEFLON
INSULATOR
1T'X ST X1"
)
o
mailman= c

2 HOLES
FOR MOUNTING
FEEOPOINT GUSSET PLATE
CL COPPER WIRE (C)

i
1
STRA
18+COPPER
44-40 SCREW 'NIX SHEET METAL SCREW
200-0+451 BALANCED FEED
(B) ______
BOOM

ELEMENT (END VIEW1

Yagis for 220 and 432 MHz
211
11 -Element 220 -MHz Yagi
An 11 -element 220 -MHz version of the Yagi
described for 432 can be made using the dimen- TOP VIEW

sions of Fig. 9-59, and adding four more directors

spaced about 19 inches apart. This requires a
12 -foot boom, which should be heavier stock than
the 1 by 1 used for the 6 -foot models. The added
directors can be all the same length as D5, or
tapered 1/8 inch per element, as the others are in
Fig. 9-59. An 11 -element Yagi of this type was
made for 220, using a 3/4 -inch aluminum boom,
and a combination delta and universal -stub match.
If the matching was readjusted for frequency DRIVEN ELEMENT
changes it could be made to work uniformly over
at least 2 MHz. The effect of the metal boom could X 16
be observed. Elements had to be about 1/8 inch
longer than with the wood boom, for the center
frequency of the array to remain the same, but this
is not critical if the matching is adjustable.

WOEYE 15 -Element 432-MHz Yagi

Demonstrating the truth of an earlier statement
that there are many ways to make an effective
Yagi, WOEYE built and tested the 432 -MHz an-
tenna of Fig. 9-60, using quite different element
Fig. 9-61 - I
Gain of the 15 -element Yagi can be
ectors above
increased about 0.5 dB by adding
spacings. All directors are spaced the same, 8-7/16 and below the plane of the other elements. Note
inches, on a 116-inch boom, of 3/4-inch aluminum that the in -line reflector, R1, is spaced closer to the
tubing. Element lengths are tapered, except for the driven element when this is done.
five forward directors, not shown in the drawing,
which are all 11-5/16 inches long. Aluminum rod
3/16 inch diameter is used for all elements. 2 -BAY AND 4 -BAY ARRAYS FOR 432 MHz
Several interesting mechanical and electrical
features are included. The delta arms are brought The 432 -MHz Yagis described above can be
to an insulating block mounted on the boom. used effectively in stacked pairs, or in a 4-bay
Elements are held in position by means of self- system, as shown in Fig. 9-62. For convenience in
tapping screws so placed that they just bear against stacking, delta-matched driven elements are used,
the elements. A gusset plate and two U clamps with open-wire phasing lines and a universal stub.
hold the boom in position on the vertical support. All director elements are as in Fig. 9-58. The driven
Gain has been measured carefully, at over 14 element is 13 inches long, and the reflector is
dB over a reference dipole. This can be raised 0.5 13-1/4 inches. Essential mechanical details of the
dB by the addition of a 3 -element reflector, details supporting frame are shown in Fig. 9-63.
of which are shown in Fig. 9-61. Note that The dimensions of the fanned-out portions of
reflectors 2 and 3, above and below the boom, are the phasing lines are not particularly critical, so
13 inches long, while reflector 1, mounted through long as all four are the same. The sides of the delta
the boom, is 13-1/2 inches. are about 2 inches long, and they fasten to the
Stacking in pairs or sets of four can use driven element about 1-3/8 inches each side of the
dimensions and methods of Fig. 9-62. Feed is set center point. The phasing lines can be 1/2 -inch
up for 50 ohms and a 4:1 balun. spaced TV line, though No. 12 or No. 14, with a

NO.14 SPACED
I/2, 52. LONG
10.14 SPACED Vt.

9-62 -
52' LONG
Fig. Phasing arrange- NO.12 SPACED
ments for two and four I,' 15'LONG
432 -MHz 11 -element Yagis.
Bay spacing of approximately NO.12 SPACED
two wave lengths is set by the ANY COAX ANY COAX V, IS -LONG
length of the phasing lines. AND BALUN AND BALUN
The universal stub matching
device may be used with any S.w.R. BRIDGE
type of transmission line, as
well as with the coaxial line
and balun as shown.
 212 BUILDING AND USING VHF ANTENNAS
Boom\

51»APPROX. ,7 1 X I' CROSS

BRACE

VERTICAL I'X 2' MAIN

-4Z x I' BRACE
SUPPORT HORIZONTAL
SUPPORT

I'X I» CROSS
BRACE
I'X I

BOOM VERTICAL
SUPPORT
6

Fig. 9-63 -
Mechanical details of the 432 -MHz arrays. At the left is a side view of the 44 -element
system. The Masonite gusset plates used to hold the array in alignment are made as shown at the right.
The array is supported on a round wooden closet pole, fastened to the three horizontal members in
the sketch at the left, by means of U clamps.

minimum of spreaders, preferably Teflon rod, will When four bays are to be used additional
be much better and more durable. The lines should bracing is needed, and the gusset plates and
be supported on standoffs at several places, to forward bracing become necessary. The front brace
prevent flexing at the connection points. is 1/2 X 1 -inch stock, bolted between the two
booms to keep them in alignment. The two vertical
Construction supports with the gusset plates are tied together
All -wood construction was used for low cost, horizontally with two 1 X 1 -inch cross braces and a
ease of assembly, and freedom from worry over 1 X 2 -inch main support, as shown. Not shown in
large amounts of metal in the field of the array. the sketch are two 1/2 X 1 -inch wood sway braces
Lightweight wood design would be none too strong that run from the mid-points of the two forward
for large arrays on lower frequencies, but at 432 vertical braces to the 1 X 2 -inch main horizontal
MHz the wood frames are sturdy enough to stand member. These are held in place by small brackets
up longer than most uhf enthusiasts will want to cut to fit from sheet aluminum. The main vertical
stay with one array. support, not shown, is 1-1/4- or 1 -1/2 -inch round
The wood is mostly 1 X 1 stock. Like all closet-pole stock. This is inexpensive and strong,
lumber dimensions for width and thickness, this is and there is no extraneous metal in the array
a misnomer. The actual size is likely to be more proper.
nearly 7/8 X 7/8 inches, but this is not important To make the wood members reasonably durable
for our purposes. It merely makes it impossible to and waterproof they were sprayed with Krylon
give precise dimensions for the supporting frame. before assembly. The Masonite gusset plates were
Get good-quality dry wood, free of knots, and also well soaked with lacquer spray. The whole
preferably a kind that is not subject to severe assembly was painted with ordinary outside white
warping. Most lumber dealers will be glad to advise house paint.
you on the best materials for outdoor use, and
available woods vary around the country.
The holes for the elements are drilled the size Adjustment
of the elements or slightly smaller, and the Matching the array should be done with the
elements are forced into place. Half-inch brass bottom bay at least four feet above ground, if in
wood screws that run in from the top or bottom the position that it will be in use; that is, with the
hold the elements in position firmly. booms horizontal. The region in front of the array
Bracing can be whatever the wind and weather should be free of trees, buildings, wires, or any
conditions in your locality demand. The principal other materials or objects that can reflect 432 -MHz
details of the array frame are given in Fig. 9-63. At energy. A high -gain array has a strong field out
the left is the assembly for two of the 11 -element front. An appreciable reflection back has a marked
bays. The main vertical member, also 1 X 1, is held effect on its impedance. If you don't have a good
perpendicular to the booms by means of gusset large open area, prop the array up with the vertical
plates of 1/4 -inch Masonite, as shown at the right. support in a horizontal position, and the four
If only an 11 -over-11 is to be built, this vertical booms pointing straight up. Ground under the
member can be dispensed with, and the bays array will have little or no effect on its impedance
clamped to the main vertical support by means of in this position, as the power radiated off the back
U clamps. is negligible, for this purpose.
A 48 -Element Collinear for 432 MHz 213
Fig. 9-64 -Element details
and phasing harness for the
48-element 432 -MHz collinear
array. Reflectors, not shown,
are 13-1/4 inches long, with
their inner element ends 3/4
inches apart. Phasing harness,
B, should be spaced no more
than 1/2 inch. Main trans- (A)
mission line can be coax or
balanced line.

-a f --- 3..Max.
a

With an SWR bridge In the coaxial line near the are 13-1/4 inches long, and placed approximately
antenna (preferably some small multiple of a 5-1/2 inches in back of the driven elements. The
wavelength away), adjust the short on the universal latter dimension has little significance as far as
stub and the point of connection of the balun for performance is concerned.
zero reflected power. Once the proper points are Most details of the supporting frame are in Fig.
found, permanent connections can be made. 9-65. The four vertical members, A, the two
smaller horizontal braces, B, and the 24 element
A 48 -ELEMENT COLLINEAR FOR 432 MHz supports, C, are all "1 by 1" stock. This won't cost
much, so get the best you can; free of knots and
The high -gain 432 -MHz collinear of Figs. 9-64 well dried. Take your lumberman's advice as to the
and 65 is inexpensive and easy to build and adjust. kind of wood; the best available varies from one
Though light in weight and mostly made of wood, section of the country to another. As with all
it survived two rough winters in a hilltop location. lumber, the actual size is less than the trade size, so
Materials can be obtained anywhere, and only the the assembling can be done with 2 -inch brass 6-32
simplest tools are needed in its construction. screws, washers, and nuts. These are indicated in
Basically it is made up of four 12 -element the front view.
collinears, each having six half-wave driven ele- The element supports, C, are held in place with
ments, with reflectors. These are arranged in a 2 -inch brass wood screws. Drill holes in the A
square and connected by 2 -wavelength phasing pieces just large enough to pass the screws, and in
lines of open TV line, spaced 1/2 inch. To help to the ends of the C pieces about half the screw
clarify this, in Fig. 9-64, the driven-element sets are diameter. Now drill all the element holes. Paint all
numbered 1 and 2, at the left, and 3 and 4, at the surfaces with a priming coat, and let it dry
right. The phasing harness is shown separately, thoroughly. Coat the inner end of the C pieces
with terminals correspondingly numbered. At the with a good glue or cement, and screw them in
center is a universal stub (see Fig. 8-18), permitting place firmly - but don't overdo it. After the glue is
the array to be fed with any balanced line, or coax dry, paint the frame again.
and a balun. The elements can be about any conducting
We have ignored the 24 reflector elements so material: aluminum wire, welding rod, or what -
far, as they are not connected electrically, and have -you. Put in the reflectors first. They are not
showing them only complicates any drawing. They centered exactly in the supports, but rather are

i3"-±-258 13H
15Z

C
13a"
l --i"
C
13é

Fig. 9.65 - Details of the wooden 13i

supporting frame for the 432-MHz col-
linear array. Vertical members, A, are
1 X 1 X 69 inches; horizontal braces, B,
1 X 1 X 54 inches. Element supports, C,

are 1 X 1 X7 inches. Center brace, D, is

1 X 2 X 54 inches. Assembly dimensions
are given on the drawing.
214 BUILDING AND USING VHF ANTENNAS
placed so that their inner ends are about 3/4 inch needed, physically, but the extra length permits
apart. Put a 1/2 -inch brass wood screw into each looping them back to supports attached to the
support to bite lightly into the element and hold it frame. TV hardware is useful here. Make all
in place. This is also done with each driven phasing-line bends on as large a radius as possible,
element. and try to keep junctions perpendicular.
Each set of six half-wave elements and inter- The universal stub can be any convenient
connecting lines requires only four pieces of wire. material. The element wire was used in the original,
At the upper left of Fig. 9-64, E, F, G, and H are 16 inches long, with the conductors spaced about 1
one piece, and J, K, L, and M are another. inch. When the proper points for the sliding short
Elements N and O are made from pieces of wire and the line connection are found, be sure that
about 13-1/2 inches long. Pound one end of each tight permanent connections are made. The excess
flat on an anvil, wrap it around a wire or rod of the of the stub below the short can be cut off, and the
same diameter, and drill through the overlapping stub grounded for lightning protection, if desired.
flat portions for a 4-40 screw. This makes a loop In the original the stub is mounted on brackets
that can be clamped tightly to the midpoint of the attached to the center vertical support, not shown
interconnecting lines. in the sketch.
For E -F -G -H and J -K -L -M, cut pieces about 52 A U-shaped clamp of sheet aluminum holds the
inches long, feed them through the proper holes in brace D to the supporting mast. Sheets of metal on
the C supports, and bend so that the elements are each side of the brace prevent damage to the
12-1/2 inches long, with inner ends about 1/2 inch wood when it is clamped tightly in its final
apart, and the phasing sections are arranged as position. A similar but smaller clamp and plate
shown in the sketch. Spacing at the cross-over arrangement was made for the lower B member, to
points can be done with 3/8 -inch poly or Teflon steady the assembly.
rods, about 1-1/4 inches long, and either drilled or Adjustment for matching was done with the
notched to pass the phasing wires. Drilling is best, beam in a horizontal position, aiming straight up.
but it creates something of a "threading" problem Two different feed systems have been used: a
in assembling the elements and phasing leads. We coaxial line and balun, and a balanced line. With
gave it up after an initial try, and resorted to the latter, tubular transmitting Twin-Lead runs
notched insulators, cemented in place with epoxy from the stub to an anchor at the top of the tower.
glue. They did not come loose in nearly two years Here it joins a 100 -foot run of No. 12 wire,
on the tower. Spacing of the lines should be 1/2 -inch spaced on Teflon spreaders, one every 6
maintained at 3/4 inch or less. feet. This line is pulled up tightly and is straight
Elements N and O are wrapped around the throughout. At the house end there is a grounded
phasing leads and clamped in place. Trim each to adjustable short, and a balun connection, to
12-1/2 inches over-all length, if necessary. When all complete the run to the antenna relay. The short
elements are in place and they and their phasing and balun connection are adjusted for zero reflec-
sections are lined up properly, put in 1/2 -inch brass ted power on this short section of coax. A mild
wood screws bearing on each element to hold it SWR exists on the balanced line, but this is of no
permanently in position in its C support. consequence, the total loss having been measured
The phasing harness, 1-2, 3-4, and connecting at under 1.6 dB.
line, is made of 1/2-inch open TV line, each piece The array has a very sharp main lobe, with two
53-1/2 inches long, this being 2 wavelengths as fairly large minor lobes on either side - an
checked out for resonance with a dip oscillator at inescapable consequence of the wide spacing be-
432 MHz. Connection is made to the four sets of tween bays. Response is sufficiently broad for
driven elements by soldering to lugs under the nuts uniform results over at least 4 MHz in the vicinity
that hold the center elements in contact with the of 432 MHz, when the stub is adjusted at the
aluminum phasing leads. The lines are longer than middle of the desired frequency range.

THE CORNER REFLECTOR

Corner and plane reflector principles were and spacings was given in Fig. 8-9. Since the
discussed in some detail in Chapter 8. The corner spacing is not critical as to gain, variation of it may
requires but one driven element, and is capable of be used to achieve impedance matching.
giving a very clean pattern and moderate gain, with Gain with a 60- or 90-degree comer with
very broad frequency response. At 220 and 420 1 -wavelength sides runs around 10 dB. A 60 -degree
MHz its size assumes practical proportions, and it comer with 2 -wavelength sides has about 12 dB
can even be used at 144, though usually at less gain. It will be seen that this is not outstanding for
than optimum size. the size of such an array, but there are other
The corner angle can be 90, 60, or even 45 advantages. A corner may be used for several
degrees, but the side length must be increased as bands, for example, or perhaps for uhf television
the angle Is narrowed. The driven element spacing reception, as well as for amateur uhf work.
from the comer can be anything from 0.25 to 0.7 A suggested arrangement for a comer reflector
wavelength for a 90 -degree comer, 0.35 to 0.75 for system is shown in Fig. 9-66. Sheet metal or wire
a 60 -degree one, and 0.5 to 1 for a 45 -degree mesh may be used with equal effectiveness for the
comer. Feed impedance for various corner angles reflecting plane. A series of spines, as shown, is
The Corner Reflector 215
Fig. 9-66 - Con-
struction of a cor-
ner-reflector array.
Frame can be wood
or metal. Reflector
elements are stiff
wire or tubing. Di-
mensions for three
bands are given in
Table 9-11. Reflector
element spacing, G,
is the maximum
that should be used
for the frequency;
closer spacings are
optional. Hinge per-
mits folding for por-
table use.

equally good, if the space between them is kept tower. Principal dimensions for corner reflector
under 0.06 wavelength at the highest frequency for arrays for 144, 220, and 420 MHz are given in
which the reflector is to be used. The frame may Table 9 -II. These dimensions are not critical,
be made of wood, with a hinge at the comer to because of the broad frequency response of any
facilitate portable work or assembling atop a plane -reflector system.

TABLE 9 -II

Dimensions of Corner -Reflector Arrays for 144, 220, and 420 MHz
Dipole
Side to Reflector Reflector Corner Feed
Length Vertex Length Spacing Angle Im-
Band "S" "D" pedance
(MHz) (Inches) (Inches) (Inches) (Inches) (Degrees) (Ohms)
144 65 27.5 48 7-3/4 90 70
144 80 40 48 4 90 150
220` 42 18 30 5 90 70
220 52 25 30 3 90 150
220 100 25 30 screen 60 70
420 27 8-3/4 16-1/4 2-5/8 90 70
420 54 13-1/2 16-1/4 screen 60 70
"Side length and number of reflector elements somewhat below optimum -
slight reduction in gain.

BROAD -BAND ANTENNAS

In addition to utilizing harmonic resonances, as conical. Its pattern will be far from perfect, and its
in most long-wire antennas, there are several ways gain low, but it will work after a fashion. Similarly,
to make an antenna work on more than one band. uhf bow -ties and corner -reflector antennas have
Most are variations of the broad -band dipole been pressed into service on 432 MHz in a pinch.
principle, in which the radiating element Is modi- Performance of the TV conical on 50 MHz may be
fied in shape so that it has no precise electrical improved slightly by extending the elements 3 or 4
length. The "conical" and "bow -tie" antennas of inches.
vhf and uhf TV usage are familiar examples. In all
such antennas the net effect is to make the A 3 -Band Log -Periodic Array
transmission line gradually become the antenna, Amore esoteric idea is the basis of the
the point at which this happens varying with "log-periodic antenna," now widely used in mili-
frequency. tary and commercial stations where many frequen-
A would-be vhf enthusiast who must have an cies must be employed to maintain communication
unobtrusive antenna that will work on 50 through in the hf range. In theory, the frequency range of
220 MHz could do worse than to put up a TV this type of antenna is almost unlimited, and in
216 BUILDING AND USING VHF ANTENNAS

t L

Fig. 9-67 -
The log -periodic array for 140 through 450 MHz looks like a Yagi when viewed from top
or bottom. Actually it has two electrically separate booms, each with a set of elements arranged as
shown in Fig. 9-68. Black objects are wood -block spacers for the booms. Design is by K7RTY.

practice a spread of 4 or 5 to 1 is not uncommon. bled so that when the antenna is viewed from
Arrays of this kind can take many forms. A simple directly above or below it appears somewhat like a
version by K7RTY is shown in Figs. 9-67 through long Yagi, as in the photograph. The two booms
70. are maintained 1-1/4 inches apart, by means of
As author Heslin put it in his June, 1963, QST wooden blocks.
article: "This is an antenna whose resonance Frequency response is determined by the short-
transfers smoothly from one element to the next, est and longest elements. The example is quite
as the frequency is raised." More on the principle uniform in gain and feed impedance, from 140 to
of the log -periodic antenna may be found in a 450 MHz. Gain over this entire range is roughly
November, 1959, QST treatment by Milner, what would be expected from a 3 -element Yagi for
W1FVY. any one frequency.
The version described here is not readily drawn The element mounts were made from inexpen-
or photographed in its complete form, to show full sive TV antenna parts, modified to take an element
details. It has two booms one above the other, as with a threaded end, as shown. K7RTY used
shown in the sketch of the short (front) end, Fig. 1/4 -inch rod for elements, but other sizes are
9-69. Elements are progressively longer and wider suitable if a mounting method is available. Note
spaced as we move toward the back of the array. that two assemblies like that shown in Fig. 9-68 are
The array is fed with coaxial line, which runs inside required. These must be held in alignment, but
the lower boom its entire length. The outer shield insulated from each other.
connects to that boom, and the inner conductor to An array of similar electrical properties was
the upper. Each boom has a set of staggered made by W1CUT, using 1/2 -inch aluminum chan-
elements, as shown in Fig. 9-68. These are assem- nel stock for the booms, and threaded -end 1/8-inch

74 -

D/RECTIDN OF

TRANSMISSION

,9yi

Fig. 9-68 - Dimensions for one

section of the log -periodic vhf
array. The finished array consists
of two of these sections, mounted
one above the other, to give the
appearance of Fig. 9-67.
Broad -Band Antennas 217
Saw

A
conductor
connects
to tkisboont i
yÍ '
Skield connects
to this boom.

Coal' ft?edlisie

Fig. 9-69 -The two booms of the wide -band array are kept in alignment and insulated from each
other by three wooden blocks, left. Short end of the array, right, shows how the array is fed. Lower
boom, with coax inside, acts as an infinite balun.

aluminum rod for elements. The coaxial line ran

the length of one of the channels, apparently
serving as an "infinite balun" in the same way as in
the K7RTY version where the coax runs inside the
boom. This array was tested and found to have an
SWR under 2:1 at 144, 220, and 432 MHz, and
gain averaging 6 dB over this range.
The log -periodic, in common with all broad-
band arrays, does not give something for nothing.
Gain is very low, compared with what a Yagi of the
same physical size would deliver on one frequency, ,41
but the principal weakness is its broad -band nature.
Being almost equally effective across more than a 1 34 I

3 -to-1 frequency range, the antenna presents very

much more of a problem with spurious receiver
responses than does a Yagi, with the latter's
inherent selectivity. By the same token, greater
care must be exercised in keeping down spurious
products in the transmitter. The log-periodic will
accept power on any frequency in its wide range,
and radiate it with some gain. This is quite a
different matter from working with a Yagi that has
gain only over a narrow frequency range, and
considerable rejection ability on most other fre-
quencies.
Fig. 9-70 - Element mounts are made from TV
line standoffs with stainless -steel straps, top. The
Still, if one antenna, with one feedline and one insert is knocked or dulled out, and the clamp
rotator, for several vhf and uhf bands (and perhaps portion bent as shown at the center. Complete
the TV channels in between) is your requirement, mount, with 1/4 -inch element in place, has lock
the log-periodic will probably fill it as well as any and tension nut inside the clamp, to avoid need for
one antenna can. drilling the latter.

STIFFENING ANTENNA TUBING

Lightweight tubing used in antenna work can may not be so readily available, but it is worth
be made much stronger by inserting a wooden plug shopping around for.
at the point where strain is expected. This is Also good for this application, though heavy:
particularly effective for a light vertical support for ordinary thick -wall water pipe. It may bend, but it
beams, where it runs through the tower bearing. will not collapse or break. The so-called 1 -inch
This is the point where the tubing is most likely to water pipe (inside diameter) is fine. Lightweight
collapse. Stuffing it with a wood dowel that is a steel tubing is probably the least desirable of all,
close fit will strengthen it greatly. A foot or two because of its susceptibility to collapsing. Alum-
either side of the tower bearing is all that is inum or dural tubing offers the best compromise
needed. Rug-pole stock is available at most lumber- between strength and light weight, of readily
yards. Get hard wood for greatest strength. This available materials.
 Chapter 10

Fm -Theory and Techniques

Methods of radiotelephone communication by Fm has some impressive advantages for vhf
frequency modulation were developed in the 1930s operation, especially when compared to a-m. With
by Major Edwin Armstrong in an attempt to fm the modulation process takes place in a
reduce the problems of static and noise associated low-level stage. The modulation equipment re-
with receiving a-m broadcast transmissions. The quired is the same, regardless of transmitter power.
primary advantage of fm, the ability to produce a The signal may be frequency multiplied after
high signal-to-noise ratio when receiving a signal of modulation, and the PA stage can be operated
only moderate strength, has made fm the ideal class C for best efficiency, as the "final" need not
mode for mobile communications services and be linear.
quality broadcasting. The disadvantages, the wide In recent years there has been increasing use of
bandwidth required, and the poor results obtained fm by amateurs operating around 29.6 MHz in the
when an fm signal is propagated via the ionosphere 10 -meter band. The vhf spectrum now in popular
(because of phase distortion), has limited the use use includes 52 to 53 MHz, 146 to 147.5 MHz, and
of frequency modulation to the 10 -meter band and 440 to 450 MHz.
the vhf/uhf section of the spectrum.

FREQUENCY AND PHASE MODULATION

It is possible to convey intelligence by modula-
ting any property of a carrier, including its
frequency and phase. When the frequency of the
carrier is varied in accordance with the variations in
a modulating signal, the result is frequency modu-
lation (fm). Similarly, varying the phase of the
carrier current is called phase modulation (pm).
Frequency and phase modulation are not indepen-
dent, since the frequency cannot be varied without
also varying the phase, and vice versa.
The effectiveness of fm and pm for communica-
tion purposes depends almost entirely on the
receiving methods. If the receiver will respond to
frequency and phase changes but is insensitive to
'k
+ amplitude changes, it will discriminate against most
forms of noise, particularly impulse noise such as is
set up by ignition systems and other sparking
á :
devices. Special methods of detection are required
to accomplish this result.
Modulation methods for fm and pm are simple
and require practically no audio power. There is
also the advantage that, since there is no amplitude
variation in the signal, interference to broadcast
reception resulting from rectification of the trans-
mitted signal in the audio circuits of the be receiver
is substantially eliminated.
Fig. 10-1 - One advantage of the fm mode is the
Frequency Modulation
large amount of surplus equipment which can be
converted for amateur use. AR EC and civil -defense Fig. 10-2 is a representation of frequency
groups increasingly are using vhf fm for local
communications networks because of the low cost modulation. When a modulating signal is applied,
and high reliability of surplus commercial gear. the carrier frequency is increased during one half
Here WA6ECK checks Motorola Handi-Talkies cycle of the modulating signal and decreased
which have been modified for 6-meter operation at during the half cycle of opposite polarity. This is
the San Fernando Emergency Operating Center indicated in the drawing by the rf cycles occupying
during an earthquake emergency. less time (higher frequency) when the modulating

218
Frequency and Phase Modulation 219
signal is positive, and more time (lower frequency)
when the modulating signal is negative. The change l MINIM II I IIIDlillllllilIÍIIIIillll
in the carrier frequency (frequency" deviation) is
proportional to the instantaneous amplitude of the IIIIIIIIIIIIIIIII I I
9I I
' 91II1
modulating signal, so the deviation is small when
the instantaneous amplitude of the modulating
signal is small and is greatest when the modulating Nfores/wafNadu/atingS\
signal reaches its peak, either positive or negative.
As shown by the drawing, the amplitude of the
signal does not change during modulation.

Phase Modulation
If the phase of the current in a circuit is
changed there is an instantaneous frequency
change during the time that the phase is being
shifted. The amount of frequency change, or
deviation, depends on how rapidly the phase shift
is accomplished. It is also dependent upon the total
amount of the phase shift. In a properly operating Fig. 10-2 -
Graphical representation of frequency
pm system the amount of phase shift is propor- modulation. In the unmodulated carrier at A, each
tional to the instantaneous amplitude of the rf cycle occupies the same amount of time. When
the modulating signal, B, is applied, the radio
modulating signal. The rapidity of the phase shift is frequency is increased and decreased according to
directly proportional to the frequency of the the amplitude and polarity of the modulating
modulating signal. Consequently, the frequency signal.
deviation in pm is proportional to both the
amplitude and frequency of the modulating signal.
The latter represents the outstanding difference Fig. 10-3 shows how the amplitudes of the
between fm and pm, since in fm the frequency carrier and the various sidebands vary with the
deviation is proportional only to the amplitude of modulation index. This is for single -tone modula-
the modulating signal. tion; the first sideband (actually a pair, one above
and one below the carrier) is displaced from the
Fm and Pm Sidebands carrier by an amount equal to the modulating
frequency, the second is twice the modulating
The sidebands set up by fm and pm differ from frequency away from the carrier, and so on. For
those resulting from a-m in that they occur at example, if the modulating frequency is 2000 Hz
integral multiples of the modulating frequency on and the carrier frequency is 53,000 kHz, the first
either side of the carrier rather than, as in a -m, sideband pair is at 52,998 kHz and 53,002 kHz,
consisting of a single set of side frequencies for the second pair is at 52,996 kHz and 53,004 kHz,
each modulating frequency. An fm or pm signal the third at 52,994 kHz and 53,006 kHz, etc. The
therefore inherently occupies a wider channel than amplitudes of these sidebands depend on the
a-m. modulation index, not on the frequency deviation.
The number of "extra" sidebands that occur in Note that, as shown by Fig. 10-3, the carrier
fm and pm depends on the relationship between strength varies with the modulation index. (In
the modulating frequency and the frequency devia- amplitude modulation the carrier strength is con-
tion. The ratio between the frequency deviation, in stant; only the sideband amplitude varies.) At a
hertz, and the modulating frequency, also in hertz,
is called the modulation index. That is,

Modulation index - Carrier frequency deviation

Modulating frequency
Examp1s; The maximum frequency devia-
tion in an fin transmitter is 3000 Hz either
side of the carrier frequency. The modulation
index when the modulating frequency is 1000
Hz is
Modulation index3000 -3
At the same deviation with 3000 -Hz modula
tion the index would be 1; at 100 Hz it so sc xo
would be 30, and so on. aoouurar. iNocx
In pm the modulation index is constant regard-
less of the modulating frequency; in fm it varies Fig. 10-3 -
How the amplitude of the pairs of
sidebands varies with the modulation index In an
with the modulating frequency, as shown in the fm or pm signal. If the curves were extended for
above example. In an fm system the ratio of the greater values of modulation index it would be
maximum carrier -frequency deviation to the high- seen that the carrier amplitude goes through zero
est modulating frequency used is called the devia- at several points. The same statement also applies
tion ratio. to the sidebands.
220 FM -THEORY AND TECHNIQUES
modulation index of approximately 2.4 the carrier or pm signal is about equivalent to a well -
disappears entirely. It then becomes "negative" at modulated a-m signal of one-fourth the carrier
a higher index, meaning that its phase is reversed as power. But with a suitable fm receiver, fm is equal
compared to the phase without modulation. In fm to or better than a-m, watt for watt, in the kinds of
and pm the energy that goes into the sidebands is voice communication for which fm is now widely
taken from the carrier, the total power remaining used.
the same regardless of the modulation index. Three deviation amounts are now common
Since there is no change in amplitude with practice: 15, 5, and 2.5 kHz, which in the current
modulation, an fm or pm signal can be amplified vernacular of fm users, are known as wide-band,
without distortion by an ordinary class C ampli- narrow -band, and silver -band, respectively. (See
fier. The modulation can take place in a low-level box above.) The 2.5-3 kHz deviation (called nbfm
stage and the signal can then be amplified by either by OTs) was popular for a time on the vhf bands
frequency multipliers or straight -through ampli- and 10 meters after World War II. Deviation figures
fiers. are given for the frequency swing in one direction.
If the modulated signal is passed through one or The rule -of -thumb for determination of bandwidth
more frequency multipliers, the modulation index requirements for an fm system is:
is multiplied by the same factor that the carrier 2 (oF + FAmax)
frequency is multiplied. For example, if modula- where OF is one half of the total frequency
tion is applied on 8150 kHz and the final output is deviation, and FAmax is the maximum audio
on 146.7 MHz, the total frequency multiplication frequency (3 kHz for communications purposes).
is 18 times, so if the frequency deviation is 500 Hz Thus, for narrow-band fm, the bandwidth equals
at 8150 kHz it will be 9000 Hz at 146.7 MHz. 2 (5 + 3) or 16 kHz. Wide -band systems need a
Frequency multiplication offers a means for ob- 36 -kHz receiver bandwidth.
taining practically any desired amount of frequen- It should be emphasized that "narrow -band"
cy deviation, whether or not the modulator itself is and "wide -band" as used here are vernacular terms.
capable of giving that much deviation without In interpreting amateur regulations, anything that
distortion. does not qualify for use below 52.5 or 29.0 MHz
Bandwidth (see box) is wide -band fm. Future plans for
amateur use of fm for 220 MHz and higher bands
FCC amateur regulations (Part 97.61) limit could well employ bandwidths comparable to fm
the bandwidth of F3 (frequency or phase broadcasting (150 kHz), especially in the uhf
modulation) to that of an a -m transmission portion of the spectrum.
having the same audio characteristics below
29.0 MHz and in the 50.1- to 52.5 -MHz Comparison of Fm and Pm
frequency segment. Greater bandwidths are
allowed from 29.0 to 29.7 MHz and above Frequency modulation cannot be applied to an
52.5 MHz. amplifier stage, but phase moduhation can; pm is
therefore readily adaptable to transmitters em-
If the modulation index (with single -tone ploying oscillators of high stability such as the
modulation) does not exceed 0.6 or 0.7, the most crystal-controlled type. The amount of phase shift
important extra sideband, the second, will be at that can be obtained with good linearity is such
least 20 dB below the unmodulated carrier level, that the maximum practicable modulation index is
and this should represent an effective channel about 0.5. Because the phase shift is proportional
width about equivalent to that of an a -m signal. In to the modulating frequency, this index can be
the case of speech, a somewhat higher modulation used only at the highest frequency present in the
index can be used. This is because the energy modulating signal, assuming that all frequencies
distribution in a complex wave is such that the will at one time or another have equal amplitudes.
modulation index for any one frequency compon- Taking 3000 Hz as a suitable upper limit for voice
ent is reduced as compared to the index with a sine work, and setting the modulation index at 0.5 for
wave having the same peak amplitude as the voice 3000 Hz, the frequency response of the speech-
wave. amplifier system above 3000 Hz must be sharply
A major advantage of fm or pm, especially for attenuated, to prevent excess splatter. (See Fig.
vhf operators living in areas of high population 10-4.) Also, if the "tinny" quality of pm as
density, is the elimination or great reduction of received on an fm receiver is to be avoided, the pm
interference problems commonly associated with must be changed to fm, in which the modulation
any form of amplitude modulation, including ssb. index decreases in inverse proportion to the modu-
Also, the modulating equipment is relatively simple lating frequency. This requires shaping the speech -
and inexpensive. However, assuming the same amplifier frequency -response curve in such a way
unmodulated carrier power, narrow -band fm or pm that the output voltage is inversely proportional to
is not as effective as a -m unless a receiver designed frequency over most of the voice range. When this
especially for fm reception is employed. is done the maximum modulation index can only
As shown in Fig. 10-3, -at an index of 0.6 the be used to some relatively low audio frequency,
amplitude of the first sideband is about 25 percent perhaps 300 to 400 Hz in voice transmission, and
of the unmodulated-carrier amplitude. This com- must decrease in proportion to the increase in
pares with a sideband amplitude of 50 percent in frequency. The result is that the maximum linear
the case of an a-m signal modulated 100 percent. freqúency deviation is only one or two hundred
When copied on an a-m receiver, a narrow -band fm Hz, when pm is changed to fm. To increase the
Methods of Frequency Modulation 221
Qt
Fig. 10-4 - Output frequency spectrum of a
narrow -band fm transmitter modulated by a 1 -kHz
tone. O
.- a
nlug Bum
SPECTRUM ANALYZER
MODEL -1L

u
deviation for narrow -band requires a frequency 1.111.
multiplication of 8 times or more.

'
It is relatively easy to secure a fairly large
frequency deviation when a self-controlled oscil-
lator is frequency modulated directly. (True fre- 1 IRE
quency modulation of a crystal -controlled oscil-
lator results in only very small deviations and so
I 1111 111'Ul
requires a great deal of frequency multiplication.)
The chief problem is to maintain a satisfactory i nnñin
degree of carrier stability, since the greater the
inherent stability of the oscillator the more diffi- n I< I t} lf--
0 `-®+
Llvfi 1 Ilfliti
_fii1f11iR11IiYQiiiiiiY
05
'.

cult it is to secure a wide frequency swing with

linearity. .f.Efi- ;jIDTH FACTOR _

METHODS OF FREQUENCY MODULATION

Direct Fm transistor. It increases when R1 is made smaller in
comparison with C3. It also increases with an
A simple and satisfactory device for producing
increase in LC ratio in the oscillator tank circuit.
fm in the amateur transmitter is the reactance However, for highest carrier stability it is desirable
modulator. This is a vacuum tube or transistor to use the largest tank capacitance that will permit
connected to the rf tank circuit of an oscillator in the desired deviation to be secured while keeping
such a way as to act as a variable inductance or within the limits of linear operation.
capacitance.
Fig. 10-5A is a representative circuit. Gate 1 of
REACTANCE MODULATOR
the modulator MOSFET is connected across the
oscillator tank circuit, C1 -L1, through resistor RI OSC. TANK
TO 05C.
and blocking capacitor C2. C3 represents the input
-
.001 C2 50
capacitance of the modulator transistor. The resis-
tance of R1 is made large compared to the
CI
reactance of C3, so the rf current through R1 -C3 RI
will be practically in phase with the rf voltage
appearing at the terminals of the tank circuit.
However, the voltage across C3 will lag the current RFC
LnH
by 90 degrees. The rf current in the drain circuit of +12v
the modulator will be in phase with the gate RFC ...C3 +_º5/if
ImH
voltage, and consequently is 90 degrees behind the AUDIO
INPUT 2ro-005,115V
current through C3, or 90 degrees behind the rf
tank voltage. This lagging current is drawn through T.005
4711
the oscillator tank, giving the same effect as though (A)
an inductance were connected across the tank. The
frequency increases in proportion to the amplitude
of the lagging plate current of the modulator. The
audio voltage, introduced through a radio - VARACTOR REACTANCE
frequency choke, varies the transconductance of MODULATOR
the transistor and thereby varies the rf drain OUTPUT
current. 151E°
The modulated oscillator usually is operated on
a relatively low frequency, so that a high order of
5uF 50,ÚW RFC
+1sv
carrier stability can be secured. Frequency multi- 50/Aí
pliers are used to raise the frequency to the output
470 390
frequency desired.
A reactance modulator can be connected to a
crystal oscillator as well as to the self-controlled
type as shown in Fig. 10-5B. However, the result- +12V
ing signal can be more phase -modulated than it is
frequency -modulated, for the reason that the 9.1V
frequency deviation that can be secured by varying (B)
the frequency of a crystal oscillator is quite small. m
The sensitivity of the modulator, Fig. 10-5A, Fig. 10-5 -
Reactance modulators using (A) a
(frequency change per unit change in gate voltage) high-transconductance MOSFET and (B) a varactor
depends on the transconductance of the modulator diode.
222 FM -THEORY AND TECHNIQUES
PHASE MODULATOR sensitivity and rf power output from the modula-
RF
INPUT .001
RF ted amplifier. An optimum figure for Q appears to
75 47 OUTPUT
(
be about 20; this allows reasonable loading of the
MPF102
modulated amplifier and the necessary tuning
RFC variation can be secured from a reactance modula-
1mH
tor without difficulty. It is advisable to modulate
RFC
at a low power level.
+12V Reactance modulation of an amplifier stage
I m.H
usually results in simultaneous amplitude modula-
(A) 1ov tion because the modulated stage is detuned from
resonance as the phase is shifted. This must be
AUDIO INPUT
eliminated by feeding the modulated signal
through an amplitude limiter or one or more
"saturating" stages - that is, amplifiers that are
operated class C and driven hard enough so that
PRE-EMPHASIS variations in the amplitude of the input excitation
DE -EMPHASIS produce no appreciable variations in the output
7500 amplitude.
.75H
5 o TO For the same type of reactance modulator, the
II AUDIO speech -amplifier gain required is the same for pm
FROM AMP.
DISCRIMINATOR 1,01 as for fm. However, as pointed out earlier, the fact
lOK
1µF that the actual frequency deviation increases with
the modulating audio frequency in pm makes it
AF necessary to cut off the frequencies above about
OUT (C) 3000 Hz before modulation takes place. If this is
(B) not done, unnecessary sidebands will be generated
at frequencies considerably away from the carrier.
Fig. 10-6 -
(A) The phase -shifter type of phase SPEECH PROCESSING FOR FM
modulator. (B) Pre -emphasis and (C) de -emphasis
circuits. The speech amplifier preceding the modulator
follows ordinary .design, except that no power is
taken from it and the of voltage required by the
A change in any of the voltages on the -
modulator input usually is small not more than
modulator transistor will cause a change in rf drain 10 or 15 volts, even with large modulator tubes,
current, and consequently a frequency change. and only a volt or two for transistors. Because of
Therefore it is advisable to use a regulated power these modest requirements, only a few speech
supply for both modulator and oscillator. stages are needed; a two -stage amplifier consisting
of two bipolar transistors, both resistance coupled,
Indirect Fm will more than suffice for crystal, ceramic or high -Z
The same type of reactance modulator circuit dynamic microphones. For more information on
that is used to vary the tuning of the oscillator speech amplifiers see The Radio Amateur's Hand-
tank in fm can be used to vary the tuning of an book, Chapter 13.
amplifier tank and thus vary the phase of the tank Several forms of speech processing produce
current for pm. Hence the modulator circuit of worthwhile improvements in fm system perfor-
Fig. 10-5A or 10-6A can be used for pm if the mance. It is desirable to limit the peak amplitude
reactance transistor or tube works on an amplifier of the audio signal applied to an fm or pm
tank instead of directly on a self-controlled oscil- modulator, so that the deviation of the fm trans-
lator. If audio shaping is used in the speech mitter will not exceed a preset value. This peak
amplifier, as described above, fm instead of pm will limiting is usually accomplished with a simple
be generated by the phase modulator. audio clipper which is placed between the speech
The phase shift that occurs when a circuit is amplifier and modulator. The clipping process
detuned from resonance depends on the amount of produces high -order harmonics which, if allowed to
detuning and the Q of the circuit. The higher the pass through to the modulator stage, would create
Q, the smaller the amount of detuning needed to unwanted sidebands. Therefore, an audio low-pass
secure a given number of degrees of phase shift. If filter with a cut-off frequency between 2.5 and 3
the Q is at least 10, the relationship between phase kHz is needed at the output of the clipper. Excess
shift and detuning (in kHz either side of the clipping can cause severe distortion of the voice
resonant frequency) will be substantially linear signal. An audio processor consisting of a com-
over a phase -shift range of about 25 degrees. From pressor and a clipper, such as described in Chapter
the standpoint of modulator sensitivity, the Q of 13 of The Radio Amateur's Handbook, has been
the tuned circuit on which the modulator operates found to produce audio with a better sound (i.e.,
should be as high as possible. On the other hand, less distortion) than a clipper alone.
the effective Q of the circuit will not be very high To reduce the amount of noise in some fm
if the amplifier is delivering power to a load since communications systems, an audio shaping net-
the load resistance reduces the Q. There must work called pre -emphasis is added at the trans-
therefore be a compromise between modulator mitter to proportionally attenuate the lower audio
Testing an Fm Transmitter 223
frequencies, giving an even spread to the energy in when operating through a repeater. The rigors of
the audio band. This results in an fm signal of mobile and portable operation make a frequency
nearly constant energy distribution. The reverse is check of a channelized transceiver at regular
done at the receiver, called de -emphasis, to restore intervals a recommended procedure.
the audio to its original relative proportions. Frequency meters generally fall into two cate-
Sample circuits are shown in Fig. 10-6. gories: the heterodyne type and the digital coun-
ter. For amateur use, the vhf/uhf counterparts of
FM EXCITERS
the popular BC-221 frequency meter, the TS -174
and TS -175, will provide sufficient accuracy. Fre-
Fm exciters and transmitters take two general quency counters that will work directly up to 500
forms. One, shown at Fig. 10-7A, consists of a MHz and higher are available, but their cost is high.
reactance modulator which shifts the frequency of The less expensive low -frequency counters can be
an oscillator to generate an fm signal directly. employed using a scaler, a device which divides an
Successive multiplier stages provide output on the input frequency by a preset ratio, usually 10 or
desired frequency, which is amplified by a PA 100. The Heathkit IB-102 scaler may be used up to
stage. This system has a disadvantage in that, if the 175 MHz, using a counter with a 2 -MHz (or more)
oscillator is free running, it is difficult to achieve upper frequency limit. If the counting system does
sufficient stability for vhf use. If a crystal - not have a sufficient upper frequency limit to
controlled oscillator is employed, unless the am- measure the output of an fm transmitter directly,
ount that the crystal frequency is changed is kept one of the frequency -multiplier stages can be
small, it is difficult to achieve equal amounts of sampled to provide a signal in the range of the
frequency swing. measurement device. Alternatively, a crystal -
The indirect method of generating fm shown in controlled converter feeding an hf receiver which
Fig. 10-7B is currently popular. Shaped audio is has accurate frequency readout can be employed,
applied to a phase modulator to generate fm. As if a secondary standard is available to calibrate the
the amount of deviation produced is very small, receiving system.
then a large number of multiplier stages is needed
to achieve wide -band deviation at the operating Deviation and Deviation Linearity
frequency. In general, the system shown at A will Asimple deviation meter can be assembled
require a less complex circuit than that at B, but following the diagram of Fig. 10-8A. This circuit
the indirect method (B) often produces superior
was designed by K6VKZ. The output of a wide -
results.
band receiver discriminator (before any de -
emphasis) is fed to two amplifier transistors. The
TESTING AN FM TRANSMITTER output of the amplifier section is transformer
coupled to a pair of rectifier diodes to develop a dc
Accurate checking of the operation of an fm or voltage for the meter, Ml. There will be an
pm transmitter requires different methods than the indication on the meter with no signal input
corresponding checks on an a -m or ssb set. The because of detected noise, so the accuracy of
common forms of measuring devices either indicate the instrument will be poor on weak signals.
amplitude variations only (a milliammeter, for To calibrate the unit, signals of known devia-
example), or their indications are most easily tion will be required. If the meter is to be set to
interpreted in terms of amplitude. read 0-15 kHz, then a 7.5-kHz deviation test signal
The quantities to be checked in an fm trans- should be employed. RI is then adjusted until M1
mitter are the linearity and frequency deviation reads half scale, 50 µA. To check the peak
and the output frequency, if the unit uses crystal deviation of an incoming signal, close both Si and
control. The methods of checking differ in detail. S2. Then, read the meter. Opening first one switch
and then the other will indicate the amount of
Frequency Checking positive and negative deviation of the signal, a
The crystal -controlled, channelized operation check of deviation linearity.
that is now popular with amateur fm users requires
that a transmitter be held close to the desired Measurement of Deviation Using
channel, at least within a few hundred hertz, even Bessel Functions
in a wide -band system. Having the transmitter on Using a math relationship known as the Bessel
the proper frequency is particularly important function it is possible to predict the points at

ANT.
MIC. AUDIO CLIPPER REACTANCE
o-AMP. FILTER MODULATOR
05C. MULT. MOLT. PA

(A)

" milAUDIO
AMP
L
fT CLIPPER
FILTER
XTA L
OSC.
PHASE
MODULATOR
MULT. MULT. MULT PA -o
ANT.

(B)
Fig. 10-7 - Block diagrams of typical fm exciters.
224 FM -THEORY AND TECHNIQUES
PEAK DEVIATION METER

TO
DISCRIMINATOR
o

100K

.1

(A)

Audio Deviation Produced

Frequency 1st Null 2nd Null 3rd Null
905.8 Hz ±2.18 kHz ± 5.00 kHz ± 7.84 kHz
1000.0 Hz ±2.40 kHz ± 5.52 kHz ± 8.65 kHz
1500.0 Hz ±3.61 kHz ± 8.28 kHz ±12.98 kHz
1811.0 Hz ±4.35 kHz ±10.00 kHz ±15.67 kHz
(B) 2000.0 Hz ±4.81 kHz ±11.04 kHz ±17.31 kHz
2079.2 Hz ±5.00 kHz ±11.48 kHz ±17.99 kHz
2805.0 Hz ±6.75 kHz ±15.48 kHz ±24.27 kHz

Fig. 10-8 -(A) Schematic diagram of the deviation meter. Resistors are 1/2 -watt composition and
capacitors are ceramic, except those with polarity marked, which are electrolytic. CR1-CR3, incl., are
high-speed silicon switching diodes. R1 is a linear -taper composition control, and Si, S2 are spst toggle
switches. T1 is a miniature audio transformer with 10,000 -ohm primary and 20,000 -ohm center -tapped
secondary (Triad A31X). (B) Chart of audio frequencies which will produce a carrier null when the
deviation of an fm transmitter is set for the values given.

which, with certain audio -input frequencies and circuit for applying the adjustable voltage is shown
predetermined deviation settings, the carrier out- in Fig. 10-9A. The battery voltage is 3 to 6 volts
put of an fm transmitter will disappear completely. (two or more dry cells in series). The arrows
Thus, by monitoring the carrier frequency with a indicate clip connections so that the battery
receiver, it will be possible by ear to identify the polarity can be reversed.
deviation at which the carrier is nulled. A hetero-
dyne signal at either the input or receiver i -f is
required so that the carrier will produce a beat
note which can easily be identified. Other tones TO MOO. GRID
RESISTOR
will be produced in the modulation process, so
some concentration is required by the operator (A)
when making the test. With an audio tone selected
from the chart (Fig. 10-8B), advance the deviation
control slowly until the first null is heard. If a 2000
higher order null is desired, continue advancing the
control further until the second, and then the
third, null is heard. Using a carrier null beyond the
third is generally not practical.
For example, if a 905.8 -Hz tone is used, the
transmitter will be set for 5 -kHz deviation when
the second null is reached. The second null
achieved with a 2805 -Hz audio input will set the
transmitter deviation at 15.48 kHz. The Bessel-
function approach can be used to calibrate a
deviation meter, such as the unit shown in Fig. 2000
10-8A. -a .a to
Grid -Voltage Deviation
Reactance -Modulator Fm Calibration (B) From Operating Bias

It is possible to calibrate a reactance modulator

by applying an adjustable dc voltage to the
Fig. 10-9 - (A) Dc method of checking frequency
deviation. R1 is 500 to 1000 ohms. (B) A typical
modulator input and noting the change in oscil- curve of frequency deviation versus gate voltage for
lator frequency as the voltage is varied. A suitable an fm modulator.
Reception of Fm Signals 225
The oscillator frequency deviation should be polarity positive with respect to ground the radio
measured by using a receiver in conjunction with frequency will move in one direction when the
an accurately calibrated frequency meter, or by voltage is increased, and in the other direction
any means that will permit accurate measurement when the polarity is reversed. When several read-
of frequency differences of a few hundred hertz. ings have been taken a curve may be plotted to
One simple method is to tune in the oscillator on show the relationship between input voltage and
the receiver (disconnecting the receiving antenna, if frequency deviation.
necessary, to keep the signal strength well below A sample curve is shown in Fig. 10-9B. The
the overload point) and then set the receiver BFO usable portion of the curve is the center part which
to zero beat. Then increase the dc voltage applied is essentially a straight line. The bending at the
to the modulator input from zero in steps of about ends indicates that the modulator is no longer
1/2 volt and note the beat frequency at each linear; this departure from linearity will cause
change. Then reverse the battery terminals and harmonic distortion and will broaden the channel
repeat. The frequency of the beat note may be occupied by the signal. In the example, the
measured by comparison with a calibrated audio - characteristic is linear 1.5 kHz on either side of the
frequency oscillator. Note that with the battery center or carrier frequency.

RECEPTION OF FM SIGNALS
Receivers for fm signals differ from others The Fm Receiver
principally in two features: there is no need for Block diagrams of an a-m/ssb and an fm
linearity preceding detection; (in fact, it is advan- receiver are shown in Fig. 10-11. Fundamentally,
tageous if amplitude variations in signal and back- to achieve a sensitivity of less than one microvolt,
ground noise can be "washed out"), and the an fm receiver requires a gain of several million -
detector must be capable of converting frequency
variations in the incoming signal into amplitude eSVIATION
LIMITS
variations.
Frequency -modulated signals can be received I I

after a fashion on any ordinary receiver. The

receiver is tuned to put the carrier frequency part
way down on one side of the selectivity curve.
When the frequency of the signal varies with
modulation it swings as indicated in Fig. 10-10,
resulting in an a-m output varying between X and J
Y. This is then rectified as an a -m signal. á
With receivers having steep-sided selectivity a
curves, the method is not very satisfactory because
the distortion is quite severe unless the frequency
FREQUENCY
deviation is small, since the frequency deviation
and output amplitude are linear over only a small
part of the selectivity curve. Slope detection is also
Fig. 10-10 -
detector characteristics. Slope
Fm
detection, using the sloping side of the receiver
unsatisfactory if the receiver selectivity does not selectivity curve to convert fm to a -m for subse-
match the signal deviation closely. quent detection.

A -M RECEIVER
SPEAKER
ANT.
C-= R F
AMP
MIXER
1 LCIDEJ_
DET. f 1
UDIO
AMP.

Fig. 10-11 - Block diagrams of (A) an a-m (B) an

OSC. fm receiver. Dark borders outline the sections that
are different in the fm set.

F M RECEIVER

ANT.
o-- RF
AM P.
MIXER
WIDE
FILTER
I

AMP.
-F
LIMITER - FREQ.
DISCM.
AUDIO
AMP.
226 FM -THEORY AND TECHNIQUES
FM FILTERS

Center Nonimal Ultimate Impedance (r) Insertion Crystal

Manufacturer Model Frequency Bandwidth Rejection In Out Loss Disaiminator
KVG (1) XF-9E 9.0 MHz 12 kHz 90dB 1200 1200 3 dB XD9-02
KVG (1) XF-107A 10.7 MHz 12 kHz 90dB 820 820 3.5 dB XD107-01
KVG (1) XF-107B 10.7 MHz 15 kHz 90dB 910 910 3.5 dB XD107-01
KVG (1) XF-107C 10.7 MHz 30 kHz 90dB 2000 2000 4.5 dB XD 107-01
Heath Dynamics (2) 21.5 MHz 15 kHz 90dB 550 550 3 dB
Heath Dynamics (2) 21.5 MHz 30 kHz 90 dB 1100 1100 2 dB
E.S. (3) FB-6D 10.7 MHz 15 kHz 80 dB 950 950 2 dB AB-1C
ES. (3) 10-MA 10.7 MHz 30 kHz 80dB 2000 2000 4 dB AB -1C
E.S. (3) EL-3A 11.5 MHz 36 kHz 70 dB 50 50 4 dB AL -1
E.S. (3) DR-9 21.4 MHz 20 kHz 40dB 750 750 5 dB AR -10
Clevite (4) TCF4-12D3CA 455 kHz 12 kHz 60 dB 40k 2200 6 dB

Clevite (4) TCF4-18G45A 455 kHz 18 kHz 50 dB 40k 2200 6 dB

Clevite (4) TCF6-30DS5A 455 kHz 30 kHz 60 dB 20k 1000 5 dB

Fig. 10-12 -A list of fm -bandwidth

filters that are available to amateurs. Manufacturer's addresses are
as follows:
3) E. S. Electronic Labs, 301 Augustus, Excelsior
1) Spectrum International, P.O.Box 87, Topsfield, Springs, MO 64024.
MA 01983. 4) Semiconductor Specialists, Inc., P.O. Box
2) Heath Dynamics, Inc., 6050 N. 52nd Avenue, 66125, O'Hare International Airport, Chicago,
Glendale, AZ 85301. L 60666. (Minimum order $5.00.)
I

too much total gain to be accomplished with For the average ham, the use of a high -selec-
stability on a single frequency. Thus, the use of the tivity filter in a homemade receiver offers some
superheterodyne circuit has become standard prac- simplification of the alignment task. Following the
tice. Three major differences will be apparent from techniques used in ssb receivers, a crystal or
a comparison of the two block diagrams. The fm ceramic filter should be placed in the circuit as
receiver employs a wider bandwidth filter and a close as possible to the front end -
at the output
different detector and has a limiter stage added of the first mixer, in most cases. Fig. 10-12 lists a
between the i -f amplifier and the detector. Other- number of suitable filters that are available to
wise the functions, and often the circuits, of the if, amateurs. Prices for these filters are in the range of
oscillator, mixer, and audio stages will be the same $10 to $30. Experimenters who wish to "roll their
in either receiver. own" can use surplus hf crystals, as outlined in
In operation, the noticeable difference between ARRL's Single Sideband for the Radio Amateur,
the two receivers is the effect of noise and or ceramic resonators.
interference on an incoming signal. From the time One item of concern to every amateur fm user
of the first spark transmitters, "rotten QRM" has is the choice of i -f bandwidth for his receiver, as
been a major problem for amateurs. The limiter both 15- and 5 -kHz deviation are now in common
and discriminator stages in an fm set can eliminate use on the amateur bands. A wide -band receiver
a good deal of impulse noise, except that noise can receive narrow-band signals, suffering only
which manages to acquire a frequency -modulation some loss of audio in the detection process.
characteristic. Accurate alignment of the receiver However, a wideband signal will be badly distorted
i -f system and phase tuning of the detector are when received on a narrow -band receiver.
required to achieve good noise suppression. Fm At this point it seems reasonable to assume that
receivers perform in an unusual manner when an increasing fm activity and continued production of
interfering signal is present, exhibiting a character- commercial narrow-band transceivers 'may grad-
istic known as the capture effect. The loudest ually shift amateur operation to a 5 -kHz deviation
signal received, even if it is only two or three times standard. But, as with the a-m operators, the
stronger than other stations on the same frequen-
cy, will be the only transmission demodulated. By
comparison, an S9 a -m or cw signal can suffer
noticeable heterodyne interference from an S2
carrier.

Bandwidth
Most fm sets that use tubes achieve i -f selec-
tivity by using a number of over -coupled trans-
formers. The wide bandwidth and phase-response
characteristic needed in the i -f system dictate
Fig. 10-13 -
Representation of limiter action.
Amplitude variations on the signal are removed by
careful design and alignment of all interstage the diode action of the grid and plate -current
transformers. saturation .
Reception of Fm Signals 227
wide -band enthusiasts will be around for some time
to come, lured by inexpensive surplus wide -band
gear. It should be remembered that true wide -band
fm has merit - on frequencies where there is room
r
for it. . I

(A)
Limiters 1-r :
1 ! _
When fm was first introduced, the main selling lI
point used for the new mode was the noise -free
reception possibilities. The circuit in the fm recei-
ver that has the task of chopping off noise and
amplitude modulation from an incoming signal is
the limiter. Most types of fm detectors respond to
both frequency and amplitude variations of the
signal. Thus, the limiter stages preceding the
detector are included to "cleanse" the signal so
that only the desired frequency modulation will be
demodulated. This action can be seed in Fig.
10-14.
Limiter stages can be designed using tubes,
transistors, or ICs. For a tube to act as a limiter,
the applied B voltages are chosen so that the stage
will overload easily, even with a small amount of
signal input. A sharp -cutoff pentode such as the
6BH6 is usually employed, with little or no bias
applied. As shown In Fig. 10-13, the input signal Fig. 10-14 -(A) Input wave form to a limiter stage
limits when it is of sufficient amplitude so that shows a -m and noise. (B) The same signal, after
diode action of the grid and plate -current satura- passing through two limiter stages, is devoid of a-m
tion clip both sides of the input signal, producing a conponents.
constant -amplitude output voltage.
Obviously, a signal of considerable strength is limiting at a desired level. The input -signal voltage
required at the input of the limiter to assure full required to start limiting action is called the
clipping, typically several volts for tubes, one volt limiting knee, referring to the point at which
for transistors, and several hundred microvolts for collector (or plate) current ceases to rise with
ICs. Limiting action should start with an rf input increased input signal. Modern ICs have limiting
of 0.2 µV or less, so a large amount of gain is knees of 100 mV for the circuit shown in Fig.
required between the antenna terminal and the 10-15C, using the CA3028A or MC1550G, or 200
limiter stages. For example, the Motorola 80D has µV for the Motorola MC1590G of Fig. 10-15D.
eight tubes before the limiter, and the solid-state Because the high -gain ICs such as the CA3076 and
MOTRAC receivers use nine transistor stages to get MC1590G contain as many as six or eight active
sufficient gain before the first limiter. The new ICs stages which will saturate with sufficient input, one
offer some simplification of the i -f system as they of these devices provides superior limiter perfor-
pack a lot of gain into a single package. mance compared to a pair of tubes or transistors.
When sufficient signal arrives at the receiver to
start limiting action, the set quiets - that is, the Detectors
background noise disappears. The sensitivity of an
fm receiver is rated in terms of the amount of The first type of fm detector to gain popularity
input signal required to produce a given amount of was the frequency discriminator. The characteristic
quieting, usually 20 dB. Current practice using the of such a detector is shown in Fig. 10-16. When the
new solid-state devices can produce receivers which fm signal has no modulation, and the carrier is at
achieve 20 dB quieting with 0.15 to 0.5 µV of point O, the detector has no output. When audio
input signal. input to the fm transmitter swings the signal higher
A single tube or transistor stage will not provide in frequency, the rectified output increases in the
good limiting over a wide range of input signals. positive direction. When the frequency swings
Two stages, with different input time constants, lower the output amplitude increases in the nega-
are a minimum requirement. The first stage is set tive direction. Over a range where the discriminator
to handle impulse noise satisfactorily while the is linear (shown as the straight portion of the line),
second is designed to limit the range of signals the conversion of fm to a-m which is taking place
passed on by the first. At frequencies below 1 MHz will be linear.
it is useful to employ untuned RC-coupled limiters in the search for a simplified fm detector, RCA
which provide sufficient gain without a tendency developed a circuit that has now become standard
toward oscillation. in entertainment radios which eliminated the need
Fig. 10-15A shows a two -stage limiter using for a preceding limiter stage. Known as the ratio
sharp -cutoff tubes, and 10-15B has transistors in detector, this circuit is based on the idea of
two stages biased for limiter service. The base bias dividing a do voltage into a ratio which is equal to
on either transistor may be varied to provide the ratio of the amplitudes from either side of a
228 FM -THEORY AND TECHNIQUES
+250V

1sr LIMITER 2ND LIMITER

Fig. 10-15
iter
- Typical
circuits using
lim- 4700
(A)
tubes, (B) transistors, (C) a 68146 6BH6
INPUT OF
DISC. TRANS.
differential IC, (D) a high - VI V2
r
gain linear IC. FROM 470
I -F
AMP. RFC1 x 1Í 1a
2.5 'NH
6BH6 (2)
3/ \4 100 47K

3300
6.3V
AC (A)
+250 V
1ST LIMITER 2ND LIMITER
2N64I - 2N641 PART OF
DISC. TRANS.

EXCEPT AS INDICATED, DECIMAL

VALUES OF CAPACITANCE ARE
IN MICROFARADS 10 I; OTHERS
APE IN PICOFARA OS I PF OR yyF I;
RESISTANCES ARE IN OHMS,
11.1000,1~000000

(B)

LIMITER
+9V PART OF
DISCRIMINATOR
U2
S
TRANSFORMER
r- --
I- F TRANS

.001
TO
DETECTOR

(C) +IIV

discriminator-transformer secondary. With a detec- rather than series opposing, as in the standard
tor that responds only to ratios, the input signal discriminator circuit. The recovered audio is taken
may vary in strength over a wide range without from a tertiary winding which is tightly coupled to
causing a change in the level of output voltage - the primary of the transformer. Diode-load resistor
fm can be detected, but not a-m. In an actual ratio values are selected to be lower (5000 ohms or less)
detector, Fig. 10-18, the do voltage required Is than for the discriminator.
developed across two load resistors, shunted by an A practical discriminator circuit is shown In
electrolytic capacitor. Other differences include Fig. 10-17. The fm signal is converted to a -m by
the two diodes, which are wired in series aiding transformer Tl. The voltage induced in the Ti
DEVIATION secondary is 90 degrees out of phase with the
LIMITS current In the primary. The primary signal is
introduced through a center tap on the secondary,
coupled through a capacitor. The secondary volt-
ages combine on each side of the center tap so that
the voltage on one side leads the primary signal
while the other side lags by the same amount.
When retained, these two voltages are equal and of
opposite polarity, resulting in zero -voltage output.
A shift in input frequency causes a shift in the
phase of the voltage components that results in an
increase of output amplitude on one side of the
Fig. 10-16
inator.
- The characteristic of an fm discrim- secondary, and a corresponding decrease on the
other side. The differences in the two changing
Reception of Fm Signals 229
DISCRIMINATOR
Ti
-1
FROM
I -F
AMP.
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
IN MICROFARADS IyF I; OTHERS
ARE IN PICOFARA DSIPF OR y3+F1;
RESISTANCES ARE IN OHMS;
1.1000, M91000000

+B RFC

Fig. 10-17 - Typical frequency-discriminator circuit used for fm detection. T1 1s a Miller 12-C45.

FROM
voltages, after rectification, constitute the audio I -F RATIO DETECTOR
output. AMP. PLATE
The sensitivity of the ratio detector is one half 1000
that of the discriminator. In general, however, the -NWT
transformer design values for Q, primary-secondary 3.30= I
16800
coupling, and load will vary greatly, so the actual 25V ..._
performance differences between these two types _- 330
of fns detectors are usually not significant. Either 650007
circuit can provide excellent results. In operation,
the ratio detector will not provide sufficient
limiting for communications service, so this detec-
tor also is usually preceded by at least a single
limiting stage. AUDIO
Another popular discriminator circuit which +B
OUTPUT

does not require a special transformer is shown in

Fig. 10-19. Often called the Travis discriminator, Fig. 10-18 -A
ratio detector of the type often
used in entertainment radio and TV sets. T1 is a
this arrangement uses two tuned circuits, one set ratio -detector transformer such as the Miller 1606.
above the i -f center frequency and the other set
below by the same offset amount, feeding separate
free fm detector designs. The crystal discriminator
detectors. The outputs of the two detectors are utilizes a quartz resonator, shunted by an inductor,
combined by R1, which is adjusted for maximum
in place of the tuned -circuit secondary used in a
a-m rejection. Because it can be built for any i -f,
discriminator transformer. A typical circuit is
the Travis discriminator is useful as an fm detector shown in Fig. 10-20. Some commercially made
for the Heath SB-series receivers, ARC -5s, and crystal discriminators have the input -circuit induc-
others with nonstandard i-fs. tor, L1, built in (C1 must be added) while in other
types both L1 and Cl must be supplied by the
New Detector Designs
builder. Fig. 10-20 shows typical component val-
The difficulties often encountered in building ues; unmarked parts are chosen to give the desired
and aligning LC discriminators have inspired re- bandwidth. Sources for crystal discriminators are
search that has resulted in a number of adjustment- listed in Fig. 10-12.

1N67A
.001
HEP802
.001

1N67A

R1
OUTPUT
.01 1M
500K
INPUT 2200 2200 BALANCE

HEP802 +12V
Fig. 10-19 - The Travis discriminator. .01
D
.001 1N67A

.001
1M 1N67A
2200
,77-7
2200

+12V
 230 FM -THEORY AND TECHNIQUES
CRYSTAL DISCRIMINATOR
r 100K

10K To
Audio
Amp.

From
Limiter
Cl

- 100K -j
Fig. 10-20 -
Crystal discriminator. Cl and L1 are resonate at the intermediate frequency. C2 is equal in
value to C3. C4 corrects any circuit imbalance so that equal amounts of signal are fed to the detector
diodes.

two stages provide "hard" limiting (Fig. 10-23B) of

the i -f signal (Fig. 10-23A). The output pulse train
from UM. (Fig. 10-23C) is fed to a divide -by -four
circuit consisting of two flip-flops, U2A and U2B.
The output of U2A (Fig. 10-23D) triggers a
monostable multivibrator, which consists of R2,
.4 Cl and U1D.
.. v The period of the multivibrator Is set to be less
.....-- M re
-
than half of the period of the i -f signal. For an i-f
of 455 kHz, the period is set at 800 nanoseconds.
^
Output from the multivibrator consists of 800 -
nanosecond pulses whose repetition rate varies in
_. direct proportion to the variation in frequency of
the fm signal (Fig. 10-23E). The output of U1D is
amplified by U1E and converted to an audio signal
by the de -emphasis network (Fig. 10-23F). Maxi-
mum recovered audio occurs at a 50 -percent duty
cycle.
Fig. 10-21 - The digital detector. Advantages of the pulse -counting detector in-
clude linear detection over wide frequency ranges
and inherent quieting. The self-squelching action
Digital Detector results because the digital circuits remain inactive
until the input signal reaches a threshold sufficient
The integrated circuit, allowing complex func- to trigger U2B. As the ICs are operating in the
tions to be performed in a very small package, has saturated mode, they provide excellent limiting,
caused a revolution in fm receiver design. For sheer and no preceding limiter stage is required. The
simplicity, the pulse -counting detector, shown in primary disadvantages are a low -voltage, high -
Figs. 10-21 and 10-22, has intrinsic appeal. Using current power requirement (3.6 V at 50 mA for
inexpensive RTL logic, the detector requires 200 the circuit shown in Fig. 10-22), and the very low
mV of signal from the recéiver i-f system. The level of recovered audio. Upper frequency limit for
oscilloscope photographs of Fig. 10-23 show this the RTL circuits is about 2 MHz, but other faster
digital detector at work. The first inverter, U 1 A, is logic families (TTL or MECL II) can be employed
biased to operate as a linear amplifier. The next for i -f systems in the 2- to 20 -MHz range.

AMPLIFIER LIMITER DIVIDER MULTIVIBRATOR

AUDIO
U28 U2A OUTPUT
0
001
C 9
66K
w 13 6 9 10

oH FF FF
R3
FROM 330
I -F

6800 :
10K
TO PIN 4 TO PIN 11
+3.6 V
ON U1 AND U2 ON WAND U2

-
Fig. 10-22 Diagram of the integrated -circuit
detector. Resistors are 1/2 -watt composition and
capacitors are disk ceramic. Labeled components
U1
U2 -- Motorola MC789P hex inverter.
Motorola MC790P dual flip-flop.
not listed below are marked for text reference.
Reception of Fm Signals 231
Fig. 10-23 - Waveforms showing the operation of
a digital detector. (A & B) Input and output of the
limiter section. (C & D) Input and output of the
divide -by -four circuit. (E) Output of the multi - (A)
vibrator. (F) Recovered audio after de -emphasis.

The PLL
Now that the phase-locked loop (PLL) has been
reduced to a single IC package, this circuit is
destined to revolutionize some facets of receiver (B)
design. Introduction by Signetics of a PLL in a
single flat -pack IC, followed by Motorola and
Fairchild (who are making the PLL in separate
building-block ICs), allows a builder to get to work
with a minimum of bother.
A basic phase-locked loop (Fig. 10-24A)consists
of a phase detector, a filter, a dc amplifier, and a (C)
voltage -controlled oscillator (VCO). The VCO runs
at a frequency close to that of an incoming signal.
The phase detector produces an error voltage if any
difference in frequency exists between the VCO
and the i -f signal. This error voltage is applied to D)
the VCO. Any changes in the frequency of the
incoming signal are sensed at the detector and the
error voltage readjusts the VCO frequency so that
it remains locked to the intermediate frequency.
The bandwidth of the system is determined by a
filter on the error -voltage line.
Because the error voltage is a copy of the audio
variations originally used to shift the frequency of
the transmitter, the PLL functions directly as an
fm detector. The sensitivity achieved with the (E)
Signetics NE565 PLL is good - about 1 mV for
the circuit shown in Fig. 10-24B. No transformers
or tuned circuits are required. The PLL bandwidth
is usually two to ten percent of the i-f for fm
detection. Components R1 -C1 set the VCO to near (F)
the desired frequency. C2 is the loop -filter capaci-
tor which determines the capture range -
that
range of frequencies over which the loop will
acquire lock with an input signal, initially starting
out of lock. The NE565 has an upper frequency
limit of 500 kHz; for higher frequencies, the PLL DETECTOR

1101.1PHASE
I
o-
-F DETECTOR
AUDIO
OUTPUT
(A)

+5V

Fig. 10-24- (A) Block diagram of a PLL demodu-

lator. (B) Complete PLL circuit.
FROM AUDIO
1- F OUTPUT.

(B)
-5V
 232 FM-THEORY AND TECHNIQUES
NE561, which is usable up to 30 MHz, can be FM RECEIVER DESIGN
employed.
Until recently, fm receivers have followed the
Squelch Considerations design shown in block -diagram form in Fig. 10-26.
One or two rf amplifier stages and a double -
The high gain used in fm receivers accounts for conversion frequency scheme were used. Greater
another requisite, an audio squelch. With 140 dB band occupancy has inspired both commercial and
or more of amplification before the detector, the amateur receiver designers to work on the dynamic
noise level is high when no signal is being received. range and strong -signal-handling capabilities of fin
To allow monitoring for long periods without receivers. As cross modulation and overload effects
fatigue, a squelch is employed to quiet the noise are primarily caused by the rf amplifier and first
until a signal is heard. mixer, a good deal of research has gone into the
Squelch circuits are generally classified as application of various solid-state devices for use in
carrier operated and noise operated Early com- receiver front ends. The result of this work has
mercial fm squelch designs used the noise -operated been a vast improvement in fm receiver signal -
circuit, while later models in tube receivers com- handling capability.
bined in the carrier- and noise -operated squelch in Two devices, the FET and the hot-carrier diode,
an effort to achieve extra sensitivity. Carrier signal are responsible for the revolution in front-end
levels were sensed by the voltage change produced circuits. Both can be operated to provide square -
across the grid-return resistor of the limiter stage. law response as mixers. The sensitivity of either
However, the excellent quieting sensitivities of type is such that the rf amplifier can be eliminated
solid-state receivers have allowed a return to the in many cases. Although the hot -carrier diode has
noise -operated squelch, which can be opened relia- been used by the amateur fraternity, the device
bly by a 0.1-µV rf signal, when the receiver hasn't been popular with fm receiver designers for
designer employs the latest techniques. two reasons. To assure linear mixing, the level of
A number of squelch circuits were described by oscillator injection to a diode mixer must be at
Danz in QST for September, 1969. All of the least 10 dB above the strongest signal to be
designs noted in this article, however, were inten- received. Even with a balanced bridge of hot -carrier
ded for use with high -output detectors. Many of diodes, the power required from the local oscillator
the newer detectors require a squelch circuit with is considerable. Also, the diode mixer must be
high gain, such as shown in Fig. 10-25. Here, the followed by a low -noise i-f preamplifier for best
output from the detector is split into two compon- overall receiver noise figure, introducing a new area
ents by Ql, one of which is fed to the audio gate for cross-modulation effects to appear. FET de-
while the other is amplified by Q2. Only the noise vices exhibit a slightly better noise figure (by 1 to
component of the audio signal (frequencies above 3 dB), and thus, are usually chosen over hot -carrier
5 kHz) is passed by Ll-Cl to the noise rectifier. diodes.
The do output of the noise rectifier is amplified by The first field-effect transistors to gain wide
Q3. When noise is present, Q4 and Q5 will be held acceptance were the junction types (JFET).
on by the output from Q3. When the receiver Though still the leader in the low -noise-figure
quiets, Q4 and QS will shut off, opening the audio competition, the JFET also requires a rather large
-
gate. Duration of the squelch "tail" the length of
time that the audio gate remains open after the
amount of power from the oscillator chain when
used as a mixer. Early metal -oxide semiconductors
-
input signal disappears is determined by C. (MOSFET) were easily damaged by static charges

SPLITTER NOISE AMP. NOISE RECTIFIER SWITCH AUDIO GATE

470 100
+ p+12V
2200 25)uF
V
15V

HEP801 1500
FROM
Disc. 01 SQUELCH GAIN
CI
.047 1N67Á HEP51 1,2F
3300 15V
I MPS-A10
1M
J Q3 + ILAUDIO
OUTPUT
MPS -A10

2200
39K
tr 100
88 MI-1 fi 10K
1N67A
3300
EXCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE ARE IN MICROFARADS I yF I ;
OTHERS ARE IN PICOFARADS ( PF OR yyFr,
RESISTANCES ARE IN OHMS',
k=IOOO,M 1000 000.

Fig. 10-25 - Diagram of the high-sensitivity squelch. L1 is an 88-mH telephone loading coil.
Fm Receiver Design 233
ANT.
o--R- RFAMP. MIXER --d-- FILTER MIXER I -F AMP. DISC.

OSC (A)
OSC.

ANT.
o--- RFAMP. MIXER FILTER I -F AMP DISC

(B)
OSC.

ANT.
7 FILTER MIXER ----~ I F AMP. - -{ DISC.

OSC.
(C)
Fig. 10-26 - Block diagrams of the fm receiver designs discussed in the text.
and voltage spikes, and were considered too deli- wide-band fm will provide greater noise sup-
cate for mobile service. However, the development pression in local work. However, in practice, vhf
of the diode -protected dual-gate models has made fm mobiles experience greater range than pre-
the MOSFET as hardy as the bipolar transistor. viously found on a-m due to the output powers
Only one volt of oscillator injection is required to employed which are considerably higher than those
the high-impedance gate 2 of a typical dual -gate common on a -m.
MOSFET mixer, which eliminates the high injec-
tion level necessary for a JFET. Operating Practices
The designer has a choice of two basic ap- Amateur fm practice has been to retain the
proaches to the layout of a new fm receiver. He fixed -frequency channelized capability of the com-
can use single conversion (Fig. 10-26B). But, to mercial equipment. VFOs and tunable receivers
provide sufficient gain before the limiter, he must tend to be unsatisfactory because of the require-
employ an rf amplifier, and worse, use a bipolar - ment for precise frequency netting. An off-
transistor mixer to achieve high conversion gain. frequency signal will be received with distortion
Even with an rf amplifier stage, getting sufficient and will not have full noise rejection.
i-f gain with stability can be a problem. Alter- Channelized operation with squelched receivers
natively, a dual -conversion scheme can be em- permits continuous monitoring of the active fre-
ployed where sufficient overall gain can be ob- quencies. Long, time-consuming calls and CQs are
tained in the i -f stages. With this design, Fig. not necessary (or appreciated) to establish com-
10-26C, the rf stage can be eliminated if sufficient munications, as all receivers on the channel "come
rf selectivity can be achieved before the first mixer alive" with the operator's first word. Natural, short
without seriously degrading the sensitivity of the transmissions are usually encouraged. The old
receiver.
monopoly switch routine, where the operator gabs
to himself for 10 minutes at a time, will get him
FM COMMUNICATIONS invited off a busy fm channel. Some channels are
calling channels on which extended ragchewing is
Though information on fm theory and practice
discouraged, whereas other channels, or the same
has been available to the amateur for many years, channel in another area, may be alive with chatter.
this mode was largely neglected until the early This is a matter of local determination, influenced
1960s. Then large quantities of used commercial by the amount of activity, and should be respected
fm mobile equipment became available for amateur by the new operators and the transient mobile
use, creating new interest. Originally designed to operator alike. Some groups have adopted the use
cover frequency ranges adjacent to amateur bands, of the "10 code" which was originated for law -
this equipment is easily retuned for amateur use. enforcement communications. However, plain lan-
One feature of fm is its noise -suppression guage in most cases is as fast and requires no
capability. For signals above the receiver threshold, clarification or explanation to anyone.
wide -band fm has a signal-to-noise ratio advantage
over a -m as a result of its greater "intelligence Standards
bandwidth." This same increased bandwidth, how-
ever, results in a much more abrupt signal thres- Standard channel frequencies have been agreed
hold effect, causing weak signals to suddenly upon to permit orderly growth and to permit
disappear. The generality can be made that a-m has communications from one area to another. On two
a greater range in weak signal work but that meters, it has been agreed that any frequency used
 234 FM -THEORY AND TECHNIQUES
will fall on increments of 30 kHz,
beginning at Private Line, Channel Guard, and Quiet Chan-
146.01 MHz. 146.94 MHz (or "nine -four") is the
national calling frequency. On six meters, nel.
national calling frequency is 52.525 MHz, with
the -
Down channel Communications circuit from the
other channels having a 40 -kHz spacing beginning machine to th e control point.
Up channel - Communications
at 52.56 MHz. Ten -meter fm activity can be and/or control
found circuit from the control point to the machine.
on 29.6 MHz. Usage of the 420 -MHz band
varies
from area to area, as it is used for control channels, Open repeater - A machine where
transient
repeaters, and remote bases, as will be discussed operators are welcome.
later. Subdivision of the 220- and 420 -MHz bands Closed repeater - A machine where use by
by modes and channels, on a "gentlemen's agree- nonmembers is not encouraged. (When heavy
ment" basis was under discussion at the time of expenditures are involved, freeloaders are not
publication of this book. Watch QST for develop- popular.)
ments.
Fm Bibliography
APPENDIX Armstrong, "A Method of Reducing Distur-
bances in Radio Signalling by a System
Fm Jargon of Frequency Modulation," Proc. IRE,
May, 1936.
Diplexer -A device to allow two transmitters to Crosby, "Frequency Modulation Noise
use a single antenna. Characteristics," Proc. IRE, April, 1937.
Duplex - Simultaneous transmissions between two Noble, "Frequency Modulation Funda-
stations using two frequencies. mentals," QST, August, 1939.
Duplexer - A device to allow simultaneous trans- Grammer and Goodman, "Wide -Band FM In
mission and reception on a single antenna. Amateur Communication," QST, Jan-
Simplex - Alternating transmission between two
uary and February, 1940.
Hund, Frequency Modulation, McGraw-Hill
or more stations using one frequency. Book Company, 1942.
Low band -30 to 50 MHz. Also, the six-meter Grammer, "Amateur FM," Technical Topics,
amateur band. QST, October, 1946. (First discussion of
High band - 148 to 174 MHz. Also, the two compatible narrow -band fm and a -m.)
-meter
amateur band. Goldsmith et al, Frequency Modulation, in
Remote base -
A remotely controlled station, two volumes, RCA Review, RCA, 1948.
Rider and Uslan, FM Transmission and
usually simplex.
Machine - Either a repeater or a remote base. Also Reception, John F. Rider Publisher,
1948.
called a "box."
Lytel, Two -Way Radio, McGraw-Hill Book
-
COR Carrier-operated relay.
Company, 1959.
CTCSS - Continuous tone -controlled squelch Hadlock, "Wide -Band FM Gear for 220
system. Continuous subaudible tone (250 Hz or Mc.," QST, March, 1961.
lower) transmitted along with the audio to Pre-Progress Line Diagrams, in two volumes,
allow actuation of a repeater or receiver only Mobile Radio Department, General Elec-
by transmitters so equipped. More frequently tric Company, 1968.
referred to by various trade names such as Wolf, FM Schematic Digest, Two -Way Radio
Engineers, 1970.
 Chapter 11

Fm Transmitters, Receivers
and Accessories
Fm is now in regular use on the 28- to
The construction of equipment and accessories
420 -MHz bands. Equipment for the 10 -meter
for the fm mode will be covered in this chapter.
band is described in The Radio Amateur's Hand-
The reader should review the material in Chapter
10 to obtain the theoretical background
of book, as are power supplies suitable for use with
the projects described here. Vertical polarization
frequency -modulation techniques before attemp-
is standard practice for most fm communications.
ting to build individual projects from this chap-
Any of the popular antennas described in this
ter. Chapter 10 also contains information about
the adjustment of fm gear which will be needed book will be suitable when mounted to provide
vertical radiation. References for additional infor-
when testing a piece of homemade equipment. mation on the fm mode are listed at the end of
in this
Many of the projects described elsewhere
Chapter 10.
book are applicable to fm communications as
well as a -m, cw, and ssb.

RECEIVING FM
standard chassis is also suitable as a base. The
FM RECEIVING ADAPTERS
layout of the stages should be kept in a straight
line so that rf feedback paths can be avoided.
To put the older tube receivers such as the
Point-to-point wiring is used throughout.
75A, HRO, and Super Pro models into fm ser-
vice, the receiving adapter shown in Fig. 11-2 was Alignment
designed. Filament and plus B voltages are taken
from the companion receiver. Obviously, the bet- "Lining up" the adapter takes time and test
ter the basic receiver, the better will be the equipment. A VTVM or microammeter plus a
performance of the fm receiving system. For this signal generator are required. Good alignment
application sets with high -gain i-f amplifier sec- cannot be accomplished by ear; if the necessary
tions and a broad-band selectivity position (such test instruments aren't available, they should be
are
as the SP -400, SP-600, SX-73, and R-390) borrowed.
excellent choices. Receivers that have only a
To start, check the alignment of the communi-
6 -kHz or narrower bandwidth may need an
extra
cations receiver, following the manufacturer's in-
i -f amplifier stage in the fm adapter
in order to structions, to be sure that the rf and i -f stages are
tap the receiver i -f at the output of the second "peaked" before the fm adapter is installed. Two
mixer. Of course, a converter will also be re- simple internal modifications are required in the
quired with the basic receiver to copy vhf fm receiver, as shown in Fig. 11-2B and C. If the
signals.
A sample of the receiver i -f signal is passed to
Tl, a 455 -kHz i -f transformer, which feeds ampli-
fier/limiter VI. A low screen voltage and signal
bias enhance the limiting characteristic of the
tube. Further "hard" limiting action is provided
by the two sections of V2, a 12AT7. A sample of
the grid current of V2A is available at TP1, a test
point used during alignment. A commercially
made discriminator transformer converts the fm
signal to a-m; the a -m is detected by CR1 and
CR2. An RC de -emphasis network is included to
match the standard pre -emphasis used on fm Y

transmitters. Audio amplification is provided by

V3 - in some receivers with high -gain audio
systems this stage may not be necessary.
The adapter is constructed on an aluminum
channel which is 11 inches long, 2 inches wide,
and 1-3/4 inches high. A 1/4-inch lip is included Fig. 11-1 - The fm adapter, wired for connection
on one side as a mounting foot. A Minibox or a to a Collins 75A2.

235
 236 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
t-F AMPLIFIER
455 KHz
LIMITER DISCRIMINATOR
455 KHz
6DC6 1/212AT7 33 '/J2AT7 455 KHz
CRt
JI TI T2 IN 67A

+250V

EXCEPT AS INDICATED, DECIMAL VALUES OF

6.3V
CAPACITANCE ARE IN MICROFARADS l yF 1 ;
AC OTHERS ARE IN PICOFARADS 1 PF OR PF1.
RESISTANCES ARE IN OHMS;
t+1000, M+1000 000.

Fig. 11-2 -
(A) Schematic diagram of the 455-kHz
fm adapter. Resistors are 1/2 -watt composition;
capacitors are disk ceramic, except those with
polarity marked, which are electrolytic.

J2 L1 --
J1, J2 Phono jack, panel mount.
430-850-µH slug -tuned variable inductor
(Miller 42A684CBI).
R1
Ti
T2
---Audio -taper composition control.
-f transformer, 455 kHz (Miller 913-C1).
I

Discriminator transformer, 455 kHz (Miller

(B) 913-CD).
FM
-
TP1, TP2 Tip jack (Johnson 105 -XX).
FROM
REC.M
TO AUDIO
AMP. (B) Diagram of the connections to use the fm
adapter with a communications receiver. The tap
DET.
to the -f stage is through a 50-pF disk -ceramic
i

capacitor. If the receiver has a wide -band -f i

(C) system, the connection should be made to the last

intermediate -frequency amplifier; for narrow i-fs,
tap the first i -f stage. (C) Audio connections.

receiver has a wide i -f bandwidth, a sample of the To align the discriminator, set the receiver
i -f signalcan be taken from the plate of the last i-f selectivity at the broad position, and connect the
stage. Otherwise, the tap should be made at the voltmeter to TP2. Voltage at this test point will
plate of the first i -f amplifier, and an extra stage, a swing both plus and minus, so a zero -center
duplicate of VI, included in the adapter. Short meter or VTVM with a lead -reversing switch
lengths of shielded cable are used to carry the i -f should be employed. Set the secondary of the
signal to the adapter and to return audio to the
discriminator transformer for a zero -voltage in-
receiver -
see Fig. 11-2C. Some units (75A2, dication on the meter. Then vary the signal -
HRO-50) which have provision for fm adapters generator frequency plus or minus 15 kHz. Going
already have a front-panel switch wired for this off center frequency in one direction will pro-
purpose. duce positive voltage at TP2, while going in the
Connect the signal generator to the receiver, other direction generates negative voltage. The
and set the generator to produce an S9 reading primary of the transformer must be set so that,
on the receiver signal-strength meter. The receiver for example, if a shift down in frequency by 5
crystal filter should be switched to its most kHz produces plus 2 volts, then a change of 5
selective position to assure that the incoming kHz in the other direction should produce minus
signal is being heterodyned to exactly 455 kHz. 2 volts. Unfortunately, the two adjustments on
Then, with a voltmeter or microammeter con- the discriminator transformer are interlocking, so
nected to TP1, adjust both sections of Ti and Ll considerable experimentation is required. Also,
for maximum limiter current. The receiver i -f the tuning of the preceding stages, if not centered
stage being "tapped should also be realigned to on 455 kHz, will affect the discriminator lin-
compensate for the capacitance of the adapter earity. The first time around, a half hour or more
cable. of alignment and realignment is usually required
 237
A Solid -State Adapter

-5'
_
r.e a *
- -
--
- .
°e

Fig. 11-3 -
In this bottom view, the input
transformer is to the left, followed by the i -f
amplifier, limiter, and detector. On the far right are Fíg. 11-4 -
The solid-state fm adapter is construc-
mounted
the audio-amplifier stage and gain control. ted on a 6 x 2 -inch etched -circuit board,
on a homemade chassis.

needed to
voltage for half a turn or so of the slugs will be
to achieve equal swings in output the phase tuning of the discriminator.
equal swings in frequency a linear response. - is
complete
One further check of the discriminator
as an A SOLID-STATE ADAPTER
required. An impulse-generating device, such
shaver, should be switched on, and the
electric
to point in Tubes are seldom used in current designs. For
receiver, set for a -m detection, tuned
is strong. Then, those builders who prefer to be "up with the
the spectrum where the noise a solid-state version of the 455 -kHz adap-
the dis- times,"
switch to the fm adapter and adjust constructed. Using IC limiter/amplifier,
criminator transformer for best suppression
of the ter was
signal and miniature i -f transformers, the unit
requires
noise pulses. If the alignment with the
25 mA at 12 V for power. See Fig. 11-5.
has been completed properly, only a only
generator
LIMITER DISCRIMINATOR
AMPLIFIER/LIMITER
455 KHz 455 KHz T3 3
CR11
1N67A
6
Ti T2 2N1306
J1
4n,5 4n5 01
C .L
C6
FROM>
I -F
( L88-5-11-
MILLER
e,
3
2
UI

8807 MILLER 8806

330
6.8K
R8
AMP 2
G
470
07

EXCEPT AS INDICATED, DECIMAL R2

VALUES OF CAPACITANCE ARE 1000
IN MICROFARAOS (pF I;OTHERS
R9
ARE IN PICOFARADS(pr OR yyF1; 47
ARE IN ONMs; AUDIO
RESISTANCES
tI000.M 1000000
(A)
.1 c.3 R3 R10
GAIN
500K
T
2200
o AUDIO
V
+12V OUTPUT
Qt
ELD C5 C6 R8 8
Ci +32V
U1JUMPER

R9
R6
Q2
Ji- OUTPUT J2
C1t
Pit

C9
JUMPER
lii R2 C3 RC4 Ell R7 Rg
\
R5
(B)
C2 R3 -
o

are 1/4- or 1/2-watt composition

Fig. 11-5 -
(A) Diagram of the 455 -kHz narrow -band adapter. Resistorswhich are electrolytic. Components
and capacitors are disk ceramic, except those with polarity marked, -board location.
with reference numbers that are not listed below are noted for circuit
J1,J2-Miniature 1/2 -watt composition control.
Phono receptacle, panel mount. T3 Miniature discriminator transformer, 455 -
R1 -
T1,T2 - Miniature 455 -kHz -f transformer, 455 i U1
kHz (Miller 8806).
-
Motorola MC1590G.
kHz (Miller 8807).
(B) Template and parts -layout diagram (not to scale).
238 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
The Motorola MC1590G provides 70 dB gain, and the same as for the tube version. The bandwidth
hard limiting action superior to that obtained of the miniature transformers restricts this adap-
with the tube version. ter to narrow -band reception. However, builders
The unit is built on a 2 X 6 -1/2-inch circuit wishing a wideband version can use the J. W.
board; a template is given in Fig. 11-5B. Because Miller 8811 miniature coils combined with a
of the high gain of the IC stage, a shield is 12-pF coupling capacitor to form a wide -band
required across pins 4 and 6 to isolate the input transformer.
from the output. Alignment and installation are

A RECEIVER FOR 6 -METER FM

A block diagram of the 6 -meter receiver is
outlined in Fig. 11-7. The front-end circuit uses
40673 dual -gate MOSFETs for the if amplifier
and mixer. Toroid coils were chosen for the
signal circuits because of their self-shielding prop-
erties. A crystal -controlled JFET oscillator pro-
vides injection voltage for Q2. Trimmer C3 is
included to move the oscillator frequency "on
channel."
A single -conversion scheme using a high i -f
was chosen for simplicity, although having 120
dB of gain at 11.5 MHz requires careful layout
and good bypassing techniques to assure stability. Fig. 11-6 -
Front view of the complete receiver for
6- or 2 -meter fm. The unit is housed in an LMB
Three Motorola MC1590s are used, although two 10 x 5 X 2-1nch cabinet. Knobs are Kurz-Kasch
can achieve the required gain. Three stages, each S-748-1 L.

RF AMP. MIXER CRYSTAL -F AMP. -F AMP LIMITER

FILTER
I I
CRYSTAL
Q2 U1 U2 U3 DISCRIMINATOR
ANT.
U4
52.2MHz
11.5 MHz 11.5 MHZ
OSCILLATOR
Q3

41.025
MHz
0 AUDIO AMP. AUDIO AUDIO
SWITCH
ZI GAIN

SPKR.
AUDIO
BUFFER

ARROWS INDICATE
NOISE NOISE NOISE
SIGNAL D/RECT/ON SWITCH DET. AMP. AMP.
SQUELCH
GAIN

Fig. 11-7 - Block diagram of the 6-meter fm receiver using a crystal discriminator.
RF AMPLIFIER MIXER
100K 52 MH, RFC 100 11.5 MHz
50/W'
tILY10
"
133K 001
9.2V r7a 40673 6.8
52MHz
62
L2 .001 L3
J1 I '

V
1AV
ANT' L14 A01
C
OSC.
HEP002
Q3
a
VI
4r
MHZ
 J I17 d
A Receiver for 6 -Meter Fm -y...
Fig. 11-9 -
Schematic diagram of the fm receiver. 1
Unless otherwise noted, resistors are 1/2 -watt I1
composition and capacitors are disk ceramic, ex- .
cept those with polarity marked, which are elec-
trolytic.
C1 -C6, incl. - Miniature air variable (Johnson
189-507-5).
C7 -Miniature air variable (Johnson 189-509-5).
Feedthrough type.-
C8 -C10, incl.
FL1 -
For a wide -band 11.5 -MHz i -f, ESEL Fig. 11-8 Top view of the receiver. The etched
ESEL- circuit board containing the front-end components
-
MEL-3A; narrow -band 11.5-MHz -f, i

DL-11A; wide -band 10,7 -MHz -f, KVG is

at the lower left, while the -f amplifier/detector
i
i

XF-107D; narrow-band 10.7 -MHz i -f, KVG circuit board is bolted to the rear wall of the
XF-107B. If 10.7 -MHz i -f is used, add 15-pF
cabinet.
ceramic capacitors across L5, L7, L9, L11 and
L13. (See Fig. 10-12 for a complete list of filter L13 24 turns No. 22 enam. on Amidon T-50-2 -
choices and addresses of the filter manufac- (red code) core.
turers). L14 88-mH surplus telephone loading coil. -
J1, J2 -
Phono type, panel mount. 01, Q2 RCA dual -gate MOSFET. -
L1 -2 turns No. 22 enam. over L2. Q3 Motorola rf JFET. --
L2 - 11 turns No. 22 enam. on Amidon T-50-6 04 Motorola audio JFET.
-
toroid core (yellow code). (Amidon Associates, 05 GE audio bipolar.
12033 Otsego St., North Hollywood, CA Q6-09, incl. Motorola audio bipolar. -
R1 Linear -taper composition control. -
L3
91607.)
-4 turns No. 22 enam. on L4. RFC1 50-1./H miniature choke (Millen 34300). - -
L4 -
6 turns No. 22 enam. on Amidon T-50-6
(yellow code).
RFC2-RFC7, incl.
W. Miller 70F504A1).
500-11H miniature choke (J.

L 5 -
2.96 - 3.15 -µFí variable inductor (J. W. S1
TP1
Spst miniature toggle.
Vector T2.8 terminal.
--
Miller 46A336CPC).
L6 -
0.37 to 0.47-11H variable inductor (J. W. U1 -U3, incl. Motorola IC. -
Miller 46A397CPC). U4 ESEL AL -1A for 11.5 -MHz -f, KVG 107-01 - i

L7, L9, L11

T-50-2 (red code) core.
-
36 turns No. 26 enam. on Amidon
Y1
for 10.7 MHz.
International Crystal type EX. --
L8, L10, L12
L9, and L11, respectively.
-
16 turns No. 26 enam. over L7, Z1 2 -watt audio amplifier, 1 -volt sensitivity
(Amperex PCA-1-14).

I -F AMP I -F AMP LIMITER CRYSTAL DISCRIMINATOR

r t
11.5 MHz TAI
11.5 MHz 11.5 MHr
MHz U3 18 100
U1
(

LIS

T DISC
I 01-T- 500N As TUNE
3300 rh 003 O01
.001
.001 J
SH/ELD¡
RFC3 RFC2
500fJ1 100
500fUi

z,1
2.3850 AUDIO NOISE
ov SE
GATE SQUELCH NOISE AMPLIFIER

.i Z F DETECTOR
MPS -A55
158 Q8 8.9V 5O
8.8V SE 331( 6800
AUDIO
V\^''''--1r
3300 O1
AMPLIFIER ON 04
L8V
1000 I V
Z1 MPS -A55 MPS -ASS
AUDIO
.0033 Q6 }+
GAIN 3.5V IM
2.58 SOK
8-O SPEAKER
10K Í
SQUELCH
GAIN
5600
15v
1800 112v 10

680 Z
J2 si
SOLaF
OM 33
.
12V
I T.001 NOTE: SO= SQUELCH OPEN
SE. SQUELCH ENGAGED
 240 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES

iJZry
,.
''';Y
_=-
o ."t-x
Fig. 11-10 -
The bottom view of the i -f board
shows the shielding and filtering used.

systems can have voltages as high as 16, a series

regulator should be included for mobile opera-
tion.
To set the receiver for the 6 -meter calling
frequency, 52.525 MHz, a 41.025 -MHz crystal is
required. When built for 52 MHz, this unit can be
used for 144- or 220 -MHz reception by adding a
converter having 6 -meter i-f output ahead of the
with resistive loading, exhibited far better stabil- fm receiver. Suitable designs are shown elsewhere
ity than two '1590s running "flat out." in this manual.
The i-f amplifier circuit board uses extensive
shielding. The finished product looks more like a Alignment
piece of vhf gear than an i -f amp., but with high A signal generator and a VTVM are required
gain special precautions are necessary. The front for receiver alignment. The VTVM should be
end and i-f amplifier proved to be easy to build connected to terminal TP1 and set to read 0-3 V
and adjust. The squelch circuit was a bother from dc. The generator should be set to about 52.5
the beginning. The basic circuit used for the MHz, and connected to J1 on the receiver. Ro-
noise -operated squelch is borrowed from the tate R1 fully counter clockwise, and you should
Motorola MOTRAC series. Output from the hear noise in the speaker. If no noise is heard,
crystal discriminator proved to be quite a bit place your finger on the input terminal of Zl. If
lower in level than would be obtained from an the audio module is working, a loud hum will be
equivalent LC discriminator. Also, Z1 (Fig. 11-9) heard. No noise output indicates that the squelch
wants to "look into" an impedance of 100,000 may be locked up. Check the voltages given in
ohms or more. The first version of the squelch Fig. 11-9 against those found on the audio gate.
circuit had far too little gain, so a redesign was With a cw input signal, adjust C4, C5, and C6
required. The final version of the circuit is shown for a maximum reading on the VTVM. Then set
in Fig. 11-9. The voltages noted at the audio Cl and C2 for maximum indicated output volt-
gate, Q8, are critical; any variation from the age. Set the signal to 52.510 and note the VTVM
indicated values will cause a full squelch con- reading. Then set the generator to 52.540 MHz
dition where the audio gate will not open, or, no and adjust C7 for a VTVM reading of the same
squelch operation at all. magnitude, but opposite polarity, from that ob-
A commercial amplifier module was chosen tained on the "low side" of 52.525 MHz. Repeat
for the audio output section. It delivers two the procedure several times to assure a correct
watts (rms) to an 8 -ohm load when using a adjustment.
12 -volt supply. Sufficient audio output is avail- With the basic alignment completed, inject a
able for comfortable copy of weak stations when 52.525 -MHz signal. Then set C3 so that the
operating mobile. An inexpensive imported VTVM reads zero voltage. This completes the rf
0.5 -watt audio amplifier may be substituted if the and i-f alignment. Connect an antenna and ad-
receiver will be used only in a ham shack. Power vance the squelch control, Rl, until the back-
requirements for the receiver are 12 to 13.5 volts ground noise disappears. The squelch should close
at 70-mA drain (squelch engaged) and 200 mA the audio gate at one-third to one-half scale
(at full audio output). Since automobile electrical rotation of Rl.

THE MK -II FM RECEIVER FOR 146 MHz

The simple receiver described here provides

true fm reception. It uses a two-stage limiter and
discriminator, a crystal-controlled converter for
good stability, and a vernier drive for smooth
tuning. Reception with the version shown is
excellent. Selectivity is adequate for separating
three local repeater output signals which fall at
146.79, 146.88, and 146.94 MHz. Sensitivity is
such that a 0.2-pV signal with 5 -kHz deviation is
plainly audible. A 0.7-µV signal provides 20 dB
of quieting. Limiting action is good on all but the
weakest of signals.
\l
Fig. 11-11
fm receiver.
- Front view of the completed 2-meter
The Mk -II Fm Receiver for 146 MHz
241
Fig. 11-12 -
Block SSAN:
diagram of the as- 156.7-147MHI
sembled MK -II fm ANT.
receiver. BT1 should
be a large -size 9 -volt AUDIO
HELICAL LINE
battery, or 7 or 8 RESONATOR
EXT. AP
penlite cells con- GAIN
RS
nected in series.
Cl -
Two -gang min-
LOA

iature variable,
one section 40
pF and the other
having 20 pF.
See text. (J. W.
Miller No. 1640
three -section
unit used here,
two gangs being
connected in
parallel.)

RF 4MP. MIXER
146-147 MHZ 1350-1650 kHz
146-147 MHz
JI C20 C4 01 C8
22 E C

100
ANT, C12

L2 001 0T0 LINK ON

.C13 BC -SET ANTENNA
o
40235
B

OSC. CASE

4145 MHZ Y1
E B C
48.45 MHz 000
33
'CIS MP53563
2N4124
DSG
R12 000
lOK

T MPF102
(80rrOM
viEWS
C18 1000
.001 ''12f

O
+9V -
EXCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE ARE IN MICROPARADS 1 yF I ,

OTHERS ARE IN PICOFARAOS ( Pr OR NYrl.

RESISTANCES ARE IN OHMS;
1000, Y1000 000,

Fig. 11-13 -
Circuit of the 2 -meter converter. Parts L5 -15 turns No. 20 enam. wire, 3/16 -inch ID,
not listed below are designated for pc-board layout closewound.
purposes. Drawing A is discussed in the text. L6 -100-/1H rf choke (Millen J302-100 or J. W.
Miller 70F104A1 suitable). Must be 100 NH in
Fixed -value capacitors are disk ceramic. Resistors
are 1/4- or 1 /2 -watt carbon. value.
L7 - 5 turns No. 16 tinned copper wire, 1/4 -inch
ID, 3/8 -inch long.
L8 - 10 turns No. 26 enam. wire, closewound on
dia slug-tuned form (J. W. Miller
C16 - 10-pF piston trimmer. Ceramic or pc -board 1/4 -inch
4500-2 blank suitable). Variable inductance
air trimmer can be substituted.
J1 -- Not used.
SO -239 -style chassis fitting. 0.35 to 0.6 pH required_
Q1 -Q4, incl. -
Npn bipolar transistor, RCA 40235
L1
L2 - 3 turns No.
1/8-inch long.
20 enam. wire, 3/16 -inch ID, or 40637. Motorola 2N4124 or MPS3563 suit-
able, or any high -beta npn with fT of 250 MHz
L3 -
4 turns No. 20 enam. wire, 3/16 -inch ID, or greater. (Q2 of inset A is Motorola MPF102,
2N5484, or HEP802.)
L4
5/16-inch long.
-
5 turns No. 20 enam. wire, 3/16 -inch ID, Y1 -
48.45 -MHz 3rd -overtone crystal (Inter-
national Crystal Co. type GP in FM -2 holder).
5/16 -inch long.
 242 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
Fig. 11-14 - Inside view of the MK -II. The vacant
area at the bottom has been reserved for a small
transmitter and modulator. The limiter/
discriminator board is mounted vertically by means
of an L bracket. Four metal posts support the
2 -meter converter at the upper left. The bc radio
and Its new tuning capacitor and gain control are
housed in the shield compartment at the upper
center. An aluminum cover is used to enclose the
shield box during operation.

, an mawI chassis ground through L6). The values of resis-

Ir
- tors R4 and R5 will have to be changed as
indicated. R6 will be deleted if the JFET is used.
r No other changes are required. Though it was not
tried, a common-gate JFET rf amplifier might
work nicely at Ql, but with reduced gain over
the bipolar amplifier shown.
Those wishing to use the bipolar arrangement
for all four stages of the converter, but with a
negative ground system (such as mobile), can
simply substitute pnp transistors for Ql through
Q4. The Motorola 2N4126 would be suitable for
this purpose. The positive-ground approach was
adopted here to make the two additional circuit
boards compatible with the bc receiver, which has
a positive -ground hookup. (The foregoing tech-
nique applies only to the use of the converter in
combination with the car radio, and with the
remainder of the circuits described here.
Circuit Features
A block diagram of the MK -II is shown in Fig. Bc-Set Modifications
11-12. Two options are offered FL1 for re- - A 6 -turn link of small -diameter hookup wire
ducing images and cross-modulation from nearby must be wound over the low-impedance end of
commercial services, and MI, a zero -center tuning the built-in ferrite-bar antenna of the be set. This
meter. The jack labeled EXTERNAL CHARGE will connect to the i -f output of the converter by
should be included only if a nickel -cadmium means of twisted wire or shielded conductor. The
battery is used. two-section tuning capacitor should be set at the
Fig. 11-13 illustrates the converter circuit. high end of the be band (1650 kHz or higher),
Inset drawing A shows how the mixer can be then glued in position. Alternatively, it can be
changed to a JFET. The modification was tried in removed and discarded. A two -gang bandspread
the receiver. It reduced cross -modulation effects capacitor of approximately 20-pF maximum capa-
caused by a nearby commercial repeater. Since citance (oscillator section), and 40 pF in capaci-
the MPF102 is an N -channel device it will be tance (maximum) for the mixer section is wired
necessary to employ the hookup shown (gate in parallel with the tuning capacitor in the set. A
returned to the minus bus, and drain returned to three -section J. W. Miller component was used.
Ti 55",, 2.1-F
o
1sT I-F
A r AMP r
50.N1
DE T,
AMP. R1
COLLECTOR CR 1 470
CO
U

AGC
LINE

R2
10
10. Ell>TO
AF AUDIO AMR,
GAIN SECTION

Fig. 11-15 -
Typical circuit of second i -f amplifier and a-m
Points marked A and B indicate where circuit is opened for detector of imported bc-band radios.
addition of the limiter/discriminator.
Components between points A and B are discarded. R2 is replaced by a panel -accessible
(see text). Values for R1, Cl, and C2 are not necessarily the same in gain control
all imported sets.
The Mk -II Fm Receiver for 146 MHz
Each section is 20-pF maximum capacitance, so
two of the gangs are wired in parallel to provide
40 pF for the mixer tuning. A two-section broad-
cast variable can be used by removing the proper
number of plates to secure the desired band -
spread. Other styles of two-gang variables can be
modified in a like manner.
Fig. 11-15 shows the typical configuration
used at the second í-f amplifier and detector of
most imported pocket-size bc sets. After locating
the detector diode, CR1, the modifications shown ó
can be made. Remove all components between
the letters A and B. Audio gain control R2 can
be replaced by a panel -mount type of the same
value (with switch), or the ingenious builder can
retain the original part and mount it on the front
panel of the MK-II. Terminals A and B will
connect to the limiter/detector board of Fig. c =
.
:(...) E I g
11-16. Li- ll
The speaker from the be radio is mounted on
the front panel of the MK-II. In fact, the speaker
grille from the bc set was pried off the plastic
case and used between the speaker and the front s
=s8
i'iI1J--0ñO
panel in true miserly fashion. Some sets do not
have separate grilles, so perforated metal or cir-
cuit board can be used to protect the speaker ó
cone. Z
E75 m
Limiter/Discriminator '- ó
The circuit of Fig. 11-16 shows that two
21 á
ÉLo=ó
bipolar transistors are used in an RC-coupled dL x2
limiter strip, followed by a diode discriminator.
RFC1 and RFC2 are used to prevent vhf para-
o
ó o
ó°:
d H
d LO

sitics. Two Amidon ferrite beads are used for ar

V O C
each choke. Alternatively, a 22 -ohm 1/2 -watt omQó
resistor can be substituted for each choke. It can d E .Ln
LULn U M
be seen that here, again, the npn transistors are co
connected in the circuit for a positive ground mz
system. Type 2N4126 transistors can be substi- Q
tuted if one wishes to use this assembly with (mmUcc
Ew a;
receivers that have a negative ground.
Diodes CR1 and CR2 function in a self-
¢>a,I
U?_ N a+

adjusting limiter arrangement. They can be omit- Mfn Qd 7

ted if the bc set used ahead of the limiter board CC
U d
is low in i-f gain. Matched diodes should be used
at CR3 and CR4. Those having an assortment of óE
= d
germanium diodes can select two that have nearly EJá c
equal dc resistance readings (when checked with
an ohmmeter). Match them in resistance in both a U a) {o

the forward (low dc resistance) and reverse (high r`yo

dc resistance) directions. Ó HOÑ
YO
Ú o_
Terminal El of Fig. 11-16 provides a test Ex ^.D v
voltage (dc) for aligning the discriminator. After
alignment, those wishing to can connect a zero- óLtr
a7
Ñ
y N - .
E
7
center microammeter between El and ground for V row .+

use as a tuning meter. Lafayette Radio sells a t m

E
low-cost miniature meter for this purpose. ºd; crg)
o óL
m `-o

Assembling the MK -II

cvo

03óa v
CnV ? ó Z
The photographs show that the MK-II is not a I

ó_
1° 22
b
r
miniature unit. A large portion of the cabinet has
I

CO U a N
been reserved for a 2 -watt fm transmitter. How-
ever, the builder can certainly shrink the
10-1/2 X 8 X 3 -inch dimensions by a considerable
E
m o
.30
mwU -

margin if he is skilled at packing many parts into iiv T:=-18 U

 244 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
2 -METER HELICAL
RESONATOR
Fig. 11-17 - Circuit of the helical resonator.
Double -sided pc board is used for the box walls
5 C2
and is soldered along each seam by means of a
pencil -type iron. E. F. Johnson silver-plated air -
variable capacitors (No. 189-563) were used in this
model, but high -dielectric piston trimmers can be
INPUT.
substituted. Coils L1 and L2 consist of 4-1/2 turns
of No. 12 copper wire, 3/4 inch long. Each coil is
centered in its compartment and tapped 1/4 turn
(A) from ground to obtain a bilateral 50 -ohm imped-
ance. A 3/8 x 5/8 -inch aperature is cut in the
center divider, and is opposite the second and third
turns of each coil. Cl and C2 are tuned for peak
ENCLOSURE OF DOUBLE -SIDED
CIRCUIT BOARD signal response with the resonator connected be-
tween the receiver input and the feed line. The
2%4 1%a" bottom end of the assembly need not be enclosed.
T Ideally, the enclosure should have no soldered
seams, and both it and the coils should be silver
SOLDER plated. This suggests the use of large -diameter
SEAMS copper tubing for the outer shield if maximum Q is
desired.
lB%

a small space. This enclosure is fashioned from The next step is to connect the limiter/
1/16-inch -thick aluminum stock. The panel is discriminator board to the circuit in place of the
painted dark green. components temporarily connected between
To reduce unwanted pickup of be stations, points A and B. Attach a VTVM or zero -center
the pocket radio is enclosed in a box made from µA meter between El and ground (Fig. 11-16). If
double-sided pc board. A press -fit U-shaped alum- a VTVM is used, set it for the 1.5 -volt scale and
inum lid encloses the top of the compartment. adjust the meter -set control so that the needle is
The bandspread tuning capacitor and gain control exactly on zero. Now, apply a strong signal
are contained in the same box. No. 6 spade bolts (several hundred µV) to the mixer input, using
hold the assembly to the main chassis. Templates 1500 kHz as a test frequency. Vary the signal
for the circuit boards appear in QST for August 1971, generator plus and minus 15 kHz while observing
page 16. the meter. Thé signal should swing plus and
Tune -Up and Use minus in a linear fashion. (If a VTVM is used for
the test, it will be necessary to switch the
It is suggested that the builder test the be set polarity back and forth with the meter reversing
independently after it is modified and mounted switch.) The meter should return to zero when
in its shield box. The components which were no signal is present. If the foregoing conditions
removed between points A and B of Fig. 11-15 are not met it will be necessary to adjust T1 of
can be clipped between those points for the test. Fig. 11-16 for a linear response. The alignment of
Connect a 9 -volt battery and tune in a be station. Ti will be a tedious task, so plan to spend some
Make certain that the new volume control and time in the adjustment process. Alternately tweak
tuning capacitor are functioning as intended. Ad- the pink and blue cores of the transformer, a few
just the trimmers on the tuning capacitor to degrees of rotation each time, then sweep across
assure tracking of the mixer and oscillator sec- the 30 -kHz range and observe the meter response.
tions of the radio. Next, connect a signal gener- Try various settings of the cores until proper
ator to the base of the mixer through a 10-pF alignment is achieved. If a signal generator is not
blocking capacitor. Set the generator for 455 -kHz available, tune across a strong be station and
output and align the i -f transformers for peak adjust Ti for linear response of the discriminator.
response at that frequency. After Ti is correctly adjusted, there should be no
a-m detection of the be signal when the station is
tuned in to its center frequency. Readable audio
will be heard, however, when tuning to either
side of center frequency.
Install the be set and the limiter/discriminator
- assembly in the chassis. Next, connect the 2 -
meter converter to the circuit. Tune in a weak
2 -meter fm signal, or use the output from a vhf
_1r
signal generator. Adjust L8 and C16 of Fig. 11-13

43
Fig. 11-18 - View of the homemade helical
resonator shown in Fig. 11-17. Double -sided pc
board is used for all walls but the top one in this
version. Flashing copper was used for the top
surface in this model, but pc board would have
been suitable.
A Tunable 440 -MHz Fm Receiver 245
for maximum signal response at 146.8 MHz. band may appear as images. The helical-resonator
Spread or compress the turns of L2, L3, and L4 filter shown in Fig. 11-17 can be built and used
for peak resporse. Use an insulated rod while ahead of the receiver to reduce or eliminate
making these tests. This completes the adjust- unwanted responses. FL1 is easy to assemble and
ments. The converter can now be bolted in place is very inexpensive, yet is superior in performance
in the cabinet. to most strip -line filters. Design data for these
Since the MK-II represents one of the least filters are given in ¡77 Reference Data for Radio
sophisticated approaches to fm reception, some Engineers, 5th Edition, Chapter 22. The filter
ills may become manifest. Strong signals from consists of two very high -Q tuned circuits, coup-
nearby two-way commercial services may show led through a small aperture in the metal wall
up in the tuning range of the receiver. Also, between them. Input and output taps are set for
signals from the lower portion of the 2 -meter a 50 -ohm impedance.

A TUNABLE 440 -MHz FM RECEIVER

Though originally intended for checking signal

paths between proposed uhf fm sites, this receiver
also provides convenient eavesdropping on local
440-MHz fm activity, and monitoring of com- v
mercial frequencies adjacent to the high end of
our 420-MHz band. It or something quite similar
to it is readily put together, and can be assem-
bled quite inexpensively. Though a duplicate can
be built from the information supplied here, the
reader should treat this discourse mainly as a
collection of ideas. The exact units used for the
receiver may not always be available, so other
components may have to be adapted to the job.
In most modern TV receivers the uhf tuner is
solid-state, designed to feed the 43-MHz TV i-f
strip directly. The mixer output is untuned, so it
is apparent that any intermediate frequency can Fig. 11-19 - Panel view of the 440 -MHz fm
be accommodated, depending solely on the fre- receiver. The vernier dial drives the tuning shaft of
quency range of the tuner oscillator. One advan- a converted uhf TV front end. At the rear is a
tage of the 43 -MHz i -f is the great number of 450 -MHz Motorola fm receiver, the i -f system of
TV tuner.
low -band fm monitor receivers for this frequency which takes the output of the converted
The meter can be connected in the limiter or
range available at low cost. discriminator circuit.
It is best to avoid the narrow -band monitors
that won't accept more than 5 -kHz deviation.
Tuning will be too sharp for the stability and plates in the oscillator compartment. The mixer
tuning rate of uhf converters, and most amateur and antenna circuits will not require tuning across
fm in the 420 -MHz band is still wide -band devia- our small intended frequency range, so the rotor
tion. The wide -band receiver will not be bothered plates in the two upper sections of the tuner can
seriously by any drift in the TV tuner. all be removed.
The uhf TV tuner used here is a Sickles Model Mount three 16-pF glass trimmers on the
228. Many tuners use the same basic electrical tuner walls, parallel to the three lines and as close
and mechanical layout, the only differences being as. possible to them. See Fig. 11-20. Connect the
in the dial -drive arrangement made for a partic- trimmers to their respective lines with short
ular TV receiver. These tuners do not have trim- pieces of heavy wire or copper strip. By bending
mer capacitors for alignment, such as were used the one oscillator rotor plate away from its stator
in earlier uhf converters. Alignment is done with carefully, the tuning range can be reduced to as
specialized test equipment, by precisely bending
little as 15 MHz.
the rotor plates in the tuner. Most tuners are
one channel, over Remove the 300-ohm connector and coupling
linear within plus or minus TV
from the antenna section by drilling out the
loop
the entire uhf TV spectrum, when they leave the mounting rivets in its insulating support. Enlarge
factory.
the hole, and mount a phono jack or a BNC
Uhf Tuner Modifications fitting for antenna connection. Run a 1/8 -inch
copper strip from a point on the antenna line,
Getting this type of tuner to cover the upper 1/2 inch from the wall, down to as close to the
10 MHz of the amateur band is best done by coupling port to the mixer compartment as pos-
adding trimmer capacitors across the tuner cir- sible, then up to the connector, as shown in Fig.
cuits. The sketch of the tuner as modified, Fig. 11-20. If the port is not merely an open hole in
11-20, should help to make clear how this is the wall separating the sections, but is a Faraday
done. Start by removing one of the two rotor shield, as in the 228 tuner, remove the Faraday
246 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES

Fig. 11-20 -Bot-

tom view of the uhf
TV converter, as
modified for use in
the 420 -MHz band.
_, STATOR PLATES Only the oscillator
RAN K POST circuit is tuned by
_IS ANT LINE the vernier dial, all
P rotor plates in the
P mixer and preselec-
ANT
tor circuits having

'
ía1
been removed. One
rotor plate is left in
P -
/
STATOR PLATES
CERAMIC POST FARADAY
the
tion.
oscillator sec-
Positions of
MIXER LINE SHIELD
P
in O plates to be re-
P :I:~ MIXER
DIODE
.--1/2
I

OUT
-F
moved are indicated
by the letter P.
_g GLA55 IRIMMER E
STATOR PLATE a,
CERAMIC POST

. Q
/ / - OSC TRANSISTOR AND
BIAS RESISTORS NOTSHOWN
P '- OSC. INDUCTOR
LINE BYPASS CAP

GLASS TRIMMER

the receiver to be used does not provide a dc

path, a 50 -MHz rf choke should be connected
shield to improve interstage coupling. This shield from the tuner i -f output to ground to provide a
is not visible in the tuner photo. If used, it is a dc path for the mixer crystal current. The ground
comb -like insert in the port, as shown in enlarged end of the choke can be lifted to measure crystal
form in Fig. 11-20. current, which should be at least 100 µA.
One lead of the mixer diode in the 228 tuner In the receiver shown here the tuner feeds the
runs through a small port between the oscillator 75 -MHz i -f of a Motorola 450 -MHz crystal-
and mixer sections. The other diode lead, barely controlled fm receiver. The two 6AK5 frequency
visible in the tuner photo, is parallel to the mixer multipliers which provide injection to the first
line and is connected to the right-hand wall of mixer, the front-end rf tubes and afc tube were
the tuner. Cut it at this point and solder the lead removed to conserve power, since they are not
to the mixer line, about 1/2 inch from the wall, used in this application. A National vernier dial
as shown in Fig. 11-20. In some of the better drives the 2-plate tuning capacitor in the tuner,
tuners the mixer diode is reverse biased, and does through a shaft extension and a flexible coupling.
not go directly to ground in the mixer section. If there is backlash in the drive system it is not
This was done for improved noise figure and evident in the operation of the receiver. Signals
should be left the way it is. tune in smoothly and easily.
Most tuners are designed to terminate in an i -f A meter indicates limiter current, for signal
input circuit which is at dc ground potential. If strength observation, and as an aid in adjusting
antenna systems and beam headings in fm work.
The nature of fm reception makes small changes

º in signal level undetectable by ear. The discrimi-

nator is also metered, for ease of station tuning.
Supply voltage for the tuner is dropped from the
- receiver B -plus line, and is regulated with a Zener
diode.
Visible in the upper left portions of the
photographs of the complete receiver is a
grounded -base uhf preamplifier using a TIXM101
transistor. A preamp suitable for this purpose is
shown later in this chapter.
The tuning range of this receiver is roughly

I +I1
- I
.
440 to 454 MHz. No images or other spurious
responses are heard. Tuning is smooth, and sta-

Fig. 11-21 - Looking into the bottom of the

tunable 440-MHz fm receiver shows the converted
TV tuner below the vernier dial. A transistorized rf
preamplifier, also built into a TV tuner case, is
visible just to the left of the tuning shaft.
Transmitting Fm 247
tions are easily "zeroed" on the discriminator MHz. The TV tuner alone is very poor; on the
meter. order of 10 microvolts for 10 -dB quieting. If one
The sensitivity is adequate for monitoring on wishes to be able to hear the weak ones, a much
line -of-sight paths, but not for weak -signal DX better mixer is needed, and it should be preceded
work such as is commonly done around 432 by a good low -noise rf amplifier.

TRANSMITTING FM
AN FM TRANSMITTING ADAPTER
The circuit given in Fig. 11-23 is a trans-
mitting adapter which will provide frequency
modulation for older 6- and 2 -meter exciters,
such as the Clegg series. Audio input from a
microphone is amplified and clipped to increase
speech effectiveness and to provide constant de-
viation. The output of the speech clipper is tr
passed through a 6 -dB -per -octave filter to shape
the audio response so that the phase modulator
will produce fm directly. Crystals cut for 8 MHz '1 ,1
are employed. For transmitters such as the Clegg
22er that require a high drive level, the alter-
native output circuit (Fig. 11-23B) should be
used. Fig. 11-22 -
The fm transmitting adapter. The unit
may be constructed on an etched board, as shown,
or built on a metal chassis using point-to-point
OSC. PHASE wiring.
RFC MODULATOR
8 MHz
+200V 100 loo. OUTPUT

390

47 Fig. 11-23 -
Diagram of
the fm adapter. Resistors
are 1/2 -watt composition

30
5.M,j
- (A)
and capacitors are disk
ceramic, except for those
with polarity marked,
which are electrolytic. L1

(
11000 is a Miller 42A686CBI
220 slug -tuned coil. For exci-
AUDIO AMP. CLIPPER AUDIO RFC ters which need only a
2.5mH volt or so of drive use the
'100K SHAPING
1
-1-4 270K 12A%7Á
.01
circuit at A, while for
I2AU7A 82K high-level drive use the
V2A 1 h128 V38 alternate oscillator/
ryRyFCy .001 82K 8200
z
t
modulator section shown
250K at B.
MIC.

220* .

:,,/,,,+ ` 02 2200
oEVIATrON rh.
MIC.GAIN 1200 LEVEL
10M
.097 3300
V

CRYSTAL OSC. PHASE MODULATOR

3300
+200V0
470 tw 8MHz
1r
.01-r /
100 .. 100
EXCEPT AS INDICATED, DECIMAL
.h 346B1Y11
OUTPUT
VALUES OF CAPACITANCE ARE .001 27K
IN MICROFARADS (yF I;
OTHERS
ARE IN PICOFARADS OR , l,F Fl:
RESISTANCES ARE IN OHMS; 8MHz
FRCQ OI 4'47
AOJ.
t 1000. U. 1000000
S.M..SILVER MICA 5M.
tüvL--
T100K 75 +
S.M. 39K
RFC
I mN
(B)
AUDIO
 248 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES

6V6
TO TIE
STRIP
Fig. 11-24 -
Gonset fm conversion.
miniature dpdt toggle switch.
S1 is a

(0/d R3 connection)

changes are needed in the receiver, as the wide

bandwidth of the Gonset i -f allows slope detec-
tion of strong fm signals, such as from a local
TO 2E26
PLATE repeater.
The modifications are shown in Fig. 11-24.
The adapter may be assembled on either a pc Drill a 1/4 -inch diameter hole in the rear deck of
board or metal chassis. The crystal -frequency the transmitter chassis, to the right of the micro-
trimmer and deviation control should be set in phone jack, 1/2 inch above J4. Install the mini-
accordance with the instructions outlined in ature switch in the hole. The red wire from pin 3
Chapter 10. The microphone gain control is ad- of the 6V6 socket should be disconnected and
justed to provide 10 to 15 dB of speech clipping. run to the switch, as shown in Fig. 11-24. Then,
remove the end of resistor R3 which connects to
Ll, the output circuit, should be set for maxi- the terminal strip and connect it to S 1. Add a
mum drive to the associated transmitter.
.01-µF capacitor across R3. Set the switch for fm
operation, and adjust the MIC GAIN control for
USING A GONSET II FOR FM COM- approximately 1/4 rotation, which will produce
MUNICATIONS approximately 5 -kHz deviation. A higher setting
will be needed for 15 -kHz deviation. The Gonset
A simple conversion of the Gonset Communi- has no speech clipping in the modulator, so the
cator II (or similar a -m rigs) allows fm operation amount of deviation produced will be directly
with a few wiring changes that should take less proportional to audio input. To improve the
than an hour. To produce narrow-band deviation, performance of the unit for both a-m and fm
the oscillator is screen modulated with audio operation, an outboard speech clipper such as
from the a -m modulator section. A toggle switch described in The Radio Amateur's Handbook can
allows an instant return to a -m operation. No be added.

AN FM TRANSMITTER FOR 2 METERS

possible a variety of amateur applications, and

á;
Fj. ,EN
,!. ; , ;-r
4 j..,art.
noteworthy among them, emergency/portable
operation.

Rf Circuit
1g; .
Four low-cost bipolar transistors are used in
1. the circuit of Fig. 11-26. Ql is the oscillator,
' . , i + KI
which uses 18 -MHz fundamental crystals ground
for a load capacitance of 20 pF. Output from the
first stage is taken at 73 MHz, a frequency
multiplication of 4. The second stage, Q2,
doubles the frequency to 146 MHz. The re-
maining stages operate as amplifiers at 146 MHz.
Frequency modulation is effected by applying
Fig. 11-25 -
Top view of the rf module showing its audio to a voltage-variable diode (Varicap) CR1. As
relative size. The crystal sockets and Vector push - the amplitude of the audio varies, the junction
in terminals for connection to the crystal switch
are at the upper left. The coils near the crystal capacitance of CR1 changes, and this change pulls
sockets are for adjusting the crystals to frequency. the crystal frequency above and below its preset
The oscillator is at the lower left, and the PA stage frequency to provide fm. The amount of devia-
is at the far right. The small loop of wire at the tion, or swing, is determined by the audio level
lower center is a B-plusjumper. impressed across CR1. Normally, this will be set
for 5- or 15 -kHz deviation, depending upon the
Whether you're a would-be fm-er, or a person bandwidth in vogue for a given area. Approxi-
who has already explored the world of fm and mately 1.5 volts of reverse bias is developed
repeaters, this little 2 -watt solid-state transmitter within the circuit and appears across CR1. This
can be the key to new operating enjoyment. No eliminates the need to provide back bias from the
need to scoff at the QRP aspect of this project, 12 -volt line.
because here we have a piece of gear that can be Crystals Y1 through Y4 are adjusted to the
operated from the 12 -volt automotive system, a desired frequency by means of trimmer capaci-
dry -battery pack (10 size -D flashlight batteries in tors. Approximately 3 kHz of shift is possible
series, or a 12 -volt lantern battery), or a simple with the value given. Regulated voltage is sup-
ac-operated 12 -volt dc supply. This feature makes plied to Ql (and to the bias line of Q2) by
 DRIVER PA
04 146MHa
DOUBLER 03 274

02 40637
146 Ha L12
7
4037 C23
OSC. 6O C17
'73"' 25

4g37
1] ...L C5201
50
11
rn330 í1t7 S M
CB SM

m C16
RFC 2 116 390
10p I4 390
4700 _C12 C22
390 390
R7
R4 560 C21
MC3 .05
470
4.72H 1- ( RFC 7
390
RFCS
1.5yH I I
R5 .,UMPE R 12 -13.5V DC

RFC 6
SOpH
L20 o-
390 EXCEPT AS INDICATED, DECIMAL
93V C14
R1 1W 02 VALUES OF CAPACITANCE ARE /-77
226 01-04INCI.
BOTTOM VIEW IN MICROFARADS 1911 OTHERS
¡

r-n ARE IN PICOFARADS IPr OR y9F1;

SMSILVER MICA
CR1 C2 RESISTANCES ARE IN OHMS,
AF NPO. ZERO -TEMP COEFF.
120 .1000.
INPUT

Fig. 11-26
-
Schematic diagram of the rf sec tion of the transmitter. Most of the parts are available
from Allied Industrial Electronics (Catalog No. 710), Chicago, I L. Numbered components not given in
the parts list are so labeled for circuit -board identification. Fixed -value capacitors are disk ceramic
unless noted differently. Fixed -value resistors are 1/2 -watt carbon. L1 to L4 are not used in this
C5, C9, C17
- 1.7- to 14.1-pF pc -type air variable version.
-
(Johnson 189-505-5). Compression trimmers or RFC7, RFC8 104.1H low -Q rf choke. 4 turns No.
ceramic padders can be substituted. L11 - 12 turns No. 22 enam., closewound, 30 enam. wound on Amidon miniature ferrite
- Phenolic -base ceramic padder (Erie bead (Amidon Assoc., 12033 Otsego St., N.
C23, C24 3/16-in. dia (0.4351.1H).
- Hollywood, CA 91607).
type 557 or equiv.). Cut off phenolic flange. L12 - 5 turns No. 16 bus wire, 5/16 -in. ID x 1/2 -
- Miniature ceramic trimmer. in. long (0.14 µH). RFC9 4 Amidon miniature ferrite beads on a
C11, C25 -C28, incl.
CR1 - Varicap diode, 82-pF nominal capacitance, Q1 -Q4, incl. - RCA bipolar transistor. Substitutes short length of hookup wire.
73.8- to 90.2-pF total range (Motorola MV839 should have equal or higher voltage, wattage, S1 -
2 -pole, 4 -position nonshorting phenolic or
pr nearest equiv.). and fT ratings. ceramic rotary switch.
CR2 - Zener regulator diode (Motorola HEP-104 R9 -1000 -ohm linear -taper pc -board carbon Y1 -Y4, incl. -
18 -MHz fundamental crystal cut for
or equivalent). control (I RC type X-201, R102B or equiv.). 20-pF load capacitance. International Crystal
L5 -5 turns No. 16 bus wire, 5/16 -in. ID x 5/8 in. RFC1 - 1-mH pc -board rf choke (James Millen Co., .0025 percent commercial standard, FI -4
J302-1000). holder. High -accuracy .0025 percent type pre-
long. Tap 1-1/2 turns from C5 end (0.11 µH).
L6, L10
- 3 turns No. 16 bus wire, 5/16 -in. RFC2 - 10-4H molded rf choke (J. W. Miller ferred for best stability. FI -4 pc -board sockets
ID X 1/2 in. long (.075 µH). 9310-36). See text. used ín this transmitter. Crystals from other
L7, L8
- 6 turns No. 22 enam., closewound, RFC3, RFC4, RFC6 -
4.7-1.1H molded rf choke (J. manufacturers may work satisfactorily if
3/16-in. dia. See text. W. Miller 9310-28). See text. ground for 20-pF load. (International Crystal
L9 -4 turns No. 22 enam., 3/16 -in. dia, spaced to RFCS -
1.5-µH molded rf choke (J. W. Miller Mfg., Inc., 10 North Lee, Oklahoma City, OK
occupy 3/8 in. on form (.06 µH). 9310-16). See text. 73102).
 250 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
PREAMP. CLIPPER AMP

C34V 06
05 C323-5- LO -2
C29 CR4 CR5 R19
3V 4 .05V
C- -V^V 2700 AP'

C28
=j- too
I

R12 T1 HI -2 1PF
56K 3715V
O
500 -OHM
WC. .18
(TTV.I
/
C31
10p1
158
// R2
66
AF
0-20 V

V\Ar
^R13

560K
t

EXCEPT AS INDICATED, DECIMAL 12-13,5V DC

VALUES OF CAPACITANCE ARE
C39 C38-100MF
15V
IN MICROFARADS I yF , OTHERS
1

ARE IN PICOFARADSIPF OR VPF);

VOLTAGES LISTED ARE PK.-PK. Au00
RESISTANCES ARE IN OHMS; NC. NO CONNECTION
t 1000.

Fig. 11-27 -
Schematic diagram of the clipper/
modulator. Numbered components not listed be-
Q5, Q6
40231
- Audio -type npn bipolar transistor. RCA
or equiv. Also, Motorola 2N4123 or
low are for circuit -board identification. Capacitors MPS -A10 suitable.
are disk ceramic or paper. Polarized capacitors are -
R11 See text.
electrolytic. Fixed-value resistors are 1/2 -watt
carbon.
- 50,000 -ohm linear -taper carbon control.
R23
Pc -board type IRC X-201, R503B.
T1 - Miniature audio transformer, 10,000-ohm
primary, 2000-ohm secondary, ct not used.
CR4, CR5 - Silicon diode. 1N914 or top -hat Radio Shack/Archer 273-1378 or equiv. Con-
rectifier suitable. nect low -Z winding to 06 collector.

means of Zener diode CR2. This measure helps Some of the inductors are wound on 3/16 -inch
insure against oscillator instability. diameter phenolic rod. Any low-loss rod material
A drive control, R9, is connected in the (Plexiglas or polystyrene) can be used, or the
emitter lead of Q2 to permit the operator to coils can be wound on a 3/16 -inch form, slipped
reduce power to the minimum amount needed. off and used as air -wound inductors. If this is
This measure helps to prolong the life of dry done, put a drop of coil dope on each inductor
batteries during portable operation. to hold the turns in place. The large air-wound
coils can be wound over a drill bit of appropriate
The Modulator size. Ferrite beads are used as cores for chokes
RFC7 and 8. Each choke has four turns of
Only a few peak-to -peak volts of audio are No. 30 enameled wire looped through its ferrite
needed to provide fm. A two -stage audio channel bead. The chokes, after being installed on the
is shown in Fig. 11-27. This circuit amplifies the circuit board, are glued in place with china
microphone output to a suitable level for clipping cement.
at diodes CR3 and CR4. A small amount of Glass -epoxy circuit board is recommended for
forward bias is used on the diodes to permit the rf module. Low-cost phenolic board is suit-
clipping action at relatively low audio level. The able for the modulator. Poor-quality circuit board
10,000 -ohm resistor and .05-/.ZF capacitor used can cause problems in the rf module.
after the clipper diodes serve as a filter to reduce Heat sinks should be used on transistors Q3
the harmonics caused during clipping. Output and Q4. Small clip -on types are suitable. Wake-
stage Q6 amplifies the clipped audio to a maxi- field Engineering Co. makes a wide variety of
mum level of 20 volts peak to peak. The devia- these devices. They are listed in most mail-order
tion control, R23, is adjusted to provide the catalogs. Alternatively, homemade sinks can be
amount of frequency swing needed. A value of made from 16-guage aluminum or brass stock.
approximately 3 volts pk-pk- is typical for 5 -kHz The transmitter can be housed in any metal
deviation with the circuit of Fig. 11-26. box that suits the builder's fancy. The transmitter
board measures 3 X '7-1/2 inches. The modulator
Construction Data dimensions are 1-3/4 X 4-1/2 inches.
If you have built a commercial kit, you can
assemble this transmitter easily. Circuit -board Tune -Up
templates for the transmitter and modulator are Connect the modulator output to the audio
available from ARRL for 50 cents and a large input terminals on the transmitter board. Use
self-addressed, stamped envelope. A parts - shielded audio cable or small -diameter coax. At-
placement overlay is included in the package. tach a No. 47 pilot lamp across the transmitter
Several phanges included in Fig. 11-26 have not output jack. This will serve as a visual -indicator
been made on the template. Li to L4 have been dummy .,load of approximately 50-ohms imped-
replaced by C25 to C28, and C27 has been ance. Adjust the drive control to nearly full
eliminated. resistance (low powel. Plug in a crystal and
An Fm Transmitter for 2 Meters 251
DRIVER FINAL OUTPUT FILTER
40973 40975 L5

.--n1-?-1"
L1 Q1 L6
Q2 L8 L9

2`T2
INPUT L7
CI C5 ANT.
25 ? YYl
10 L4 --= 22T
cb
S.M.
401
RFC2
.001 C9
001
C NARROW
LEAD

2N5995 +12V
g 2N5996 RFC3 RFC4
TOP VIEW

Fig. 11-28 -
Schematic diagram of the 25 -watt
solid-state amplifier. Capacitors are disk ceramic
L6 -
dia.
2 turns No. 20 enam., closewound, 1/4 -inch

unless otherwise noted. L7 -L9, incl. -

3 turns No. 20 enam., closewound,
C1, C2, C4, C6 -
Compression trimmer. 1/4 -inch dia.
-
C3, C5 -
Air variable. Q1, Q2 RCA power semiconductor, used with
C7, C8 -
Feedthrough type. appropriate heat sink.
-
L1 -
1/4 x 2/3-inch copper strap. RFC1, RFC2 4 turns No. 30 enam. wound on a
L2, L4 -
9 turns No. 20 enam., closewound, single ferrite bead (turns looped through the
bead).
3/16-tn. dia.
L3, L5 -
3 turns No. 14 tinned bus wire, 3/8 -inch
RFC3, RFC4 -
Four ferrite beads on a 1/2 -inch
ID X 1/2 -inch long. length of No. 20 wire.

apply +12 volts to the B-plus terminal of the second harmonic. The wavemeter response at 293
transmitter module (negative lead to ground foil). MHz should be very low, but some energy will be
Couple a wavemeter to L5 and adjust C5 for a present.
maximum reading at 73 MHz. Modulator Checkout
The next step is to set the wavemeter for 146
MHz and couple it to tank coil L6. Tune C9 for The circuit of Fig. 11-27 is designed for a
maximum output indication. The same technique low -impedance dynamic microphone (500 to
is used to adjust the tuned circuits of Q3 and Q4. 1000 ohms). If a high -impedance microphone is
Now, advance the drive control to obtain maxi- to be used, replace R11 with a 100,000 -ohm
mum power. The dummy load should light at this unit. This will reduce the audio drive to Q5, thus
point. Retune each stage for maximum lamp preventing saturation of that stage. Also, the
brilliance. Alternately adjust the tune and load high -value resistor will give the high -impedance
trimmers, C23 and C24, for maximum glow of microphone a more suitable impedance to look
the lamp. Normal operation should cause the into. Adjust R23 for the amount of deviation
lamp to light to full brightness or slightly more. required. This can be done best by checking with
At 13.5 volts one should be able to obtain another amateur who has an fm receiver of the
above -normal lamp brilliance. Couple a wavemeter desired bandwidth. The crystal can be rubbered
to the output tank and check for second - to the desired frequency by adjusting its series
harmonic energy. Choose a setting for C23 and capacitor.
C24 that provides maximum rf output at 146 A final word of caution: Never operate this
MHz with the lowest possible reading at the transmitter into a highly reactive load. The SWR

^ r

s A 1/
1

Fig. 11-29 -
Top view of
- e-
the solid-state 1-W trans-
mitter for 220 MHz.
252 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES

zr
°
p
3 m
ó
có
ó
óy
O
d Ó p
C V -0C Ya
7
'' .1-
7 O '

_7 2p
d Ñ
r,
2U _v
O i °%

cLi w: vI,

L CO C

I
>o.,

o
C`)
C)
x. 2._ ,EN>:::
Ú0

N a> a> 1 tr 0
m 3
IgIVÚÑ-ÚÚ1
O
TI
,J CC
LL 1.1.
Cr Cr
LLLL.-
Cr Cr ).-
ro
0.
CD

2
,
CO
p p
W

To
7 -oO>d
C
C
> 7
O
.- Q c
3 Jm 3
\+
É a;
N. E

-
d O`

o93°' EN?.3,- ¡1
n H1 OO >O
z au,
`
-o
wÑ C
a, a, m ce m
al
lis
c a E'in c'-
NycJ..n .cJ
CO
E1
Ó C
'_^
C

O
z iÑ
w
CC
MJ Zp C Ñ
tLj
z
m

O d
C- 7(:0 O,i,L
E
wM
m
in M 7
w Ó
'.
P..

o
C1
jó «8
ál d « -- N7 C
I

m
co -v

. 3
a, J Ó J L co c coi I aC,
I

p N
J J J J
:':9,1- CID d(y
E m
L Ó L-00in -o
d

¡IO
Cn w C ÚQ15 +
I
Ñ VC CO

O
MÚ
2 m

.
'-E
LL
E'm
Ú
Ñ
E
o

Ea
C N
° E d
111

U)
E
7
U
E

"O
a,

EM
Ñ 1E
a

a
in
00

m Y
Z CE
o

u, Ú
V 1

1
._
Ñ
d
$ pmE
á' a. iC a Z
Cb

N
I
1Ó C
oun
L
v-.E.-`
6, .2
E
. 7
.

UNal 3ºIJ
ÚUJ J
Converting Surplus 253
should be no higher than 2:1 to prevent damage L5 -8 turns, tap 1-1/2 turns from C5 end.
to PA stage Q4. The output circuit of Q4 is L6 - 4 turns, 1/2 -in. long.
L7 - 9 turns, closewound.
designed to work into a 50- to 75 -ohm load. L8 - 5 turns, closewound.
Suitable substitutes for Ql, Q2, and Q3 are L9 -
3 turns, 3/8-in. long.
2N4427 and 2N3866. L10 - 2 turns, 1/4 -in. long.
2-METER SOLID-STATE AMPLIFIER
L11 -10 turns, closewound.
L12 - 5 turns, 5/8 -in. long.
Also, change C19 to a 1.7- to 14.1-pF air -variable
The circuit of Fig. 11-28 was developed for
use with the new RCA 2N5995 and 2N5996 npn
capacitor (Johnson 189-505-5). Tune-up pro-
cedure for the 220 -MHz model is the same as
overlay rf power transistors. Low-inductance
described earlier for the 146 -MHz version.
radial -type leads are used for the base, emitter,
and collector connections. The mounting stud is A SOLID-STATE 50 -MHz
independent from the transistor elements. This
ideal feature makes it possible to bolt the tran- FM TRANSMITTER
sistor to the wall of the equipment case, thus
using the cabinet or chassis as a heat sink. The four-transistor unit shown in Fig. 11-30
Both transistors are "mismatch tested" by the can be used as an exciter, or as a 1 -watt trans-
manufacturer. This means that an infinite load mitter. It operates from 12 volts dc, and can be
mismatch from short to open can be tolerated by modulated by the audio clipper/filter shown in
the transistors during driven periods, provided the Fig. 11-27.
maximum junction temperature is not exceeded. The circuit is built on a small Minibox. Trans-
The circuit shown here was tested for both short istor sockets are used to provide for point-to-
and open conditions, and no damage to the point wiring. Layout should follow the same
transistor resulted. pattern used with vhf vacuum-tube transmitters -
Driving power for the 2N5995 is 3/4 watt. short, direct connections.
The 2N5996 requires approximately 5 watts drive Checkout should be done one stage at a time,
to develop its rated power output. The 2N5995 starting with the oscillator. The remainder of the
can be used as a driver for the 2N5996 to obtain transistors should be removed from their sockets
approximately 20 watts output at 146 MHz. and a No. 49 lamp connected across L2. A faint
Similar circuit constants can be used for both glow should show when Ll is peaked at 25 MHz.
stages. Check each stage in the same manner, terminating
the output link of each stage with the dummy
CONVERSION OF THE 146 -MHz lamp. A No. 47 lamp will be needed at L8
because of the higher power at the transmitter
FM TRANSMITTER TO 220 MHz
output. Full brilliance should be noted on the
The transmitter shown in Fig. 11-26 was No. 47 lamp if all stages are working properly.
designed for 2 -meter operation. However, by Finally, tune each stage for maximum transmitter
making a few changes in the circuit, it can be output. Power output from this transmitter is
made to operate at 220 MHz. The oscillator, Ql, approximately 1 watt.
is changed from a quadrupler to a tripler. Using Once the transmitter has been adjusted for
an 18 -MHz fundamental -type crystal, the output maximum output, the modulator should be con-
network of Q1 is tuned to 55 MHz. The next nected. The check-out procedure for the modula-
stage, Q2, works as a doubler to 110 MHz, which tor section is contained in the description of the
in turn drives Q3. Q3 also works as a doubler, 2 -watt, 2 -meter transmitter which appears earlier
producing output at 220 MHz. There is enough in this chapter. Once the modulator is function-
drive from Q3 to provide about 1 watt output ing properly, adjust R23 of Fig. 11-27 for the
from Q4, the final stage. A 2N3866 should be desired deviation. The Bessel-function method or
used for Q3. The coil specifications given in Fig. a deviation meter (both described in Chapter 10)
11-26 are the same for 220 -MHz operation, ex- can be used to set the amount of frequency
cept as follows: swing.

CONVERTING SURPLUS
IMPROVING FM RECEIVER A simple preamplifier, such as shown in Fig.
PERFORMANCE 11-32 for 146 MHz and in Fig. 11-34 for 440
MHz, may be added to a receiver to increase its
Many older fm receivers, and some new sensitivity and to improve limiting (as the overall
models, do not have sufficient sensitivity or limit- gain before the limiter will be Increased by 10-15
ing capability. Also, the transceivers designed for dB). The 2 -meter version uses a dual -gate MOS-
the mobile telephone service do not have a FET while the 440-MHz unit employs two JFETs
squelch or audio power-amplifier circuit. Suitable in a grounded -gate circuit. Both amplifiers are
accessory units can be constructed easily to im- adjusted by peaking all tuned circuits for maxi-
prove the performance of a rig deficient in any of mum limiter current while receiving a weak sig-
these areas. nal.
 254 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES

ally cut off. The second half is used as a switch

in the audio path, turning off the audio output
when the noise from the discriminator exceeds a
preset level. The 7061 tube serves as an audio
output stage. Insufficient bias on the 7061 will
C^ result in high tube current and low, distorted
audio output. It should have a negative grid bias
1
of approximately 10 volts. A template and parts -
'=""
layout diagram for the circuit are available from
o.o the ARRL Technical Department for 50 cents
and a business-size self-addressed stamped envel-
ope.
ADDING ADDITIONAL CHANNELS
Many fm units currently available from sur-
plus sources are single -channel models. Others
Fig. 11-31 -
The 2 -meter preamp. may be moun- provide 2 -channel operation by switching between
ted in a small Minibox or connected directly inside
an fm receiver.

A receiver will have a poor limiting character-

istic if the gain before the limiter circuit is
Insufficient, or if the limiter itself is of poor
design. The circuit of Fig. 11-35 can be added to
a receiver to replace an existing limiter stage. The
new limiter uses an RCA CA3012 integrated
circuit. Care must be used in the installation and
layout of this high-gain IC to assure stability. The
CA3012 will provide a "hard" limiting character-
istic with about 100 mV of signal input.
If a receiver does not have a squelch or audio
output circuit, the unit shown in Fig. 11-36 may
be added. The circuit function is fairly simple.
Noise from the discriminator is fed to the pen-
tode section of a 7060 tube (Fig. 11-36). The L.
gain of this noise amplifier is controlled through
the action of the squelch control. Amplified noise
is detected in the other half of the 7060; the dc
Fig. 11-33 -
The 440 -MHz preamplifier is con-
structed in a 3 X 3-1/2 X 1 -inch box made of
output is amplified by half of a 7058 tube, the double -sided circuit board. All abutting edges are
output of which is connected to its other half in soldered to complete the enclosure. Two
such a way that the second half can be electric - 3 X 15/16 -inch shields separate the tuned lines.
R4
PREAMP. too CT TAP
.00i L2
+12V
C2 R2 C6
R3 R5 C4 L2
150k 220 C5
146 MHz

TO J2
LS
Cl L3
L3 TO J1
G1 R5
L1 IJ2
J1 146MHz OUTPUT +12V
INPUT

<I R2 TAP
220 C3 R1 R3 R4
1.1
Q1

(A)
m (B) FOIL SIDE
(HALF SCALE)

Fig. 11-32 - Circuit diagram (A) and pc -board

J1, J2 - PhonoNo.type, panel mount.
- 5 turns
L1 16, 5/16 -inch dia, 1/2 -inch long.
layout (B) for the 2 -meter preamplifier. Resistors Tapped at 2 turns for the antenna connection
are 1/4 -watt composition and capacitors are disk and 4 turns for G1.
ceramic unless otherwise noted. Components not
listed below are given designators for circuit -board
L2 -4 turns No. 16, 5/16 -inch dia, 3/8 -inch long.
Tapped at 2 turns.
location purposes.
C2, C6 -Air variable (Johnson 189-506-5).
L3 -
1 turn, plastic-covered hookup wire, 5/16 -
inch dia, placed between two turns of L2.
Adding Additional Channels 255
440 MHz PREAMP

440 MHz 440 MHZ 440 MHz

r T

(on 1...1 and L3) are 1/2 -inch up from the

Fig. 11-34 - Schematic diagram of the uhf ground end. Drain taps for Q1 and Q2 on L2
preamplifier. Capacitors are disk ceramic unless and L3, respectively, are made just below C2
otherwise noted. and C3.
C1 -C3, incl. - 1.4- to 9.2-pF miniature variable
RFC3, RFC4
-
RFC1, RFC2 420 -MHz choke (Miller 4584).
- Two ferrite beads on a short piece
(Johnson 189-0563-0011.
J1, J2 - BNC type, chassis mount. of No. 20 hookup wire. (Beads are available
L1 -L3, incl. - 2-5/8 X 1/4 -inch strip of brass, from Amidon Associates, 12033 Otsego St., N.
Hollywood, CA 91607.)
soldered to the enclosure on one end and to the
capacitor at the other. Input and output taps Q1, Q2 -Motorola JFET.

two first -oscillator crystals with a spdt relay. In cathodes to 8 to 9 volts positive. The full supply
some, a dpdt relay is employed, so that the voltage is used to reverse -bias the diodes not
unused crystal and its netting capacitor are selected. (There will be no voltage drop across
grounded. Most often, separate oscillators are R5 -R2 and R6 -R3, as the only current flow in
used for each channel, selection being made by these essentially open circuits is the minute leak-
grounding the cathode of the appropriate oscilla- age current through CR2 and CR3.) The diodes
tor stage. being reverse -biased, their junction capacitances
A convenient solution to the crystal -switching are quite low. The switch does two things: it
problem lies in the use of diodes, changing the grounds R4, removing the supply voltage from
bias on the diode in the desired channel so that it the cathode of CR1, and it completes the for-
goes from an open -circuit to a conducting con- ward -bias circuit from R9 through RFC4, CR1,
dition. This is particularly convenient for remote RFC1, and RI to ground.
control of the channel selection in a mobile The same power source and switching may be
installation, as up to four crystals can be used, used for both transmitter and receiver, if it is
and the system works quietly and with very low desired to switch both simultaneously. Only sep-
current drain. It is most readily applied to equip- arate 1000 -ohm isolating resistors, R8 and R9 in
ment in which one side of the crystal is ground- Fig. 11-37A, are required.
ed. The arrangements discussed thus far require
A circuit for crystal switching, for use in any one control wire between the operating position
equipment where one side of the crystal is groun- and the equipment for each channel to be switch-
ded, is given in Fig. 11-37A. An example ís the ed. A two -channel system using but one control
Marconi DT45. Three crystals are shown, but up wire is shown in Fig. 11-37B. It requires that 12
to four have been used. The do is shown here
being obtained from the 12 -volt ac line in the
receiver, with CR4, R7, and C6 comprising the
rectifier-filter circuit. The isolation capacitor, C4,
should be added if one is not already in the
circuit. The grid-to -ground capacitor, C5, usually
about 10 pF, should be removed if there is a
capacitor of this type in the circuit.
The 12 to 14 volts dc from the supply is fed
through R9 and RFC4 to the anodes of CR1,
CR2, and CR3. One diode, in this instance CR1,
will conduct, its circuit to ground being com-
pleted through Si, 111, and RFC1. Its current, Fig. 11-35 -Diagram of a limiter which may be
approximately 5 mA, and the resultant voltage added between the last i -f stage and the detector of
drop across R9, brings the voltage on the diode a receiver.
256 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
+200V NOISE AMR

8122 NOISE AUDIO

1500 9105 R106
1006 RECT. GATING
104 2701
180 dos
100K vie V2A 1
17060
VIA 7060 / 1
2 77058
NOISE 107
INPUT
908 ( -
R109 R110
FROM
9100 C1O1 9 3 1M 2208
DISC, 478 120 3
C1500106í-s
0100
75 9101
688
TO
SQUELCH
CONTROL
C107
O1 8108
LBM T C108
02
8111
8200
GROUND

EXCEPT AS INDICATED, DECIMAL T100

AUDIO
VALUES OF CAPACITANCE ARE COMM. OUTPUT
IN MICROFARADS I AF I; OTHERS
32 5000f1 AMPLIFIER
ARE IN PICOFARADS (PF OR yyF);
AUDIO
RESISTANCES ARE IN OHMS, OUTPUT
8 0000 . M 1000 000 AUDIO AMR

9113 8114 V3
3906 398 7061
R112
0111
14

AUDIO
d09
INPUT
,01
7058
7060 C116
7061 1500
4/ \5 +12V
R11S
13.5V

C110
0113 1
.01

R120
3308
C114T
C.
R121
910L
C715

560
m
-385 (TRANS.)
9145035V(REC.) 5%

Fig. 11-36 -
Schematic diagram of the audio/squelch adapter. Resistors are
1/2 -watt, five -percent tolerance composition, unless otherwise noted. Capacitors
may be disk ceramic or paper, except those with polarity marked, which are
electrolytic.
T100 -
Audio output type, 5000 -ohm primary, 3.2 -ohm secondary.

volts do be available at the control position, as cult with GE Progress-Line transmitters and re-
well as in the equipment itself. A voltage divider, ceivers. The GE units have their crystals working
R3 -R4 -R5 is tapped at 4 and 8 volts. The remote into a lower capacitance (10 pF) than the other
control switch S1 connects the switching matrix two, and the modification introduces some stray
to either +12 volts or ground. In the position capacitance, which tends to lower the frequency
shown, the do voltage is applied through RFC3 to of oscillation. Motorola crystals, for 24 -times
the diodes. Note that it is applied to the cathode frequency multiplication, may be used in the
of CR1 and the anode of CR2. The anode of Progress -Line transmitters, to get around this
CR1 has 8 volts on it, from the voltage divider, problem. It may also be necessary to increase the
so it is reverse -biased. The cathode of CR2 has 4 value of the screen-to -cathode capacitor in these
volts on it, so it is forward -biased, connecting units.
crystal Y2 into the circuit. The situation reverses Diode switching can be used in circuits where
when Si is in the other position, CR1 being both sides of the crystal are above ground, but
forward-biased and CR2 reverse -biased. providing for individual frequency adjustment
The rf chokes, RFC1 and RFC2, affect the may be difficult. A modification for the Motorola
operation. It is easiest to get working if 2.5-mH 5V transmitter is shown in Fig. 11-37C. The
chokes arc used, but smaller oncs, or TV peaking tuned circuit L1 -C4 is as in the original, with C3
coils, may work. Oscillator output level can be added to tune down into the 2 -meter band. The
checked by measuring bias developed at the first extra capacitance required is 30 to 40 pF. The
multiplier grid (or grid current) and chokes selec- netting capacitors Cl and C2 are added for
ted for best results. individual crystal -frequency adjustment. The
Thís modification is easily used with Motorola chokes and electrolytic capacitors, L2-C6 and
or Marconi transmitters, but may be more diffi- L3 -C7 are to be used only if necessary to remove
An Inexpensive Vhf Fm Receiver 257
1000
TO COMPANION UNIT, xTALC..
RCVR . OR XMTR. OS
Re
2.5mN
CR4 RT 10 R9 .0011 CA /
12V
oc l
C6
.2.CR2

C2 ` RFC
3

?! ImN

R2
1500
R3
1500 (A)

TO COMPANION UNIT

O ;
D

(B)
RI
1500'

R3

t2V 100/Iw eV
KO/1W

.0011L
ao+
%TAL.
O5C.

Fig. 11-37 -
(A) Diode -switching of crystals having paralleled "netting" capacitors, in circuits where
one side of the crystal is grounded. Three positions are shown, but up to four have been used. The dc
source and the switching circuit can be used for the crystals in the companion unit of a transceiver.
Power is shown here taken from the 12 -volt ac line in one of the units, though an external do source
can be used. (B) Switching circuit for oscillators in which the crystal is above ground on both sides.
Filtering in the 12 -volt leads may be dispensed with if the voltage source is free of hum or other
objectionable modulation. (C) Crystal switching of two positions with a single control wire. Values of
the rf chokes RFC1 and RFC2 can be set up by experiment, though 2.5 mH is satisfactory.

hum or other modulation from the dc leads. that are designed for 50- or 146 -MHz operation.1
Otherwise, the circuit is similar to that of Fig. The reason for the lower cost is twofold: much
11-37B. more 30- to 40 -MHz equipment is available, and
there is less demand for the older low-band gear.
AN INEXPENSIVE VHF FM RECEIVER 1 Dealers who offer surplus fm gear in their
catalogs include:
For whatever amateur fm frequency you may Gregory Electronics Corp. Spectronics
have in mind, a 10 -meter commercial fm receiver 249 Route 46 1009 Garfield Ave.
is a "best buy." A review of the advertisements Saddle Brook, NJ 07662 Oak Park, IL 60304
and flyers from fm dealers in used equipment will Mann Communications
P.O. Box 138
show that 30- to 40 -MHz crystal -controlled re- 18669 Ventura Blvd.
ceiver strips are priced far below those models Tarzana, CA 91356
258 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
Fig. 11-38 - Front view of the modified Pre -
Progress GE receiver. The volume and squelch
controls have been brought out to the front panel.
The panel has been refinished, and a new speaker
grill, cut from perforated aluminum stock, has
been added.

who works for a "two-way" radio service com-

pany (such firms usually keep files on all popular
models).

Crystals
The GE receiver required a 23.6 -MHz "rock."
A receiver strip that can be converted to 10 - A few words about the selection of crystals may
meter operation should cost $5 to $20, depend- help the beginner to avoid costly mistakes. The
ing on condition, age, and model. various manufacturers of fm equipment use dif-
For monitor service, a wide -band (36 -kHz - ferent oscillator circuits; thus, the capacitance
wide i -f) receiver is preferable to a narrow -band loading on the crystal also varies. Often a crystal
(13 -kHz bandwidth) model, as amateurs currently that is "on channel" in one rig can be quite far
use both wide and narrow fm deviation. A wide - off frequency in another. The crystal manufac-
band receiver can copy signals of either deviation turers can grind a crystal for any of the com-
while a narrow-band receiver is suited only for mercial fm rigs - if you supply the necessary
narrow-band transmissions. The 30- to 40 -MHz information -the make, model, and serial num-
receiver, once set on 29.6 MHz can be used to ber of your rig. Give the crystal and operating
monitor that channel, or, using a converter, to frequencies desired, and mention whether or not
receive either 6- or 2 -meter fm signals. the crystal will be used in an oven.
The scheme to use a 10 -meter receiver for vhf For a 10-meter receiver, where the crystal
fm reception is shown in Fig. 11-39. Any of the frequency is not multiplied, and no oven is used,
popular converter designs may be adopted - any you can order a general-purpose crystal and save
of the converters described in this book would be a few dollars -
"if you're a gambler. One of
an excellent choice. Instead of using the specified International Crystal's inexpensive EX crystals
conversion -oscillator crystal, however, one is chos- was tried with the 4ER7A2, and it "hit" fre-
en that will heterodyne the desired channel - quency. Anyone taking this approach does so at
149.94 MHz for example, down to 29.6 MHz. his own risk, though.
Using a 28 -MHz i -f converter, a 58.57 -MHz crystal
will be required. Alignment
A trial conversion was made on a somewhat
dilapidated GE 4ER7A2 receiver. With a receiver With crystal in hand, the next step is to get
in hand, the next job 'was to find out what the receiver going. Though fm receiver alignment
crystal would be required for reception on 29.6 can be a complicated procedure, only a few
MHz. To obtain details on receivers made for the simple adjustments are required to set up most
land mobile service, one can purchase a book of fm receivers. Test jacks are found on most recei-
schematics,2 or can consult with a local amateur vers, and these points can be used to check
operation of the various stages. The oscillator
2 Two-way Radio Engineers, Inc., 110 Tre-
mont Street, Boston, MA 02120, has a booklet of should be adjusted for a point just below maxi-
schematic diagrams covering Motorola equipment mum output. Then the discriminator must be
up to the early Motrac models. Gregory Elec- checked for a zero-voltage output with no signal
tronics (footnote 1) sells similar books of dia- input to the receiver. Then, injecting a signal
grams on the GE pre-Prog units.
from a crystal calibrator, the trimmer across the
crystal in the hf oscillator should be adjusted so
that the calibrator harmonic on 29.6 MHz also

--> 2 -METER -3. 10 -METER XTAL

FREQ.
OPERATING FRED. -29.6 MHz
2
14694MHz CONVERTER 29.6MHz RECEIVER

Fig. 11-39 - Frequency scheme to receive the 2-

(A) and 6- (B) meter national fm calling frequen-
[-158.67MHz
cies. Often a 6- to 30-pF trimmer must be added
/77 across the converter crystal to "pull" it exactly
"on channel."
6 -METER 10- METER XTAL
52. 525 M H z CONVERTER 29.6 MHz RECEIVER FREQ.."OPERATING FREQ.-29.6MHz

= 22.925MHz
Using Fm Surplus in the Home Station 259
1sT LIMITER DISCRIMINATOR
TY

Fig. 11-40 - Connections to meter (A) limiter grid voltage and (B) discriminator output voltage. M1 may
be a 0-50, 0-200, or 0-500 microammeter or 0-1 milliammeter, with R1 chosen to provide a half-scale
reading with a signal input of 50 microvolts. M2 is a zero -center (50.0-50-/lA) meter with an appropriate
multiplier resistor (usually 0.22 to 0.47 megohms). T1 is the discriminator transformer.

produces a zero voltage reading at the discrimina- nator meter will indicate to which side, and
tor output. Metering the grid voltage developed at relatively how far off, a particular station may
the second limiter, any rf-stage adjustments be. Of course to remain accurate, the receiver
should be peaked for maximum voltage at the must be checked against a frequency standard on
limiter. a regular basis.
If a variable -frequency signal generator is avail-
able, it can be used to further check the discrim- USING FM SURPLUS IN THE HOME
inator. Again metering the discriminator output, STATION
check to see that a cw signal 15 kHz on either An fm station for home use can be construc-
side of 29.6 MHz produces the same absolute ted easily and inexpensively from surplus trans-
voltage reading. (Note that the voltages read will mitter and receiver strips. Usually, the individual
have different polarities.) If one side or the other strips cost less than a complete unit because they
is off a bit, adjusting the input trimmer on the have little resale value to commercial users. Two -
discriminator transformer will usually correct the meter transmitter and receiver strips which have
situation.3 been removed from mobile rigs are now priced at
As can be seen from alignment procedure about $30 each, while the 6 -meter and 450-MHz
outlined above, voltages measured at the limiter models are $8 to $12. To make a home station
and discriminator stages tell quite a bit about (often called a base station by fm-ers) requires an
how a receiver is working. Constant metering of ac -operated power supply, a control relay, and a
these stages will make the receiver more useful in few panel controls.
giving reports to other stations. The first meter, A heavy-duty homemade power supply (Fig.
Fig. 11-40, indicates the relative strength of in- 11-42) provides the voltages required by the
coming signals. One peculiarity of the fm mode is decks. The same circuit can be employed to
that, once a signal produces full quieting of the power many types of retired mobile rigs. It has
receiver noise, further increases in signal strength been tried with a Motorola 80D and a Motorola
will not be detectable by ear. This limiter grid- T44AAV, as well as the GE MTS transceiver. The
current meter serves as an S -meter, allowing an large power transformer, Ti, was garnered from
operator to observe changes in strength of in- an old TV set. T2 powers the bias circuit, while
coming signals. T3 supplies voltage to the tube filaments and
A second meter, connected to the discrimina- oven heaters. RI sets the receiver B plus at 220
tor, indicates if a station being received is "on V, and R2 adjusts the level of audio fed from the
channel." Because of the temperature changes receiver to the transmitter when the REPEAT
and vibration encountered in mobile operation, mode is selected. This feature may be omitted if
and because of the crystal prblems mentioned operation as a repeater is not desired.
earlier, many stations can get off frequency. (On As solid-state rigs continue to replace units
loud signals, off-frequency operation is also diffi- using tubes in mobile -service equipment, amateurs
cult to detect by ear.) Once the receiver is will have a chance to acquire some first-class gear.
properly aligned with a crystal calibrator, the A few simple modifications are all that is re-
plus -or -minus voltage indication on the discrimi-
3 A complete alignment should only be attem-
pted if the required test equipment is available.
The manufacturer's instructions for testing and
adjustment should be followed implicitly. Trial -
and -error alignment will not work on most fm
receivers.
-
quired to ready a retired unit for amateur fm use.

°ti L

Fig. 11-41 -G E Progress Line decks are used in

this station. A 50-0-50-1/A meter (Simpson
2123-27507) is employed to monitor various stages
in the transmitter and receiver. Sampling is done at
the existing test jacks on the GE decks.
 260 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES

K18I
153505 TRANS
OF INAL

.01
K1C1
+ 280V TRANS.

T2
¡ ^\^
V V
4706 4706
LS
1011
1000
SOW
2500
50 1.4 X1D

4
NC
375V CR1 CR2 R7
1.6,1F=
375V
/M CR3 CR4 600V
m
O,NF
450V
m 400 Ti. 1 o
+220V TRANS.
OSC.

m7
16pF 001
+220V REC.
H4 N 01."./\./"\r 6005 450V
SV VOLUME
21 LS -1
01
22 560 REC. AUDIO
T3
1W CR5 Sw
70V TO TRANS.
PTT
40pF 40MF ,O1 FIL.
REMOTE
100V
r
63V
250;-/ 2505 +
4700 OFF O REPEAT
J /T17 SW
AUDIOFROM,
O
NC, S18
AM P.

OS K1A NC
PTT

CR6 -12V
CONTROL
O1QO
NC
ISiCMRELELAT
000
t KZS MC.

IA
T4 n D51
25V +
-L_ AUDIO MIC. O
NC
SID
TO TRAN51 J4 REMOTE CONTROL
/
12V 12V AC
FIL, AND OVENS
NCO.
14,0 YIC. INPUT
500
R8 0 REC.AUDIO

D
1
-001 SSE 1D0

a © OTT

EXCEPT AS INDICATED, DECIMAL VAUJES OF LEVEL © MIC.

117V
CAPACITANCE ARE IN MICROFARADS l J1F I ;
OTHERS ARE IN PICOFARADS (pF OR yyF); 10
SET
e GND.

AC RESISTANCES ARE IN OHMS ;

k I 000.
NCNO CONNECTION

Fig. 11-42 - Homemade power supply for the G.E. transceiver, plus the audio and control switching.
Unless otherwise noted, resistors are 1/2-watt composition and capacitors are disk ceramic, except those
with polarity marked, which are electrolytic.
CR1-CR5, incl. - Silicon rectifier, 1000 PRV, 1A R7 Wire -wound adjustable -tap power resistor. --
--
CR6 Silicon rectifier, 200 PRV, 1.5 A. R8 Linear -taper composition control.
CR7 Thyrector transient suppressor (G.E. S1
6RS2OSP4B4).
-pole (5 used), 5 -position, 3 -wafer phenolic -6
rotary switch (Centralab PA -1020).
DS1
J3 -- 12 -volt lamp.
Microphone jack, 3 circuit, nonshorting.
-4
T2 TV power transformer (Allied 6K91VG
suitable substitute).
-
.J4
K1 -3
-circuit jack, panel mount.
-pole, double -throw relay, 12 -volt coil, 2-A
T3 Filament transformer, 6.3-V, 0.6-A second- -
ary (Allied 6K32HF).
K2
contacts (Potter & Brumfield KA14AY).
-Antenna relay.
T4 Control transformer, 12 -volt, 8-A secondary -
(Allied 6K8OVBR).
L5 - Power choke, 10 H, 200 mA (Allied Z1 pad (Calectro S2-175). -L
6X37VG).
ACCESSORIES
AN FM SWEEP GENERATOR
In the past a sweep generator was such an
expensive piece of test equipment that it was
rarely found in amateurs' workshops. Today,
however, because of a new low-cost integrated -
circuit function generator, anyone willing to
spend an evening building a simple project can
enjoy the advantages of a sweep oscillator. Such a
generator is useful for aligning fm receiver i -f
strips, for checking homemade i -f amplifiers and
filters, and for determining the response charac-
teristics of band-pass tuned circuits.
Fig. 11-43
alignment.
- The sweep generator for fm i -f The heart of the sweep generator is a Signetics
NE566 integrated-circuit voltage-controlled oscil-
An Fm Sweep Generator 261
REGULATOR
BOTTOM
CR 1 +16v VIEW
Ti 4
U3 U2
3

15V - U3
N3723í
Fig. 11-44 -
Circuit diagram of the 6.3V CR2 300 F

sweep generator. Resistors are 1/2 -watt

composition and capacitors are disk ce- U3
ramic (except those with polarity
marked, which are electrolytic) unless
otherwise noted.
- -
Cl, C2 See Table 1.
CR1, CR2 Silicon diode, 100 PRV, 500 mA.
SWEEP OSc.
RI 10k OSc.
1500 100-1000kHt
DS1 - Neon indicator, panel mount, for 117 V ac.
SWEEP FRED

J1, J2 -5 -way binding post. r--4,-'VV v 4 A toX y3D R2

20k
Q1, 02 - GE bipolar transistor, most low -power
npn amplifier or switching types with medium
10
FREO
ADJ.

5
beta should be suitable.
R1 - -
R2, R3
Linear taper, pc mount.
Linear taper, panel mount.
U1
NE566T
iv
Ii C
SWEEP
5

- -
ON 1

S1, S2 Spst toggle. ,--.-. R3

Jl

T1 Filament type, 117-V primary, secondary 6.3

V at 300 mA.
10k
01
o03T
+,
C1

02
úó 1M
OUTPUT
LE
OUTPUT

U1 -U3, incl. - Signetics integrated circuit (avail-

2113414 2N3414
S M
J2

able from Compar, 2531 Whitney Avenue, EXCEPT AS INDICATED, DECIMAL

Hamden, CT 06518; the NE566T [TO-991 VALUES OF CAPACITANCE ARE
costs $10.25 and the NE566V [DIP package] is IN MICROFARADS IYF); OTHERS

$9.50; the N5723L is $2.00). 1000, M 1000 000 ARE IN PICOFARADS I AF OR »F);
N.C.- NO CONNECTION RESISTANCES ARE IN OHMS;
S.M. SILVER MICA

lator. The '566 produces square- and triangular-

wave outputs simultaneously. The frequency of
Power for the generator is provided by a
oscillation is determined by an external resistor, a 6.3 -volt filament transformer. A voltage doubler
capacitor, and the voltage applied to the control and an N5723 integrated-circuit regulator are
terminal. The device can be made to shift fre- employed to deliver 12 volts to the NE566s.
quency over a ten -to -one range with exceptional Approximately 40 mA of current is needed.
linearity. The upper frequency limit of the
NE566 is approximately 1 MHz. Construction
Circuit Information The sweep generator is assembled on elec-
tronic pegboard. A pc board can be employed, if
A schematic diagram of the sweep generator is desired, although making a circuit board would
shown in Fig. 11-44. Integrated circuit U2 func- probably double the amount of time needed to
tions as the main oscillator whose frequency can complete the project. A small 3 x 5 x 4 -inch
be varied from 100 kHz to 1 MHz. Control R2 cowl -type Minibox is used as an enclosure.
provides the means of frequency adjustment. The After checking the completed unit for wiring
output level from the oscillator may be varied by errors, apply line voltage and measure the do
adjustment of R3. When Si is closed, a sawtooth voltage at pin 6 of U3. The reading should be
wave is applied to the control terminal of U2, approximately 12 volts. With Si open, the output
sweeping the output frequency. The sweep fre- from the oscillator can be checked by monitoring
quency is determined by the setting of R1. The the second harmonic with a be radio. A short
triangle -wave output from U1 is modified to a piece of hookup wire connected to J1 will serve
sawtooth wave using Q1 and Q2. Square -wave as an antenna. With Si closed, R1 should be set
output from pin 3 of U1 is applied to Q2 via Ql, to produce the desired sweep frequency.
which is connected to function as a Zener diode.
When the voltage reaches sufficient level to turn
Q2 on, the timing capacitor, Cl, is immediately
discharged. This discharge occurs just as the tri-
angle wave form reaches its peak voltage, prevent-
ing the down -slope side of the wave form from
appearing at the output. Either the sweep timing
or the frequency range of the generator may be
modified by changing the value of the timing
capacitor, Cl for U1 and C2 for U2.

Fig. 11.45 - Inside view of the sweep generator.

Small components are mounted on Vector T2.8
terminals which have been inserted in a piece of
electronic pegboard. The voltage-regulator IC is
i 1
located to the far right, just above the power
transformer. The small pc -mount control to the far
left is used to set the sweep frequency.
 262 FM TRANSMITTERS, RECEIVERS AND ACCESSORIES
AN FM-ER'S FREQUENCY STANDARD

The fm channels now in use have been spaced

at 30- or 60-kHz increments by informal agree-
ment of those using fm repeaters. Although not
all fm operation is on these channels, in most
cases repeaters do operate on "standard" frequen-
cies. The frequency standard shown in Fig. 11-47
delivers a spectrum of 30-kHz markers up to
approximately 300 MHz. These markers can be
used to calibrate the frequency of a 50-, 144-, or
220 -MHz narrow-band receiver with an accuracy
of 50 Hz or less. Of course, the standard must be
first calibrated against WWV or WWVH at 15 ry
MHz if high accuracy is to be obtained.
The frequency standard uses a high-speed UL
gate package for the oscillator, buffer, and output
amplifier. A Shottky-barrier IC, the SN74S00,
can be used if uhf output is desired. Two divide - Fig. 11-46 The fm-er's frequency standard is -
by -10 ICs produce the 300- and 30 -kHz markers. built in a Ten-Tec JW-4 enclosure. The panel
A National Semiconductor IC, the LM109, regu- switches control the ac power and the frequency of
lates the power supply voltage for the TTL logic. the output markers.
If desired, the unit can be battery operated from the precision with which this calibration is made.
a 9- or 12 -volt source by connecting the battery Allow time for Cl to "settle down" mechanically
to point A of Fig. 11-47. after each adjustment. At very close to zero beat
The frequency standard is constructed on an the incoming signal will vary rapidly in signal
etched circuit board which is housed in a Ten-Tec strength. The number of variations per second is
JW-4 enclosure. Point-to-point wiring may be the difference in frequency between the two
used, if desired. When the unit is finished and all carriers. Adjust Cl until the pulsation rate is as
wiring has been checked for errors, turn on the slow as you can make it.
ac power and connect a voltmeter to pin 2 of the A very short patch cable should be used to
LM109. The reading obtained should be 5 volts couple the standard to the receiver to be cali-
plus or minus 0.1 volt. Then, monitoring the brated. With a voltmeter connected to the re-
WWV transmission on 15 MHz, adjust Cl for zero ceiver discriminator, adjust the inductor or trim-
beat with the standard-frequency transmission. mer capacitor which calibrates the receiver crystal
The accuracy of the standard is dependent upon until a zero reading is obtained.
3003 OUTPUT

i2---- I
X52
U1A 3MHr. u2 i {z U3 I;
Kc,l, 5-25 RF.=

9
sR74H00
e
IQI It I:
U1
WC SN7490 11
4
SN7490
Y1 CI I 2.21 3,a t,10
2200
680

rz
2200 EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
IN MICROFARADS IJF 1; OTHERS
ARE IN PICOFARADS l pF ORjI F l-
CRI-CR4 RESISTANCES ARE IN OHMS,
U4 +SV
REG. BOTTOM VIEW .1000.

512V LM309 N.C.= NO CONNECTION

i woo T T
.!r
.22 CR1-CR4, incl.
mA.
- Silicon diode, 100 PRV, 500
,-1-735V

DS1
J1
- Neon indicator, 117 V, panel mount.
- Phono type.
S1, S2 - Spht toggle.
Fig. 11-47 -
Diagram of the frequency standard. T1 - 300-mA filament type, pc mount.
Resistors are 1/2 -watt composition and capacitors
are disk ceramic, except as noted otherwise.
U1 - 6.3-V,
TTL quad gate (Motorola, Signetics, or TI
SN74H00).
A pc board is available from Spectrum Research U2, U3 - TTL MSI decade counter (Motorola,
Labs, P.O. Box 5824, Tucson, Arizona 85702, for Signetics, or TI SN7490).
$3.50. A template and parts layout appeared in U4 - IC regulator (National Semiconductor
QST for April, 1972. LM 109)

Cl - 5- to 25-pF ceramic trimmer. Y1 - 3-MHz crystal

EX).
.

(International Crystal type

 Chapter 12

Repeaters - Theory and Practice

remote-control operation must be obtained and
provisions made to control the equipment over a
telephone line or a radio circuit on 220 MHz or
higher. The licensed operator must then be on
hand at an authorized control point. Fig. 12-2B
shows the simplest system of this type. The control
decoder may be variously designed to respond to
simple audio tones, dial pulsed tones, or even
n3- Touch -Tone signals. If a leased telephone line with
rr dc continuity is used, control voltages may be sent
directly, requiring no decoder. A 3 -minute timer to
disable the repeater transmitter is provided for
fail-safe operation. This timer resets during pauses
between transmissions and does not interfere with
normal communications. The system just outlined
is suitable where all operation is to be through the
:1;,-0«¿I¿ . slolí repeater and where the frequencies to be used have
.t.e.; no other activity.
Fig. 12-1 - When selecting a repeater site, height
Remote Base Stations
above the surrounding area is usually an important
consideration. Here, K7UDG and W7FHZ investi- The remote base, like the repeater, utilizes a
gate a proposed repeater location near Wenatchie, superior location for transmission and reception,
Washington.
but is basically a simplex device. That is, it
transmits and receives on a single frequency in
A repeater is a device which retransmits re- order to communicate with other stations also
ceived signals in order to provide improved com- operating on that frequency. The operator of the
munications range and coverage. This communica- remote base listens to his hilltop receiver and keys
tions enhancement is possible because the repeater his hilltop transmitter over his 220 -MHz or higher
can be located at an elevated site which has control channels (or telephone line). Fig. 12-3A
coverage superior to that of lower-lying stations. A shows such a system. Control and keying features
major improvement is usually found when a
repeater is used between vhf mobile stations, which
normally are severely limited by their low antenna
heights, especially in rough terrain. Aupo
The simplest repeater consists of a receiver with
} RECEMER xNTR

its audio output directly connected to the audio

RESINO
input of an associated transmitter tuned to a CONTROL
SWITCH LINE
(A)
second frequency. In this way, everything received
on the first frequency is retransmitted on the
second frequency. But, certain additional features
are required to produce a workable repeater. These
are shown in Fig. 12-2A. The "COR" or carrier-
operated relay is a device connected to the receiver
squelch circuit which provides a relay -contact
closure to key the transmitter when an input signal
of adequate strength is present. As all amateur
transmissions require a licensed operator to control
the emissions, a "control" switch is provided in the IRESET
keying path so that the operator may exercise his
duties. This repeater, as shown, is suitable for HCONTROL
RECEIVER
CONTROL
DECODER (Bl
installation where an operator is present, such as at
the home of a local amateur with a superior ow

location. It would require no special licensing

LINE
E
under existing rules.
In the case of a repeater located where no Fig. 12-2 -Simple repeaters. The system at A is
for local control. Remote control is shown at B.
licensed operator is available, a special license for

263
264 REPEATERS - THEORY AND PRACTICE
V
-' DOWN
CHANNEL
XYTR
--1-
CONTROL
STATION
O

(A)
UP
CHANNEL
RECEIVER
%YTR
V
UP F2
CHANNEL XN TR
RECEIVER

Fig. 12-3 -
A remote base is shown
at A. A repeater with remote-base
F2
operating capability is shown at B. CONTROL

1
STATION RECEIVER
Control and keying circuits are not
shown. Telephone-line control may 'V V
be substituted for the radio -control DOWN
channels shown. CHANNEL
FI
(B) ACTS RECEIVER

have been omitted for clarity. In some areas of frequencies, such as 146.94 MHz, because it
high activity, repeaters have all but disappeared, in minimizes interference to simplex operation and
favor of remote bases, because of the interference permits simplex communications through the
to simplex activity caused by repeaters unable to system with passing mobiles who may not have
monitor their output frequency from the trans- facilities for the repeater-input frequency.
mitter location. The audio interface between the repeater re-
ceivers and transmitters can, with some equipment,
Complete System consist of a direct connection bridging the trans-
Fig. 12-3B shows a repeater that combines the mitter microphone inputs across the receiver speak-
best features of the simple repeater and the remote er outputs. This is not recommended, however,
base. Again, necessary control and keying features because of the degradation of the audio quality in
have not been shown in order to simplify the the receiver -output stages. A cathode follower
drawing and make it easier to follow. This repeater connected to each receiver's first squelch -con-
is compatible with simplex operation on the trolled audio amplifier stage provides the best
output frequency because the operator in control results. A repeater should maintain a flat response
monitors the output frequency from a receiver at across its audio passband to maintain the repeater
the repeater site between transmissions. The con- intelligibility at the same level as direct trans-
trol operator may also operate the system as a missions. There should be no noticeable difference
remote base. This type of system is almost man- between repeated and direct transmissions, The
datory for operation on one of the national calling intelligibility of some repeaters suffers because of
improper level settings, which cause excessive
clipping and distortion. The clipper in the repeat-
er transmitter should be set for the maximum
system deviation, for example, 10 kHz. Then the
receiver level driving the transmitter should be set
by applying an input signal of known deviation
below the maximum, such as 5 kHz, and adjusting
the receiver audio gain to produce the same
deviation at the repeater output. Signals will then
be repeated linearly up to the maximum desired
deviation. The only incoming signal that should be
clipped in a properly adjusted repeater ís an
over -deviated signal.
1
The choice of repeater input and output fre-
quencies must be made carefully. On two meters,
600-kHz spacing between the input and output
frequencies is common. Closer spacing makes
interference problems between the repeater trans-
o
mitter and receiver more severe. Greater spacing is
not recommended if the user's transmitters must
be switched between the two frequencies, as
happens when the output frequency is also used

Fig. 12-4 - W1FBY tries local control on a

450-MHz repeater. This "machine" uses transmit-
ter and receiver decks from the Motorola T-44
series mobile transceivers.
Practical Repeater Circuits 265
choices of input
for simplex operation, either for short-range com- alternative. On six meters, several
to maintain communications when frequencies have been paired with 52.525 MHz.
munications, or
the repeater is not functioning. A 5 -MHz spacing is The choice and usage are matters for local agree-
recommended on 440 MHz. ment.
Careful consideration of other activity in the In some cases where there is overlapping geo-
area should be made to prevent interference to or graphical coverage of repeaters using the same
from the repeater. Many "open" or general -use frequencies, special methods for selecting the
employed. One of the
repeaters have been installed on one of the national desired repeater have been requires the user to
calling frequencies. On two meters, a 146.94 -MHz most common techniques
a 0.5 -second burst of a
output is usually paired with a 146.34 -MHz input, automatically transmit start of each transmis-
and many travelers have made good use of this specific audio tone at the
tones are used to select different
combination where it is found. Where 146.94 -MHz sion. Different
tone frequencies are 1800,
simplex activity has not permitted a repeater on repeaters. Standard 2400 Hz.
1950, 2100, 2250, and
this frequency, 146.76 MHz has been used as an

PRACTICAL REPEATER CIRCUITS

Choice of Equipment which are placed in the receiver and transmitter
feeders, as shown in Fig. 12-5A. Construction
Because of their proven reliability, com- details of a simple duplexer using this design are
mercially made transmitter and receiver decks are given below. The second type of duplexer uses the
generally used in repeater installations. Units de- hybrid ring circuit of Fig. 12-5B. The isolation
signed for repeater or duplex service are preferred provided by a hybrid ring depends on the arrival of
because they have the extra shielding and filtering two signals which are of equal magnitude, but 180
necessary to hold mutual interference to a mini- degrees out of phase, at the output terminal. The
mum when both the receiver and transmitter are signal is split into two parts at the input. One signal
operated simultaneously. Those who wish to build path is made one-half wavelength longer than the
their own fm transmitters and receivers will find other to achieve the required 180 -degree phase
suitable circuits in Chapter 11 of this manual and shift.
in The Radio Amateur's Handbook. For duplex operation, the hybrid ring must be
Wide-band noise produced by the transmitter is made to pass the desired frequency while atten-
a major factor in the design of any repeater. The uating the unwanted signal. A high -Q cavity is
use of high -Q tuned circuits between each stage of placed in the ring to act as a switch to take one of
the transmitter, plus shielding and filtering the signal paths out of the circuit. The cavity is
throughout the repeater installation, will hold the resonant at the pass frequency of the hybrid ring
wide -band noise to approximately 80 dB below the
output carrier. However, this is not sufficient to
prevent desensitization - the reduction in sensi-
CAVITIES

tivity of the receiver caused by noise or rf overload

from the nearby transmitter - if the antennas for
the two units are placed physically close together.
Desensitization can be checked easily by moni-
toring the limiter current of the receiver with the
transmitter switched off, then on. If the limiter
current increases when the transmitter is turned
on, then the problem is present. Only physical
(A)
isolation of the antennas or the use of high -Q
tuned circuits in the transmitter and receiver
feeders will improve the situation.
Duplexers
The ultimate answer to the problem of receiver
desensing is to locate the repeater transmitter a
mile or more away from the receiver. The two can
be interconnected by telephone line or uhf link.
Another effective approach is to use a single
antenna with a duplexer, a device that provides up
to 120 dB of isolation between the transmitter and
receiver. High -Q cavities in the duplexer prevent
transmitted signal energy and wide -band noise
from degrading the sensitivity of the receiver, even
though the transmitter and receiver are operating (B)
on a single antenna simultaneously.
Duplexers fall into two general categories. The Fig. 12-5 - (A) Simple cavity and (B) hybrid -ring
first, and simplest, consists of a number of cavities duplexers.
 266 REPEATERS - THEORY AND PRACTICE
Fig. 12-6 -
The 5 -cavity duplexer used at the
K1IGF repeater. The cavities are made to the
specifications of Fig. 12-7.

:Y construction and alignment of a hybrid -ring du-

PP7
:, SFi plexer is beyond the scope of this text. Excellent
information on this subject is contained in bulletin
EMI No. 104 from Sinclair Radio Laboratories,
Box 23, Tonawanda, NY 14150. The same firm
manufactures a complete line of vhf duplexers.
A Homemade Duplexer
Repeater groups can easily construct a simple
146 -MHz duplexer suitable for transmitters run-
ning up to 100 watts output. Insertion loss
measured on a sample unit built for the K1IGF
repeater was approximately 2 dB for the trans-
mitter leg and 2.2 dB for the receiver section. The
original design for this duplexer was provided by
WIGAN.
All cavities are made from 4 -inch dia copper
drain pipe, as shown in Fig. 12-7. The cavity is
tuned by means of a plunger made of 1-1/4- and
1 -inch copper pipe. The top plate is machined from
and thus will act as a short circuit at this
copper or brass sheet stock to fit tightly into the
frequency. Quarter-wavelength lines isolate the
cavity from the signal so that the desired energy is cylinder. The input and output coupling links,
not attenuated. which enter the cavity by way of Teflon insulators,
At the frequency to be canceled, the cavity is are made using No. 14 wire. A l-1/8 x 4 x
1 -1/2 -inch Minibox is used to enclose the input and
not resonant and therefore it acts as a high
impedance in parallel with one branch of the output connections. The transmitter cavities use
hybrid ring. Because the cavity does contribute 12-pF air-variable capacitors for coupling (Fig.
some phase shift, a stub tuner is included in the 12-7A) while the receiver units employ inductors
opposite leg as a means to adjust the ring for made from 2 -1/2-inch lengths of No. 14 wire (Fig.
maximum attenuation of the unwanted signal. The 12-7B). Otherwise, the cavities are identical.
Rough alignment of each cavity can be accom-
plished using a transmitter, wattmeter and dummy
TRANSMITTER CAVITY RECEIVER CAVITY load. The transmitter cavities are tuned to deliver
maximum power to the 50-ohm load. The same
O procedure is used for the receiver cavities, but with
1 "4 the transmitter set on the receive frequency. Then,

0111
W PIPE SLEEVE TO PREVENT
BENDING MINIBOX
-CONNECTOR
all cavities should be connected with short lengths
of double -shielded coaxial cable (RG-9B or
RG-55A), as shown in Fig. 12-5A. The cavities
should be mounted in a wooden frame so that
BRASS TEFLON
OR COPPER FEEDTHROUGH individual units do not touch each other. The only
PLATE ground path should be through the feeder cables.
The quarter- and half-wavelength cables that con-
nect to the antenna should be cut carefully and
checked for resonance using a dipper.
Oki
0-
60 -.
1 --(WO 1~y-, S° M0-My'
+1. ..
40 41 40 SZMl~
.. ti
ml

-
0 100 200 300 400
o HORIZONTAL SPACING (FEET)

k- 4" DIA CAPPER

OftAIN PIPE
0 20 40
VERTICAL SPACING (FEET)
(A)
60 80
(B)

Fig. 12-7 - Details of the 146 -MHz cavity filter to

be used as part of a duplexer. Connections for the
Fig. 12-8 -
Charts to calculate the amount of
isolation achieved by (A) vertical and (B) horizon-
transmitter type are shown at A and for the tal spacing of repeater antennas. If 600 -kHz separa-
receiver cavity at B. Cl is a 1.9- to 12-pF air tion between the transmitted and received frequen-
variable (Johnson 189-505-5). L1 consists of a cies is used, approximately 58-dB attenuation
2 -1/2 -inch piece of No. 16 silver-plated wire. (indicated by the dotted line) will be needed.
 267

DIPOLE ELEMENT

GAMMA MATCH DETAIL

S

(c)
SEAL WITH SHRINK
RG-8A/U / TUBING AFTER ADJUSTMENT
CENTER CONDUCTOR

Fig. 12-10 - (A) Phasing harness, (B) dipole

elements, and (C) feed -point detail of the dipole
array. The gamma -match capacitor consists of a
7 -inch length of the inner conductor of RG-BA/U
coaxial cable over which is placed a 7 -inch piece of
Fig. 12-9 - An array of dipole elements which 1/2 -inch dia aluminum tubing. The gamma match
of each dipole should be adjusted for minimum
provides approximately 6 dB of gain.
SWR when fed with 50 -ohm cable, before the
phasing harness is attached. Each phasing line
Once the duplexer has been installed, connect a should be checked with a dipper to assure it has
signal generator to the antenna connector and been cut to resonance before the harness is
recheck the alignment of the receiver cavities for assembled. If more than 100 watts will be fed to
minimum insertion loss. Then, remove the gener- the antenna, the phasing lines should be made from
RG-11 A/U.
ator and replace it with a power meter and dummy
load. Turn on the transmitter and adjust the
transmitter cavities for maximum power output. coverage area desired for their machine and the
When these adjustments are complete, connect the type of antenna pattern which is needed to provide
antenna. Repeak the cavities for best reception of a this coverage. For directional coverage where the
weak signal source, consistant with maximum repeater is located at one end of the desired area, a
power output. Several rounds of adjustments are Yagi, corner reflector, or collinear array may be
usually needed to achieve optimum performance suitable.
from the duplexer. Desensing of the receiver by the If omnidirectional coverage or a cardioid pat-
transmitter is an indication of improper duplexer tern is chosen, the antenna shown in Fig. 12-9 can
adjustment or a lack of shielding of individual be used. It consists of four dipoles, each fed by
sections of the repeater. means of a gamma match. The dipoles may be
mounted around a tower or pole to provide
ANTENNA CONSIDERATIONS approximately 6 dB gain. If all of the elements are
mounted on one side of the supporting mast, a
If two antennas are used at a single site, there cardioid pattern with approximately 9 dB gain will
will be a minimum spacing of the two antennas result. The lengths needed for the coaxial phasing
required to prevent desensing. Fig. 12-8 indicates harness are shown in Fig. 12-10. A commercially
the spacing necessary for repeaters operating in the made version of this antenna is available from
50-, 144-, and 420 -MHz bands. An examination of Cush -Craft, Manchester, NH 03103.
12-8 will show that vertical spacing is far more
effective than is horizontal separation. The chart
assumes unity -gain antennas will be used. If some
type of gain antenna is employed, the pattern of
the antennas will be a modifying factor.
The type of antenna chosen for a repeater may
be as simple as a quarter-wave ground plane or as
complex as the requirements for a particular
repeater dictate. A repeater group must decide the fir
*
Fig. 12-11 -
Close-up view of a single dipole
element. A TV -type U clamp holds the assembly to
a supporting mast or tower leg.
 268 REPEATERS - THEORY AND PRACTICE

i/ 41,
LAW

/ V Sr Ir V Ir 7i
9 9
MOP

.4,

Fig. 12-12 -
Front and inside (right) views of the
r
control unit built by WA1 DMX for the WA1 KGQ TLi_..,

repeater. Plug-in relays and timers are employed

for easy maintenance, should a failure occur.
CONTROL
AMP DELAY SWITCH Two connections are needed between the re-
peater receiver and transmitter: audio and trans-
mitter control. The audio should be fed through an
INPUT
impedance-matching network to assure that the
receiver output circuit has a constant load while
the transmitter receives the proper input imped-
MPS -A55
ance. Filters limiting the audio response to the
K2B
-o 300- to 3000 -Hz band are desirable, and with some
gear an audio-compensation network may be re-
Lo quired. A typical COR (carrier -operated relay)
-o
t
K2C
-o
circuit is shown in Fig. 12-13. This unit may be
operated by the grid current of a tube limiter or
the dc output of the noise detector in a solid-state
Fig. 12-13 -
COR circuit for repeater use. R2 sets
the length of time that K2 will stay closed after the
receiver.
Timers
input voltage disappears. K2 may be any relay with
a 12 -volt coil, although the long -life reed type is Normally a repeater is given a "tail"; a timer
preferred. CR1 Is a silicon diode. holds the repeater transmitter on for a few seconds
after the input signal disappears. This delay pre-
+12V vents the repeater from being keyed on and off by
a rapidly fading signal. Other timers keep each
o REMOTE
RESET
transmission to less than three minutes duration
(an FCC requirement for stations using a remote -
control license), turn on identification, and control
logging functions. A simple timer circuit is shown
in Fig. 12-14A. A unijunction transistor sets the
timing cycle while a silicon-controlled rectifier
activates the control relay. If desired, an all -solid-
state control system can be constructed using a
flip-flop IC In place of the relay.
Two other timer circuits are shown in Fig.
12-14B and C. At B, a silicon -controlled switch,

+9V
Fig. 12-14 -
Timer Cir-
cuits. (A) Cl should be
a low -leakage capacitor;
TIME
15M K1 may be any minia-
RI ture relay with a 12 -
3K58 CRI
volt coil. Reset of the
timer is accomplished
by interrupting the sup-
ply voltage momentar-
ily. (B) SCS timer. K1
may be any relay with a
6 -volt coil. (C) FET
N.C.NO CONNECTION timer. S1 can be a mo-
mentary -contact switch
(B) or a relay contact.
Logging and Identification
269
CR1, is employed in conjunction with R1 to 1505
+15V

provide up to 4 minutes of time delay. When CR1

fires, Q2 activates Kl. The amount of time delay -15V 30V
may be adjusted using R1. For circuits which
require a pulse output rather than relay control, 15Y
OUTPUT
the circuit of C may be used. Q1 and Q3 are
connected to form a multivibrator. The timing 12K

cycle is started by closing Si. R2 and C2 set the

length of the timing cycle. Pulse output is taken
from the collector of Q3.
68K HEP
803
An Audio Mixer
It is often necessary to combine audio signals at 681(

a repeater when several receivers or transmitters are 560

to be cross connected. The mixer circuit shown at vK
00K
Fig. 12-15 was designed by W1ELU and W1IRH.
Inputs to the mixer are activated by FET switches. 1N914

When a 15 -volt potential is applied to the ENABLE

AUDIO ENABLE
connection, the audio signal is passed to the adder AUDIO ENABLE AUDIO ENABLE

circuit which uses an operational amplifier con-

nected as a "summer." The sum of all of the inputs Fig. 12-15 -
An audio mixer which can be
appears at the output of the op amp. More than controlled from a remote location. U1 is an
three input sections may be used, if desired. op -amp IC (Motorola MC1709G).

LOGGING AND I DENTI FICATION

Current FCC rules require that a log be kept of Identification of the repeater itself may be
repeater operations, just as with any other licensed done by users, but lest a forgetful operator leave
amateur station. The most effective .method of the repeater unknown, some form of automatic
logging has proven to be a tape recorder equipped "ID" is preferred. A tape deck with a short loop
with a 10- to 20 -second timer. The recorder is tape for voice "ID" or a digital cw generator has
turned on for a short time (set by the timer) at the proven to be effective. A suitable digital unit was
beginning of each series of transmissions, so that described in QST for June 1970.
the calls of using stations may be recorded. In
some systems the time must be given by the user, THE REPEATER -A GOOD NEIGHBOR
while on others time signals from WWV or CHU are
recorded simultaneously on a second track of the Because a repeater is usually situated in an
tape. excellent location for vhf operation and is equip -
C2
.05íf R7
2N4123 2N4123
C4
-r`I
BAUDIO OUT
2200
R3

Fig. 12-17 - Schematic diagram of the "electronic

whistle." The main diagram is for high -impedance
output. All values of capacitance are in µF;
polarity indicates electrolytic.

iz
ó

Fig. 12-18 - The single -tone encoder is con-

structed on an etched'circuit board which fits into
an Amphenol 141-006-01 edge -contact connector.
10
270
Mk .4%i;.:.
~
NIiiik

r,nk
211114::
aii a
REPEATERS - THEORY AND PRACTICE
!CENTER FREQUENCY

II ! CI, C2 ® C.3 RECEIVER

PASS ELAND

,
I

+14V . mm..

s6K/tkw

.. 100K4W,r
6.8ic/y4w
OUTPUT

T !
.-.
Fig. 12-21 -
WINDOW

The "window" concept. The selec-

tivity curve of the receiver slopes, becoming wider
for strong signals. The device described effectively
establishes a window 15 kHz either side of the
center frequency. Any signal deviating more than
r-+
C,

IOK/y4W
2
R7
,
:

C1,C2,C3 -SEE NOMOGRAPH(

t

!!!r
- --
15 kHz or more than 3 kHz off the center
frequency is rejected.

-
RI SEE TEXT
with a good ear for frequency can use a short
Q1-MP5 6521 OR 214.339 I I I
1

.0001
1.,
whistle instead of an electronically generated tone
.001 .01 0.1
to key the repeater. A better approach, however, is
T CAPACITORS fcF a simple transistor tone generator, such as shown in
Fig. 12-17.
Fig. 12-19 - Circuit diagram of the audio oscillator
(courtesy of Electronics World magazine). The
The whistle -on device was built for use with a
Motorola 30D transmitter on a 1-1/2 x 2 -1/2 -inch
nomogram provides the values needed for Cl, C2, piece of Vectorbord. It is nothing more than an
and C3 for any frequency from 10 Hz to 10 astable multivibrator, triggered by a one-shot.
kilohertz.
When the push -to-talk switch is closed, actuating
the transmitter relay, K1, Ql goes from saturation
ped with first-class gear and antenna, the potential
to cutoff, and the multivibrator, Q2 -Q3, begins
for interference with other stations is great. Re- oscillating with a period dependent on the values
peater owners should assure that stations on the of R3, R5, C2, and C3. Values given result in a
input frequency, but not wishing to use the "whistle" of roughly 650 Hz.
repeater, do not key the machine. A simple form
of tone coding may be employed so that distant Oscillation ceases when Ql turns on again. This
is regulated by the values of R2 and Cl, and is
stations do not inadvertently turn on the repeater roughly 0.25 second with the values shown. The
during band openings. Provision should also be 470-ohm resistor, Rl, protects the base of Ql from
included so that badly over -deviated or off- current surges when the PTT switch is released.
frequency signals will not operate the machine.
Suitably devices which can be used to achieve these Most of the component values are not critical,
objectives are described below. except the RC products which determine timing.
Since the frequency is low, almost any bipolar
Tone Control transistors can be used. Npn types are shown, but
pnp will work with opposite voltage polarity. The
Many repeaters use a form of tone control so beta rating should be at least twice R3/R4, to
that a carrier on the input frequency will not assure saturation.
inadvertently key the transmitter. The most popu-
lar form of tone control ís known as tone burst, Tone Generator
often called "whistle on," because an operator For single -tone control, each station needs a
simple, stable, audio oscillator. The design for such
an audio oscillator was drawn up based on articles
+6V
written by Maynard, "Twin-T Oscillator," Elec-
TONE tronics World, August, 1968, and Antanaitis, "A
DECODER Simple Two-Transistor A.F.S. K. Generator," QST,
September, 1969. The criteria for the oscillator
were that it should be small, and that a number of
them could be interchanged quickly from a given
connector. The audio oscillator was built on a
circuit board designed so that it could be plugged
into an Amphenol 143-006-01 circuit -board socket
(see Fig. 12-18).
Ki B T.
EXWr
._---: The circuit is shown in Fig. 12-19, and the
attached nomogram is used to help choose the
design frequency which depends upon the chosen
Fig. 12-20
112 -watt
-
Tone -burst decoder. Resistors are values of Cl, C2, and C3. Several transistors were
composition and capacitors are mylar. K1 tried and the choice of transistor did not seem
is an spst reed relay with a 6-volt coil (C. P.
Clare critical. Of those that were tried, the 2N339
PRA -2010).
appeared to be a good substitute for the MPS6521.
The Repeater - A Good Neighbor 271
2143644

RI
TO
DISCRIMINATOR 470

Fig. 12-22 -Circuit diagram and parts Information

for the 30 -kHz window for vhf fm repeaters.
Capacitor values are in microfarads (µF1. Polarity
marking indicates electrolytic. Parts not described
below are marked for identification in text.
CR1-CR4, incl. -1-A rectifier (Motorola 1N4001).
K1 - Miniature reed relay (Dunco MRR-1 B -M) 12
volts dc, 576-ohm coil.
R2 -250,000 -ohm miniature trimmer -type con-
trol, linear taper (Mallory MTC-4).
Several audio units using the circuit of Fig.
12-19 have been built. The frequency of each was
checked with a frequency counter. It was found
that their frequencies would vary approximately
0.1 Hz around the design frequency. If RI were window, as seen in Fig. 12-21. As modulation is
replaced with a 50,000 -ohm control, the oscillator applied the carrier deviates around the center
would tune as much as 1500 Hz from the design frequency in a symmetrical manner. If the devia-
frequency. tion is too wide, distortion results. If the center
frequency is not in the center of the window, the
Tone Decoder signal is also distorted. Both effects are all too
Most narrow -bandwidth tone decoders cur- familiar to operators of fm repeaters.
rently used in amateur repeater and remote -station A repeater turns on whenever a signal is of
applications employ several bulky LC circuits to sufficient strength to cause quieting of its receiver,
achieve the required audio selectivity. The phase- thus some way
of turning off the receiver is needed
that ís more than 3
locked loop (PLL) ICs, pioneered by Signetics, whenever it picks up a signal
have simplified the design and reduced the size of kHz off
the intended center frequency, or is
the bandwidth of the repeater
tone decoders so that a complete Touch -Tone deviating more than
x -1/2 -inch receiver. A simple solution to this problem is
demodulator can be built on a 3 5
shown schematically in Fig. 12-22. Its operation is
etched circuit board (about the size for a single -
tone decoder using LC components). based on the fact that the discriminator current in
a signal is centered in
A typical PLL single -tone decoder, such as an fm receiver is zero when
for tone-burst entry control at the receiver passband, but rises if the signal is
might be employed
a repeater, is shown in Fig. 12-20. One RC network
mistuned. All that is needed, then, is a dc amplifier
current to a level that
establishes the frequency to which the PLL is to raise the discriminator
tuned, according to the relationship:
will operate a relay to cut off the repeater receiver.
Observations at the repeater site showed that,
1 in a Motorola Sensicon receiver, nearly all inter-
frequency - ference came from signals that produced 10 micro-
R1C1
amperes or more of discriminator current, so the
The PLL, a Signetics NE567, may be operated circuit was set up to respond to this value of
from 0.1 Hz to 500 kHz. C2 establishes the current. With no signal, there is no discriminator
bandwidth of the decoder, which can be set current. There is still no current when a signal
between one and fourteen percent of the operating centered in the passband is received. When the
frequency. C3 smooths the output signal, and, signal is modulated, the output of the discrimina -
when this capacitor is made a high value, provides a
delay in the turn -on function when a tone is +¡2v
received. Up to 100 mA may be drawn by the '567
output circuit, enough to key a relay directly or to
drive TTL logic. The PLL contains 62 transistors.

Off -Frequency Controller

Two common causes of unsatisfactory quality
of voice transmissions through a vhf fm repeater
are excessive deviation and off-frequency trans-
mission. The simple device shown in Fig. 12-22 will
cut off a repeater receiver if signals having either of Fig. 12-23- Circuit diagram of the oscillator
these undesired characteristics are fed into it. which is used to provide an audible indication of
The reader can think of this device as a off-frequency operation. Y1 Is chosen for the
"window." Any signal outside the window no center of the receiver i -f passband, and it may be
longer activates the receiver. The transmitter car- set to the exact frequency desired by means of C1,
rier frequency should be at the center of this a mica -insulated trimmer.
 272 REPEATERS - THEORY AND PRACTICE
tor is audio, around a zero dc level, if the signal is level which will energize the relay, while the gain
on frequency. The filter, R1/C1, removes audio of the circuit remains constant.
above 15 Hz from the cut-off circuit, without It must be realized that R1 and Cl do not make
affecting the operation of the receiver and the a perfect audio filter. Some audio leaks through,
system. and the sum of off-frequency current and the
When an off-frequency signal is received that is attenuated audio causes unsymmetrical drive to the
within the receiver passband, current is produced circuit and actuates the relay. The final factor to
in the discriminator in proportion to the amount add in is that modulation in excess of normal
that the signal is off-frequency. If the current is (usually between 5- and 15 -kHz deviation) causes
positive it turns on the first transistor, Ql. This the discriminator to become overloaded, and its
turns on Q4, whose emitter current then actuates output "kicks." This kick is added to the above
the normally closed relay, K1. These contacts are factors, and further encourages operation of the
in series with the repeater transmitter control. relay, to prevent over deviation of the repeater
When they open, the energizing signal for the transmitter.
transmitter is removed, and after the delay built The operation of the transistor circuitry is very
into the transmitter is passed, the system will drop. fast, and the relay is a miniature reed type, with an
If the transmitter was not energiznd to start with, operating time of one or two milliseconds, so
the open contacts will prevent it from starting up. circuit response is almost instantaneous. At the
Thus the system prevents an off -frequency signal most, less than a second of offending signal will be
from starting the repeater, or will cut it off if an heard through the repeater.
off-frequency signal takes over the receiver. A simple gadget at the repeater will allow users
If the signal is off-frequency in the other to adjust their transmitters to the input frequency
direction, the discriminator current is negative, and of the repeater. It consists of a crystal -controlled
transistor Q2 will be turned on. This turns on Q4 oscillator set for the center of the i -f passband of
as before and operates the relay to shut down the the receiver. The circuit diagram is shown in Fig.
system. The selection of the three resistors, R3, 12-23. During a "netting party," one of the
R4, and R5, will balance the gain so that a given repeater control stations turns on the oscillator by
distance off frequency in either direction will means of the control link. Then, by listening to the
produce the same discriminator current. The value beat note transmitted by the repeater, an operator
of R1 and the setting of R2 serve to control the can "zero" his transmitter. Of course, no modula-
drive signal to the circuit, and thus set the current tion should be used on the transmitter being
adjusted.
AUTOPATCH A ND TOUCH TONE
PROTECTIVE CONNECTION functions, as well as telephone dialing. Because all
EQUIPMENT BLOCK
of the Touch-Tone frequencies are within the voice
TRANSMISSION
band, they can be transmitted by any amateur
CIRCUIT, 2 -WAY voice transmitter.
Z-60011 The Touch -Tone control system consists of
LINE TO pairs of tones (see Fig. 12-25) for each of 10
TEL. CO. TOLL CALL¡ AVAILABLE
OFFICE INDICATOR ONLY WITH numbers and the two special functions. One tone
RELAY
5830
CI RCUIT PBX TRUNKS from the high -frequency group is generated simul-
24v iF rOHMS
OFF HOOK taneously with one tone from the low-frequency
group to represent each number or function. The
( DIAL PULSING Touch -Tone generator pad from a standard tele-
117V AC phone instrument is usually employed. See Fig.

LOCAL
TEL. CO
PROVIDED
-INCOMING (RINGING)
SIGNAL INDICATION
AMATEUR PROVIDED
12-26 for connections.
For more information on the, requirements for
and construction of phone patches, review the
following articles in QST: "Phone Patching -
Legitimately," March, 1969; "Legalize Your Phone
GROUND
{lndirntsc a normally cpen. contact Patch," May, 1969; "Phone Patching - One Year
Later," November, 1970; "An Improved Phone
Fig. 12-24 -
Typical connections for a repeater
and an interface device provided by the telephone
Patch," Hints and Kinks, November, 1970; and
"Phone Patching and the Telephone Network,"
company. May, 1971.
Some repeater groups have provided an inter-
connection to the public telephone network Low High Tone
through a device called an autopatch. Details on all Tone (Hz) 1209 Hz 1336 Hz 1447 Hz
phases of phone patching are contained in Chapter 697 1 2 3
15 of The Radio Amateur's Handbook. Typical 770 4 5 6
interconnections to a telephone company interface 852 7 8 9
device for an unattended repeater axe shown in Fig.
12-24. Such interconnection has led to the wide-
941 * 0 #
spread use of the telephone company's Touch -
Tone system of tone signaling for repeater control
Fig. 12-25 -Standard Touch-Tone frequencies for
the 12-digit pad.
Autopatch and Touch Tone 273
Fig. 12-26 - Typical connections to use a Touch - TOUCH TONE
Tone pad for repeater control. Resistances are in P AD

ohms. R1 is a linear -taper composition control and

J1 is a panel -mounted phono jack. Capacitors are
electrolytic; color coding on the wire leads from
the pad ís shown in parentheses.
A Touch -Tone Encoder
To generate the Touch -Tone codes listed in Fig. <
Ji
12-25, two special integrated -circuit function gen-
erators may be used. The ICs are Signetics NE566s,
voltage -controlled oscillators that were outgrowths
of phase -lock -loop technology. The circuit shown
here was developed by Jim Wyland at Signetics. Fig. 12-27 -
An L-
This design, Fig. 12-28, provides 11 number codes. shaped piece of alumi-
If the additional frequency of the 12 -button num, bolted to the top
telephone pad, or the five additional frequencies of of the utility box, pro-
the 16 -button generator are needed, one can add vides a convenient un-
appropriate decoding diodes and timing resistors. der -the -dash mount for
a mobile installation of
The operating frequency of an NE566 is approx- the Touch -Tone pad.
imately:
2 V8- V5
Fo x
RIC1 V8-VI
where V8, V5, and V1 are the voltages at pins 8, 5,
and 1, respectively. diodes is required to prevent one of the output
In the practical circuit shown in Fig. 12-28, the tones from varying slightly in frequency. A diode
value of the timing resistor (Rl) is selected to package such as the Signetics NE301A provides a
generate the highest frequency desired. Additional suitable set of matched diodes. Alternatively, a
resistors are switched in to lower the frequency of keyboard with double -pole contacts or a design
oscillation to the other Touch-Tone frequencies. such as the Chromerics EF-20271, which provides
Both oscillators use a .022-µF capacitor (C2) for the matrix necessary for Touch-Tone coding (thus
timing. Calibration is needed only at one frequency eliminating the need for diodes), may be used.
for each oscillator since the other three tones are
set by the resistor ratios used. A Touch -Tone Decoder
If a single -contact push-button switch is used,
the diodes shown in Fig. 12-28 will be necessary. The phase -locked loop described earlier in Fig.
Close matching of the forward -voltage drops of the 12-20 may be used as the basis of a Touch-Tone

; OUTPUT

Fig. 12-28 -Schematic diagram

of the tone generator. All resistors
.022

9100
are 1/2 -watt composition. Capaci-
tors are mylar.
BT1 --9-V miniature Battery.
Linear-taper, 1/2 -watt 768
CAL.
1000
R1,R2
composition control. T%' 1200

S1 - Chromerics EF-20457 key- 135óH,

board (available from Chro- 852 Hz
576
merics, 77 Dragon Court, 1447Hz \ L1

Woburn, MA 01801). Note:

The Chromerics EF-20271
o-f: 6800
74-
U2
NE566V
4

1.6K
Touch -Tone encoded key- 5%
board may also be used, elimi- 0-41
nating the need for all 1N914
diodes.
U1,U2 - Signetics IC (available
941 Hz

from Compar, Inc., 2537

Whitney Ave., Hamden, CT o-{ y
06518).
 274
TONE DECODERS
697 Hz

U1

770 Hz

U2
'f
5

135ic1-
R1

15-11.114

R2
NE567

E567
N
PLL
PLL

6
6V

11<

2,u
s
)
a
t
41

L+-
+1

,4-, 6V
6Y
8

8
+5V

2200

+51/

2200

""-
zLn,,.4?00

45V
REPEATERS - THEORY AND PRACTICE

TO PIN 14 OF
U8,U9AND U10
NOR GATES

12
1 U8A

U8B

10

13
01

02

-0

04
3
Fig.

1
1 2-30
Touch -Tone decod-
er. Resistors are
/ 2 -watt composi-
tion and capacitors
are mylar. The val-
ues indicated for
R1 -R7, incl., are ap-
proximate. Final ad-
justment of the fre-
quency of each PLL
should be made by
trimming the resis-
tor values as needed
to center the phase -
locked loop on the
to e phone -company
-

852 Hz I

frequency. A fre-
mrJj NE567 2200 U9A b' 0 5 quency counter fa-
cilitates the adjust-
31 PLL 8
tl5
4700 ment procedure.
R3 6 14
AUDIO
U3 Z
U1 -U7 incl., are Sig-
INPUT .47 56 U9B 4 b 6
netics NE567Vs and
12KM 6V
l
2/4F U8 -U10, incl., are
(200mnV) +5V
941 Hz 6V 4l Motorola MC7402Ls
9 U9C 0 7 (2 NOR gates are
N E567 2200 not used).
PLL b

vvv--
3

U4 5 R4
6 +11
U9D p13 08
1F
12

10.8K Tt^¡1
T'/
F
r},6/ 2
1204 Hz m 6V +5V
3 U10A o 9
220
NE567 0
72.1 PLL 8 5

U5 s R5 +114700 6 U108ó' o 0

8600
1

+5V
1336 Hz 26VF

NE567 2200
fi 3
PLL 8

U6 4700
5 R6 6 1l+
1
7500W:14
1477Hz "11-W +5V
Fig. 12-29 -
The homemade Touch -Tone gener-
ator is housed in a plastic box. The keyboard is a
41
Chromerics EF-20457 which has no moving parts.

, .j
NE567 2200
Electrical contact is made by chemical means.
3 PLL e

U7 s
R7
+ 250,2F decoder. Seven PLLs are employed in the circuit
6800 ./ .h 6V T 6V shown at Fig. 12-30. The resistor values indicated
.i, 2e in the diagram for R1 through R7 are approximate;
the exact value required for the standard Touch-
Tone frequencies (Fig. 12-25) will depend on the
exact value of the 0.1-µF capacitor. When a pair of
tones arrives at the input of the decoder, 697 and
1207 Hz for example, after several cycles of audio
7 have been received, PLLs Ul and US will activate,
changing from 5 volts to about 0.4 volt at the
-;t;a? `,. output pins. These two "lows" will cause NOR
gate U8A to change to a "high" (about 5 volts) at
its output, the digit output line of the number
represented by the two audio tones.

Fig. 12-31
5 -volt
-
The Touch -Tone decoder, including a
power supply and a solid-state numerical
display which is driven by the digit output lines
through a diode matrix. As numbers are received
and decoded, they are displayed, providing a visual
check on the operation of the decoder.
 Chapter 13

U H F and Microwaves
Segments of the radio spectrum have labels. new, or at least out of the ordinary. For all our
From 3 to 30 MHz is "high frequency," 30 to 300 emphasis on friendships made and maintained by
is "very high," 300 to 3000 "ultra high," and all radio, a strange voice, a new country, or even a
higher "super high." But all through this book new state stirs us. For reliable communication we
we've had trouble keeping within these semantic tend to rely on the telephone, or the local vhf
lines. This trouble with labels is, in a way, a capsule repeater. Worth for point-to-point has never done
history of the radio art. Today's "ultra" is tomor- too well as a sales argument for new bands.
row's commonplace. But much communication above 1000 MHz
Webster defines ultra as "Going beyond others, must, of necessity, be largely on an organized basis.
or beyond due limit," yet within this writer's Bands are incredibly wide. Beam patterns must be
memory everything above 30 MHz was called sharp, if communication over interesting distances
"ultra high." Our QST column started as "On The is to be maintained. Cooperation in the matter of
Ultra Highs" in 1939, but it wasn't long before we operating times, frequencies, and beam headings is
began to think of frequencies up to at least 220 the only alternative to spending fruitless hours
MHz as something rather below the "ultra" class. scanning the radio horizon for signs of life.
Today we tend to take the 420 -MHz band out of Routine "activity" as we know it on lower bands
this category. seems thus unlikely ever to develop in our micro-
Working even above 1000 MHz is today ap- wave bands, at least until we have space well filled
proaching a routine business, but it is well to with communications satellites, and frequencies for
remember that the techniques we think of as them are sure to be in short supply.
"conventional" are that way because of continuing We already have the know-how to transfer to
advances, amateur and commercial. We tend now the uhf and microwave bands much of the talking
to put the frontier somewhere around 1000 MHz. we do on all-too -crowded lower frequencies. Will a
Who can say where it will be tomorrow? new generation of amateurs seize upon the oppor-
Any frequency "frontier" is more philosophical tunities that the microwaves offer? It could be that
than technical, challenging basic concepts of ama- the future of our avocation hinges on a positive
teur radio communication. Throughout our his- answer.
tory, amateurs have cherished the element of Equipment and methods are already well within
surprise. We call CQ to see who will come back. We our capabilities. What we need most, in the world
listen, listen - everlastingly listen - for something above 420 MHz,'is people!

UHF LINES A ND CIRCUITS

The changing nature of tuned circuits as we and outer conductor can be moved up or down the
move higher in frequency was discussed at length line, as shown at the bottom of the sketch.
in Chapter 5. It may be well to review this material The line may be tuned with a variable capa-
before going too far into uhf circuitry. Earlier we citor, as in 'B. Adding capacitance in a tuning
were concerned with "lumped -constant" circuits; device, or in the form of the input or output
those where inductance and capacitance exist as
separate entities. In uhf work we have to think in
terms of sections of transmission line.
Still higher in frequency we will abandon the
line idea. Our circuits become resonant cavities,
and an allied change occurs in methods used to
transfer power from one circuit to another, and to
an antenna. The parallel -conductor line disappears,
giving way first to low -loss coax, and eventually to
waveguide.
In Fig. 13-1 we have typical coaxial circuits,
shorted at one end and open at the other. The line
need not be cylindrical. A rectangular tube and a
flat inner conductor (strip -line) work equally well. (A)
Such a line section an electrical quarter-wavelength (B)
long can replace a coil -and -capacitor circuit as the Fig. 13-1 -A coaxial tank circuit can be tuned by
tuning element of an oscillator, amplifier, or other means of a "false bottom," as at A,or by variable
uhf device. Its resonant frequency can be varied in capacitance across the open end, as in B. In A the
several ways. Its length can be adjusted, if two effective length is adjusted by means of the
pieces of tubing are made to telescope one inside shorting disk, which is movable. Good electrical
the other, as in the upper portion of circuit A. A contact to both inner and outer conductors is
movable disk making firm contact to both inner important at this point.

275
 276 UHF AND MICROWAVES
E9 wave exists along a transmission line, the rf voltage
and impedance are repeated every half wavelength.
If tube capacitance and lead inductance tend to
(A) make us "run out of tank circuit" with a quarter-
wave line, and a half -wave line is not convenient

I -4`
Max,
y 4

Max
for our purposes, we can make the line any odd
number of quarter wavelengths, as for example the
3/4 -wave line of B. This may have a quarter -wave
resonance lower in frequency, but because of the
E9 different loading effect of tube and circuit capa-
Mix citance at the two frequencies it will not be exactly
one-third that of the 314 -wave mode.
(B) L Coaxial and Strip -Line Circuits
x a x
There is no need for the conductors to be
round in cross-section, or truly coaxial in nature.
Fig. 13-2 -
The useful frequency limit of a The strip line is often convenient for the amateur
line -type tank circuit can be extended by making it
a half -wave line, A, or three-quarter -wave line, B.
Rf voltage distribution along the lines is shown by V
curved line, Eg.
AAA
`
A A II

capacitance of a transistor or tube, lowers the 9

resonant frequency for a given length of line. The
line must, therefore, be made physically shorter
than a quarter wavelength. This loading limits the 8 { I

frequency over which a quarter-wave line can be

used, just as in circuits discussed in Chapter 5. 7

The same steps can be taken to extend the

frequency limit. A half-wave line is shown schema- 6
tically in Fig. 13-2A. A push-pull version could
have a tube at each end. Rf voltage distribution
along the line is shown above it. Where a standing 5
u C=2.$C=o/f
I

`+
C -20 wO

90 4
e - ---- --- ---
60
70-
60
so
40-
30
a

b
3

pliall1111
1

; - C=Of

.2
20 1

)
MN:
I
= -
-_
C=5/
wOp.
100 200; 300 400 1 500 600 700 ...r SCI LEA
800 900 1000 1100 }200 LE B
log FREQUENCY IN MC.

7
Fig. 13-4 -
Graph for determining the length of a
capacity -loaded quarter -wave coaxial line of 71
6
ohms impedance, for frequencies from 150 to
5 1300 MHz. The value C includes tube output and
4 tuning capacitances.
3
builder, and many 'examples appear in this book.
2
Other shapes could be used, but coaxial or strip
lines are most common.
The Q of these circuits is nearly always impor-
tant, so the conductors should be of large size, and
of metals having high conductivity. A coaxial line
00
0 WO ISO 20O 2S0 St with a No. 20 wire inner conductor would be little
CHARACTERISTIC IMPEDANCE. 20 better than a coil of the same wire size, for its
Fig. 13-3 - Characteristic impedance of coaxial
lines for various conductor diameter ratios. The
ohmic resistance would be as high as the same wire
wound into a coil. Electrical conductivity is par-
outside diameter of the inner conductor and the ticularly important at points of high rf current
inside diameter of the outer conductor are used. (lowest rf voltage), notably at the shorted end of
Waveguides and Cavity Resonators 277
coaxial or strip-line circuits. Insulation should be ance can be obtained to an accuracy sufficient for
kept to a minimum, and preferably avoided en- most purposes from Fig. 13-3.1 The impedance of
tirely at or near the points of high rf voltage. a strip line can be obtained from the formula:
Insulation loss is introduced by a tuning capa- S
citor, as it must be at a point of appreciable rf Zo =377
W
voltage. Movable -disk capacitors are favored, as
they do not require insulating supports or metal where S is the spacing between the strip and the
frames that often Introduce parasitic resonances. outer conductor, and W is the width of the strip.
Impedance of the line may be important in Preferably W should be several times S. This
some applications. This can be obtained for coaxial information is from Brayley.2
lines from the formula: The same author gave a formula for figuring the
b length of coaxial tank circuits when the total
Zo = 138 log circuit capacitance is known. Solution of this
a formula for various capacitances, at frequencies
where Zo is the impedance of the line, b is the from 150 to 1300 MHz, was worked out graphi-
inside diameter of the outer conductor, and a is the cally by Garrett and Manly, as shown in Fig. 13-4.3
outside diameter of the inner conductor. Knowing The top curve in each set (C equals zero pF) is for
the dimensions of available materials, the imped- lines not capacitively loaded.

WAVEGUIDES AND C AVITY RESONATORS

Because of the loading effects of capacitance on Waveguide can be any cross-section, but only
parallel-line and coaxial circuits the useful fre- rectangular and circular are common. In the
quency limit for any type of line circuit is rectangular, Fig. 13-5, the width, x, is the critical
somewhere around 1000 to 2500 MHz, depending dimension. It must be more than a half wavelength
on the tube or solid-state device used. At higher for the lowest frequency to be transmitted. Gener-
frequencies something different is needed, for ally the height, y, is made about a half wavelength.
transmission lines as well, for losses in even the It can be seen that waveguide has another advan-
best coaxial line become prohibitive. The answer tage over other kinds of line, in addition to lower
lies in the use of cavity resonators and waveguides; losses; by its very nature it can be an effective
devices in which the dimensions of conducting high-pass filter.
boxes or tubes determine the frequency at which Five factors should be kept in mind in dealing
they will operate. with waveguide dimensions: free -space wavelength,
A; guide wavelength, Ag, the actual length of the
Waveguide Principles wave as it travels through the guide; cut-off
A waveguide is a metallic tube of circular or wavelength, 4, longest usable wavelength, Au, that
rectangular cross-section, through which electro- can be transmitted without excessive attenuation;
magnetic waves can be transmitted. The walls of and the shortest wavelength, AS, that can be
the waveguide are not considered as carrying transmitted before the next mode develops. These
current, in the sense of the conductors of a 2 -wire are obtained, for the dominant mode, as shown
line, but rather as a boundary confining the waves below:
to the enclosed space. Energy injected at one end 11811
by capacitive or inductive coupling, or by radia- A., inches

tion, flows through the guide to the load, by means Freq., MHz
of reflections from its inner walls. A
There is an infinite number of ways in which Ag, inches
the electric and magnetic fields can arrange them-
selves in a waveguide, depending on guide dimen-
sions in wavelengths. These modes are separated
into two groups: transverse magnetic (TM), and Rectangular Circular
transverse electric (TE). The mode is identified by
these letters, followed by subscript numerals, as 3.41 r
TEL p, TM1,1 etc. The number of possible modes 1.6x 3.2r
increases with frequency, for a given size wave - 1.1s 2.8r
guide. The dominant mode (the only one for the
lowest frequency the guide will pass) is generally Typical Inside dimensions of rectangular wave -
used in practical work, as there is little point in guide for the various amateur bands are as follows:
using a larger guide than necessary for a given 1For this and other numbered references, see
frequency. bibliography at the end of this chapter.

Fig. 13-5 - Waveguide cross-section can

be rectangular or circular. In the rect-
angular, the width, X is the critical J_
dimension. Cutoff wavelength is 2x or
3.41 r.
 278 UHF AND MICROWAVES
3500

3000

Square Box Cylinder Sphere

Fig. 13-6 - Forms of cavity resonators. The square 2500
and cylindrical types are merely closed -off sections
of waveguide.
2000
2300 MHz -
1.34 by 2.84 inches; 3300 MHz -
same; 5650 MHz -0.622 by 1.372 inches; 10,000
MHz - 0.375 by 0.75 inch; 21,000 MHz 0.17 by - 1500
0.42 inch. These are standard waveguide sizes. If
you were to make your own, the dimensions would
not have to be this precise.
1000

The Cavity Resonator

Now suppose that we cut off a section of 500
waveguide, and seal off the ends. We then have 0.500 1.000 1.500 2.000 2.500
what is known as a cavity resonator. The term RADIUS IN INCHES
"cavity" is frequently applied to coaxial or strip-
line circuits that are completely enclosed, but the
Fig. 13-8 -
Radius of a cylindrical cavity for a
2C39 tube, for frequencies from 500 to 3500 MHz.
name should be reserved for resonant boxes with From information by Ramo and Whinnery, Fields
no inner conductors, as shown in Fig. 13-6. The and Waves in Modem Radio.
resonant frequency of these typical cavity shapes
depends on the inside dimensions of the box and A common form of cavity is the re-entrant
the mode of oscillation. The latter is comparable to cylinder type of Fig. 13-7. It resembles a coaxial
transmission modes in waveguide. For the lowest line, with both ends closed and capacitive loading
frequency mode, the resonant wavelengths are as at the top, but the mode of oscillation varies
follows: considerably from that in coaxial circuits. The
resonant frequency of the re-entrant cylinder
Square box: 1.41 1 cavity depends on the diameters of the two
Cylinder: 2.61 r cylinders, and the distance, d, between the cylinder
Sphere: 2.28 r ends.
A tube commonly used in amateur work in the
The resonant wavelength of the cylinder or box
uhf region is the 2C39A or 7289/3CX100A5. Fig.
is independent of the height, when this is less than 13-8 makes possible a quick estimate of the size of
half a wavelength. In other modes of oscillation
a cylindrical cavity for this tube. The graph form
the height must be a multiple of a half-wavelength,
was supplied by Garrett and Manly.3

d Coupling to Waveguides and Cavity

Resonators
Energy may be introduced into or extracted
from a waveguide or resonator by means of either
Cross-sectional view the electric or the magnetic field. The energy
Fig. 13-7 - Re-entrant cylindrical cavity resonator. transfer frequently is through a coaxial line, two
as measured inside the cavity. A cylindrical cavity methods for coupling to which are shown in Fig.
can be tuned by means of an adjustable false 13-9. The probe shown at A is simply a short
bottom, when operating in such a mode. Other extension of the inner conductor of the coaxial
tuning methods include placing adjustable tuning line, so oriented that it is parallel to the electric
paddles or slugs inside the cavity, so that the lines of force. The loop shown at B is arranged so
standing -wave pattern of the electric and magnetic that it encloses some of the magnetic lines of force.
fields can be altered. The point at which maximum coupling will be
Just as coaxial lines represented improvement secured depends upon the particular mode of
in Q over coil -and -capacitor circuits near the upper propagation in the guide or cavity; the coupling
limit of the useful frequency range of the latter,
the resonant cavity is a means of obtaining very
much better circuit efficiency at frequencies where
the coaxial circuit begins to be inefficient or
impractical. With care in the construction of a
cavity as to silver plating and the best possible
electrical conductivity where surfaces join, a circuit
Q of several thousand is possible. A Q of 1000 or (A) 03)
more is readily obtainable. Fig. 13-9 - Coupling to waveguides and resonators.
Transistors and Tubes for the Higher Frequencies 279
will be maximum when the coupling device is in perpendicular to the electric lines the coupling will
the most intense field. be minimum; similarly, when the plane of the loop
Coupling can be varied by turning the probe or is parallel to the magnetic lines the coupling will
loop through a 90 -degree angle. When the probe is have its minimum value.

TRANSISTORS AND TUBES FOR THE HIGHER FREQUENCIES

Even with the best possible circuits, the upper Conventional Uhf Transmitting Tubes
useful limit of frequency is determined finally by Several glass-envelope transmitting tubes work
the characteristics of the devices connected to moderately well in the 420 -MHz band. The 6939 is
them. Capacitance between elements and induc- a small dual tetrode much like the 6360, but with
tance of leads brought out from them load down more compact design to extend the upper frequen-
the circuit in the same way as wires and capacitors cy limit for uhf service. It will deliver about 6
connected externally. In the tube, there is the watts output. The 6524 and 6252 are dual tetrodes
seemingly Infinitesimal time required for passage of with plate leads brought out through the top of the
electrons between cathode and plate. This may be envelope. They are capable of around 15 watts
only 0.001 microsecond, and of no importance at output in the 420-MHz band. The 5894 is a larger
1000 kHz, but It is the equivalent of a full version of the 6252, widely used some years ago
oscillation cycle at 1000 MHz. These facts set for 420 -MHz work. It will give up to 40 watts
something around 3000 MHz as the practical output at 432 MHz. Any of these tubes operated as
frequency ceiling for negative-grid tubes, and 1000 a tripler from 144 MHz will drive another as a
MHz for inexpensive transistors. straight -through amplifier, and they can be used
Only specially designed tube types will go with more -or-less conventional circuitry.
anything like this high. At 420 MHz we have left The best transmitting tubes for use at high
all but a very few types behind, and at 1215 MHz amateur power levels in the 420-MHz band are the
no tube commonly used on lower bands is con- external -anode types beginning with the 4X150A,
sidered. and continuing through later versions such as the
Solid-state techniques for generating power all 4CX250B and R, and the 4CX300A. These tubes
through the uhf and microwave spectrum are are well adapted to coaxial- and strip-line circuitry.
available, but suitable devices are costly at present. They require forced-air cooling. A conduction -
Mass applications are lowering manufacturing cooled version, the 8072, has considerable appeal
costs, and power transistors and other devices for for medium -power applications where use of a
uhf and microwave communication should be more blower is not convenient. Many other types of this
within our financial reach as markets widen. general tube family are available, but the ones
Uhf Receiving Techniques mentioned above are most often seen in amateur
uhf circles. All require special sockets, which may
In receiving, where power -handling capability is include built-in bypassing, where desired.
not a primary consideration, the useful frequency Tubes generally available to amateurs for trans-
limit is somewhat higher than in transmitting, for mitting use above 1000 MHz include the "light-
more -or -less conventional designs. Probably the house" triode, Fig. 13-10. This old but still useful
first mass-produced uhf tube was the "acorn," a type can be picked up inexpensively on the surplus
small triode with leads brought out radially
market. It comes in several styles, of which the
through the glass envelope. In suitable circuits it
2C40 is the most plentiful. The "pencil triode"
will oscillate up to around 500 MHz.
series of tubes, not widely used by amateurs, work
More recently, the Nuvistor carried reduction
up to about 2000 MHz. The 2C39 planar triode,
of electrode size and lead length close to the
and later versions such as the 7289/3CX100A5,
practical minimum. The inexpensive 6CW4 and
6DS4 work well in our 420-MHz band. The
double-ended 8058, made for grounded -grid ampli-
fier service, does even better.
Very high-transconductance planar triodes Plate
(416B, 7768, and others) are the best rf amplifier Gna
tubes available for 420-MHz service, but they have
been replaced almost entirely by moderately priced cathode
transistors, which work better and are very easy to
use. H FLatl,,ade
Actually there is little point in attempting to connection
use vacuum tubes for receiving above 400 MHz, for Neater
even the best are incapable of satisfactory noise
D.0 CatbaCa
figure, compared with a well-designed transistor connection
front end. Even at 432 MHz a good crystal mixer
followed by a low -noise i -f amplifier will do just
about as well as the best vacuum -tube designs.
Fig. 13-10 - Sectional view of the "lighthouse"
tube. Electrode spacing and lead inductance are
Transistor if amplifiers are taking over the re- held to near the practical minimum, making the
ceiving job in amateur uhf circles. tube useful in the uhf range.
 280 UHF AND MICROWAVES
FOCUSING quencies from a few hundred megahertz to many
KV ELECTRODE
gigaherts, and for power levels from milliwatts to
CATHODE)
J) ELECTRON GUN
megawatts.
ANODE
The Klystron Amplifier
RESONANT
INPUT
CAVITY
r_ f R.F INPUT Referring to Fig. 13-11, the essential parts of an
amplifier klystron are the electron gun, the drift
DRIFT TUBE tube, the resonant cavities, and the collector. The
ELECTRON BEAM cathode, anode, and focussing electrodes of the
RESONANT gun form an electron beam. Current flows as in a
OUTPUT diode, when high voltage is applied to the anode.
CAVITY
The gun electrodes focus the accelerated electrons
through a hole in the center of the anode, into a
cylindrical beam which flows at constant velocity
Fig. 13-11 - Essential parts of the klystron tube,
shown in simplified form to illustrate the theory of
through the hollow drift tube and resonant cavi-
ties, to the collector.
"bunching." Rf drive applied to the input resonant cavity
causes an rf voltage to exist between adjacent
sections of the drift tube. (Spaces between these
can be used up to around 3000 MHz in suitable
sections are called interaction gaps.) Electrons
circuits. The 2C39 tnpler to the 1215 -MHz band, flowing past the input gap are affected by the rf
described in QST by W6DQJ,4 is still widely voltage across it. When the voltage is positive (in
copied. WB6IOM described a 2 -tube amplifrer.5 the direction of electron flow) electrons in the gap
The 41611, more familiar to amateurs in receiving are accelerated slightly. During the alternate half of
applications, is capable of operation up through the rf cycle they are retarded. Some of the
the 3300-MHz band in suitable circuits. electrons in the beam are thus moving faster and
Even these uhf types work poorly above our
2300-MHz band. In our 3300-MHz band and at all
some are moving slower than the average rate the
beam is velocity modulated.
-
higher frequencies the negative -grid tube gives way As the beam moves toward the output cavity,
to specialized types such as the klystron, travelling- the fast electrons tend to catch up with the slower
wave tube, and magnetron. Costly, except as
ones, by a process called bunching. When the
surplus, even these types are being outmoded by
bunches reach the output cavity gap they are well
solid-state developments.
formed, and they charge the output resonant
VELOCITY MODULATION - / circuit as sharp pulses of rf current. The cavity acts
as a resonant coupling device, and the power is fed
THE KLYSTRON to the transmission line, and on to the antenna.
The action is not unlike that in a conventional
At some point in the uhf range, just where tetrode at lower frequencies, the essential differ-
depending on power level and application, transit - ence being that in the klystron the electron
time effects render conventional tubes unusable. bunches are not formed by a control grid, but
Electron flow from cathode to anode, in tubes rather by this velocity modulation of a continuous
we're familiar with, is in the form of short bursts, electron beam.
regulated by the rf charge on the grid. When transit There is more to a klystron amplifier than this,
time is an appreciable part of the rf cycle these and some of it is suggested in the simplified
pulses become poorly defined, and performance schematic of the klystron dc circuits, Fig. 13-12.
falls off. Most power klystrons have 3 or more cavities,
Of the many devices developed to get around because multiple -cavity designs provide higher gain
this problem, the power klystron6 is probably the and efficiency than the simple two -cavity type
most interesting to the amateur. The klystron is described. Klystrons have been built with six or
complex and costly, and amateurs know it mainly more cavities, capable of power gain in excess of
as an occasional surplus market item, but if we are 90 dB. Amplifier klystrons have not been used
to do much with our assignments in the micro- extensively in amateur uhf communication to date,
wave region some understanding of its operation is their principal employment having been in
almost mandatory. In one form or another, it has 1296 -MHz moonbounce experiments,7 where high
been used in nearly all amateur work above 3000 power and extreme stability are mandatory.
MHz.
The klystron uses the phenomenon of transit The Klystron Oscillator
time to advantage, through a technique known as Most amateur work with klystrons has been
velocity modulation. It is capable of reasonable with the simpler oscillator types, at power levels in
efficiency, high gain, good linearity, and high the milllwatt range. Basic principles are similar to
stability. Its chief fault from the amateur point of those described above, but the oscillator is of the
view, other than high cost, is that the frequency - two -cavity type. Buncher and catcher cavities are
determining circuits are usually part of the tube connected by a feedback loop to sustain oscil-
itself. All too often, klystrons that look like lation. The catcher cavity is made resonant at the
surplus bargains turn out to be built for other than frequency of the velocity modulation of the
amateur frequencies. They are available for fre- electron beam, by changing the shape of the cavity
Velocity Modulation -The Klystron 281
Fig. 13-12 -Basic circuit of the
klystron, identifying the various volt-
ages and currents: Ef - filament
voltage; If -
filament current; Eb -
beam voltage: Ib -
beam current:
/by -body current; Ic - collector
current; O.L. -
overload relay coil;
Im -
magnet current; Pd -driving
-
power; Po output power.

physically and by adjustment of electrode voltages.

The bunched beam current is rich in harmonics,
but the high Q of the catcher cavity suppresses the
unwanted harmonics and keeps the output wave
form pure.
Practical Microwave
designs will include a speech amplifier, preferably
Communication with Klystrons with some provision for automatic level control, to
Though klystrons and other microwave devices keep the modulation at a constant level.
seem strange to the uninitiated, and admittedly Keeping the transmitting and receiving func-
they are costly when purchased new, microwave tions separate may be handled in several ways.
communication of a practical nature can be Separate antennas can be used for the transmitting
achieved with equipment that is surprisingly sim- oscillator and the receiving mixer. This was done in
ple. Thanks to the surplus market and traditional some of our earlier microwave work, but more
ham ingenuity, it can be relatively inexpensive as recently it has been supplanted by the simpler
well. The techniques employed vary in detail, but expedient of "polarization duplexing," first used
the principles are basically similar -and elemen- by the San Bernardino Microwave Society, in the
tary. 1950s.8 Here an open-ended cylindrical waveguide
The method is applicable to any microwave section has probes for transmitting and receiving,
band for which klystrons are available. Two placed at angles of 90 degrees from each other. The
klystron oscillators are built and mounted in "waveguide" may be nothing more than a beer can,
parabolic antennas. Each serves the dual role of as shown in Fig. 13-13. The i -f system can be an fm
transmitting oscillator and local oscillator for receiver, or even a simple superregenerative detec-
receiving. A crystal mixer is built into each unit, tor operating at the selected intermediate fre-
and some of the klystron oscillator output is quency.
diverted for injection to this mixer. The two This is inherently a duplex system, and one in
stations are identical, except for the frequency of which the actual operating frequencies are unim-
oscillation. The two oscillators are separated in portant, so long as they are within an amateur
frequency by an amount equal to the intermediate band, and separated by the chosen intermediate
frequency to be used in reception. frequency. There could hardly be a simpler com-
Let us assume that the band is to be the one at
5650 MHz. One oscillator will be on, say, 5700
MHz. We will use fm broadcast receivers for our i-f
systems. This will require a mixer output fre-
quency of about 100 MHz, so our other oscillator
will be on 5800 MHz.
A klystron oscillator is very readily frequency
modulated. Just a small audio voltage applied to
the dc voltage on the klystron repellor element
does the trick. This can be as simple a device as a
microphone transformer connected in the repellor
lead, with an ordinary carbon microphone the only
other "speech equipment" needed. More advanced

Fig. 13-13 - Looking into the open end

San Bernardino Microwave Society 3300 -MHz
polaplexer. The klystron oscillator is mounted at a
45 -degree angle, with Its "antenna" in the same
plane inside the beer-can waveguide section. Per-
pendicular to it is the receiving probe, connected
to the crystal mixer, upper left. The vertical screw,
upper center, is the mixer injection adjustment.
 282 UHF AND MICROWAVES
3365 MHz
(3335 MHz POLARIZATION 45 DEG.
YSTRON C C W FROM VERTICAL.
POLAPLEXER
3335 MHz
RF
SAMPLE IN21E POLARIZATION 45 DEG.
MIXER C W FROM VERTICAL.

30 MHz 30 -MHz FM
PREAMP REC. AUDIO
BC -683
DISH

DISCRIMINATOR
OUTPUT

A FC REC.
CRYSTAL RELATIVE
-600V i±l O
AFC 21.45 MHa
20-28 MHz LOCK LOCK ERROR VOLTAGE
MIXER TO KLYSTRON
IN2IB FM
BC -603 REFLECTOR
AFC'
n 3356,15 MN SWITCH A FC AMP. K LYSTRON

DIODE CATH. FOLLOWER PWR SUPPLY

HARMONIC -6005
GEN. IN2IB

Fig. 13-14 -
Block diagram of the ROCLOC

M U LT.
CHAIN
XTAL OSC.
RANSISTOR)
n
MIC
frequency -stabilized microwave communications
system developed by the San Bernardino Micro-
wave Society.
305.13 25 MHz
Wit

munications system, yet "polaplexers" of this of copper, with a series of identical "keyholes" in
general type have been used in all our amateur the wall around the inner diameter.11 Each key-
bands above 3000 MHz, and for communication hole represents a transmitter circuit, the hole itself
over some quite remarkable distances. Examples of providing the inductance and the slot, or base of
this technique, too numerous for inclusion in this the keyhole, the capacitance. These keyholes ring a
book, may be found in many issues of QST8 central emitting cylinder or cathode.
A magnetic field is applied axially, causing
ROCLOC -A
Stabilized Klystron System electrons to describe circular paths about the
A marked improvement over the self-controlled cathode when a high -voltage pulse is applied
polaplexer described above was developed by between anode and cathode. The critical velocity
members of the San Bernardino Microwave So- of the electron stream is reached when adjacent
ciety, primarily the late D. L. Thompson, W6IFE, cavities represent positive and negative portions of
and George Tillitson, K6MBL.9 Called ROCLOC the output wave. This is an oversimplified explana-
(Relative Or Crystal Local Oscillator Control) it tion, but it will suffice for our purposes, in view of
allows the klystron oscillator to be frequency - the limited application of the magentron to our
locked to either the incoming signal (relative) or to kinds of communication. The main uses of magne-
a harmonic of a stable crystal oscillator (crystal). trons are for pulsed service, where very high peak
Use of a tunable afc receiver allows the klystron voltages are applied for very short periods, at high
oscillator to be moved accurately in frequency repetition rates. Magnetron peak power of the
without losing frequency lock. order of a megawatt is common, but there has been
ROCLOC equipment has been used by SBMS limited use of the device for continuous -wave
members in several record -setting expeditions, in- applications.
cluding one that resulted in a 214-mile DX record
for the 3300- and 5650 -MHz bands.10 A typical The Travelling -Wave Tube
system is shown in block -diagram form in Fig. Gains as high as 50 dB over very wide bands in
13-14. the microwave region are possible with the travel-
ling -wave tube, shown schematically in Fig. 13-15.
OTHER MICROWAVE DEVICES An electromagnetic wave travels down the helix,
Demand for microwave radar during World War and an electron beam is shot through the helix
II resulted in many devices being developed under
very high priority. The klystron, the magnetron,
INPUT
and the travelling-wave tube had existed in prin- yBEAM
ciple for some time but the wartime emergency ¡-HELIX
OUTPUT
brought them into mass production and use. J3G¢ 0000,
ELECTRON GUN ,ATTENUATOR ¿COLLECTOR
The Magnetron
The magnetron and travelling-wave tube had FOCUSING STRUCTURE

little application to amateur communication, so

they are dealt with only briefly in these pages. Fig. 13-15 -
Basic components of the travel-
ling -wave tube. Device is actually a complete
Essentially a magnetron is a thick-walled cylinder amplifier in a vacuum envelope.
The Varactor Diode 283
from the electron gun at the left end, in the devices has advanced rapidly in recent years.
direction of the wave propagation. When the Manufacture of power transistors capable of opera-
electron velocity is about the same as the wave tion up to at least 3000 MHz is becoming practical.
velocity in the absence of electrons, turning on the The varactor-diode multiplier has been refined to
electron beam causes a power gain for the wave the point where useful power is obtainable up to at
propagation in the direction of the electron mo- least 35 GHz. Transferred -electron (Gunn Effect)
tion. diodes are presently available for oscillator and
The input and output ports are coaxial lines to amplifier applications from roughly 5 to 18 GHz,
which the ends of the helix are coupled. The beam at power levels competitive with or superior to
is fbcussed electrically at the gun end, and magnet- familiar low -power klystrons.13
ically along the helix, by a series of opposing- The Gunn Diode oscillator (up to 200 mW
polarity magnets stacked between ferrous pole output) and amplifier (up to 1 watt output) appear
pieces. to be particularly promising. They require only a
Outstanding features of the TWT are great simple 12 -volt supply of good regulation, and they
bandwidth and large power gain. Efficiency and are well -adapted to frequency modulation. Power
power output are both rather low. The term amplification is possible in either the linear or the
"tube" is really a misnomer; the TWT is actually a locked -oscillator mode. These factors would make
complete broadband rf amplifier in a vacuum it appear that a solid-state version of the ROCLOC
envelope.' 2 system (see Fig. 13-14) could provide a practical
and effective portable microwave communications
system for operation from a car battery supply.
Solid -State Power Generation Gunn Diode oscillators are currently used for
Microwave power generation, amplification, traffic -control radar systems operating just above
and frequency multiplication with semiconductor the 10-GHz amateur band.

THE VARACTOR DIODE

Varactor Principles
Few developments have had more impact on
uhf and microwave communication than the varia- Consider a small block of germanium that has
ble reactance device known as the varactor diode. been doped with impurities, so that one half is
This offshoot of the semiconductor industry made p -type material (contains free positive charges) and
possible many improvements in receiver and trans- the other half n-type material (contains free
mitter design, some offering advantages never negative charges). The result is a p -n junction
hoped for with vacuum tubes. diode. If a voltage of matching polarity is con-
Notable among the varactor's contributions to nected across it, free charges in the material will be
uhf progress are the parametric amplifier, proposed repelled from the terminals and move toward the
in theory earlier but made practical by the varac- junction boundary. This nets an exchange of
tor, and the solid-state frequency multiplier for charge, or forward conduction.
transmitting applications. Diode multipliers were If the applied voltage is reversed the free
also known for many years, of course, but varactor charges are drawn away from the junction boun-
multipliers work at power levels useful for trans- dary, leaving a neutral region called the depletion
mitting. We will discuss these applications in some layer. No exchange of charge is possible; hence the
detail later, but first let us see how the varactor condition of high back resistance. We are interested
diode works. This section is mainly the work of in the latter condition. The depletion layer is a
Wayne Taft, W1WID. dielectric (no free charges) and the regions outside
- +
IIIII- c

P D N

+ - Cio
+ -

DEPLETION LAYER
(A)
EQUIVALENT CAPACITOR

I I

L c' RS J (C) ci ^s (6)

Fig. 13-16 -
Characteristics of the varactor diode. Changing depletion layer with variation of applied
voltage is shown at A. Capacitance decrease with increasing back -bias voltages is shown in Curve B.
The complete equivalent circuit «of a varactor and its mount is given in C, and the practical effect of a
varactor, within its design frequency range, at D.
284 UHF AND MICROWAVES
it are conductors (they have free charges). The two Cj 0 -
Junction capacitance at zero bias.
conductors thus act as the plates of a capacitor,
whose plate spacing (capacitance) is dependent on
Cj.6 -Junction capacitance at some specified
value of reverse bias, in this instance -6 volts.
the applied back -bias voltage. Rs- Series resistance, sometimes called
In Fig. 13-16A we see the depletion layer for "spreading" resistance.
two conditions of back bias: low voltage, for close VB - Reverse breakdown voltage.
spacing and high capacitance; and high voltage, for
B - Thermal resistance in Degrees C per watt.
wide spacing and low capacitance. A typical curve
Useful for power dissipation calculations.
of capacitance vs. applied back bias is shown at B.
This curve is good between the limits of zero bias Junction capacitance is usually measured with a
and the reverse -breakdown voltage (VB) of the bridge at some low frequency, on the order of one
diode. Capacitance is inversely proportional to the megahertz. The value of Rs is usually determined
square root or cube root of the voltage, depending indirectly, by Q measurements at 500 MHz or
on how the semiconductor is doped. The most higher. In addition, two commonly used terms
common type of varactor follows the square-root involve combinations of the above:
law, the result of an abrupt change in doping at the Cutoff Frequency, fc, at a specified value of
junction. bias, and hence C.
Practical Varactors
Up to this point we have considered the
Normalization Power, P0.- (VB)2
Rs
reverse -biased p-n junction as a lossless voltage - These terms equate roughly with maximum
variable capacitor. Unfortunately the regions of the usable frequency and plate dissipation as the latter
semiconductor containing free charges are not would be used with vacuum tubes. In general, for
perfect conductors. Unavoidably they have a built- equal cutoff frequencies, varactors with higher
in fixed series resistance (Rs) usually between 0.1
Pnorm will handle higher power in multiplier
and 10 ohms, which degrades the performance of service. In parametric amplifiers, varactors with
varactor circuits from that obtainable if there were higher Pnorm require more pump power.
a completely lossless varactor. Further complica-
tions arise from mounting the device, as any
practical package adds two important parasitic Applications and Availability
reactances. These are the internal series lead Varactors are useful for a variety of frequency -
inductance and the shunt case capacitance. changing and amplifying applications, including
The resultant varactor equivalent circuit is electronic tuning, phase or frequency modulators
shown in Fig. 13-16C. The semiconductor chip in place of a reactance modulator, parametric
itself is shown inside the broken line, with its amplifiers, and frequency multipliers. Practical
variable capacitance, Cj, and series resistance, Rs. exploration was made of the first two possibilities
In series with it is the lead inductance, Lc, and in in discussion of fm in earlier chapters of this book.
parallel with this combination is the case capa- We will look into the latter two below.
citance, Cc. The parasitic reactances limit the Once it was necessary to test various rectifier
maximum usable frequency of the varactor. Pack- diodes in order to find one that would make a
ages are available, however, for frequencies well good varactor. Now many companies are producing
into the higher microwave range. By choosing diodes specifically for varactor purposes. These
varactor packaging for the frequency range of include such familiar names as Microwave Asso-
interest, package reactances can be neglected. The ciates, Sylvania, Motorola, Amperex, and others.
simple equivalent circuit of 13-16D is then suf- Most varactors now are made from silicon, rather
ficient to describe the varactor. than germanium, for better high -temperature per-
Varactor units look very much like other formance. Varactors made from gallium arsenide
diodes. The small glass -case version with pig-tail are also available. These have extremely high
leads is useful only at frequencies below about 100 cut-off frequencies, but are somewhat more expen-
MHz, and at low power levels. A stud -mounted sive than the silicon types.
varactor of the type used in the frequency multi- Familiarity with the terminology outlined
pliers shown in this chapter could be mistaken for above will enable the would-be user to sort out
a silicon rectifier diode, except for its price tag. It units of greatest interest from the catalog listings.
is useful up to 1500 MHz or so, and at power levels Presently available varactors have characteristics as
up to 50 watts. Microwave packages commonly follows:
used for parametric applications include'the 1N21 VB - ranges from -6 volts for most para-
style and a related double-ended unit. Then there metric -amplifier diodes
to -250 volts for the
are tiny "pill" varactors for strip -line circuits, and higher-power frequency -multiplier types.
various other mountings capable of working well
up into the microwave region.
Ci_6 -from 0.1 pF (microwave types) to over
100 pF for vhf frequency multipliers.
Varactor Terminology -
fc 10 to 300 GHz for silicon, and up to 800
GHz for gallium arsenide.
In order to specify a varactor, certain meas- RS - 0.1 to 10 ohms; usually higher -
urable "parameters" are now in use: capacitance units have lower Rs.
CjVB or Cj min - Junction capacitance at Price is roughly proportional to cutoff fre-
reverse breakdown. quency and Pnorm
The Parametric Amplifier 285
THE PARAMETRIC AMPLIFIER

Lowest -noise devices for uhf reception include "pump frequency" are, in effect, merely new
the maser, the travelling -wave tube, and the para- names for the more familiar local oscillator and its
metric amplifier. There is little point in dwelling on output frequency, in this case.
the first two here. The maser must operate in a In an up -converter (output frequency higher
strong magnetic field. It requires certain gases, or than the pump frequency) a stable power gain
exotic substances like rubies or garnets. Worst of
all, it must be cooled to very low temperatures,
requiring cryogenic techniques quite beyond the
equal to
f
s
could be realized with ideal diodes and
lossless circuits. If the output circuit is tuned to
reach of the amateur worker. We have already
discussed the TWT briefly in this chapter. It, too, is the pump frequency minus the signal frequency
an expensive device; its strong point, very wide - (though still an up -converter) the gain relationship
band amplification, is not required in amateur
service.
This leaves the parametric amplifier, a develop-
is -f
Ts
. The minus sign implies that regeneration is

involved and, depending on conditions, very high

ment of some potential worth to the amateur, but gains can be achieved.
one that has been little used and even less
As a down -converter the output frequency is
understood by most of us. Probably the best
treatment of this admittedly involved subject, for always lower than the signal frequency. Where the
the amateur reader, was a series of QST articles by signal is higher than the pump the relationship fo
W4AO and W4LTU14 written at the time that fs
early development work on paramps was just remains, but since fo is smaller than f5, the device
getting underway. These articles are still required is an attenuator. When the signal frequency is
reading for anyone who would understand noise
problems and the potential and partial solutions.
Readers without engineering background or exten-
below the pump frequency, and -is still less than
fs
sive electronics experience may not find them easy unity, the actual gain may be very high, because of
to assimilate, but they are about as simple as they regeneration, as in the up -converter.
could be without leaving the authors open to the In the regenerative arrangements the pump
charge of over -simplification. What follows is, to a frequency is always the highest in the system, and
large extent, a condensation of their excellent is equal to the sum of the signal and output
work. frequencies. In the regenerative conditions the
Bateman and Bain made a valiant effort to signal in the input circuit is amplified by regen-
stamp out the almost meaningless term, "para- erative action, and the device may be used as an rf
metric," in favor of "reactance amplifier," a name amplifier merely by taking the output from this
more indicative of the way the amplifier works, point, instead of from the output circuit, Tank 2.
but "paramp" seems to have won out in the years The difference frequency must still appear in the
since. There is a rather large family of parametric output circuit, however. The terms "idler" and
devices, and many mechanical and electrical an- "idler frequency" have become standard names for
alogies have been used to explain their operation. the output tank and the energy therein. They have
We will not go into these here, but they run all the no purpose in our life, but they must exist.
way up from the children's swing, which is For practical purposes, the approximate noise
probably as apt (and as confusing) as any. Many figure of the amplifier of Fig. 13-17 can be
systems are "pumped" in one way or another; we'll obtained from the formula:
leave the analogies at this point.
The varactor is in effect a capacitor, the value F=1+Ra+fs
of which changes with applied voltage. It can thus -R1 .fi
be used to modulate power from an external where F is the noise figure, Ra is the shunt
source, in relation to a signal voltage, and therefore resistance across the input circuit represented by
amplify a signal applied to it. The parametric the antenna, R1 is the shunt resistance represented
amplifier of most interest to amateurs is physically by the losses directly associated with the tank
quite simple, being mainly a diode, fed at the signal
frequency, and pumped at a higher frequency
simultaneously. The basic circuit, Fig. 13-17, is
quite similar to a conventional crystal mixer, and it
may be used for either frequency conversion or
straight-through amplification.
The signal frequency, fs, Is applied to Tank 1.
In a frequency converter, Tank 2 is tuned to the ts

output frequency, fo, which may be either higher

(up-converter) or lower (down-converter) than the
signal frequency. The pump tank, top, has only the
job of providing an efficient means for exciting the
diode capacitor (varactor). The terms "pump" and
Fig. 13-17 -
Basic circuit of a parametric amplifier
shows its similarity to a crystal mixer.
 286 UHF AND MICROWAVES
3A
1. High idler and pump frequencies relative to
25 signal frequency.
2. High tank -circuit Q.
2A 3. High-Q semiconductor capacitor, or varac-
N tor, CR1.
J
m 1.5 4. High available capacitance variation, zsC, in
W
0
the varactor.
1.0 5. Small values for Cl and C2.

Q5 Practical Considerations
For a given gain, the regenerative amplifier
06 configuration (basic circuit, Fig. 13-17) is the least
2 4 6 8 0 12
D
14 16 18 20 stable of the arrangements outlined above. Its noise
fs performance, however, is quite good. Furthermore,
it may be used directly ahead of an existing
Fig. 13-18 - Noise figure
amplifier of Fíg. 13-17,
of the parametric receiver or converter. Another big advantage is that
as a function of frequency instability in the pump does not affect the fre-
and antenna loading.
quency stability of the output. Typically, 20 dB of
fairly stable gain is available over a bandwidth of
100 to 200 kHz at 432 MHz. This ís enough
circuit and diode, fa is the signal frequency, and fi bandwidth for weak-signal communication as
the idler frequency.
presently done on this band.
The last two terms of the equation added A practical paramp consists of a varactor
together are a measure of the noise generated by properly coupled to the necessary tuned circuits
the amplifier. Each should be kept small, so
that and pump. In the vhf range lumped circuits are
their sum is a minimum. The second term can be useful, but at uhf and higher, coaxial, strip -line, or
kept small by coupling tightly to the antenna, so cavity resonators are necessary. In the microwave
that Ra is much less than R1. The third term may region these can be constructed from waveguide.
be kept small by using an idler frequency much Resonators should always be high -Q, to prevent
higher than the signal frequency. This means, of losses and poor noise figure. Mechanical stability
course, a still higher pump frequency. is
extremely important in a regenerative setup, as
The way that the noise figure varies with pump small mechanical variations can completely upset
the tuning.
frequency and various values of Ra is shown in Fig.
Varactors designed for paramp service have low
1
13-18. The bottom curve, for as = 0, represents breakdown voltage, usually between -5 and -10
R1 an volts. Typical zero -bias capacitances are on
the
idealized case in which R1 is considered infinitely order of 0.2 to 5 pF. Cutoff frequencies range
large. This curve illustrates the value of a from 20 GHz up, but in general the higher the
high
pump frequency. For example, if a pump fre- cutoff frequency the better the noise figure.
quency S times the signal frequency is used, the Typical diodes are the Microwave Associates
contribution from idler noise will be 0.2. The noise MA -450 and 460 series, and the Sylvania
D4075,
figure under idealized conditions would then be D4140, and D4141 series.
1.2, or about 1 dB. In any practical circuit,
Paramp Limitations
however, the contribution from Ra will add to this.
Someone may be wondering about the noise
Ri
Thus, when you are straining for lowest possible associated with the amplifier output load, and
noise figure it would be more practical to use a whether it is amplified along with the signal by the
pump frequency in the range of 7 to 10 times the regenerative action of the circuit. It is, and the
signal frequency. The contribution from idler coupling problems involved can be solved only by
noise the use of an esoteric device known as a circulator.
will then be in the range of 0.11 to 0.17,
leaving The circulator has the unique property
some room to maneuver in with respect to the of per-
mitting power to flow in one direction only,
contribution from Ra between certain pairs of terminals. By properly
Rl connecting a circulator in a receiving system using
a paramp, the noise generated in the
From here on, analysis of the parametric .load can be
amplifier is very involved, and will not be dealt made harmless by dissipating it in a resistive
with in detail in this text. Though the noise figure sy stem.
equation, even in the simplified form given above, Without a circulator, checking noise figure by
gives us indications on how to keep noise means of a noise generator can lead the worker
figure to astray, in that apparent noise figures
a minimum, it is by no means the
whole story. much lower
Nothing has been said as to how much capacitance than the actual are indicated. Another "fudge
variation is required from the varactor and its factor" in noise generator measurement with a
pump, but it may be said that the following paramp is the low indicated noise figure obtained
conditions are desirable in a practical device: when a pump frequency only twice that of the
signal frequency is used. Such an arrangement is
Frequency Multiplication with Power Varactors 287
fundamentally limited to a minimum noise figure Measurement of the various "parameters," an
of 3 dB, but noise generator measurements may over -worked word we'll use this once, since we're
indicate a noise figure of zero dB. These factors talking about parametric amplifiers, is all but
were undoubtedly at the bottom of some early impossible. Adjustment for optimum results is
amateur enthusiasm for paramps. cut -and -try, to a degree probably not encountered
All this is not to say that the parametric in any other amateur electronic endeavor.
amplifier has no place in the amateur tihf picture. Results can be worth the trouble. Even without
It certainly does have, for those amateurs suf- the circulator (and not many amateurs have access
ficiently skilled in receiver work to assess what is to one) it should be possible to develop noise
being accomplished with the many adjustments figures around 3 dB at 432 MHz. This is better
required. The paramp as generally used in the than is likely with vacuum tubes or crystal mixers,
amateur field is a very tricky item. The pump but no improvement over the better uhf transistor
frequency and power level should both preferably amplifiers. They pretty well take over the burden
be adjustable, in the interest of precise adjustment, of low -noise reception at 432 MHz and should do
yet both should be reasonably stable, so that they the same for higher frequencies eventually.
will stay put when other adjustments are made. Until such times as they do, there is con-
These two attributes are not readily combined in rf siderable to be gained from use of the paramp for
power sources at several GHz! 1000 MHz and up. There is little practical value in
Most paramps built by amateurs have used a paramp for lower amateur bands, except for
klystrons for pump sources. A 432 -MHz paramp practice and experience. The principles are appli-
requires something around 4000 MHz or higher, cable at any frequency, and suitable pump sources
for best results. Since the power and frequency for lower bands are readily obtained. The amateur
stability of a klystron oscillator are both relatively who wants to learn more about paramp con-
poor, adjustment of a paramp using one becomes struction and adjustment can work with them at
something approaching black magic. The pump 50 or 144 MHz, where measurement of results is
frequency and the diode bias must be adjusted, and considerably easier than at uhf.
then the pump power increased, while fiddling The Bateman -Bain series14 describes practical
with the other two items. All three react on each paramp construction for 144 MHz. An effective
other. If the operator finally does get things paramp for 1296 MHz was described in January,
peaked up for optimum results, a slight change in 1961, QST.15 A modification of this for 432 MHz
load impedance (such as may occur when the appears in October of the same year.16 A 432 -MHz
antenna is rotated and objects of differing re- paramp with crystal -controlled pump, the work of
flecting properties appear in its pattern) will throw K2CBA and W1WID, was described in Edition 1 of
the adjustments off, and the work starts all over. this Manual.

FREQUENCY MULTIPLICATION WITH POWER VARACTORS

We are indebted to Henry H. Cross, W100P, input-output curve similar to the upper curve of
for the first practical information on use of Fig. 13-20 is observed. For one tuning condition,
varactors for frequency multipliers in transmitters the lower curve applies, and this is the case where
for 432 MHz. The following is mainly from his a-m is applied to the input. The function is not
QST treatment of this subject.17 perfectly linear, but in on-the -air tests the
Power varactors now available to amateurs will 432 -MHz signal from a 2 -meter phone rig and a
give up to 15 watts output on 432 MHz when varactor multiplier sounds quite satisfactory; better
driven with 30 watts on 216 MHz. They will do in fact than some 432 -MHz plate -modulated set-
almost as well tripling from 144 MHz. The tripler
described below will give a substantial signal on 1*--- VARACTOR CHARACTERISTIC
432 when driven by nothing more than any of the
popular a-m transmitters such as the Communica-
tor. No auxilliary power or audio is required.
The dc voltage -capacitance characteristics and
the output voltage as a function of time, for .-- V
sine -wave current input, are shown in Fig. 13-19.
Once the diode draws conduction current, the
theory gets more complicated, but harmonic out-
put does not cease, so the complications can be APPLIED CHARGE CURVE
ignored for small currents. If the multiplier is
retuned each time the drive level is changed, an

Fig. 13-19 - Dc voltage -capacitance characteristics,

and output voltage as a function of time, of a
TIME

varactor multiplier for sine-wave input current. VOLTAGE PRODUCED

 288 UHF AND MICROWAVES

Fig. 13-20 - Power output from a

450 -MHz tripler using a Type
MA -4060A power varactor. The
150 Mc. INPUT solid line shows the power avail-
450 Mc OUTPUT
able at various drive levels, when
-%-RETUNED EACH POINT
-0-TUNED AT 20W. IN
the tripler is tuned for maximum
output with 20 watts drive. Uni-
form efficiency, upper curve, is
possible if the system is retuned
for each power level change.
2 4 6 8 10 12 14 18

INPUT POWER, WATTS

Fig. 13-21 - Interior of a

doubler stage using a power
varactor. Driven with 20 watts
on 216 MHz, it delivers 10
watts on 432, yet it requires
1
no power supply or modu-
lator.

ups. Doubling from 216 MHz to 432, with a unit about 8 MHz even, and retuning the various stages
like that in Fig. 13-21, the varactor does even to the lower frequency.
better. The trap, L5C5, is mostly to simplify tuning;
without it changing the output capacitor, C3,
VARACTOR DOUBLER FOR 432 MHz would also change the tuning of the 216 -MHz
circuit. The double -tuned input and output circuits
The circuit of the varactor doubler is given in help to establish that the measured output is on
Fig. 13-22. This works well with any 220 -MHz the desired frequency. The circuits are rather
transmitter of moderate power. To operate in the low-Q, however, and use of a coaxial or strip -line
segment of the 420 -MHz band usually reserved for filter in the line to the antenna or following
narrow-band work, 432 to 436 MHz, the 220 -MHz amplifier is recommended, to prevent radiation on
rig is tuned down to 216 MHz, by using a crystal at the driving frequency.

216 Mc.
276 Mc.
432 Mc.

Fig. 13-22 -
Schematic diagram and parts infor-
mation for the varactor doubler.
Cl, C2 -
8.7-pF miniature variable (Hammarlund
MAC -10).
L1 -Tap4-atturns
J1, J2 BNC coaxial fitting.
No. 20. 3/8-inch dia, 3/8 inch long.
C3, C4 -
5-pF miniature variable (Hammarlund L2
turn.
- 5 turns No. 20, 3/8 -inch dia, 3/8 inch long.
1

C5
MAC-5).
- 9-pF subminiature variable (Johnson
L3 - 2-1/2 turns No. 20, 3/8-u. dia, 1/4 inch
h

CR1
189-503-4)
- Power varactor (Microwave Associates
L4 -long.
Tap
2 turns No. 20, 3/8-inch dia, 1/4 inch long.
MA -4060A). L5 - 3 atturnsturn.No. 22, 1/4-inch dia,
1

1/4 inch long.

Varactor Tripler, 144 to 432 MHz 289
amplifier. Up to 14 watts output at 432 MHz is
possible with 20 watts drive on 144 MHz, with this
tripler.

Construction
The tripler is built in a 5 x 7 x 2 -inch chassis.
A shield runs the length of the chassis 2 inches
from one wall, forming a 2 -inch square trough
inside the chassis. A National TPB polystyrene
feedthrough connects the varactor to L3.
Details of the strip -line circuit construction are
shown in Fig. 13-23A. The line is a 5 -inch brass
strip 1/2 inch wide, having a 1/2 -inch "foot" at the
bottom for bolting the strip to the chassis. The
input and output links are tuned with cylindrical
ceramic trimmers. The low -potential ends of L3
and LS are soldered directly to the tops of these.
Fig. 13-23
input circuit
-is The 432 -MHz varactor tripler. The
at the lower right and the varactor
CS is made by cutting two 1 -inch disks from sheet
brass. One is soldered to the end of L4, and a
with its biasing resistor is at the center. Details of mount for the other made from a Miller 4400 coil
the 432 -MHz strip -line tank circuit are given in Fig. form. The ceramic form itself is broken off the
13-23A. mount, and the slug removed from the end of the

VARACTOR TRIPLER, 144 TO 432 MHz

A simple varactor tripler with improved effi-
ciency and reduced unwanted -frequency output Is
shown in Fig. 13-23, with the circuit in Fig. 13-24.
Early varactor multipliers used resonant traps to
remove the driving-signal and idler frequencies
from the output. This tripler, built by W10ER,
omits the traps and uses a strip -line output circuit,
as shown in Fig. 13-23A, to give higher selectivity
at 432 MHz. Only the input circuit, LIC1, and the
input matching network, C1C2, are on 144 MHz.
The series circuit, L2C3, tunes to the idler frequen-
cy, in this case the second harmonic, 288 MHz.
The third harmonic, 432 MHz, is taken off through
coupling loops L3 and LS, and the quarter -wave
strip -line, L4C5. II
Overall efficiency is considerably higher than 1-2:1 Ls V;
when the extra trap circuits are included, and the
output ís relatively free of unwanted frequency
components. It is still desirable to follow any
varactor multiplier with a separate coaxial or
Fig. 13-23A -
432 -MHz tank -circuit details for the
varactor tripler. L3 and L5 are coupling loops
strip -line filter, to be sure that unwanted frequen- made from No. 14 wire, and L4 is a 1/2 -inch -wide
cies are not passed on to the antenna or following brass strip.

Fig. 13-24 -Circuit diagram of the 432 -MHz CR1 - Varactor diode (Amperex H4A/1N4885).
varactor tripler. The strip -line output circuit gives J1, J2
- - BNC coaxial receptacle, chassis mounting.
6 turns No. 16, 3/8 -inch dia, 1/2 inch long.
better attenuation of unwanted harmonics than is L1
possible with lumped -constant circuits. L2 - 3 turns No. 12, 3/8 -inch dia, 3/4 inch long.
Cl -15-pF variable (Hammarlund MAPC-15). L3, L4, L5 - See Fig. 13-23A.
-
C2 25-pF variable (Hammarlund MAPC-25).
-
C3 15-pF variable (Johnson 160-107).
C4, C6 -
10-pF ceramic trimmer (Centralab 144 Mc.
LI
829-10).
C5 - See Fig. 13-23A.

INPUT J2

OUTPUT
 290 UHF AND MICROWAVES
5014e. Llyz

_'J
LI

1
3f

50 Mc.
IN
C 50 Mc. 432 Me.
331 311.5 L7
J Mc Me. 432 Mc.

110.75 Mc.
R
5 432 Mc.
OUT

11 5
190.75
Mc. C7 C
IN
Ja

Fig. 13-25 -
Circuit of the parametric converter.
- -
Resistances are in ohms (K = 1000); capacitances J1, J2 BNC female.
are in pF. J3 Type N female.
-
Cl, C2 30-pF variable (Johnson 160-130).
L1 - 10 turns No. 20, 1/2 -inch dia; tap at 3 turns
- (B&W 3003).

--
C3, C4 15-pF variable (Johnson 160-107).
C7 -
11-pF variable (Johnson 189).
L2
L3
10 turns No. 20, 5/8 -inch dia (B&W 3007).
3-1/2 turns No. 18, 1/2-in. dia; tap at 1 turn
- -
C8, C9, C10 5-pF variable (Johnson
189). (B&W 3003).
C5, C6 Gimmick, 3 turns No. 18 solid plastic - L4 -3 turns, same as L1, without tap.
covered hookup wire twisted together; 1/2
length for 190 MHz; 3/4 -inch length for-inch
L5 -4 turns No. 18, 1/4 -inch dia, spaced wire dia.
MHz (Johnson trimmers may be used as
50 L6 -3 turns No. 18, 1/4 -inch dia, spaced 2 times
wire dia.
shown
C11
in
photos).
-
Gimmick, 2 pieces 1/8 X 7/16 copper
L7 -2 turns No. 18, 1/4 -inch dia, spaced 2 times
wire dia.
CR1
ribbon overlapped 3/8 inch, spaced 0.020 inch.
-
Varactor diode (Amperex H4A/1N4885).
L8 - 2 turns No. 18, 1/4 -inch dia, spaced, tap
1/2 turn. at

threaded rod. The disk is then soldered to the end tune C3 and C5 for maximum output at 432 MHz.
of the rod. The coil-form base is mounted on the If the two cylindrical trimmers, C4 and C6, are
chassis so that the two disks are opposite each set
near maximum capacitance when this is done,
other. For better mechanical stability of the tuning they
will require only a slight repeaking for
shaft, a 6-32 lock nut can be placed on the shaft. maximum
output.
The input capacitors need not be repeaked for
Adjustment maximum 432-MHz output. If the driver stage is
The varactor multiplier will operate most effec- set up for operation into a 50 -ohm load,
tively if the input circuit represents a S0 -ohm load and the
tripler input circuit is adjusted for zero reflected
on the driver. Start tune-up with an SWR power at 144 MHz, optimum 432-MHz output will
bridge
connected between the 144-MHz energy source follow with proper adjustment of the 288- and
and J1. Adjust Cl and C2 for lowest 432 -MHz circuits.
reflected
power at 144 MHz. It is well to use lower The tuning and efficiency of a varactor
than full multi-
144 -MHz drive for this initial adjustment. plier are related to drive level, so final
Now adjustment
should be made at the power level where
maximum
efficiency is desired. The curves of Fig. 13-20
illustrate this.

Operation with Modulated Drive

Adjustment of the varactor multiplier while
checking the modulation linearity at the output
frequency will permit the multiplier to be used
somewhat in the manner of a linear amplifier. Like
proper adjustment of the linear, tuning
for lin-
earity in a multiplier results in much lower
output.
Typically, a multiplier capable of delivering 8
on cw or fm will yield about 2 watts
watts
or reasonably
linear a -m output. With cw or fm, the
process is
merely a matter of "tuning for max."
r.
TRANSMITTING CONVERTER FOR
50 TO 432 MHz
Fig. 13-26 -Parametric up
ferring a 50 -MHz ssb signal to -converter for trans-
432 MHz. The large A varactor diode can be used as a
black object, a heat -dissipating cap for the
tor, taken from a 2C39A tube, is not neecbdvarac- mixer and multiplier, to produce a 432combination
-MHz signal
all -metal chassis is used.
if an from a 50-MHz source. The transmitting
converter
of Figs. 13-25 to 13-27 originally described
by
Transmitting Converter for 50 to 432 MHz 291
Fig. 13-27 - Under side of
the 50 -to -432 converter.
The 50 -MHz circuits are at
the lower left, 190.75-MHz
circuits at the upper left,
and 432 -MHz output cir-
1
cuits with idler tanks along
the right side. The varactor
diode is at the top just to
right of center.

. `N

Adjustment and Operation

W1IGJ in QST for March, 1966, uses 50 -MHz Adjustment of the converter is a little tedious
energy from a Heathkit HX-30 (ssb exciter of since there are eight interacting controls. First
about 6 watts output) and 10 to 15 watts of apply about 10 to 15 watts of pump power
energy on 190.75 MHz to produce about 3 watts through an SWR indicator, and adjust C3 and C4
of ssb output on 432 MHz. With the HX-30's for minimum SWR. A field -strength meter tuned to
tuning range of 50 to 51 MHz, the resulting output 381 MHz placed nearby will serve to detect the
frequency range is 431.5 to 432.5 MHz. doubling operation. Tune C8 for maximum
The 190.75 -MHz pump energy is doubled in the 381 -MHz signal. Go back and forth a few times
varactor to 381.5 MHz. The unwanted product, between C3, C4, and C8, adjusting for maximum
381.5 - 50.5 = 331 MHz, is also generated, and 381 -MHz signal and minimum SWR (they should
must be supported by the idler circuit, L5, C7 in coincide).
Fig. 13-25. This frequency does not appear in the Now connect a load, preferably a wattmeter, to
output. There is some 381.5 -MHz energy in the the output and apply about 1 watt (30 -percent
output, but this is removed easily with a coaxial or scale on the HX-30) to the 50 -MHz input. Tune the
strip-line filter. wavemeter to 331 MHz and adjust Cl, C2 and C7
for maximum as above. Now tune the wavemeter
or other output indicator. Note that a high
Circuit Description wavemeter indication at 432 MHz indicates only
The input parallel -tuned circuits, L3C3 for the circulating current in L7C9 - not output. At this
pump frequency, and L1C1 for the signal frequen- point it is well to go back and start again. Since all
cy, are lightly coupled to the varactor via pi the adjustments interact to some extent you
networks. Although this appears unsatisfactory at should go through at least three times. Do not be
first glance, it will be noted that the reactance of upset if the output indicator on the HX-30 goes up
the 50 -MHz pi-network inductor, L2, at the pump when connecting to the converter; this is normal. A
frequency is so high as to constitute an open 6 -dB pad between the HX-30 and the converter
circuit. Conversely, the pump inductor, L4, pre- gives better carrier suppression since you can use
sents essentially no reactance at the signal frequen- more audio (sideband power) while the carrier level
cy but the capacitor to ground, C4, does. There- output of the HX-30 remains essentially constant.
fore essentially no loading of the signal frequency The unit exhibits good linearity when used to
occurs. drive a 2C39 grounded-grid amplifier to about
The output has series-tuned idler circuits for 12-15 watts output. There is some leakage of the
381.5 Mhz (pump frequency x 2) and 331 MHz 381.5 -MHz signal into the output; this is removed
(pump x 2 minus signal), and two resonant circuits by a couple of tuned amplifiers or by a simple
for 432 MHz (pump x 2 plus signal). Output is high -Q filter (see Chapter 15). With no filter the
taken from a tap on the output tuned circuit. 381 -MHz signal is at least 20 dB down from the
As may be seen in the photographs, the unit is 432 output.
built in an inverted 4 x 6 x 2-inch chassis. The This same scheme can be used to convert from
mounting plate is 1/16 -inch-thick double-sided a 28 -MHz ssb exciter, with appropriate changes in
printed-circuit board (0.040 copper may be used if pump frequency and idler resonances. It is possible
desired). All components are mounted on this to triple from the pump source instead of doub-
plate. Wiring is done with at least No. 18 wire ling, with little change in efficiency. The overall
going point-to-point. Most leads are inherent in the performance is nearly the same except for slightly
components. The varactor heat sink shown is not higher pump power requirements to make up for
necessary if a solid metal mounting plate is used. additional loss.
 292 UHF AND MICROWAVES
VARACTOR TRIPLER FOR 432 TO 1296 MHz
The line circuits are cut from flashing copper
3/8 inch wide. In the model shown they are made
of separate strips soldered together, but they could
be cut as shown ín Fig. 13-30. L3 is an "L" in
shape as well as in function. The 1/8 -inch hole in
one end fits over the varactor post. At the other
end is C4, which Is merely a piece of 3/8 -inch brass
or copper tubing, soldered to the strip, and facing
#71
toward the top of the chassis. Running through the

IF

Nr.>t11;
11
Fig. 13-28 - Varactor multiplier for 1296 -MHz
output with 432 -MHz drive, designed and built by
W1WID. Case is a 3-1/8 X 1 -inch brass box. Large
screws at the left are the movable elements of
capacitors C4, C5, and C7.

Happily varactor multipliers work almost as á

well on higher frequencies as in the 432 -MHz
applications just described. A varactor tripler for
1296 -MHz output with 432-MHz drive is shown in
Fig. 13-28. It is the work of Wayne Taft, W1WID,
who also contributed the basic information on
varactors earlier in this chapter.
Except for the 432 -MHz circuits, coils and i Y 1
capacitors are out of the question for this applica- p
tion. Strip lines are used in an ingenious and
relatively simple manner. The circuit, Fig. 13-31, is
almost identical to that of the 432 -MHz triplers
Fig. 13-29 - Interior of the 1296-MHz varactor
tripler. Coils and variable capacitors are the
described early in this chapter, but the circuits will 432 -MHz circuits. Inductances for 1296 MHz are
require explaining, to the reader accustomed to the copper strips. L-shaped shield of brass isolates
way such things look on lower frequencies. input and output circuits.
The varactor is mounted in the center of a brass
box 3-1/8 inches square and 1 inch high. Adjacent chassis is a No. 10 brass screw, which is the "rotor"
to the BNC input fitting near one corner of the of C4. It runs into the cup formed at the end of L3
box is the 432 -MHz input circuit, L1C1. A small by the brass tubing.
piston -type trimmer, C2, couples energy to C3 and
L2. The latter may be seen connected to the
Construction of L4 is somewhat similar, except
that coupling capacitor C6 is built into it. The
varactor at the center, though the varactor itself is capacitor C5 is merely another No. 10 screw that
out of sight under the strip-line circuits for 1296 runs down so that its end makes a small variable
MHz.
capacitance to ground at the right-angle turn in L4.

I+ ---I Vi-{ --I Vi -1

BEND UP
{ r-reENo UP

I I'--I T I/32 BRASS

7-:-,-
T;
1-
-4.1
-r-}.1

IV4p ©
L Vi
/µi
7
1

IV¡ BEND Va.

yII
IB N0.9 t B UP
-
1
NO. BEND DOWN
+

,
29
r
II
-19,
31/¡ 34; NO.27
F4N..1 1----1946'
i
N291+
N0. 7jI
K, N0.9I
T
I,I
I
N0.9
1 ¡N_O_9
- -4 1-"3/,i
29 Li ":4'1-1411.--1-i
SENO OP
3/S^ -FL, C4 Ili
L3

I/32 BRASS
MATERIAL: FLASHING COPPER

Fig. 13-30 - Details of the case and copper strip lines for the 1296-MHz tripler.
Reception Above 420 MHz 293
Fig. 13-31 -
Schematic diagram of
C6
12% Mc.
the 1296 -MHz tripler.
Cl, C3, -5-pF, miniature trimmer
(Hammarlund MAC -5).
C2 - 0.5- to 5-pF piston trimmer.
JI

C4, C5, C7 -
10-32 brass screws, run-
ning through brass nuts soldered to top of case.
Locknuts are nylon. C4 has 1/4 -inch length of
3/8-inch brass or copper tubing soldered to un-
- -3 BNC
derside of L3, to increase maximum capacitance. J1, J2 fittings.
C6 - Bent -up tabs on L4 and L5, approximately
3/32 inch apart. Bend for adjustable capaci-
L1
ct.
turns No. 18, 1/4 -inch dia, 3/8 inch long,

tance. -
L2 Like L3, but 2 turns.
-
CR1 - Varactor diode (Microwave Associates L3, L4, L5 Copper strip lines. See text and Fig.
MA4062 D 1. 13-30.

There is no brass cup at this point, as only a very of which provides cariable capacitance in the same
small capacitance is required. Coupling between L4 manner as described for CS.
and LS (C6 in the schematic) is made by bending All this is an involved way of saying that tuned
up the ends of the short arms of L4 and LS. These circuits really reach an elementary simplicity at
1/8 -inch wide surfaces then face each other about frequencies this high. They are confusing only
3/32 inch apart. when we think of "coils" and "capacitors" in their
The output inductance, L5, is the most com- 3 -to -30 -MHz connotation. The small shield visible
plex piece. It is bent into U shape at one end to in the photograph is the full height of the box. It
support itself at the same height from the chassis as isolates the 432 -MHz circuits from the output,
the other inductances. The output tap for the BNC thereby keeping the level of the unwanted
connector is made at a point 13/16 inch from this 432 -MHz energy in the output lower than it would
end. Capacitor C7 is the third brass screw, the end be with an open layout.

RECEPTION A BOVE 420 MHz

Except where fm or TV is involved, most the best crystal diodes and circuits, the mixer noise
reception in the 420 -MHz band is done with figure is almost independent of frequency, up to
converters and communications receivers. The cir- around 10,000 MHz. Unlike tube and transistor
cuitry and construction of such converters vary mixers, which usually have some conversion gain,
little from similar devices for the vhf bands, so the diode mixer has a conversion loss. This must be
most of our information on 420 -MHz reception is added to the noise figure of the system following
in Chapters 3 and 4. Fm receivers can use quite the mixer, to obtain the overall noise figure of the
similar front-end stages and circuits. Their special uhf receiver having a crystal mixer.
requirements in the i -f and detection stages are The 1N21 -series crystal diodes, frequently used
dealt with in Chapters 10 and 11. in uhf and microwave mixers, have suffixes A
In ATV work, the "i-f system" is usually a through. F, indicating progressively better mixer
conventional home TV receiver. The uhf front end action and higher price. Nothing below a 1N21C is
can be either a modified uhf TV tuner or a desirable, if noise figure of the mixer is important.
converter (crystal -controlled or tunable) built for The 1N21F (most costly) is capable of 6 dB,
the purpose. Because of the great bandwidth of a compared with 9.3 dB for the 1N21C. Few, if any,
TV signal, stability requirements in the oscillator tubes will match the 1N21F in mixer service above
are not severe. A tunable converter can be made about 500 MHz, especially if the diode Is followed
satisfactorily stable for this use. Such a converter, by a good low -noise i -f amplifier. Some fairly
made from a common uhf TV tuner, is described in expensive uhf transistors are probably at least as
Chapter 11. It is shown in use with a low-band fm good, up through the 1215 -MHz band, and their
receiver, but it could work into a home TV receiver use will make the quality of the following amplifier
on any of the lower channels equally well. Choose less important. No simple receiving system for
a channel that is not in use locally. frequencies above 1000 MHz is good enough so
Most conversions of surplus uhf fm gear or TV that the noise figure of the following i -f amplifier
front ends leave something to be desired in the can be ignored entirely.
matter of sensitivity. Any of the rf amplifier
circuits described in this book should take care of Choosing the Intermediate Frequency
this deficiency adequately. A simple low-cost
The i -f should be as low as practical, as low
amplifier shown immediately following this section
noise figure is more readily obtainable below about
was developed by W6ORG for ATV use. It should
30 MHz than at higher frequencies. It cannot be
serve equally well for other modes.
too low, however, or image response (including
noise at the image frequency) becomes an adverse
The Crystal Mixer factor. A desirable i -f for a 432 -MHz converter may
Many 420 -MHz receiving systems, and nearly all be 14, 21, or 28 MHz, since communications
receivers for higher bands, use crystal mixers. With receivers for these ranges are universally available.
 294 UHF AND MICROWAVES
Nearly all receivers have rather poor noise figure (it Injection Considerations
is not really important in hf work) so a low -noise
Stability is critical if narrow -band uhf reception
i -f or if amplifier, or both, should be provided in
is undertaken. The converter crystal oscillator
any uhf receiving system. The low -noise i -f is should run at the lowest practical input, and the
particularly important at frequencies above 1000 crystal should be isolated from temperature varia-
MHz, where rf amplifier performance deteriorates tions due to component heating and cooling or air
,rapidly with increasing frequency. circulation. Putting the crystal inside the box is
Often 50 or even 144 MHz will be used for the recommended. Injection should be as free of
converter output frequency. This is useful where a harmonic content as possible. Any injection makes
good converter for the vhf band in question is mixer noise, and injection on other than the
already a part of the station equipment. The i -f desired frequency is sure to degrade the noise
need not be tunable in this case; a typical setup figure of a crystal mixer. Finally, the injection
might be a 432 -MHz converter working into a system should be coupled loosely to the mixer, to
50 -MHz one, which then works into a tunable prevent loss of signal energy through the injection
receiver at 14 MHz. circuits.
Two 1296 -MHz converters of excellent design The latter two points are often neglected in
have appeared in QST, The Radio Amateur's amateur designs, with the result that mixers for
Handbook, and previous editions of this Manual. 420 MHz and up having noise figures in excess of
One -by K6AXN18 has a selective trough -line front 15 dB are not uncommon. When a good rf
end, a crystal mixer, and 14 -MHz output. The amplifier is used ahead of the mixer, the amplifier
other, by W6GGV and K6UQH19 features aperture establishes the noise figure of the system, but with
coupling in the mixer system, and 144 -MHz out- a crystal mixer alone the above considerations
put. A 2300 -MHz converter with 144 -MHz output must be handled with care if satisfactory reception
was described by W4HHK.20 is to be achieved.

INEXPENSIVE 440 -MHz PREAMPLIFIER

The simple preamplifier of Fig. 13-32 was built approximately as shown in the sketch included
from information supplied by Tom O'Hara, with the schematic diagram. Parts layout is not
W6ORG, of the Southern California ATV Club. particularly critical, and ours is slightly different
Tom makes and sells similar amplifiers for a from that used by W6ORG, to permit experimen-
nominal sum, primarily for use with commercial ting with input and output loading adjustments.
fm gear that has been converted to 440 -MHz ATV In its simplest form the W6ORG amplifier has
reception. The amplifier shown was built and only one adjustment, the collector circuit capaci-
tested in the ARRL Lab, and used in 432-MHz tor, CS in Fig. 13-33. Two others are shown in our
reception. Its performance is adequate for most diagram: the input loading control, Cl, and C4,
purposes, being probably as good as can be which serves a similar purpose in the output
obtained with inexpensive transistors. circuit. These are not necessary for good results,
The amplifier is assembled on a 2 -1/4-inch but they are helpful in adapting the preamp to
square circuit board having one copper side, etched various antenna and receiver situations.
Another variation from the original layout is
the inclusion of a decoupling resistor, R1, and the
second feedthrough bypass capacitor, C3. Without
these (and with 12 volts connected to C2) there
was instability resulting from rf voltage on the
supply lead. If lower supply voltage is used, a
lower-value resistor can be substituted for R1, as
regenerative effects increase with increasing supply
voltage.
w--- Adjustment
.'.í ri1
When Cl is a fixed value (5 pF in W6ORG's
units) and the lead to J2 is tapped directly onto
Ll, eliminating C4, regenerative effects develop
under certain conditions of loading, both input and
I! output. This is not necessarily bad, as controlled
regeneration makes for higher gain than would
normally be possible with a single stage. With a
very poor receiver you may need all the gain you
can get. If the receiver front end needs only a
little
help from the preamp, Cl can be set so that the

Fig. 13-32 - Simple one -transistor

preamplifier for
432 -MHz receivers. The variable loading
C4, was added after the picture was made.capacitor,
Inexpensive 440 -MHz Preamplifier 295
Fig. 13-33 - Schematic diagram of the uhf
preamplifier. Circuit -board etching is shown at the
MRS -H83

lower left. Coupling between L1 and J2 can be JI

direct, or through variable capacitor C4. All num-
bered components are described in the text.

antenna loads the input heavily, resulting in stable

and uncritical operation over a wide band. Similar- ALTERNATE
ly, using the output capacitor C4, in place of direct COL/RUNG
METHOD
tapping of J2 onto Ll, permits adjustment of
loading for various receiver input conditions.
It is well to start adjustment with a low supply +12V-
voltage, increasing to normal after the amplifier is
working well. Fairly good results are possible with
6 volts or less. Tune in a weak signal with the main
receiver, and peak C5 for best reception. That's all
there is to it, if Cl is a fixed capacitor and the
direct connection is used between L1 and J2. If a
variable -voltage supply is available, it would be well
to try different voltages, as gain should be greater
with higher voltage, up to a maximum of about 13.
If the tuning of C5 seems extremely critical, or
if there is actual oscillation in the stage, the tabs of C4 to Ll and J2
as directly as possible, in
variable capacitors should be tried. Increasing Cl such position that C4
taps on L1 at its approxi-
should broaden the stage tuning, as it loads the mate midpoint. Subminiature printed -circuit
air
amplifier more heavily. It can be used, in effect, as trimmers are ideal for Cl, C4, and C5.
a bandwidth control, if amplifier bandwidth is
important. The purpose of C4, if used, is to tune Results
out any reactive effect in the circuit between Ll With a supply voltage of 12 to 13, and 2700
and the receiver. If the tuning of C5 changes ohms for R1, current drain is about 2 mA. If the
appreciably with differing lengths of coax between stage is stable, R1 can be reduced in value. Current
J2 and the receiver input, you need the capacitor drain should be 3 mA or less. Gain, with Cl
C4 in the circuit. adjusted for good stability, can be as much as 20
With direct coupling to Ll, the circuit is merely dB. More is available, at the expense of more
a 1 -3/4 -inch piece of No. 20 wire. It is soldered to critical tuning and narrower bandwidth. The pre-
the collector tab on the circuit board, and it rises amp was tested ahead of a good transistor conver-
vertically about 3/8 inch. It then bends 90 degrees ter for 432, and also with a very poor crystal mixer
and runs horizontally for one inch, bending down often used as a trial horse in such circumstances.
at a 45 -degree angle to solder to the circuit board As much as 35 dB could be obtained with usable
surface near the BNC mounting flange. The inner stability, but this much is seldom needed. It did
conductor tip of J2 is soldered to L1 at this make the poor mixer show up quite well, however.
45 -degree bend. In the stable condition, the preamp gave results
To install C4, unsolder Ll from the tip of J2 equal to anything we've tried, other than amplifiers
and bend the wire slightly away from it. Solder the with quite expensive uhf transistors.

A TWO -STAGE PREAM PLIFIER FOR 1296 MHz

At around 1000 MHz, even the most expensive make the rest of the system less critical. The
uhf transistors drop off in performance very two -stage amplifier of Fig. 13-34, built by WA2-
rapidly. Carefully built preamplifiers using tran- VTR,21 has a gain of around 19 dB, which should
sistors are still capable of improving the perfor- take care of all but the worst mixers. The best
mance of the best crystal mixers, up through at transistor should be used in the first stage. If the
least 1300 MHz. On higher bands the parametric first stage works well, the noise figure of the
amplifier is about the only low -noise receiving
system that is within the capabilities of the
amateur.
At 1296 MHz, where most weak -signal com-
munication is done in the 1215 -MHz band, a single
transistor stage of optimum design will mask the
noise of the better crystal -mixer receiving systems.
Two stages are advisable, however, as their gain will

Fig. 13-34 - The two -stage preamplifier for 1296

MHz is built in separate units. The first stage is at
r
the right. A jack for plugging in a small 9 -volt
transistor radio battery is shown in the foreground.
 296 UHF AND MICROWAVES
In the two photographs the first stage is shown
at the right side. The boxes are handmade of thin
sheet brass. Standard aluminum Miniboxes could
be used, though brass or copper facilitates solder-
ing direct to the case. The shield in the first stage
should extend nearly the full width and height of
the box. This is not so important in the second
C stage, which has a tuned circuit only on the output
side. The bent brass mounting plate in the second
Fig. 13-35 -
Details of the thin brass shield plate
used to support the transistor in the first
stage is primarily to insure minimum emitter lead
inductance.
rf
amplifier stage. Dimensions will depend on the case The interior views show the input sides at the
size and height of the tuning capacitors
used. The bottom. It will be seen that the strip for the input
emitter leads are soldered to the horizontal "shelf" circuit, Ll, is wider than that for the output, L2.
made by bending the plate as seen in the The transistor has higher input than output capa-
end view.
citance, requiring less inductance in the input
second is not quite so important, so
long as it circuit. All strip inductors are brass and are 3/4
builds up the total gain to an adequate level.
inch long. They are soldered directly to the tops of
the tuning capacitors and are pi -networks.
Construction
Transistors used in early work with 1296 -MHz Adjustment
amplifiers had wire leads. The KMC KC5200 and A signal source is necessary in tuning up the
K5500 used here have flat ribbon leads, making
possible a mounting having substantially no lead
preamplifier. Most small two-meter transmitters
put out enough energy on the 9th harmonic to be
inductance. Most transistor makers now produce
types capable of working well above 1000 MHz. plainly audible at 1296 MHz. Transistorized "bea-
cons" commonly used by amateur uhf experimen-
Usually these have small tablet -shaped cases and ters are fine. Anything strong enough to be heard
ribbon leads. All are fairly expensive. The "accor- on the converter, without the preamplifier, will
dion -pleated" shield plate shown in Fig. 13-35 serve. Just be sure that, if you are listening to
suspends the transistor on its emitter leads, with a harmonic, ít is the right one.
the base lead on one side and the collector lead on
the other. These two leads are soldered to their Initial peaking can be done with no voltage on
the preamp. If a 50 -ohm antenna is used the tuned
respective strip lines, LI and L2, with the mini-
mum possible length.
circuits will be close to optimum adjustment if
The input and output coupling capacitors are peaked first in this way. The same is true íf one is
no -lead disks, though conventional disk fortunate enough to have a 1296 -MHz signal.
ceramics generator with 50 -ohm termination.
may be used if the minimum possible lead length
is Now apply about 5 to 6 volts, and check the
assured. Their value is not particularly critical. The
tuning capacitors, Cl through C6, should be current on each transistor. Adjust the bias controls,
high -quality short piston or coaxial capacitors,
RI and R2, for 1 to 2 mA on the 5200 and 2 to 15
inch center to center. 3/4 mA on the 5500. Now reduce the
strength of the
signal and adjust the tuned circuits for
K3200
maximum
470
470 LI

Fig. 13-36 -Circuit diagram and

for the preamplifier stages. Theparts information
upper
should be used for a single stage. Capacitorportion
are given in pF. values
Cl -C6, incl.- 1 to a pF, high
-quality short piston
or coaxial trimmer. (Johnson used here.)
/'i.7
C7-C10, incl.
or ceramic.
- 500-pF feedthrough, button mica
CR1 - Protective diode. 10 mA or more.
K5500 L3 470
J1 -J4, incl.
625/U.
- BNC receptacle, UG-290/U or
470
L1 - Brass or copper strip, 5/8 X 3/4
J4 inch.
J3
L2 - Brass or copper strip, 1/4 X 3/4
inch.
L3 -inch.
Brass or copper strip, 1/2 X 3/4

R1, R2
trol.
- 25,000 -ohm miniature con-
Power Tripler or Amplifier for 432 MHz 297
Fig. 13-37 - Interior views of the two preamplifier
stages, again with the first stage at the right. The
input ends are toward the bottom of the picture.

response. Readjust the bias for mínimum noise on

the first stage and maximum gain on the second.
Do not exceed 4 mA on 5200.
The preamplifier as shown has a socket for
plugging in a small 9 -volt transistor radio battery. s 4,

This may now be used, and a final peaking and bias

adjustment made for best results. Bear in mind that
optimum signal-to-noise ratio is the objective. This
can be achieved by careful adjustment of the first
stage, and it is not likely to be the same as for

it
maximum signal level. The second stage can be
used as a gain control, to some extent, though this i
is best done in the first i -f amplifier. The gain of
the two stages is about 19 dB, when the system is
adjusted for best noise figure. Not many amateurs
will be able to measure noise figure accurately at
this frequency, but it should be under 5 dB. A transistors at 1000 MHz, but at 1296 MHz it may
system noise figure of 3 dB is possible with these be slightly higher.

POWER TRIFLER OR AMPLIFIER FOR 432 MHz

The power required to drive a 432 -MHz kilo-
watt amplifier to maximum output is more than is
readily obtained with a varactor multiplier. Also,
the cost of an amplifier capable of taking a full
kilowatt input may be a bit high for many users of
the 420 -MHz band. Thus there is continuing
interest in power triplets, and in single -ended
amplifiers, using the external -anode type of tube.
The tripler of Fig. 13-38 and the amplifier of
Fig. 13-40 were built by Dick Stevens, W1QWJ, to
satisfy these two requirements. They are nearly
identical in construction, except for the grid
circuits, and even these are quite similar schema-
tically. Details of both are given under the diagram,
Fig. 13-39. Previous editions of this Manual carried
Fig. 13-38 - Interior of the 144-432 power tripler.
The plate compartment and circuit are similar to
descriptions of a similar design, with a cylindrical those of the straight -through amplifier of Fig.
plate circuit. There is nothing sacred about coaxial 13-40. The main portion of the case is a single
construction, so W1QWJ (who was also partly U-shaped piece of sheet aluminum. The cover
plate, not shown, fastens to the end plates and
responsible for the coaxial version) decided to try a
divider with self-tapping screws.
sheet -aluminum housing, in the interest of simpler
metal work for the average builder. Results with
the two versions were practically identical, so the The line terminates in a 3 -inch-square piece of
aluminum-box form is shown here. The coaxial stiff brass or copper, which forms one surface of
model is in the 1972 ARRL Handbook. the bypass capacitor, C3. Insulation for C3 in the
Construction tripler is 10-mil Teflon. This will do for the
amplifier if it operates at low plate voltages, but
The plate circuits of the tripler and amplifier for use at full plate -voltage rating the Teflon
are identical, for all practical purposes. Each uses a thickness should be 1/32 inch. The assembly
quarter-wave line, tuned at the plate end with a should be held in place with screws passing through
1 -1/2-inch brass-disk capacitor, Cl, driven by ceramic or Teflon shodder washers in the end
10-32 threaded brass rod. The rod passes through a plate of the box. The insulation requirements of
panel bushing (1/8-inch shaft bushing tapped for the tripler are less stringent, as it probably will
10-32) mounted in the front surface of the always be used at moderate plate voltages, and
aluminum box. The plate inductor, L1, is 2-3/8 without modulation.
inches of 1 -5/8 -inch OD copper pipe, fitted with The output-coupling loop, L2, and its serles
1/2-inch forger stock at the plate end. If finger capacitor, C2, are plainly visible in both photo-
stock is not readily available, a removable wrap- graphs. The portion of the loop parallel to Ll is
around brass or copper strip, held in position with about 5/8 inch long. The spacing from Ll, for
a hose clamp, will work equally well. optimum output, will vary slightly, depending on
 298 UHF AND MICROWAVES
J2
C2 OUTPUT
432 MHz 4X150A OR
144 -MHz DRIVE 4CX2506 u
INPUT CIRCUIT
GRID J1 i
RFCI
4
IOKV

Cl2
1W
+

:+
600 TO 2000V
C10 500
500 RFC2

--gH C8
BIAS
500
+
250 TO 350V
C9

6.0V

--Yr
Fig. 13-39 - Schematic diagramand parts infor- J2 -N -type fitting.
mation for the
432 -MHz tripler and
Principal difference between the two unitsamplifier.
is in the
L1 -
Copper pipe, 1 -1/2 -inch 10, 2-3/8 Inches
grid circuit. long. 1/2 -inch finger stock at anode end. C3
Cl, C5 -
Copper or brass disk, 1 -1/2 -inch dia. L2
soldered to far end.
-
1-1/2 inches No. 14, bent in L shape. Sides
Stationary plate fastened to Li 2-1/8 inch from are 5/8 inch each after soldering to L2 and J2.
C2
C3.
- 10-pF shaft -type trimmer, grounded -rotor
L3 -
Copper pipe, 1/2 -Inch ID and matching pipe
cap, total length 3-1/8 inches. Center of C5 7/8
type. inch from open end. Shaft of C5 1-1/8 inch
C3 -
3 X 3 -inch plate soldered to end of L1. from inside right edge of box.
Center hole 1 -1/2 -inch día. Insulated from case
with 10 -mil or 1/32-inch Teflon. See text.
L4 -4
turns No. 20, 5/8 -inch dia, 1 inch long. Tap
C4 - 5-pF miniature trimmer. Solder direct be- L5
RFC 2-1/4 turns from grid end.
-
2 turns No. 22 plastic -insulated, 1/2-inch dia.
tween grid terminal and J1.
C6 - 25-pF trimmer, shaft type.
C7 - 20-pF miniature trimmer, shaft type.
Twist ends 2 turns and insert between turns of
L4.
C11 - Built into socket (not important in
RFC1, RFC2, RFC3 -
10 turns No. 22 enam,
C12 - 500-pF 10kV TV "doorknob."
tripler). closewound 1/4 -inch diS.
J1- BNC fitting. RFC4 -
30 turns No. 22 enam, 1/4 -inch dia, 1-1/8
inch long. Ready-made 2 -meter choke usable.
the operating conditions, but will average 1/4
inch. dimensions are not particularly critical, except
Because it is in a low -impedance circuit, the
series where given in the parts list.
capacitor, C2, can be a miniature type with
.017 -inch spacing (Johnson 160 series). The socket used for the tripler is not a critical
The high voltage is brought through the end item. An inexpensive one without built-in by-
the assembly, on one of the screws
of passing should work satisfactorily in this stage.
holding C3 in The
place. A TV-type door -knob capacitor, latest type, having a raisedscreen ring, is advisable
C12, by- for the amplifier, though an earlier version without
passes the B+ side of the choke,
capacitor, the choke, and the terminal for
RFC1. The it is used here. The screen ring and socket in
the
should be covered by some kind of
C3 tripler are separate units (E. F. Johnson) that are
protective somewhat less expensive than complete single
enclosure, to prevent accidental contact with
the assemblies. No tube chimney is needed, as the
high -voltage line. only
path for the cooling air is into the grid
The sheet-aluminum cases are made from compart-
U- ment, through the socket and anode fins, and
shaped pieces to give a 3 -1/2-inch square
cross-
out
section. The amplifier is 9 inches overall, and through the open end of Ll.
the Varying contact resistance to ground in tuning
tripler 7-1/2 inches. The cover, not shown in the
mechanisms using rotating -disk capacitors is
photographs, fastens to the bent -over end pieces
a problem, especially after
often
and the divider, and has 3/8 -inch bent protracted use. This ís
-over flanges solved in the W1QWJ units quite simply, by
on the long sides. The interior length means
of the plate of a tension system at the panel. A
compartment is 4 inches in both units. Other piece of
aluminum or brass 1/8 x 1 x 1-1/2 inches in size
is
drilled and tapped to pass the drive screw.
Holes
tapped for 6-32 or 4-40 are made either
side of the
center hole. The plate is threaded onto
the shaft,
and rubber grommets are slipped over
the mount-
ing screws, between the panel and the metal
plate.
The tension on these is set firm enough to
-1 , maintain
o
Fig. 13-40 -
500-watt power amplifier for 432
MHz. Grid and plate circuits are half
-wave lines,
tuned with disk capacitors. Plumbing
are used for both lines. components
High Efficiency 432 -MHz Kilowatt Amplifier 299
Fig. 13-41 - This is not a view of a partly finished
project. It is the interior of a strip -line kilowatt
amplifier built by K2RIW. The "line," actually a
5 X 9 -inch piece of circuit board stock, is tuned at
the right side by a beryllium -copper spring capa-
citor plate, controlled by means of a fish -line drive.
Small collars soldered to the cover fit into mylar
chimneys attached to the tube anodes, to confine
the airflow for greatest cooling effectiveness.

good shaft-to -ground contact, but not enough to

make the threads bind or strip. Other mechanical
details of the units should be apparent from the
photographs.

Operation
Normally the tripler is operated well below 432 PARALLEL KW
maximum ratings for the tube used. Using it as a
driver for a kilowatt amplifier calls for something a
around 25 watts output, so the plate voltage can be
800 volts or less. The 2 -meter driver also should be
capable of about 25 watts output. Operating
conditions can be anything within reason, so long
as the dissipation ratings of the elements concerned
are not exceeded. Screen dissipation can rise to on the surplus market. A glass -Insulated tube of the
dangerous limits with high driving power and low early 4X150A variety should be adequate for this
plate voltage, so this is something to watch in use.
setting up a tripler with this type of tube. The Operation of the 4CX2S0B and similar tubes as
screen voltage circuit shown for the K2RIW ampli- amplifiers is covered in detail in Chapter 6, for
fier, Fig. 13-45, is recommended, except that 250 both linear and Class -C conditions. This amplifier
volts might be safer than 300 in the tripler has been operated for considerable periods at 500
application. watts input, with little sign of the varying output
Because relatively low input is required for the and tuning drift that may occur when excessive
tripler, the tube can be the sort occasionally found heating of tube and components is present.

HIGH EFFICIENCY 432-MH z KILOWATT AMPLIFIER

The section on power amplifier design, Chapter The strip -line tank circuit is so simple that one
5, goes to some length in showing why "parallel gets the impression, upon lifting the cover of the
operation is not practical for the higher vhf K2RIW amplifier for the first time, that the builder
bands." How then can there be a 432 -MHz hasn't finished the job. The basic idea is that the
amplifier in which two 4CX250B-series tubes are inductance of a strip -line can be made as low as
used in parallel? The answer lies in a qualifying desired, simply by making the inductor wider. Any
statement added in this edition to accommodate number of tubes can be paralleled and resonated,
developments like the 432 -MHz amplifier described so long as the intertube resonances - push-pull
herewith: "It is possible to use certain vhf and uhf modes - are controlled. Details of the "mode -
tubes in parallel with properly designed strip -line killing" are not given here, nor are several other
and cavity circuits. These require a complete break interesting and useful ideas incorporated in this
with conventional coil -and -capacitor concepts." unique design. Study of the full treatments by
The amplifier of Fig. 13-41, designed and built
by Dick Knadle, K2RIW,22 not only makes para-
llel connection practical; it is the most effective
kilowatt amplifier configuration yet described for
432 -MHz operation, in literature available to the
average uhf enthusiast. Parallel use of 2C39As in a
1296 -MHz resonant cavity was described by
WB6IOM in QST and the ARRL Handbook.5

Fig. 13-42 -
Interior of the 432 -MHz amplifier,
with the half -wave strip-line removed. The tuning
. -
flapper, C5, is fastened to the right sidewall. A
similar string -drive capacitor plate, C4, is soldered
to the power output jack on the rear wall. The
plate -circuit rf choke, RFC1, is mounted on a
feedthrough bypass capacitor, C7, in the upper left
corner of the picture. Positions of C4 and C5 are
those for maximum power output. r
 300 UHF AND MICROWAVES
a Fig. 13-43 -
Details of the plate compartment,
J2 RrC2 i AIR ! showing placement and dimensions of all major
256H components.
O 1.3yf Ai rI g7j, !a
dNPUr,

VK2y¢,.
VW/ i

E,
1
SPACERSy DERED
TO UNDER SIDE I CS:
313" }
a"
2% o
0\W/i
' O
L1: 5%9"
OiS tube chimneys are sections of 1 -5/8 -inch brass
LP OI
tubing, 3/8 inch high, also soldered in place. The
<fiOJ four butt joint corners of the upper chassis have
spots of solder ín the centers of the cracks, on the
inside, to insure good rf shielding.
u u
Placement of holes for the sockets on the
chassis and the holes in the strip -line, L1, should be
12"
SK60O done carefully. Misalignment will cause skewing of
TUNING
SK61 0
SHAFTS J2, C4,C5 MOUNT the tubes in the sockets, which may result in poor
2"BELOW TOP OF CHASSIS
contact and make tube insertion difficult. The
K2RIW in QST22 and Ham Radio23 is strongly resonator is made of a 5 x 9 -inch piece of 1/16 -
recommended. inch double -sided glass-epoxy circuit board. This is
In this amplifier the parallel grid and plate easy to cut and solder, and the conductor thickness
networks force the voltages on the two tubes to be is more than adequate, since rf current skin depth
identical. If one tube has higher emission it may in copper is 0.125 mil at 432 MHz. Any rigid good
draw slightly higher plate current, but this ís of conductor greater than 1 -mil thickness should do
little consequence because this happens to be the equally well.
condition under which the amplifier will produce Finger stock around the tube holes is soldered
the greatest output with these two tubes. Such to both sides of the copper -clad board. Intrinsic
tube unbalance could be troublesome in a capacitance between the two surfaces eliminates
push- need for
pull amplifier, or in a parallel one of any other connections. Some builders of
conventional the
design. amplifier have made anode connectors of
Troublesome stray resonances, almost inevita- flashing copper, clamped to the tubes in various
ble with conventional coils, tuning capacitors ways. Such connectors are satisfactory, but
and they
coupling loops, are avoided by use of a flapper- hamper tube replacement. The strip-line is mount-
type tuning capacitor, and a similarly ed above the groundplane on two 1 -1/2-inch
constructed
capacitive output probe. These are C5 and C4, porcelain or Teflon spacers.
respectively, in the schematic diagram, Fig. 13-44. Cooling air comes into the upper chassis
The fish -line method of controlling these
capaci- through a 1 -1/2 -inch screened hole and divides into
tors has a further advantage: it eliminates
variable two paths. About three -fourths goes through
contact resistance to ground, a problem that often tube anodes and chimneys and out through the the
develops with rotating -disk' capacitors using screw- screen top -cover holes. The rest passes through the
thread drive. tube sockets and out through a 3/4-inch
screened
hole in the side of the lower chassis. This
flow
Construction pattern is beneficial in several ways. The cool air
The amplifier is built in two first flows around the plate -circuit
Premier which are suspected sources of components,
8 x 12 x 3 -inch chassis, screwed detuning due to
together top to heating in many closed -in
bottom. The top surface is cut away, leaving a amplifiers. Secondly, full
3/4 -inch rim all around for mounting the top air pressure is available at the
plate. need cooling most. The sockets tube anodes, which
The cover ís 0.091 -inch aluminum, though create greater back
would probably have been equally good. thinner pressure than the anodes, resulting in the 3:1 flow
tion of Pem-nuts, Rivnuts, or other blind Installa- division in favor of the anodes.
on this rim Is recommended. Any builder fasteners is
The anode chimneys are 10-mil Mylar, 1-5/16
to be called upon to show the "innards.. likely inch wide. The material is cut long enough to
of the overlap about one-fourth turn
amplifier frequently, and holes for self when wrapped
-tapping around the tube anode. Two tubes
screws won't stand this kind of use. are placed end
The two 1 -5/8 -inch air -exhaust holes to end, and the Mylar wrapped around their
directly and anodes
above the tube anodes have copper held in place with rubber bands. The overlap is
screening then glued with silicone rubber
soldered across them, on the inside. The air
intake rubber bands left on until the adhesive, with the
hole on the back wall is similarly adhesive is dry. Be
screened. The careful that the chimneys
surface of the box around the hole are not so long as to
edges was separate the finger stock
tinned with Kester aluminum solder, using an from the
tip on a propane torch. The iron when the cover is tightened down. Tentube anodes,
-mil Mylar is
screening is tinned often used by draftsmen, so it
with ordinary solder, where it is to should not be hard
contact the to obtain. If the air flow is not sufficient
aluminum, and then tacked in place using the iron there may
attachment and torch. Such a soldered screen gives be melting of the Mylar. More air Is then called for,
better shielding than most compression or chimneys can be made of sheet Teflon.
methods. The narrow collars that align mounting The capacitor plates, C4 and C5, are made of
with the 0.008 beryllium copper, but many
other springy
High Efficiency 432 -MHz Kilowatt Amplifier 301
conductive materials are usable. The natural rest drilled with small holes to take the string. Slight
position of each is made so that it is just below LI, misalignment of the shaft bushings causes enough
so they do not come dangerously close to it, in any friction so that the capacitors hold their settings.
position. This is preferable to placing any kind of The power -sampling capacitor, C6, is a metal
insulating material on the plates to prevent elec- tab suspended 3/16 inch below the center pin of
trical contact, as dielectric losses can be appre- J2. Semirigid coax carries the rf power to the lower
ciable. The mounting end of C5 is bent 90 degrees, chassis, where the diode CR1, and associated
1/2 inch from the end. A stiffening plate 1/16 -inch components are located. The outer conductor is
thick is cut to fit this portion where it is bolted to soldered to the chassis at the point where it runs
the chassis wall. The edge of C4 is soldered directly through. If semirigid coax is not availab5., an
to the center pin of J2. insulated wire inside small copper tubing should do
Fishing string is fastened to the free end of each equally well, as only a very small amount of power
flapper capacitor. Each string goes through a small is needed. If the inner conductor is stiff, the
chassis hole and is wrapped around a Bakelite capacitor plate can be supported on the wire alone.
adjustment rod in the lower chassis. The rods are Its size and position should be such that about a

J2 OUTPUT

14CX25013 érr, RFC2 C4

Y1 LOAD

TUNE
CR1
Ct TUNE .001

CS '47
3
10K
RFC3
LOAD

C3 14CX25013
RFC!
L _

J1 C J
INPUT i
.O<

390

10 <2000V
390 J4 I

J3 EXCEPT AS INDICATED, DECIMAL VAWES OF

CAPACITANCE ARE IN MICROFARADS I PF I ;
OTHERS ARE IN PICOFARADS I Pr OR J9F1;
RESISTANCES ARE IN OHMS;
t. 1000, M.1000 000.

C7 - .0011.IF 2500-volt feedthrough (Erie

CSK-711).
Fig. 13-44 -
Schematic diagram and parts infor- C8 -
.001 -/IF 3000 -volt disk ceramic.
mation for the 432 -MHz kilowatt amplifier. In the CR1 -
Hewlett-Packard HP-2301 diode.
interest of conveying mechanical information, J1, J2 -N
-type coaxial fitting.
some parts are not shown in conventional schem- J3 -4
-pin male power connector.
atic form. J4 -
High -voltage connector (MHV type).
Cl - 5-pF per section butterfly (Johnson L1 -
5 X 9 -inch double -sided glass -epoxy circuit
160-205). board. Round corners 5/16 -inch radius. (Half-
C2 - 3-pF per section butterfly (Johnson wave strip -line.)

C3
160-203).
-
4.5-pF per section printed -circuit butterfly
L2 -
Half-wave grid circuit, flashing copper. See
Fig. 1347.
(Johnson 189-251-5). RFC1 - 5 turns No. 18, 1/4 -inch dia, 1 inch long,
C4 -Fig. 13-43.
String -driven loading capacitor. See text and axis vertical.
RFC2 - Like RFC1, but 1/2 inch long.
C5 - String -driven tuning capacitor. text and See RFC3 - rf choke (Ohmite Z-460).
0.1-1.1H
Meter functions: Position A -B - grid current, 100
Fig. 13-43.
C6 - Copper disk, 1/4 -inch dia, 3/16 inch below mA; -E - Screen current, V1, 50 mA; - D -E
screen current, V2; -H - relative power output.
C
J2center conductor. text. See F
 302 UHF AND MICROWAVES
10 REQ'D
2W
12,uF
J5

,
TI
3000V
117V
III 2000V
C10 .5A
4800V CR2-CR9
.5A
í00K
100W

11111114111111
120 R2- R10
CR10 CR11
3000V
C9
L3
11H R11 R12

750

2K -90V
15W
+300V -IOOV TO -200V
RIO 300V

J
@50mÁ RI415K @ITOIOmA

r77
10W
-Oh- V3

Fig. 13-45 - Details of power supplies for the L3 - 11-H 500-mA choke D3227, surplus).
432 -MHz amplifier. Capacitors with polarity shown P1 -4-pin plug to match J3.(UTC
are electrolytic. Parts not described are numbered Q1 - Npn transistor, 10 watt, 100 VCE, beta at
for text reference. Only the external circuits of the least (Delco
screen and bias supplies are shown.
-
CR2-CR11, inc. 3000-V 500-mA diode, 20-A
R1
DTS-401).
-V1 5014-ohm 15 -watt control. Adjust for 5.5 V at
and V2.
surge rating. T1 - 4800 V, ct, 500 mA (UTC D3221,surplus).
11 - 117 -volt pilot lamp. T2 - 6.3V 5-A filament transformer (UTC S-55).
J5 - MHV connector. V3 - 5651, V R-90 or 90-V -watt Zener diode.1

3/4 -scale reading is obtained on the meter, at full 3/8 -inch threaded ceramic insulator, fastened to
power output. the chassis. The outer conductor of the RG58
The grid circuit tuning and loading capacitors, input cable is soldered to a ground lug under the
Cl and C2, are butterfly types, to avoid sliding insulator, and the inner conductor to the other
contacts to grounded rotors. C3 is a screw -driver stator of C3. Bakelite rods extending through the
adjustment trimmer, used to set the range of C2. front are used for adjustment of Cl and C2. Proper
One stator of Cl ís soldered to a ground lug on the settings will make the amplifier look like a 50-ohm
chassis floor. The other stator is soldered to a lug load over all reasonable drive levels.
on the lower side of L2, 4 inches from the grid Careful shielding of all leads entering the grid
end. One stator of C2 is connected similarly to L2, compartment is required, if the drive requirement
at a point 2-1/4 inches from the grid end. The is to be kept low. There is a strong rf field in the
other C2 stator and one stator of C3 are soldered area of the 4CX250B grid circuit, and any leads
to an ungrounded lug bolted to the top of a not completely shielded and grounded will couple
power out. Note that no wiring of any kind is
IP
mm 411.
placed in the area under or near L2.
The grid compartment shield is a modified
5 x 7 x 3 -inch aluminum box, held in place with
sheet-metal screws. Slots are cut in the box around
the tuning shafts and input cable, for easy removal.
i Fig. 13-46 -
Bottom of the 432 -MHz amplifier,
showing grid -circuit details. Butterfly capacitor
mounted at 45 -degree angle, lower center of the
grid compartment, is C3. C2 is just above C3, out
of sight. Cl is near the right end of L2. Bakelite
shafts, right, carry the fish -line to C5, right, and
C4. Their bushing mounting plates are slightly
misaligned to create tension on the shafts.
 High Efficiency 432 -MHz Kilowatt Amplifier 303
screen current, so R13 normallydrops the 300 -volt
supply only about 10 volts.
The grid bias supply need not deliver much
current; rather it must act as a load for the 20 mA
C2 LUG
the tubes normally deliver when driven into Class -C
conditions. This calls for a shunt regulator, able to
tolerate 100 mA without failure, in case drive is
momentanly applied without plate voltage. With-
31/í' out a shunt -regulated supply, bias can increase
when drive is applied. This may increase the drive
requirement for full power output, and cause
back -bombardment of the tube cathodes.
Fig. 13-47 -
Details of the half -wave grid line, L2.
Material is flashing copper or similar sheet brass.
Performance

Five 1/2-inch holes are drilled in the end adjacent

The builder of the amplifier tested it very
thoroughly using the best laboratory equipment
to Cl, to permit free air flow.
available for this kind of work. He used a 20-watt
Power Supplies 144 -MHz source and his 144-432 tripler23 to drive

The amplifier requires 2000 volts at 0.5 am-

pere, 300 volts at 50 mA, and -90 volts, with a
50-mA current sinking capacity. The choke -input E)

solid-state supply for the high voltage (Fig. 13-45)

tion, and a built-in spike arrestor consisting of
CR10, CR11, and C9. The choke and meter in the
negative lead contribute to safe operation and
reliability. The bleeder resistor is connected so that
the meter does not read the bleeder current. Use of
RG59 coax and MHV connectors for running the
high voltage from the supply to the amplifier is a
good safety measure. (MHV' connectors are similar
to BNC, except for having an extended center
insulator. They are rated to 5000 volts, and they
automatically maintain a ground through the co-
axial shield.)
The amplifier is protected from high screen
voltage by R14, which drains off negative screen Fig. 13-49 - Rear
kilowatt amplifier.
view of the K2RIW 432 -MHz
current that can occur during some operating
conditions. With R13 in the line to the screens, the amplifier. Maximum efficiency (with fixed
dissipation cannot exceed 12 watts under any conditions of plate, screen, and bias voltage) was

//
condition. Most tubes used in this amplifier deliver obtained with around 22 watts drive. The curves of
maximum power output with about 5 mA positive Fig. 13-48 show the power output (solid line) and
plate efficiency (broken line) for various drive
900
levels. It will be seen that nearly 350 watts output
can be obtained with only 10 watts drive. With the
800 bias and screen supplies described, plate current
700 EFFICIENCY drops with decreasing drive levels, so the plate
600 OUTPUT
dissipation remains at a safe level at any power
500- / output condition, down to one watt or less.
i The operating conditions described by K2RIW
400 - are convenient for anyone who wants a simple way
300- of transmitting cw or fm, over a wide range of
200 -
power-output capability without changing supply
voltages. If the amplifier is to be used as a linear,
100 with a-m or ssb, provision for changing at least the
10 15 20 25 .30
screen and plate voltages would be desirable, to
5
POINTAT WHICH C4 DRIVING POWER, WATTS
preserve linearity and efficiency at low -input con-
RUNS OUT OF RANGE ditions. Something like the operating conditions
described for the 50- and 144 -MHz linear ampli-
Fig. 13-48 -
Output and plate efficiency of the
432-MHz amplifier, when operated Class C, with
fiers in Chapter 6 would then be in order. It should
be emphasized that the series -shunt resistor net-
bias, screen voltage, and plate voltage fixed at -90, work in the screen supply and the shunt-regulated
+300, and +2000 volts, respectively, varying the bias supply systems used by K2RIW and shown in
drive level from 8 to 30 watts. Input drops with Fig. 13-45 have great merit, and could well be
decreasing drive, so operation is always at safe
levels of plate dissipation. Suggestions for linear incorporated in any amplifier using this series of
operation are given in the text. tubes. They should help materially to correct
 304 UHF AND MICROWAVES
problems reported by some users, and to prolong that band. The parallel amplifier, running at 700
what has been normally a record of excellent tube watts output on 432, reduces the 144MHz output
life, with this tube family. by 50 dB. Referred to the 432 -MHz output, the
The testing and measurement program con- spurious outputs are down as follows: 144 MHz -
ducted by K2RIW also included a thorough check
for spurious outputs, using instrumentation not
71 dB, 288 MHz - -
69 dB, 864 MHz 45 dB, and
normally available for amateur work. His tripler,
1296 MHz -
50 dB. With the amplifier running
into a tuned antenna system, the rejection of the
though probably more effective than most in unwanted frequencies should be even higher than
reducing radiation at the driving frequency, does the above exceptional figures. Spectrum analyzer
put out 28 milliwatts on 144 MHz, which is examination showed no other significant outputs,
enough to be heard over a considerable area on from dc to 12.4 GHz.

PULSE COMMUNICATIO N ON 2300 MHz AND UP

Microwave communication with simple oscil- components, and they have shown repeatedly that
lator-type transmitters requires a very wide band- they are capable of providing reliable communica-
width, transmitting and receiving. While much tion over distances comparable to those covered on
good work has been done with such inherently the vhf bands.
low -efficiency systems, something better is needed The rf portions of the transmitter and receiver
if the full amateur potential of the microwave can be identical with those used for simpler
region is to be realized. systems.25 The effectiveness of pulse is obtained
The usual alternative, use of crystal control and through special characteristics of the modulator in
narrow -band receiving techniques, makes a very the transmitter, Fig. 13-46A, and the threshold
marked improvement but it entails considerable detector in the receiver, B. Both involve relatively
effort and expense. The pulse system shown in simple circuitry.
block -diagram form in Fig. 13-50 represents a
desirable compromise between these two extremes.
SURPLUS GEAR FOR THE
Stability requirements are no greater than with the
simple oscillator approach, yet the communica- MICROWAVE BANDS
tions range can be nearly equal to that obtained Amateur operation on the frequencies above
with narrow -band cw. Cost and complication are 1000 MHz would hardly have been practical, had it
considerably lower than with narrow-band meth- not been for many items of equipment obtained on
ods. the surplus market at a fraction of their original
Pulse is a wide -band mode by nature, so it is cost. To describe in detail how all this gear can be
permitted only in the microwave region, where used would take a book larger than this one. Much
amateur assignments are wide enough to accom- of the necessary information for making use of
modate it. The system shown was developed by microwave surplus is in QST. Nearly all the
John T. Zimmer, W2BVU, and Robert F. Guba, references in the bibliography below employ sur-
W1QMN, for use in the 2300 -MHz band, but it is plus in one way or another.
applicable to higher bands, except 10,000 MHz, Some surplus items that have not been treated
where pulse is not permitted. Their full treatment extensively in QST are described briefly below.
in QST24 and condensations in the 1965 and 1968 There are many others, and as uhf and microwave
editions of this Manual describe a system well communication converts gradually from tubes to
within the reach of the experienced amateur. Many solid-state devices there will undoubtedly be more.
versions have been built, mostly from surplus
APR -4, with tuning head TN-54 - Radar
receiver, 2150 to 4000 MHz- LO in tuning head.

KEY -r I uroR
MDPUaLSE
PULSED
2300 Mc
OSCILLATOR
(I ANTENNA
APR -5A
in one unit.
- Radar receiver, 2150 to 5000 MHz,
APR-9, with tuning head TN-128 - Radar
receiver, 1000 to 2600 MHz. LO in tuning head.
APG-5 or APG-I5 - Tail -gun radar, 2700 to
2900 MHz. Has 2C43 pulse oscillator and 2C40 LO,
both in coaxial circuits readily convertible to
POWER
SUPPLY
2300-MHz band. Have been used in W2BVU-
W 1QMN pulse system of Fig. 13-50.
(A) TRANSMITTER

1 INC
ANT. MIXER -^ 2w0 TADESNOLD
PRF
I IPN- DET DETECTOR FILTER

LOCAL AUDIO
Fig. 13-50 - Block diagram of a complete pulse
OSCILLATOR OM PLI FI ER communications system for the 2300 -MHz band.
To speaker Communications range approximates that obtain-
or earphones able with narrow -band methods, with
(B) RECEIVER much sim-
pler and less expensive equipment.
Pulse Communication on 2300 MHz and Up 305
14 Bateman and Bain, "New Thresholds in
2J39 Integral -magnet magnetron, 9 -kW peak-
power output, 3267 to 3333 MHz. V.h.f. and U.h.f. Reception," December, 1958, and
FPN-13 - Naval radar, crystal -controlled, 8 to January, February, and March, 1959, QST.
15 Troetschel and Heuer, "A Parametric Ampli-
9 GHz. Output stage has V-45 klystron, useful as
quintupler to 10 GHz. Has cavities useful for 2300; fier for 1296 Mc.," January, 1961, QST.
3 with 2C39s. 16 Sager, "Parametric Amplifier for 432 Mc.,"
UPX-6 - Crystal-controlled transmitter and Hints and Kinks, October, 1961, QST. A complete
receiver (base unit for use with APX-6) 1215-1300 432 -MHz paramp was described in Edition 1 of this
MHz. Power and pulse circuits useful for any Manual.
microwave band. 17 Cross, "Frequency Multiplication with
Power Varactors," October, 1962, QST.
BIBLIOGRAPHY 18 Krivohlavek, "1296 -Mc. Converter Without
Complications," March, 1961, QST. Also ARRL
1 Holladay and Farwell, "Beer -Can Baluns for
Handbook, 1962-1968.
144, 220 and 432 Mc.," February, 1965, QST. 19 Meyer, "Crystal -Controlled 1296 -Mc. Con-
2 Brayley, "Coaxial -Tank Amplifier for 220 verter," September, 1962, QST; ARRL Handbook,
and 420 Mc.," May, 1951, QST. 1970-1972; 1965 and 1968 editions of this Man-
3 Garrett and Manly, "Crystal Control on ual.
10,000 Mc.," November, 1963, QST. "Wilson, "2.3-GHz Crystal -Controlled Con-
4 Robertson, "Tripler for the 1215 -Mc. Band," verter," April, 1971, QST, p. 34; May, p. 87.
July, 1955, QST. 21Vilardi, "Two -Stage Transistor Preamplifier
5 Laakmann, "Cavity Amplifier for 1296 Mc.," for 1296 Mc.," December, 1968, QST.
January, 1968, QST; ARRL Handbook, 22Knadle, "High -Efficiency Parallel Kilowatt
1968-1972. for 432 MHz," QST, two parts, beginning April,
6 Badger, "An Introduction to the Klystron," 1972.
August, 1961, QST. 23Knadle, "Dual -Band Stripline Amplifier-
7 Orr, Harris, "Project Moonbounce," Septem- Tripler for 144 and 432 MHz," Ham Radio,
ber, 1960, QST. February, 1970.
"W3GKP and W4HHK Work on 2300 MHz!" 24Guba and Zimmer, "Pulse: A Practical Tech-
World Above 50 Mc., December, 1970, QST. nique for Amateur Microwave Work," February
8 Simple duplex phone equipment for all ama- through May, 1963, QST. Essential details repro-
teur microwave bands has been described many duced in 1965 and 1968 editions of this Manual.
times in QST. The following references should be Photocopy of the Manual text, 15 pages, $1.50,
helpful to anyone interested in this approach to from ARRL.
microwave communication. 25Koch, "Simplified Oscillators for 2300
3300 MHz.: Mc.," February, 1948, QST. Basic information in
Baird, "Radio Club for Microwave Enthu- ARRL Handbook, 1949-54, and in reference 22,
siasts," December, 1957, QST. above.
Bredon, "Let's Go Microwave," June, 1958,
QST. Other useful QST references to uhf and micro-
Peterson, "Practical Gear for Amateur Micro- wave equipment and techniques include the follow-
wave Communication," June, 1963, QST. ing, listed by bands.
5650 MHz.: 420 MHz:
Merchant and Harrison, "Duplex Phone on Poland, "Converting FM Equipment," August,
5300 Mc.," (Temporary band, later changed to 1968.
5600 Mc.) January, 1946, QST. Clement, "Using Motorola TU -110 Series on
Prechtel, "Experimental Transceivers for 5650 420 MHz," September, 1971.
Mc.," August, 1960, QST. 1215 MHz:
10,000 MHz.: APX-6 conversion - September, 1960; Feb-
McGregor, "Dishing Out the Milliwatts on 10 ruary, 1961; February, 1968, p. 82.
kMc.," February, 1947, QST. Basic information Fisher and Turrin, "UHF Directional Coup-
repeated in several ARRL Handbook editions, lers," September, 1970.
1948-1954. Scott and Banta, "Using the Helical Antenna on
21,000 Mc.: 1215 Mc.," July, 1962.
-
Sharbaugh and Watters, "Our DX 800 Feet!" Troetschel, "Quad Helix for 1215," August,
August, 1946, QST. Same Authors, "World Above 1963.
20,000 Mc.," May, 1959, QST., includes infor- Fisher, "Interdigital Bandpass Filters," March,
mation on equipment for 50,000 Mc. 1968.
Vilardi, "Easily -Constructed 1296 -MHz An-
9 Jensby, "Stable Microwave Oscillators," July, tennas," June, 1969.
1966, QST. 2300 MHz:
10 Kolbly and Munn, "Microwave DX, Cali- 2300 -MHz Harmonic Generator, October,
fornia Style," September, 1970, QST. 1968, p. 94.
it Argento, "Centimeter Wave Magnetrons," Moonbounce, July, 1969, p. 54. Pulse Record,
December, 1945, QST. February, 1969, p. 53. Tropo Record, September,
12 Scott, "The Travelling -Wave Tube," July, 1970, p. 96.
1963, QST. Many useful publications devoted to microwave
13 "Gunn Diode Circuit Handbook," Micro- techniques and theory are available from the U.S.
wave Associates, Inc., Burlington, MA 01803. Government Printing Office, Washington, DC
Copyright 1971. 20402. Write for list of titles and prices.
 Chapter 14

Test Equipment for the V H F

Station
Many vhf men tend to be experimenters at to test equipment. Yet whether we buy our
heart. Much of the pleasure and satisfaction to be operating gear or build it ourselves, only by having
derived from vhf work comes from striving for the adequate testing and measuring equipment, and
best possible performance from our stations, using it regularly and well, can we be sure that we
whether they are lavish or modest in nature. Thus are getting maximum return from the money spent
it is important that we acquire suitable test on transmitters, receivers, and antennas. Nowhere
equipment, and the know-how to use it effectively. is this more true than in the vhf field.
This does not necessarily mean a laboratory full of Most amateur test equipment is essentially the
expensive gear. Much of what is needed can be same, regardless of the frequencies we intend to
provided at moderate cost, particularly if we are concentrate on. Things like volt -ohmmeters, oscil-
willing to take the time and trouble to make some loscopes, meters in general, the audio oscillator,
of it ourselves. the grid-dip meter - these are standard items that
All ham gear costs money, and the average we will not dwell on extensively in these pages.
amateur investment in equipment is rising every Rather, we will devote our space mostly to simple
year. Regrettably, this does not always apply to gadget-type items, tailored to the vhf man's needs.
the percentage of the total budget that is devoted We'll start with the simplest of them all.

LAMPS - POOR MAN'S METERS

Good meters are a very worthwhile investment. number of ways. Most of the lamps listed are
Every ham should have a few, but they do not have available in radio or even hardware stores. The
to be built into everything. Knowing what to test rating in amperes, right column, tells us that these
and measure, and being able to make a meaningful lamps draw 0.06, 0.1, 0.12, 0.15, 0.17, 0.2, 0.25,
interpretation of what the meters tell us, is more 0.4, and 0.5 amperes, at normal brilliance. These
important than having shelves full of expensive test figures (60 to 500 milliamperes) are close to ranges
equipment. An experienced and knowing ham can of milliammeters we would buy if we had an
do good work with a meter or two, a few pilot unlimited budget. Obviously, lamps can replace
lamps, neon bulbs, and some simple wavemeters. meters in many uncritical circuits where we need
There is no substitute for experience in this merely an approximate indication of current being
department, and the way to get the experience and drawn. Lamps are fuses of a sort, too; they'll burn
knowhow is to start right in, building simple out if current goes very high for any reason.
equipment and making it work. Lamps serving as current and power -output
indicators are shown in Fig. 14-1A. Lamp Il is a
Current and Power Indicators plate -current "meter." Connected in the line from
More can be done with pilot lamps as current the plate circuit, it gives an indication of plate
and power indicators than most hams realize. The current being drawn. If the stage is supposed to
information of Table 14-I can be put to use in a draw 30 to 60 mA, a 2 -volt 60-mA pilot lamp (No.
48 or 49) will do. It can be mounted in a socket, to
make it easy to replace if burned out. You can
bum out a lot of them for the price of one good
meter.
Lamp I2 is a dummy load and power -output
indicator. The coil Ll is a single turn of insulated
hookup wire about one inch in diameter. With the
loop coupled to an rf circuit carrying power, the
lamp will light due to power absorbed. With a very
low -power circuit the loop can be hung over the
coil in the plate circuit, or inserted between its
Fig. 14-1 - Pilot lamps can be used as current and turns. For looser coupling, needed where the
power output indicators. Lamp 11 can substitute power output is much over 1/10 watt, the lamp
for a plate -current meter, where only a rough can be taped
indication is needed. 12, inductively coupled to the handle to to an insulating rod to be used as a
plate circuit, provides an indication of rf power hold the loop at the desired distance
output. Lamps can be calibrated roughly for from the tank circuit.
brilliance, as shown at B. Here a 2 -volt 60-mA pilot We can tell several things from these two lamps.
lamp is connected in series with a single flashlight They can be calibrated roughly for both current
cell and a 100-mA meter, for calibration purposes. and power from known dc sources. At a
current of

306
Lamps-Poor Man's Meters 307
50 mA, which it will draw from a 1.4 -volt cell TABLE 14-I - PILOT -LAMP DATA
connected as in 14-1B, a No. 48 or 49 bulb gives
off considerably less light than at its rated 60 mA, Lamp Bead Base Bulb RATING
but still enough to be clearly visible. From its No. Color (Miniature) Type Volts Amp.
normal rating, we know that full brilliance indi- 0.15
40 Brown Screw T-31/4 6-8
cates 120 milliwatts of power. We thus have four Bayonet 6-8 0.15
40A1 Brown T-31/4
useful references with which to work: 50 mA and 41 White Screw T-3''/ 2.5 0.5
60 mA of current, and 70 and 120 milliwatts of 42 Green Screw T -3'h 3.2
power. Similar references can be set up easily for 43 White Bayonet T -3'h 2.5 0.5
other lamps in the table. These are rough indica- 44 Blue Bayonet T-31/4 6-8 0.25
tions, of course, but there are plenty of places 45 Bayonet T-3'4 3.2
where they are good enough. 462 Blue Screw T-3'4 6-8 0.25
471 Brown Bayonet T-3''/ 6-9 0.15
Using Neon Lamps 48 Pink Screw T-3'h 2.0 0.06
493 Pink Bayonet T-3'Á 2.0 0.06
Neon lamps of various sizes are handy as 49A3 White Bayonet T-3'/ 2.1 0.12
indicators of rf fields, as they will ignite when 50 White Screw G-31/2 6-8 0.2
512 White Bayonet 6-8 0.2
merely held near a hot circuit, if the field is strong G-31/2

enough. Held close to a transmitter tank circuit, or 53

55
--
White
Bayonet
Bayonet
G-31/2
G-4'/
14.4
6-8
0.12
0.4
to the end of an antenna carrying rf power, they 2925 White Screw T -3''h 2.9 0.17
will glow with a brilliance and color that conveys 292A5 White Bayonet T -3'h 2.9 0.17
several bits of information. The brilliance is a 1455 Brown Screw G-5 18.0 0.25
relative rf power indicator, and the color is 1455A Brown Bayonet G-5 18.0 0.25
somewhat of an indication of frequency. At the 1487 -- Screw
Bayonet
T-3'/ 12-16
14
0.20
0.15
low end of the rf range the glow is a bright orange, 1488 T-31/4
1813 Bayonet T -3'h 14.4 0.10
changing gradually to purple with increasing radio Bayonet 12-16 0.20
1815 T-31/4
frequency. In the vhf range the neon glow is
markedly more toward the purple than the orange. 140A and 47 are interchangeable.
2Frosted bulb.
Glow color can thus be an indication of the 349 and 49A are interchangeable.
frequency range of a parasitic oscillation, as an 5Use in 2.5 -volt sets where regular bulb burns out
too frequently.
example. * White in G.E. and Sylvania; green in National Un-
A neon lamp or pilot lamp connected at the ion, Raytheon, and Tung -Sol.
center of a half-wave dipole gives an indication of **0.35 in G.E. and Sylvania; 0.5 in National Union,
Raytheon, and Tung -Sol.
relative power at some distance from a transmitting
antenna. The pilot lamp is the better for this
purpose, for the neon must be subjected to a fairly years ago in QST.1 It works on the principle of
high rf voltage to ignite the gas, whereas the separating the outgoing from the reflected power
filament of a lamp glows at once from the heat on a transmission line, by a combination of
being dissipated. inductive and capacitive coupling. The Twin -Lamp
Any of the lamps mentioned can be an aid to is merely two 60-mA pilot lamps connected to two
making a rough check on frequency, when absorp- coupling loops, and to one side of the transmission
tion wavemeters or Lecher wires are used. We'll line, as shown in Fig. 14-2. The schematic circuit is
cover these applications later in this section. given in A, and the mechanical details in B.
Neons or pilot lamps can be used to give a The Twin -Lamp can be made up permanently
rough indication of standing -wave ratio on open - attached to a short section of Twin -Lead, which
wire or other balanced transmission line. A lamp can be inserted in the transmission line as needed,
with a coupling loop attached, or a neon bulb held or left permanently connected. A similar arrange-
close to the line, will brighten or dim in proportion ment can be made for open -wire lines. Length of
to the rf voltage along the line. Two lamps, one at the loops, Ll and L2, depends on the frequency
the minimum and another at the maximum point and power level. Two to four inches in length
of rf voltage, thus provide a direct indication of 1For this and other numbered references, see
SWR that can be roughly calibrated against brill- bibliography at the end of this chapter.
iance indications from a similar lamp in a metered
dc circuit.

The Twin -Lamp SWR Indicator

A fine example of the use of lamps to do a
much -needed job is the Twin-Lamp SWR indicator
originally described by vhf enthusiast W4HVV

Fig. 14-2 - The Twin -Lamp ís a simple stand-

ing -wave indicator for use with balanced line.
Lamps I1 and 12, in the schematic diagram, A, are
2-volt 60-mA pilot lamps, with their tips soldered
together and connected to one side of the 300 -ohm
line. Loops L1 and L2 can be made from short LSolSe.uL
`----Transmission Line
Soldered J (B)
pieces of Twin -Lead, as shown at B.
 308 TEST EQUIPMENT FOR THE VHF STATION
should be suitable for vhf work at average power glow, and with a bad mismatch the two will be
levels. about the same brilliance. The device should not be
The lamp nearest the transmitter end will light thought of in terms of measurement of SWR, but it
when power is fed to the line. If the SWR is low, does give a usable indication of adjustment, and it
the lamp at the other end will remain dark. At tells whether the SWR is high or low, which is all
moderate ratios the lamp on the antenna side will that is really important ín most instances.

ABSORPTION WAVEMETERS
One of the most useful items of test equipment quencies go by so rapidly as the tuning capacitor is
in a vhf station is the absorption wavemeter. rotated that the frequency indication is necessarily
Fortunately, it is also simple and inexpensive. only approximate, but it is a very useful tool. Two
Anyone who does much work with vhf equipment of our wavemeters are this kind. The first tunes
should have a few of them. Three that take care of from 45 to 155 MHz, giving coverage of the
basic vhf requirements are shown in Figs. 14-3 to 50-MHz band, the 144 -MHz band, and other highly
6. useful spots such as 48, 72, 96, and 100 Mhz, as
A wavemeter can be a sophisticated device, but well. The other spans 113 to 255 MHz, passing
in its simplest form it is merely a tuned circuit through the 144- and 220 -MHz bands along the
whose resonant frequency for various capacitor way. The third is the bandspread type, having a
settings has been calibrated. When the wavemeter is fixed capacitor in parallel with the variable one.
coupled to a circuit carrying if power it will absorb Covering 120 to 160 MHz, it spreads the 144 -MHz
some of the power when tuned through the band enough so that a fairly good indication of
frequency of the source. Indication that this is actual frequency is possible, if the wavemeter is
happening can be obtained in several ways. made, calibrated, and used with care.
If the stage to be checked is an oscillator,
tuning the wavemeter through its frequency may
cause a shift in frequency. This can be heard on a Construction
receiver tuned to the oscillator frequency. Any The wavemeter should be solidly built, to retain
current drawn by an oscillator or amplifier stage its calibration in normal handling and use. Ready-
will nearly always show some change, and the made coil stock provides the required rigidity, if
power output will dip, so any of these indications, connected with the absolute minimum of lead
or a combination of them, can be employed. The length. Some kind of indicating knob and scale
frequency of an oscillating detector can be checked arrangement is desirable, so that a permanent
with the wavemeter in a similar manner. The hiss calibration can be made. Construction should be
level of a superregenerative receiver will dip when such that the instrument can be held in the hands
the wavemeter tunes through its receiving frequen- without affecting the calibration, and used with
cy. The regenerative detector may go out of reasonable safety around live circuits.
oscillation, or the frequency of the beat note will Many arrangements are possible; those shown
change, due to wavemeter absorption. are merely one approach to the problem. Each is
The approximate frequency of any unwanted mounted on two pieces of aluminum, a handle and
oscillation in a transmitter or receiver can be a panel area. The handle is about 1 x 6 inches and
learned with the wavemeter in much the same way. the panel is roughly 2-1/2 inches in diameter. One
Check the current drawn by the oscillating stage, of the screws holding these pieces together has a
and watch for a quick change as the wavemeter Is soldering lug attached, with a clip lead soldered to
tuned. In a receiver, listen to the effects of the it for grounding the wavemeter frame, in the
oscillation, and note when they change abruptly in interest of safety. The variable capacitors are the
any way as the wavemeter is adjusted. type in which the rotor is grounded directly by the
Wavemeters can be made to cover side ranges of mounting shaft and nut, these being less subject to
frequency with rough calibration, or narrower hand -capacitance effects than variables with un-
ranges with more accurate calibration. We show grounded mounting studs. One type of coil stock
both types here. With the wide -range type, fre- (B&W Miniductor No. 3002, having No. 20 tinned

e j:e -
?Y

Fig. 14-3
the third
- Three simple wavemeters for vhf use. The two at the left have wide-range circuits, while
is a band -spread version. Each has a clip lead connected to the handle, for grounding to
chassis of the equipment being tested, in the interest of safety. the
Absorption Wavemeters 309
Fig. 14-4 - Rear
view of the wave -
meters, in the same
order as in Fig.
14-3.
3

wire, 1/2 inch in diameter, 8 turns per inch) is used Known frequencies of transmitter or receiver stages
in all three. may also be used. Calibration points throughout
The wavemeter for 50 and 144 MHz has a the range can be marked on one half of the
50-pF variable (Hammarlund HF-50) and 4-3/4 capacitor range, giving calibration points at 48, 50,
turns of coil stock. The bottom of the coil is 72, 90, 100, 120, and 140 MHz on Unit 1. The
soldered to the right stator post (as we look at the other half of the full circle of rotation can be
back of the wavemeter) with less than 1/8 inch of marked for 50 to 54 and 144 to 148 MHz, the two
exposed lead. The wire end of the top of the coil is amateur bands we're interested in.
7/8 inch long, its last 1/4 inch being soldered to The second unit was calibrated at 115, 125,
the rotor lug. The position of the coil is such that 150, 175, 200, 220, and 250 MHz on one half of
it is coaxial with the rotor shaft. Knobs are the scale, and the 144- and 220 -MHz bands on the
Johnson 116-222-5, though any indicating pointer other. The bandspread version has nicely spread
or dial-type knob will do. points at 120, 125, 130, 140, 150, and 160 MHz
The unit for 144 and 220 MHz has a 15-pF on one side and 144 to 148 on the other.
variable (Hammarlund HF-15) and roughly 2-1/2
turns of coil stock. The top edge of the coil lines
up with the center fo the rotor shaft in this model. Cs
The top lead is about 3/4 inch long, and the
bottom one substantially nil.
The band -spread model uses a double-spaced Fig. 14-5 - An absorption-type wavemeter can be
capacitor, originally 15 pF (Hammarlund HF-15X) the simplest form of tuned Circuit, as in L1 -C1, at
with all but the back two rotor plates removed. A. Circuit B has less capacitance in Cl, a fixed
The stator is left intact. The coil is practically capacitor, C2, in parallel, and a smaller value of
identical to that in the second unit, except that the inductance in L2, for a band -spread effect.
top lead is 1-1/8 inches long. The coil is connected
between the left stator post and the rotor lug. Laid If care is used in duplicating the originals, you
tightly against the ceramic end plate of Cl, and can copy the dial scales reproduced here, Fig. 14-6,
connected with the shortest possible leads between and you'll be close enough for ordinary work.
the stator post and rotor lug is a 5-pF ceramic Direct calibration with a reliable frequency source
fixed capacitor, C2. It is out of sight behind the is much better, of course. You very likely can
variable capacitor in Fig. 14-4. borrow a grid -dip meter for this purpose, if you do
not own one.
Calibration
Using the Wavemeters
Pieces of white card cut to 2 -1/4 -inch diameter
disks are mounted under the nuts holding the We have already described some of the uses for
capacitors to the frames. With the capacitor full the instruments, but here are practical examples.
out or full-in, the knob is set on the shaft so that Suppose you have a 50-MHz crystal oscillator. You
its indicator points straight up when the wavemeter know it is oscillating, but you're not sure whether
is held in the left hand. A mark is made on the it is on the right overtone. Or perhaps you want to
white card to indicate that this is one end of the check a frequency multiplier to see if it is
tuning range. Swinging the capacitor around 180 operating on the right harmonic. Connect a meter
degrees, another mark is made to indicate the other in the plate circuit, or in the grid circuit of the
end. following stage, if any. Lacking a meter, use a lamp
A calibration accurate enough for most pur- load or a neon indication, as in Fig. 14-1. Now,
poses can be made with a calibrated grid-dip meter. connect the wave -meter clip lead to a ground point

Fig. 14-6 - These wavemeter scales may

be copied, if construction
is exactly like
the original units. White cards are 2-1/4
inches in diameter. Calibration should be
regarded as rough only, until checked
against a source of known frequency
accuracy.
310 TEST EQUIPMENT FOR THE VHF STATION
in the equipment, and hold the wavemeter so that Watch out for live circuits. Remember that even
it couples to the tuned circuit in question. Rotate low voltages can be lethal. Large amounts of rf
the wavemeter dial slowly, watching the meter or power absorbed by the wavemeter may cause rf
output indication. There will be a quick dip bums on sensitive skin. Always be on the alert to
downward as the wavemeter tunes through the keep hands and other parts of the body away from
frequency at which the stage is operating. Hold the any circuits carrying ac or dc voltages. The
wavemeter as far as possible from the circuit being wavemeter circuits can be enclosed in transparent
checked, and still get the dip indication. This will plastic, if desired, but constant attention to the
show the frequency of the rf power with the best ever-present danger of high voltage is the only real
accuracy. protection.

SIMPLE FIELD -STRENGTH INDICATOR AND WAVEMETER

be copied with accuracy sufficient for most pur-
poses, if the parts and layout of the original are
duplicated. With values given the range is roughly
47 to 175 MHz, covering the 50- and 144 -MHz
bands with some leeway on both ends of the scale.
The various frequencies from 48 MHz up, com-
monly encountered in vhf transmitters, are also
marked on the white scale, making the meter
handy in checking oscillator and multiplier stages.
Our brethren on lower bands may find this gadget
useful for checking the frequency and source of
vhf parasitic oscillations or harmonics in their
equipment for the "dc bands."
A small plastic meter box with a sloping front
panel is used for a case. If a metal box is used, the
instrument will have less hand-capacity effect than
with a case of insulating material, but it will
interfere with its use as a simple wavemeter. The
y0. ground connection shown in broken lines in Fig.
No 14-8 is for a metal-box version, and represents the
47
o
t
Í7S
grounding of the rotor of Cl. Parts may be
arranged to suit the maker, so long as rf leads are
kept short. Note the copper -strap short across the
meter terminals. This should be kept in place
Fig. 14-7 -A field -strength indicator and wave -
meter for 50 and 144 MHz. The coil size is set so
whenever the plastic-box instrument is not in use,
to prevent damage to the meter in strong if fields.
that the tuning range is approximately 47 to 175
MHz.
Uses
Some sort of field-strength indicator is a must A rough idea of the gain, pattern, and front-to -
for vhf antenna work, and it can be useful in other back ratio of a vhf beam can be obtained by setting
ways. The instrument of Fig. 14-7 doubles as a the field-strength meter up at a distance out in
wavemeter. In this role it is more sensitive and
versatile than tuned -circuit absorption wavemeters 75
J,
just described. The diode CR1 rectifies rf current, CI
and the meter reads the resultant direct current.
You can get visual indications of rf field strength
and frequency without metering the circuits under ,Ti
470
examination. PROTECTIVE
SHORT
With a pickup antenna connected to J1, the
meter will indicate relative power being radiated by
a vhf antenna, for use in evaluation or adjustment.
Sensitivity depends on size of the pickup antenna, Fig. 14-8- Schematic diagram of the field -strength
meter. Ground connection shown in broken lines is
closeness of coupling between Ll and L2, type and for metal -case construction.
condition of the diode, and range of the indicating
meter. The lower the meter range the better. With
Cl - 75-pF miniature variable (Hammarlund APC-
75B).
a 100-µA meter the instrument will give full-scale
deflection at 100 feet or more from a vhf beam, if
CR1
J1 --- 1N34 or other vhf diode.
Insulated tip jack.
a half-wave pickup antenna is used. With a 1-mA L1 1 turn insulated hookup wire, 1/2 -inch
meter the separation is considerably less, for a diameter, spaced 1 turn from L2.
given transmitter power and antenna gain. L2 - 3 turns insulated hookup wire, 1/2 -Inch
The tuning range of Cl -L2 can be calibrated diameter, 1/4 inch long. Adjust turn spacing for
desired tuning range on C1.
with a grid -dip meter or any other source of rf Meter is 100 microamperes, though up to 1 mA is
power of known frequency, or the scale shown can usable, with reduced sensitivity.
Simple Field -Strength Indicator and Wavemeter 311
Fig. 14-9 - Rear view of the field -strength meter.
Note copper strap across meter terminals. This
should be kept in place when the instrument is not r
in use, to protect the meter.

front of the antenna under test and then perfor-

ming substitutions or adjustments. No diode is
linear, so changes in meter reading are only
indications, not accurate measurements. A favored
method of evaluation, with a device of this kind, is
to measure the power input to the antenna that
will give a constant reading as conditions are
changed. If power in the line can be read accur-
ately the power ratios thus obtained give a fair
indication of the results of the antenna adjust-
ments or changes. More on this in the chapter on
antennas. 4
In using the instrument as a wavemeter a
pickup antenna may or may not be required.
Usually an rf indication can be obtained by holding
the wavemeter near the circuit to be checked. Tune
the capacitor slowly until a meter indication begins
to show, then move the instrument away to keep
l
the meter from going off scale. If the diode is in a
strong enough rf field there will be rectified
current regardless of the setting of Cl. In such A sensitive microammeter is costly and fragile.
circumstances it is advisable to use a pickup Especially when working around high power, it is
antenna a few inches long or more and move the easy to run the 100-µA meter off scale without
instrument farther away, in order to prevent direct realizing it. To prevent damage to the meter it is
pickup by the diode. Varying the coupling between advisable to keep it shorted at all times when the
Ll and L2 is helpful in eliminating this blocking instrument is not in use. A strip of copper about
effect. Be sure to take all due precautions against 1/4 inch wide is kept wrapped around the meter
high-voltage shock and rf burns in using this and terminals for this purpose, and is removed only
other items of rf test equipment. when work is to be done.

REMOTE -INDICATING Fl ELD-STRENGTH METER

A limitation on the usefulness of a simple length. The meter box itself is made usable for
field -strength meter of the type shown in Fig. 14-7 other measuring jobs by building a switch and
is that when it is in a good position for doing its appropriate multiplier circuits into the case. As a
job the instrument may be far from the man who general-purpose test meter it can read 0 to 50 or 0
wants to observe its fluctuations as he makes
adjustments. The ideal spot for a field-strength
pickup device is as far out in front of the antenna
being worked on as possible, and preferably in
about the same plane. At this distance and height it
must be observed by a separate worker, or the man
making adjustments will have to use field glasses to IOC. tóooi
read it, if he can see it at all.
The logical remedy is to make the pickup
portion of a field-strength indicating device a
separating unit, put it up where it will do the job
` ,rA

required, and then run a line down to a meter at or

near the adjusting position. The field -strength
meter of Fig. 14-10 can be used with the meter
box and pickup unit plugged together, as shown, or d
connected by means of a cable of any desired
Fig. 14-10 - A vhf field -strength indicator and
general-purpose test instrument for dc measure- .o`"`"~ ^
ment. The small pickup unit can be plugged into
the indicator as shown, or connected to it by a
2 -wire cable of any desired length. Antenna rods or
wires plug into jacks at each end of the small box. i
The meter can be used with test leads for meas-
uring direct current and voltage. Ranges are 0 to 50 f
and 500 microamperes, 5, 50, and 500 milli-
amperes, and 0 to 500 volts dc.
 312 TEST EQUIPMENT FOR THE VHF STATION
Fig. 14-11 - Schematic diagram of the remote-
indicating field -strength meter. Resistors in the
indicatorunit should be 5 percent or better for
good meter accuracy.
i20
CR1 - Any small rf diode, 1 N34 or equivalent.
8.211 J1, J2 - Insulated tip jack (Johnson 105-601 to
611, depending on color).
IOC M
J3, J4 - Insulated "banana jack" (Johnson
108-901 to 907).
lOMrc
.15, .16 - Insulated tip jack and sleeve (Johnson
105-701 to 711).
P3 P4

PI P2
P1, P2 - Insulated tip plug, male (Johnson
105-601 to 611, with 5/8 inch No. 12 wire
soldered in jack).
P3, P4
311).
- Insulated tip plug (Johnson 105-301 to

J5 J6
R FC, RFC2
RFC1, RFC2 - 1 -watt resistor wound full length
with No. 30 enamel.
IN34
-2
J3

¡J4. J
S1 -section 6 -position switch.

4'7

to 500 microamperes, or 5, 50, or 500 milli- lower the meter range used, the more sensitive the
amperes, and 0 to 500 volts, dc. The multiplier instrument becomes. A fairly inexpensive imported
circuit also allows different levels of sensitivity in 50-microampere meter (AMD MRA-38) is shown.
the field-strength indicating role. With this meter it is possible to get a usable
As may be seen from the lower portion of Fig. indication of rf at up to several hundred feet away
14-11, the pickup unit, a 1-5/8 x 2-1/8 x 3 -1/4 - from a good beam, if a half -wave pickup antenna is
inch Minibox, contains only a diode, a resistor to plugged into 33 and J4. It is used with a 100 -foot
complete the dc circuit, a bypass capacitor, two rf cord by the author.
chokes, and two "banana jacks," one at each end There is really only one important rule in using
of the box. These take plug-in wires or rods, which the instrument: Be sure that you have the multi-
make up a half -wave dipole. Something shorter is plier switch, Si, set in such a position that the
usable, if maximum sensitivity is not important. meter movement will not be banged off -scale.
There are no tuned circuits, so the system will When there is the slightest doubt about the
work on any frequency. The only selectivity is the magnitude of the reading to be expected, start on
very slight amount introduced by the length of the the highest current range and work down. This
pickup dipole. applies whether you are reading plate current in a
On the back of the small case are two pin plugs. transmitter or using the instrument as a field -
These take matching pin jacks on the end of the strength meter. A 50-microampere meter can be
remote cable, or they can plug directly into the pin burned out in a tiny fraction of a second. Play
jacks on the 3 x 4 x 5 -inch Minibox used for the safe!
meter case. The two -conductor connecting cable The meter ranges other than 0-50 MA are
can be any available wire. Ordinary zip cord for obtained by introducing various resistances in the
home-appliance wiring is convenient. Coax or meter circuit, as shown in Fig. 14-11. This means
Twin -Lead that may have seen better days is also that shunts ín circuits being checked may affect
suitable. the accuracy of current readings. For more on
Layout of parts in both units is generally field-strength -meter techniques see the previous
uncritical, and obvious from the pictures. The item, and Chapter 8.

I` III Fig. 14-12 - Interiors of the

field -strength meter pick-up unit,
right, and meter box, left.
Remote -Indicating Field -Strength Meter 313
A TRANSMATCH FOR 50 AND
144 MHz WITH SWR INDICATOR

The antenna coupler (Transmatch) shown in

Fig. 14-13 will permit unbalanced transmitter
output lines (50-75 ohms) to be matched to
balanced feeders in the 300- to 450 -ohm imped-
ance range. Also, coax-to -coax matching Is possible
with this circuit, permitting 50 -ohm lines to be
matched to 75 -ohm lines, or vice versa. In situ-
*a
ations where a high SWR condition exists, this
coupler will enable the transmitter to look into a
matched load.
The built-in Monimatch-type SWR indicator
Fig. 14-13 - Antenna couplers for 50 and 144
MHz combined with an SWR and relative -power
enables the operator to tune the Transmatch for indicator, in one package.
minimum reflected power, assuring a good match
between the transmitter and the feed line. Remem-
ber that the use of devices of this kind will not
correct for any mismatch that exists at the antenna be possible by relying upon the physical contact
end of the line. Although it assures a good match provided by the shaft bushings.
between the transmitter and the line, it can only Make coil taps by bending No. 6 solder lugs
disguise the fact that a mismatch exists at the around the coil wire at the proper spots, then
antenna. soldering the lugs in place. No. 20 wire is used to
connect the taps of Ll to jacks J1 and J2. A short
The Circuit piece of 300 -ohm twin line connects the taps of L4
Balanced circuits are used for both bands. to J4 and J5. A No. 6 solder lug is bolted to the
Butterfly capacitors are employed for good circuit outside (back) of the cabinet as near to J1 as
symmetry. The links of tuned circuits L2 and L3 in possible. Another such lug is placed adjacent to J4.
Fig. 14-14 are series-tuned to tune out reactance in When operating coax -to -coax style, a short jumper
the line. Switch Si transfers the SWR bridge wire connects J1 to its ground lug, or J4 to its
element from one tuned circuit to the other, ground lug, depending on the band being operated.
providing visual indication of the matching adjust- The jumper must be removed for balanced-feeder
ments. Section SIB shorts out the unused tuned operation.
circuit to prevent interaction. Switch S2 selects
either the forward- or reflected -power sampling The Bridge Element
circuits from the bridge and supplies their rectified The SWR element is of the Monimatch variety.
dc voltages to R1, the meter sensitivity control The circuit is given in Fig. 14-14A, with its
which is adjusted for full-scale meter reading when physical layout shown in B. The inner line, L6, is a
S2 is set to read forward power. 4 -inch length of 1/4 -inch OD copper tubing. One
end of L6 is soldered directly t" the center lug of
Construction J7, the remaining end supported by a small
A homemade 12 x 5 x 5 -inch aluminum cab- standoff insulator. The line is mounted in plastic
inet is used. (See September, 1966, QST, p. 17 for blocks for additional support, making sure that it is
construction details.) If a similar layout is followed centered within the walls of L8, the aluminum
keeping all leads as short as practical, the complete outer channel. J7 should be mounted on the back
unit can be housed in a commercially available wall of the box so as to be centered on the axis of
chassis or cabinet. The rf tuning controls are L6 when it is in position. The pickup lines, L5 and
mounted in a straight line across the front of the L7, are made from No. 16 wire, each 3 inches in
cabinet. The SWR bridge element is bolted to the length, and are spaced 1/8 inch away from L6,
bottom of the case (inside) between the input jack, being supported by the plastic blocks. Once they
J7, and the band -change switch, Si. Shielded audio are in place, a drop of Duco cement should be
cable is used to connect the output of the bridge to added at each point where they pass through the
the lugs on S2. Short lengths of RG-58/U coax plastic blocks.
cable connect L2 and L3 to S1A. The shield braids The 150-ohm terminating resistors (1/2 -watt
of both cables should be grounded to the chassis at units) are mounted inside the channel, L8, and are
each end. soldered to ground lugs. Diodes CR1 and CR2
A 2 -lug terminal strip is bolted to the chassis attach to the remaining ends of wires and are
directly under the center of LI. Similarly, a second routed out through small holes in the walls of L8.
terminal strip with two lugs is mounted under the It is important that the physical placement of the
midpoint of L4. These strips serve as mounting diodes, the resistors, and the pickup wires be
points for links L2 and L3. No. 12 wire connects symmetrical. The better the symmetry, the better
the rotors of all four tuning capacitors in to one will be the balance of the bridge, electrically. The
another. The ground bus is also connected to the diodes and their related 0.001-µF bypass capacitors
main chassis at one point. This procedure assures a are attached to small terminal strips that are
better ground return for the capacitors than might mounted near the holes in L8. If matched resistors
 314 TEST EQUIPMENT FOR THE VHF STATION

50-54 MHz 144-148 MHz

J1 J4
300- 300-
450 450
OHM OHM SAL
51A L3
SAL INPUT
INPUT 35
J2 J5

50-75 J3 J6 50-75
OHM
UNBAL (A1
L8 150
OHM
UNBAL
INPUT INPUT
A01 I

IL5 L7
CR2

J7 TO
50-75
OHM TRANS
OUTPUT
REF
OOStf
25KS ENS

(A)
No.16-GAUGE

o 3 OALUMINUM
A A
TB
fBEND ¡,
4"

8,,
,
}
A
o 1
A=Ne.33DRILL PLASTIC BLOCK
8=41ií DRILL (2 USED)
L8 LAYOUT (B)
Fig. 14-14 - At A, the schematic diagram of the vhf Transmatch. Capacitance is in pF unless
otherwise noted. Resistance is in ohms, K = 1000. At B, physical layout of the bridge element and the
plastic insulating blocks.
Cl - 26-pF per section butterfly (E. F. Johnson
167-22).
L3 - 2 turns -1/2
No. 14 enam. or spaghetti-covered
bare wire, over center of L4.
C2
C3
-
-
100-pF miniature variable (Millen 20100).
35-pF miniature variable (Millen 20035).
L4 - 5 turns No.-inch10 dia,copper
1
wire, -inch dia,
1

spaced one wire thickness between turns. Tap

C4 - 10-pF per section butterfly (E. F. Johnson 1-1/2 turns from each end.
167-21). L5 -3
-inch length of No. 16 solid wire.
CR1, CR2 --
J1 -J4, incl.
Germanium diode, 1N34A or equiv.
Insulated binding post.
L6 -4 -inch length of 1/4 -inch dia copper tubing.
-
L7 Same L5.
J5 -J7, Incl.
- - SO -239-style chassis connector. L8 - See drawing.
as

L1 7 turns No. 10 copper wire, 1 -1/2 -inch dia,

spaced one wire thickness between turns. Tap
R1
S1
--25,000 -ohm control, linear taper.
2 -pole 2 -position rotary, single section,
2-1/2 turns from each end. phenolic switch
L2 - 2 turns No. 14 enam. or spaghetti -covered
bare wire, 1/2-inch dia, over center of L1.
S2 - Spst rotary, (Centralab 1462).
single section, phenolic switch
(Centralab 1460).

and matched diodes are used in the bridge circuit, meter sensitivity. Apply power from the trans-
electrical balance will be even better than is mitter - low power until initial tuning is com-
possible with random-selected components. Since pleted - and adjust R1 for full-scale meter reading.
the bridge is a relative-reading instrument, the For 144 -MHz operation, tune C3 and C4 in the
latter condition is not vital. same manner. Repeat the tuning until no further
reduction in reflected power is possible. The meter
Operation should fall to zero, indicating a 1:1 match. Switch
S2 back to the forward position and set R1 for a
Attach the vhf transmitter to J7 wíth a short full-scale meter reading. No further adjustments
length of coax cable. Connect a balanced feeder to will be needed unless the transmitter frequency is
J1 and J2 (for 50 -MHz operation), or to J4 and J5 changed appreciably. The tuning procedure is
(for 144 -MHz operation). Set S1 to the desired identical for matching coax to coax. In doing so,
band position and switch S2 to read forward however, the antenna feed line (coax) is connected
power. Initially, R1 should be set for minimum to either J3 or J6 and the shorting strap (discussed
 A Vhf Impedance Bridge 315

.i
1_ I

1
¡161

K. 11-j i

t1

Fig. 14-15 - Inside view of the Transmatch. The 6 -meter circuit is at the left, the SWR bridge element
is at the center, and the 2 -meter circuit is to the right of the bridge element. The meter, S2, and R1
are at the far right.

earlier) must be connected to J1 or J4. In some Reduced power (less than 50 watts) should be used
situations, it may be possible to get a better match during initial tuneup, thus preventing parts from
by leaving the shorting strap off. being damaged by heating or arcing. The coupler
After the coupler is tuned up, the transmitter should never be operated without a load connected
power can be increased to its normal level. This to its output terminals. Such operation will usually
unit will handle power levels up to 500 watts destroy the 150 -ohm resistors and the diodes, CR1
(transmitter output power) provided the coupler is and CR2, in addition to causing arcs in the
tuned for a matched condition at all times. Transmatch.

A VHF IMPEDANCE BRIDGE

Itis often helpful to be able to measure
impedances of input and output circuits of vhf
converters, transmitters, cavity filters, dummy
loads, and antennas. Most impedance -matching
devices described in amateur literature are in-
effective above about 30 MHz, because of in-
accuracies resulting from excessive internal capaci-
tance and inductance. The bridge of Figs. 14-16
through 18 was tailored to the needs of the vhf
operator. Minimum lead inductance and com-
ponents chosen for vhf qualities permit reasonable
accuracy in the vhf range. ,
An SWR bridge can be used to secure a proper
match between circuits, or between a circuit and
its load, but this does not allow measurement of
the terminal impedance when a mismatch is pre-
sent. This impedance bridge will enable the user to
make direct readings of impedance in the 10 -to -
500 -ohm range, thus permitting the solution of a
variety of matching problems
Construction
It is suggested that the builder duplicate, as
nearly as possible, the physical layout. Flashing
Fig. 14-16 - Simple vhf impedance bridge. Settings
of the variable control, marked on the front of the
copper 1/4 inch wide is used for leads in the bridge case, represent values of impedance corresponding
portion of the circuit, to keep down lead in- to nulls in meter indication.
 316 TEST EQUIPMENT FOR THE VHF STATION
bridge meter reading, making certain that the null
is as deep as possible. Mark the dial scale, and jot
down the value of the load resistor used. Repeat
this process with various resistor values unitl a
complete set of calibration points has been ob-
tained. A new scale can now be made and marked
permanently with India ink, indicating the resis-
tance values at the check points.

Uses
The more nearly a load is to being purely
resistive, the deeper the null will be. As a load
i becomes reactive, as frequently happens in rf
measurements, the null is less pronounced. Even
though the null is poor, the readings remain useful,
Fig. 14-17 - Interior of the impedance bridge.
Inside the shield, right, are the coaxial terminals,
as they approximate the actual value of the
impedance under measurement.
the variable control, R3, and the load resistor, R2. Vhf matching networks may be checked by
Feedthrough insulators may be improvised from presetting the bridge to the desired value of
1/4 -inch Teflon rod. Leads thereto are 1/4 -inch -
resistance, and then making adjustments to the
wide copper strips. Load resistor R2 is connected
directly between the terminals of J1 and J2. matching netwoik until the null is obtained on the
bridge meter. Adjust carefully for best null. A
Gamma-match system can be adjusted without a
ductance. A shield of copper, brass, or aluminum transmitter in this way. Insert a length of coax, of
divides the metering and bridge circuits. The the impedance of the line to be used, between the
potentiometer R3 is a 2 -watt carbon control, with antenna and the bridge. This line should be a
linear taper, so that the resistance scale will not be multiple of a half wavelength long electrically, so
cramped at one end. It is mounted on an insulating that the antenna impedance will be repeated at the
plate, rather than on the metal box, to reduce
output terminal. This cable permits the operator to
capacitance between the chassis and the metal take readings without being in the immediate field
shield on the control. The insulated mounting plate
of the antenna. A 2 -wavelength section should be
is then fastened to the minibox, with the control
about right for work in the 144 -MHz band. If the
bushing centered in a 5/8 -inch hole. coax is solid-dielectric (not foam) RG-8 or similar,
The bridge resistor RI is a 5 -percent 1 -watt a test cable of 107 inches will do.
carbon. It is made of two 1000 -ohm 1/2 -watt Other devices such as the coaxial -line filters of
resistors in parallel, in the example shown. R2 Chapter 15 may be checked for input and output
should be 50 ohms, a value not readily obtainable. impedance by using line sections that are multiples
This is within the possible range of 47-ohm of a half wavelength, and a dummy load of the
10 -percent tolerance resistors, so a stock of these correct impedance. The filter input or output
can be checked for the nearest to the desired 50 circuit can then be adjusted for a null at the
ohms. The value of a resistor can afso be raised by desired impedance. Transmitter or converter input
fling into the carbon element. Be sure that the or output circuits can be adjusted in a similar
ohmmeter used for checking the resistance value is manner. Unknown values of impedance can be
reliable, if accuracy is desired.
The jacks J1 and J2 are mounted as closely
together as their flanges permit. The meter is a
100 -microampere type, for maximum sensitivity
when using the bridge with rf sources of low power
output. It is an inexpensive 1 -3/4 -inch square
imported model (Calrad).

Checking and Calibration

A pointer knob on R3 and a paper dial scale
pasted to the end of the Minibox provide for
calibration. Select a number of 1 -watt carbon
resistors of values between 10 and 500 ohms, to be
Fig. 14-18 - Schematic diagram of the vhf
impedance bridge.
used as calibration loads for the bridge. Attach the -
Cl 0.002-1.1.F disk ceramic.
link L1 to the input jack J1. Couple a grid -dip C2 - 0.001-µF disk ceramic.
meter or some other low -power source of rf power CR1 - 1N82A diode. 1N34 also usable.
to the link. Use a frequency around 145 MHz If
best accuracy is wanted in the 2 -meter band. L1
P1
-- -
Ji, J2 Coaxial connector, SO-239.
1 turn No. 12 enamel, 1 -inch diameter.

Coaxial plug, PL -259.

Attach a low -value resistor to 32 with the
shortest possible leads. Adjust the coupling be-
tween the rf source and L1 for full-scale reading on
R1
R2
R3
---500 -ohm carbon, 1 watt.
50-ohm carbon, 1 watt.
500 -ohm 2 -watt control, linear taper (Allen
the bridge meter, or as high a reading as possible if Bradley). Case is 2-1/4 X 2-1/4 x 4 -inch Mini -
less than full-scale. Next, adjust R3 for a null in the box.
An Inexpensive Directional Coupler 317
determined by attaching the bridge to the circuit frequency stability of the transmitter is an advan-
being tested, and then sweeping across the bridge tage, but be sure that it has output only in the
range for a null. desired frequency range. Use of a coaxial filter in
Though a grid-dip meter is mentioned as a the transmitter output may be desirable, in order
power source, a low -powered transmitter can, of to prevent harmonics and subharmonics from
course, be used in these various applications. The reaching the bridge.

AN INEXPENSIVE DIR ECTIONAL COUPLER

Precision in-line metering devices that are capa-
ble of reading forward and reflected power over a
wide range of frequencies are very useful in
amateur vhf and uhf work, but their rather high
cost puts them out of the reach of many vhf
enthusiasts. This device by Tom McMullen, W1SL,
is an inexpensive adaptation of their basic prin-
ciples. You can make it yourself for the cost of a
meter, a few small parts, and bits of copper pipe
.
and fittings that can be found in the plumbing
stocks at many hardware stores. It can be as
accurate as you want to make it. _A
Such an instrument is known by several names.
directional coupler, in -line wattmeter, SWR indi-
cator, and so on. W1SL calls his a "line sampler."2 ,rt

It can be left in the antenna line at all times,

without consuming appreciable power or affecting
reception. Commercially available units are also
simple. It is their precision workmanship and
reliable calibration over wide frequency ranges that
make them costly to buy, though a joy to own.
Fig. 14-19 -
Two versions of the line sampler. The
single unit described in detail herewith is in the
foreground. Two sections in a single assembly,
Construction made by W1NTH, provide for monitoring forward
The sampler consists of a short section of and reflected power without probe reversal.
hand -made coaxial line, In this instance of 50 ohms
impedance, with a reversible probe coupled to it. A ical and electrical bond will result. If a torch is
small pickup loop built into the probe is termin- used, go easy with the heat, as an over -heated and
ated with a resistor at one end and a diode at the discolored fitting will not accept solder well.
other. The resistor matches the Impedance of the The inside of the line section must be smooth
loop, not the impedance of the line section. Energy and without gaps or burrs. Any protrusions into
picked up by the loop is rectified by the diode, and the line must be removed with a file or reamer. If a
the resultant current is fed to a meter equipped tubing cutter is used, apply minimum pressure, or
with a calibration control. large burrs will be formed and will be hard to
The principal metal parts of the device are a remove.
brass plumbing T, a pipe cap, short pieces of Coaxial connectors with Teflon or other heat -
3/4 -inch ID and 5/16 -inch OD copper pipe, and resistant insulation are recommended. Type N,
two coaxial fittings. Other available tubing com- with split -ring retainers for the center conductors,
binations for 50 -ohm line may be usable. The are preferred. Pry the split -ring washers out with a
ratio of outer-conductor ID to inner-conductor OD knife point or small screwdriver. Don't lose them,
should be 2.4/1. For a sampler to be used with as they'll be needed in the final assembly.
other impedances of transmission line, see Fig.
8-28 for suitable ratios of conductor sizes. The
photographs and Figs. 14-19 and 20 just about tell
the rest of the story.
Soldering of the large parts can be done with a
300 -watt iron or a small torch. A neat job can be
done if the inside of the T and the outside of the
pipe are tinned before assembling. When the pieces
are reheated and pushed together, a good mechan -

Fig. 14-20 - Major components of the line

sampler. The brass T and two end sections are at
the back of the picture. A completed probe
assembly is at the right. The N connectors have
their center pins removed. The pins are shown with
one inserted in the left end of the inner conductor
and the other lying in the right foreground.
 318 TEST EQUIPMENT FOR THE VHF STATION
CALIBRATE
1000pF 50K
C2

1000 R2 _ 0-100

500pF
Fig. 14-21 - Circuit diagram for the line sampler.
Cl - 500-pF feedthrough capacitor, solder-in type.
J1 CR1 J2
C2
type.
-1000-pF feedthrough capacitor, threaded
CR1 -
Germanium diode 1N34, 1N60, 1N270,
1 N295, or similar.

J1, J2 -
Coaxial connector, type N (UG-58 A/U).

The inner conductor is prepared by making

L1 -
Pickup loop, copper strap 1 inch long X 3/16
inch wide. Bend into "C" shape with flat
eight radial cuts in one end, using a coping saw portion 5/8 -inch long.
with a fine-toothed blade, to a depth of 1/2 inch. M1 -
0-100/1A meter.
The fingers so made are then bent together, R1 -Composition resistor, 82 to 100 ohms. See
text.
forming a tapered end, as seen in Fig. 14-22.
Solder the center pin of a coaxial fitting into this,
R3 - 50,000 -ohm composition control, linear
taper.
again being careful not to overheat the work.
In preparation for soldering the body of the inside the pipe. The disk has four small tabs around
coax connector to the copper pipe, it is convenient the edge for soldering inside the pipe. The diode,
to use a similar fitting clamped into a vise as a CR1, is connected between one end of the loop
holding fixture, with the T assembly resting on and a 500-pF feedthrough capacitor, Cl, soldered
top, held in place by its own weight. Use the into the disk. The terminating resistor, R1, is
partially prepared center conductor to insure that connected between the other end of the loop and
the coax connector is concentric with the outer ground, as directly as possible.
conductor. After being sure that the ends of the When the disk assembly is completed, insert it
pipe are cut exactly perpendicular to the axis, into the pipe, apply heat to the outside, and solder
apply heat to the coax fitting, using just enough so the tabs in place by melting solder into the
that a smooth fillet of solder can be formed where assembly at the tabs. The position of the loop with
the flange and pipe meet. respect to the end of the pipe will determine the
Before completing the center conductor, check sensitivity of a given probe. For power levels up to
its length. It should clear the inner surface of the 200 watts the loop should extend beyond the face
connector by the thickness of the split ring on the of the pipe about 5/32 inch. For use at higher
center pin. File to length, if necessary, slot as with power levels the loop should protrude only 3/32
the other end, and solder the center pin in place. inch. For operation with very low power levels the
The fitting can now be soldered onto the pipe, to probe position can be determined by experiment.
complete the 50-ohm line section. The decoupling resistor, R2, and feedthrough
The probe assembly is made from a 1 -1/2 -inch capacitor, C2, can be connected, and the pipe cap
length of the copper pipe, with a pipe cap on the put in place. The threaded portion of the capacitor
top to support the upper feedthrough capacitor, extends through the cap. Put a solder lug over it
C2. The coupling loop is mounted by means of before tightening its nut in place. Fasten the cap
small Teflon standoffs on a copper disk, cut to fit with two small screws that go into threaded holes
in the pipe.
FEEDTHROUGH
CAPACITOR
SOLDER LUG Calibration
PIPE CAP
The sampler is very useful for many jobs, even
if it is not accurately calibrated, though it is
11/2- INCH LENGTH desirable to calibrate it against a wattmeter of
COPPER PIPE LOCKING SCREW known accuracy. A good 50 -ohm dummy load is a
must. (More on loads elsewhere in this chapter.)
3/4 -INCH
BRASS PIPE T
11/4 -INCH LENGTH OF
COPPER PIPE,3,4-INCH I

TEFLON STANDOFF
TYPE N COAXIAL
/CONNECTOR

Fig. 14-22 - Cross-section view of

the line sampler. The pickup loop is
supported by two Teflon standoff
CENTER
'SPLIT -RING" insulators. The probe body is se-
SOLDER CO DUCTOR RETAINER cured in place with one or more
INCH 0 D SOLDER locking screws through holes in the
PICKUP
LOOP brass T.
Silicon Diode Noise Generator 319
The first step is to adjust the inductance of the On higher bands the meter reading will be higher
loop or the value of the terminating resistor, for for a given power level, and it will be lower for
lowest reflected-power reading. The loop is the lower-frequency bands. Calibration for two or
easier to change. Filing it to reduce its width will three adjacent bands can be achieved by making
increase its impedance. Increasing the cross-section the probe depth adjustable, with stops or marks to
of the loop will lower it, and this can be done by aid in resetting for a given band. And, of course,
coating it with solder. When the reflected -power more probes can be made, with each calibrated for
reading is reduced as far as possible, reverse the a given band, as is done in some of the com-
probe and calibrate for forward power, by in- mercially available units.
creasing the transmitter power output in steps and Other sizes of pipe and fittings can be used, by
making a graph of the meter readings obtained. Use making use of information given in Chapter 8 to
the calibration control, R3, to set the maximum select conductor sizes required for the desired
reading. impedances. (Since it is occasionally possible to
pick up good bargains in 72 -ohm line, you might
like to make up a sampler for this impedance.)
Variations Type N fittings were used because of their
Rather than use one sampler for monitoring constant impedance, and their ease of assembly.
both forward and reflected power by repeatedly Most have the split -ring retainer, which is simple to
reversing the probe, it is better to make two use in this application. Some have a crimping
assemblies by mounting two T fittings end-to -end, method, as do apparently all BNC connectors. If a
using one for forward and one for reflected power. fitting must be used that cannot be taken apart,
The meter can be switched between the probes, or drill a hole large enough to clear a soldering iron
two meters can be used. tip in the copper -pipe outer conductor. A hole of
The sampler described was calibrated at 146 up to 3/8 -inch diameter will have very little effect
MHz, as it was intended for 2 -meter repeater use. on the operation of the sampler.

SILICON DIODE NO ISE GENERATOR

One of the most useful tools in adjusting vhf Ordinary disk ceramics are not suitable above 50
receivers is a noise generator.3 In its simplest form a MHz.
noise generator is a diode drawing current, and The coaxial plug should match the connectors
therefore making noise. This noise extends all on the receiving equipment with which the gener-
across the rf spectrum, up to a frequency deter- ator is to be most often used. The PL-259 plug
mined mainly by the circuitry and the diode used. with UG176/U adapter was used here. The flange
Such a crystal -diode noise generator is shown in on the adapter is only very slightly larger in
Figs. 14-23 through 25. Noise figure cannot be diameter than the threaded portion, so using it as a
"measured" with a device of this kind, but it is means of clamping the assembly to the generator
handy as a noise source for adjusting a vhf receiver case is not very satisfactory. To give more binding
for best noise figure. The lower the diode current surface, washers of flashing copper were made for
for a given margin of diode noise over receiver both sides of the mounting hole. One is shown in
noise, the better the receiver is working. Fig. 14-25. Cut with shears from one edge to the
washer hole, and bend the washer at the break
Construction slightly, so that it can be threade -nto the adapter
The noise generator is built in an aluminum sleeve. Use one of these washers on each side of the
Minibox 3-1/4 x 2-1/8 x 1-5/8 inches in size (Bud box, which must be drilled for a 3/8 -inch hole to
CU -3001-A). Only the load resistor, R1, the diode, pass the threaded portion of the adapter. One
CR1, and the bypass capacitor, Cl, are critical as washer can be soldered to the adapter flange and
to mounting position. These should be connected the other to the end of the plug sleeve, to make the
with absolutely minimum leads, if the generator is whole assembly less likely to work loose in using
to be useful above 100 MHz or so. R1 is inside the the noise generator.
adapter sleeve of Pl. Cl should be a button -mica The battery is the 9 -volt type commonly used
or other capacitor having good uhf dharacteristics. in small transistor radios. A mercury battery is

p
_

7 A tivic

Fig. 14-23 -Silicon -diode noise

generator for vhf receiver testing,
with its audio detector, left, for
smoothing out noise readings. Tip
jacks in the noise generator per-
mit taking diode current readings.
 320 TEST EQUIPMENT FOR THE VHF STATION

C 1- Fig. 14-24 -
Interior of the diode noise generator.
The diode, load resistor and bypass capacitor
should be connected with the shortest possible
leads.
mercury type are used, but the diode current is
directly related to noise output, and can be
calibrated roughly in noise figure.
The load resistor, Rl, should be a value equal
to the line impedance of the antenna system to be
used. If a 51 -ohm resistor is not available, 47 ohms
is close enough for ordinary purposes in work with
52-ohm antenna systems. Use a 75 -ohm resistor if
worth the difference in price as its voltage will the line is 72 ohms. A 68 -ohm load may be used if
remain practically constant throughout its useful both 52- and 72 -ohm lines are to be encountered.
life. This makes possible a reasonably accurate
The value is not particularlyaitical, as the noise
calibration of the generator's noise output in terms
generator is not a precise test instrument.
of the setting of the series control, R2.
Connection to the battery is made by means of Using the Generator
a terminal block removed from the top of a dead
battery. The fiber insulating plate on which the In receiver work the generator is best connected
terminals are mounted is fastened to the front wall directly to the receiver or converter antenna jack.
of the box. The smaller of its terminals is grounded Run the receiver with its avc off, if possible. Turn
to the case, and the other is backed up by two up the rf and audio gain controls until receiver
layers of plastic insulating tape. The assembly thus noise is heard. Keep the rf gain control as low as
acts as both connector and mounting plate for the possible, in order to prevent overload, except
battery. where the gain of the first stage in the receiving
The large end of the diode is held in a plate clip system is affected by the gain control setting. This
of the type used for metal tubes. The smaller end is rare in vhf receiving setups, but it may be found
contact was removed from an old octal wafer in lower band gear. Note the level of the receiver
socket. Many different diodes are usable if the noise, by ear or by connecting an ac voltmeter, dB
upper limit of frequency is not important. Silicon meter, or the audio detector described later, across
rectifier diodes of the kind used in power-supply the speaker or earphone terminals. Now turn on
work can be used at 50 or 144 MHz, but they draw the noise generator, starting with R2 at its maxi-
considerable current, and do not work well at mum resistance setting. If no increase in noise is
higher frequencies. Uhf mixer diodes of the 1N21 heard, reduce the resistance slowly until noise
series are recommended. Also used with good begins to rise.
results: 1N25 and 1N32. Germanium diodes such Use a noise increase that you can remember or
as the 1N34 are not satisfactory. A good diode will measure. The lower the crystal current required to
give plenty of noise with no more than about 2 mA give this noise increase, the better the receiver. This
diode current. You may want to put a fixed gives a rough comparison of one vhf converter to
resistor in series with R2 to keep the current below another, provided that they are for the same band.
this point, if all your work is going to be with Preferably, the converters so compared should be
receiver front ends known to be in quite good used with the same receiver. Adjustment of an-
working order. tenna coupling, oscillator injection, tests on various
The tip jacks, J1 and J2, are for measuring tubes or transistors, or checking any other factor
current through the diode. This will run about 10 affecting receiver performance, can be done with a
mA maximum with the 1N21 series diodes, but noise generator of this type. Line loss in a length of
may be higher with other types. The setting of R2 coaxial line can also be measured, by connecting
is not meaningful when batteries other than the the line between the noise generator and the
CRI converter, noting the difference in noise through
TIP BASE Hsi:-j the line, and with direct connection of the gener-
ator.
Rz SHEAR In receivers having no provision for removing
CUT avc, some other method of measuring noise output
oi
500
must be made. The receiver S meter can be used, if
2 Si
9v
it responds to the receiver's noise level with the
generator turned off. If it doesn't, it may be
Fig. 14-25 -
Schematic diagram of the noise necessary to run the noise-generator output higher
generator. Two washers, right, are used to mount than normally would be the case, in order to get a
the coaxial plug to the case. meter rise indication on generator noise.
-
Cl 500-pF button mica. Any receiver adjustment that makes it possible
CR1 --
J1, J2
Silicon mixer diode, 1N21, etc.
Insulated típ jack.
to obtain a given noise increase with a lower diode

---
P1 Coaxial plug, PL-259, with UG-176 adapter.
current, or a greater S -meter reading increase with
S1 Toggle or pushbutton switch. the same level of diode current, is an improvement.
R1 51 -ohm 1/2 -watt carbon, mounted inside As shown, the noise generator produces in
adapter sleeve of P1. excess of 20 dB of noise at 50 and 144 MHz, and
R2 - 50,000 -ohm control. enough to be usable with any fairly good receiver
Low -Power Dummy Loads 321
at 220 and 420 MHz. The amount of noise and the
upper useful frequency limit depend on the diode
used, and on its condition. Avoid subjecting the
diode to strong rf fields, or to excessive current. If
you buy a good diode, it will probably come
encased in metal foil, or otherwise shielded. Keep
it that way until it is installed in the noise Fig. 14-26 - Schematic diagram of the average -
type audio detector shown in Fig. 14-23. Parts
generator, and then use the lowest diode current arrangement is uncritical.
that will give satisfactory noise output. -
J1, J2, J3, J4 Tip jack.
A refinement some users like is to substitute a
pushbutton switch for the toggle type for Si. If
T1 - Small audio output transformer. Low -
impedance winding connects to J1, J2.
this is the microswitch type it can be closed with
light finger pressure, making it easy to take Originally described by KIDJP in QST,4 it is
readings without the likelihood of disturbing the merely connected to the speaker or earphone
setting of the diode -current control, R2. terminals of the receiver, and the meter is then
connected to its output terminals. The detector is
Audio Detector for Noise -Generator Work shown schematically in Fig. 14-26. Use is the same
In using a vacuum -tube voltmeter or other ac as if the meter were connected to the receiver
output meter in noise -generator work the erratic directly. The above reference is well worth reading
nature of the meter indication is often a problem. by anyone interested in improving receiver perfor-
Noise is random in nature, and unless the meter is mance.
highly damped, the needle will fluctuate con- Though the audio detector is shown with the
stantly, making it difficult to establish a reference. crystal-diode noise generator, it is helpful with
The device shown with the silicon -diode noise other types of noise generators, wherever the
generator in Fig. 14-23 is a simple audio detector flickering of the meter indication may be trouble-
to smooth out such meter readings. some.

LOW -POWER D UMMY LOADS

In order to test a transmitter legally, an Lamps in between these power levels are highly
amateur must use a dummy load. Most of us put reactive, and the impedances they represent vary
our rigs on the air for test purposes for brief greatly with the power being dissipated. Low -
periods, but the considerable running of a trans- wattage incandescent lamps, 15 to 40 watts, are
mitter that is usually required during construction particularly poor. Lamps larger than the 100 -watt
and trouble -shooting should not be done with the size nearly always develop "hot spots" in a portion
transmitter feeding an antenna. When an antenna of their filaments, making them unreliable as
test is unavoidable, the operator should be sure power indicators. They also tend to go gassy and
that the frequency to be occupied is not being used burn out before reaching their normal power level.
for communication at the time. If an appreciable Lamps of intermediate wattage can be improved
transmitter -on time is contemplated, a non- somewhat as rf loads by connecting a variable
radiating load is the only considerate (and legal) capacitor in series to tune out the reactance they
approach to testing. and their leads represent. See Fig. 14-27A.
Once it was the usual thing to hook a lamp of Put this combination on the output of your
suitable wattage across the output terminals. This SWR bridge, and feed in some rf power. Tune the
may still suffice, but it leaves much to be desired in capacitor for lowest reflected power, not maxi-
most instances. Few lamps or combinations thereof mum lamp brilliance. Readjust transmitter output
come anywhere near to being 50 -ohm loads, in the coupling for maximum output, after the load is
vhf range, and consequently they may be all but matched as well as you can get it. A maximum
useless for any meaningful testing. Exceptions are value of 100 pF should serve for 50 or 144 -MHz
some small pilot lamps (within their limited power - loads. Use 50 pF if nothing lower than 144 MHz is
handling capabilities), and incandescent lamps of to be used, and 25 pF is enough for higher bands,
around 100 -watt rating. Several blue -bead pilot if the combination will work at all.
lamps in parallel may make a fair vhf load, and The resistive load, B, is much better, within its
some 100 -watt lamps singly or in parallel are usable power capabilities. Two such dummy loads using
at frequencies in the vhf range. paralleled resistors are shown in Fig. 14-28. Three

LITD
(A) (B) (C)
Fig. 14-27 - Three dummy loads for vhf use. Lamp In A has variable capacitor in series to tune out
reactance. See text. Load B has carbon resistors in parallel. Total resistance should equal impedance
the transmitter is designed to work into. A considerable length of lossy coax, C, makes an excellent rf
load. Since the line dissipates most of the power, the load resistor need not be of high wattage rating.
 322 TEST EQUIPMENT FOR THE VHF STATION
flashing copper are used, and the resistors mounted
in a circle, to keep inductance down. Nine 470 -
ohm resistors are used in the unit pictured, but
thirteen 680-ohm or nineteen 1000-ohm resistors
r' would probably do equally well.
These are not perfect loads, but they are much
better than lamps. They show no greater than

I s;
L.7
IS

/4 . \
>
1.2:1 SWR at 50, 144, or 220 MHz, with the best
match (close to 1:1) at 50 MHz. Don't take the
wattage ratings of the resistors too literally. They
get warm at a dissipation of 1 watt each, and when
they get more than just warm to the touch, the
resistance value may begin to change. This is not
too much of a problem ordinarily, as the power is
usually on for only a fraction of a minute at a
Fig. 14-28 - Two vhf dummy loads of the type time, just long enough to take a
shown schematically in Fig. 14-27B. Copper fins reading or make a
on unit at right aid in dissipating heat, and disks
quick adjustment. The cooling fins on the unit at
provide low-inductance parallel connectors. the right help to keep the heat flowing out of the
resistors, and it will dissipate more than 10 watts
considerations are important here: the resistors safely for brief periods; no more than a few
must be the composition (carbon) type, the induc- seconds at a time.
tance of the leads must be kept to an absolute The load at C is the best of all. In fact, the
minimum, and the power to be dissipated should principle is used in some of the best rf wattmeters
be kept below the rated total wattage of the and dummy loads made. If you've got 100 feet of
resistors used, except for brief tests. coax that is too lossy for use on an antenna, don't
Any number of resistors can be connected in throw it away; it is an ideal dummy load. Looking
parallel, so long as they are all the same wattage back to Table 8 -III, we see that 100 feet of
and the same resistance. Six 330 -ohm 1 -watt RG58/U (even in new condition!) has a loss of 6
carbon resistors are paralleled by soldering their dB per 100 feet at 144 MHz. This means that you
ends to straps of flashing copper in the unit at the can put the circular load of Fig. 14-28 on the end
left. One strap solders to the coaxial fitting sleeve, of it, feed 40 watts into the other end, and the
and the other is narrowed down to fit inside the resistors will just reach their rated dissipation. At
sleeve and make contact to the pin. The portion 420 MHz you could run 60 watts into the line and
inside the fitting is covered with insulating spa- it wouldn't hurt the load. You can even short the
ghetti or plastic tape, to prevent shorts. end of the line, or leave it open, and it will make
A somewhat better load, where higher dissi- hardly any difference in the SWR reading at 432
pation is wanted, is seen at the right. Here disks of MHz.
A load of this kind is a good match at any
frequency where the loss is 6 dB or more, and its
power -handling capability is considerable, for short

( I >l
T test periods. If the coax can be coiled loosely and
subjected to a cooling air blast, it can be made to
take just about any amateur power for short
periods.
\1 Improved Resistive Load for 432 MHz
The resistor -bank load of Fig. 14-29 has a
detachable tuned circuit, which can be connected
to improve its performance in the 420 -MHz range.
Six 330 -ohm 1 -watt resistors are soldered between
the centerconductdr tip and the edges of the
flange of an SO -239 coaxial fitting. Without tuning
the load is close to a pure resistance at 50 MHz,
but it becomes more reactive at progressively.
higher frequencies. At 144 MHz the indicated SWR
is 1.3:1. It rises to about 1.5:1 at 220 MHz, and to
1.9:1 at 432. None of these figures is too bad for
most purposes, but it is possible to make a
near-perfect load out of the resistor bank, by
adding the tuned circuit shown.

Fig. 14-29 - 5 -watt dummy load, with detachable

circuit for tuning out the reactance of the load at
432 MHz. The trimmer is disconnected from the
resistor junction when the load is used on bands
below 420 MHz.
Frequency Measurement with Lecher Wires 323
The loop of wire and trimmer capacitor provide connected, tune the trimmer for minimum re-
means for cancelling out the reactance of the flected-power indication. -

resistors and their leads at 432 MHz. The LC Similar tuning can be used for lower frequen-
circuit can work on only one band conveniently, cies, but turns of wire, rather than a small loop,
and it is really needed only at 432, so a connection may be needed. Or tuned loads can be made for
between the trimmer and the resistor junction is each band, for accurate work.
made when an accurate match is needed for It is well to check a batch of resistors with an
420 -MHz band work. The loop is 2-1/4 inches of ohmmeter before selecting those to be used in such
No. 16 wire, bent into C shape, and soldered to a load. Six resistors actually 330 ohms each would
one corner of the flange. The top end is soldered to give a load of 55 ohms. About 20 330 -ohm
the movable plate of the trimmer. The stationary resistors were measured with án ohmmeter and the
plate is soldered to the center pin, or not, as lowest 6 of them were used. The resulting total
needed. When using the load with the circuit resistance of the load is 53 ohms, very close to the
figure needed.
FREQUENCY MEASUREME NT WITH LECHER WIRES
Here is a measuring instrument that is almost as easily and inexpensively in numerous ways, one of
old as radio communication, but it is still a handy which is shown in our practical example, Figs.
item for the vhf or uhf experimenter. The length of 14-30 and 31.
an electromagnetic wave can be measured directly The construction requires little explanation,
on a transmission line, by observing the distance and dimensions are not critical. The base of the
between points of maximum or minimum rf assembly is made from two straight pieces of
1 x 2 -inch pine, fastened together in a T-shaped
voltage. Lecher Wires are a means of doing this
reliably and accurately. If you are accustomed to cross-section, and supported on two blocks of
using metric scales you can read wavelength di- wood. The anchors for the measuring line are of
rectly, the distance between the voltage peaks or similar material. The wires are held tight with
nodes being a half wavelength at the frequency turnbuckles at the left end, and are supported on
being measured. (A meter is 39.37 inches, which as insulators at the right end.
every vhf man knows is a half wavelength at 150 How long you make the line depends on the
MHz.) lowest frequency you want to measure. The model
shown is 7 feet long, which will take care of
Construction measurement from the 144-MHz band up well into
the microwave region. If you want to start at the
The wires in the portion of the instrument to 220 -MHz band an overall length of about four feet
be used for measurement must be without insula- will suffice.
ting material in direct contact. Provision must be A rough Lecher -Wire measurement of wave-
made for holding the wires taut, and in uniform length can be made by running a knife or screw-
spacing. The shorting device must make firm driver blade along any bare -wire transmission line,
contact, and the distance between peaks (or nulls) but if you want to measure accurately something
must be measured precisely, if accurate measuring like the shorting blade and carrier shown here must
is to be undertaken. These objectives can be met
be incorporated, to give repeatable results. The

-Turnbuckles Fig. 14-30 - Structural details of a Lecher Wire

system for measuring wavelength from the vhf
range up into the microwaves. Wood parts are all
1 X 2 -inch pine. Enlarged view, left, shows the
sliding shorting assembly and its metric scale.

Shorting blade
}
Metric Scale
Insulators
Coupling
loop
 324 TEST EQUIPMENT FOR THE VHF STATION
Fig. 14-31 -Close-up view of the
coupling end of the Lecher Wire
assembly. Overall length depends
on how low the user wishes to be
1111011.
able to go in frequency.

block rides on the base strip, with two metal side circuit so as to couple some energy from it. The
plates keeping it in alignment. These plates need loosest coupling that will work is the best.
not be metal, but it is convenient that way. At the Now, run the sliding short along the Lecher
right end of the travelling block is a notched metal Wires, watching for a change in the indicator,
short. The wires are kept tight enough so that a whatever it may be. When the change occurs, note
good electrical contact is made by this plate. the reading on the base and block scale. Let's say
The base is marked off in tenths of a meter, it's 0.255 meter. Now move the carrier along until
beginning at a point directly under the coupling a second dip is found, and note the scale reading.
end of the line. The travelling short has a trans- Suppose it is 0.937. Subtract the first reading from
parent metric scale (most stationery stores have the second, the answer in our example being 0.682
them) fastened to its underside, so that readings meter. This is a half wavelength at the frequency
can be taken directly in metric units of length. being measured. To convert this to frequency in
(Inches and feet don't enter into this at all.) megahertz, divide 150 by the wavelength just
Now we have to couple to the rf source in some measured. Our answer shows that we have just
way. The propagation factor of the coupling line is missed the 220-MHz band, and we're on 219.9
of no importance, so ít can be Twin-Lead or M.Hz.
anything in the way of a balanced line that may be Making use of the Lecher-Wire principle does
handy. From here on the operation is much like not always require a system specially built for this
that with an absorption-type wavemeter, except purpose. Any two parallel wires or rods with one
that the Lecher Wires are much more accurate if end coupled to an rf power source can be pressed
made and used properly. into service, so long as there is little or no solid
Measuring Wavelength and Frequency insulation between them, in the area used for
measurement. Just running a screwdriver along the
The energy source can have any of several wires, and noting the approximate position of
indicators: a grid current meter, plate current peaks or nulls, will serve to show, for example,
meter, rf voltmeter, field -strength meter, lamp whether the right harmonic is being picked off in a
load, or whatever. The coupling loop of Twin - multiplier system. The assembly described is capa-
Lead, at the right end of the Lecher Wires, is ble of quite accurate frequency measurement,
shorted at the end. This loop is placed near the rf which is quite a different matter.

UHF SLOT TED LINE

A slotted line, as its name implies, is merely a If a diode is used for rectifying the rf current,
section of transmission line, fitted with a movable varying readings on the meter in series with the
probe to permit sampling the rf voltage along the diode show only that there is some SWR on the
line, through at least one maximum and one line. When you adjust the load or matching device
minimum point. If more peaks can be checked, so for the least change in reading with probe move-
much the better, but you can get by with just one ment, you have made the best adjustment you can,
"max and min." At least two of one and one of the with the matching system or load in use. The
other is better, and the mínimum length for a good numbers themselves are not meaningful otherwise,
420 -MHz slotted line is about three feet. Some until the device is calibrated in some way.
useful work can be done with a 3 -footer on 220 or The line shown was made from information
even 144, however, if the length of line feeding given by Pitt Arnold, WOIPE, in a talk at a Central
into it is adjusted so that at least one peak and one States Vhf Conference some years ago. Admittedly
valley can be observed. this is a "make -do" device, not to be compared
In addition to measuring standing -wave ratio, with laboratory -type slotted lines in accuracy of
the slotted line is useful for measuring wavelength measurement, but it is a highly useful and in-
in the manner of Lecher wires. The scale mounted structive tool in many ways. Its sensitivity in SWR
on the front of the instrument is for this purpose. measurement is markedly better than that of even
Uhf Slotted Line 325
Fig. 14-32 -Input end of the slotted
line. Peaks or nulls of diode current,
indicated on the meter as the probe is
moved along the line, show half -wave
intervals. Scale measures half -wave-
lengths in inches, or in centimeters if a
suitable scale is available. Three dummy
loads are shown in the foreground: a
sealed resistor load in a BNC fitting, left;
a tuned frequency -sensitive load, center;
and a No. 47 pilot lamp and fitting,
right.

quite expensive equipment of the calibrated - Rubber feet with 8-32 screws are used to fasten
wattmeter type. the base plate to the side plates. The holes in the
flanges of the latter are tapped for 8-32 thread.
Construction A 36 -inch scale mounted along the front
surface of the line measures the distance between
The inner conductor is half -inch OD aluminum voltage peaks or valleys. A millimeter scale simpli-
tubing, centered between two plates about four fies the mathematics involved in frequency meas-
inches high. The space between the inner surfaces urement. The scale shown in a replacement unit for
of the plates is 0.9 inch. We used two L-shaped side use in a pocket tape, and was originally 6 feet long.
plates, bolted to a base plate. All are 3/32 -inch It is mounted on standoff washers, so that the edge
sheet aluminum. A better arrangement might be to of the probe mount will slide just inside its top
use thicker side plates, with tapped holes in their
edge.
bottom edges. The principal item of importance is The size of the coupling loop on the probe and
that the assembly be sturdy enough so that the
its position with respect to the inner conductor of
spacing between the side plates will remain con- the line will depend on the power input to the line,
stant when the instrument is used. on the sensitivity of the meter used. If a
If solid rod is available for the inner conductor, and self-controlled oscillator is used as an energy source
the ends can be drilled and the center pins of the it is desirable to couple as loosely between the
coaxial fittings inserted therein. BNC or N-type source and the line as possible. Loose coupling
fittings can be connected to the line in this way. between the probe and the line is also desirable, as
Do not use the so-called uhf fitting, SO-239. BNC too much coupling will broaden out the peak and
fittings are convenient for use of the instrument in valley indications and make frequency measure-
demonstrations. Solid rod of the right size was not ments inaccurate. Make the loop on the probe of
at hand, so half-inch aluminum tubing was used. such size and shape that it can be used for close
Thin strips of brass about 3/8 inch wide were coupling, and then bend it up to operate with the
bolted vertically to both sides of each end of the loosest coupling that will give a satisfactory meter
tubing, bent toward each other, and soldered to
indication. See Fig. 14-35.
the center conductor of the coaxial fitting. That it
was not possible to eliminate rise and fall of rf If a self-controlled energy source such as a uhf
current completely shows that this makeshift is not dipper is used with the line, it can be coupled to a
small loop of insulated wire, mounted in a BNC
perfect, but in spite of this the instrument is more
sensitive to SWR variations than other uhf test fitting and plugged into the input end of the line.
With a transmitter having 50 -ohm output, a coaxial
equipment available to the average amateur.
The end plates and the mounting for the probe cable is used between the transmitter and the line
are made of 1/32 -inch sheet aluminum, bent in U input. When the transmitter has one watt or more
shape, with BNC sockets (UG-290/U) mounted output the coupling between the line and the
with their diagonals perpendicular to the long axis probe can be quite loose. As the instrument is
of the plate, and centered in the large surface. The actually used you will find it possible to work out
tip -to -tip dimension of this fitting's flange is just optimum coupling for various purposes. For the
time being, a loop of the type shown and described
over 0.9 inch, so the corners should be filed down
to make the assembly a smooth fit over the top or will be adequate for most uses. For greatest
end surface of the line. Occasional light applica- versatility, make up several probes.
tions of silicone -base lubricant will permit the
probe assembly to slide easily along the top of the Uses
line. The end plates are made in a similar manner, A good way to become familiar with the slotted
and are fastened to the front and back surfaces of line is to check dummy loads at 432 MHz. The first
the line with self-tapping screws. The dimensions requirement for this is that the energy fed into the
of these parts are given in Fig. 14-33. Mounting line be free of harmonics. Use a coaxial or strip -line
holes need be drilled only in the end plates, not in filter between the rf source and the line. Filtering
the probe assembly. is especially important if the energy source is a
 326 TEST EQUIPMENT FOR THE VHF STATION
36

SIDES ALUMINUM
(A) 3/32 BEND UP 90

(B) BASE 3/S2 :ALUMINUM

37/
e Fig. 14-33 - Details of the principal metal parts of
the uhf slotted line. Schematic diagrams and parts
s/
information for the probe and meter assembly are

(C)
MOUNT
BNC
1-r 3/O
BENO9Ó
T
2348'
given in the lower portion. The two sides, A,
should be mounted on the base, B, so that the
FITTING
spacing between the inner walls is 0.9 inch. The
PROBE AND END PLATES 3/ .,
probe and end plates should be bent to provide a
o
I/32.ALUMINUM
o B sliding fit over the line assembly. The inner
conductor, not shown, is 1/2-inch diameter, 35
inches long.
METER ASSEMBLY
CR1 - Any uhf diode.
BNC LI L1 -4 -inch strip of 3/16 -inch thin copper or brass,
PROBE
bent into loop about 1-1/2 X 3/4 inch ín size.
See photograph and text.

A much better one-watt load can be made by

varactor multiplier. A simple varactor system is soldering a 51 -ohm 1 -watt composition resistor
almost certain to have enough power at unwanted inside a BNC fitting. Cut a copper or brass disk the
frequencies to foul up the slotted line indications. size of the end of the fitting, and drill a hole that
Start with a load that will represent some will just pass the resistor lead, at the center of the
mismatch. A No. 47 pilot light soldered into a BNC disk. Solder the resistor lead to the disk, and then
fitting will do, and it has the advantage of giving solder the disk to the end of the fitting. With this
some visible indication of relative power. Put such load in place you'll probably find not much more
a load on the output end of the slotted line, and than 10 percent variation in probe current along
connect the input to the energy source capable of the line, at 432 MHz, and perhaps less at 220 or
about one watt output at 432 MHz. A 2 -meter 144 MHz. This is a pretty fair load, but remember
transmitter of 2 to 5 watts output, a varactor that its dissipation capability is limited. Do not 'run
multiplier and a strip-line filter (such as described it at one-watt input or more for any longer than
in this Manual) will make a fine beginner's demon- necessary to get readings. Continued heating of the
stration setup. resistor is likely to cause it to change value and
Adjust the input to the line so that the lamp destroy the usefulness of the load.
shows some glow. Now insert the probe and watch A good load, but one that is frequency sensi-
the meter indication carefully. It may be found tive, can be made by mounting the resistor in the
that with the coupling loop as shown in the fitting, with a variable capacitor in series. Select
photograph the meter will go off scale before the the resistor, if possible, so that its do resistance will
probe is resting in its normal operating position. If be between 50 and 52 ohms. Adjust the capacitor
so, bend the loop back until a satisfactory reading with an insulated screwdriver until a setting is
is obtained. Now slide the probe slowly along the found that shows the smallest variation in current
line, and record the maximum and minimum as the probe is moved along the line. The best
current readings. You'll probably find that the observed with this arrangement is a variation of
meter will go to zero at the minimum rf voltage about 2 microamperes in a reading of about 50 µA.
points, when the coupling is adjusted for about a Use readings that are available at some distance in
3/4-scale reading at the peaks. Obviously, a No. 47 from the ends of the line for this adjustment, as
lamp is not a very good load at 432 MHz! coupling to the line changes appreciably in the last
few inches at each end.
Wavelength measurements should now be tried.
The distance between two nulls or two peaks of
probe current is a half-wavelength. Be careful not
to overcouple with the probe for this work, or the
indication will be excessively broad, or even
double -humped. If a self-controlled energy source
is used, tight coupling will pull the frequency
enough to render measurements highly inaccurate.

-
Fig. 14-34- Interior of one end of the slotted line.
Strips of thin brass, barely visible here, are bolted
to the center conductor and soldered to the BNC
fitting centered on the end plate of the line.
Uhf Slotted Line 327
A uhf oscillator, set at an unknown frequency,
can be coupled to the input end of the line by
means of a 1 -inch diameter loop of wire soldered
into a BNC fitting. Operate the slotted line without
a load, or with a load having an impedance other
than 50 ohms, so that an appreciable SWR will be
seen with the probe. Measure the distance between
nulls. Bend the probe until zero reading is ob-
tained. Now move slowly each way until one
microampere is seen on the meter. Observe these
two 1-µA points, and take the midpoint between
them. Now move along the line to the next null,
and find its exact location in the same manner. We
find the points to be 12-5/8 inches apart, let's say. Fig. 14-35 -
Underside of the slotted -line probe.
Position of the coupling loop with respect to the
The frequency of oscillation is then found from
the formula: center conductor should be adjusted for the
minimum coupling usable at the power level of the
5905 energy fed into the line, if maximum accuracy is to
F- S be achieved. A light coating of silicone -base lub-
ricant on the inner edges will assure smooth
where F is the frequency in MHz and S is the operation and good electrical contact.
spacing of the nulls in inches. So
5905 With a 100-µA meter and a single probe
F= - 468 MHz position with respect to the inner conductor of the
12.625
line the highest SWR that could be measured was
If you have a millimeter scale the numbers are just over 3:1. At this point the max -and-min
simpler, for readings were 100 and zero. The best SWR
300,000 sensitivity that could be indicated reliably was
F= 2S represented by meter readings of 100 and 95.
Making a rough graph of the range where reflected -
where S is the spacing of the nulls in meters. A
power wattmeter readings and slotted-line meter
millimeter scale would show these same nulls to be
readings could be compared indicates that reason-
32 millimeters apart, which ís a half wavelength.
able SWR-reading accuracy down to at least 1.07
to 1 is not difficult, with this addmittedly imper-
F=300,000_ 468 MHz
fect instrument.
2 x 32
If coupling is held sufficiently loose that nulls can How important is it to be able to read SWR
be read accurately, frequencies in the vicinity of
under 1.2:1? Not very, if all you are interested in is
the 420 -MHz band should be well within one getting an antenna to work as well as it is practical
percent when measured in this way. to make it. But being able to see low values is
important, if you are attempting to measure
Calibration antenna performance. Forward -power readings
needed to achieve certain reference readings on a
Because of the characteristics of the diode, the field -strength meter can be translated into gain
relative currents indicated on the meter are not measurements in experimental work with antennas,
directly translatable into SWR. The meter reads if the reflected power is down very close to zero.
rectified current, whereas the SWR is the ratio of Unless it is, all such readings are suspect. The
>maa to /min in rf current. If we have a wattmeter slotted line will let you measure closer to zero-
capable of reading rf power output with a fair
reflected than any conventional in-line meter sys-
degree of accuracy we can make a calibration of tem.
the dc meter readings in terms of rf power. This
was done using a Bird Thruline Wattmeter, Model BIBLIOGRAPHY
43, with a 10-E (400 to 1000 MHz) plug-in unit.
1 Wright, "The Twin -Lamp," QST, October,
With this instrument, and one plug-in unit, the 1947, p. 22.
highest forward -reflector power ratio that can be 2McMullen, "The Line Sampler," QST, April,
read with appreciable accuracy is 100 to 1 (10 1972.
watts forward, 0.1 reflected) corresponding to an 3 Tilton, "Noise Generators-Their Uses and
SWR of 1.22:1. A dummy load was deliberately Limitations," QST, July, 1953, p. 10. Detailed
mismatched to give this 1 -percent reflected-power information on vacuum -tube noise generators is.
condition, using the wattmeter. The load was then also in the ARRL Handbook, 41st edition and
checked with the slotted line probe adjusted to later.
give a maximum current indication of 100 µA. The Noise generator information was brought up to
date and equipment for 420 MHz and higher
minimum was measured at 65 µA, showing that the frequencies was described in a three-part sympo-
slotted -line was capable of indicating SWR values sium, QST, February; 1964, pp. 23-35.
far lower than could be measured by other means 4Frye, "Adjustment Procedures for V.H.F. Con-
available. verters." QST, October, 1958, p. 24.
 Chapter 15

Interference Causes and Cures

In one respect amateur radio is vastly different neighbor troubles are beneficial in our vhf receiving
from most other hobby -type activities: from its
situations as well.
earliest days it has existed in competition with In handling TVI and related interference prob-
other services. Because we occupy frequencies that lems two cardinal points should be kept in mind:
are under constant pressure from other users of
radio, amicable and successful solution of our Interference is primarily a public -relations
interference problems is vital to our very existence. problem, not a technical one. Every form of it
With occupancy of the radio spectrum rising can be cured; it is getting the job done amicably
daily, electronic devices of infinite variety in that is difficult.
almost every home, and population density in- Being able to demonstrate that the trans-
creasing almost everywhere, interference problems mitter is not at fault is important, but it is not
inevitably tend to multiply. Interference is a enough. The amateur must understand the
two-way affair. We both cause it and suffer from it, factors involved, and be able to take or recom-
but the first is our major concern in these pages. mend corrective measures. Nobody is going to
Fortunately some of the steps we take to cure our do this for us.

THE NATURE OF TVI

With these facts of community life established, 4) Harmonics of Oscillator or Exciter Frequen-
let's look at the causes and cures of TVI, the vhf cies Usually this shows up as a crosshatch pattern,
man's major interference problem. The principal independent of modulation, changing or disappear-
forms of TVI from vhf transmitters are listed ing when the transmitter frequency is shifted.
below, in the approximate order of their impor- Usually trouble develops only when the harmonic
tance: falls in a sensitive part of the TV channel. See Fig.
1) Blocking Every user of the 50-MHz band in 15-1 and 2. Examples of common combinations
a Channel 2 area knows about this. It may run all are the 9th harmonic of 6-MHz stages and the 7th
the way from a light cross -hatching of the picture harmonic of 8 -MHz ones, falling in Channel 2; the
to complete blackout. Nearly always the visual 10th harmonic of 8 MHz in Channel 6; 7th
effects change with transmitter modulation. Usu- harmonic of 25 MHz in Channel 7; 4th harmonic
ally there is audio interference along with the of 48 MHz in Channel 9 or 10, depending on the
picture trouble. The vhf man's worst interference operating frequency. There are others, but these
problem, it has been responsible for a high per- are the most common sources of trouble. Exciter
centage of all complaints reaching FCC in recent frequencies may also get directly into the receiver's
years. It is much worse in Channel 2 than on higher i -f system, in some instances.
channels, but it is possible on all channels, 2 5) Final -Stage Harmonics The 4th harmonic of
through 6, in receivers near to a 50 -MHz station. 50 MHz falls in Channels 11 to 13, depending on
Blocking of Channel 13 by 220 -MHz energy is the operating frequency. Various harmonics of 50,
similar, but by no means so severe. 144, and 220 MHz fall in the uhf TV band, though
2) Audio Troubles Rf pickup by the detector usually they are not strong enough to cause much
or audio circuits of a receiver (TV or radio) results trouble. The 2nd harmonic of 50 MHz falls in the
in your voice riding through almost regardless of fm broadcast band, and while this did not give us
the tuning of the receiver or even the setting of the much trouble in the past, increasing use of fm is
audio gain control. Picture reception may be clear. changing this picture.
This problem is not confined to TV receivers; it is There are many other possible sources of TVI
common in all devices having audio amplifiers: from the operation of amateur vhf transmitters,
hearing aids, hi-fi and public-address systems, but it is safe to say that at least 95 percent of our
record and tape players, musical instrument ampli- problems are covered by the above list. Items 1 and
fiers, and so on. 2 are by far the most troublesome. We will look
3) Image Response This basic weakness of all into each in some detail, but first go back over the
superheterodyne receivers is explained in Chapter list and be sure that you understand each one.
3. As most commonly encountered in amateur vhf Then remember, despite all you may have heard,
circles it is responsible for reception of 2 -meter these interference problems can be solved.
signals in Channel 2, in TV receivers having the The frequencies assigned to television and fm
currently used high intermediate frequency. If an broadcasting in the vhf range are shown in Fig.
old TV set with a 21 -MHz i-f does not show 15-1, together with the harmonics of the 50 -MHz
interference from your 144-M1Iz transmitter, and a band that fall in this range. Prevention of radiation
newer one does, this is likely to be the cause. in these assignments, Items 4 and 5, is the
Along with it there will probably be some audio amateur's responsibility. Items 1, 2, and 3 are
trouble (2), if you are using amplitude -modulated receiver defects, which must be corrected at the
phone. receiver.

328
 Correcting TV Receiver Deficiencies 329
66 MHz 766111 82114: 88 MHz 100MHz loe MHz
50MHz 54 MHz 60MHz 726112
[4lr-
AMATEUR TV 2 TV 3 TV 4 TV 5 TV 6 FM BROADCAST.

5 V S V S V
V i 2nd Hormonle
of 50MHz

1Mo MHz 112 MHz 198MHz 204 MHz 210MHz 216MHz 220MHz 225MHz
174 MHz 180MHz

-
,..
TV7 TV 8 TV 9 TV 10 TV II 1A-;TV 12
.
J Py.l'W/p;h
a`S
TV 13,5 AMATEUR

5 V 5 V 5 V 5 VI f _-- 5 v 5 V 51

1¡ -4th
1

Hormonic of 50MHz
I

Fig. 15-1 -
Frequencies assigned to vhf television and fm broadcasting in the United States. The
approximate positions of the video and sound carriers are indicated on each TV channel. Crosshatched
areas show second and fourth harmonics of the 50 -MHz band. Positions of the amateur 50- and
220 -MHz bands with respect to the TV channels are also indicated.

It is of utmost importance that the amateur measures to be taken at the receiver. The earlier
know how to recognize the nature of the inter- the amateur gets into the matter the better, for
ference, so that he can correct the trouble if it lies friendly relations between him and the set owner
in his transmitter, or recommend the corrective are vital, if a solution is to be reached.

CORRECTING TV RECEIVER DEFICIENCIES

Fundamental blocking (1) is a receiver problem. nearly all TV sets are. With such receivers a signal
The rf circuits of a TV set are broad in frequency 4.5 MHz below the picture carrier can cause severe
response. If they were not, picture quality would sound interference, the severity depending on the
suffer, since the television picture and its sound selectivity and alignment of the TV receiver's i -f
occupy a channel 6 MHz wide. It is not easy to system. 55.25 MHz 4.5 MHz = 50.75 MHz. This -
build a receiver that will pass 54 to 60 MHz and makes 50.75 MHz the worst possible spot on which
reject rf from a nearby amateur station operating to operate in the 50 -MHz band with a -m phone,
on 50 to 54 MHz. It is unlikely that TV manu- from the standpoint of sound interference. Proper
facturers will ever mass-produce receivers that do it alignment of the TV set makes a big difference
effectively. with this trouble,. but a practical fact of life in a
Though blocking troubles are much more severe Channel 2 area is that staying well away from
in Channel 2, if the 50 -MHz signal is strong enough 50.75 MHz is very desirable. Any operating fre-
it may block the receiver on all low-band channels, quency above 50.4 MHz makes interference very
2 through 6. Clearing any channel above 2 is much more likely.
usually done quite readily. The simple stub to be The approximate range over which 50 -MHz
described later will nearly always handle it. Why signals are likely to overload TV receivers in
clearing Channel 2 of 50 -MHz interference is more Channel 2 was shown graphically in QST by
difficult is obvious from Fig. 15-2, which shows 50-MHz pioneer W2IDZ. Fig. 15-3 is from his
the locations of the sound and picture carriers in a now -classic treatment of the 50 -MHz TVI problem
TV channel, as well as the interference potential of which paved the way for today's thousands of
any signal falling in the channel. In Channel 2 the 50 -MHz enthusiasts who now manage to live with
picture carrier is at 55.25 MHz, which is just too Channel 2.1 This is a matter of effective radiated
close to the 50 -MHz amateur signals to make it a power. If your transmitter puts out 100 watts, and
simple matter to keep the latter below the over- your antenna has a gain of 10 dB (10 times) your
loading level. ERP is 1000 watts. If the antenna's main lobe fires
There is a related problem in connection with into the TV antennas your sphere of evil influence
50 -MHz interference to the TV sound in Channel will be roughly 400 feet in radius.
2, in receivers of the intercarrier sound type, which 1For this and other numbered references, see
bibliography at the end of this chapter.

l 1.25 Mc.
PICTURE
CARRIER
SOUND
CARRIER
2i!
é Mc

MEGACYCLES FROM LOW EDGE OF TV CHANNEL

Fig. 15-2
relative
- Locations of the video and sound carriers in black -and -white TV signal, showing
severity of interference caused by harmonics falling in various parts of the channel. This
a

information can be put to good use by vhf amateurs in instances where mild harmonic interference
is encountered. The trouble may be corrected, or at least alleviated considerably, by shifting the
operating frequency so that the offending harmonic is moved out of a sensitive frequency range.
 330 INTERFERENCE CAUSES AND CURES
101,100-
8.000 J measures used on the receivers. A high-pass filter
6,000 installed on the receiver may be relatively ineffec-
5,000 tive if there is a strong rf field around the receiver
4,000
3,000
itself. Complete shielding of the receiver, a difficult
and seldom-taken step, may then be the only
2,000
interference cure.
Finally, it is important to remember that
`.-1,000 interference range is directly related to the trans-
800 mitter power. Nobody has to run high power all
600
500 the time. On the vhf bands a high percentage of all
C 400 communication can be carried out just as well with
c 300 10 watts as with 1000. Dropping power that much,
200 when we don't need all we can run, means the
difference between a 40 -foot and a 400 -foot
100 interference radius. Even 0.1 watt is effective for
80 local chats, and it will cure practically every
60 interference problem.
V 50
40 !i Though what has been said thus far is mainly
30 concerned with 50 -MHz blocking -type inter-
20
ference, the principles apply equally to image and
audio problems, regardless of the amateur trans-
mitting frequency.
to
10 20 30 40 6480 200 400 600100 Using Stubs and Traps
50 00 300 500 1000
OVERLOAD DISTANCE IN FEET If the fundamental interference is mild and the
Fig. 15-3 -
Average overload distance for a TV
receiver on Channel 2 and an amateur station
TV signals are strong, a simple quarter -wave stub of
Twin -Lead cut to the transmitter frequency and
between 50 and 51 MHz. The effective radiated connected to the TV receiver antenna terminals
power is the transmitter output multiplied by the will take care of it. The stub is a good first step in
antenna gain (not in decibels). The supersensitivity any case, as it costs almost nothing, is easy to try,
of misaligned inter -carrier -type receivers to and ordinarily has little or no effect on TV
50.75 -MHz signals is not included. reception. Such a stub is an electrical quarter -wave-
You can cut your interference potential in length at the transmitter frequency, open at the far
several ways without touching a TV installation. end. If it is fitted with open-end lugs at the other
Raising the 50 -MHz antenna to the point where its end it can be slipped under the receiver antenna
main lobe of radiation is well above the TV terminals readily.
receivers and antennas can knock the ERP at the
receiver down by 20 dB or more. This would be Start with a piece a bit more than 0.824 in
the same as using 10 watts instead of 1000 at the length, about 50 inches for 50 MHz or 17-1/2 for
left side of Fig. 15-3. The net effect of a 20 -dB 144. Connect the stub at the antenna terminals,
reduction in signal level at the TV antenna is a and trim it for length while watching the inter-
reduction in interference radius by roughly a factor ference. When the interference level drops trim in
of 10 in distance; a very large difference in a small increments until interference disappears. The
built-up residential area. Blocking interference has stub should have a negligible effect on the TV
a sort of threshold; raising the antenna may put reception where a reasonably strong TV signal is
you on the safe side of it. available, and ít will be effective for any but the
Where the receivers and antennas are close to worst cases of interference.
the transmitter the latter should be well shielded Another type of stub, this one tunable and
and the transmission line nonradiating, if the high requiring no electrical connection to the TV set, is
amateur antenna is to pay off to the greatest shown in Fig. 15-4. It is a double stub, used
extent. This is particularly important in the multi- sandwich -fashion on both sides of the line to the
family dwelling. The indoor dipole (often the timid TV receiver, and tuned for resonance at the
soul's last resort) may be the worst possible transmitting frequency. It is thus a tuned trap,
approach in such circumstances. coupled to the line of the receiver. It.is somewhat
Getting the radiated power well above the TV more effective than the self -resonant stub just
sets also helps the effectiveness of any corrective described, and it may have less effect on the TV
3-30ypf. A

To
Ant. Fig. 15.4 - Sandwich -type
trap for installation in the
To 300 -ohm line to the TV re-
Rcvr. ceiver. Approximate lengths
Short
Porcine' open ends these ends (dimension A) are 40 inches
and connect to one Put sections flat against for 50 MHz and 11 for 144.
terminal of condenser. 300-ohm line from ant. Two traps are in parallel, one
Some on other side. and tape In place.
on each side of the TV line.
Pickup by Audio Circuits 331
reception. It is convenient to make one up on a In this discussion the accent has been on
section of line that can be connected between the 50 -MHz applications but the principles apply to
receiver antenna terminal board and the line to the any vhf problem where the interference is coming
TV antenna. The setup can be pretuned to the in on the antenna or transmission line to the
transmitter frequency, and thus be ready for a receiver. Direct pickup of rf by the receiver, or
quick test. by its ac line, will not be affected by stubs, traps,
The stub is shorter than the self-resonant type, or filters at the antenna terminals. All these
to allow for the capacitive loading. About 38 to 40 devices are more effective if connected right
inches is suitable for 50 MHz, and 10 to 11 inches where the antenna line enters the receiver chassis,
for 144. A worthwhile refinement for a stub that rather than át the terminal board on the back of
will be used for test purposes is substitution of a the cabinet, if there is an unshielded run of
split -stator variable capacitor for the mica trim- Twin -Lead from the terminal board to the tuner
mer shown. This allows adjustment for stub re- input of any appreciable length.
sonance without introducing hand -capacity ef- It is well to have any treatment ready for
fects, and interference can be nulled out much quick application, and to have the actual work
more effectively. Once it is determined that this done by the owner's serviceman, or by the local
type of stub does the job, an inexpensive trimmer TVI Committee representative. Even when the
like that shown can be put on. It will work just neighbor is friendly he may be a little nervous
as well, but is harder to tune accurately. about your working on the receiver. The quicker
' Traps tuned to the transmitter frequency can and more effectively the job is done, the better.
be inserted in the receiver line. Often a single If you have a functioning TVI Committee,
trap in one side of the line will do the trick, or they probably already have a demonstration filter
one can be connected in each leg. These can be for this purpose. If none is available, you are not
resonated with the aid of a grid-dip meter to the required to supply it. The set owner should be
transmitting frequency. For highest selectivity use encouraged to take up the matter with the dealer
the smallest amount of inductance that will tune from whom the set was purchasel, as many
to the transmitting frequency, with the capacitor manufacturers make provision for supplying fil-
available. Typical tuned circuits in vhf equipment ters where needed.
in this Manual can serve as models.

PICKUP BY A UDIO CIRCUITS

Next to TVI of a visual nature, the most transmitter, these modes cause no trouble. Using
common interference problem for vhf operators them has been a way of living in harmony with
using amplitude modulation ís rf pickup and one's neighbors in many densely populated areas,
demodulation by audio amplifiers and other elec- where a-m would have set off immediate neigh-
tronic devices using similar circuits. Much of this borhood strife.
is audio-only interference to reception of TV and Stub and trap remedies discussed above are
radio programs, but it occurs in every kind of rarely useful with audio interference, as the en-
electronic device using high-gain amplifiers. It is ergy gets into the detector or amplifier stages
increasingly troublesome with high-fidelity record directly, rather than by way of the antenna. Long
and tape systems, as these rarely have any inter- unshielded leads to speakers, remote gain con-
ference -prevention measures built-in. trols, phono pickups, tape heads, and the like are
Where the amateur is using conventional am- always suspect. Converting them to shielded wire,
plitude modulation the interference is usually or to better shielding than they may have had
all -too -clear reproduction of the operator's voice, originally, is recommended. Usually circuits con-
often at very high levels, unaffected by the nected to the first amplifier are the trouble spots,
amplifier gain control. With a sideband signal, the but any long lead may pick up rf energy and
interference may sound like ssb as received on an conduct it into the amplifier. Later stages give
a -m detector. Sideband is a form of amplitude less trouble. Though they may have rf pickup
modulation and allows the offending amateur potential, the gain from there on is far lower
only a slight breathing spell. than when the first stage is included.
Because only amplitude changes at an audio Once the source of the troublesome rf pickup
rate are involved, there is an obvious cure: going is traced down, remedies are simple, though often
over to fm or cw. So long as there is no blocking subject to some cut -and -try. Recommendations
of the program material by the rf power from the shown for vacuum tube amplifiers, Fig. 15-5,

(B) (C)

Fig. 15-5 - Treatment of audio stages for pickup of rf energy. Capacitor Cl, resistor R2, or rf choke
RFC1 should be connected right at the amplifier input. All circuits normally connected to the tube or
transistor input should be connected to the left side of these filtering devices.
 332 INTERFERENCE CAUSES AND CURES
should be applicable to transistor stages. Rf Clearing up audio problems is simple in prin-
chokes used should be for the band in question, ciple, but the set-owner may not take kindly to
though a single choke may work for both 6 and the amateur's digging into his equipment. If there
2 meters. Where bypassing is called for, use values is any doubt, the wise approach is to give the
that will not affect the audio quality. For- owner's serviceman the necessary information,
tunately, effective vhf bypass values (.001 /IF and and have him do the job. Servicemen who are not
lower) are in this category. Series-resonant by- amateurs may have little knowledge of the prob-
passing (see Chapter 16) may be the best and lem, so you may have to use diplomacy in two
simplest treatment of all, for one -band problems. directions. Thus it is doubly important that you
Whatever corrective measure is finally used, it know precisely what you are talking about, in
should be applied as close to the offending tube recommending corrective measures.
or transistor Input element as possible. Remove Pickup by the receiver's ac line may be a
all normal connections, insert the rf choke or factor, though not too often in vhf work. Rf
resistor, and reconnect the circuits at the left end filtering of the ac line where it enters the receiver
of the choke or resistor. is the answer here. Heavy -wire chokes (No. 18 or
Decoupling of the heater circuits of the audio so) and good rf bypassing are the treatments.
stages may be necessary, though this is unlikely. Filters on the plug end of the ac cord, where it
Series -resonant bypassing or ferrite -bead chokes plugs into a wall outlet are almost never of any
are fine for heater decoupling, if needed. value.

KEEPING HARM ONICS AT HOME

So far we've been concerned mainly with already outlined do not clear up the interference,
troubles that arise as a result of receiver de- find out why, at once. Just because you have a
ficiencies. With vhf TVI, at least, they are in the 500 -dollar superwhatsis transmitter does not guar-
vast majority. The possibility that the transmitter antee that it is free of the troubles described in
may be at fault should not be overlooked, how- Items 4 and 5. Running it indiscriminately on the
ever, and every possible check should be made on air can only bring down the righteous anger of
this before operating extensively in an area where the neighborhood around you. By then you may
there are TV sets nearby. The importance of have lost the war, but the fighting will drag on
doing this before the TVI complaints begin to and on.
roll in cannot be over-emphasized. If you have If your own TV receiver does not respond to
demonstrated to your own complete satisfaction treatment for fundamental -frequency and audio -
that your transmitter is "clean," you can face rectification ills, Items 1, 2, and 3, you've got
your neighbors with confidence and good humor. harmonic problems. The first step is to find out
These personal attributes are of inestimable value, where the offending harmonic energy is coming
for this TVI business, remember, is a public - from. Put a non -radiating dummy load on the
relations problem. transmitter, and check again. Use a good load
(see Chapter 14), preferably shielded. There are
CHECKING FOR HARMONICS some good ones available ready made and in kit
The first order of business is to be sure that form, if you don't want to make one from
all available. channels can be received clearly on scratch. Do not use a lamp load; it can radiate
your own TV receiver. If there is interference plenty of energy to cause interference.
from your transmitter in any of them, don't wait If there is no interference with the dummy
for the angry phone calls. Find the trouble, and load on, the harmonic radiation is from the
fix it - right now! If the various treatments antenna, and your problems are well on the way
to solution. There are several simple and practical
corrective measures. One of the best is a tuned
antenna coupler, particularly if your antenna is
fed with balanced line of any kind. Details in
Chapter 8. Also very worthwhile is a high-Q
coaxial or strip -line filter. More on these later in
this chapter, and in QST.2
A low-pass filter connected in your antenna
line is good harmonic radiation insurance, but
such filters are rather difficult to make and adjust
properly. You can buy them ready made, and
i... there have been good 50 -MHz designs in every
edition of the ARRL Handbook for many years.
-1/ Fig. 15-5 - Method for bypassing the end of a
shielded power lead. Leads to the 0.001-11F disk
capacitor shouldbe soldered as close as possible to
the capacitor body. Shield over the wire should be
grounded to the chassis at frequent Intervals. This
method is suitable for harmonics only up to about
100 MHz.
Keeping Harmonics at Home 333
There is little point in repeating such information
here, when it is so widely available already'.3
Both the antenna coupler and the high -Q filter
have an important advantage over the low-pass
filter: they protect the receiver more effectively,
preventing overloading from strong signals below
the amateur band in use, as well as above it. II CHASSIS

They might be quite helpful if you have a near

neighbor who runs high power on the lower RFC
frequency ham bands, or on some higher one.
Harmonic Sniffing
Harmonics that get out by way of the antenna
disappear when the dummy load test is made. If
the interference persists you have work to be
done on the transmitter or its power circuits. - Most effective filtering for harmonics
Fig. 15-7
Some kind of harmonic "sniffer" is now required. up through the high TV channels is accomplished
The simple field -strength indicator of Fig. 14-7 by use of the method shown in Fig. 15.6, plus an rf
may be enough. It will cover TV Channels 2 choke and feedthrough capacitor, RFC and C2, for
through 6 as it stands. A smaller coil for L2 will bringing power leads out of the chassis of a vhf
permit it to tune up through Channel 13, if need transmitter.
be. Plug a stiff wire or rod into J1 for a pickup Cl - 0.001-µF ceramic disk (see Fig. 15-61.
antenna. C2 - 500-pF or 0.001-µF feedthrough capacitor.
With the transmitter running into a dummy RFC - 14 inches No. 26 enam., closewound on
high -value 1 -watt resistor or 3/16 -inch form.
load, place the pickup antenna close to the ac
leads, power cable, any unshielded tubes or cir-
cuits, exposed meters, variable capacitor shafts, or ready have it, but adding it is no great chore.
any other part of the transmitter that could be Just be sure that the shielding completely en-
radiating harmonic energy. If you find some you closes every part of the rf portion of the rig.
have a shielding or filtering job ahead; perhaps Then, if the harmonics still come out, you can
both. find the leaks and stop them. Here are the
Another effective harmonic radiation detector common leaky spots:
is the TV set itself. Cut a piece of Twin -Lead Power Cabling Even with complete shielding,
long enough to reach from the TV set to any leads coming out of the rf portion of a trans-
part of the transmitter you want to check. Con- mitter are likely to have harmonic rf on them.
nect one end to the receiver antenna terminals. Getting rid of it is no great problem. Shielded
Short the other end to make a coupling loop. wiring in the transmitter is good insurance. Where
Tape bare wires so that there will be no shorting the lead comes out of the transmitter housing it
of high voltage into the TV set. Now use the should be filtered. The simple device of Fig. 15-6
Twin-Lead as a probe, coupling it to any sus- will take care of harmonics and other spurious
pected part, wire, or circuit. If there is harmonic radiations in all the low TV channels, 2 through
energy present the interference level will increase 6. Ground the shield on the wire at intervals
markedly as the probe is placed near the guilty inside the rig, and at the point where it leaves the
component. enclosure.
We used to build in complete harmonic pro- If exciter or final-stage harmonics, such as
tection into every transmitter. Experience has 4 x 48 or 4 x 50, are radiated by power leads, the
shown that so much TVI is the result of receiver 15-6 method may not work, since disk ceramics are
deficiencies that we no longer do this. The ineffective above about 100 MHz. Bringing out
chances are that any reasonably well-designed vhf leads on feedthrough capacitors is much better. See
transmitter will be practically TVI-free, and that Fig. 15-7.
the receivers will be the culprits - but you Chassis Leaks Harmonics, especially those in
cannot rely on it. If your own TV set shows the upper vhf and uhf TV bands, can leak out of
evidence of harmonic interference, particularly strange places. One exciter for 50 and 144 MHz
with the transmitter on a dummy load, the built by the author showed harmonic interference
chances are good that some of your neighbors in Channels 10 through 13. This got no worse
will see the same evidence, unless it is visible in when the exciter drove a kilowatt amplifier. Some
your own receiver only when in very close prox- harmonic energy was found on the power leads.
imity to the transmitter. Decoupling as shown in Fig. 15-7 helped, but there
was a faint pattern left.
Harmonic Suppression Using a TV set for the visual indicator, It was
Curing TVI is not a black -magic operation, found that the metal rings on the exciter tuning
whether it is the fault of the receivers or your knobs were hot with harmonic energy. The receiver
transmitter. All transmitters generate harmonics. blacked out when the Twin-Lead probe was
Yours is a veritable Pandora's Box full of them; brought near to them. The variable capacitors
you just have to keep the lid down - tight. tuning the exciter stages were the type having small
Shielding is relatively easy. Most transmitters al - rectangular studs for mounting, providing no way
 334 INTERFERENCE CAUSES AND CURES
of grounding the rotors directly to the panel. men in Channel 10 areas have found relief by
Substituting variable capacitors having threaded changing the order of frequency multiplication
bushings on the rotor shafts, permitting direct from 8-24-48-144 to 8-24-72-144. There can still
grounding to the panel or chassis cleared this be energy in Channel 10 (8th harmonic of 24
trouble completely. MHz), but it is almost certain to be far lower than
Long cracks in a chassis, or between the chassis when the 48-144 sequence is used.
and its cover plate, can act like slot radiators for Another example of taking the easy way out is
harmonics. This is why transmitter shielding is the elimination of the 10th harmonic of 8.4 MHz
fastened with so many screws. in Channel 6 in 50 -MHz transmitters by going to
6.3 MHz or 12.6 MHz in the oscillator stage. Again,
Harmonic Generators this is not the best solution, but it may be a
A transmitter with perfectly clean output can practical one in some circumstances. The right way
still have harmonic troubles, for harmonics can be to do the job is to fix the installation, so that the
generated in strong rf fields. Crystal diodes and offending harmonics are not allowed to get
many other rectifiers, intended or accidental, can through to the transmitting antenna, or to theTV
do it. Look out for them, wherever they may be. receiver.
Check for corroded connections in the antenna A sure cure for most of these troubles is a high
system, in your own array, or in the TV antenna. starting frequency in the exciter. With a 72 -MHz
Watch for poor metal -to-metal contacts not di- oscillator in a 2 -meter rig there is no chance of a
rectly connected to either your antenna system or harmonic In Channel 10. (A 48 -MHz oscillator
that on the TV set being interfered with. This would be no help.) In a 50-MHz transmitter many
condition is found fairly often on apartment house of the troubles can be cleared by starting with a
roofs, where the litter from years of erection and 50-MHz oscillator. It should be emphasized that
decay of TV antennas may be strewn, and metal this does nothing for the fundamental -overload
oxides are turned into harmonic generators by the problem in the low TV channels, however, and
impact of appreciable amounts of transmitter rf most 50 -MHz TVI is of this nature. The 50 -MHz
power. Try for shipshape installation of the ama- oscillator also does not prevent radiation of a 4th
teur antenna, and for antenna height that puts the harmonic in Channel 11, 12, or 13. This must be
main lobe of radiation completely above the TV suppressed by techniques discussed a few para-
antennas and rooftops. graphs back.
Antenna -Mounted TV Boosters
Designing Around Harmonic Problems
Antenna-mounted broad-band boosters using
Where radiation of harmonics of oscillator or bipolar transistors, currently popular in TV fringe
exciter frequencies is causing trouble in the high areas, overload very readily, adding greatly to the
vhf TV channels, as in the 4th harmonic of tripler vhf man's interference problems. This is especially
stages working from 48 to 144 MHz, it is often troublesome, as it results from the viewer's having
possible to use a different frequency multiplying purchased an expensive antenna system, usually
sequence and avoid the problem. This is not an when a color TV receiver is installed. Conversion to
ideal cure, since radiation on anything but the field-effect transistors, raising the amateur vhf
wanted frequency should be held to the practical antenna, and installation of a stub on the booster
minimum, but it can be an easy solution to a local input are known cures. This problem, currently
problem. There is nothing sacred about common one of the most difficult encountered by vhf men,
frequency multiplying practices, and many 2 -meter is likely to be with us for some time.
COAXIAL AND ST RIP -LINE FILTERS
If harmonics or other spurious frequencies Quality of the filter elements is important for
appear in the output of an amateur vhf transmitter best results. Use large conductors and the best
they can be kept out of the antenna by a high -Q possible connections, particularly in high -current
tuned circuit inserted in the line between the areas. Copper or brass, preferably silver-plated, is
transmitter and the antenna. Such a "filter" will fine. Aluminum is satisfactory, and even fruit juice
pass only a very narrow band of frequencies, cans can be used, if all metal -to -metal contacts are
offering a substantially impassible barrier to most clean and solid. Insulation should be kept to a
others. The tuned filter can be helpful in receiving minimum, especially at or near the high -impedance
as well, since it will reject energy on frequencies end of the line. A movable -disk capacitor, requiring
other than the desired ones, and thus prevent no supporting frame or insulating material, is good
overloading from out -of-band signals. for line tuning. If conventional variable capacitors
Antenna couplers described in Chapters 8 and are unavoidable, use types with high -quality insula-
14 perform this function, but higher rejection of tion, and preferably no metal frame other than the
unwanted frequencies is possible with the tuned - minimum needed to support the plates. The type
line filters of Fig. 15-8_ Examples are shown for with threaded shaft bearing, permitting direct
each band from 50 through 450 MHz. Con- grounding of the rotor, is preferable.
struction is relatively simple, and the cost is low. The filter is not a magical device. To get high
Standard boxes are used, for ease of duplication. selectivity and rejection of unwanted frequencies it
Coaxial-line filters, also using low-cost com- should not be loaded too heavily. A properly
ponents, may be found in QST for October. adjusted filter has some insertion loss, and its
1964.2 tuning is critical. If the rejection need net be
Coaxial and Strip-Line Filters 335
Fig. 15-8 - High -Q strip -line filters for 50 MHz
(top), 220, 144, and 420 MHz. Those for the two
highest bands have half -wave line circuits. All use
standard chassis.
extremely high the coupling Into and out of the
filter can be adjusted to broaden response and
e
reduce insertion loss. Two filters can be used in
series, for very high rejection of unwanted fre-
quencies. What you want to do with a filter
determines how you adjust and operate it.
A typical use for a coaxial or strip -line filter is
to prevent radiation of unwanted harmonics of the
exciter frequencies in a 50-MHz transmitter. The
filter of Fig. 15-10 is selective enough to pass
50 -MHz energy and attenuate the 7th harmonic of
an 8 -MHz oscillator, that falls in TV Channel 2.
With an insertion loss at 50 MHz of about 1 dB, it
can provide up to 40 dB of attenuation to energy
at 57 MHz in the same line. This should be more
than enough to take care of the worst situations,
provided that the radiation is by way of the
transmitter output coax only. The filter will not
eliminate interfering energy that gets out from
power cables, the ac line, or from the transmitter marked "input" and "output" but the decals could
circuits themselves. It also will do nothing for TVI be exchanged.
that results from deficiencies in the TV receiver, The 50 -MHz filter case is a 6 x 17 x 3 -inch
such as the various problems we have already chassis (Bud AC -433) with a cover plate that
discussed. fastens in place with self-tapping screws. An
Building the Filters aluminum partition down the middle of the assem-
bly is 14 inches long, and the full height of the
When information on the strip-line filters chassis, 3 inches. Construction should be clear
shown here was first published, considerable mail from the photograph.
was received from prospective users, who did not
understand just how they are made. This was true
particularly of the 50 -MHz model, mechanical INPUT
details of which are not immediately apparent
from the photograph, Fig. 15-10. In an attempt to r-
answer these questions, an equivalent circuit is
shown for the two general types.
L2 ILI
Though they look very different mechanically, 14% 1GASE
the 50- and 144 -MHz filters are similar electrically,
as seen in Fig. 15-9A. The 144-MHz model, Fig.
15-11, is readily understandable as a conventional
quarter -wave line, grounded at one end and tuned OUTPUT
at the other. The 50 -MHz filter, Fig. 15-10, is also
a quarter -wave line, but it is folded back on itself,
to conserve space and permit use of a standard
chassis. Its Ll is a strip of aluminum, with a
mounting lip at the lower right end bolted to that

-
end surface of the chassis. The two coaxial fittings
are above and below the strip, the upper one being
out of sight in the picture. What appears to be a
fitting on the bottom surface is actually a re-
CASE
flection. The ends of the two coupling loops and
the end of L1 all ground to the chassis in the lower
right corner of the picture. The partition down
---I
through the middle is grounded to the main chassis
surface, and to the right end. The line element is
isolated from the chassis, except at the ground (B)
point, lower right. The tuning capacitor is in the
upper right corner. Fig. 15-9 -- Equivalent circuits for the strip -line
The 220- and 432 -MHz filters are half-wave filters. At A, the circuit for the 6- and 2 -meter
lines, in which L1 is grounded to the case at both filters is shown. 12 and L3 are the input and
ends, and tuned by a variable capacitor, C, at the output links. At B, the representative circuit for
middle. The output and input loops, L2 and L3, the 220- and 432 -MHz filters. All four filters are
are at opposite ends of the line, for balance. All bilateral, permitting interchanging of the input and
filters shown are bilateral. The coaxial fittings are output terminals.
 336 INTERFERENCE CAUSES AND CURES
Fig. 15-10 -
Interior of
the 50 -MHz strip -line filter.
Inner conductor of alum-
inum strip is bent into U
shape, to fit inside a stan-
dard 17 -inch chassis. Coup-
ling is by L-shaped loops
about 1/4 inch above and
below the tuned line, lower
O right.

Fig. 15-11 - The 144 -MHz

filter has an inner con-
ductor of 1/2 -inch copper
tubing, grounded to the
left end of the case and
supported at the right end 4111.
_
by the tuning capacitor.

The inner conductor óf the line is 32 inches The 144 -MHz model, second from the bottom
long and 13/16 inch wide, of 1/16 -inch brass, in Fig. 15-8, is housed in a 2-1/4 x 2-1/2 x 12 -inch
copper, or aluminum. In the model shown this was Minibox (Bud CU -2114-A). The inner conductor
made from two pieces of aluminum spliced to- (see Fig. 15-11) is 1/2 -inch copper tubing 10 inches
gether to provide the 32-inch length. Splicing long. One end is slotted 1/4 inch deep with a
(visible at the left end of the U-shaped inner hacksaw. This slot takes a brass angle bracket 1-1/2
conductor) seemed to have no ill effect on the inches wide, 1/4 inch high, with a 1/2 -inch
circuit Q. The sides of the "U" are 2-7/8 inches mounting lip. The 1/4 -inch lip is soldered into the
apart with the partition at the center. The line is tubing slot, and the bracket is then bolted to the
supported on ceramic standoffs. As may be seen end of the box, so as to be centered on the end
from Fig. 15-10, these were shimmed up with plate.
sections of hard wood or bakelite rod, to give the The tuning capacitor (Hammarlund HF-15-X) is
required 1 -1/2-inch height. mounted 1-1/4 inches from the other end of the
The tuning capacitor is a double-spaced variable box, in such a position that the inner conductor
(Hammarlund HF-30-X) mounted 1-1/2 inches can be soldered to the two stator bars, as seen in
from the right end of the chassis. Input and output Fig. 15-11.
coupling loops, visible on each side of the line, The two coaxial fittings (SO -239) are 11/16
lower right of Fig. 15-10, are of No. 10 or 12 wire, inch in from each side of the box, 3-1/2 inches
10 inches long. Spacing away from the line is from the left end. The coupling loops are No. 12
adjusted to about 1/4 inch. This may be increased wire, bent so that each is parallel to the center line
for higher rejection, but this will result in increased of the inner conductor, and about 1/8 inch from
insertion loss. The position of the input and output its surface. Their cold ends are soldered to the
coaxial connectors is shown in Fig. 15-8. brass mounting bracket.

Fig. 15-12
strip line
- used
A half -wave
is in the
220 -MHz filter. It is
grounded at both ends and
tuned at the center.
Coaxial and Strip -Line Filters 337
Fig. 15-13 - Construction
of the 420-MHz filter is
similar to the 220 -MHz
one, except that it is shor-
ter, and a disk -type tuning Ac
capacitor is used.

The 220-MHz filter uses the same size box as RELAY BRIDGE FILTER
XMTR.
the 144-MHz model just described, but the circuit
is a half-wave line, grounded to each end of the
box and tuned at the center. The inner conductor
is 1/16 -inch brass or copper, 5/8 inch wide, just RCVR.

long enough to fold over at each end for bolting to

the box. It is positioned so that there will be 1/8 Fig. 15-14 - Preferred method of connecting a
inch clearance between it and the rotor plates of tuned filter in the antenna line of a vhf station
the tuning capacitor. The latter is a Hammarlund makes the selectivity of the filter available for both
HF-15-X, mounted slightly off-center in the box, transmitting and receiving.
so that its stator plates connect to the exact
midpoint of the line. The 5/16 -inch mounting hole
in the case is 5-1/2 inches from one end. Two small When the filter is properly adjusted (with the
holes drilled in the inner conductor allow it to slip SWR bridge) you may find that reception can be
over the stator posts, for soldering in place. improved by retuning the filter. Don't do it, if you
The links for input and output coupling are at want the filter to work best on the job it was
opposite ends of the box, as seen in Fig. 15-12. intended to do: the rejection of unwanted energy,
The SO-239 coaxial fittings are 1 inch in from transmitting or receiving. If you want to improve
opposite sides of the box, 2 inches from the ends. reception with the filter in the circuit, work on the
Their coupling links are No. 14 wire, 1/8 inch from receiver input circuit. To get maximum power out
the inner conductor of the line. of the transmitter and into the line, adjust the
The 420 -MHz filter is similar in design, using a transmitter output coupling, not the filter. If the
1-5/8 x 2 x 10-inch Minibox (Bud CU-2113-A). A effect of the filter on reception bothers you,
half-wave line is used, with disk tuning at the connect it in the line to the transmitter only.
center. The disks are 1/16 -inch brass, 1 -1/4-inch Don't expect the filter, or any other device you
diameter. The fixed one is centered on the inner can connect onto your station, to be a TVI
conductor, the other mounted on a No. 6 brass cure-all. There is no such magic box available, at
lead-screw. This passes through a threaded bushing, any price. Curing TVI calls for some understanding
which can be taken from the end of a discarded of what goes on in transmitters, antennas, and TV
slug -tuned form. An advantage of these is that receivers. There is no easy way out, but by the
some kind of tension device is usually included. If same token, there is no completely hopeless
there is none, a lock nut can be used. situation. Every form of interference can be cured.
Type N coaxial connectors were used on the
420 -MHz model. They are 5/8 inch in from each
side of the box, and 1-3/8 inches in from the ends. BIBLIOGRAPHY
Their coupling links of No. 14 wire 1/16 inch from
the inner conductor are visible in Fig. 15-13. 1Ladd, "50 -Mc. TVI - Causes and Cures, "
June and July, 1954, QST.
Adjustment and Use 2Tilton, "Coaxial -Tank V.H.F. Filters," Octo-
ber, 1964,QST.
If you want the filter to work on both 3ARRL Handbook, Interference Chapter, all
transmitting and receiving, connect up your system modern editions. Also Tilton, "TVI Hints for the
as shown in Fig. 15-14. With this arrangement you V.H.F. Man," April, 1953, QST.
need merely adjust the filter for minimum re- Other references of interest to the vhf worker
flected power reading on the SWR bridge. This include the following QST items:
should be zero, or close to it, if the antenna is well U.H.F. Strip TVI-Proofing the ARC -5 V.H.F.
matched. The bridge should be used, as there is no
Transmitter," Johnson, November, 1950, QST.
Techniques for dealing with various forms of
way to adjust the filter properly without it. If you TVI were included in scores of QST articles of the
try, adjust for best reception of signals on fre- early 1950s. They are of historical as well as
quencies close to the ones you expect to transmit technical interest, since they tell the month -by -
on. This works reasonably well only if the antenna month story of the TVI battle that was eventually
is well matched. won by the amateur.
 Chapter 16
Bits and Pieces
Many vhf enthusiasts are experimenters at last nut and bolt what someone else has already
heart. Their first projects, when they are bitten by worked out.
the vhf bug may be kits or duplicates of QST or This last section of our book is for these
Handbook items, but soon the urge develops to amateurs. It will be something of a hodge-podge of
custom design and build vhf gear. This can take the ideas and techniques that might have been worked
form of studying published constructional articles into other chapters, but which more logically fall
for ideas, and then adapting them to one's own into the "Hints and Kinks" category. The QST
needs. We like to think that much of the material section under that title has been a most -read
in this book will be used in this way. Eventually, feature for generations. We hope that our version
with accumulated knowledge and experience, most of it will find equal acceptance. Our thanks go to
vhf men get to the point of designing for their own the scores of vhf men who supplied the items you
requirements, rather than merely duplicating to the will find here and elsewhere in this manual.

IMPARTING THE "C OMMERCIAL LOOK"

Well-built ham gear of good design usually If you've not used aerosol sprays before it may
works at least as well as equipment purchased be well to practice a bit with some metal scraps.
ready made, but it seldom looks the part. Even the Read the directions; don't assume that you know
simplest equipment can be given a quality look, if how to handle these sprays. The manufacturer
the builder will devote a little time and thought to probably knows more than you do, and he wants
appearance of the final product of his building your results to be good.
efforts. Expensive cabinets are not necessary; even Shake the can thoroughly. Keeping the nozzle
simple chassis-mounted units having no front panel at least 12 inches from the work, spray with a
or cabinet, in the usual sense, can be made neat sweeping motion, using just enough to cover. More
and attractive in appearance with the use of a little will surely cause runs of paint, destroying the
paint, decals, and care in layout. appearance. Allow several minutes for complete
drying, and spray again, evenly and lightly. Now
put the woík aside for at least 24 hours. This
Painting
allows the paint to age, and greatly lessens the
Every hardware store today carries aerosol - chance of damage in handling.
spray enamels, in a wide variety of colors. Black, Two-tone finishes can be made neatly by
grey, and white are favored for ham gear, but other masking off any area that ís to be painted a
colors have their uses. Matching or contrasting different color, or, in the case of aluminum, left its
colors can give many nice effects. If this book were natural finish. Wherever metal is to be natural In
printed in color the reader could better appreciate color, a coating of clear lacquer will keep it looking
the value of judicious use of paints, in other than its best much longer than if it is left uncoated.
the conventional grey and black. This trend to Large areas can be masked off with newspaper,
color in homemade ham gear should be credited to with masking tape only at the edges. Press the tape
Doug DeMaw, W10ER, who pioneered use of color firmly along the paint boundary desired, to prevent
freely in ARRL construction projects. He con-
seepage under the edge of the tape. Keep the tape
tributed the following suggestions. on until the paint has dried thoroughly.
The best paint job is usually possible if the After using spray paint, turn the can upside
metal parts are drilled, ready for assembly, but the down and press the nozzle for a few seconds to
equipment is not actually put together. This is clean out the spray jet. This simple precaution,
practical on all but the most experimental items, often ignored though it is included in the direc-
and even these can usually be rebuilt in finished tions, will make the next job much simpler than if
form, once the bugs are taken care of. Clean the it is omitted.
metal with fine steel wool, to remove rough spots When using two different colors on a surface be
and dirt. With new aluminum this treatment is sure that they are compatible. Test them in
desirable to remove the high gloss, making a better advance; some different paint bases may react on
base for paint. For exceptional durability, spray a one another.
first coat of zinc chromate, an undercoating finish Old transformers and chokes can be made to
also available in spray cans. look like new by painting. Clean them thoroughly
After the rub -down, clean with a grease solvent. of rust, loose paint, grease, etc. before spray
Avoid touching the metal with the bare hands, as painting. Sanding or scrubbing with steel wool may
skin oils and acids can cause blemishes in the be needed.
finished work. Prop the work up with a large area Highlighting can be applied to cabinets and
of newspaper or other protective covering under it. panels by painting with a base color such as grey or
Spray paints dry quickly so dust is no problem, but black, and then spraying over lightly with gold,
select a place that is well ventilated and clean. silver, or copper. For an effect of depth, use clear

338
Imparting the "Commercial Look" 339
spray over the finished product. Take plenty of The solvent should be tested on a paint sample,
drying time between these operations. as some solvent -paint combinations cause wrinkling
and peeling. Paint should be allowed to dry for at
Applying Decals least 48 hours before applying the solvent.
Neat labelling provides the final touch, and it is Other Appearance Factors
of practical value when other amateurs may want
to use your equipment. Typewritten or hand -inked Choice of knobs can make or break the
labels pasted onto equipment make it look like the appearance of homebuilt gear. Occasionally an
work of a rank beginner, regardless of how amateur will devote a lot of time and effort to
a neat outfit, and then spoil the whole
skillfully the electrical and mechanical work has building
are easy to apply. A book of effect with a random collection of knobs. Parts
been done. Decals
many projects arrangement is important, too. Controls don't have
them with enough to last through
costs less than $2.00. They are available in black, to be perfectly balanced in their distribution over a
white, and gold. panel area, but pleasing arrangements nearly always
resorting to string drives,
The label desired is cut from the sheet and then can be made without
remote controls, and various other mechanical
soaked in water to separate it from the paper
backing. Slide the decal onto the metal surface and devices.
move it into the desired position. If you get it Speaker grills offer an opportunity for appear-
slightly awry, dampen it and move it with a small ance highlights. A grill of
perforated aluminum
up properly blot the painted black, mounted against a grey panel, will
brush. When it is lined thought and
The give a pleasing effect. With some
moisture with absorbent cloth or paper towel.
label can be moved again by moistening, until it is advance planning, and
the expenditure of a little
permanently dissolved with Tekni-Solv or lacquer extra time, the final product
of your handiwork
applied with small brush, can be something you'll be proud to show off to
thinner. This should be a
using just enough to moisten the label area. your friends.

SILVER PLATING - WHAT IT DOES, HOW TO DO IT

Silver is one of the best conductors known. Plating can be done in several ways. First, you
Where very high conductivity is important, silver can take your parts to a plating shop. This costs
plating will improve almost any other metal. In money, but assures a good job. There are at least
addition, silver has a special attribute: it remains a three do-it-yourself methods now available, in-
good conductor when oxidized, whereas few other cluding a home version of the process the plating
metals do. For these reasons many items of shops use.
military uhf gear are silver plated throughout, and For this you need a silver anode and a quart of
copious silver plating has come to be almost concentrated plating solution. Both are available
synonymous with quality in the minds of vhf and from distributors of plating materials. They cost
uhf workers. $6.00 each from Hoover & Strong Co., Tupper
But silver is expensive, so it is not so widely Bldg., Buffalo, NY. Other items required before
used in amateur applications. Just how much does you set up in the plating business are a voltage
it do for us, and is it worth the cost? There is no source, 1 to 3 volts dc; a 2 -quart plastic dish, a
single answer, except that silver plating probably 5 -quart rinsing bucket, degreasing solvent, a pair of
never did any harm, and it may be helpful. It clip leads, and some fine steel wool. The plating
makes soldering much easier, and it certainly solution will enable you to plate with other metals
improves metal-to-metal contacts, especially sliding as well as silver. The plastic containers can be
ones. It is well worthwhile in the portions of obtained from any hardware store.
circuits where rf current is high, as in the shorted Preparing the Work
end of a coaxial or parallel-line rf circuit.
Silver plating makes a measurable improvement Copper, brass, and bronze are most suitable for
in the Q of a vhf circuit; 5 to 10 percent increase in silver plating. Steel can be plated, if it is first plated
200 -MHz coils wound of copper ribbon resulted with copper. Whatever the metal it should be
from before -and -after measurements in the ARRL cleaned and polished before immersion in the
Lab. It is probable that copper and brass tank plating bath. Rub it down with fine steel wool, and
circuits of the type used in the 144- and 432 -MHz clean in a degreasing solution. Chemical houses
amplifiers described in this book would be slightly supply degreasers, or you can boil the work in a
better after plating. Tests on typical items have mild solution of laundry detergent. Rinse thor-
shown no measurable improvement in transmitter oughly in clean hot water. Handle only with rubber
efficiencies through plating, but these have not gloves; finger oils and acids will prevent the metal
been made on enough circuits to be sure that no from plating properly.
benefit is obtainable. Certainly the long-term con- Plating
duc tivity of silver-plated items, as compared with
copper or brass counterparts that oxidize quickly Use distilled water to dilute the plating solu-
with handling and use, should have some bearing tion, usually 3 quarts of water to 1 of solution.
on the value of plating to the amateur. This must be at room temperature. Too warm a
 340 BITS AND PIECES
bath will cause discoloration, and too cold will
make for spotty plating. Connect the metal to be Other Methods
plated to the negative side of a 1 -1/2 -volt cell, and Plating kits are available in several forms. An
slide it into one end of the plating tank. Connect inexpensive one is made by Miniplating, Box 161,
the silver anode to the positive terminal, and Middleboro, MA. This consists of a plastic cylinder
submerge it at the opposite end. Maintain a spacing for holding two penlite cells, an electrode that fits
of at least 6 inches between anode and work. Too in the end of the cylinder, clip leads, and a jar of
close spacing causes excessive current flow and plating solution in jelly form. The electrode is
discoloration. Agitate the work frequently to covered with a spongelike plastic. The silver kit is
prevent bubble formation on it. inexpensive; extra jars of plating concentrate are
Immersion time is usually 5 to 10 minutes. available separately.
Longer will give heavier coating, and it is best to To use the kit the electrode is dipped in the
err on this side, as far as the rf quality of the jelly, the clip lead connected to the work, and then
plating is concerned. The higher the voltage the the surface to be plated is rubbed with the coated
rougher the finish. Something between 0.5 and electrode until silver is deposited.
1-1/2 volts is best. Another method, very simple to use, involves a
After plating is completed rinse immediately in plating powder. It is applied with a damp cloth
fresh clean water, preferably lukewarm. Do not dipped in the powder, and then rubbed onto the
touch with the bare hands if you want a clean surface to be plated. Because some rubbing is
surface. To preserve the finish, spray with clear required, the resulting surface comes out very nice
lacquer after the work is thoroughly dry. A lacquer and smooth. The material, called COOL-AMP, is
spray does not affect the ability of the surface to made by a company of that name, 8603 S.W. 17th
take solder. If incomplete plating is found near Ave., Portland, OR. The powder is sold only in
solder areas it is probably due to the presence of jars, minimum order pound, but a little goes a
1

flux. Such areas can be scrubbed with a stiff brush long way. Several would-be platers could do quite a
and xylol or alcohol. Replating can be done as bit of work each with one pound, which covers
needed, in the manner already outlined. about 6000 square inches!
Caution: Silver plating solutions contain cyan- Both the above methods are best used with
ide. Avoid breathing the vapors from the bath. In rubber gloves. The plating materials are a little
mixing, pour the plating solution into the water, rough on the skin otherwise, and neater work is
not vice -versa. Wash hands thoroughly after any possible if the fingers are kept from direct contact
contact with the fluid. Do the plating in a with the work or the plating substances. Several of
well-ventilated room. Store the chemicals in clearly the items described in this book were plated using
marked containers, out of the reach of children. the kit or the powder.

VARIABLE -FREQUENCY CRYSTAL HOLDER

The frequency at which a crystal oscillates is
affected by the capacitance of metal pressure
plates either side of the crystal in the holder, in
mounts such as the FT-243, shown in its original
form at A in Fig. 16-1. In B, a flexible top
electrode is substituted, and provision is made for
varying the pressure this exerts on the crystal, thus
varying the capacitance.
r;rijf T!!
!;,É :r

(A)
4 rNCN SCREW
RUBBER GASKET ú
COPPER PLATE
THREADED ELATE FLEXIBLE PLATE

CRYSTAL
BLANK
Fig. 16-2 - An FT -243 surplus crystal, modified
for variable frequency control.

In this system by W4RMU for swinging the

COPPER LOWER PLATE crystal oscillator frequency, a spring electrode of
(B) PLATE
0.004-inch brass or 0.003 -inch steel shim stock is
Fig. 16-1 - Cutaway views of the FT-243 crystal used as the top plate, in place of the usual top
electrode of the holder. The top cover of the
holder, in its original form, A, and modified, B, for
variable frequency control. to the latter a spring crystal holder is drilled and tapped to take a
with adjustable pressure is substituted for the 1/4 -inch screw (fine thread preferred, but 1/4-20 is
upper electrode.' Spring tension is adjusted by usable) which provides the pressure adjustment.
means of a small screw in the cover plate. The copper plate that made contact with the
Variable -Frequency Crystal Holder 341
original electrode in the holder serves the same A modified FT -243 crystal is shown in Fig.
purpose in the revised one. The spring plate is bent 16-2. The amount of frequency shift will vary from
around a 5/8-inch rod to give the curvature needed. one crystal to another, with the types that are
It is important that the corners of the pressure convex -ground giving the greatest usable swing.
plate be completely free of roughness or burrs. Activity varies somewhat over the frequency swing,
Polish them carefully with an emery stone or very dropping off quickly at each end. A typical surplus
fine file. The range of frequency shift will be from crystal on 6006.667 kHz, which originally gave a
the point where the spring plate touches the crystal 2 -meter frequency of 144.16' MHz now gives
at its center (low end of the range) to the point coverage from that frequency to 144.25 MHz, with
where the pressure no longer holds the crystal substantially no change in final grid drive. This is in
firmly in place. a conventional 6CX8 tetrode oscillator.

CAPACITOR ROTORS - TO GR OUND OR NOT TO GROUND?

The question of grounding the rotor of a tuning or not depends on several factors. If the center-tap
capacitor, whether to do it, and if so, how, bothers of the plate coil is bypassed to ground there is no
many builders of vhf gear. In single -ended circuits, need to ground the rotor, and it may not be
Fig. 16-3A and B, grounding of the rotor, as in A, necessary in other circuits. Especially in the upper
is usually the preferred method. The bypass capa- vhf and into the uhf range, grounding the rotor of
citor, Cl, may be far from perfect, with the result C3 may unbalance the circuit severely, though this
that the rotor will have some rf voltage on it, and it depends to some extent on the capacitor con-
may radiate harmonics, or make the tuning sensi- struction. With small butterfly types, as in a 6360
tive to hand capacity. amplifier stage, it is almost impossible to avoid
Choice of the tuning capacitor, C2, may be rotor grounding. In these circuits, with their
important, too. Some variables have metal studs well-balanced miniature capacitors, there is no
embedded in the ceramic end plate for mounting. reason for doing otherwise, but the coil center -tap
With these a connection, usually to a rotor spider, should not be bypassed if the rotor is grounded. In
must be made for grounding. The resultant lead split-stator capacitors with two sections in line on a
inductance may be enough to leave the rotor above single rotor shaft, some unbalance almost always
ground for rf voltage, at frequencies above 100 results from rotor grounding. In one 432-MHz
MHz or so. The capacitor having a threaded rotor tripler-amplifier formerly in the ARRL Handbook,
bushing, for direct grounding to the panel or running the tuning capacitor rotors above ground
chassis, is much better for grounded -rotor circuits. was a necessity. So great was the unbalance with
In the push-pull circuit, C, or any other where a the rotors grounded that neither tripler nor ampli-
split -stator capacitor is used, grounding the rotor fier stage would operate at all in that condition.

(A) (B) (C)

Fig. 16-3 -Because the bypass, Cl, may not be completely effective, the grounded-rotor circuit, A, is
preferred to that in B, which bypasses both the rotor and the low end of the plate coil. In the
push-pull circuit, C, the rotor is best left ungrounded, unless the design of C3 is such that good balance
to ground is assured.
SERIES -RESONANT BYPASSING
It is well known that the inexpensive disk- TABLE 16-I
ceramic and "dog -bone" types of capacitors are
relatively ineffective for bypassing above about Values of capacitance in pF required for reso-
100 MHz or so. This is due mainly to their nance at frequencies commonly encountered in
considerable lead inductance, even when they are amateur -band vhf work, for leads of 1/4, 1/2, and
connected as close to the elements to be bypassed 1 inch in length.

as possible. Actually this lead inductance can be Frequency 1/4 Inch 1/2 -Inch 1 -Inch
used to advantage, by selecting lead lengths that MHz Leads Leads Leads
make the capacitor series-resonant at the frequency 48-50 800 400 200
to be bypassed. 72 390 180 91
This approach is recommended by WA2KYF, 96 220 100 56
who supplied the information in Table 16-1, 144 100 47 25
220 39 20 10
showing capacitor and lead-length combinations
for effective bypassing of rf energy at frequencies
 342 BITS AND PIECES
commonly encountered in vhf work. The values are because the former will be less likely to allow
not particularly critical, as a series -resonant circuit unwanted coupling to other circuits. For example,
Is broad by nature. The impedance of a series - a 100-pF capacitor with 1/4-inch leads is a better
resonant by-pass is very close to zero ohms at the bet than a 25-pF with 1 -inch leads, for bypassing at
frequency of resonance, and it will be lower than 144 MHz. The series-resonant bypass is worth a try
most conventional capacitors for a considerable in any circuit where instability is troublesome, and
range of frequency. conventional bypassing has been shown to be
A high -capacitance short -lead combination is ineffective. Screen, heater, and cathode circuits are
preferable to a lower value with longer leads, usually good candidates.

HINTS ON PARABOLIC ANTENNA DESIGN AND CONSTRUCTION

Effective work above 1000 MHz almost re- INCNES
quires a parabolic antenna. The usual solution is a 30

trip to a surplus depot or military junkyard, but

25
parabolas of moderate size are not too difficult to
make. A 5 -footer (23 dB gain at 2300 MHz) built
for the first amateur work on 2300 MHz ever, and 20

still a good design, is shown in Fig. 16-4. This

picture is from a description by Koch and Floyd of Is

their pioneering 2300 -MHz work, in QST for July,

1946. How the reflector was made is fairly obvious I0
from the photograph, and Fig. 16-5. The dimen-
S
sions need not be in inches; the proportions can be
used for any size dish. Focus (Antena
Much interesting information on practical para- Yo 9oesne.e)
bolas was given by WA9HUV in "The World Above
50 Mc," June, 1971, QST. His homemade 12 - s

footer for 432 and 1296 MHz is shown in Fig.

10
16-6. The following is from the 'QST report. A
10 -page photocopy of the full WA9HUV con-
Is
struction and design information is available from
ARRL Headquarters for $1.00.
7.0

Illumination and Element Spacing

25
It will be seen that spacing between reflector
elements is appreciable, and that not all the 30
reflector area is "filled" with tubing. Would-be dish 0 5 10 IS INCNES
builders may not realize that though the contour X
must be kept close to the ideal, the density of the

g.
reflector need not be high. Reduction of the 6 or ,, wire
Clamp made
of thin petal
CENTER
DETAIL
END DETAIL

Baakin9 aid
support plate

Fig. 16-5
be
-
The frame for a parabolic reflector can
cut from plywood or pine shelving, using the
dimensions or proportions given above. The shape
of the curve can be computed from the formula Y2
= 4AX, where A is the distance from the center to
the focal point, in this case 17 inches.

density in the four areas around the outer portion

of the dish has a negligible effect on the overall
efficiency of the system.
The limit for spacing of reflector components
has been shown by theoretical analysis and leakage-
loss measurement to be greater than many readers
might expect.1 The 3 -inch spacing used in this
Fig. 16-4 -A homemade parabolic reflector for
2300 MHz. The model shown is 5 feet in diameter,
array is closer than is needed for optimum results
1Microwave Theory and Design, Vol.
using proportions shown in Fig. 16-5. Radiation Laboratory Series, BIT-Edition, 12, MIT
p. 449.
Hints on Parabolic Antenna Design and Construction 343
at 1296 MHz. Theory indicates that the dish will
work at 2300 MHz with a loss of only 3 dB,
compared to a solid reflector. If it were to be used
for 432 only, the spacing could be as much as 12
inches without appreciable loss.
Placement of the feed is determined mainly by
the beamwidth of the feed antenna and the size
and shape of the parabola. The pattern of the feed
should be wide enough to cover the entire para-
bola, yet not so wide that appreciable power spills
over the edges of the dish. Remembering the shape
of the pattern of a dipole and reflector, and the
way the power drops off beyond the 3 -dB points,
it can be seen having the 3 -dB points just inside the
edges is a fairly good approximation. It happens
that the patterns of the dipole-and -reflector at 432
and the cylindrical horn at 1296 are quite similar,
so one position is satisfactory for both. The feed
assemblies are about 60 inches out from the dish
center.
The curvature of the parabola is then deter-
mined by the optical requirement for beam -
forming (collimation) that the distance from the
source to any point on the reflector, and back out Fig. 16-6 -The WA9HUV 12 -foot dish for 432
to an imaginary plane in front of the dish, be the and 1296 MHz atop its 50-foot tower.
same. All energy from the source must arrive at the
imaginary plane simultaneously, producing what is 9.6, 17, and 26.7 inches lower, progressively, for
known as a plane phase front, and the narrowest each nail out from the center point of the curve.
possible beam. The 60-inch focal length of this The curve should be smooth, and the final
design is critical to about plus or minus one inch. result should just touch each nail as the element
rests on the jig. The shape of the dish can be
Bending the Metal controlled to some extent during assembly. Final
positioning can be adjusted to match the curve of a
Reflector components of the WA9HUV dish test gauge at all points.
were bent to the desired shape with the aid of a
very simple "jig." Once the desired curve was Results
determined, large nails were driven into the stud-
ding in a garage wall, and the dish elements bent so By early 1972 the WA9HUV 12 -foot dish had
that they just touched each nail. This is done quite been in service more than two years, on 432 and
readily with the smaller sizes, and it is not too 1296 MHz, with excellent results on both bands.
difficult with the larger sizes, if done a small Beamwidths are 14 degrees on 432 and 6 degrees
amount at a time, and with great care. The garage on 1296 MHz. These result in gains of 21 dBi and
studs in this instance are 16 inches apart center to 29 dBi, respectively, assuming 2 -dB illumination
center. This called for fixture nails at 1.07, 4.25, loss.

MAKING AND US ING RF CHOKES

General-purpose rf chokes can be bought al- ohms. They can be used for voltage -dropping as
most as cheaply as they can be made, but winding well, as in the case of Nuvistor stages that must run
your own has its points. If choke efficiency is at around 70 volts, with a power source that may
important, you can probably make a better one be as high as 200 or 250 volts.
than you can buy, and for random applications a For decoupling the heater circuits chokes must
few turns of wire on a resistor, or self-supported, be used. RFC1 and RFC2 can be wound on small
may suffice. resistors, preferably of 10,000 ohms or more.
RFC3 and RFC4 prevent signal loss to ground,
Applications when low -value cathode resistors are used. This is
A choke is used to keep rf out of a circuit, or in not a critical application; almost any choke will do.
another. For vhf applications something between a The shunt-feed plate choke in a transmitter
quarter and a half wavelength of wire wound up amplifier, RFC6 in Fig. 16-8, is a more critical
into a coil' will do. It should preferably be of small matter. This choke has the whole rf power of the
diameter (usually 1/4 inch or less) and 3 to 6 times amplifier across it, at high impedance, and it had
as long as it is wide. In some circuits a carbon better be a good one. On the other hand, the
resistor will work just as well; maybe better, if output choke, RFC7, never sees much over SO
there is not a heating problem. Fig. 16-7 shows ohms impedance, so long as the transmitter is
resistors used for decoupling the plate -power leads working into a well -matched load. Its quality is not
of familiar rf amplifier circuits. R1 and R2 can be a matter for great concern, and it .is mainly a
anything from a few hundred to several thousand protective device, to prevent dc voltage from
 344 BITS AND PIECES

Rr
150V.

Fig. 16-7 :-
Typical rf amplifier circuits for vhf receivers or
converters. Resistors R1 and R2 are used for decoupling of the
RFC= power leads. Rf chokes could replace them, but resistors do the job
sav well enough. In the heater circuit the current is too high for resistors
so simple rf chokes, RFC1 and RFC2, must be used. In the
grounded-grid amplifier, right, chokes RFC3 and RFC4 are inserted
in the cathode leads, to prevent signal loss to ground. Heater circuit
chokes are the same as for the cascode circuit at the left. The
quality of the chokes is not particularly critical in either application.
appearing on the antenna line, should Cl break shows right and wrong positions for the rf choke
down. The grid choke, RFCS, is not particularly ona 2 -meter plate line. The pipe-line amplifier of
critical. Chapter 6 was built originally as shown at the left
Placement of an rf choke may have con- of Fig. 16-9. When power was applied the rf choke
siderable bearing on íts performance. Fig. 16-9 went up in smoke. Moved to a position outside the
"U" of the plate line its replacement has run coolly
ever since.
TABLE 16-I1 There is no "good" place to put a choke used in
the manner of RFC6 in Fig. 16-8. It has to be close
Rf Chokes for 50-, 144-, and to the tube, so it is subjected to considerable heat,
220 -MHz Service as well as to high rf voltage and heavy do flow.
Frequency Inductance Description Consequently this rf choke must have large built-in
50 MHz 7.8 to B&W Miniductor No. 3004,
safety factors in all categories.
9.5 µH 1-3/8 to 1-9/16 inch long.'
50 MHz 8.3 µH No. 28 dsc, spacewound on Designing for the Job
1/2-inch Teflon rod. Wind- To handle the de load without overheating, No.
ing 1-314 inch long. See text
50 MHz 7.2 pH
28 wire is about as small as it is safe to go in an rf
No. 28 dsc, closewound on
1/4 -inch Teflon rod. Wind- choke for heavy-duty tra nsmitting applications.
ing 1-7/16 inch long. Larger is better, if there is room. Space -winding the
144 MHz 2.15µH No. 22 Nyclad, closewound turns increases the heat -dissipating qualities, mak-
1-3/16 Inch on 1/4 -inch Tef- ing it possible to use smaller wire than when the
lon rod. turns are closewound. Most heat trouble in if
144 MHz 1.42 µH 31 turns No. 28 dsc, space - chokes develops from their being used in hot
wound on 1/4 -inch dia self- places, and being subjected to high rf voltages,
supporting. rather than to excessive do flow alone.
(Above 144 -MHz chokes work well on 220 MHz.)
220 MHz 0.6 µH 13 turns No. 22 Nyclad on Wire size is important in heater chokes, espec-
1/4 -inch Teflon rod. ially where the current to several tubes runs
220 MHz 0.75 pH 17 turns No. 28 dsc space - through a single choke. No. 22 or 24 wire is about
wound on 1/4 -inch Teflon as small as should be used in heater leads,
rod. Winding 5/8 inch long. ordinarily. These or larger sizes can be used for
220 MHz 0.52µH 22 turns No. 22 Nyclad self-supporting chokes for the higher bands.
closewound on No. 24 Distributed capacitance limits the range over
drill, self-supporting. which an rf choke will work. This makes the
Excellent for use except where high tempera-
tures are involved.

Fig. 16-8 - Transmitter applications for

rf chokes vary markedly in regard to the
quality of choke needed. In the grid
circuit, RFC5 has no very difficult job to
do, and any choke suitable for low -
power use is suitable. The shunt -feed
choke, RFC6, must meet severe require-
ments, especially in high-powered ampli-
fiers. It is effectively across the trans-
mitter tank circuit, and is subjected to
high temperature, current, and voltage.
The output choke, RFC7, is mainly a
BIAS
safety device and it operates under much
less stringent circumstances.
Making and Using Rf Chokes 345
Fig. 16-9 - How a choke is positioned RFC
rigkt
with respect to other circuits may be
important. The choke at the left is
coupled to the plate line of the trans-
mitter tuned circuit. Outside the loop, as
at the right, makes the choke far less
subject to rf breakdown.

space -wound choke superior to the close -wound range first. The pieces unwound will be useful for
one. A minimum of cement on the windings is also higher frequency production later.
desirable. The space -wound 50 -MHz choke in Table Self-supporting chokes of excellent quality can
16 -II, shown in the upper right of Fig. 16-10 is as be made by winding No. 22 or 24 wire tightly on
good as you can make for that band, and better various drill sizes, and then slipping the drill or
than most chokes you could buy. It is good at 144 other winding form out. If wound under tension
MHz as well, and even serviceable at 220 MHz. A the coil will hold its shape when slipped off the
closewound choke of fine wire, heavily doped with form. Turns can be spaced by running a thin knife
lacquer, might be usable on only one vhf band, and blade between them. You can't make a better
very likely it would not be too good, even there. choke than this.

Making Your Own

To set up in the rf choke business we need
some wire: No. 22 enamel (Nyclad or Formvar
preferred); No. 28 enamel, silk or cotton covered;
and No. 30 or 32, of any similar insulation. Silk- or
cotton -covered wires take cement nicely, but ena-
mel is OK otherwise, and it is usually most readily
available.
High -value 1/2- or 1 -watt resistors make good
winding forms for use at 144 MHz and higher. A
2 -watt resistor is big enough for a 50 -MHz choke,
but Teflon or Nylon rod stock is better. Do not use
polystyrene or lucite, if any heat is to be involved.
These materals will melt in the heat of an average
transmitter enclosure. Teflon rod can be found in
plastics supply houses, in 1/4- and 1/2 -inch dia-
meter. It drills and taps nicely, it won't melt, and
Fig. 16-10 - Typical handmade vhf chokes. At the
rear are closewound and spacewound chokes for 50
its insulating quality is excellent. Bakelite rod or MHz wound on 1/4 -inch and 1/2 -inch Teflon rod,
even wood dowelling is good enough for the drilled and tapped for end -mounting. Three
less -critical choke applications. The smallest - 144 -MHz chokes are seen in the center row, the
diameter prepared coil stock is usable for 50 -MHz two at the left being excellent for high -current
chokes, but it won't stand much heat. applications. Similar types to these, but for the
Space -winding rf chokes is easy. First drill 220 -MHz band, are in front.
through the rod at spacings indicated under wind-
ing length in Table 16 -II. Now measure off slightly You can tell a good choke from an inferior one
more than a half wavelength of wire. Double it easily enough. Connect it across your driver-stage
back on itself and feed the end through one of the tuned circuit, and see what it does to your
holes in the rod. Now wind the coil as if it were to final -stage grid current. Also note how much you
be bifilar. If you clamp the other end of the double have to retune the driver circuit to restore reso-
wire in a vise, or tie it down firmly otherwise, this nance. A perfect choke would have no harmful
can be done easily. Keep the wires under tension, effects, and it would not heat up. You won't find
and be sure that they are not twisted at any point. one that good, but a well -designed choke will come
Wind tightly and then feed the end through the close. If the choke is not a good one, don't run the
other rod hole. test too long at any appreciable power level, or you
Now remove one of the wires by unwinding -
won't have to look for indications you'll smell
carefully, keeping it under tension throughout. The them!
remaining wire will be space-wound as neatly as if A recent development m the rf choke field is
done by machine. Apply a thin coating of poly- the ferrite bead. These are small beads of ferrous
styrene cement, using a bit more around the lead materials that can be slipped over wires wherever
holes, and your choke is done. It will be dry and the effect of an rf choke is needed. They are
ready for use in a few minutes. If having all those particularly effective for heater decoupling pur-
wire scraps left over runs against your Scotch poses, as they can be placed directly on the heater
instincts, make chokes for the lower end of the lead, adding no dc resistance to the circuit.
 346 BITS AND PIECES
"JUST LI KE QST, EXCEPT . . ."*
These words are voiced or written almost daily
in telling the owner's sad story of instability in a
transmitter or receiver he has built. Investigation
nearly always shows that the builder made his own
troubles, as a result of common misconceptions
about bypassing and grounding. Every new project
is to some extent a design problem, and a certain
amount of debugging is almost inevitable, but there
are right and wrong ways to do the basic jobs
encountered in amateur equipment construction.
Take the matter of sockets and bypassing. Fig.
16-11 shows wrong and right methods. The wrong,
upper, has Pins 4 and 9 supposedly grounded by a
small wire running through the socket center ring
(which is supposed to be a shield) to a lug under
the mounting screw at the left. Pin 3 is bypassed to
"ground" by running the low side of a disk ceramic
capacitor to the center ring. What happens here?
The only place that is really at ground potential is
the portion of the lug that is under the mounting
nut. The rest of the lug, the wire, the socket pins, Fig. 16-11 - Ríght and wrong treatments of 9 -pin
and the bypass lead could all be above ground for sockets. Instability is likely in the long wire -lead
rf. The path to ground from Pin 3 is a long one, version, upper, providing common coupling
and the other pins are common to it. This is an through use of the center ring for the bypass
capacitor perched above Pin 3. In the lower, the
invitation to feedback, due to common-lead coup- lug provides low-inductance path to ground for the
ling. ring and Pins 4 and 9. The bypass, Pin 8, is
Now look at the lower. The same pins are horizontal, 'and connected to the lug as close to the
grounded by bending the socket lugs against the chassis as possible.
center ring and soldering them there. The lug under
the mounting nut is also bent over to the center
ring. This puts Pins 4 and 9 very much closer to
ground potential than the wire method at the left.
maker's invitation to use those ring extensions for
The capacitor from Pin 8 is in a horizontal
position, and is grounded to the lug, close to the the cold (you hope!) end of your bypass capa-
mounting nut. If the disk ceramic is capable of citors, you can very easily build in common
being an effective bypass, this wiring arrangement coupling through the ring's inductance. The socket
will give it a chance. There is almost no common at the right, with no such conveniences, forces the
path to ground between Pin 8 and the other user to bypass to ground, at the mounting nut,
which is the right place to do it.
circuits, and the center ring is much nearer to being
a shield than before.
The best amplifiers, transmitting and receiving,
have very high gain and power sensitivity. Only a
Another demon for the vhf man is the socket little bit of feedback can make them take off. If
shown at the left of Fig. 16-12. The manufacturer
very kindly provided an elevated mounting ring on
this feedback is within the device itself, or if it is
due to coupling between tuned circuits, you can
this gem. Fine for use on lower frequencies, or in a
neutralize or shield it out. If the coupling is built in
hi-fi amplifier, perhaps, but it caused all manner of
through common ground leads and ineffective
trouble in a 6146 amplifier for 50 MHz built by
bypassing, as illustrated in these typical examples,
the author of this book.
it can grow you a lot of grey hairs.
This socket makes reliable contact to the * From a longer article, under the same title,in
chassis only at the mounting ears. If you accept the March, 1959, QST

Fig.16-12- Socket at the left with its built-in "grounding" ring is an invitation to trouble in vhf
circuits. The one at the right necessitates use of grounding lugs at the mounting holes and encourages
good bypassing and grounding practice.

T
 More Versatility with the Heath Sixer and Twoer 347
MORE VERSATILITY WITH THE HEATH SIXER AND TWOER
The well-known "Benton Harbor Lunchbox" is are connected together and a lead run from this
a mainstay of a -m activity on 6 or 2 meters in connection to the chassis of the transceiver. This
many localities. Here are several modifications of provides a common ground for the oscillator. The
these popular little rigs that will add to their key jack, J101, must be insulated from the panel,
versatility. They are the work of Lew McCoy, and either insulating washers or electricians' plastic
W lICP. tape can be used for this purpose. The jack is
mounted on the panel between the microphone
Adding A2 for Code Communication connector and the volume control. The oscillator
assembly is supported by its own leads. When
The 50- and 144 -MHz bands are ideal for code installing the board, be careful that none of the
communication for -
practice or for improved
connections on the bottom touch.any leads in the
signal readability when the going is rough. Unfor- transceiver.
tunately very few readymade a -m transceivers now In order to monitor your own sending, a
available provide for code work of any kind, so 330 -ohm, 1/2 -watt resistor should be connected
much interesting potential of the vhf bands is lost
between terminal 4 of the transmit-receive switch
to owners of such equipment. The superregener- and the chassis, as shown in Fig. 16-13C. This feeds
ative receivers in the Sixer and Twoer make it a very small amount of audio from the transmitter
impractical to copy keyed cw, but they are fine for to the speaker. When transmitting A2 turn the
reception of tone modulation. volume control full on; otherwise, the audio
The simple transistor tone oscillator of Fig. oscillator output will be short-circuited to ground.
16-13 can be built into either unit easily. By
For receiving or using the oscillator for code
connecting the output of the oscillator to the arm practice, the volume control should be set at a
of the volume control the keyed tone will modu- comfortable listening level. A switching circuit
late the transmitter, and when the transceiver is could be used so that the volume control setting
switched to the receive position the oscillator can wouldn't have to be changed, but this would have
be used for code practice. Also, in the transmit complicated the conversion and didn't seem worth
position a slight amount of the audio tone is fed to the expense or crowding of components.
the speaker, permitting the operator to monitor his
own "fist." Metering Transmitter Output
The tone oscillator is mounted on a 2 x 2 -1/4 -
inch piece of perforated unclad circuit board. This One problem with the Twoer and Sixer is that
is 1/16 inch thick and is perforated with 1/16-inch external meters are required for tune-up, and there
diameter holes spaced approximately 1/4 inch is no constant metering of the output. A low-cost
apart. Push-in clips are available for making con- milliammeter connected as a relative output indica-
nections, but in the units shown the connections tor can be installed in each unit, for constant
were made by soldering the component leads monitoring of the power going to the antenna.
together. The meters in Fig. 16-15 are edgewise miniature
The emitter of Ql, and one side each of C101, S meters. There is adequate space on the panel for
R101, and R102 (and C102 in the 2 -meter unit) both the meter and control R103 just below the
TO PIN 6 Of V2
TO PIN 6 OF V2
"101 3101
KEY 5e21C KEY
lo e2K
01 OI
e
3300
7y. RI02
C102 E R10254700 REY 4700
DOWN
.001
RFC101 TO TERMINAL 2 TO TERMINAL 2
OF OF
VOL'R6E CONTROL "X vOLOYE CONTROL 'le
IRK I8K
(C)
.02 0101 ,05 -`02 02 .02

T
R101 (A) ( B)
11300 01
144 MC. 50 MC.

Fig. 16-13 -
Circuit diagram of the code oscillator. A is the 144 -MHz unit and B is for 50 MHz. Fixed
resistors are 1/2 watt; resistances are in ohms (K = 1000) and capacitances are in µF. Capacitors are
paper or Mylar, working voltage 25 or more. Component numbers under 100 refer to the original
Heath circuit; those over 100 are the added components. C is the circuit for monitoring one's own
sending.
C101
--0.05µF dsk ceramic, paper, or Mylar. Q1 --
Npn, RCA type 40314 or similar.
1800 ohms, 1/2 watt.
C102 0.001 µ disk ceramic F R101
.1101 - Single -circuit phone jack or phono jack. RFC101 -2.71.11-1 (Millen 34300-2.7 or similar.)
 348 BITS AND PIECES
ispresent on the tank coil and you could get a
dangerous shock.
The 6 -meter installation is slightly different.
-
The trimmer capacitor' is mounted on a 3 -lug
ff terminal strip with the center terminal grounded.
The strip is mounted between the crystal socket
4"14 i and the socket for V4, using the unused coil
mounting hole as the mounting point. A 2 -turn
link, with the turns just slightly smaller in diameter
than the tank coil, is made from insulated No. 16
ANT. CONNECTOR

IT
1
1

1
-i. D1 3300
TO METER
JACK T
t

Fig. 16-14 -
Interior of the
code oscillator installed.
"Lunchbox" with the R103
lox
1.4101
o-1
SENSITIVITY
nameplate. The ungrounded end of R103 is con-
nected to the meter-jack side of R13 (a 3300 -ohm,
1/2-watt resistor) by an insulated wire fed under
the chassis through a grommet below the meter. Fig. 16-16 -
Addition of the metering circuit.
M101 -
0-1 milliammeter (Radio Shack 22-004,
Amplifier Tank Circuit Modification World Radio Labs 99M194).
Another worthwhile improvement can be made
R103 -10,000 ohms, 1/4 -watt control.
by changing the output tank circuit from capa- or 18 solid wire. This is positioned just inside the
citive to inductive coupling to the antenna. This tank coil at the feedthrough capacitor end. The
reduces the possibility that undesired frequencies adjustment procedure is the same as with the
generated in the multiplier stages will reach the 2 -meter unit.
antenna. Remove the wire from terminal 12 of the
Remove the coupling capacitor that goes from transmit-receive switch and the antenna output
the tank coil (L3 in the Sixer and L4 in the Twoer) terminal. A length of RG-58/U is substituted for
to terminal 11 of the transmit-receive switch. In this lead, grounding the outer conductor at both
the 2 -meter unit, insert one side of a 3-30 ends.
compression trimmer capacitor under the nut that External Crystal Socket
holds the tube socket for V4, at the chassis -edge
A crystal socket on the front panel makes
side. The new coupling loop, L101 in Fig. 16-17, is
frequency changing much easier. This mounts on
made of insulated No. 14 or 16 solid wire. The the front panel alongside the meter, and a short
loop for the Twoer is one turn the same diameter
length of Twin -Lead, fitted with a crystal socket
as the tank coil, inserted between the first and
plug (Millen type 37412), is used to connect it to
second turns, at the feedthrough capacitor end. the chassis -mounted crystal socket. If you have a
One end of the loop is connected to terminal 11 of defunct crystal, it can be removed from its holder
the transmit -receive switch and the other end to and the Twin-Lead soldered to the holder pins, to
the ungrounded side of the 3-30 compression make a plug.
trimmer. Keep these lead lengths as short as
possible. ANT.
Using the lamp dummy load that comes with LIOI _ RGSA/U CONNEEcc)TOR'E'

the kit, tune the tank capacitor and the com-

pression trimmer for maximum lamp brilliance.
The output meter will read maximum when the 103
lamp is the brightest. It may be necessary to reduce
the sensitivity by means of R103 to keep the meter
from going off scale. Fig. 16-17
and Sixer.
- Tuned output circuit for the Twoer
Try moving the loop in relation to the tank
coil, for maximum brilliance of the lamp load. Be C103 - For 144 MHz, 3-30-pF compression
sure to turn off the power to the transceiver when
trimmer; for 50 MHz, 8-50-pF trimmer (Centra -
lab type 822 -AN or similar).
making this adjustment because the B -plus voltage L101 -
See text.

. ...
Fig. 16-15 -
Twoer and Sixer,
The
complete with code ^'::.
-}.:r_
SZ"_:
-

' RL.ar
oscillator, relative b
L.._.,
power meter, and
front -panel crystal
socket.
-^h _ _
9
L -Match for Coax -Fed Vhf Arrays 349
L -MATCH FOR COAX -FED VHF ARRAYS
Whenever coaxial line is used to feed a driven
element directly, there must be provision for
converting the unbalanced line (coax) to the
balanced load represented by the dipole. Some
method of raising the driven element impedance is
also required. The simple L-Match of Fig. 16-18
was devised by Ralph Campbell, W4KAE, to do
both jobs. Its derivation is shown in Fig. 16-19.
If the question of balance is ignored, an
inductive stub of U shape, solid line in the sketch,
can be used in conjunction with a shortened driven
element, to effect an impedance match between
the transmission line and a driven element of lower
impedance. Such a stub tried by W4KAE in a
2 -meter Yagi having a two-piece dipole fed with
coax gave something approaching an impedance (B)
match, but left the problem of balance unsolved.
Checks with an rf probe showed that the portion
connected to the inner conductor was hot with rf, 1J
-1 Fig. 16-19 - Evolution of the L-Match. Perpen-
dicular loop, solid line, raises feed impedance, but
leaves problem of unbalance unsolved. Moving loop
de: a to the right partially corrects balance. Boot -shaped
..1
loop, B, combines impedance-matching and balun
effects.

element Yagi, operated at 145 MHz. The mounting

screws of the dipole, which serve as connection
:TS
points for the coax and balun loop, are 2-1/4
inches apart. The loop is made of aluminum
ground wire, about 9 inches overall, including the
"eyes" at each end for slipping over the mounting
Fig. 16-18 - The L -Match as installed by W4KAE. screws.
Some variation in loop inductance, and thus in
but the other side was practically cold. In effect, impedance matching, can be made by varying the
the driven element was acting like the fed portion spacing between the upper and lower portions.
of a ground -plane antenna, with the other half and Balancing effect is related to the position of the
the metal boom acting like the radials. loop "toe" with respect to the driven element.
When the loop was bent toward the side of the Obviously these effects interlock, so a cut -and -try
dipole that was connected to the inner conductor approach is indicated. The array should be set up
of the coax, (broken line in Fig. 16-19) the balance at least one wavelength above flat ground, with no
of power in the halves of the dipole improved, and reflecting objects in the field for many wavelengths
the SWR indication on the line went lower than out in front. An alternative recommended in our
could be obtained with a perpendicular stub. This antenna chapters is to point the array straight up
led to experiments with a boot-shaped loop, Fig. while making matching adjustments.
16-19B, varying the position of the "toe" with The L -Match should be of workable dimensions
respect to the hot half of the driven element. Soon for 50- or 220 -MHz arrays. Suggested lengths of
nearly perfect balance was achieved, and the SWR wire for making the loops are 25 and 6 inches,
indication was brought down to 1.05:1. Pre- respectively, for 50 -ohm feed in Yagis of otherwise
sumably very careful adjustment of the length of conventional design.
the driven element, and the length and position of
the stub, could bring the SWR down even lower, 4 3/B
though checks on many supposedly well -matched
2 -meter arrays might show his match to be better
than most. -r 2s/is
The effect of the stub amounts to inductive P

loading at the center of the dipole, so the

end-to -end length must be physically less than that 35 3/4
of a driven element fed by other means. The dipole
length and the size and position of the balun loop
vary with frequency, and with the feed impedance a
Fig. 16-20 -
Dimensions of the L-Match used with
broken dipole in a 15 -element 145 -MHz Yagi.
of the array in question, so no one size can be right Dimensions vary with frequency and driven -
for all 2 -meter Yagis. The dimensions shown in Fig. element impedance, but those shown should be
16-20 were optimum for a "store-bought" 15- average values.
 INDEX
Charts and Tables Collinear, 12 and 16 Elements 188
Air-Dielectric Balun Dimensions 171 Corner Reflector 214
Amateur Bands Above 50 MHz 14,15 Delta Match, 144 MHz 190
Antenna Dimensions 176 Folded Dipole, 50 MHz 177
Antenna Element K Factor 154 Gamma Match, 50 MHz 183,185
Antenna Gaín, Max with 2 Elements . 153 Ground Plane 179
Antenna Isolation, Repeaters 341 Halo 197,203}204
Antenna Height -Gain Nomogram 26 "J" Antenna 178
Antenna Isolation, Repeaters 266 L -Match, 144 MHz 349
Capacitor Series Resonance 341 Log Periodic, 144-450 MHz 216
Cavity Radius, 500 to 3500 MHz 278 Long -Wire, V and Rhombic 159
Characteristic Impedance vs Conductor Mobile Antennas and Mounts . 200-206
Size and Spacing 168,169,276 Omnidirectional Antennas 195-199
Chokes, Rf (Winding Data) 344 Parabolic Antenna Construction . . 342
Coaxial Line, Length of Quarter-Wave 276 Plane and Parabolic Reflectors 159
Collinear Js Yagi Frequency Response 156 Portable Beams, 50 and 144 MHz . . 206
Corner Reflector Dimensions 215 Quads, 144 MHz 194
Corner Reflector Feed Impedance 159 Skeleton Slot, 144 MHz 195
Crystal Frequencies, Vhf Converters 44 Antenna Considerations, Repeater 267
Folded Dipole Nomogram 168 Antennas for 50 MHz 177-189
Fm Filters 226 Antennas for 144 MHz 189-199
Meteor Shower Data 23 Antennas for 220 and 420 MHz 208-217
Noise Figure, Parametric Amplifier 286 Antenna Tubing, Stiffening 195
Operating Conditions, 829B 105 Auroral Propagation 15,16,19
Path Loss vs Distance 27
Pilot Lamp Data 307 B
Radiation Resistance, Parasitic Array 154
Radiation Resistance vs Height 166 Backscatter 15,20
Receiver Sensitivity Nomogram 25 Baluns, Making and Using 170
Rhombic Antenna Data Bits and Pieces 338
159
Station Capability Nomogram 24 Bridge, Impedance 315
Standing -Wave Loss 174 Bridge, SWR 174,313
Sunspot Number History 20 Building and Using Antennas 176
TV and Fm Frequencies, Broadcasting 329
TV Overload Range, Ch. 2, 50 MHz 330 C
TV Sound and Video Channels 329 Capacitor Rotors, Ground or Not? 341
Transmission Line Characteristics 163 Capacitor Series Resonance 341
Yagi Design (2) 155 Cascode Amplifier 37
Cavity Resonator 278
A Circular Polarization. 158
Absorption Wavemeters 308 Chokes, Rf 343
Adjusting Rf Stages 36 Coaxial and Strip -Line Filters 266,334
All -Metal Construction, Antennas 189 Coaxial Fittings 164
Amateur Bands Above 50 MHz 14 Coaxial -Line Characteristics 163
Amplifier Design and Operation 73-79 Coax (Tips on Selection) 164
Amplifiers, Parametric 285 Coaxial -Sleeve Balun 171
Amplifiers, Receiving 35-38 Coaxial Tank Circuits 275
Antennas and Feed Systems, Principles 151 Collinear Arrays 156,188,195
Broadband 251 Communications Receiver Problems 41
Collinears, Directional 156 Converter Design . 34
Collinears, Omnidirectional 156 Corner Reflector Antenna 215
Couplers 169 Corrective Stub 169
Element Dimensions and Spacings 153 Coupler, Antenna 169
Frequency Response 151 Crystal Holder, Variable 340
Gain 151 Crystal Oscillators 70
Height Gain vs Line Loss 152 Crystal Oscillator, VXO 81
Helical 158
Impedance Matching 166 D
Long -Wire (Incl. V and Rhombic) 159 Danger, High Rf Field'
Omnidirectional 175
152 Decals 339
Quad 158 Decoupling Sleeve
Pattern Shape 170
152 Delta Match 167
Phasing 161 Deviation and Linearity, Fm
Plane and Parabolic Reflectors 223
159 Diode Mixers 38,295
Polarization 157 Diode Multiplier 41
Stacking 160 Diode, Varactor 283
Yagi Antenna Design 153 Dipole, Folded 168
Antennas, Construction 176-217 Discriminator, Fm 227-230
Big Wheel, 144 MHz 198 Double Conversion 33
Coaxial 178 Dummy Loads 321
Collinear, Omnidirectional 195 Duplex (AQ) 14 28

350
 E K
Element Diameter, Effect on Length . . 154 Klystron Oscillators and Amplifiers . . 280-282
Element Spacing 155 Knife -Edge Refraction 26-27
Elements, Parasitic 153
L
F
Lamps as Indicators 306
Feeding Phased Arrays 173 Lecher Wires 323
Field -Effect Transistors 34-41 Linear Amplifiers 73,113
FET Preamplifiers, 50, 144, 220 MHz . . 53 Long -Wire Antennas 158
Field -Strength Meters 310-312 Long Yagis 182,192
Frequency Modulation 28,43,79,84,87,97,218-262 Lunar Communications (EME) 16 22
Accessories (Sweep Gen, Freq, Std . . 260
Alignment, Receivers 235 M
Bandwidth 226 Magnetron
Checking Frequency 223 282
Maximum Usable Frequency (muf) 20
Converting Surplus 253 Meteor Scatter, Shower Table
Conversion to 220 (146 -MHz Trans.) 253 22,23
Mixers 38,293
Communications and Jargon 233 Modulators, Fm
Deviation and Linearity 223 98,221,250,291
Multiband Amplifier Circuit 77
Equipment 2 35-262
Filters (Table) 226 N
Heterodyne Fm Exciter 88
Limiters and Detectors 227 Narrow -Band Fm (VXO) 84
Methods 221-225 Neon Lamp (rf Indicator) 307
Preamplifiers, 440 MHz 255,294 Noise Blanker 63
Reactance Modulator 221 Noise Figure and Sensitivity 30
Receiving Adapters 235 Noise Génerator 36,319
Receiving Fm 225 Novice Operation, 145 MHz 14
Receivers, 53, 146, and 440 MHz . 238-245
Repeaters 263-274 o
Speech Processing 222
Squelch 232 Omnidirectional Antennas 156,195,198
Transmitting Adapter Open -Wire Line 162
247 Operating Modes
Transmitters 248-252 27
Transmitting Amplifier 253 Oscillator -Multiplier Exciters 68
Frequency Multiplication 68 72,287
Oscillators, Fundamental and Overtone 69,70
Frequency -Shift Keying 28
Oscillators, Tunable 40
F2 Layer 15,20 Oscilloscope Uses 114

G P
Gamma Match 167,183,185 Painting Equipment 303
Grounded -Gate (Grid) Amplifier 36,46 Parabolic Reflectors 159,342
Grounded -Grid Amplifier (Trans.) . . . 74,115 Parametric Amplifiers 285
Path Loss 24,27
Pilot Lamp Data 307
H Pi -Network or Inductive Coupling 75
Halo 194,203,204 Polarization 157
Harmonics and TVI 327,332-335 Polaplexers, Microwave 281
Heathkit her and 2er Modifications . . . 347 Portable Antennas 206
Height Gain 26 Power Amplifiers 73-79
Helical Antenna 158 Preamplifiers, Receiving 53,255,294
Heterodyne Exciter, 50 to 144 MHz 125 Propagation by Bands and Modes 15,17
Heterodyne Fm-Cw Exciter 88 Pulse Communication 304'
Heterodyning 69 Pump (Parametric Amplifier) 239
Historical References 13
How It All Started 7 Q
Q Section 169
I Quad Antennas 194
Idler (Parametric Amplifier) 285
Impedance Bridge 315 R
Impedance Matching 166
Interference, Causes and Cures 327 Radiation Resistance 154
Intermediate Frequencies 45 Reactance Modulator, Fm 221
Inversion (Tropospheric Propagation) . . 18 Receiver Principles 30-43
Ionospheric Predictions (F2 Layer) . . . 20 Noise Figure 30
Ionospheric Scatter 21 Mechanical Qualities 30
Receiving Fm 43,225
J Rejecting Unwanted Freqs 31,42
Superregenerative Detector 31
"J" Antenna 188 Transistors or Tubes? 33
"Just Like QST, Except . . !" 346 Receiver Front-End Design 34
JFET Preamplifiers 53 Low-Noise Transistors 34

351
 Preamplifiers 37 Transceiver, Transistor, 144 MHz
Protection 146
38 Transceivers, Microwave 281
Mixers 38 Transmitting Conve'er, 50 to 144 MHz
Injection Stages 125
40 Transequatorial Propagation 21
Receivers, Fm . . . See Frequency Modulation Transistors or Tubes?
Receivers, Converters, and Preamps, Vhf . 44-66 73
Transmission Lines 162-165
FET Converters, 50 MHz 45,55 Transmission -Line Table
FET Converters, 144 MHz 163
47,55 Transmitter Design 67-79
Low-Noise Converter, 432 MHz . 58 Transmitter Design, Fm
Noise Blanker 218-225
63 Transmitters and Exciters, Vhf 80-122
Receiving, Uhf 293-297 Fm-Cw Transmitter, 220 MHz
440 -MHz Preamplifiers 97
255,294 Grounded -Grid Amp, 50 MHz 115
1296 -MHz Preamplifier 295 Heterodyne Fm-Cw Exciter
Records, Two -Way Work 88
16 High -Efficiency Kilowatt, 432 MHz . 299
Reliable Coverage 24 Kilowatt, Amplifiers, 50 and 144 MHz
Repeaters, Theory and Practice 106
263-274 Kilowatt Amp, 144, Improved 118
Antenna Considerations 267 Medium -Power 144 -MHz Transmitter
Audio Mixer 89
269 Microwave Stations 281,282
Auto -Patch and Touch -Tone 272 Power Tripler or Amplifier, 432 . 297
Choice of Equipment 265 Pulse Station, 2300 MHz
Control 304
268 Varactor Frequency Multipliers 287
Duplexers 265,268 VXO, 50-450 MHz
Logging and Identification 80
269 VFO, Transistor 85
Off-Frequency Controller 271 VFO with Fm
Remote Base 87
263 829B Amplifier, 144 MHz 102
Tone Control 270 Transverter vs Separate Units
Touch -Tone Decoder 67
273 Transverter, 28 to 50 MHz 129
Window Concept 270 Transverter, 14 or 28 to 144 MHz
ROCLOC - Stable Microwave 133
282 Traps and Stubs, TVI 330
Travelling-Wave Tube 282
S Tropospheric Propagation 15,19
Selectivity and Sensitivity 30,41 TVI, Prevention and Cure 327-337
Signal -to -Noise Ratio 30
Silver Plating 339 U
Slope Detection, Fm 225 Uhf and Microwaves
Slow-Scan TV 275-305
28 Communication with Klystrons 281
Sporadic -E Skip 15,18 Kilowatt Amplifier, 432 MHz (K2RIW)
Spurious Responses in Receivers 299
31 Lines and Circuits, Uhf 275
Stability, Receivers 30 Magnetron, Travelling -Wave Tube .
Stability, Transmitters 282
69,72,78 Parametric Amplifiers 285
Stacking Antennas 160 Pulse Communication, 2300 MHz .
Station Gain 304
24,25 Reception Above 420 Mhz 293
Station Planning 124 ROCLOC, Stable Klystron System
Stiffening Aluminum Tubing . 282
195 Transistors and Tubes, Uhf 279
Strip -Line Circuits and Filters 276,299,334 Tripler or Amplifier, 432 MHz
Stub, Corrective 297
167,169 Varactor Multipliers 287-293
Stubs for TVI Prevention 330 Varactor Principles
Sunspot Cycles 283
20 Velocity Modulation, Klystron 280
Superhet and Superregen 31 32 Waveguides and Cavities
SWR Indicators 307,313,317 277
440 -MHz Preamplifier 255,294
1296 -MHz Preamplifier 295
T
Technician Class Frequencies 14 V
Test Equipment 306-327 Varactor Multipliers, 432 MHz
Absorption Wavemeters 308
287
Diode Noise Generator Varactor Tripler, 1296 MHz 292
319 Variable -Frequency Crystal 340
Directional Coupler 317 VFO Coupling Circuits
Dummy Loads 84
321 V FO, Transistor 85
Field -Strength Meters 310-312 VXO, 50 to 450 MHz
Impedance Bridge 315
80
Lamps
Lecher Wires
306 w
323
Oscilloscope 114 Waveguides and Cavity Resonators 277
Slotted Line 324 Wavemeters, Absorption Type 308
SWR Indicator and Transmatch 313 Wide -Band Fm 79,220
T -Match 167
Tone Modulation, A2 28,347 Y
Transceiver, Transistor, 50 MHz 138 Yagi Antennas 153,180,189,207,209

352
I
 The whole picture of amateur
from basic fundamentals through the rodio
phases of this appealing hobby is most complex
Nether novice or covered In ARRL publications.
will help you to keep abreastold-time amateur, student or engineer,
of the times in the League publications
ever-expanding field of electronics.
THE RADiO AMATEUR'S
Internationally recognised, HANDBOOK
universally consulted. The all- ..EARNING THE RADIOTELEGRAPH
purpose volume of radio. Packed For those who find CODE
to the amateur and with information useful It difficult to master the code. De-
signed to help the beginner.
clear, concise manner, professional
contains
alike. Written in a
For borne study and Contains practice material
grams, charts and tables. hundreds of photos, dia- classroom use.
5. 50
THE RADIO AMATEUR'S
$5.50 HOW TO BECOME A RADIO
LICENSE MANUAL Tells what amateur radio Is AMATEUR
Study guide and reference book, points and
this fascinating hobby. Special how to get started in
the coveted amateur the way toward emphasis Is given the
tions and answers tolicense. Complete with typical ques- needs of the Novice licensee,
all of the FCC amateur station featured. with o complete amateur
Novice, Technician, Cbndiional, exams-
General, Advanced and
Extra Class. Continually kept up to $1.00
date. UNDERSTANDING AMATEUR RADIO
51,00 For the beginner. Explains
In simple language the
THE ARRL ANTENNA mentary principles of electronic ele-
BOOK
Theoretical explanation and
complete instructions for cludes how -to-build -it information and rodio circuits. In-
building different types of antennas for amateur work. transmitters and antennas: A "must" on lowsost receivers,
Simple doublets, multielement comer. guide for the new-
arrays, mobile,
others. Transmission rotaries,
long wires, rhombic and
exhaustively discussed. Profusely illustrated. lines are $2.50
SING-LE SIDEBA:':D FOR THE
$3.00 A RADIO AMATEUR
digest of the best sob articles
THE RADIO Air,ATEUR'S cussions of theory and practical from OST. Includes dis-
Written for the amateur OPERATING MANUAL
who prides himself scriptions of equipment. Covers "how -to -build It" de-
operating procedures. It is a ready on good sion. reception and transmis-
guide to good operating practices. reference source and
who wishes to brush up on Ideal for the amateur
HINTS _11ID KINKS
$;i.oe
who wishes information on all operating procedures, and If you build and operate an amateur
ating, facets of amateur oper- radio station, you'll
find this a mighty valuable
11.50 . book in your shack and work-
shop. More than 300 practical
FM AND REPEATERS FOR ideas. S0
A complete THE RADIO AMATEUR
treatment at fm and repeaters. A COURSE iN RADIO
chapters on receivers, Includes FUNDAA(EFi°ALS
Complete text for home study and
repeater operation. Of transmitters, antennas as well as
great value to the fm buff as well Lions at the end of each classroom use. Quaff
as those lust becoming chapter test the reader's com-
Interested in the subject. prehension of the material presented.
$3,00 the back of the book. Answers appear In
52

The; IMERICAN tÁD1

,
ti w,...," *.
;ii

-`. ''s--
tWLAY t,LEAGÚÉ

-` -
NEWI,VGTON, CONN. 06111