WOO'/\JOlS!40!

PllJUeopewe·NWWI

THESE FIVE WAVEFORMS, all at 1 kHz, are ASK ANYONE \\HO'S EVER BEEN ABLE TO whole bit. You can also use two gener­
typlcal of the kinds of outputs the generator will afford one, an<l they'll tell you that a ators-one to control the other one for
deliver. From top to bottom they are, obvlously;.
function generator is the handiest and really wild waveforms. Typical wave­
sine wave, triangle wave, pulses, square wave,
and al the bottom, a ramp. While the patterns most versatile signal source you can form photos of the basic instrument ap­
shown are at 1 kHz, all of these signals are own, regardless of your particular elec­ pear in Fig. I.
available over the full range of the generalor-1 tronic interest. Function generators pro­
Hz to 1 MHz.
duce various test waveforms-sine, How it works
triangle, and square at the very least, A single function-generator in­
and sometimes much more . Function tegrated circuit does the whole job prac­
generators are often the number one tically by itself, helped along only with
choice for digital logic testing and ex­ a transistor used to compensate ampli­
periments, electronic music, radio, TV, tudes. Fig. 2 shows us, in simplified
and hi-fi service, servo experiments, net­ form, that we can think of the circuit as
work synthesis, general electronics ser­ made up of seven blocks. The four most
vice and repair, ultrasonics, analog complicated ones are inside the IC. The
computation, and virtually anywhere basic object of the game is to generate a
else you need a flexible and stable wide­ waveform of the desired frequency.
range signal source. Then you select or alter its shape, con­
The problem is that today's func­ trol its amplitude, and then you control
tion generators cost at least $400, and its offset, or how much de there is to be
sometimes much more. This puts them in the output. Finally, you amplify it so
beyond the range of the student, service it will drive a low-impedance output.
technician, experimenter, or serious Fig. 5 shows us more details on
electronic hobbiest. But, thanks to a each block. In Fig. 5-a, we see that the
brand new $15 integrated circuit and waveform generator is a fancy version
some simplified switching circuitry, you of a multivibrator called an emiller­
can now duplicate most, if not all, the coupled asrable. It has two stable states,
capabilities of a commercial function one with Q3 conducting heavily and
generator for less than $40. one with Q3 off and Q4 conducting
The Radio-Electronics function heavily. Let's say that Q3 has just
generator has 30 pushbutton ranges that turned on, conducting heavily. Its col-

BUILD THIS $40 IC

Build this 30-range universal signal
square, pulse, and ramp

give you sine, triangle, square, pulse, lector voltage is low. This in turn holds
and ramp outputs over a six-decade fre­ the base voltage of Q4 low, temporarily
quency range from I hertz to I me­ insuring that Q4 is indeed off. If Q4 is
gahertz. Output amplitude is adjustable off, all of the available current that nor­
from 0 to 2 volts and may be either ac mally flows through Q4 goes into the
coupled, or de coupled with an adjust­ capacitor instead, charging the right end
able offset from -2 to + 2 volts, nicely of the capacitor negatively. Since the
giving you pulses and other test signals charging takes place from a current
of either polarity. Output impedance is source instead of a resistor, the charge
around 70 ohms, and the circuit drives on the capacitor is constantly changing
TTL directly. with time, resulting in a linear ramp or a
The complete specs appear in the straight-line charging action.
table. Complete kits and any and all in­ The capacitor keeps charging until
dividual parts are commercially avail­ it gets to -1.2 volts. At this point, Q4
able and you can put this instrument to­ will begin to conduct, and through
gether in a few evenings. Thanks to snap-action feedback, the circuit will
special printed-circuit boards, switch flip over, and Q4 snaps ON and Q3
wiring is greatly simplified, eliminating snaps OFF.
one of the biggest assembly headaches Now things have reversed, and the
on any piece of test equipment. An un­ capacitor starts charging in the other di­
breakable impact plastic case is used. If rection. It does so until the left side goes
you like, you can easily add external negative enough to flip things back the
AM, FM, pulse modulation as well as way they initially were, and you have a
voltage control, remote gating, the continuous back and forth oscillation.

36 RADIO-ELECTRONICS • SEPTEMBER 1972

 We note there are three outputs when we get to the sinewave output, fier. The gain is controlled by resistor
available. A ramp output from Q4, an anything but a 50-50 duty cycle will give R., and the difference between the two
out-of-phase ramp output from Q3 and us a bunch of second harmonic dis­ input signals gets amplified and appears
a square wave from Q6. If we take the tortion in a hurry. So, we set up a high as an output. A differential amplifier is
Q6 output, we have our square wave. If impedance pot and some big resistors a linear amplifier for small input signals
we take either the Q3 or Q4 output, we (R21,R22,R23 of Fig. 6) around Q l and or low gains. Its also a limiter or clipper
have our ramp. Now, if we take the dif­ Q2 to give us a symmetry control that for large input signals or high gains, for
ference between Q3 and Q4, we get a lets us touch things up to an exactly 50- you can get no more current out than is
triangle wave. And finally, if we take 50 duty cycle. available at the bottom of the circuit,
the difference between Q3 and Q4 and If we shunt a bunch of current nor less than zero. Converted to volt­
round off the sharp comers, we come up around one side, we get a very low duty ages across the load resistor, this means
with a sinewave. cycle. This is how we get our pulse wave­ you get a smooth clipping or clamping
form. Weuse the square-waveoutput, but action for high input levels or high
Changing frequency we unbalance things badly enough to get gains.
We can obviously change fre­ a l: 5 or 20: 80 duty cycle. By the way, this So, we change both the gain and
quency by changing capacitor values, also changes frequency, so the dial read­ drive level to improve upon our basic
for the charging time for a constant cur­ ings for the pulse output will be around waveforms. Triangle and ramp are sent
rent is directly proportional to how big half the actual pulse repetition rate. through at low level and low gain; they
a capacitor you hang on the circuit. So, So, the wave generator directly come out the way they went in. The dif­
looking at Fig. 3, we switch select a se­ generates a square wave and a pair of ferential amplifier nicely converts the
ries of capacitors that are decade mul­ out-of-phase ramp waveforms. If we two out-of-phase ramps into a uniform
tiples ( I OX) of each other. This takes take the difference between the ramps, triangular wave. For square and pulse,
care of range switching. we get a triangle wave. If we polish the we heavily overdrive the differential
To get a I O: I vernier range, we tops of the triangle wave, we get a sine­ amplifier and run the gain up yery high.
have to somehow change the value of wave. Finally, if we pick the square­ This gives us a constant amplitude, lots
the current source at the bottom of the wave and unbalance the symmetry, we of output, and sharp rise and fall times
circuit. We can't change what's inside get a pulse output. for these waveforms.
an IC, but we can add to or subtract For the sine wave, we use a moder­
from this current by sourcing or sinking Wave shaping ate amount of gain and start with a
some extra current at pin 13. We do this The shaping circuit of Fig. 5-b is triangle wave generated as the differ­
with an external potentiometer and cur- called a variable-gain differential ampli- ence between the two ramps. What this

FUNCTION GENERATOR
source that generates sine, triangle,
waveforms from 1 Hz to 1 MHz.

by DON LANCASTER

rent-limiting resistor. In the middle of ' l // l< 't /( )/ �

the range, the pot does nothing. Make it

more negative, and it adds to the avail­
able circuit current, and thus increases
frequency. Make it more positive, and it
steals some of the available current,
leaving less for the circuit and lowering
the frequency. We get a I O: I operating
range with reasonable care, and, ex­
cept for some pot loading, the voltage
versus frequency curve is very
linear.
We can also shunt some cur­
rent around Q I and Q2 and get
some added features. Nor­
mally the currents through
Q l and Q2 are very nearly
50-50, but they can go one
way or the other by a few .
percent. This makes the ..f31J/(J.
square wave slightly lZ >/'. •

asymmetrical, which f/!CfftJP/CS

i s n' t too b a d. •
Trouble i s t h a t --�

·--

www.americanradiohistorv.com
does is knock the sharp corners off the in the square and pulse modes since the
triangle waveshape, giving us the famil­ shaper heavily limits, and you get a con­ SPECIFICATIONS
iar sinewave shape, with a distortion stant output level. The problem isn't too FREQUENCY RANGE: 1 Hz to 1 Mhz,
typically around 2%. This technique is bad in triangle and ramp, but in the sine pushbutton selected by decades and
far simpler than those normally used in position, the waveshape changes pretty vernier adjusted over any one decade.
function generators, and this level of drastically as you reduce the amplitude, Overall accuracy 2-5%
distortion is not even noticeable for giving you "pointy" sinewaves at the FUNCTIONS: Sine, Triangle, Square,
anything but critical audio testing. high end of any frequency range. To Pulse, and Ramp
Biasing is critical on any differ­ beat this, we raise the gain of the shaper AMPLITUDE: Variable 0-2 volts Sine,
ential amplifier. Both inputs must be re­ slightly at the far end of the frequency Square, Pulse; 0-1. 75 volts Triangle and
turned to the same de level, or the circuit pot. The drop in input amplitude gets Ramp. TTL compatable output using
goes into limiting and ignores the input made up by the increase in shaper gain, + 0. 6-volt baseline offset.
signal due to the unbalance in currents and the output sine wave amplitude and OFFSET: In DC Mode, adjustable from -2
caused by improper biasing. waveshape stays constant over the pot's to + 2 volts giving high, centered, or
There are two ways we use the dif­ range. low baseline. Capacitor coupled in ac
ferential amplifier in the Radio-Elec­ Since the pot's gomg the wrong mode.
tronics function generator. For triangle
and sine, we connect the inputs directly
AMPLITUDE
to the ramp and out-of-phase ramp out­ __ __,CORRECTOR
AC-DC
puts. These are both at the same de 01
COUPLING
=
level and thus give us proper bias. At
the same time, the differential amplifier
RANGE WAVE FUNCTION SHAPING AMPLITUDE OUTPUT
takes the difference between these two SELECTOR GENERATOR SELECTOR CIRCUIT MODULATOR DRIVER
waveforms, generating a triangle wave S2 I Cl-a S1 IC1 b IC1·c ' IC1 d
at low gain and a rounded top triangle
approximation to a sine wave at higher
gain. FREQUENCY LEVEL OFFSE1
The rest of the waveshapes must be
capacitor-coupled. To do this, we return
both inputs of the differential amplifier

I �---1- h
to ground and bypass one input. This

I
gives us a single-ended amplifier into
which we can couple these square,
0
pulse, and ramp inputs. We run the I
"Xl"
S2-a
ramp at low gain, so the same signal 0 o-J.-.o
comes out we put in. In square and
pulse, we run very high gain by shorting
Rx. This gives us a bigger output with
(
(
C16* 240µF
f
completely fiat tops and fast risetimes. 7
Two small refinements complete the 0 o-f.o 0
,--�-- "XlO"
I
square and pulse coupling. The cou­ S2-b
0 o-1-.o

r
pling capacitor is a compromise, since
one that gives no droop on the I-Hz ( C15• 24µF
range also gives obj ectionably long (

-
,
transients on the higher ranges as fre­
0
quency is suddenly changed. To beat er: 0 o-f.o
, ----
0
this, we use a relatively small coupling <( "XlOO"
0 I
S2-c
"' 0 o-J.-.o I

j
capacitor C6 for the higher ranges and u
u
Cl. I-
switch in a larger C 10 for the I-Hz
z E--- C14*2.4µF 3:
(/)
range. The pulse waveform has a large <( (

--
::;; 0
u..
de component due to its duty cycle. If
0 Cl.
we capacitor-couple, we end up with I- 0
0 o-f--o
,- 0
-- I-
linear amplification on the bottom of I "XlK"
the pulse and limiting on the top. This o-J.-.o S2-d

r
o·
reduces pulse amplitude, but worse yet,
it makes the bottom noisy and rounds ( C13* 0.24µF
(
---
the edges of the rise and fall times. To
beat this, we purposely unbalance the 7
differential amplifier slightly with R27, 0 o-f-o
, 0
-- "X10K"
I
but only during pulse operation. This S2-e
0 o-L.o

r
nicely limits both the bottom and the
top of the pulse, giving us an output as (- C12* 0.022µF
big and as clean as the square-wave. --IE
Amplitude compensation
There is a slight change of wave
0 o-f--o
,--
I
-- 0-

0
7

"X100K"
S2-f
generator output amplitude when we 0 o-L.o
operate over a 10: I current range. Most
obvious is a dropoff at the high current, *TWO CAPACITORS IN
PARALLEL SEE TEXT.
high frequency and caused by larger
drops in Q3 and Q4. The high-end volt­ S2 IS 6 STATION 4 PDT INTERLOCK! NG ALL SECTIONS SHOWN IN "UP" POSITION
age drop gets eliminated automatically

38 RADIO-ELECTRONICS • SEPTEMBER 1972

 way to do us any good, we invert it with We control the gain with an elec­
.A.IS a transistor and use this control voltage tronic modulator as shown in Fig. 5-c.
SPECIFIC 1r�que ncy sine to steal a small amount of the shaper's We can also inj ect signals here for am­
DISTORTION: over vernier current source. This increases the sha­ plitude modulation, remote volume
$i 2%; 3%
wave, less • per gain as Rx is now smaller with re­ control, keying, etc. If we like. The cir­
range cuit is a true multiplier, which means it
typ1caI ; square spect to the circuit currents than it was
DUTY CYCLES: Pulse 20%
·

. not before. By keeping the transistor satu­ doubles as a balanced modulator. The
adjustable to 50%. Dial
calibration
r ated for part of the pot's range, we circuit works by letting both the input
p u ls e ra n ge .
accurate on don't start any correction till we need it. and the control provide only a fraction
OUTPUT IMPEDANCE
: 70 Ohms. Zener of the available current to the output.
M ore and more gain is added as we get
protected Since the two fractions are cascaded, the
6%"; grey impact plastic to the end of the pot. Looking ahead at
SIZE: 2'/2" x 5" x
Fig. 6, we see that R3 l and R32 decide product of the input and control appears
case.
how much gain we add at the high end, at the output. Polarity may be con­
WEIGHT: 2 pounds
AM, while the ratio of R29 to R30 decides trolled by reversing the difference in
MODULATION: None in basic unit;
te when in the pot's rotation the gain cor­ voltage on the control inputs. Zero dif­
FM , VCO, remo gain, keying, etc. are
rection is to begin. ference gives zero gain, and one volt
easily added to the basic circuit.
gives full gain, the polarity of the input
determining which side of the wave­
FIG. 2-(left) BLOCK DIAGRAM OF GENERATOR gives an overall look at how the unit operates.
shape ends up on top. The electronic
The one IC does the lion's share of the work.
gain control is internally directly con­
FIG. 3-(bottom lett) RANGE-SWITCHING CIRCUIT Is on a separate prlnted-clreult board. An lnter­ nected to the shaper; all we have to do
locklng push-button switch, th at mounts on this board, Is used. is connect up an external gain control or
signal. Note that no waveforms have to
FIG. 4-(below) FUNCTION-SELECTOR SWITCH torrn• another subassembly on a separate circuit
board. Again an Interlocking pushbutton switch •• used and Is mounted on the board. travel through the pot and we can easily
and safely use long leads without worry­
ing about high frequency response falling
off and causing misleading conclusions
,------
from the tests you are performing.
� 0 0
OFF
I Output buffer
Sl·a
o-1--0 0 0 A Darlington differential amplifier
converts the output signal to one with
enough current drive to handle a short
circuit or another low-impedance load
without damage. To adjust the offset, or
\ ---,
how much de we get out in addition to
SINE the waveshape, we capacitor-couple the
Sfb bufef r's input and add a variable bias
0
that we control with the offset poten­
R28
tiometer. We also provide an output
22K coupling capacitor that we short out
when we want variable offset and insert
0 for ac work.
TRIANGLE
-= I
Sl-c
I o-1--0 0 Circuit of the generator
u
I- The complete schematic is shown
�
(/) in Figs. 3, 4 and 6. Capacitors C 1 1 to
u..
0 C 16 select the decade frequency ranges,
Q. ---·, while R17 selects the vernier frequency
0
I-
PULSE and drives amplitude compensator QI.
s:;-:d Pots R9 and R 1 1 put Jimits on the fre­
quency control, so we can calibrate the
0
I- upper and lower frequencies on the
dial. Selector SI picks the function we
want and controls power to the + 10,- 10
volt conventional Zener supply.
SQUARE R 19 controls the triangle wave­
s,-:e- shape, while R 19 and R20 together con­
trol the sine waveshape. R2I to R23
provide the symmetry control, while
R 13 controls offset. Square, pulse, and
ramp coupling is handled by C6, helped
along C10 on the I-Hz range. R27
RAMP
Sj':f I switches in only during pulse operation
o--l-0 0
to give a good pulse waveform.
The switching may seem com­
R27 plicated, but its really simple, partic­
1 MEG ularly since the switches mount directly
Sl - 6 STATION 4 PDT on the board, SI-a controls the ac
INTERLOCKING ALL
SECTIONS SHOWN IN CONNECT TO FRONT
power. If this button is up, the power is
"UP" POSITION "GND" � PANEL AT J2 WITH on. S 1-b handles the sine wave. It first
7
HEAVY WIRE connects the shaper differentially to the

SEPTEMBER 1972 • RADIO-ELECTRONICS 39

 +lO
16 J Oi..
..-.ate a trn.hase ramp outpu ts to
�wi tches in enough� wave. Then it
t.gi\i?crease m the
Q5
shaper with R20
to
outp ut. Finally '>e sine wave
. it prevem,,
gettmg m to the' circuit ll22 from
R22 I�. \l.sed to
·

bias the sha

pe.
·

12
Q4
_fUl_rL_
SQUARE OUT
That's all we hav
e room for In this Is­
sue. Next month we
wlll conclude the
story by presenting the
remalnlnq text on
g
-v1&/1-
construction and operation
alon with a
complete set of circu it-board patterns
,
RAMP OUT both foll and parts placem ent diagrams.
TIMING
CAPACITOR R-E
-REF Q1

�
PARTS LIST
C1-1 µF. 400 volts. mylar
C2,C7-0.22 µF, mylar
OUT-OF-PHASE C3,C9-500 µ.F, 12-volt electrolytic
RAMP OUT C4,C5-1000 µF, 25-volt electrolytic
C6-3 µF, S-volt tantalum
INTERNAL
DIFFERENCE BETWEEN CB-6000 µ.F, 10-volt electrolytic
CURRENT
RAMPS IS A TRIANGLE C10-50 µF, S-volt tantalum
SOURCE
C11-Two parallel capacitors, 2400 pF, 2%
9
+ C12-Two parallel capacitors, .022 µF, 2%
-10 C 12-Two parallel capacitors, 0.24 µ.F, 2%
·FREQUENCY
C14-Two parallel capacitors, 2.4 µ.F, 2%
C15-Two parallel capacitors, 24 µF, 2%
A C16-Two parallel capacitors, 240 µ.F, 2%
C11-C16 Value shown is total capacitance
See October issue for capacitor selection de-
tails

J1
C17-30-µ.F, 12-volt electrolytic
+
C18-0.1 µ.F, disc ceramic
Vour _
TO __ D1,D2-Silicon power diode, 1 amp, 50 PIV,
_ + ·
MODULATOR 1 N4001 or equiv.
INPUT INPUT
03,04-10-volt Zener, 1N4740 or equivalent
A B GAIN CHARACTERISTIC
05-15- to 20-volt Zener, 1N4744 or equivalent
5 6 F1-0.1-ampere fuse
IC1-XR-205 Function Generator IC (Exar)

MMM
EFFECT ON TRIANGLE WAVE:
J1,J2-5-way binding posts, 1 black, 1 blue
01-2N5129 transistor, silicon npn
R1,R25,R28-22.000 ohms, 14 watt
7 8 R2,R3,R29-4700 ohms
Rx
R4,R31,R32-47,000 ohms
LOW GAIN MED GAIN HIGH GAIN
SETS GAIN R5,R21,R22.R30-10.000 ohms
BIG Rx MED Rx LOW Rx OR
R6,R24-12 ohms
SHORT
R7-1500 ohms
RS-33,000 ohms
B
R9-1000 ohms upright PC potentiometer
R10-680 ohms, '/, watt
+10 R11-1000 ohms upright PC potentiometer
14
OPTIONAL OUTPUT
R12.R18-330 ohms, watt
OUTPUT 2 R8
R13-50,000 ohms linear potentiometer with
spst pull switch
� - R14-5000 ohms linear potentiometer
FROM R15,R16-100 ohms, '12 watt
MODULATOR R17-1000 ohms linear potentiometer
R19-5000 ohms upright PC potentiometer
R20-5000 ohms upright PC potentiometer
FROM Rl
INPUT R23-250,000 upright PC potentiometer
SHAPER 10 R26-1000 ohms, '!.watt
14
+10
R27-1 megohm, watt
16 S 1,S2-six station, 4PDT interlockir\g
S3-spst pull switch on R13
T1-Power transformer: 20 Vet @ 60 mA, PC
mount, Signal PC-20-60

-10
MJSC: Main, Selector, and Range PC boards,
punched lront panel; Impact plastlc
case; llne cord; '!.'' knobs (2); skirted
2'1•" knob with speclal callbratlon; Hat
INTERNAL
cable or wiring harness; capacitor cllp;
CURRENT no skid leet; luse cllps; PC termlnals;
SOURCE R3
angle brackets (4); mounting hardware;
9 TO plastlc machine screws lor lront panel;
OUTPUT VIA mylar button callouts; ground lug; wire;
-10 COUPLING & aleevlng; solder.
PROTECTION
R7
CIRCUITS NOTE: The lollowlng are avallable lrom South­
west Technical Products, 219 West
c D Rhapsody, San Antonio, Texas, 78216
Set ol three PC boards, etched and drllled FGb,
FIG. 5-SIMPLIFIED DIAGRAMS OF PORTIONS ol the waveform generator's circuitry. a-basic emit­ $8.25
ter-coupled astable multlvlbrator. b-shaplng circuit Is called a varlable-galn dlllerentlal ampllller. Complete kit ol all above parts FG-1, $39.95
c-electronlc modulator controls the gain. d-output driver.

40 RADIO-ELECTRONICS • SEPTEMBER 1972

www.americanradiohistorv.com
 117 VAC

Fl
0.5A

All
R10 lK
680>'1 MIN
PREO

C4 C5
lOOOµF, 25V lOOOµF, 25V
R12
-lOV
� (---.-�
- ���__,___�+ E't------i
l 33on
01 D2 =
R15 Rl6
loon IN4001 10o n
D3 +lOV
...L

C8
= IN4740 +r 5000
_L 10
-10\/ +1OV
(SOURCE)
01
2N5129
R18
33on
R21 R23 R20
R30
lOK 250K 5K R19
lOK
fil!i 5K
ADJ TRI
SYMMETRY R22
ADJ
10K +lOV

+ C9
+lOV -lOV 5 0
0

T
R31 12V
IZ.7K

-10V 0
1-

J
R32
47K

ICI
C18
XR205
o.1
TOP VIEW T
+
C6
-��� OD��
.-M
���--,
3
x y
Cl
.22

C10
R5
50µF
R3
C3 R4
C2 4.7K
500 47K
+

RS
.22
33K µF =

R7
-lOV
R6 1.5K
+lOV
12n

Cl
1 µF
400V 'GND"
CW L---'\N..--� CCVJ FRONT ::>N FIG. 3A
R13 PANEL
FIG. 6-MAIN GENERATOR cir­
50K BLUE BLK
cuit. When two switch boards
OFFSET Jl J2
are added (Fig. 3 and Fig. 4) to
this master board, the generator L.._O
___ UTPUT�
Is complete.

www.americanradiohistorv.com
R-E's Function
Generator
Part II: Final assembly and construction details
including printed-circuit patterns and parts r
placement diagrams

by DON LANCASTER

LAST MONTH WE PRESENTED THE FIRST

part of this article on a $40 function gen­
erator build around a single IC that could
deliver five different waveforms. This is­
sue we complete that article by presenting A?t@{J.
the remaining text and diagrams. fkctro11ics ,.......,,..�\
The pulse is handled by S 1-d. It •

first set<; up the shaper for single-ended

capacitor-coupled operation. Next it •
shorts the shaper's gain control. giving capacitors giving you high accuracy.
full limiting. Finally, it unbalances the For instance C 13 is supposed to be much, and even if it does. it just raises
wave generator with R26 to get the 0.241iF. Virtually any 0.22-iiF capacitor the frequency a bit. So the reasonable
pulse duty cycle. If we had a fifth deck you get will be lower than this value, so thing to do is to use ordinary high-qual­
on S 1-d, we'd use it to switch in the sha­ get cne and measure it. If it is exactly ity electrolytics. and if extreme accuracy
per unbalancer R27 needed for good 0.22µF, put a plain old 0.02-11F in paral­ is needed, check up on them every six
pulse amplitude and risetimes. We lel with it. If it's 0.23µF, use an O.O l1iF. months or so.
don't have this switch available, so we If it's 0.2 l{tF, use a parallel 0.03µF, etc. Note that C 1 1, the XIOOK capaci­
use the normally closed contacts of S 1-e If you don't have a good capacitance tor is in the circuit all the time, and that
and S 1-f, the SQUARE and RAMP select­ bridge handy, use a scope timebase and C 12 is slightly smaller as a result. This
ors. Thus if the PULSE switch is de­ adjust for equal 10: I frequency steps. prevents the function generator from
pressed, SQUARE and RAMP obviously The big capacitors are always a taking off at a very high frequency in­
are not, and the resistor is switched in. hassle on any low-frequency oscillator. between button pressings. and also
S 1-e is the SQUARE select which Ideally, you should use bipolar tan­ eliminates a minor starting problem.
connects the shaper for a single-ended talums, but these are too big and too ex­
capacitor-coupled input. It also runs the pensive to be reasonable. From a prac­ Building it
shaper gain wide open, and prevents tical standpoint, the total ac voltage in Three PC boards are needed for as­
R27 from being in the circuit. Finally, it the circuit is less than a volt rms, and sembly. Two of them support the
connects the wave generator's square­ the impedance is very high. So, even on switches, while the third holds almost
wave output to the shaper. very long term usage, an ordinary good everything else. These are commercially
S 1-f is identical to S1-e, except it quality electrolytic won't run down very available, but if you are building your
connects the ramp output to the shaper. own, complete details appear in the fol­
S2 is much simpler. All it docs is lowing pages. You can start assembly
switch in a selected capacitor for each with the switch boards, placing the
frequency range. It also switches in an smaller components and wire jumpers
anti-droop capacitor C10 only on the exactly as shown. The switches are then
XI range. soldered in place. After that, any termi­
nals that mav interfere with the main
About the capacitors circuit boarct' may be clipped short.
The instrument will be a no more Wire the switch assemblies to the main
accurate than the capacitors you select. board last.
2% tolerance or better yet I% tolerance To put together the main board,
is needed for range-to-range accuracy. start with the input and control termi­
To beat buying very expensive capaci­ nals, followed by the jumpers, fusehol­
tors, you use a parallel pair of capaci­ der, resistors, and finally the major
tors with roughly a IO: I range of value. components. Use sleeving on the jump­
M easure the big one and select the FOLLOW THIS PATTERN to make the dial for ers where indicated. IC I is identified by
small one, and you get by with plain old your unit. Final size is 2 9/16-ln. wide. (continued on paie 50)

OCTOBER 1972 • RADIO-ELECTRONICS 45

www.americanradiohistorv.com
a code dot and notch beside pin one. Its
shown top vie»· everywhere. Be sure
pins I and 2 of the power transformer
go to the 1 10 volt line side or you'll • • •
have extensive damage on your hands.
Also be sure to watch out for diode and
capacitor polarities. •
After the main board is complete,
loosely bolt the switch hoards in place
with right angle brackets. Be sure the
function selector is on the left and the
frequency selector is on the right. The
main circuit board may then be bolted
temporarily into the impact plastic case.
Now, check the front panel alignment
with the pushbuttons to make sure ev­
erything works smoothly. If you have
to, adjust the angle brackets slightly to
get smooth operation. Once everything
fits nicely, tighten the switch circuit
boards down firmly and remove every­
thing from the case.
Wire jumpers or resistor ends may
now be used to solder together the pads
on both switch boards to the main
• • •

ii
board. You can then mount things back
into the case, adding the line cord and
an anti-strain knot. No-skid feet may
also be added.
Mount the front panel components
FOIL PATTERN OF THE MAIN CIRCUIT BOARD. Actual size of this board Is S Inches high and S'I•
and wire them to the main circuit using lnches wide. Two other boards plug Into this one. They are the range-select and mode-select assem­
a small harness or flat cable if you have blies, also shown on these two pages.
it; or neatly laced cable if you do not.
The frequency dial is slightly non­ SWTP
linear; a suitable replica is on page 45,
or you may prefer to calibrate your own t M
·1 f
dial for extra accuracy. Either way, be I
CD
"
u u
N
u
sure the dial markings coincide with the
l
ii
pot's range when you push the knob in �
place.
A single heavy ground lead must S2 RANG!'= Sl"LECT
be added from the terminal provided on
the function selector board to the front
panel and output Black terminal. Be
0
sure to connect this wire exactly this
way. Other grounding schemes (or a
lack of them) can give you a little glitch
in the top and bottom center of the sine­
w ave. After applying stick-on M ylar
callouts to the pushbuttons, your unit • •
should be complete and ready for test.

Preliminary checkout RANGE-SELECTOR BOARD DETAILS. Foll pattern of board must be enlarged to S'l• Inches wide. Top
diagram shows parts placement on this circuit board.
Center all the controls, and apply
power very briely, measuring the volt­
f
ages across Zener diodes D3 and D4, that the shaper and modulator are also calibration will fix this. The same thing
and looking for the + 10 and -IO val­ working. Finally, check the output for is true of the sine; and ramp; while they
ues. If the voltages are low or if a scope the same 2-volt square wave. might look funny before calibration, the
shows hum at this point srop and find Next, check out the offset adjust­ important thing here and now is that
our whar is causing rhe excess currenr ment and proper operation of the pull they exist. If everything seems to be
flow If things look OK, apply power switch. With the switch OUT, you should there, you're ready for calibration. If
again for 15 seconds and measure any get de coupling and variable offset. there's any problem, go back to figure
temperature rise in ICI with your fin­ With the switch IN, you should get ac two and isolate where the problem
ger. It should be negligible or only very coupling. Check the frequency pot over seems to be. Don't attempt to calibrate
slightly warm. Again, if its too hot, stop its range. Don't worry if funny things an instrument that doesn't seem to be
and find out why. happen at either end unless they're still working right.
Now, set the unit to a 400-hertz there after you go through the calibra­
square wave and with a scope, check tion procedure. Check for droop on the Calibration steps
pin 12 of IC I for a 2-volt square wave. I-Hz range. Now, check the other func­ For top-notch performance, you
Check for out-of-phase ramps at pins 14 tions. Pulse should look like square, should calibrate your function gener­
and 15. Turn up the level. You should only with a 5: I duty cycle. Triangle may ator. You'll need a good oscilloscope for
get a 2-volt square wave here, verifying be rather small or may have bent tops; this and, while not essential, a counter

50 RADIO-ELECTRONICS • OCTOBER 1972

www.americanradiohistorv.com
 Your instrument is now fully cali­
brated and ready for use. If you have no
means of calibration at all, just set all
the internal pots to their mid range, and
you'll still have a reasonably useful, if
somewhat inaccurate piece of test gear.
After you've gone through the cali­
bration, any end effects that showed up
on the frequency knob should now he
w ,..1.._
___ 1..., gone. You might like to doublecheck
R13. +
the amplitude compensator QI. To do
-0 �FSET this, switch to X 100 and a triangle wave
tcwo and compare the " I " and the "IO" am­
- R SI -·
R I plitude on a scope. They should be very
-R32_::;- l-
.--R8� C8 ; nearly the same. Ir the "IO" is smaller,
I try lowering R3I and R32 slightly, and
vice versa. If there is a hump in the
middle, try lowering R 30. If there is a
dip in the middle, lower R29. Optimum
compensation should be a very slight
dip followed by a very slight rise. Fi­
nally switch to sine and note an undis­
torted and uniform sinewave over the
whole range. Adjustment of the ampli­
PWR
tude compensator is a very fine point
PWR
and shouldn't be necessary in a properly
-J - working unit.

Operating hints
Use the de position with zero offset
MAIN CIRCUIT BOARD SHOWING WHERE PARTS ARE MOUNTED. In addition, the mode-select
board and the range-select board are connected to the main "mother" board. The only wiring ex­
unless you have to couple into a high­
ternal to these boards connects to the front panel ol the Instrument. voltage, tube-type circuit or something
similar. Note also that the ac-coupling
capacitor is rather small and forms a
high-pass filter whose lower cutoff is de­
termined by the load you hang on the
generator. For very-low-frequency work
or very heavy loads, you'll want to sub-
stantially increase the value of this ca­
GMO
pacitor.
To drive TTL. use the pulse or
square output and turn it up to the full 2
volts. Now adjust the offset control for a
positive baseline oflset of 0.6 volts. This
gives you a signal that exceeds the TTL
1-0 requirements and you can easily
drive any TTL logic circuit. DTL and
RTL are also easy to drive, and a simple
interface works with MOS logic. For
• MECL logic. simply use some negative
offset and you're home free.
If you need some of the fancier ca­
MODE-SELECTOR BOARD DETAILS. Foll pattern ol this board must be enlarged to 5y, Inches wide.
Top diagram shows parts placement on this circuit board.
pabilities of professional function gen­
erators, you can easily pick them up.
For instance, you can apply a differ­
would also be handy. readjust SINE ADJUST. ential control voltage to pins 3 and 4.
To start the calibration, center all 4. Switch back toSQUARE, and set This gives you an amplitude modulator,
pots. Select de coupling with zero offset, the frequency knob to "I" on the 100- a remote volume control, a phase con­
maximum amplitude and a 400-Hz Hz range. Now adjust the MIN FREQ po­ trol (reverse the polarity for an upside
square wave. Now: tentiometer R11 for a 100-Hertz square­ down waveform, or a way of off-on key­
I. Adjust the SYMMETRY control wave or a JOO-millisecond period. ing.) You can also use this for single
for exactly a 50% duty cycle of the out­ 5. Change the FREQUENCY knob to sideband modulation or ordinary AM
put when viewed with a good scope. "IO" and adjust MAXIMl.JM FREQUENCY by capacitor-coupling a suitable signal
2. Switch to TRIANGLE and adjust potentiometer R9 for a I-KHz square­ to the present level control input. Or, a
for the largest undistorted triangle out­ wave or a I-millisecond period. voltage replacing the frequency control
put you can get and then back it off just 6. Redo steps 4 and 5 as often as gives you a vco (Voltage Controlled
a bit. Use TRIANGLE ADJUST control R l9 you have to till there's no more inter­ Oscillator). a frequency modulator, or a
for this step. action. Twice through is usually all it way of frequency shift keying. For high
3. Switch to SINE and adjust for takes. modulation frequencies, you'll want to
the best looking sine wave using SINE 7 Verify operation of the other fre­ reduce C9 or delete it entirely. Its in
ADJUST R20. If you ever readjust the quency ranges and the PULSE and RAMP there to keep from hum modulating
TRIANGLE ADJUST, you'll also have to outputs. with external pickup. R-E

OCTOBER 1972 • RADIO-ELECTRONICS 51

www.americanradiohistorv.com
 THE FIRST
letters
CROWN
PREAMPLIFIER
MORE ON THE the square and pulse modes.
FUNCTION GENERATOR All of the commercially available kits
Several problems crept into the Ra­ have picked up all of these additions
dio-Electronics Function Generator story and corrections and include a slightly
(September and October 1972). The heavier power supply than the original
pulse waveform is shown upside down in called for.
the photos. On the overlays, timing ca­ DON LANCASTER
What would happen to a preamplifier
pacitors C11 through C16 are shown in­ Phoenix, Ariz.
design. if the design engineer could free
terchanged; the largest capacitor goes
himself from stereotyped ide<1s and start
with the lowest frequency at the top of NEED SERVICE DATA
fresh with only a list of customers' re­
fhe instrument. The input POWER callouts They have opened a new Radio &
quests? Well. at CROWN that has just hap­
should, of course, go up with the input TV Repair Course here in high school
pened, and the result is the IC150. an
terminals on the fuse and power trans­ for juniors and seniors. We are in need
exciting "new concept" control center with
simplified circuitry. controls that are easy former. of a complete set of Sams Service notes
·to understand and use. several exclusive Resistor R27, the· 1-megohm pulse from 1 to date. or we would accept 500
features, unsurpassed quality, and - to top unbalancing resistor should go to + 10 to date. We are also in need of used test
it all off - a lower price tag. volts as called out in the schematic. The equipment of all kinds. We will be glad
Crown Engineers discovered that pre­ PC overlay and master has this going to to pay a reasonable price and will an­
amp switches don't need to pop that -10, which gives you one-third amplitude swer all letters.
there is something better than the stereo and very noisy pulse putput. Carve and RALPH DOROUGH, Instructor
mode switch that the phone preamp
jumper the PC board so that this resistor Killeen High School
can be dramatically improved and.
runs to + 10. Vocational Building
that by using !C's. a versatile high-qual­
There is enough linearity variation 3101 Clinkenbeard Dr.
ity, advanced- performance preamplifier
from IC to IC that you'll want to custom Killeen, Texas
can be priced to beat inflation.
calibrate your frequency dial to match
Of course. the true uniqueness of such
an innovative design cannot be appreciated the particular IC you are using. A larger WHO ARE THEY?
by reading about it. The only answer is value resistor for R 10, perhaps 1000 have a question about the article
to experience the IC 150 yourself. Let us o hms or 1200 o hms should be used un­ "State Of Solid State." Radio-Electron­
tell you where Crown's "new concept" is less your particular IC will not provide ics, October, 1972. At one point the text
being introduced in your area. Write today enough range. This improves linearity mentions Signetics and Integrated Sys­
for a list of locations. and makes th.e control setting easier and tems Inc. At another point the text men­
more stable. tions EXAR IC's. Are EXAR and In­
World's quietest phono preamp The amplitude compensator shown tegrated Systems the same? I am most
Infinitely variable stereo panora11\i.J control interested in building a low-cost
seems to work only with certain XR-205
Siient sw11cti1ng and automat.c muting
at turn-on and turn-off chips. A better circuit looks like this: speedometer and tachometer. Could you
Integrated circuit modules provide information on a source for IC
lndusuy·s lowest d1stort1on levels
5{11 \I( circuits?
Full range tone and loudness conirols
Guaranteed phase response ROBERT N. SLEGHT
3-vear parts and labor warranty
Will drtve any ampllf1er Laramie, Wyo.
... R.10
S269. walnut enclosure S33 ..
l........... , 'i:oMf".o.,1io\T10N .,
(°lj"'"""''' ,.kllE.$�0\.J) EXAR and Integrated Systems are
•o•
(wu A 1"111:6> lt_t.a.•aTolll) indeed the company. The correct full
name and address is EXAR Integrated
Note that all we've done is add a se­ Systems Inc., 733 North Pastoria Ave.,

ries resistor to the base of 01 a:nd ex­ Sunnyvale, Calif. 94086.

change the old 4700-ohm R29 fixed re­ I have not, at this date, seen any cir­
sistor to ground for an adjustable PC cuit diagrams that would meet your
pot. This setup will work with any IC needs for the speedometer. We are,
Ask your dealer also about Crown·s new com· used. You adjust things so the voltage at however, negotiating with an author for
pan1on 0150 power amplifier, which delivers 200 point A stays near ground for FREQUENCY the purchase of his article on a digital IC
watts IHF output at 8 ohms or 350 watts at 4 ohms. dial settings from 1 through 6. Above 6 it tachometer. Tentatively, this story is
No amp in this power range · however expensive -
should start smoothly swinging positive, scheduled for publication in our April
has better frequ ency response or lower hum,
noise or distortion. It offers performance equal ending up with + 8 or more when your 1973 issue. -Editor

to the famous DC300, but at medium power and frequency dial reads 10. For more cor­
price. It's worth listening intoi rection at the high end, reduce R31 and NO MORE SUPPLIERS?
R32, and vice versa. The new pot de­ recently ordered some parts from
Exported as cides when in the FREQUENCY dial rota­ Newark Electronics. Their reply to me
AM CRON tion the correction begins. Amplitude was "We no longer accept orders from
compensation only affects the sinewave individuals. Therefore we are returning
® distortion and the high frequ ency your order."
BOX 1000, ELKHART. INDIANA 46514. U.S.A. triangle and ramp size; it is bypassed in STEVEN W. RUSSELL R-E
Circle I I on reader sen·ice card
26 RADIO-ELECTRONICS • JANUARY 1973

www.americanradiohistorv.com