ELECTRONIC TECHNOLOGY SERIES

,,

a publication
 $.90

MULTIVIBRATORS

Edited by
Alexander Schure, Ph. D., Ed. D.

JOHN F. RIDER PUBLISHER, INC.

480 Canal Street • New York 13, N.Y.
 Copyright 1956 by

JOHN F. RIDER PUBLISHER, INC.

All rights reserved. This book or parts thereof may

not be reproduced in any form or in any language
without permission of the publisher.

FIRST EDITION

Library of Congress Catalog Card No. 56-8364

Printed in the United States of America

 PREFACE

Originally "multivibrator" meant only a free-running, square-

wave oscillator normally utilized to produce either harmonics or
subharmonics for test purposes. Current applications of multivi-
brator circuits are so extensive that a listing of just a few of the
specialized devices utilizing this type of oscillator indicates the im-
portance of the subject. Cathode ray oscilloscopes, television re-
ceivers, tv cameras, tv transmitters, computers, radar, electronic
switches, multiplex telegraph transmitters-these are but a few of
the specialized devices that require these particular types of relaxa-
tion oscillators.
The necessity of a careful study of this topic cannot be stre~
too highly. The book has been organized to help the student crys-
tallize the important ideas pertaining to multivibrators and their
circuit applications. Although the treatment is nonmathematical,
the analyses are sufficiently extensive to permit the technician,
practicing engineer, or advanced student to develop these funda-
mental concepts and basic applications to best advantage. No at-
tempt is made to treat or list all possible applications since this
is an obvious impossibility in a work of this size. The illustrative
circuits chosen, however, are sufficiently representative of major
design considerations.
Particular attention is given to the basic principles of the
multivibrator and the major types, namely: the bi-stable multi-
vibrator, mono-stable multivibrator, and a-stable multivibrator.
Less definitive terms (sometimes more popular in terminology)
are explained to help the reader recognize them. A sample design
problem is also presented in sufficient detail to permit an under-
standing of the procedures followed in a design analysis.
Grateful acknowledgement is made to the staff of the New
York Technical Institute for its assistance in the preparation of
the manuscript for this booklet.

New York, N. Y. A.S.

February 1956

iii
 CONTENTS

Chapter Page

1 Basic Principles

2 The Bi-Stable Multivibrator 6

3 Mono-Stable Multivibrator 24

4 A-Stable Multivibrator 36

iv
 Chapter 1

BASIC PRINCIPLES

1. What a Multivibrator is

There was a time when the word "multivibrator" referred only

to a free-running, square-wave oscillator used to produce harmonics
and subharmonics for test purposes. Today the term carries a much
deeper and broader significance. Multivibrators are found in tele-
vision, radar, computers, long distance communications systems,
diversity receiving equipment, electronic switches, radio-control,
stroboscopes, cathode-ray oscilloscopes, and an incredibly large
number of other specialized devices.
A multivibrator is a type of relaxation oscillator1 employing a
two-stage amplifier, with the output of the second stage regene-
ratively coupled to the input of the first stage. However, within
these limits there are different types. These types differ somewhat
in circuitry and considerably in function. Various terms have been
used in the past to distinguish them. Some of these terms are in-
consistent, some ambiguous or indefinite. In this book, each im-
portant type is discussed in a separate chapter. The latest accepted
terms are used in referring to these types, and each chapter heading

I A relaxation oscillator may be defined as an oscillator in which one or

more voltages or currents change abruptly and periodically to produce, as a rule,
a non-sinusoidal waveform. Relaxation oscillations are generally obtained by
utilizing· the time delay of capacitive charge or discharge through a resistor.
2 MULTIVIBRATORS

1s the proper term for the type discussed m that chapter. These
heading are:
Bi-stable Multivibrator (Chap. 2)
Mono-stable Multivibrator (Chap. 3)
A-stable Multivibrator (Chap. 4)
Other popular tenns, less authoratative or definitive, will also be
explained later, to help the reader recognize them.

2. Definitions

Before beginning the discussion and analysis of multivibrator

circuits, it is necessary to define some of the terms that will be
used in subsequent chapters.
Multivibrator. A type of relaxation oscillator consisting basic-
ally of a two-stage resistance-coupled amplifier whose output is
coupled regeneratively to its input. Types differ mainly in methods
of coupling. The type of coupling use influences mode of opera-
tion. The outstanding feature of multivibrator operation is that
electrical operation can be rapidly changed from one to the other
of two states. These two states are: (a) conduction in tube 1 and
cutoff in tube 2, and (b) conduction in tube 2 and cutoff in tube 1.
Stable state. Electrical state in which a multivibrator remains
indefinitely, unless the circuit is acted upon from an external
source (for example, by a trigger voltage) .
A-stable Multivibrator. A multivibrator in which neither of
the two tubes reaches a stable state during operation, hence a cir-
cuit capable of free-running oscillation caused by self-excitation.
The circuit continually alternates between one state and the other
at a definite rate, determined by circuit constants. In the past, this
circuit has been referred to merely as a "multivibrator."
Bi-stable Multivibrator. A multivibrator in which either tube
becomes stable in either one of its states. The circuit locks in
either condition and remains that way until reversed by the ap-
plication of a suitable external excitation pulse.
Mono-stable Multi-vibrator. A multivibrator in which only one
of the two states can become stable. The circuit returns itself to
its normal stable state after it has been temporarily shifted to the
other state by application of an external excitation pulse.
Some other names, not as specific or authoritative, but some-
times applied to the different multivibrator circuits are as follows:
 BASIC PRINCIPLES 3

A-Stable M ultivibrator. Free-running multivibrator.

Bi-Stable Multivibrator. Binary, binary counter, locking cir-
cuit, frequency divider, repeater, regenerator, flopover circuit,
Eccles-Jordan trigger circuit.
Mono-Stable Multivibrator. One-shot multivibrator, flip-flop.

3. Pulse Formation

If the reader clearly understands how pulses are formed in

vacuum tube circuits, a study of multivibrators is easier. The fol-
lowing points constitute a brief review of some of the primary ac-
tions in simple pulse circuits.
a. A positive-going pulse is one that "rises" toward its maxi-
mum (most positive) value in a positive direction, regardless of
the polarity of the absolute value of the voltage during the pulse.
For example, pulses at both (A) and (B) of Fig. 1 are positive-
going, even though the absolute value of the voltage at (B) never
becomes positive.
b. A negative-going pulse is one that "rises" to its maximum
(most negative) value in a negative direction, regardless of the
polarity of the absolute value of the voltage during the pulse. For
example, pulses at (A) and (B) of Fig. 2 are both negative-going,
even though at (A) the absolute value of the voltage never becomes
negative.
c. A sudden rise in plate current in a vacuum tube yields a
negative-going voltage pulse at the plate (between plate and cath-
ode). Similarly, a sudden drop in plate current results in a positive-
going plate voltage pulse. 2
d. A negative-going voltage pulse applied to the grid of a tube
yields a positive-going voltage pulse at the plate; a positive-going
grid voltage pulse produces a negative-going plate voltage pulse.
e. Coupling capacitors do not change the direction of rise of
the coupled pulse; the same applies to coupling resistors.
f. A negative-going voltage pulse applied to the cathode of a
tube causes a negative-going voltage pulse to appear at the plate;
a positive-going pulse on the cathode yields a positive-going plate
pulse.

2 The plate voltage is equal to the value of the B-supply minus the drop
in the plate resistor. As plate current increases, the voltage drop in the plate
resistor increases, and thus plate voltage decreases.
4 MULTIVIBRATORS

+ -1LJLJL +

01----------- 01-----------

-1LILJL
(A) (B)

Fig. 1. Examples of "positive-going" pulses.

Multivibrator circuits have two important properties that lead

to many valuable applications. The first is the abrupt waveform
generated; i.e., somewhere in the circuit there is a very rapid change
of voltage with time, say of the order of 100 volts per microsecond.
This allows the circuits to be used where high resolutions in time
are to be effected, as in accurate time measurements, in time mod-
ulation circuits, in pulse counting apparatus, and in frequency
division. The second property of the multivibrator is its ability
to generate square waves. Not only is the rise and flyback time
very short, but the circuit remains in one state for a relatively long
time so that the output is substantially constant over the interval
between transitions. This output can be used to turn another cir-
cuit on or off, a process called "gating." There are many applica-
tions of "gates" of this nature. Such a gate can be used to turn on
or off a tube for a definite time sequence relative to the rest of the
circuit. For example, a sweep can be started, a pulse appearing in
a certain time interval can be counted or rejected, a cathode ray
tube can be blanked, or a receiver can be made inoperative. The
time over which this constancy of output appears may be measured
in seconds, milliseconds, or microseconds, so flexible is the circuit.
In some multivibrators, the width (length of flat top of square
wave) is governed by the voltage at some point in the circuit or by
the value of some circuit parameter. This suggests a host of ap-
plications, such as time modulation, coding, and others involving
definite time sequence.

4. Multivibrator Action

The voltage on a capacitor cannot change instantaneously. If

there is to be an abrupt change in voltage between any two points
 BASIC PRINCIPLES 5

+ 7..nf7..t +

0,-...--------- 0,-...---------

(A) (B)

Fig. 2. Examples of "negative-going" pulses.

in a circuit, the capacitive loading between these points must be

as low as possible. For a given capacitance, the change in voltage
in a given time is proportional to the charge delivered in this time
and hence to the current. Thus large charging and discharging
currents are prime requisites of abrupt waveforms. These sudden
changes are initiated by turning a tube off or on. Doing this calls
for tubes that have a very steep transfer characteristic; that is, the
change in plate current for a given change in grid voltage is large
(a high value of transconductance) .
Throughout many of the discussions to follow, the terms "on"
and "off" will be used to describe the state of a tube. Caution
must be exercised at this time to be consistent about the meanings
of these words as used in the text. When a tube is described as
"on" or "full-on," this means that the tube is conducting with its
cathode-grid voltage essentially at zero or at such a value that the
plate circuit is fully conductive. \\Then a tube is described as "off,"
it is to be taken as being at the cutoff point with no plate current
flowing at all.

5. Review Questions

(I) Give six electronic applications of multivibra1.or circuits.

(2) Define: multivibrator, a-stable multivibrator, bi-stable multivibrator,
and mono-stable multivibrator.
(3) What is meant by a "stable state" in a multivibrator?
(4) List four salient points in pul,e formations.
(5) What is meant by a "negative-going pulse"?
(6) What is gating?
(7) Why is a high-transconductance tube desirable for steep waveform?
(8) What is meant by the terms "on" or "off' with respect to a tube operat-
ing as a multivibrator?
 Chapter 2

THE Bl-STABLE MULTIVIBRATOR

6. Fundamental Circuit

The basic bi-stable multivibrator circuit is illustrated in Fig.

3. As is evident from an inspection of the circuit diagram, each
tube is an amplifier, the plate of which is directly coupled through
a resistor to the grid of the other. The load resistances (RL) are
much smaller in value than the coupling resistances (Rl and R2).
Typical values are 15k ohms for the load resistors and I megohm
apiece for the coupling resistors.
From the symmetry of the circuit one might suspect that the
quiescent current in each tube would be the same. This is not the
case, however, as will be shown in the following text; in addition,
the starting state is an unstable condition, which almost imme-
diately causes one tube to go off and the other to go on. Suppose
there is a momentary increase of current IP 1 in tube l. This causes
a negative voltage pulse to appear at the plate of V 1. This negative
pulse is immediately transferred to the grid of V2 through Rl.
(See Par. 3.) As this grid goes more negative than previously as a
result of this pulse, plate current IP 1 of V2 decreases, causing a
positive voltage pulse to appear at its plate. Since this plate couples
to the grid of VJ through R2, a similar positive pulse instantan-
eously appears at this grid, further increasing plate current I pl·
Since it was just this effect that started the action, and since the

6
 THE Bl-STABLE MULTIVIBRATOR 7

effect is being intensified by repeated pulse transfer from tube to

tube, it consitutes a runaway situation, which continues until either
one of two things happens:
a. V2 may reach cutoff. In this case Eps becomes equal to the
supply B+ voltage. (The voltage drop that occurs either in RL,
as a result of the current flow from B- to B+ through RL, R2, RJ,
or through RL, RI, R4 is neglected in this discussion).
b. VJ may go full-on. Depending upon the supply voltages and
the circuit parameters, one of these conditions will be attained. be-
fore the other; this will be the stable state, in which the circuit will
remain. For example, if V2 is cut off, any momentary increase in
lp 1 will only drive the grid of V2 further negative, causing no
change in lps or Epi• Similarly, if lp 1 should decrease slightly, its

B+ 150V

Fig. 3. Basic bi-stable multi•

vibrator circuit. PULSE,_____\
INPUT...,-,

effect would be to make £ 02 somewhat less negative, but if the tube

is sufficiently beyond cutoff, this will have no effect. An analogous
argument may be used to show that, if one tube is full on, the
circuit is also stable.

7. Function of Bias Vohage

The statement in Par. 5 that a stable state cannot exist with

both tubes carrying the same quiescent current must be qualified
somewhat in this way: there are ranges of bias voltages for which
both tubes may be full on or both tubes may be cut off. Except
for these extreme conditions, however, the original statement still
holds.
8 MULTIVIBRATORS

Taking specific component values for, say, a 6SN7 twin-triode:

RL is 20k ohms, RI and R2 are 1 megohm, R3 and R4 are I meg-
ohm, and a supply voltage of 250 volts, it may be shown by analysis
of tube behavior, as bias voltage is changed from zero to 280 volts
or more, that there are actually five stable states (as tabulated in
Table I).

AVERAGE PLATE CHARACTERISTICS

16
TYPE 6SN7
E •6.3 VOLTS

:312--+---+---++---+t-++---,r+--t---+---+--+---l--+-+-----l
a:
~
2
C

~8
2

80 160 240 320 400 480

PLATE VOLTS

Fig. 4. Plate characteristics for the 6SN7GT, showing the load line for the condi-
tions in the circuit of Fig. 3.

How the five different stable states can exist may be better
understood by remembering the following basic facts:
a. "Full-on" (as previously defined) is the state in which the
net voltage on the grid is zero.
b. Net voltage on a grid is equal to the algebraic sum of the
voltage coupled from the other tube plate and the bias voltage.
c. A tube remains "part-on" if its net grid voltage is more
positive than the cutoff value, but more negative than zero.
d. A tube is "cut off" if it has enough negative grid voltage
(net) to reduce the plate current to zero.
For illustration of how the values and conditions of Table 1
are obtained, consider the Ev - IP curves for the 6SN7 in Fig. 4. The
plate load for each tube is 20k ohms, so a load line is drawn repre-
senting this value. It is drawn between the 250-volt plate voltage
on the horizontal axis and 12.5 ma on the vertical axis. The cur-
rent value is that resulting from application of the 250-volt plate
 THE Bl-STABLE MULTIVIBRATOR 9

voltage across the 20k ohm plate load. All possible conditions for
each tube in this circuit fall along the load line thus drawn. Cutoff
grid voltage is about -30 volts (cutoff is gradual). The following
discussion of the five possible states can be checked against th·e
graph of Fig. 4. Let us examine the conditions of Table l.
a. Bias is between zero and 42 volts. Both tubes have a min-
imum plate voltage of about 84 volts, due to the drop across plate
resistance of each during greatest conduction. This value is de-
rived by locating the point along the load line (Fig. 4) for zero
grid volts (Ee = zero). Even when the coupled voltage from the
other tube tends to be positive, the grid voltage stays close to zero,
due to the flow of grid current through RJ and R4. Thus, until
considerable bias is applied, neither tube can affect the other; they
are both so far in conduction that even moderate changes of volt-
ages cannot change this stable state.
b. When the grid is no longer positive (and thus does not
shunt the grid resistor to low impedance) the voltage at a grid is
about one-half the voltage of the plate to which it is coupled. This
is because of the voltage-dividing action of coupling and grid re-
sistors R2, RJ and RJ, R4. When the external bias Ee is raised to
about 42 volts (half the plate voltage of 84 volts) one of the tubes,
being slightly different from the other, starts to have its current
limited appreciably before the other tube. Let us say V2 plate cur-
rent starts to become less. V2 plate voltage (and thus VJ grid volt-
age) becomes more positive. V1, still in the "full-on" condition,
is not affected appreciably, maintaining about the same coupled
voltage on the grid of V2. Thus no "runaway" condition occurs, VJ
already being saturated. As bias is increased, current gradually de-
creases in V2, and VJ is full on while V2 is part on.
c. When bias Ee has been raised to 52 volts, it exceeds the pos-
itive voltage fed from V1 plate to V2 grid by enough to cut V2 off.
Further increases (up to 125 volts) keep V2 cut off and VJ full-on.
d. When the bias reaches 125 volts, it equals the portion of
B+ voltage fed to VJ grid from V2 plate during cutoff of V2.
Then the net V1 grid voltage starts to become negative and the
plate current of VJ is reduced. With such a high bias, V2 cannot
be pulled out of cutoff and further increase of bias simply reduces
VJ plate current through the part-on condition.
e. When the bias reaches 140 volts, it exceeds the full B+
voltage coupled from VJ to V2 (125 volts) by enough to cut off VJ.
Thus in this range (over 140 volts) both tubes remain cut off.
10 MULTIVIBRATORS

TABLE 1

Bias VJ V2
zero to 42 volts full-on full-on
42 to 52 volts full-on part-on
52 to 125 volts full-on cut off
125 to 140 volts part-on cut off
more than 140 volts cut off cut off

Of course, since the circuit is symmetrical, the two right-hand

columns may be interchanged to apply to either of the two tubes.
It is further evident from the table that careful consideration must
be given to the choice of the bias voltage on the basis of tube type
and circuit values if the circuit is to be truly bi-stable; i.e., one tube
full-on and the other cut off.

8. Trigger Action of the Bi-Stable Multivibrator

In considering the stability of the circuit in Par. 6, it was

pointed out that small spurious changes in current were ignored by
the multivibrator. However, if a large positive pulse is intentionally
injected into the grid of the cutoff tube, instantaneous plate cur-
rent flows in this tube and a new runaway process begins, in which
the conditions of the two tubes interchange. This is known as a
transition and is explained in exactly the same way as was used to
show why a "part-on, part-on" condition is not possible. Referring
again to Fig. 3, assume that V1 is cut off and V2 is full-on. A posi-
tive pulse is now applied to the grid of V 1, driving it into conduc-
tion. As its plate current rises, a negative pulse is transferred from
its plate to the grid of V2, causing the plate current of V2 to drop;
again, as a result of this decrease of Ipi, a positive pulse is transferred
from the plate of V2 to the grid of V 1, causing it to rise further
above cutoff. As in the previous case, the process is cumulative and
results in a transition.
The action just described acounts for the name "trigger cir-
cuit." In most applications it is desirable to have a large change in
voltage at the plate of one tube upon the injection of a pulse.
Hence the circuit is usually designed so that, in the quiescent state,
one tube is full-on and the other is cut off, as explained previously.
 THE Bl-STABLE MULTIVIBRATOR 11

The term bi-stable refers to this state and to its companion with
the conditions of the tubes interchanged.

9. Self-Biased Bi-Stable Multivibrator

The bias supply of Fig. 3 may be eliminated by using self-bias

(Fig. 5) . The cathodes are tied together and returned to ground
through a common resistor, Rk. If the two halves of the circuit were
absolutely identical, the voltage drop across Rk would remain con-
stant and there would be no need for a bypass capacitor. Assume,
however, that the on current of VJ is greater than the on current of
V2 because of different tube characteristics or differences in the re-
sistors. Then, as the transition occurs, the rise of current in one tube
is greater or smaller in rate than that of the other, causing degener-
ative action, which tends to slow down the transition. It is advisable

Fig. 5. Bi-stable multivlbra-

tor, circuit in which the ex-
ternal bias supply is elimi-
nated by use of cathode bias
resistor Rk"

to bypass Rk with a capacitor that will give a time constant consider-

ably higher than the transition time. Experience indicates that if
the product of RkCk equals 100 microseconds, it is ample; thus, if
Rk is l0k ohms (a normal value), Ck may be 0.01 µ.f.
The operation of the self-biased circuit is identical to that of
the separately (fixed bias) biased arrangement of Fig. 3. The self-
biased circuit does have two distinct advantages, however: (a) no
separate bias battery or supply is required, and (b) the quiescent
12 MULTIVIBRATORS

conditions are less dependent upon individual tube characteristics,

due to the "automatic" biasing action of Rk.
A design procedure for a self-biased, bi-stable multivibrator
will be given later in this chapter.

10. Simultaneous Application of Pulses to Both Grids

In many applications, a succession of triggers is applied simul-

taneously to both grids. This is done by connecting the grids to-
gether by means of small capacitors (25 to 100 p.µf.) and applying
the trigger pulses to the junction point of the capacitors (Fig. 6).
A positive trigger has no effect upon the on tube but will start the
off tube drawing current to begin the runaway action and resultant
transition. Assuming that there is an incoming train of positive
pulses, the next one trips the new off tube to restore the original
conditions. Thus two pulses are required to complete one multi-
vibrator cycle, each of the pulses having been "stretched" into a
constant voltage extending to the next pulse. A circuit of this kind
is well suited for use as an electronic switch, for example, to display
two waveforms simultaneously on an oscilloscope. The multivi-
brator output can be made an excellent square wave having a re-
lationship to the pulses, as shown in Fig. 7 and discussed in Par. 12.

11. Input and Output Considerations

Square-wave Output. The waveform at the plate of the multi-

vibrator is normally a good square wave, but it must be coupled
to its load through an RC network. Other types of coupling systems
(transformer or impedance coupling) are not practicable because
of their poor high frequency response and their tendency to load
the output circuit. RC coupling is thoroughly feasible, provided
the time constant of the coupling network is chosen to be sufficiently
high. If the time constant is too low, the coupling system tends to
differentiate the square wave. Normally, a time constant of ten
times the width of the square wave is considered long enough to
produce an output waveform essentially the same as that which ex-
ists on the multivibrator plate. This is the type of waveform
that is ideal for electronic switching.
Pulse Output. In binary counting or two-to-one scaling systems
(Par. 12 and Fig. 7) , it is desired that two input pulses produce one
output pulse of essentially the same waveform as the input; each of
 THE ~-STABLE MULTIVIBRATOR 13

these output pulses may then be used as the input to a succeeding

scaler, which again reduces the counting rate by a factor of two.
Such a cascaded "binary counter" is very useful in counting neural
impulses of high repetition rate, cosmic rays, radioactive particles,
etc. when the impulses arrive too fast to be counted on a slow me-
chanical register. The production of such output pulses is quite
easy. The normal square wave on the multivibrator plate is dif-
ferentiated by an RC network having a time constant that is small

RI R2

Fig. 6. How trigger pulses

may be applied to both
tubes at the same time,
through capacitors CJ and
C2.

compared to the time between pulses. "Peaking" takes place, and a

series of positive and negative pulses result, as shown in Fig. 7.
This means that for every two positive input triggers there is one
positive output pulse. If n scale-of-two circuits are used in cascade,
a frequency division of 2" is obtained. For example, six such units
provide a scale of 64, one output pulse being obtained for every
64 input pulses.
Plus and Minus Triggers. Negative triggers function as satis-
factorily as positive ones. In the case of a negative trigger, the on
tube current is decreased, and this initiates the action. As a matter
of fact, positive and I'legative triggers may be present in any mixture
-regular or random-and each pulse will trip the circuit regardless
of polarity. Nor do the pulses have to be of the same amplitude, as
long as the smallest of them have sufficient intensity to initiate the
transition. When positive pulses are used, excessive amplitude must
be avoided because a pulse that is too large has a tendency to charge
the input capacitors, and (once the pulse has passed) there may be
14 MULTIVIBRATORS

an effective negative pulse as the capacitor discharges. If the input

pulse width is narrow, the positive and negative triggers follow each
other so rapidly that they tend to cancel without activiating the
multivibrator at all; if the pulse width is great, both the true and
the spurious trigger will start transitions, giving an incorrect count.
Shunt Capacitance. Emphasis has been placed on the abrupt
voltage changes that take place in a multivibrator; this is what
makes high speed counting possible, since the resultant waveforms
are steep sided and lend themselves to high repetition rates. A ca-
+150V

A f7.......f7.__T

eYr½-
I I
I

C
ri--Llrt
I I
:

DH--Y+-++
I I I I I I
2 3 4 5 6

Fig. 7. Binary c:ounter type of multivibrator, with trigger pulses applied at the
cathode.

pacitor shunting such a voltage tends to prevent the abrupt change

and, therefore, slows down the circuit reaction. The obvious con-
clusion is, therefore, that such capacitances must be held down to
the lowest possible values. The input capacitors (Fig. 6) must be
made very small so that the input time constant may be of the
order of 15 microseconds. This in turn demands that the trigger
pulse width be even smaller than this; otherwise, input differentia-
tion may occur and cause erratic action as explained previously.
Improved performance is obtained if the input trigger is not square
but has an exponentially sloping lagging edge; such a waveform re-
duces the tendency toward differentiated overshoot and stabilizes
the tripping action. Since this is just the waveform obtained from
a multivibrator through a short time-constant RC coupling net-
work, these circuits may then be used in cascade for scaling applica-
tions with no intervening pulse shaping networks.
 THE Bl-STABLE MULTIVIBRATOR 15

For the same reason that it is desirable to have a gradually

sloping lagging edge, it is necessary for the trigger pulse to have
a very rapid rise time at its leading edge. If the rise time were long,
the short time-constant input circuit would attenuate the pulse to
the point where it could not trip the circuit at all. Experience in-
dicates that reliable triggering may be obtained by using an input
pulse having these characteristics (Fig. 8) :
Rise time: 0.l microseconds
Decay time: 1.0 microseconds, exponential
Shape: peaked
Amplitude: 20 to 50 volts
Commutating Capacitors. For fast, reliable operation, the cir-
cuit of Fig. 5 and Fig. 6 requires one minor modification: the addi-

,_ _ _ _ AMPLITUDE
I 20-50V0LTS
I
II
I I
I I
Fig. 8. Characteristics of
pulse suitable for reliable
II
triggering of multivibrator.
TIME
1.0 )JS
I ~ECAY TIME _I
I ------oif EFFECTIVT---T
I I PULSE WIDTH I
__J I.RISE TIME
1 I 0.1 µs

tion of capacitors across the coupling resistors, RI and R2. Without

these, a very fast input trigger of the order of l or 2 microseconds
can come and go without activating the circuit. The reason for this
is as follows: suppose that an ideal negative pulse shuts off V2 in
zero time. The voltage at the plate of V2 rises to B+ value instan-
taneously, unless there is heavy capacitive loading; it may be safely
assumed that the loading is light enough to allow this instantaneous
action to occur. The voltage at the grid of V1, however, cannot
follow this sudden change because of the fact that the input ca-
pacitance of the tube as an amplifier may be as high as 50 µ.µ.f and
because the input trigger capacitor, Cl (Fig. 6), is also essentially
in shunt with the grid circuit of V 1. A total capacitance of ap-
16 MULTIVIBRATORS

proximately l 00 p,p,f combined with about I megohm of resistance

(R2) yields a time constant of 100 microseconds, certainly too long
to respond to a I-microsecond trigger. This difficulty is overcome
by shunting RI and R2 by so called commutating capacitors of
about 50 p,p,f apiece, as illustrated in Fig. 7. This balances out the
effect of the shunt grid capacitances and permits the grid voltage to
follow the plate voltage changes instantaneously.

12. Practical Binary Scale-of-two Counter

The circuit shown in Fig. 7 and variations of it are utilized in

such equipment as time-division multiplex communications devices,
locking tone keyers, and frequency shift keying units, in addition
to being found in computers, radiation counters, and frequency
measuring apparatus. It differs from the circuit just discussed in
that the input triggers are introduced at the cathode rather that at
a grid or plate, and is included here to show how this variant meth-
od of excitation operates.
Assume that V 1 is full-on and V2 is cut off. A square wave
(A) is now applied lo the input terminal and is differentiated by
C3R5 to form pulses (B). First consider the case of a positive pulse.
The voltage drop already existing across R5 due to the cathode
current of tube VJ is momentarily increased, the cathodes of both
tubes being driven more positive. Tube V2, already cutoff (cathode
positive and grid relatively negative) is not affected by this in-
crease in the positive potential of its cathode. Tube VJ, however,
had its grid at a slightly higher positive potential than its cathode.
(The amount is limited by the grid current flowing through R3.)
The positive pulse excitation applied across R5, effectively inserted
into the grid circuit to make the grid relatively negative, is ampli-
fied in the conducting tube, V1. This reduces the plate current and
produces a positive pulse at the plate of VJ that is then transferred
through CJ and RI to the grid of V2. If it is of sufficient magni-
tude, it starts the runaway action, with J/2 becoming conductive
and VJ reaching cutoff. The status is thus reversed.
Upon the arrival of the next negative pulse in the pulse train,
both cathodes are driven more negative, tending to increase both
plate currents. V2, already full-on, is affected very little, develop-
ing a very slight increase in its plate current, which results in a
negative pulse at its plate; small as this may be. it is transferred to
 THE Bl-STABLE MULTIVIBRATOR 17

the grid of the cutoff tube, VI, and helps to hold VI cut off. Thus
the negative pulse (if not too large) does not trigger a transition,
and V2 remains conductive until the next positive pulse arrives.
See Fig. 7 (C) , points 1 through 3.
Since the circuit is symmetrical and the input excitation is ap-
plied to both cathodes, the foregoing discussion holds true for any
initial tube condition (either VI or V2 conductive). As the wave-
forms in Fig..7 show, two input square wave cycles are required to
produce one output cycle; similarly, if the output is differentiated
by an RC network connected to the plate of V2, one output pulse
is obtained for every two input pulses (D) .
If it is preferred, excitation may be applied to the grids through
capacitors as in Fig. 6. In this event, transitions are brought about
by negative rather than positive pulses, but the same scale-of-two
action is obtained.

13. The Bi-stable Multivibrator as a Regenerator

In reading the preceding paragraph, particularly that portion

dealing with the ineffectiveness of negative pulses as triggers, it
may have occurred to the reader 'that the cutoff tube, even though
"held" at cutoff by the small pulse fed back from the conducting
tube, tends to go into conduction when the cathode is driven neg-
ative. Whether it does or does not depends largely on the ampl-
itude of the pulse. If it is assumed that the conducting tube circuit
is such that the tube is really full-on (i.e., drawing near-maximum
plate current) , a negative pulse applied to its cathode would pro-
duce an almost negligible holding pulse to the cutoff tube, since
its plate current cannot rise significantly, even with excitation of
the right sign.
On the other hand, if the amplitude of the pulse is two or
three times greater than that required for reliable scale-of-two trig-
gering, there will come a point at which a negative pulse applied
to the cathodes will also cause a transition to take place by driving
the cutoff tube into conduction to start the runaway process once
again. Thus, the system acts not as a frequency divider but as a
repeater or regenerator, putting out square-wave voltage or pulses
at the same frequency as that of the input excitation. The compar-
ative waveforms for this result are given in Fig. 9.
18 MULTIVIBRATORS

14. The Bi-stable Multivibrator as a Locking Circuit

In this application, the multivibrator functions in exactly the

same manner as the regenerator, except that the pulses are delivered
at a slower rate, so that the circuit may stay locked-in in either of its
two stable states until the arrival of the next pulse. The arrange-
ment provides the electronic equivalent of a pulse-energized elec-
tromagnetic relay locking in whichever position it is "kicked to."

17
I l__J
17L_JI INPUT
WAVEFORM

INPUT PULSES
Fig. 9. Waveforms for re-
AFTER
DIFFERENTIATION generator, or repeater, type
of multivibrator.

REPEATER
OUTPUT
WAVEFORM

As a rule alternate positive and negative pulses of voltage are

applied to one grid instead of to the two cathodes in parallel, as
in the case just discussed. The cathode is generally connected di-
rectly to ground instead of through a resistor, bias being obtained
from a battery or other similar source.

15. Diode Trigger Feed

The basic methods of triggering a bi-stable multivibrator have

been described in the preceding paragraphs. Mention also has been
made of the fact that the multivibrator tube should not draw grid
current while being tripped, and that the input circuit should not
load down the grids. Flow of grid current may always be avoided,
regardless of input pulse amplitude, by using negative triggers
exclusively.
The grid loading problem and much of the trigger waveshape
difficulty may be overcome by feeding the pulse to the multivibrator
through diodes, as illustrated in Fig. 10. The diodes may be dual
tubes, such as the 6H6, 6AL5, or germanium (or silicon) crystal
rectifiers.
 THE Bl-STABLE MULTIVIBRATOR 19

In the arrangement of Fig. 10, the input pulse must be neg-

ative, but positive pulses may also be used if the diodes are re-
versed. The diodes function as follows: assume that VJ is full-on
and V2 is cut off. Under these conditions, the plate of diode VJ
is negative with respect to its cathode, since the voltage on this tube
is that which appears across R1, 1 • Hence VJ is not conducting. Di-
ode V4 is in the non-conducting state too, because its plate voltage
comes from the drop across Ru, and (since V2 is cut off) there is
zero voltage across this resistor; with no difference of potential be-
tween cathode and plate, V4 must also be quiescent. These condi-
tions represent a stable condition until the arrival of a negative
trigger pulse.

R5

<>-)....__
Cc +
__
TRIGGER
INPUT

Fig. 10. Circuit employing diodes to minimize trigger-source loading,

With neither diode conducting, the multivibrator is essentially

disconnected from the trigger source and is thus loaded down only
by the diode interelectrode capacitances, which are very small and
may be ignored. A negative trigger pulse is now fed through the
coupling capacitor, Cc. Unless it is very great in amplitude, it has
no effect upon VJ, because this tube's plate is so negative with re-
spect to its cathode that only an unusually large input pulse could
reverse these conditions. The instantaneous negative voltage on
V4, however, causes this tube to conduct (the path is through Ru
and R6) so that a relatively large negative trigger appears at the
20 MULTIVIBRATORS

-
18

16

14

C 12
~
.!I- IO

160 200 240 280 320 360 400 440

f Ep (VOLTS)
CUT-OFF

Fig. 11. Bi-stable multivibre>tor circuit and 6SN7 characteristic curves used in
simplified design procedure in text.

plate of V2. This is coupled to the grid of V 1 through R2C2,

starting the runaway cycle and resulting in a reversal of status, with
VJ cut off and V2 conducting.
Much less attention need be paid to the trigger shape when
diode coupling is used than in the simpler circuits. Although the
input pulse must still have a very short rise time, the shape of the
 THE Bl-STABLE MULTIVIBRATOR 21

pulse and the characteristics of its trailing edge are comparatively

unimportant. Even if the pulse has a positive overshoot, it cannot
get through the diodes and cannot cause spurious triggering. The
time constant of R5Cc and R6C" is not critical, although it should
be short compared to the interval between pulses. At the instant of
diode conduction, the time constant changes from R6Cc (assuming
V4 is already cut oft) to the product of Cc and the joint resistance
represented by R6 in parallel with the series combination of Ru
and the internal diode resistance. If this is small compared with
the pulse interval, there will be differentiation of the trigger, but
this can do no harm since the diodes are impervious to positive
overshoots. Capacitor C 0 will charge, however, and must discharge
before the next pulse arrives; otherwise, the pulse will not reach
the diodes. This accounts for the need for making R6C 0 (and R5C,,,
of course) smaller than the pulse interval.

16. Simplified Design Procedure for Bi-Stable Multivibrator

A rigorous procedure for determining the circuit parameters

of a specific bi-stable multivi.brator is rather tedious and time con-
suming. For the average worker in electronics, the simplified se-
quence given below more than suffices, even though certain funda-
mental assumptions and compromises with rigidity must be made.
Given: A 6SN7 is to be set up as a bi-stable multivibrator and
is to be used with a power supply that provides 300 volts (Fig. 11) .
Assumptions:
a. With a power supply voltage of 300 volts available, we shall
want the cathodes quite positive with respect to ground, so that we
shall have plenty of "play" between these points to establish op-
erating bias. Assume the cathode-to ground voltage desired to be
0

about JOO volts.

b. Assume a reasonable value of load resistor RI. The triode
sections of a 6SN7 have the same characteristics as the 6J5. A suit-
able load resistor might be about 25,000 ohms. This will later be
tested and, should it prove too high or too low, the process may be
repeated with a different assumed value.
c. During the quiescent condition, one triode will be cut off
and the other will be full-on. To insure complete cutoff, the re-
quired bias voltage may be assumed to be about twice the actual
cutoff bias value. For the full-on condition, the control grid will
have to be zero or slightly positive for this triode section.
22 MULTl VI BRATORS

d. The individual values of R2 and R3 are considerably higher

(IO to 20 times) than that of resistor RI, so that the voltage drop
across RI due only to voltage divider ourrent is negligible.
Procedure:
a. The load resistor, RI, is chosen as 25,000 ohms (assumption
b). A load line for this resistance is to be drawn in the 6SN7 plate
characteristic curves. Since the cathodes are to be at 100 volts pos-
itive (assumption a), the load line is to be constructed for a plate-
cathode potential of 200 volts. This load line is shown in Fig. 11.
b. The conducting tube has a control grid voltage that is close
to zero (assumption c). Find the plate current that flows in the
conducting tube at this bias voltage from the intersection of the
load line in the Eu = zero curve. This current is 5.5 ma. The
figure also shows that at this plate current the voltage across the
tube is 60 volts and the drop across RI is 140 volts.
c. 5.5 ma must also flow through the cathode resistor, because
only one tube is to conduct at a time when quiescent. To obtain
a cathode potential of 100 volts, then, the cathode resistor must be:
100
R4 = - _--X-- -_3- = 18,000 ohms
5 5 10
d. According to the average plate characteristics (Fig. 11) ,
the tube, with 200 volts plate-to-cathode potential, cuts off at -12
volts of bias. To insure against spurious triggering, assume the grid
voltage needed is -25 volts (assumption c).
e. Assume VJ to be non-conducting at the moment. Its cathode
potential is 100 volts positive and its grid potential is -25 volts
with respect to cathode. Therefore, the voltage coupled from V2
(point F) is + 75 volts, and the voltage drop across R1 must be 75
volts. But the potential at point F is 160 volts with respect to
ground (140-volt drop across RI subtracted from supply voltage of
300 volts= 160 volts at point F). Thus there must be an IR drop
of 85 volts from F to C and a 75-volt drop from C to ground. This
determines the ratio of R2 to R3 as 85/75.
f. These resistors are to be high in value (assumption d). If
R2 is arbitrarily selected as 500k ohms, R3 must be 440k ohms to
agree with the ratio derived in ( e).
g. These resistors, chosen to place the grid of the non-
conducting tube (point C) 75 volts above ground must now be
tested to determine whether or not they conform with the second
part of assumption c; i.e., that point G is at zero potential or
 THE Bl-STABLE MULTIVIBRATOR 23

slightly positive. With VJ non-conducting, and its grid at -25

volts with respect to cathode, its plate voltage with respect to
ground (point B) will be very nearly 300 volts (assumption d) .
Using resistors of 500k ohms and 440k ohms, point G would, if
isolated, have a potential of + l 35 volts with respect to ground, or
35 volts more positive than the cathode. When point G is con-
nected to the grid of V2, grid current flows and prevents the grid
from becoming more than a fraction of a volt more positive than
the cathode. Thus the initial assumption is justified, and the re-
sistor choices, including the selection of RI, are practicable.

17. Review Questions

(l) Draw the circuit of the bi-stable multivibrator, utilizing a bias supply.
Assign typical values to the components and explain the circuit operation.
(2) What role does the bias voltage play in a bi-stable multivibrator?
(3) What is a "trigger circuit"? Explain "trigger action" in a bi-stable
multivibrator.
(4) Define transition; refer to Fig. 3 and explain why this effect prevents a
"part-on, part-on" condition.
(5) How may the circuit of question I be adapted for self-biased operation?
Draw and explain the resultant changes in circuit operation.
(6) Describe the modifications necessary when a succession of triggers are
applied simultaneously to both grids in the modified circuit of question 5.
(7) What type of coupli_ng is normally utilized for square wave output?
Why are other types of coupling normally not applied?
(8) Explain the operation of plus and minus triggers in binary counters or
two-to-one scaling systems. Draw the circuit of a practical binary scale-of-two
counter.
(9) Explain the function of shunt capacitance and commutating capacitors
in multivibrator operation.
(10) Draw comparative waveforms illustrating the action of the bi-stable
multivibrator as a regenerator.
 Chapter 3

MONO-STABLE MULTIVIBRATOR

18. Conversion of Bi-Stable to Mono-Stable Multivibrator

Many multiplex telegraph transmitters and radar systems uti-

lize the mono-stable form of multivibrator for timing and "delay"
purposes. The mono-stable type (Par. 2) is distinguished from the
bi-stable variety by its self-restoring action. In the bi-stable multi-
vibrator, a correct trigger produces a stable transition from one
tube to the other; the mono-stable multivibrator may go through
the same transition, but returns to its original condition spontan-
eously. This action is often called "flip-flop" in contrast with the
bi-stable behavior which is referred to as a "flopover."
If the resistive coupling from one of the plates to the opposite
grid in the bi-stable multivibrator is changed to capacitive coupling,
a mono-stable system results. In the basic circuit, the corresponding
grid resistor is returned to a positive potential, rather than to
ground or to a bias source (Fig. 12, point A).

19. Circuit Operation

In the quiescent condition, V2 is full-on. Its control grid volt-

age is zero or very slightly positive with respect to the cathode, the

24
 MONO-STABLE MULTIVIBRATOR 25

circuit values being so chosen as to establish V 1 below cutoff. The

circuit will trigger if a positive pulse is applied to the grid of VJ
(or the plate of V2), or if a negative pulse is applied to the grid of
V2 (or the plate of VJ). In this circuit, the trigger pulse is ap-
plied to only one tube at a time, not to both tubes simultaneously
as in the bi-stable type.
Suppose that an isolated negative pulse is injected into the
grid of V2. The plate current of V2 decreases and a positive pulse
appears at the plate of V2, and is transferred through CJ to the
grid of V 1, causing the plate current of V 1 to rise. The negative
pulse at the plate of V1, which results from this rise of plate cur-
rent, is immediately passed on to the grid of V2 through coupling
capacitor C2, reinforcing the original trigger. This starts the famil-

B+

Fig. 12. Basic monostable

multivibrator circuit.

iar runaway action, with the grid voltage of V2 rapidly forced neg-
ative to a point beyond cutoff. At the same instant, V1 reaches the
full-on condition, since the transfer of voltages just described brings
the grid of VJ into the zero or slightly positive region. However,
this is only a quasi-stable state. After a fixed time interval has
elasped, the circuit, without the need for an additional trigger, re-
verts to the original condition, in which V2 is full-on and VJ is
cut off.
26 MULTIVI BRATORS

20. Explanation of Reversion Action

A thorough understanding of the reasons for the waveforms

obtained from a mono-stable multivibrator requires a careful analy-
sis of the voltage changes that occur in the circuit with the super-
imposition of the trigger pulse. (Refer to Fig. 13.)
Let t = 0 represent the instant when the trigger pulse is in-
jected, t =
0- an instant before pulse injection, and t =o+ the
time when the "flip" is completed; i.e., when V 1 has gone from
cutoff to full-on. At t = 0-, the grid potential of V2 is zero or

2
POTENTIAL E9
IA)

t=o
TIME It)
IU
<:> 0/ Fig. 13. Waveforms of grid
<I
I- CUT-OFF voltage of V2, showing re-
..J -E --------
0 lated circuit actions.
>
e
It:
<:>
-s
"'>

very close to it as a result of the grid current flow through R4

(Fig. 12). At t = o+, the voltage on the grid of V2 falls to the
point labeled -S on the axis of Fig. 13. This point is the negative
swing of the plate potential of VJ as this tube goes from cutoff to
full-on (as its plate voltage drops from a value virtually equal to
B+ to its lowermost limit at the full-on condition) . As a result of
this swing after the trigger has acted, capacitor C2 becomes charged,
but once the quasi-stable state is attained, it begins to discharge
through R4; since R4 is much larger than RI, the discharge time
constant is approximately equal to R4C2 with RI being ignored.
The equation of the discharge curve is:
Egi = A - (A+ S) l -t!R4C2
This equation is obtained from the consideration that at the be-
ginning of the discharge process Eui = -S and at t = infinity,
E02 A. =
 MONO-STABLE MULTIVIBRATOR 27

If the magnitude of the grid cutoff voltage of V2 is assumed to

be -E as shown on the curve, this tube will start to draw current
when Eu 2 = -E. At the same instant, the plate voltage of V2
begins to drop, providing the pulse that pulls the grid voltage of
VJ along with it. This is the beginning of the restoring action that
returns the system to its initial stable state.

t•O t•T
I I

r
I
TRIGGER I
I
I (A)

B+ LEVEL

Ep2 I
C2
CHARG~
l, M,. . . .
(B)

ZERO ZERO
E!J2 A

-s (C)
B+ B+
Ep1
'---ti2 CHARGING
HROUGH RI

-s- TRIGGER
(D)
ZERO

(E)

Fig. 14. Waveforms in monostable multivibrator.

The foregoing analysis assists in explaining the waveforms

shown in Fig. 14. Consider first the region near t = 0. EP, rises
very rapidly toward B+ (diagram B) with a time constant equal
to the product of R2 and the capacitance at the plate of V2, which
is always quite small. This results in a steeply rising leading edge
to the B+ level. Eu, starts at -S and increases toward voltage A
28 MULTIVIBRATORS

with a time constant of R4C2, as already explained; the trigger is,

however, superimposed at the start of the change (diagram C).
Ep 1 (diagram D) starts at B+ and drops almost instantaneously to
the lower limit of its swing -S, but again the trigger is super-
imposed on the waveform, having been transmitted through C2
from the grid of V2 to the plate of V 1. Eu 1 rises quickly from
cutoff to zero, or it may overshoot slightly into the positive region
(diagram E) .
Now consider the conditions that obtain at the end of time T,
the end of the quasi-stable state. As EP 1 rises toward B+ (diagram
D), this change of voltage appears at the grid of V2, driving it pos-
itive, and the resulting grid current charges C2 until Eui returns to
zero as shown in diagram (C) . This overshoot in Eu 2 causes a very
large negative overshoot in Eps (diagrams C and B), which drops
very quickly as a consequence of the heavy current drawn by V2.
Epi, however, rises comparatively slowly (diagram D), because the
current that charges C2 must pass through RI. In addition, the
overshoot in EP 2 reflects itself in a similar overshoot in Eu 1 (diagram
E).
Since the mono-stable multivibrator generates a rectangle for
a peaked trigger, it can be used to gate other circuits; furthermore,
since the output can be differentiated to give an output pulse that
is delayed a predetermined time T after the incidence of the trigger
pulse, it may be utilized for accurately controlled delay purposes.

21. Self.biased Mono-Stable Multivibrator

The circuit of Fig. 12 has the disadvantage that a source of
negative voltage is required. A mono-stable multivibrator in which
self-biasing is accomplished is illustrated in Fig. I 5. In addition to
this change, it should be observed that coupling resistor R3 and
its shunt capacitor (Fig. 12) are both missing; feedback takes place
through common cathode resistor R5, making other feedback com-
ponents unnecessary. Another advantage of this arrangement is
that the plate of V2 is not connected to anything except R2, i.e.,
to no significant capacitance except that of the output capacitor
(and the tube's plate-to-ground capacitance) , which may be very
small. Thus a very fast waveform is obtainable at this P-oint.
Its operation may be described as follows: in the quiescent
state, VJ is cut off, due to the high bias applied to its cathode.
 MONO-STABLE MULTIVI BRA TOR 29

This bias is the drop across R5, and results from the high plate
current flowing through V2, this tube being in the conducting
state, since its grid is effectively at cathode potential through R4.
Now a positive trigger of sufficient amplitude is applied to the
grid of VJ through coupling capacitor Cc, causing VJ to rise out
of cutoff and begin to conduct. The flow of the plate current pro-
duces a negative pulse at the plate of VJ. This pulse is transferred
to the grid of V2 through capacitor C. The plate current of V2
then decreases, the voltage drop across R5 drops, permitting VJ to
draw still more plate current, increasing the negative pulse at its

B+

f----o OUTPUT
Cc
TRIGGER o---)

Fig. 15. Self-biased monostable multivibrator.

plate, etc. The transition to the state in which VJ is full-on and V2

is cut off proceeds as before and the circuit remains in the quasi-
stable condition until C discharges sufficiently to allow the grid of
V2 to rise from the voltage beyond cutoff to the point where con-
duction may once more begin. As V2 begins to pass plate current,
the voltage drop across R5 again starts to increase, reducing the
plate current of VJ, which in turn transmits a positive pulse to the
grid of V2 to bring the circuit back to its original stable state.
In analyzing the circuit operation of this arrangement, the
question of the relative conductivities of the tubes when either is
full-on often arises. With no trigger applied, the grid of V2 is at
cathode potential so that its bias is zero; when VJ is full-on, how-
ever, its grid is always negative with respect to its cathode, since
30 MULTIVI BRATORS

the former is returned to ground; hence there is always some neg-

ative bias. Thus in the stable state the current through R5 is con-
siderably higher than in the quasi-stable condition, and it is this
variation of cathode current that accounts for the feedback.

22. Self-biased Mono-stable Multivibrator with Positive Grid Retum

For very precise timing applications, the circuit shown in Fig.

16 is preferred over that of Fig.. 15. The significant difference be-
tween the two arrangements is that R4 in the latter case is returned
to B+ rather than to the cathode; also, the grid of VJ is at a fixed
voltage, E, controllable by the setting of R3.

R4

Fig. 16. Modification of self-

V2
biased monostable multivi-
brator, for accurate timing.

Let us examine the sequence of operational events first, then

determine the advantage that the positive grid return has over the
ground grid return. In the quiescent state, V2 is full-on with its
grid-cathode voltage at zero potential or slightly positive. Voltage E
is made less positive than the cathode potential of V 1 by an amount
equal to at least the cutoff voltage of the tube, thus VJ is non-
conductive. A negative trigger injected into the grid of V2 causes
the plate current of the tube to decrease, with a consequent reduc-
tion of the drop across R5. 1£ the change is sufficient, V1 will be-
gin to conduct and start the runaway action. At the end of the
transition period, which is virtually instantaneous, the grid of V2
has been driven far below its cathode, but its potential immediately
begins to rise exponentially toward B+ with a time constant equal
to R4C. (RI is neglected since, as usual, its resistance is very much
 MONO-STABLE MULTIVIBRATOR 31

lower than R4.) As ~oon as the potential on the grid of V2 rises

above cutoff, the circuit flips back to the stable state.
The more accurate timing action of this circuit as compared
to the one using cathode return is explained as follows: referring
to the instant when the circuit has reached the quasi-stable state
with V2 cut off and VJ full-on, capacitor C in both cases begins to
discharge toward the point where V2 will come out of cutoff and
start to conduct. To cut V2 off, in the multivibrator of Fig. 16, C
must charge to a relatively high voltage, because the grid of the
tube is connected to the high positive voltage of the supply; thus,
to allow conduction to occur once again in V2, only a comparatively
small discharge need occur, taking place along the very linear por-
tion of the curve (Fig. 17 A) . Angle A indicates that the approach
to the cutoff voltage is very rapid, so that slight changes in circuit
conditions and voltages have negligible effect upon the time re-
quired for C to discharge to the conduction point. Contrast this
with the behavior of C in the circuit of Fig. 15, illustrated at (B)

/
/
/
/

Lh=--
/
Fig. 17. Comparison between
timing action in circuit of
Fig. 16(A) and in circuit of CONDUCTION
Fig. 15(8). BEGINS 1 V--~=~GINS
HERE HERE

(A) (B)

of Fig. I 7. Here C takes on a much smaller total charge at the end

of the transition period, and hence must discharge almost com-
pletely before the restoring action commences. Hence its discharge
occurs over a much greater length of the curve, approaching the
cutoff point at a relatively small angle (angle B indicating slow
approach) . In this situation, the same small variations in tube or
circuit conditions are apt to produce a serious change in the timing.
In addition to its timing accuracy, the circuit of Fig. 16 has
another marked advantage over other types. The time required
for the completion of one cycle is a linear function of voltage E,
hence this arrangement makes an excellent delay gate, whose width
is easily and linearly controlled by the setting of R3.
32 MULTIVIBRATORS

23. Frequency Division or "Counting Down"

The mono-stable multivibrator is well-suited to frequency di-

vision applications, particularly when odd divisions, difficult to
obtain with the bi-stable type, are required. Consider a regular
train of pulses spaced a time, T, apart as in Fig. 18; assume that
the output pulse width of the mono-stable multivibrator has been
adjusted to occupy a time somewhere between 2T and 3T. Refer-
ring to the positive grid return setup of Fig. 16, the first negative
pulse in the train, applied to the grid of V2, almost instantaneously
triggers the transition, so that a positive voltage appears at the
plate of V2 (point A, Fig. 18) . V2 remains cut off through the

1-----T--j
I, 12 3 4 5 6 7

I I I I
A B C D
I I I I
I I I I

Fig. 18. Waveforms illustrating frequency division (counting down) in a mono-

stable multivibrator.

interval while pulses 2 and 3 appear and, since these are negative
pulses like the first, they cannot affect the circuit, because V2 is
already cut off. After pulse 3 has passed, the charge on C decays
sufficiently to permit the circuit to return to the true stable state
quickly (between c and d) so that it is in readiness for the arrival
of pulse 4, which causes the process to repeat.
When the time constant, R4C, and the magnitude of E are
properly selected to yield this effect, the multivibrator will produce
output pulses, after differentiation, which correspond to input
pulses I, 4, 7, IO, 13, etc. This, then, is a 3: I frequency division,
as used in DuMont Image Orthicon Camera Chain TA-124-E, de-
scribed in the next paragraph. The stability of this type of fre-
quency divider depends upon the constancy of the multivibrator
period and, if it is desired to count down, say, 100: I the width
would have to lie between 99T and IOOT. If, for any reason, the
 MONO-STABLE MULTIVIBRATOR 33

width should wander outside these limits, there would be a false

count, perhaps 99: 1 or 101: 1 stable divisions as high as 25: I can be
obtained under laboratory conditions; in commercial equipment
the division is usually limited to 7: 1 or less for dependability.

24. A Commercial Three-t~e Mono-stable Mvltivibrator

The 3: 1 count multivibrator to be described is used in an

Image Orthicon Camera Chain (DuMont TA-124-E) for counting
down trigger pulses which originate in a 31.5-kc controlled block-
ing oscillator. This frequency is successively divided by a cascaded

Rl8 Cl6
----1,1---+---if--oouTPUT

-I05V

Fig. 19. Circuit of 3: 1 count-down multi-vibrator. (After DuMont)

group of four mono-stable multivibrators by counts of 3, 5, 7, and

5, the output frequency of the group being 60 cycles, the television
field-rate:
31.5 /3 10.5 kc
10.5/5 2.1 kc
2.1/7 .3 kc = 300 cps
300/5 60 cps
The circuit of the 3: 1 count multivibrator is given in Fig. 19.
V5 is conducting because a positive potential of nearly 150 volts is
applied to its control grid through RI 9 and R21, while its cathode
is at a potential of -105 volts. The grid current flowing through
34 MULTIVIBRATORS

R21 and R19 makes this grid, therefore, zero or fractionally pos-
itive. Assuming little voltage drop in the tube, this heavy conduc-
tion forces the plate of V5 (hence the grid of V7 A) to assume a
voltage close to that of the cathode of V5. For this reason, V7 A
must be at cutoff or beyond, its control grid close to -105 volts,
with cathode at ground potential. This is the stable state of the
system.
The application of a negative pulse to the suppressor of V5
produces a sharp drop in plate current, a positive pulse at its plate
and therefore at the grid of V7 A, and a rise of V7 A into conduction.
Thus a negative pulse appears at the plate of this tube and is
transferred to the control grid of V5 through C15, dropping the
plate current of V5 still further. The action ceases temporarily
when V5 is at cutoff and V7A is full-on, with capacitor CJ5 charged
to the new voltage conditions. As the charge leaks off, a point is
reached at which V5 can again begin to conduct; this, of course
starts the process in which the system returns to its stable state.
The output square wave is differentiated through CJ6 and the
grid circuit of the next stage, providing a strong negative trigger
pulse of the desired waveform for activating the succeeding counters.
Note the connection of diode V6A. In the quiescent state, the
cathode of this tube is virtually at zero potential and its plate sees
-150 volts of applied potential from the power supply; thus 150
volts of inverse potential appears between its electrodes. The reader
may recognize this as a simple diode damping system whose function
is to limit the pulse voltage applied to the control grid of V5 re-
gardless of variations that may appear at the plate of V7 A. Should
the fed-back pulse amplitude become too great, V6A conducts and
instantaneously reduces its size to its former "clamped" value.
The time constant of the system is chosen by the proper ad-
justment of R19; since this resistor governs the time required for
discharge of CJ5, this is the interval in which the circuit remains
in the quasi-stable state; it controls the frequency division ratio, in
this case 3: I.

25. Review Questions

(1) How is a bi-stable multivibrator changed to a mono-stable multi•

vibrator?
(2) Draw the circuit of a mono-stable multivibrator.
(3) Detail and explain the circuit operation of a mono-stable multivibrator.
 MONO-STABLE MULTIVIBRATOR 35

(4) Explain "reversion action" in a mono-stable. multivihrator.

(5) Diagram the waveforms obtainable from a mono-stable multivibrator.
(6) How may a mono-stable multivibrator he used to gate other circuits?
For accurately controlled delay purposes?
(7) How does the circuit of Fig. 15 act as a multivibrator?
(8) Redraw Fig. 15 so as to convert it to a self-biased mono-stable multi-
vibrator with positive grid return.
(9) Show the application of a mono-stable multivibrator to frequency di-
vision or counting down.
(10) Explain the operation, advantages, and applications of the circuit drawn
in answer to question 8.
 Chapter 4

A-STABLE MULTIVIBRATOR

26. Action of the Fundamental Circuit

The a-stable multivibrator differs from the bi-stable type in

that the plate of each tube is capacitively coupled to the grid of the
other rather than being coupled through resistors. Compare Fig.
20 with Fig. 5. This difference converts the bi-stable multivibrator
to a form in which the transition from one stable state to the other
occurs automatically and periodically, with the result that free-
running square-wave oscillations are obtained.
The circuit may be set up symmetrically with Ru = Ru,
CJ = C2, R 01= R 0 ,, and the two tubes having identical para-
meters. One might suppose that such symmetry would lead to a
stable condition in which both tubes are full-on or part-on, de-
pending upon the sizes of the grid resistors and the amount of
contact-potential bias developed. This supposition is indeed a
possibility until one considers that even if the arrangement is ab-
solutely symmetrical, instantaneous conduction in both triodes can-
not be identical at successive intervals due to inherent tube "noise,"
unequally distributed emission along the length of the cathode, and
other random effects that must inevitably contribute to an instan-
taneous difference between the two plate currents. As soon as this
difference exists, circuit action commences. In actuality, oscilla-
tions start as soon as the tubes are heated and the plate potential
is applied.

36
 A-STABLE MULTIVIBRATOR 37

Let is start by assuming both plate currents to be equal and

assuming also that a burst of electrons from the cathode of VJ in-
creases the plate current of this tube slightly; instantly, the plate
potential drops and a negative pulse appears at the grid of V2.
The resulting positive pulse at the plate of V2 is then fed back
to the grid of VJ through CJ, further inc.Teasing its plate current,
thus starting the transition, which terminates when VJ is full-on
and V2 is cut off. As in the other multivibrators discussed in pre-
vious chapters, the transition occurs almost instantaneously.
An analysis of the grid and plate waveforms is best started from
this point and, for clarity, certain reasonable figures may be as-
sumed for the circuit voltages and component values. These are
shown in Fig. 20.

B+ (200V.)

Fig. 20. Basic a-stable multi-

vibrator circuit.

8-

With V2 non-conducting, the potential at point B must be

equal to the supply voltage, since there is no drop in Ru, (i.e.,
Ep, = approximately 200 volts). At the same time, a full-on condi-
tion in tube VJ means that its grid is at or near zero potential with
respect to ground, so that CJ must be charged to about 200 volts
negative on the VJ grid side. When the grid of VJ is near zero,
however, the potential at point A is about 40 volts. This figure
is obtained by drawing the 50k ohm load line into the plate char-
acteristics of 1/2 6SN7 or a 6J5, typical multivibrator tubes. Con-
sideration must also be given to the voltage at the grid of V2 at this
instant: at cutoff, a very high negative potential must appear on
the grid; as we have done before, we might assume a voltage of at
38 MULTIYI BRA TORS

least twice cutoff, but, as we shall see, it is actually much greater

than this. As a matter of fact, it is approximately equal to the
swing of the V2 plate voltage during transition. (200 - 40 = 160
volts.)
Ep 2 200 volts
Eg 2 - l 60 volts
Ep 1 40 volts
Eg 1 zero volts
Therefore:
Charge on CI = 200 volts
Charge on C2 = 65 volts
Consider now what must be happening in the grid circuit of
V2 if there is a - 160-volt difference between the grid and cathode:
a current must be flowing through grid resistor Ru 2 under the pres-
sure of this difference of potential. Obviously, this current must
be flowing out of C2, because grid current cannot flow when the
grid is so negative with respect to cathode; thus the voltage across
C2 must be changing. This capacitor must be "moving" up along
an exponetial discharge curve, which would terminate when Rg 2
stops taking current or, in other words, when point D reaches
ground potential (if other events did not prevent it) . Something
does happen, however, that changes the smooth discharge of C2:
at the same time point D arrives at the -12-volt level, V2 begins
to conduct, sending a negative pulse to point C, producing a pos-
itive pulse at A and, therefore, a positive pulse at D. Thus the
potential at D is suddenly lifted above the zero axis (Fig. 21),
but drops quickly to approximately zero as the V2 grid draws cur-
rent and brings the potential at D to ground level.
All that remains now is to show that, at the instant when V2
begins to conduct, EP 1 (point A, Fig. 20) , jumps from its conduct-
ing plate voltage of 40 volts to its cutoff plate voltage of 200 volts,
completing the second transition. As V2 starts to conduct, point B
begins to become less positive, carrying C with it. The time re-
quired for V2 to go from the condition where conduction just
starts to the tip of the overshoot (Fig. 21) is virtually zero, hence
C goes to the -160 volt point just as quickly, with VI cutting off
almost instantaneously. The process now repeats, this time with CJ
discharging along an exponential curve until conduction in VJ
again commences. The grid voltage waveforms for VJ are, there--
fore, exact reproductions of those for V2, but are displaced by 1/2
cycle along the time axis.
 A-STABLE MULTIVIBRATOR 39

I INSTANTANEOUS OVERSHOOT AT D
• ,, GRID CURRENT
~ R I N G S D TO ZERO
0 V. - - - - - - - t - - - - - - - - , r - - - - - - - - l f - " - -

12V.----
CUT-OFF ~~--
BEGINS
CONDUCTING

GRID

------
TIME

(A)
APPROX. POTENT! AL AT
-160V. -- - D STARTS HERE

2 '\:Cl CHARGING
200V.--~,------,
r&

PLATE VI CONDUCTS
CUTTING OFF
---v2

4
4 0 v . - ~ - - - - - . , - - - . - - -.....
3 ~ 5
\ Cl
DISCHARGING
ov. - - - - - - - - - - - - - - - - - - - -
(B)

Fig. 21. V2 waveforms in a-stable multivibrator.

The output waveforms at the plate of V2 are shown in Fig.

21 (B). They begin at the point where the grid of V2 is well below
cutoff (-160 volts), no current is flowing in the plate-cathode
circuit, and the plate potential is equal to that of the supply (200
volts). The voltage at B remains at this level until the instant when
V2 begins to conduct; now the almost instantaneous change takes
place in which the grid of V2 soars above the zero line (instan-
taneous overshoot), taking the plate voltage down into an over-
shoot below the 40-volt line, as shown from 2 to 3 in Fig. 21 (B).
The plate potential of V2 is quickly established at 40 volts as grid
current brings the V2 grid to zero (3 to 4), remains at 40 volts
as CJ discharges (4 to 5), and then returns to the 200-volt level
when VI begins to conduct and cuts off V2 (.5 to 6).
40 MULTIVIBRATORS

The rounded corners at I and 6 are due to the charge time of

CJ. Since this period is extremely short, such defects in the square
waveform are not noticeable in practice. Of course, the output
waveform at the plate of VJ is the same as that for V2, again being
displaced along the time axis by 1/2 cycle.

27. The Frequency of the Multivibrator

Formulas for computing the frequency of an a-stable multi-

vibrator are available, but these are complex, difficult to apply,
and outside the scope of this book. A surprisingly accurate measure
of the frequency of a multivibrator may be obtained, however, by
applying the foregoing analysis.
Three factors are of importance in frequency determinations:
a. The RC time constant of each capacitor-resistor combina-
tion in the grid circuits of the tubes. Assume these to be sym-
metrical.
b. The actual cutoff voltage of the tubes.
c. The plate voltage swing between the conducting and the
non-conducting state.
Again, a numerical example helps to explain the procedure.
We shall use the circuit and values of Fig. 20. When a tube goes
from a non-conducting to a conducting state, its plate voltage drops
from 200 volts to 40 volts. This forces the grid of the other tube
to jump from about zero volts to a level of -160 volts. This tube
remains non-conductive until the capacitor discharges from -160
volts to the cutoff level, -12 volts in our example. While the total
possible discharge range is from -160 volts to zero volts, i.e., 160
volts of change, the actual range is only 148 volts (-160 to -12) .
The real change is then:
!!~ = .925 = 92.5 percent of the possible change
The time required for a capacitor to discharge 92.5 percent of
its voltage may be found in the universal time constant charts and
is virtually equal to 2.6 time constants. The time constant of
either RC combination in the example of Fig. 20 is (R = I megohm
C = .002 11,t):
TC = 1 X .002 seconds
2.6 TC periods = .002 X 2.6 .0052 seconds
 A-STABLE MULTIVIBRATOR 41

The time of .0052 seconds, then, specifies the interval during

which either tube is cut off while the other is conducting, so that
the time required to complete a full cycle is 2 X .ll052 = .0104
seconds. The frequency (reciprocal of period) is, therefore,
lj.0104 = 96.1 cycles per second.

28. Synchronization of the A-Stable Multivibrator

The dependence of the a-stable multivibrator period on a

number of circuit and voltage conditions acting together leads to
poor frequency stability, a condition which is advantageous when
it is desired to synchronize the oscillator with some external signal
source. A crystal oscillator, in contrast, is so stable that it is dif-
ficult if not impossible to make it synchronize with any frequency
removed from its natural one. Thus the instability of the multi-
vibrator, a characteristic having an unpleasant implication, is the
very thing that makes it so useful in most of its applications.
In modern practice, multivibrators are commonly synchronized
by peaked waveforms, although other shapes, such as square or
sinusoidal waves, may be employed if the conditions warrant. In
any case, either a negative or positive pulse of sufficiently great
amplitude may be used if correctly injected. Figure 22 illustrates
the importance of pulse amplitude (positive pulse) in triggering
a multivibrator before the end of its natural period so that it will
synchronize at a frequency somewhat higher than its free-running
repetition rate. In (A) , the pulse, at the time it appears, is too
small to pull the tube out of cutoff, so that it comes and goes with-
out affecting the multivibrator frequency in any way. The natural
half-period, as determined by the factors discussed in Par. 27, is
shown here as t. When the pulse amplitude is greater (Fig. 22 B) ,
even though it appears at the same time as before with relation to
the multivibrator cyde, switching does take place at point I rather
than at point 2 when there are no sync pulses. Note that the half-
period is now substantially shorter (ti), which means that the mul-
tivibrator frequency is higher. The action is repetitive, of course,
and synchronization holds cycle after cycle; under these conditions,
the multivibrator is said to be "locked in."
A positive trigger injected into a conducting tube cannot alter
the frequency of the multivibrator, because the tube is already con-
42 MULTIVIBRATORS

,,,:~
I
I
TRIGGER TRIGGER
(A) (B)

0 -,---~""'1~-,.--
c.o. 1
E!f I
I
I
I
I
__L_
TRIGGER

(C)
Fig. 22. Grid waveforms in a-stable multivibratar, showing trigger pulses. (A)
trigger too weak, (B) trigger controlling properly, (C) late trigger due to natural
multivibrator frequency higher than trigger frequency, (D) sync at submultiple of
trigger frequency.

ducting. The pulse rides the top of the waveform (Fig. 22C) , with-
out influencing the multivibrator one way or the other. Negative
synchronizing pulses are often present in timing equipment, and
may be utilized to lock-in a multivibrator by injecting them into
the grid circuit of the conducting tube. The action is an indirect
one in this situation; the negative pulse does not cause the con-
ducting tube to revert suddenly to the cutoff state; rather it is
amplified and inverted in phase by the conducting tube and is then
applied as a positive pulse to the grid of the tube in the cutoff
state. Hence the action is similar to that depicted in Fig. 22 at
the instant of switching.
In general, the natural period of the multivibrator must be
greater than the interval between synchronizing pulses (Fig. 22B)
if locking-in is to be realized; otherwise, the pulses arrive at the
grid of the tube after it has gone into conduction, and cannot af-
fect the frequency. Thus the sync pulse repetition rate must be
of higher frequency than that of the multivibrator for successful
synchronization. In some applications, the aim is to synchronize
 A-STABLE MULTIVIBRATOR 43

the multivibrator at some sub-multiple of the sync frequency; for

example, the synchronization of a 60-cps multivibrator by pulses
having a repetition rate of 180 cps. This is illustrated in Fig. 22 (D).
The multivibrator is tripped on pulses I, 4, 7, IO, 13, etc., the
relationship between periods being 3: I as shown in the diagram.
Synchronizing pulses are commonly injected into the multi-
vibrator system by means of coupling capacitors to the grid of one
of the tubes or through a transformer having a low impedance
secondary as shown in Fig. 23.

29. Cathode-Coupled A-stable Multivibrator

The circuit of Fig. 24 should be compared with that of Fig. 15,

the self-biased mono-stable multivibrator. It will be noted im-
mediately that the only difference between these two arrangements
is that in the mono-stable circuit R4 is returned to cathode, while
in the a-stable circuit this resistor is returned to ground. In the

B+

Fig. 23. Transformer coupling

of trigger to a-stable multi-
vibrotor.

last few sentences in Par. 21, it was pointed out that the cathode
return of R4 to V2 is capable of placing this tube in the stable
state, because the grid bias is zero under these conditions, resulting
in a heavy current through R5, the common cathode resistor. The
large voltage drop across R5 provides a sufficiently high bias for V1
to keep it cut off; hence the circuit is stable. In the case of the
ground-return grid resistor (RJ of V1 in Fig. 15), the current
through R5 is not nearly as great, and the voltage drop across it
44 MULTIVIBRATORS

is too small to keep V2 cut off after (C) discharges; therefore, this
results in the quasi-stable state.
It should be evident from this that should both grid resistors
be returned to ground as in Fig. 24 the circuit would have no stable
state at all. In short, with either tube conducting the quasi-stable
condition would obtain and, of course, must result in free-running
oscillations.
The sequence of voltage transfers that result in one cycle of
free-running oscillation is outlined below:

B+

Fig. 24. Circuit of o-stoble

cathode-cou pied
R4 multivibrator.

a. With the application of power, plate current starts to flow

in both tubes. Its size is limited by the bias applied as a voltage
drop across R5.
b. Assume that the plate current of VJ is slightly greater than
that of V2. This produces a negative pulse at its plate that is trans-
ferred instantaneously through (C) to the grid of V2, starting the
transition which terminates with V 1 full-on and V2 cut off.
c. When C discharges sufficiently, V2 rises out of cutoff,
draws plate current, and increases the drop across R5, thus raising
the bias on VJ.
d. The plate current of VJ drops slightly, generating a posi-
tive pulse at its plate, which is instantaneously coupled to the grid
of V2, increasing its plate current still further. Thus the transition
 A-STABLE MULTIVIBRATOR 45

V2 1
FULL---'
ON C CHARGING THROUGH
E':I !SMALL T.C.
FOR
V2 C CHARGING
-THROUGH LARGE
T.C.

VI STARTS
ONDUCTING
I I
V2
CUT-OFF II (A)
I I I I
I I I I
B+ - - - - - -~- - - - -- -1-+- -- -
I I
Ep I II
FOR
VI I
I

(B)

Fig. 25. Waveforms in cathode-coupled a-stable multivibrator.

to the state in which V2 is full-on and V1 is cut off takes place. VJ

is held non-conducting by the large voltage drop across R5.
e. At the instant when the grid of V2 is driven positive by the
pulse from the plate of V1, grid current begins to flow in V2, and
C begins to charge through two_ separate paths:
1. R5, cathode-to-grid resistance of V2, and RI
2. R4 and RI
The short time constant of these parallel paths permits C to charge
quickly until the grid reaches cathode potential. Now grid current
ceases, but the capacitor continues to charge through path (b) at
a slower rate than before, because the time constant of this path
alone is higher. When the voltage across R5 drops to the cutoff
voltage of V1, this tube instantly starts to conduct, renders V2 non-
conducting by passing a negative pulse from its plates through C
to the grid of V2, and the cycle is complete, with VJ full-on and
V2 cut off.
Figure 25 (A) illustrates the voltage variations at the grid of
V2 and shows the change in waveform that occurs when C transfers
46 MULTIVIBRATORS

C•
CHARGING

/
o~~----+-""'-r-------,i---.--------+'----

E9
FOR
V2

I-
r--- SHORT
TC (A)

r V2
FULL
ON

Ep
FOR
V2

(Bl

Fig. 26. Asymetrical waveforms in cathode-coupled a-stable multivibrator in which

V2 grid draws current over most af cycle.

from the low to the high TC charge path. The effect this has upon
the voltage at the plate of VJ is pictured in Fig. 25 (B). Since the
output square-wave from a cathode-coupled multivibrator is almost
always taken from the plate circuit of V2 (in which square waves
like those in Fig. 21 (B) are produced) the distorted voltages at
the VJ plate are of no significance.
The cutoff interval for V2 is proportional to the time constant
on discharge for capacitor C, while VJ remains cut off for a period
that is a function of the charge time constant of the same capacitor.
The latter, as the foregoing discussion points out, may be subdi-
vided into two distinct TC periods depending upon the presence
or absence of grid current in V2. It is, therefore, quite easy to make
the charge time constant much shorter than that of discharge by
proportioning the circuit constants in such a way as to have V2
draw grid current for most of the half-cycle. When this is done,
the output of the multivibrator becomes asymmetrical, as dia-
 A-STABLE MULTIVIBRATOR 47

grammed in Fig. 26. The property of asymmetry 1s often useful

in specialized radar and television applications.

30. Review Questions

(1) Differentiate between the a-stable multivibrator and the bi-stable type.
(2) Draw a circuit representing an a-stable multivibrator.
(3) Explain in detail the circuit operation of the a-stable multivibrator.
(4) Enumerate three factors of importance in the frequency determination
of the a-stable multivibrator.
(5) Explain how an a-stable multivibrator may be synchronized with ex-
ternal signal sources.
(6) What is meant by "locked in" when the term is applied to a multi-
vibrator?
(7) What changes are necessary in Fig. 15 to convert the circuit to a
cathode-coupled a-stable multivibrator?
(8) Explain the operation of the cathode ooupled a-stable multivibrator.
(9) What type of waveform is most commonly employed to synchronize the
a-stable multivibrator?
(IO) Recapitulate briefly the differences between the fundamental classes of
multivibrators reviewed to this point.
 INDEX
A-stable multivibrator, 2, 3, 36 Grid waveforms in a-stable multi-
A-stable multivibrator, action of, 36, vibrator, 42
37, 38, 39
A-stable multivibrator, frequency of,
locking circuit, 3, 18
40, 41
A-stable multivibrator, synchronization of,
41, 42, 43 Mono-stable multivibrator, 2, 3, 24
Asymetrical waveforms, 46 Mono-stable multivibrator, circuit
operation, 24, 25
Multivibrator action, 4
Bias voltage, function of, 7 Multivibrator, definition of, 1, 2
Binary, 3 Multivibrator, properties of, 4
Binary counter, 3
Binary scale-of-two counter, 16
Binary scale-of-two counter, analysis of, Negative going pulse, 3
16, 17 Negative triggers, 13
Bi-stable multivibrator, 2, 3 Net voltage, 8
Bi-stable multivibrator, analysis of, 8,
9, 10 "Off," 5
Bi-stable multivibrator as a regenerator, "On," 5
17 One-shot multivibrator, 3
Bi-stable multivibrator, circuit diagram
of, 7 Part-on, 8
Bi-stable multivibrator, circuit of, 6 Peaked trigger, 28
Bi-stable multivibrator, design procedure, Plus and minus triggers, 13
21, 22, 23 Positive going pulse, 3
Bi-stable multivibrator, locking circuit, 18 Positive trigger, 12
Bi-stable multivibrator, trigger action of, 10 Pulse output, 12
Bi-stable ta mono-stable conversion, 24 Pulse width, 14

Cascaded binary counter, 13 Regenerator, 3, 18

Cathode-coupled a-stable multivibrator, 43 Relaxation oscillator,
Cathode-coupled a-stable multivibrator, Repeater, 3
circuit of, 44 Reversion action, 26
Cathode-coupled a-stable multivibrator, Rise time, 15
waveforms in, 45, 46
Commutating capacitor, 15 Self-biased bi-stable multivibrator, 11
Count-down multivibrator, 33, 34 Self-biased bi-stable multivibrator,
Counting dawn, 32, 33 circuit of, 11
Cut off, 8 Self-biased mono-stable multivibrator,
28, 29, 30
Decay time, 15 Self-biased mono-stable multivibrator
Design procedure, bi-stable multivibrator, positive grid return, 30, 31
21, 22, 23 Shunt capacitance, 14
Diode trigger feed, 18, 19 Square-wave output, 12
Stab le state, 2
Synchronization of a-stable multivibrator
Eccles-Jordan trigger circuit, 3 41, 42, 43

Flip-flop, 3 Three-to-one mono-stable multivibrator,

Flopover circuit, 3 33, 34
Frequency divider, 3 Trigger circuit, 10
Frequency division, 32 Trigger pulses, simultaneous applications
Full-on, 5, 8 of, 13

Gating, 4, 28 Waveforms, mono-stable multivibrator, 27

48