WM ‘1

”MINIMUM"

W

t

r
b

l
\

l

_s
N
N

 

(.

Q35? A fifiNERATQR as: sf‘AmAm

DESKECW. ANALYSifi AND '?E$?§NG

F‘iéQUENCEfl

Thasix {9r i‘ha 9% inf M. 3.
MSCHIGAN WATT: COLLEGE

i

Aésivm? F‘Warran Raickei‘é

i949

This is to certify that the

thesis entitled

"Design, Analysis and Testing
of a Generator of Standard

Frequenf'es"
presente g

 

Adelbert Warren Reickord

has been accepted towards fulfillment
of the requirements for

Q "‘ 7"
Met... degree in b.1110

 

 

 

Date May 17L1949 _

 

Bil-795

 

DESIGN, ANALYSIS AND TESTING
OF A GENERATOR OF STANDARD FREQUENCIES

By
ADELBERT WARREN REICKORID

A THESIS
Submitted to the School of Graduate Studies of
Michigan State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

MS TER OF SO IENCE

DEPARTLENT OF ELECTRICAL ENGINEERING
1949

THESIS

m.11-r-v-u 011 C \Irm'm'yvmc

.Léln.lJ-c._l .n—lfl—d‘ .LL:

3‘".- ‘T")‘ ~35 02‘ F 1:171:11?)

1.1 Introductory Discussion

1.2 Stanlerds of Frequency - Classification
1.3 Xetnods of Pregaency Iegsnrement

midrd 14

r1"* 3712,"??? II
‘v'..~'.‘- *M-b

01

THE OS”IQL;EOR

2.1 General Dis01ssion

2.2 Crystal Oscillators

2.3 Fecative Resistance

9.4 Conditions for Oscillation
Fig. 2.1 Enaivalent Circuit
9.5 utility

OkOQOlolUlfifl

2.6 oabiliza tion 1
2.7 1.0n—zrys sta 1 Fe3ative-Resistence
Sccillators 12

H
{‘0

2.8 Tne Dvnrtron Oscilletor
F13. 2.2. syn tron- -oscill tor Circuit
2.9 The $1 nsitron.Qs scillc‘

FHA
©<fi

1.0]?

Fig. 2.3 T1.;sitron Oscillator

Fig. 2.4 The resative nesistgnce Transi-
tron circuit

#1 14
a A»

or Vin-‘1) III

‘J- .{Z-A. .L..—.Lh

73 IULT YIFgngR 18
3.1 inllu”ulve Dis sion 18
5.2 14m17ul cl Dis01ssion 22
5.3 Ialtivib rrtor Xaveforms 23
3.4 Synchlonization 2
Fig. 5.1 The Bositive-grid ;_u1tivil nirtor 26
F13. 5.2 Lultivihrator have Luiges 27
Fig. 3.5 Lultivibrntor Tests 28
I1 * r“ T’T
12-..--- 1.51 .LV
CInCVIT 212132 29

1 .‘w-frj-fi-fi-rr-1 l‘
‘ ..‘

1 .
J v.1. --v..J.a _...L.J

Fig. 4.1 Stability Wests
Fig. 4.2 Harmonic ”olto es

(11’ ' --*,‘r:1~—“:‘.~ V

"—1 . .-x.~ 4-..4‘L

"1 “"T”1‘.":'f.;"r 111-1: 177‘ .‘\ 3. _ "1‘T‘7T/‘1 *"r ' m- “far-1T ‘\‘1~T(;
J‘d‘!d’—a‘.Jv'+\/.LIKI -kllJ U*. -.- —!.$J.\ \4' .LLI"J‘—AL'Jl\/—-LVI.A~H

‘

. —‘

rcait Diagram

"1

”chenatic C

i...

Biblio§r3phy

3&“8
37
38

59

C H A P T E R I
STANDARDS OF FREQUENCY
1.1 Introductory Discussion

In the universe, as we know it, there have
been observed many phenomena which have a recOgnized
periodicity or standard frequency. Several of these
have come into prominence through their continued use as
fundamental constants.

As is well known, our fundamental standard
of frequency is the period of the earth's rotation about
its axis. This period is, to a very high degree, con-
stant. Obviously, this period could not be exactly a
constant due to the action of the lunar and solar gravi-
tational forces, to name but a few.

The mean, or average, period of rotation can
be very accurately measured by astronomical methods and
is known as "one cycle per day". Almost all standards
of frequency are referred to this fundamental source
for calibration purposes.

1.2 Standards of Frequency - Classification

 

 

 

A frequency standard may be classified as
either a primary or secondary standard.

A primary standard of frequency is an in-
strument capable of generating a highly accurate, con-
stant frequency which is regularly checked against the

earth's rotation. USually, this frequency is generated

to

by a quartz crystal which is used to regulate the time
shown on a clock and which also activates harmonic and
sub-harmonic generators thus providing a series of high-
ly accurate known frequencies. Experimentation is also
being done using atomic and molecular spectral lines

as standard frequency sources.

A secondary standard of frequency is the
frequency of a stable oscillator that is regularly
checked against the frequency of a primary standard.
1.3 Methods .93 Frequency Measurement

There are several well known methods of
frequency measurement which are currently in use, of
which an outline follows.

By the use of suitable circuits, harmonic
multiples of the standard frequency may be obtained as
was indicated above. Therefore, known frequencies may
be obtained in the region of practically any unknown
frequency, and the problem reduces to one of interpola-
tion.

(a) Direct - Beating methods

In these methods voltages of the unknown
frequency and of tm known frequency are impressed
upon the input of a detector and the value of the re-
sulting difference beat-frequency is determined by

methods such as the following:

1. Matching the beat frequency with that
of an adjustable, calibrated interpola-
tion oscillator.

2. Measuring the difference beat frequency
on a frequency bridge or on a cathode
ray tube by means of Lissajous patterns.

3. Measuring the difference beat-frequency
on a pulse or frequency counter or com--
ter circuit.

4. Measuring the difference beat-frequency
on a direct reading frequency meter.

(b) Direct - Interpolation method
The frequency of an interpolation oscillator
of suitable range is ~ zero beat against the unknown fre-
quency and the adjacent harmonics of the standard fre-
quency in turn, with the resulting information yielding
the value of the unknown frequency.
(c) Harmonic - Interpolation method
This method consists of beating together the
unknown frequency and a harmonic multiple of the inter-
polation oscillator's frequency or, alternately, of
beating together tn. interpolation oscillator's fre-
quency and a harmonic sub-multiple of the unknown fre-
quency which leads to the value of the unknown frequency.
(d) Subtraction method
1 This method consists of combining the un-
known frequency with a suitable known frequency and ob-

-11..
taining a resultant which has one less digit than the
original. Then another frequency is combined with the
above resultant and a second resultant is obtained
which differs from the original by two digits. This
process is repeated until the resulting frequency
difference is negligible.

(e) Direct measurement method

This method consists of directly beating
the unknown against a calibrated variable frequency
oscillator until a null point is obtained. When this
is obtained the frequency of the oscillator is the
same as that of the unknown.and thus the unknown fre-
quency is measured.

There are other methods, examples of'which
are leoher wire measurements, direct reading frequency
meters, etc.

This thesis is concerned with the con-
struction of a standard frequency generator which may
be used as a frequency source for frequency measure-

ment purposes.

C H A P T E R II
THE OSCILLATOR

2.1 General Discussion

 

In any device for the precision measurement
of frequency the most critical feature is, of course,
the frequency controlling device. In frequency stand-
ards this section is the oscillator and any oscillator
controlled harmonic or sub-harmonic generators. In-
vestigations carried out by numerous researchers over
the last score or more years have indicated that, of
the several types of known oscillators, the piezo-
electric crystal oscillator is the best.

The crystal oscillator has been found to be

the most successful master oscillator. Frequencies
harmonically related to a submultiple of the base fre-
quency are best generated by crystal controlled os-
cillators.
2.2 Crystal Oscillators

Oscillators employing crystals may be clas-
sified in a number of ways. One classification is
based upon.whether or not the circuit without the
crystal is in itself an oscillator.

If the circuit, when the crystal is not
actively included in it, is not inherently capable

of sustained oscillation; then it is termed a "crystal"

oscillator.

If the circuit is capable of sustained os-
cillations when the crystal is not actively included
in the circuit,a1though not necessarily at the crystal's
resonant frequency, then it is termed a "crystal con-
trolled" oscillator.

2.3 Negative Resistance

Oscillators are in general classified as
negative resistance type oscillators. This negative
resistance is in general due to several circuit
features: Positive feed-back which is introduced to
balance the circuit losses, and the vacuum tube which
over a part of its Operational range is inherently a
negative resistance element.

The negative resistance is a necessity since
the alternating-current resonant-frequency ohmic los-
ses must be zero. If these losses were not identi-
cally zero, the amplitude of the oscillation would
exponentially decrease until its value be less than a
usable minimum.

2.4 Conditions for Oscillation

 

 

The analysis of the oscillator is based upon
the fact that the circuit may be considered to consist
of a anti-resonant RLC combination in parallel with the

negative resistance element. This is Justified because

-7-
the crystal itself performs as if it were a high-Q,LC
anti-resonant circuit. Numerous investigators have
empirically develOped its equivalent circuit, and the
Operation of crystal oscillators may be determined by
using the established empirical relationships. The
analysis is performed exactly the same as that for an
LC oscillator circuit.

3; * (z 4::

7t:
f if

Fig. 2.1 Equivalent Circuit.
By using the Kirchhof laws we may write
pang) +114,” + £94 = 0
4(2

aKf
whereI/9 , algebraically positive, is the negative

41‘}+,Q,,'+C.§ 4/5 = 0

resistance element.
By methods of differential equations we may
eliminate one of the unknown currents and obtain a

single equation in one unknown

2.
we: Wm»; WM

-3-

or after dividing by /oLC
JZZ,+ R iz’,‘ 4 Q / ' z
2;: CT ,ZE/Z zc $07: 4

This resultant expression in one unknown is

a standard form.whose solution is known, which is
=7Ag " ’0‘) 52')? w!
w/zere
+
60.71/ch "Hf-[05+ 29/2

Thus it is seen that if w is a real quantity

 

then the determination of the type of oscillation, i.e.,
.constant, increasing or decreasing with time, is ex-
plicitly dependent upon the magnitude of If”.

Reich‘l) has summarized the results of these
expressions as follows:

A. Sinusoidal oscillation may occur if u)

is real, 1.6. if

 

A,
// r“
F >RC~2flC
l. The amplitude of oscillation decreases

with time if

5+l>o

l. 232“

 

(1) Reich, H. J., Theory and Applications of
Electron Tubes, MbGrawéHill, 1944, 2nd Edition, p. 376

-9-

2. The amplitude is constant if

M:- ———-

3. The amplitude increases with time if

z,
/’0/<755

B. The current is eXponential in form if

is imaginary, i.e., if

1,
RC - ZVLC

l. Sustained oscillation occurs if

wee—i=5

a. The current wave is continuous, but

//0/ <

non-sinusoidal if [pl > R-

b. Triggering occurs and the wave has
discontinuities, i.e., relaxation
oscillation takes place, if ]f>'<'F?.

2. The current is an exyonential pulse if
I/OI>§% and //O/2 R.
2.5 Stability
In any source of constant frequency output the
stability of the oscillator section is of major impor—
tance. If the instantaneous or Operating frequency should
be grossly affected by incremental variations in the

Operating conditions, the utility of the instrument

-10..
would be drastically reduced.
The frequency stability of the oscillator has
been found to be dependent upon several factors:
(a) mechanical arrangement of the parts,
(b) Circuit parameters,'
(c) Tube parameters,
(d) Operating conditions, i.e., temperature,
vibration, etc.

2.6 Stabilization

 

By careful consideration of the mechanical
arrangement, i.e., using electrically and themmonical-
1y stable parts, rigid construction, thermostatically
controlled crystal housing, and by using voltage and
current regulation, moSt of the difficulties in items
(a) and (d) above can be effectively eliminated. The
lines of research to reduce the effects of the tube
and circuit parameter variations upon frequency sta-
bility have proceeded in several directions, an out-
line of which follows:

(a) PrOper choice of electrical parameters

of the oscillatory circuit, i.e.,
(1) Resistance stabilization
(2) Capacitance or inductance stabili-

zation

(b) ‘Use of selective filters as the oscil-
latory circuit

-11-

(c) Elimination of harmonic content by

means of tuned filters

One of the simplest ways of improving the
frequency stability of an oscillator is by resistive
stabilization, which consists of the addition of a
high resistance between the plate and the oscillatory
circuit. The purpose Of this is to make the total ef-
fective resistance in the plate circuit so high that
variations in the dynamic plate resistance will have
little effect upon the oscillatory circuit.

L1ewellyn‘2) has deveIOped a more general
type of stabilization. By deriving the analytical
eXpression of the requirements of oscillation for a
general type circuit he found that by using capaci-
tance or inductance in series with the grid or plate
of the oscillator, or both, complete independence of
oscillation frequency from variation in tube para-
meters resulted.

A number of investigators have shown that
frequency variation and harmonic control are inter-
dependent. The factors that ensure low harmonic
content, such as a linear Operating characteristic

and small amplitude, therefore also tend to improve

 

(2) Llewellyn, F. B., Constant Frequency Oscil-
lators, Proc. I.R.E., V01. 19, p. 2063, Dec., 1931

-13-
frequency stability. Stability can also be improved
by the use of series filter sections tuned to the har-

monic frequencies, shunted across the tube circuit.

2.? Non-Crystal Negative-Resistance Oscillators

 

Chakravart1(3) has analyzed and classified
negative resistance oscillators into three classes as
follows:

A. DYNATRON TYPE, in which the internal
resistance of a triode or screen-grid
tube under secondary emission condition
has been used to obtain negative
resistance.

B. TRANSITRON TYPE, in which a five element
or double-grid tube employing negative
transconductance has been used.

c. FEED-BACK or REGENERATIVE TYPE, in which
the input and the output terminals of a
one-way amplifier are connected in series

or parallel.

 

2.8 The Dynatron Oscillator
In a screen-grid tube if the electrode is
Operated at a higher voltage than the plate, and there

is sufficient secondary emission from the plate, there

 

(3) Chakravarti, S. P., On the Nature of
Negative Resistance and Negative Resistance bections,
Phil. mag., Vol. 30, p. 294

-15..
is a range of plate voltages over which the net re-
sultant plate current will decrease with increasing
plate current.

The negative resistance results from the fact
that while the number of primary electrons that the
plate receives is independent of the plate voltage, the
number of secondary electrons produced at the plate in-
creases with increasing plate voltage.

It is seen.that the magnitude of the negative
resistance can be controlled by varying the control-
grid potential. Fig. 2.2 shows the schematic circuit

diagram of the dynatron oscillator.

 

 

 

 

 

Fig. 2.2 Dynatron-oscillator Circuit
This type of oscillator has one general
type of disadvantage. This is its dependence upon
secondary emission. Secondary emission changes with
use of the tube and large differences have been ob-
served in the shapes of the characteristics of indi-
vidual tubes of the same type. Since for stable Opera-

tion the influence of tube parameter variations should

-14..
be as small as possible this type of oscillator is ef-
fectively eliminated.
2.9 The Transitron Oscillator

 

The transitron oscillator is also known as
the retarding field or negative-transconductanoe os-
cillator‘4).

The pentode is connected as indicated in
Fig. 2.3 and the potentials so adjusted that there
is a virtual cathode between the screen and the sup-
pressor. Under these conditions a fraction of the
space current drawn from the cathode is returned back
toward the cathode, to be ultimately collected by the

BOIBODO

 

#-
F

 

 

Lb

 

RC. " } 7—9- /
i L‘]

A w" A L h c

} - r, "P- 2F

Fig. 2.3 Transitron Oscillator.

 

 

 

 

 

 

 

 

 

 

 

 

(4) Reich, H J Theory and A
. . p lication of
Electron Tubes,.McGraw Hill: 1944, 2nd Edition, p. 381

-15-

In the transitron oscillator, and in the
circuits derived from it, both the suppressor and
screen voltages are permitted to have a.c. components.

Another distinguishing feature of this type
circuit is that neither the anode nor the first grid
plays a part in the a.c. circuits. These electrodes
are maintained at pure direct current potentials al-
though in many practical instances the synchronizing
voltage is applied through the first or control grid.

Instead of using the first grid voltage to
control the anode current, as in a triode, we have
here the novel device of using the suppressor voltage
to control the screen current. Thus in a transitron-
connected pentode-amplifier the load is connected in
the screen circuit, and the input voltage is con-
nected in the suppressor circuit.

The variations of the anode voltage of the
pentode have very little effect on the total space
current. This is because the screen grid effectively
screens the cathode region from.the influence of the
anode-voltage.

With the suppressor connected to the cathode
in the orthodox way, the proportion of the Space cur-
rent going to the anode increases with anode-voltage
until finally nearly the whole space current becomes

anode-current. If new the suppressor voltage is made

~16-
increasingly negative, the repulsion of electrons by
this electrode increases, and the prOportion of the
space current which reaches the anode is reduced. The
prOportion going to the screen is thus increased and
it follows that making the suppressor voltage negative
increases the screen current.

Contrast this with the behavior of a triode,
where making the control-grid voltage negative de?
creases the controlled current. Herein lies the key
to the prOperties of the transitron, positive feed-
back may be applied simply by R-C coupling the output to
the input.

If, in Fig. 2.4, Co and R0 are large enough so
that a change in voltage of the screen grid G2 is ac-
companied by a practically equal change in voltage of
the suppressor grid G3, the action is mathematically as

follows:
. I
AZCQ = Lea + A 8c3 632. = Aecz (732 +632)

’32
where AZ}; is the change in the current of the
screen grid G2,

Aecz is the change in the voltage of the
screen grid G2,

figz is the screen grid resistance,

A863 is the change in the voltage of the sup-
pressor grid G5, and

-17-
C;32 is the mutual conductance between G2 and
G3.
Since G32 is negative, it follows that, if the
magnitude of G52 exceeds the magnitude of l/{az , an
increase of screen voltage is accompanied by a decrease

of screen current.

 

 

 

 

 

 

 

 

 

 

Fig. 2.4 The negative resistance
transitron circuit.

Thus it follows that the transitron type
oscillator is more stable with time and will have a
lower negative resistance. Because the control grid,
61, is at a direct current pOtential the synchroniz-
ing voltage may be applied through this electrode.

-13-

C H A P T E R III
THE MDLTIVIBRATOR
3.1 Qualitative Discussion

The zero bias multivibrator is well known
and much.has been written.about its Operation. How-
ever, the extremely useful characteristics of multi-
vibrators Operating with positive grid return are not
as well known. The Operation of a multivibrator using
a positive grid return will be described and compared
to that of a conventional multivibrator.

The multivibrator'shown in Figure 3.1 con-
sists of a two stage resistance-capacitance coupled
amplifier with.the output returned to the input so
that the circuit is regenerative and satisfies the
Barkhausen criterion for sustained oscillations.

The method of oscillation of both types of
multivibratore is essentially the same. This method
of oscillation follows. If the plate current of
tube T2 is momentarily increased while that of tube
T1 remains constant or decreases, it would start the
following chain of reactions:

(a) The plate voltage of T2 would decrease:

(b) the grid of T1 would become more negative;
(c) the plate current of T1 would decrease, al-

-19..

lowing its phgte voltage increase;

(d) the :rid of T3 would increase, adding to the
original change that started the reaction. Fecuuse the
Barhhausen relationship for oscillators is satisfied,
the reaction will prOgress at a rapid rate until tube
T1 is cutoff, placing its plate voltage at substan-
tially that of the supply, while T2 has its grid posi-
tive with reapect to its cathode and considerable drOp
across its plate load.

The voltage on the grid of T1 now rises as
the coupling capacitor charges, approaching the volt-
age of the supply grid Ec asymptotically. Uhen.the
grid of T1 nears the cutoff voltage, T1 starts to con-
duct, so that its plate becomes increasingly negative.
This reduces the voltage on the grid of T2, causing
its plate to become increasingly positive and thus
accelerating the already rising voltage on the grid of
T1. When T1 becomes sufficiently conducting to make
the combined gain of T and T3 greater than unity, the
circuit "flips over"; i.e., the two tubes ch nge plcces,
the grid of T1 becoming positive with reapect to its

cathode while the grid of T. is driven Lelow the cut-

N

.L.

off voltage. The Operation is then repeated, the grid

of T3 rising exgonentiully until T3 becomes conducting,

when a second flip-over ozcurs, etc.

-03-

.L ‘ .-
u.-e "1

J's-1‘-
L

In tic abore anflyzis lects of the
grid current on the maltivibrctor's Operation h"ve been
ne“lected. These effects to a large extent defy enact

analytic l and quantitative exyression. a qualit tive

discussion followsla) ht the revere l saints, the yrid

y

L
.-

becomes positive with rescect to the cathode thus
monenturily lim'ting tn: plate voltage to a value less
than the plate Sipply voltage. is the coupling
capacitor charges, the grid voltage of the conducting
tube decreases allowing its plate voltage to increase;
this change in plate voltage is carried over to the
grid of the nonconducting tube, modifying the rcte of

1

rise of its grid voltage. Thus it is important that
the grid-circuit time constant be large compared to
the plate-circuit time constant so that the grid cur-
rent will decrease to a low value at the reversal

_-.w

points and therefore not affect the stability of one

A”!

multivibrator. Another factor ailecting multivibra-

tor stability is the input capacitance of the tubes

 

(5) Kiebert, Lartin V., Jr., end
Lultivibrator Circuits, 1.3.3., Vol. 0 , HO. d,

3

O

.‘3 (p
’ JL‘l-o.

-21..
which capacitance reduces the signal applied to the
grids.

The period of the multivibrator is depen-
dent upon the amplitude of the oscillations, the
resistance capacitance combination in.both the grid
and plate circuits, the synchronizing coupling circuit
and voltage, and the voltage to which.the grids are
returned.

In.the positive grid bias multivibrator the
period is increased by:

(a) Increasing the amplitude of oscillations
(by increasing the plate load resistance or decreas-
ing the cathode resistance);

(b) increasing either the resistance or capaci-
tance in the timing circuit, thus decreasing the charg-
ing rate of the capacitor; and

(c) decreasing the grid-return voltage, thus
decreasing the rate at which the capacitor charges.

In the zero grid bias multivibrator all
save the last statement above hold. Statement "c”
does not since the grid bias voltage is an invariant.

It is immediately obvious that of the two
aforementioned circuits the positive grid bias one is
more inherently stable due to the increased slepe of
the grid voltage within the operating range. Thus,
any incremental change occuring’in.any or the several

component values or Operating conditions will not
cause so great a change in the repetition frequency of
the positive grid bias multivibrator as in the zero
grid bias circuit.

This in itself seems sufficient justifica-
tion fer its inclusion as a crystal controlled genera-
tor of standard frequencies.

3.2 Analytical Discussion

The frequency of a multivibrator can be ap-
proximated in terms of the constants of Figures 3.1
and 3.2. When a grid is rising exponentially from its
maximum negative value (taken at time t=0), its in-

stantaneous voltage can be eXpressed as

-t/¥? C‘
‘393§;[;55 "’(£}“'£ifl)‘s 5? Cd]

where Ce =C+Cin is the effective capacitance in
the discharge circuit. The circuit will reverse when
elg reaches the cutoff value 592 =" [El/1U) ; the
value of t for this cutoff condition is the half-
period of the multivibrator, and thus determines the
frequency. The voltage E8, is found as follows:
C‘
‘éii ‘”‘£;b ’65;‘.‘£;3')(;'»c;;‘/

where E- __ 5’be #

73 RF {/3 {-Rk

_ at;
and Eb-EP3 ~ RP *Rl. +RK

The frequency of the multivibrator can now be written

-83...

 

f: 1

/ —r)

ZRQCeéh 15*"!
'where - Ec ana’ )’= 5:2!
F‘s; at

In deriving the frequency equation, the
following simplifying approximations have been made.
It is assumed that, before a change ever occurs, an
equilibrium condition is reached in which the grid
of the conducting tube is at the same voltage as its
cathode; the plate current is thus determined by the
intersection of the lead line for the combined slate
and cathode leads with the zero-bias plate current.
The effective amplication factor is determined by the
grid voltage at the effective cutoff point-othe point
where the over-all gain is just unity. Actually, a
very good approximation is obtained if the values for
[X and 9,, given in the tube manuals are used. It is
also assumed that all the grid voltage drop is across
the resistance R3 in the grid circuit.

3.3 multivibrator waveforms

The wave forms of the multivibrator are
greatly affected by grid conduction, since this puts
a heavy load on the plate circuit of the tube which
drives the grid. When the grid is positive, there

is a reasonably linear relationship between the grid

voltage and the grid current, the ratio for most

-34-
small receiver tubes ranging from a maximum of 2000
ohms to a minimum of 500 ohms. In an equivalent cir-
cuit diagram, therefore, the grid can be represented
by a resister and a switch in series, the switch being
open when the grid is negative and closed when the grid
is positive. Because of its small size the equivalent
grid-wire resistance need not be known exactly as long
as it is of the correct order of'magnitude. many
authors have neglected it completely in their mathe-
matical analysis of the multivibrator circuit.

In case the time constant of the circuit in-
volving the non-conducting slate is comparable to that
involving the non-conducting grid, the wave forms may
differ greatly from.the ideal square wave output.

This is because the non-conducting plate will not have
time to relax to Eb, and the conducting grid will not
have time to relax to zero. An exact calculation of
the wave form in this case is impossible. The reasons
for this are that the cutoff voltage on the non-
conducting tube is constantly changing, due to the
changing plate voltage, and the grid voltage of the
non-conducting tube does not follow an eXponential
curve.

3.4 Synchronization

 

When a voltage is injected into the multie

vibrator this voltage tends to cause the multivibrator

O I.
-53 e.)-

to lock itself to a frequency which bears integral
relationship to that of the injected voltage. This
ratio is entirely dependent upon the injected voltage's
magnitude. Exact details of the control depend upon
the way in.which the synchronizing voltage is injected
and the degree of symmetry between the circuit constants
of the two multivibrator tubes.

Increasing the amplitude of’the synchroniz-
ing frequency causes the multivibrator frequency to be
"drawn" in discontinuous steps toward the synchroniz-'
ing frequency, with the ratio being expressible always
as an integer of progressively smaller magnitude. See
Figure 3.3. This action is caused by the fact that in-
creasing the amplitude of the injected voltage enables
it to neutralize larger condenser charges, and hence to
cause a reversal of grid potential before the full
charge has had time to leak away.

When the multivibrator is to be controlled,
as it is in the present case, the uncontrolled fre-
quency of the oscillator should be slightly less than
the value when controlled, and in order to obtain the
maximum stability of the control the injected frequency
must have the proper amplitude, and the circuit should

favor 10:1 division.

 

 

{DE-b

 

 

 

 

 

R1 A}
C c
'5 -—- r f 7;
F——’ 7-- '_—OEC
Rx ii Rt
‘0

Fig. 3.1 The positive-grid multivibrator

 

Voltage on plate of T1

Fig. 3.2 .multivibrator wave shapes

:59
-LL -

1.2

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e p .,
§ 1)
t * °
N
H
b E
t *2
‘J c A
31% ’2 EJEUT‘]
N : I0 1‘
n "a
.o '6
‘3 4
b
0c -z , _
1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4 8 02I62024283z36

F180 303
Experimental results showing the effect of in-
creasing the amplitude of the synchronizing voltage, and

how certain methods of injecting the voltage tend to

favor either even or odd frequency ratios. These re-

sults were obtained for a zero bias multivibrator.

-29..

CIRCUIT DnSIGX
Considerable work was done in a two month

period on the multivibrator described in the previous
chapter. Several circuits were set up and qualitative
comparison tests were made between the zero bias and
positive bia multivibrators with particular attention
being paid to their relative stabilities and syn-
chronizing capabilities.

It mas founi tgat t;e stability of tue posi-
tive bias multivibrator n-s totally undesireble at the
required output level, a level which should provide a
usable lO-kc lODOth harmonic. The "breadboard" setup
used to test the circuit, although it consisted of the
components arranged upon a metal chassis, Les extremely
susceptible to variations in hand capacitance. The
action of the circuit at these times being completely
unpredictable, i.e., he fundamental frequency of the
circuit suffered extreme variations from the norm, al-
though a synchronizing voltage was applied at the time
and adjusted to an Optimum value.

The output of the multivibrator was coupled
to an oscillOSCOpe in an attempt to determine the wave

shape and harmonic content of the output of the multi-

vibrator. Various methods of coupling were devised
Which consisted essentially of condenser isolation but
nevertheless the oscilloscope at all times loaded the
circuit and at times the multivibrator was completely
thrown out of oscillation. bheu the c'rcuit did os-
cillate it nos overloaded so much when coupled to the
oscilloscOpe that the voltage output curve was char-
acterized by its extreme flatness of slepe. This type
of output is completely useless as it has negligible
high hrrmonic content.

A surplus BC 221 Frequency Heter was used
throughout in attempting to determine the Operating fre-
quency by the harmonic location method. Due to the
obvious limitations of the instrument, i.e., its sev-
eral coincident and simultaneous Operating ranges, com-
pletely contradictory and misleading results were ob-
tained and it never was determined at what frequency
the fundamental oscillation occured.

ociated circuits could have been devised

(1‘.

As
that would provide the required output and stability
using pentode type multivibrators, clippers, etc.,
but since this would yield results which could be ob-
tained only with more intricate, elaborate circuits
and also could not be incorporated into the chassis

available. Therefore the multivibrator was discarded

-31-
in toto in favor of a transitron type, sub-harmonic,
synchronized generator which would yield the some re-
sults with less trouble.

The transitron oscill tor, because the oscil-
lation occurs between the second and third grids, can
have its synchronizing voltage rpplied to the first or
control grid without additional loading or its accompany-
ing unpleesant results. Secondly, since the circuit
containing the negative resistance has across it the
inductive-capacitative combinHtion and not the RC
combination the synchronizing voltage me have a con-
stant amplitude at all Operatin: frequencies and is not
dependent upon the slepe of the grid voltage curve.

This wus considered an additional adventfge over the
multivibrator. Lastly, the LC oscillator circuit h's
in itself an inherent stability and selectivity
res;onee which the RC type circuit lacks. This was

I“

ered an aid in allevi ting the efiects of field

:2,

i

U)

coll
variations wh ch so noticibly filtered the exuected re-
Sponse of the multivibrator.

After the transitron circuit wis set up the
results improved considerably. The output LfiS recorded
to freluencies as high as 10 me #cycles with the CO
and 50 kc harmonic circuits in Operation on a super—

heterodyne receiver.

r* (3
-“)~ .-

The circuit control by the synchronizing
volta3e is not too critical 3n: it tus determined that
the optimum synchronizing voltage for the Sdhc “nd
l-kc oscilletors were W 1 use Which did riot differ by
two grett on amount and so a compromise res established
t1:t e2i ebled the transitron to oscill te at all three
sub—hormonics.

The sub-hormonic generator was tested and
found to be capable of oscillation et the fundencntal
of lOch end the sub—harmonic fundrmezt ls of SOkc,
EOhc, find lOkc. Due to the feet thet the one, fixed
tuned, HIV receiver hid not been completed at the time
consideruble difficulty has eXgerienced in the finil eo~
justnent of the o; cillitors usin: this receiver but mes
done on the superheterodyne.

Tunin; the freguency meter caused :5: multitude
of null points to occur and since the lowest irc.ie1
estable of :ein3 rec d on tiiS me er is lEhkc odditionn
error was thus introduced. llso it seemed very inlihely
thet this instrumen' should h:ve on exactly linear cali-
b etion curve.

Toise wfs veiy prevalent end increfsed in angli-
tude es the frequency increased until it completely sup-
pressed the output signal.

surnised tmit this ozcured bec: use of

cf
E“.
m

7.
.L

beet notes between the lOkc hernonics and the synchroniz-

in3 signal. fhese would occur ii the l'kc oscillator
were not adjusted exactly to lOkc.

y generator xas obtained

A veriable fie'uer'
which was capable of giving an output up to lOmc. it
this time the WUV receiver was eveile ble and therefore
the 50th hormonic of the lOOkc oscillator was zero
beat against the one fundamental standard of frequency
with the aid of the variable frequency generator. Error
was likely to lave been introduced at this step since
this receiver had not had incorporated within it any
means of eliminating the audio modulation present on
the carrier si 3nal. Also it did not have any meter with
which zero beat with the Smc ce rrier could be exec tly
determined.

St bility tests xere run on the 100 kc oscil-
lator by comparin3 its 50th hzrxuonic with the 5mc
stanfiard with the results tabulated. (Lee 313. 4.1).

It is seen from the schematic circuit dia ram
tnrt the freuiency determining elements are connected

.1

through a rotary switch to the second grid of the vacuum
tube V2 and that these control the fundamental freguency
of the transitron circuit. ihus the tron nsitron oscilla-
tor is only 0 er’tive when other than the lOch harmonic
points are desired. Eecadse of this the stability of

the ti :18 itron os .illutor should be ideitic l with 3m1t

of the lJOkc oscillator.

Effective, or 3.2.3., volt~ge UGLSJIEUEfltS
were made usin; a Vacuum tube voltmeter which had vari-
ous values of caowcitance inserted in its ingut le;d in

- L

be pt to deternine dLe energy content of various

A

an at
portions of the m rhonic cutout shectrum. (See graoh,
Fig. 4.2).

Jr m this date it tvs concluded that the
energy are fairly well distributed throughout the
soectrum.nith e fairly constant rate of decrease,
with increasing frequencies, of energy content in the
low and medium range of harmonics. The abrupt decrease
in the high portion of the sgectrum was attributed to
the effects of strey cap citanc and of field varic-
tions for these freguencies.

Vacuum tube Tl, : eye, is the oscill;tor tube.
The circuit incorporating this tube is essentially a
tuned-grid-tuned-plete oscillator circuit. The grid
circuit has as its freguency determining element a
crystal ground to oscillate at aggroximately lOOkc.

'he cutout of the oscillator is t ken from the induc-

H]

ive caiacititive combin:tion across the anode 5nd

.,

(4‘

ground. The capacitor is variable and grovides the
fine adjustment (of about 2:40 cycles) necessary to
set the oscillator at exactly 100 kc.

Vacuum tube V3, a bin , is the buffer ampli-

fier tube. The circuit incorporeting this is the

-55-
standard form of a resistxnce caoacitance coupled
amelifier. It serves to stabilise the output of the
oscillator circuit dieinst variations in loading and
circuit parameter variations.

Vacuum tube V3, a EbJ , is the transitron
oscillator tube. The circuit incorporating this tube
is the transitron oscillator section. This oscillator
is cageble of oscillating at each of the three regiired
sub-harmonics of lOOkc. Switch "1" determines which
combination of inductance and capacitance will be pre-
sent in the circuit and thus determines the frequency
of oscillation. It is seen that a synchronizing voltag
which synchronizes the oscillations of this tube to the
preper harmonic relationship with the loOkc oscillator
will be present on the control grid of V3.

Vacuum tube V4, a 6337, is an amplifier tube.
It serves to amolify its inout to the reguired output

level necessary to drive tube V5. The resistive—

\1

capacitative input network serves uS a voltage divider
which, at each of its four possible values, serves to
provide an inout of the preper level.

Vacuum tube V5, an 802, is the harmonic power
amplifier tube. The circuit incorporating the tube is
one of the many possible circuits which provides graded

amplification for the harnonically rich outyut. This

circuit attenuates the lower freguehcy components and

-35-

peaks the higher fr quency components thus providing

a reasonably conStunt enplitude for the harmonics of
the output vo ts e. The switch in the output circuit
ending in the jacks is in series with e capecitutive-
inductive voltage dividing network to provide a fairly
constant output inpedence. This is a high impedance
output end care should be taken in coupling to it es
improper matching may completely eliminate the hi her
hz'rr;:onics of the two lowest sub-herruonic fundamentals
by causing an r.f. snort circuit.

Vacuum tube vs, a 63J7, is the transitron
modulator oscillutor tube. The circuit consists of the
standard trsnsitron oscillator network whose output is
used to modulite the hzrmonic po er amplifier, V5,
with e ESOcps audio signal so :8 to make the harmonics
identifiable.

Vucuum tube V7, a bio , is the sinusoidal
amplifier tube. This tube and its essocisted circuit
provide amplification for the lowest frequencies and is
not designed to intentionally introduce harmonic dis-
tortion of uppreciable amplitude. The circuit is a
standard resistance cepucitence network and its output
is stable and may be used as a stsndard source of the

same stability as the master oscillator.

 

Aéu "”7 7/ 7* 4.57/05 7%
JW/LIIT’ 7715‘ me
72: [‘3 ¢32 . W
/ 7Z@€OILI-vfirpi€ Egyflx 4 By
,- “”7. WA: 1 (93’6/‘L/4mfl.
Me- 6;-
I/‘L/C” (91/ 71¢
727?
W
m

 

§
Qt
Qg§
t
‘éig
§
mg W
“UL, ‘
xvi};
2%: E
i! Q
Ngt \
k“ .\
3§§
Q
3.0“
Egg
SE.

ll/K/V 951/297; -
57m:

”—1 _

'I " 1"”"1 : ,
,HARflOA/YC ’l/dLIflQf-s‘r
Specrelim ‘ Maw/ax?

fa

 

 

1

¢-

; . _ .
W ...-——--"'-"’
‘fl.

as!

‘

‘0

..__-_._..,

ii:-
1

‘ f
Y

'50 A: 5 39 . zo__

149.47.534.15 1/9.” 5'

' ”zoo ' no 290 260

Ida

we [60-
l

g ; V
Ade/Cr

‘
‘4‘

A)? c} r

i, w
omfim
.1‘
. i T
. ‘ \J
o
‘o
\
o
o
o
\I
~0
a.

-59-

n “"',' ~w ”yr: - nvv h... —-.‘~—~ '.H1T~'n w? m “m... 9'
_)_A,'}_._._I..,I 3‘.“ .... D ‘ ) I- IJ-h~s ”\TJ—L-L 1-:‘..|

J1 u...._.-..L_.|.-.-§

(1')

k3

dis frequency generator will grovide very

strong lOOkc points up to and above lOmc at thich fwe-

quency a UTV stendgrd fregdency modulated carrier is
located. This sigm l and not the 5mc WLV Sig
occasionally have to he used in u'king any adjustments
that mny he necessary in correcting the fundemertel
i‘rewdency of the lOOkc oscillator es the Smc KEY sig-
nal has ceen found to suffer quite severely from fading.
Fig. 4.1 shows the results of the stcoility test and
it is evident that co: re ction should be o;1ly occrsionz1l-
ly neces,';dry.

The following proceddre is neces cry to zero
beat the lCCkc oscillator with e T'V sijnel:

1. Switch to "On" the switch labeled "Dower”.

2. Switch to "an" the switch labeled "flete".

50 Couple 8.
l

dlmay entegne from terminal "l"
01 the :1

r terninel panel to 9 receiver.

4. Adjust the receiver dgtil the desired KEV
signal is heard. (5mc or 1033)-

5. Incre se the anglitdde of modulation and
adfldst receiver dial until a lOOkc har—
monic is heard. It may be identified by
its 530 0.9.8. tone.

6. Adjust ti‘ie "ldOk crdgustment” screw until
the ploper dermonic is saperimposed on the

-40-
MTV carrier. Tnis mvy be done during the
quiet periods, when tgere is no audio
modulation present on the HIV carrier, by
tirning the switch 1 beled "Iodulntion"
to the "33?" gosition end zero heating the

J.

two unuodul.ted carriers.

7. The SOkc, 20kc, end lOkc corrections mry
be made in 3 similar LPMder.

8. To avoid frequent adjustment the switch
.1.

lnbeled "Boner" should be left in the "On"
DOS iti On.

Tn various methods of Specifically mefisuring
en unknown signal will be ibund in the Zunucl of Operat-
ing Instrictions.

Tni‘ t-nfierd snoild not be used for at le st

U‘
(0

one hour after tne gower has been turned on as this is
the period of greatest instibility. After tuis time the

.‘

standard has surficient stability for most applications.

031k.“ S‘Rflk NVWW

 
  

a

- Inlntrcluuntininutinuatl

h.

 

 

 

 

 

 

 

+4aoy .aoov/ .0060/

 

 

 

L

 

 

 

jl
II

 

 

 

 

 

«a

 

 

 

 

 

 

 

+400

 

 

 

 

 

 

 

 

 

 

 

 

L’D’IP

 

BIBLIOGRAPHY
1948

 

Clapp, J. K., Frequency.Measurement by Sliding Har-
monies, Proc. I. R. E., Vol. 56, No. 10, p. 1285,
Get.

Terlecki, R., and J. W. Whitehead, Two Portable Sub-
standards of Frequency, Jour. Sci. Instr., Vol. 25,
p. 237. July

Abbot, A. E., multivibrator Design by Graphic methods,
Electronics, Vol. 21, p. 118, June

Silver, 1.1., and A. Shadowitz, High-Ratio Multivibrato'r
Frequency Divider, Elec. Commun. (London), 701. 25,
p. 160, June

Feinberg, R., On the Performance of the Push-Pull Re-
laxation Oscillator (multivibrator), Phil. mag.,
Vol. 59, p. 268, Apr.

Davis, K. H., multivibrator Step-Down by Fractional
Ratios, Bell Lab. Rec., 701. 26, p. 114, march

Bertram, Sidney, The Degenerative Positive-Bias
multivibrator, Proc. I. R. E., Vol. 56, No. 2,
p. 277, Feb.

Phillips, F. C. F., A Direct Reading Frequency measur-
ing Set, Jour. I. R.E. (Brit. ), Vol. 8, p. 4, Jan. &
Fe 0

1945

Kiebert, martin 7., Jr., and Andrew F. Inglis, multi-
vibrator dircuits, Proc. I.R.E.. Vol. 53, no. 8,
P. 534, Aug.

1944

Lovering, W. F., A Push Pull Resistance-Capacity
Coupled Oscillator, London, Edinburgh and Dublin
PhilosOphical magazine and Journal of Science, 7th
Series, Vol. 55, No. 250, p. 715, Nov.

Greenidge, R..M. 0., The MOunting and Fabrication of
Plated Quartz Crystal Units, Bell Sys. Tech. Jour.,
Vol. 25, p. 254

Sykes, R. A., Principles of Mounting Quartz Plates,
BOSOTOJO. V01. 23, P. 178

Sykes, R. A., Modes onMotion in Quartz Crystals, the
Effects of Coupling and Methods of Design, B.S.T.J.,
V01. 23’ p. 52

Meson, W. P., and R. A. Sykes, Low-Frequency Quartz
Crystal Cuts Having Low Temperature Coefficients,
Proc. I.R.E., Vol. 32, p. 208

Bokovoy, J. A., Quartz Crystals-Development & Appli-
cation, Elec. Comm., Vol. 21, p. 233

1943

 

Druesne, M5 A. A., Quartz Crystals, Communications,
V01. 23. D. 46f. Sept.

mason, W. P., Quartz Crystal Applications, B.S.T.J.,
Vol. 22, p. 178

1942

 

Bedford, A. V., and G. L. Fredendall, Analysis, Syn-
thesis and Evaluation of the Transient Response of
Television Apparatus, Proc. I.R.E., p. 440, Oct.

1941

 

Booth, C. F., The Application and Use of Quartz Cry-
stals in Telecommunications, Jour. I.E.E., Vol. 88,
Part III, p. 97, June

Bartelink, E. H. B., A Wide-band Square-wave Generator,
Trans. A.I.E.E., Vol. 60, p. 371

1940

 

Mason, W. P., A New Quartz Crystal Plate, Designated
the GT, Which Produces a Very Constant Frequency over
a Wide Temperature Range, Proc. I.R.E., Vol. 28,
p. 220, may

Roberts, Walter Van P., The Limits of Inherent Fre-
quency Stability, R.C.A. Rev., 701. 4, p. 478, April

mason, W. P., Low Temperature Coefficient Quartz Cry-
8138.15, BOSOTOJO, V01. 19, p. 74’ Jan.

-44-

Baldwin, C. F., Quartz Crystals, G. E. Review, Vol. 43,
p. 188

Chakravarti, S. P., On the Nature of Negative Resis-
tance and Negative Resistance Sections, Phil. mag.,
Vol. 30, p. 294

1939

 

Herr, Donald L., Oscillations in Certain Non-Linear
Driven Systems, Proc. I.R.E., Vol. 27, p. 396, June

Lampkin, G. F., An Improvement in Constant Frequency
Oscillators, Proc. I.R.E., Vol. 27, p. 199, march

Brunetti, Cledo, The Transitron Oscillator, Proc. I.R.E.,
v01. 27. NO. 1' p. 88

1958

Meachem, L. A., The Bridge Stabilized Oscillator,
Proc. I.R.E., Vol. 26, p. 1278, Oct.

Anderson, J. E., Frequency Characteristics of Piezo-
electric Oscillators, Electronics, p. 22, Aug.

Stevenson, G. H., Stabilized Feedback Oscillators,
BOSOTOJO, v01. 17, p. 458. July

1937

 

Seeley, S. W., and C. N. Kimball, Analysis and Design
of Video Amplifiers, R.C.A. Review, p. 171, Oct.

Sabaroff, S. A., A Voltage Stabilized High Frequency
Crystal Oscillator Circuit, Proc. I.R.E., Vol. 25,
p. 623

Right, S. C., and G. W. Willard, A Simplified Circuit
for Frequency Substandards Employing a New Type of
Low-frequency Zero-temperature Coefficient Quartz
Crystal, Proc. I.R.E., Vol. 25, p. 549, may

Lamb, J. J., A Practical Survey of Pentode and Beam
Tube Crystal Oscillators for Fundamental and Second
Harmonie Output, QST, Vol. 21, p. 31, Apr.

Harnwell, G. P., & J. B. H. Kuper, Laboratory Fre-
quency Standard, The Review of Scientific Instru-
ments, Vol. 8, march, p. 83

.—

-43-

A

Right, S. C., QuartzPlates for Frequency Sub-
Standards, Bell Labs. Rec., Vol. 16, p. 21

1936

 

Diehl, W. F., A New Piezo-electric Quartz Crystal Hold-
er with Thermal Compensation, R.C.A. Rev., Vol. 1,
P0 86, OOtO

VanDyke, Karl S., A Determination of Some of the PrOp-
erties of the Piezoelectric Quartz Resonator, Proc.
IOROEO’ v01. 23, p. 386, Apr.

KOga, Isaac, Notes on Piezoelectric Quartz Crystals,
Proc. I.R.E., Vol. 24, p. 510, march

leCorbeiller, Ph., The Non-linear Theory of the main-
tenance of Oscillations, Jour. I.E.E., Vol. 79,
p. 361

1935

Gager, F. M., and J. B. Russell, Jr., A Quantitative
Study of the Dynatron, Proc. I.R.E., Vol. 23, No.
12, p. 1536, D800

mason, W. P., An Electromechanical Representation of
a Piezo-electric Crystal Used as a Transducer, Proc.
IOROEO, V01. 23, p. 1252, Get.

Herold, E. W., Negative Resistance and Devices for
Obtaining It, Proc. I.R.E., Vol. 23, p. 1201, Oct.

Gager, F. malcolm, The Grid-Coupled Dynatron, Proc.
IOROEO’ V01. 23, N0. 9, P. 1048, Sept.

Baldwin, C. F., and S. A. Bokovoy, Practical Operating
Advantages of Low Temperature Frequency Coefficient
Crystals, QST, V01. 19, p. 26, Jan.

Bechmann, R., Researches on Natural Elastic Vibrations
of Piezo-Electrically Excited Quartz Plates, Zeit.
f. Technisch Physick, Vol. 16, No. 12, p. 525

1934

Lack, F. R., G. W. Willard, and I. E. Fair, Some Im-
provements in Quartz Crystals Circuit Elements,
BOSOTOJO. V01. 13, p. 453, July

mason, W. P., Electrical Wave Filters Employing Quartz
Crystals as Elements, B.S.T.J., V01. 13, p. 405,

July

-46-

Hayasi, Tatuo, The Inner-Grid Dynatron and Duodynatron,
PrOCo I.R.E., V01. 22, N0. 6, p. 751, June

Meahl, M. R., Quartz Crystal Controlled Oscillator
Circuits, Proc. I.R.E., Vol. 22, p. 732

1933

 

Williams, N. R., mcdes of Vibration of Piezoelectric
Crystal, Proc. I.R.E., Vol. 21, p. 990, July

Groszkowski, J., The Interdependence of Frequency
Variation and Harmonic Content and the Problem of
Constant Frequency Oscillators, Proc. I.R.E., Vol.
21, N00 7, p. 958, July

Lamb, J. J., A mere Stable Crystal Oscillator of High
Harmonic Output, QST, Vol. 17, p. 30, June

ScrOggie, M. G., Applications of the Dynatron, Wireless
Eng. and Eng. Wireless, Vol. 10, p. 529

Colebrook, F..M., Voltage amplification with high
selectivity by means of the dynatron circuit, Wire-
less Eng. and Eng. Wireless, Vol. 10, p. 69

1932

 

Mackinnon, K. A., Crystal Control Applied to the Dyna-
tron Oscillator, Proc. I.R.E., Vol. 20, p. 1689

Peterson, H. 0., and A. M. Braaton, The Precision
Frequency Measuring System of R.C.A. Communications,
Ino.,.Proo. I.R.E., V01. 20, N00‘6, p. 941, June

Colwell, R. C., The Vibrations of Quartz Plates, Proc.
I.R.E., V01. 20, N0. 5, Do 808. may

Hougaard, 0. M., Application of Quartz Plates to Radio
Transmitters, Proc. I.R.E., V01. 20, p. 767. may

Kusunese, Y., and S. Ishikawa, Frequency Stabilization
of dadio Transmitters, Proc. I.R.E., Vol. 20, p. 310,
FGbo

Heaton, Vincent E., and E. G. Lapham, Quartz Plate
Mountings and Temperature Control for Piezo Oscil-
lators, Proc. I.R.E., V01. 20, p. 261, Feb.

1931

 

Llewellyn, F. B., Constant Frequency Oscillators, Proc.
I.R.E., V01. 19, p. 2063, Dec.

-47-

Dingley, Edward N., Jr., DevelOpment of a Circuit for
Measuring the Negative Resistance of Pliodynatrons,
Proc. I.R.E., V01. 19, N00 11, P0 1948, NOV.

Andrew, Victor J., The Adjustment of the multivibrator
for Frequency Division, Proc. I.R.E., Vol. 19, No. 11,
p0 1911’ NOV.

Boella, M., Performance of Piezo Oscillators and the
Influence of Decrement of Quartz on the Frequency of
Oscillation, Proc. I.R.E., Vol. 19, No. 7, p. 1252,

July

Kaga, I., Note on the Piezoelectric Quartz Oscillating
Crystal Regarded from the Principle of Similitude,
Proc. I. R.E., Vol. 19, p. 1022

Wheeler, L. P., An Analysis of a Piezoelectric Oscil-
lator Circuit, Proc. I.R.E., Vol. 19, p. 627

MOgel, Hans, Some methods of measuring the Frequency of
Short Waves, Proc. I.R.E., Vol. 19, No. 2, p. 195,
Feb.

Colebrook, F. N., The dynatron oscillator, Wireless Eng.
and Eng. Wireless, Vol. 8, p. 581

1930

 

Clapp, James K., Temperature Control for Frequency
Standards, Proc. I.R.E., Vol. 18, p. 2003, Dec.

Koga, I., Characteristics of Piezoelectric Quartz
Oscillators, Proc. I.R.E., Vol. 18, p. 1935

Clapp, J. K., Interpolation methods for Use with Har-
monic Frequency Standards, Proc. I.R.E., Vol. 18,
p. 1575, Sept.

Keaton, V. E., and W. H. Brattain, Design of a Portable
Temperature Controlled Piezo Oscillator, Proc. I.R.E.,
v01. 18. N0. 7’ July .

Watanabe, Yasusi, Some Remarks on the multivibrator,
Proc. I.R.E., V01. 18, N00 2, FBbo

Harrison, J. R., Push-Pull Piezoelectric Oscillator
Circuits, £7000 I.R.E., V01. 18, p. 95

-48-

1929

 

Jimbo, Seikichi, measurement of Frequency, Proc. I.R.E.,
A Vol. 27, No. 11, p. 2011, Nov.

Lack, F. R., Observations on MOdes of Vibration and
Temperature Coefficient of Quartz Crystal Plates,
Proc. I.R.E., V01. 17, P. 1123, July

Harrison, W. A., A High Precision Standard of Frequency,
Proc. I.R.E., V01. 17’ p. 1103’ July

Hall, E. L., A System for Frequency measurements Based
on a Single Frequency, Proc. I.R.E., V01. 17, N00 2.
p. 272, Feb.

Hull, L.Ms, and J. K. Clapp, A Convenient method for
Referring Secondary Frequency Standards to a Standard
Time Interval, Proc. I.R.E., Vol. 17, p. 252, Feb.

1928

 

Hund, August, Notes on Quartz Plates, Air Gaps Effect,
and Radio-frequency Generation, Proc. I.R.E., Vol.
16’ P. 1072’ Aug.

Wheeler, L. P., and W. E. Bower, A New Type of Stand-
ard Frequency Piezoelectric Oscillator, Proc. I.R.E.,
v01. 16’ Do 1035

marrison, W. A., Thermostat Design for Frequency
Standard, Proc. I.R.E., Vol. 16, p. 976, July

Worrall, Rabert R., and Raymond B. Owens, The Navy's
Primary Frequency Standards, Proc. I.R.E., Vol. 16,
No. 6, p. 778, June

VanDyke, K. S., The Piezo-electric Resonator and Its
Equivalent Network, Proc. I.R.E., Vol. 16, p. 742,
June

Cady, W. G., BibliOgraphy on Piezoelectricity, Proc.
I.R.E., v01. 16’ PO 521’ NO. 4’ April

Crossley, A., Mbdes of Vibration in Piezo-Electric
Crystals, Proc. I.R.E., Vol. 16, No. 4, p. 416,
April

Horton, J. W., and W. A. Harrison, Precision Determina-
tion of Frequency, Proc. I.R.E., Vol. 16, p. 137

-49..

1927

 

Harrison, J. R., Piezo-Electric Resonance and Oscilla-
tory Phenomena with Flexural Vibrations in Quartz
Plates, Proc. I.R.Eo. Vol. 15, NO. 12, p. 104:0, D600

Hitchcock, R. 0., Mounting Quartz Oscillator Crystals,
Proc. I.R.E., Vol. 15, p. 802, Nov.

Hund, A., Notes on Piezoelectric Generators with Small
Back Action, Proc. I.R.E., Vol. 15, No. 8, Aug.

Strock ML 8., Standard Frequency Dissemination. Proc.
I.R.E., Vol. 15, No. 8, p. 727, Aug.

Royden, George T., The Frequency Checking Station at
Mare Island, Proc. I.R.E., V01. 15, N00 4:. p. 313,
Apr.

meissner, A., Piezo-Electric Crystals at Radio Fre-
quencies, Proc. I.R.E., Vol. 15, No. 4, p. 281, Apr.

1926

 

Dye, D. W., Piezo Electric Quartz Resonator and Its
Equivalent Electrical Circuit, Proc. Phys. Sec.
(London), Vol. 38, p. 399

Terry, E. M., The Dependence of the Frequency of Quartz
Piezoelectric Oscillators upon Circuit Constants,
Proc. I.R.E., Vol. 16, p. 1486

Hund, A., Uses and Possibilities of Piezoelectric
0801llat0r8, Proc. I.R.E., v01. 14. NO. 4. p. 447’
Aug.

van der Pol, Balth, On Relaxation Oscillations, Phil.
Mag., V01. 2, p. 978, NOV.

1925

 

Hund, A., A.thhod of measuring Radio Frequency by
means of a Harmonic Generation, Proc. I.R.E., Vol.
13. NO. 2’ Apr.

1924

Cady, W. G., An International Comparison of Radio Wave-
length Standards by Means of Piezo-Electric Resonators,
Proc. I.R.E., Vol. 12, No. 6, p. 805, Dec.

.‘I

1“

II

I.

-50-

van der P01, 3., Non Linear Theory of Electric Oscil-
lations, Proc. I.R.E., v01. 22, P0 1051

1923

 

Dunmore, Francis W., and Francis H. Engel, a method of
Measuring Very Short Radio Wave Lengths and Their
Use in Frequency Standardization, Proc. I.R.E.,

V01. 11, N00 5, p. 467, Get.

Pierce, G. W., Piezo-electric Crystal ResOnator and‘
Crystal Oscillators Applied to the Precision Cali—
bration of Wavemeters, Proc. Am. Acad., Vol. 59,
p. 81

1922

 

Cady, W. G., The Piezo-electric Resonator, Proc. I.R.E.,
v01. 10, NO. 2, p. 83

1919

 

Abraham, H., and E. Block, Mesure en valeur absoluéj
des periods oscillations electriques de haute fre-
quence, Annal d. Phys., Vol. 12, p. 237, Sept.-Oct.

1918

 

Cordes, H. G., Theory of Free and Sustained Oscilla-
tions, Proc. I.R.E., Vol. 6, No. 3, p. 167, June

1916

 

Kennelly, A. E., The Impendances, Angular Velocities,
and Frequencies of Oscillating Current Circuits,
PrOOQ I.R.E., Vol. 4, N0. 1, p. 47, Feb.

 

 

 

 

R

 

”'Tl'iifixfiifltfilfliflWflfiflflflifi\iiMLIWI\Fs