ELECTRICAL COMMUNICATIONS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-01043R002000220002-9
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 137
• Document Creation Date: December 23, 2016
• Document Release Date: June 3, 2013
• Sequence Number: 2
• Publication Date: January 1, 1957
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
? Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Peinrich t;ert,,;, On the Cen;enar ? of his :3irLI? (l8;:.. - lo.";i),
b?? G.: I,evin .............................................
.1 .;chemat.ic for I ecention o'' ;igna:~.s, b 1.A.l.i:zrkevi.ch..........
improvement o:' the i?.ethods of 'c?easuring :;onl inear Distortions,
b:? :. I,. Bezladnov......... ................................. 1!.
,he .,t.ate of i?olor Television :+broad, b:' A.h..I ustar:?ev........... 31
'.ethnical Calculations of I?:rrors of ,1i.tenuat.or?s, b:? D.D.Voe:Tkuv..
Pulse Distortions in Voice-Freglienc- Telegraph Channels with
?
Phase I;odulati on During the Ac Lion of Interference, b:
A.i:.:?inFerenko ........................................... 52
Problems of Computing Circuits of T-Filters with Overlap, b:?
Z.I.I'etr?ov ............................................... 67
Use of ::oncont.act Elements in Dial Office Control Circuits, b
V.Id.I,oginski;????????????????????????????????????......?.? 73
Balancing Circuits of Common Batter:? Telephone Sets with Semni-
iconductor Amplifiers, b:-.'1.S.Sadovskiy ..................? 90
Calculaition of Active Resistances of Conductors of 't'ubular
Form, by G.P.Delektorski:.?...............................? 105
t
From Foreign Journals, Brief Reports, New Zlectronic Tubes and
113
,Semiconductor Devices....................................
I-Jew American Mass-Produced TV Jet ................................ 118
Author Certificates ...............'............................... 126
ForeignIPatents.........?........................................ 128
State Publishing House for Engineering. Brief report of 1957
;Publishing Program of Technical Literature ............... 13/i
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
HEINRICH HERTZ
On the Centenary of his Birth
(7857 - 1957)
by
G. A. Levin
A hundred years have passed since the birth of Heinrich Rudolf Hertz.
"Heinrich Hertz" - was the content of the first radio message in the history of
mankind, transmitted by A.S.Popov in 1895.
The name of Hertz was used for denoting the unit of oscillation frequency, the
most important characteristic of any oscillatory processes.
For.what did Heinrich Hertz merit his immortality,?
In the Eighteen Seventies of the last century, advocates and opponents of the
theory of action-at-a-distance in the science of electricity and magnetism were en-
gaged in stubborn controversy.
On the theoretical plane, Jame's Clark Maxwell struck the most serious blow at
the remote action theory, having created his own amazingly orderly theory of electro-
magnetic phenomena, from which it followed of necessity that no kind of action-at-a-
distance exists and that electromagnetic energy is propagated in wave form with a
velocity equal to the velocity of light. However, so long as the new theory was not
confirmed experimentally, abundant opportunities remained for every kind of skeptical
judgements about Maxwell's work.
Maxwell himself did not make such an experimental verification. To do this fell
to the lot of Heinrich Hertz.
By his famous experiments in 1888 Hertz proved indisputably the existence of
electromagnetic waves with finite velocity of propagation in space. It was further
found that this velocity was equal to the velocity of light and that in general elec-
1 STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
?
0
?
?
tromagnetic waves are, in their properties, almost completely identical to light
waves. They are reflected from conductors, are refracted during transition from one
dielectric to another, are polarized, and are capable to produce interference
phenomena.
The experimental results of Hertz became instantly sensational. The idealistic
theory of action-at-a-distance suffered a crushing defeat. Simultaneously, the elec-
tromagnetic nature of light became obvious. The services of Hertz to radio technol-
ogy are, however, not exhausted by this.
For his experiments, Hertz used equipment which in a large measure anticipated
the apparatus of the simplest radio stations.
This equipment comprised an open oscillatory loop, the prototype of future an-
tennas; a generator of damped oscillations of high frequency, which greatly resembles
the oscillator of future spark-radio stations; a receiving device in which a spark
served as the indicator of the presence of oscillations.
Thus, without setting for himself the task of accomplishing radio communication,
and evidently not even suspecting such a possibility, Hertz put into our hands some-
thing very much like a radio station, in which the primitive open loop was to be re-
placed by a real antenna, and a more modern indicator of the presence of oscillations
was to be employed in the receiver instead of a spark.
The greatest service rendered by A.S.Popov in this respect was that he was first,
to perceive how close Hertz, without himself being aware of the fact, was to a solu-
tion of the problem of radio communication and along what path the research by Hertz
should be continued in order to realize radio communication.
In his experiments, Hertz made extensive use of the phenomenon of resonance, so
characteristic for all radio technology. Hertz was a great theorizer. He is the
originator of the theory of the so-called Hertz dipole, in which the problem of emis-
sion of radio waves is solved by an open oscillator of the simplest form. To the
present time, this theory is the basis on which any theory of the emission of radio
STAT-
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
waves b: r antennas is built.
From the modern viewpoint, Hertz's theory of emission should be close-J., examined
also as to possible linking of the emission phenorre;:on with the behavior of electrons
t ? in the atom. The state of physics in Hertz's times, however, excluded the possibil-
140
of such development of the emission theory by Hertz himself. This proved possi-
ble only at a lat.er time, on the basis of quantum mechanics. _
Hertz was the first in the history of physics to investigate the photoelectric
effect. In the last years of his life, Hertz studied cathode rays. Early death -
Hertz died in 189/l at the age of 37 years - cut short the fruitful scientific activ-
ity of this gifted man.
In its development, radio technology has repeatedly deviated greatly from the
ideas of Hertz. The wave band which Hertz used for his experiments (ultrashort and
decimeter waves) proved of little use for radio communications over long distances.
At first, therefore, the radio technology of long waves saw rapid development.
The invention of the cathode-ray tube produced an opportunity to solve many such
problems of which Hertz could not even dream. The development of radio broadcasting
entailed full mastery of the medium-wave band. Extreme-distance radio communication
was reliably realized by the use of short waves, with their special mechanism of
propagation.
In recent decades, however, radio technology has again returned to the iiertzian
wave band. The development of television, radar, and radio-relay lines of communica-
tion required utilization of ultrashort decimeter and centimeter waves.
The transmission and reception equipment for waves of superhigh frequencies, of
course, has so far developed in the meantime in comparison with the equipment of
Hertz, that there cannot be even the remotest quantitative or qualitative comparison.
Certain elements of Hertzian equipment, however, have been retained to the present
time. These include the Hertz doublet itself, parabolic reflectors, and so forth.
?
The point is not at all that contemporary radio technology has, in~;Icertain re-
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
spects, gone back to hIertz.
Important is the tremendous impetus which the work of hertz as a whole gave
later generations of scientists and mainly to A.S.Popov, the creator of radio tech-
nology of wireless communication. From here began the rapid scientific progress
which led to present-day radio technology with its astonishing achievements.
Hertz is indisputably one of the classic leaders of physics in the field of the
science of electromagnetic oscillations. In this consists his great role in the his-
tory of science.
In the days of celebrating the centenary of the birth of Heinrich Rudolf Hertz,
we remember with gratitude the name of the great scientist and render due respects to
his remarkable creative work in science.
STAT
4
? Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
1
A SCHEMATIC FOR RECEPTION OF SIGNALS
by
A.A.Kharkevich
One of the possibilities of ?total?f reception of a binary
signal is discussed, with a description of the basic diagram of
a device which accomplishes this possibility. The reception is
done visually, by observation of the screen of a cathode-ray tube.
To realize the possibilities offered by interference-killing codes, it is neces-?
sary to receive the signal (in the form'of some code combination) as a whole, i.e.,
the entire combination in toto, and then to collate it with a great number of trans-
mitted signals and identify it with those from which the received signal differs
least. This permits detecting and correcting of errors.
From the viewpoint of the geometric theory, any signal is a point in space of a
corresponding number of measurements. The superimposition of interference-displaces
this point. To avoid errors, i.e., to identify the received signal not with the ac-
tually transmitted signal but with other possible signals, it is necessary to in-
crease the distance between points that represent possible transmitted signals.
The technology of receiving the signal in its "entirety?? must in general consist
of the following: 1) the. received signal is remembered; 2) the signal is checked
with all possible transmitted signals which must be known in the reception side of
the system of communications and must be stored in a certain memory device; 3) of the
possible transmitted signals the one from which the received signal differs least is
recorded - this signal is also considered as the transmitted signal (the expression
"differs least" determines the operating system of the receiver).
The general wiring diagram of a receiver of ?!total" signals is pictured in the
following manner: There is a memory in the form, for example, ofmagnetic recording,
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
in which all possible transmitted signals are recorded in advance; there also is the
possibility of separately recording the received signal. After that the received
signal is reproduced many times; simultaneously, each time one of the possible trans-
mitted signals is reproduced.
The comparison can be done in various ways; either by subtracting, squaring the
difference, and summarizing according to the formula
where y(t) is the received signal;
xk(t) is one of the possible transmitted signals;
and the transmitted signal is determined by the minimum dk;
or by multiplying the signals and then summarizing according to the formula
and finally determining the transmitted signal according to the maximum R. Both var-
iants of the method of checking are theoretically equivalent*. The quantity dk in
transmitted signal. The quantity R is'the:.factor of cross correlation between the
received signal and any possible transmitted signal. 'Equations (1) and (2) refer to
the geometric sense is the distance between. the received signal and any possible
continuous signals; T designates the duration of the signals. For discrete signals,
r r
eqs.(l) and (2) are replaced by corresponding sums:
*Forsignals of equal energy both variants simply coincide.
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
M
where yi, xik are the discrete values ("signs", s:znbols), respectively, of the re-
ceived signal and any of the possible transmitted signals, while n is the number of
signs.
The above-described system has a sufficiently universal character and is applic-
able for any signals. But the resulting technical solution is rather cumbersome. In
addition, the operation of checking and selection requires time; this time increases
rapidly with an increase in the dura.ion of signals, since an increase in the dura-
tion (or the number of signs, in the case of discrete signals) causes the total ntun-
ber of possible signals to increase according to an exponential law. The necessary
capacity of the memory also grows correspondingly. Meanwhile the advantages of
interference-killing codes are realized precisely for large segments of signal. The
most advantageous correlations are derived in the range when T (or n) tends toward
infinity.
Consequently, such a universal system is not of such interest as a simpler sys-
tem for reception of t"total" signals. We will describe this circuit applicably to
the reception of discrete and, in particular, to binary signals.
The idea consists of transfering the multidimensional image of a set of trans-
signals to a plane, i.e., to a space of two dimensions. This permits using an
mitted
ordinary cathode-ray tube as the terminal link of the receiving device; the set of
transmitted signals is presented as a system of points arranged in a certain plane
Any n-digit binary number of the total quantity
?
AI =a? N,
1
(4)
STAT
N = 2'
(3)
can be replaced by a two-digit number according to a number system with the base a.
For complete mapping of the set N, it is necessary.only to satisfy the condition
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
from where also a is determined. On the basis of egs.(3) and (1,) we rind
For the start, we will analyze the case of n = 3, a = 3. The three-digit binary
signal is geometricall;r represented by the vertices of a cube; the set of transmitted
signals contains I-1 = 23 - 8 combinations.
Each combination is expressed by a three-digit binary number. On the other
hand, these binary numbers can be expressed by two-digit ternary numbers; one is
found in excess, since
Thus, we have the following one-to-one correspondence:
Number of Combinations
vertices of cube)
Binary Recording
Ternary Recording
Pictured in Fig.la is a cube with correspondingly numbered vertices, and in
Fig.lb, a two-dimensional table for ternary recording.
/00 0
fit 1
001 2
Fig.l Fig.2
Let us assume that the binary number 101 is converted to the ternary number 12
(problems of the technology are discussed below). We consider this number as a point
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 on a plane with the coordinates x = 1, y = 2. This point coincides with the lattice
cell for No-5 and depicts the vertex No-5 of the three-dimensional cube (with the
coordinates x = 1, y = 0, z = 1).'
It is clear that, if during the transmission of the binary signal a single error
occurred (i.e. an error in one sign), then the received combination coincides with
one of the other possible combinations. For example, a single error may convert sig-
nal No.5 into one of three others, according to the diagram in Fig.2a, which corre-
sponds to the situation marked by points in Fig.2b. Consequently, an error under
such conditions cannot be corrected or detected.
In order to detect a single error, it is obviously sufficient to select binary
011
101
Fig-3
010
100
cube vertices Nos.0,3,5,6 (Fig-3a). A single error transforms the allowed combina-
combinations so that they differ by not less than two signs. In a geometric model in
the form of a three-dimensional cube, this corresponds to a selection of vertices
spaced at a distance of two edges (i.e. lying on diagonals of the faces). This means
that we use only half of all possible combinations; the remaining combinations are
forbidden. Let the combinations 000, 011, 101, 110 be allowed, corresponding to the
- tions into forbidden ones according to the diagram in Fig-3b. In the two-dimensional
Table in Fig-3c, the sites of the forbidden combinations are hatched; the incidence
of the point of the received signal on one of the hatched cells indicates an error,
9 -.translates any allowed combination into another allowed one.
-which is thus automatically detected. A binary error cannot be detected-since it STAT
In order that a single error can be corrected, the allowed combinations must
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
M
oot repeated in other cells to which the
010 '
off OQ 0 0 signals might be-transmitted as a
100
fit rot result of a single error. Thus, a
Q7
7
1to single error is corrected.
a) b) C~ t
?
differ in not less than three signs. We can construct only two three-digit binary
signals that satisfy this condition. The corresponding points are lying in the diag-
onals of the cube (for example, 000 and 111, as shown in,Fig.4a). The transition
circuit in a single error is shown in Fig-4b. Thus each of the forbidden combina-
tions formed as a result of a single mistake, is connected only with one of the two
allowed combinations. Therefore, the two-dimensional Table assumes the' form of
Fig.hc. The base positions (in the absence of an error) of signals Nos.0 and 7 are
marked by circles; these numbers are
We are intereste ,:however, in
Fig-4 ~
a larger set of signals requiring
combinations with a great number of signs during binary recording. As an example,,
let us take the telegraph code used for the transmission of 32 letters. The ordinary
code is five-digit (Baudot code), but we at once construct the simplest code detect=
ing a single error, after supplementing the five-digit combination by another 0 or 1,
with a calculation such that the derived six-digit combinations have-an even number
of units (or zeros). For conversion of the six-dimensional space to a plane we em-
ploy an octonary number system. Besides,.
N= 26=A9=82=6?f.
In the Table below are given the letters, their binary recording in the form of
?
six-digit combinations developed from the ordinary Baudot code, and the corresponding
two-digit octonary recording.
The flat Table corresponding to the last line has 8 x 8 squares like an ordinary
chess board. Half of the squares are occupied by letters, while the remaining half
lb
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
corresponds to forbidden combinations (with an uneven number of zero3 or units);
these squares are hatched (see Fig-5). In a similar manner, the binary code which
Letter
Binary Recording
Octonary Recording
C I D I E I F'
100001 001100 011011 010100 111100 010001
11 I 11 :3'i 21 71 21
corrects the single error can also be constructed. This will be a nine-digit code
(Bibl.l); a Table of 23 x 23 = 529 squares will serve as its two-dimensional mapping.
The basic diagram shown in Fig.6 is useful for a technical realization of the
reception device, with reflection of signals in a
0 1 2 3 4 5 6 7
Y plane. The binary signal, received together with
19 the superposed interference, enters the quantizing
1
- 2
S
device Kv which directs the received signal to one
777~ 777
4
of two levels and determines what, actually, is
6 received: 0 or 1. Here also occurs a possible
Fig.5
error whose probability is estimated in the usual
manner if the probability distribution of the in-
terference is known. From the quantizing device,
the binary signal goes to the coincidence circuit CC; here also enter the pulses from
the pulse generator IG, operating in synchronism with the signal (which can be
started with starting pulses). At the output of CC a binary signal of standard level
is obtained. This signal goes to one of two decoding devices Dl or D2, depending on
the position of the switch K. This latter is controlled from the pulse counter Co.
- The action of the system consists in that, for example, in the six-digit code
the first set of three binary signs determines the first figure of the two-digit num-
ber. When this has been determined, the switch is reset, and the next set of three
binary ,signs, arriving at the second decoding device, determine the second figure of
STAT
11
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
TuiIiIi
1-- ~ Al I FM-474
MMW
0
the two-digit number. The decoding devices are conventional decoding circuits KB4,
containing an RC segment with a time constant, selected so that during the cycle
period the potential in the capacitor is smoothly cut in two. This principle is
Fig.6
?
?
?
a) Signal plus interference; b) Quantizing device Kv; c) Coincidence
circuit CC; d) Switch K; e) Decoder D1; f) Storage cell N1; g) De-
coder D2; h) Storage cell 112; i) Pulse generator IG; J) Pulse
counter Co; k) Dropping circuit Cd
widely known. At the output of the decoding devices the storage cells N are con-
nected, which maintain the definitive value of the potentials resulting at the end
of the operating cycle of the decoding devices. From the storage cells the voltage
is fed directly to two pairs of deflecting plates. By means of the dropping cir-
cuit Cd, the voltage is taken from the storage cells at the required instant which
latter is likewise set by the pulse counter. The screen of the tube is superposed
by a transparency in the form of a grid (like Fig.5).
The above description is a tentative outline of an apparatus conceived primari-
ly as a demonstration unit, but also useful for certain investigations. The future.
will show whether these or other elements of a similar device can find application
in the technology of communications.
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
BIBLIOGRAPHY
1. Hanani.ng,R.ll. - Error Detecting and Error Correcting Codes. Bell Syst. Tech. J.,
Vol.29, No.2 (1950)
Article received by the Editors 19 June 1956.
STAT
0
13
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
II4PROVEMIJT OF THE METHODS OF MEASURING NONLINEAR DISTORTIONS
by
N.L.Bezladnov
Means are examined for obtaining the greatest correspondence
between the results of measurement and the degree of subjective
perception of nonlinear distortion. For this purpose,'measure-
ments are recommended of relative distortion levels realized dur-
ing the reproduction of a sound transmission or when using test
spectra which simulate the transmission. It is proposed that,
during the measurements, corrections in the distortion volume be
taken into account and also their possible masking by signals of
the transmission and noises in the reproduction rooms. The prin-
ciples of constructing the corresponding measuring circuits are
examined.
Introduction
At present, measurements of the nonlinearity of the reproducing circuit and
measurements of nonlinear distortions introduced by it are in use.
One of the main shortcomings of the existing methods of measuring nonlinear dis-
tortions in sound-reproduction devices is the discrepancy of the results of measuring
the-degree of subjective perception of distortions. This shortcoming in fact lessens
the value of the measurement. To overcome it, the principles used, as basis for dis-
tortion measurement must be established, proceeding from factors which determine the
subjective perception of distortions. Such factors are chiefly the distortion volume
? and their masking by signals of the transmission and also by noises in the rooms
where the reproduction is made. In turn, these factors are determined by the charac-
ter of the nonlinearity of the devices being examined and also by the character of
111
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
the sound transmission beinr; reproduced (distribution of the d??namic levels of trans-
mission in time and according to frequency).
T t. should be especially emphasized that, inasmuch as the point under discussion
? t ' f the re mollt1c i nfT devices
-
?
is
?
?
p(co)dw 1 u2(w)d,,
15
teris.~cs o l
is not an investigation of the nonlinear charac
but. o. the nonlinear distortions introduced by their., the character of the reproduced
transmission acquires decisive iriportance, and failure to take thin r i.rcumstar c
account. inevitably, leads to substantial errors.
Basic Principles of Measurement of Nonlinear Distortions
In accordance with the foregoing, the following principles must be used as basis
of measurements:
1. The measurements must be made in the process of reproducing the sound trans-
mission (dynamic conditions) and during the admission, at the input of the device un-
der stud;', of specially selected test spectra which simulate definite fragments of
sound transmission. For purposes of simplifying the measurements, discrete frequency
spectra can be used, which correspond to periodic oscillations of constant magnitude
(static conditions).
2. For a numerical characteristic of the distortions, it is expedient to meas-
ure the relative level of the distortion volume. For this, it is sufficient to meas-
ure the relative level of the power of nonlinear distortions and to introduce correc-
tions in the curves of equal volume, carried out in accordance with the normal level
of reproduction (for example) applicably to reproductions with a maximum loudness
level +90 decibels). The expressions for the relative levels of the distortion power
can be represented in the following form:
uli2 '12
S Pi (W) d,: J ttj(?,)dw
I -- lUl -- =20Ig ---
~tl g ,n2 wp
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
? R=Wig 20i'
? !' EU2
Here l'd is the dynamic relative level of the distortion power;
?I' is the static relative level of the distortion power;
f'i, III are the effective values of the power and voltage of the products of
nonlinear distortions;
pi(.), ui(?) are the continuous frequency spectra of the power and voltage of the
products of nonlinear distortions;
p(t)),
u(.,)
,,ill i2 are the edges of the spectra of distortions;
01, ?'2 are the edges of the spectra of fundamental oscillations.
The transformation to the expressions for the dynamic and static level of the
distortion volume R' and lit can be made by the introduction, respectively, of the
bi(0 ) b(o )
correction functions mi(,'') = 10 10 , m(O) 10 10 or of the correction fac-
bi b
tors mi = 10 10 m = 10 10 . here bi(o), b(~~), bi, b are the differences between
the intensity levels and the volume levels for corresponding frequencies (determined
according to curves of equal volume). Thus the dynamic relative level of the distor-
tion volume is
?
is
P In (w) do)
16
1'' U }is the same for the fundamental oscillations;
1 p, ((0) wt, (co) dco
R,, _ I 0 Ig
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
while the static relative level of the distortion volume is
r
It is evident from the foregoing that, in the measuring circuit, the change-over
to the levels of distortion volume can be made b;, supplementing the zones of distor-
tions and transmission with attenuations depending on the power levels in these zones
and on the frequency.
3. In view of the fact that, in the process of reproducing a sound transmis-
sion, the level of nonlinear distortions varies continuously it is useful, for the
case of a full-value characteristic of the distortions, to determine the instantane-
ous and mean value of the dynamic level of the volume of nonlinear distortions. By
instantaneous value is meant the distortion level measured by an instrument with a
time of integration, responding to the minimum signal duration sufficient for percep-
tion of nonlinear distortions (tens of milliseconds). Of greatest interest is the
measurement of the instantaneous dynamic distortion levels, which correspond to the
ma;d mum level of transmission. This permits a more reliable control of the limit of
the allowable use of the equipment than by measurement of the maximum d narrc voltage
levels. Actually in the latter case only an indirect judgement as to the magnitude
of nonlinear distortions is possible. However, the rare and transitory increases of
distortion at transmission peaks cannot have a substantial effect on the quality of
reproduction as a whole, and in this respect the mean value of the dynamic distortion
level Rd mean is more characteristic. This means the average statistical value of
the distortion level, determined on the basis of a distribution in time of the dynam-
ic Levels of sound transmission.
It. For augmenting the correspondence between the measurements and the subjec-
17
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
tive perception of distortions, the masking of distortions by the transmission sig-
nals and also b;- noises in the reproduction rooms must be taken into account.
For the most accurate determination of the masking of distortions, the frequency
? range of sound reproduction must be divided into a series of frequency bands and the
40 Z
level of perception in each of the bands must be reduced by the magnitude of the
masking. The latter is determined with consideration of the effect of the masking
spectrum within the limits of the given band and of all underlying frequency bands.
As is generally known, the portion of the spectra which corresponds to the overlying
frequencies has no practical effect on the masking.
A well-know?m method of rating the masking of spectra by spectrum (Bibl.l) con-
sists in finding, for each of the bands, the equivalent level of perception of the
masking spectrum, which also determines the masking in the given band. This level is
equal to the sum
/3,,I (-}-)13,,.,(-1 ) ... (-4-) RnK (--I - ) .(-4-) fan - I.n ( +) Bna,
0 where BnK = BK - QnK, BK being the integral level of perception of the masking spec-
0
?
trum within the limits of the given band, while QnK is the constant interband attenu-
ation which characterizes the reduction of the masking effect in the band n of the
integral level in the band K, in comparison with its effect in the same band K. The
BnK
signs (+) signify- that, rather than the levels Brj{, the quantities 10 10 are sum-
marized, which are proportional to the power of the masking oscillations.
The rating cited for the case of the masking of distortions by transmission sig-
nals can be modeled in the circuit; by suppltiying the zone, divided into distortion
frequency bands IJepending on the transmission levels) with attenuations equal to the
masking in the given band and lowering the level of perception for the masking spec-
trum of distortions in this band by its magnitude. The indicated attenuations must
be determined by the summary power of the transmission spectrum for the given and un-
derlying frequency bands. At the same time, the action of transmission powers of the
:18
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
underling fregitenc. bands rrtst. he obtaine.i through crn ,cant int erband at t enuat ions
(','nh,)?
! o:ever at some lowering of' the
such a circuit is obviousl;? fairl complex.
r e^.ordinr; acellrac:7 of masking, considerable siriplificaLi ens are anpar ent 1; possible.
?
"'hits, he number o`' t'requenc?? bands ran be redulcei, the efrect or the r'a:Ti inr levels
need t'e considered onl;? in he given l and and in one of two closest underl;: ing t're-
qpienc ' bands (or even on!-- in the given frequenc- band), and finail; , instead of in-
roClllr_.lll a smo01:11 varying attenuIt_'.on, the blanking oC the frequenr; channels
the distortion zone can be accomplished when suffirient.l'? high transmission levels
are present in the corresponding frequency band.
The rating of the masking ac-ion of noises in the reproduction rooms can be re-
alized in a similar manner, /Iit h the difference that. all introduced aLt?enuat.ions
rave a constant magnitude.
t'rinclples of 1)esigning 'l'est \trcill.Ls
s
?
?
The principal problem in the development of test, circuits is the selection of
means for isolating the distortion products at the output of the device being in-
vestigated.
it, is obvious that the method of filtration customaril,r used i-rith multifrequen-
c?? oscillations, such as a sound transmission or the test spectra simulating it are,
is not always applicable inasmuch as the distortion spectra and the fundamental fre-
quencr spectra may considerably overlap one another.
In the latter case, the following means of isolating distortion spectra are
possible: a) compensation of the fundamental frequencies in the test channel at. the
output of the device being investigated; b) blanking of the test channel for funda-
mental frequencies at the output of the device being investigated.
The compensation method is apparently the only method which permits isolating
the entire effective distortion spectrum, which is formed under dynamic conditions,
19
STAT
Ir ~~
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 without an" vn-i n t i nn and without: lowering the (3uali t;' of !1)e transmi ssi or repr'oduc-
Lion. In principle, the compensation method can also be applied in static ireasure-
rren ts. !'he coirnens.ation channel (i? i F.1 ), within the limits of the reproduced I're-
cu
nuene" range, must have t the same t.ransm.ission `ac to" I as does the 'iindl-
Fi.g.1
a) Device being measured; b) Com-
pensator K; c) Amplitude-frequenc,;-
correc for AK; d) Phase-frequenc;
corrector FK ; e) Compensation
channel.
in
mental channel. .since, unrier the examined condi-
ions, the transiris,ion i'actor ~iniquel??, deter-
mines the transition :'unction, the compensation
coral i Lions will be retained also under d:-nap-!c
cLndr.ti.on s. !-'or a control of the transmission
+';io:' ()" of the cc,r,peri.;.i : i r?Ir ,h^nnel, a suitable
extension arr and amplit.uuie-frenuenc?? and phase-
frequency correc'ors (Al, and Fh) must be provided.
,:compensation is realized in the circuits of the
compensator K.
It, is difficult in practice to obtain an ex-
act compensation within the limits of a wide fre-
expedient to have several cc-,,, ???rrsation channels a-
vailable, each of which accomplishes compensation within the limits of a re]ativel;?
narrow frequent-- band. .;uch a structure of the compensation zone facilitates swi ach-
ing to the levels of distortion volume, and also permits an easier evaluation of the
effect of masking.
Blanking the test channel for fundamental frequencies can be accomplished with
a consecutive reproduction of the amplitudes of the frequency components at the out-
put of the measuring device, by means of an analyzer with a sliding frequency of the
type of a spectroscope. This method can be applied in static measurements.
Apart from the reviewed methods of isolating the distortion spectra in multi-
frequency oscillations, another useful method is that proposed by V.i,.Voltf (i3ibl.2),
permitting an isolation of the products of distortion within the limits of a narrow
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
frequency band at the output of the examined device during cut-out of this band from
the spectra of the input signal.
? Dynamic Method of Measuring Nonlinear Distortions
? This method is designed for measuring nonlinear distortions under dynamic con-
ditions of reproduction of a sound transmission and is based on the compensation of
fundamental frequencies in the measuring zone, which is divided b.L filters into a
series of frequency channels. The method considers the correction in correlations
of volunes, and also the effect of the masking of distortion spectra by the trans-
mission spectrum.
The block diagram of the measuring unit is shown in Fig.2, where F1 and Fl, F2
and F2, F3 and F3 are band-pass filters, AK1, AK2 and AK3 are amplitude-frequency
EH
Fig.2
correctors, FK], FK2 and FK3 are phase-frequency correctors, k1, K2 and K3 designate
compensators, Z denote nonlinear loops for introducing attenuations in the low-
frequency channels at low transmission l eves s, ZP]_, ZP2 and 7.P3 represent, blanking
loops, 11; is a logarithmic ]ogometer (electronic), while INN is a meter ror maxi-
21.
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 mum d; ?namic levels of nonlinear distortion volumes, ISI`: is a meter for the mean dy-
namic levels of nonlinear distortion volumes.
The filters of each pair (F1 and Fl, F2 and Fz, etc.) must have strictly iden-
tical damping characteristics and phase-frequency characteristics, in order that the
amplitude and phase correctors AK and FK take into account only the corresponding
distortions in the device being examined. At the output of the compensators K re-
main only the products of nonlinearity (and background) formed by the device being
measured. The nonlinear loops Z, connected in the low-frequency channels of distor-
tions and of fundamental frequencies, introduce, at low levels, attenuations which
take into account corrections in the curves of equal volume. Estimates show that,
under other conditions, corrections in volume can be disregarded. The effect of
masking, in a simplified way, is taken into account by the loops ZP1, ZPand ZP2, 3
which blank corresponding frequency channels when sufficient levels of fundamental
Qscillations are present in these channels and in channels of lower frequencies.
The loops Z connected in the low-frequency channels effectively suppress the
background harmonics of the device being examined (also during the pauses). The
loops ZP suppress products of nonlinearity with frequencies that coincide with the
fundamentals and therefore do not cause nonlinear distortions.
The blanking loops must be inertial to avoid additional nonlinear distortions.
Their time parameters must be selected on the basis of a compromise between the re-
quirement for a low magnitude of additional nonlinear distortions and sufficiently
fast reaction with respect to recording the distortion dynamics. It should be men-
tioned that certain inevitable additional distortions will not greatly affect the
accuracy of measurement inasmuch as the products of nonlinearity rather than the
fundamental frequencies are subject to these secondary distortions.
As far as the amplifiers of the compensation channels are concerned, the non-
linear distortions introduced by them can easily be made negligibly small, since
their output power will not exceed tenths of a watt, and since the input is fed only
,s2
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
?
?
?
with the portion of the fundamental frequency spectra which corresponds to one of
the frequency bands of the compensation zone.
The minimum number of frequency channels must be defined as that at which the
recording of the masking is sufficiently accurate. At the same time, the required
accuracy of compensation of the fundamental-frequency voltage must-be taken into
consideration.
The entire spectrum of the distortion products and also the fundamental-
frequenc;- spectrum are fed from the output of the device being measured to the elec-
tronic logarithmic logometer LG*. The latter gives a rectified current, proportion-
al to the logarithm of the ratio of integral volumes of nonmasked products of dis-
tortions to the fundamental frequencies or, in other words, it measures the dynamic
levels of the volume of nonmasked distortion products. The logarithmic logometer
must also be of the inertia type. The time of integration is selected on the basis
of the necessity of measuring instantaneous dynamic distortion levels.
As a meter for mean dynamic distortion levels (ISN), an automatic recording in-
strument giving a level-gram of distortions can be used or an ordinary electrodynam-
ic meter$-. it is desirable to use-the meter for maximum dynamic voltage levels W.
In using the proposed circuit, the compensation may be disrupted not so much
because of the nonlinearity of the device being examined but because of a variation
in its amplitude and phase-frequency characteristics. It is, therefore, desirable
to have a possibility of checking the transmission factor of the device not only
during the intermissions but also during actual reproduction.
The described measuring unit can be used primarily for rating the single-valued
*More precisely, both the fundamental-frequency and the distortion spectra are pres-
ent at the output of the device being measured; however, at relatively small distor-
tions, this is known to have very little effect on the accuracy of measurement.
*Wfhe sufficient accuracy of the meter readings, when it was used for a similar pur-
pose, was verified experimentally under the direction of the author.
23
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
correspondence between the Fivcn subjective opinion of 1.1'e ::pert. and the numerical
indices of nonlinear distortions.
Tn addition, a similar measuring unit, possibl:T somewhat simplified (reduced
number of compensat ion frequency channels, ma?amall;? simplified recording of masking,
etc.) can be emplo:?ed in most cases for effective operating contr-rol of distortions.
The presence of such a control permits judging any possible defects of the equipment
and any disturbance of the fixed diagram of levels because of imperfections of op-
erative and nonoperative adjustment. of the transmission levels. At the same Lime,
an instrument may be required for continuous tracking of any variations in the trans-
mission factor and for automatic restoration of the balancing of the fundamental-
frequenc:: voltage.
The proposed dynamic method of measuring nonlinear distortions, in comparison
with the method of V.fl.Vollf (I3ibl.2), has the main advantage that the entire devel-
oping spectrum of nonlinear distortions is taken into account by it simultaneously,
whereas the method of Voltf requires the carrying out of a series of consecutive
measurements for the various frequency bands and therefore cannot be used under dy-
namic conditions, without repetition of the reproduced transmission.
Static Method of Measuring Nonlinear Distortions
In the development, with industrial and episodic operational control of sound-
reproducing equipment, speed and accuracy of measurements are essential under defi-
nite conditions stipulated b;- COST (State Standards) or technical requirements.
Moreover, the measuring equipment must be sufficiently simple to handle. It is
therefore expedient to measure nonlinear distortions during the input of discrete
frequency- spectra, simulating the real oscillations of the transmission, i.e., to
measure the static distortion levels.
In selecting such test spectra it is necessary to determine:
a) The number of frequencies entering the composition of the spectrum;
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
b) The edges of the spectrum and the values of the separate frequencies form-
ing it;
c) The voltage amplitudes of the frequency components of the spectrum and
? their distribution in frequency.
? It is evident that. measurements must take into account the conditions most un-
favorable with respect to the formation of nonlinear distortions, and this corre-
sponds to a transmission with a large number of frequencies, i.e. to a multivoice
performance (orchestra, chorus). But even in speech transmission, the number of
frequency components is considerable (measured in tens).
Therefore, with the object of maximum approkLmation to real conditions relative
?
to the quantity, to the power of the products of nonlinear distortions being formed,
and to the character of their distribution in the frequency spectrum, it is desira-
ble to select a sufficiently large number of frequencies of the test spectrum. At a
large number of the indicated frequencies, however, the resulting measuring unit
would be extremely complex and cumbersome. To avoid this, it is necessary to deter-
mine the minimum number of frequencies of the test spectrum at which the spectrum of
emerging distortions still sufficiently corresponds in character to the distortion
spectrum, formed by. real multifrequency oscillations.
The criteria of this correspondence can be selected in the following manner:
It is known (Bibl.3,1t.) that, during the action of multifrequency oscillations
on a nonlinear quadripole, a spectrum is formed of nonlinear distortions, whose com-
ponents are divided into primary and secondary products. The frequency of the dis-
product Ili has the form a'.2 }- b,2 :t a,)3 I ... zak where {.J,
'
tortion 1 ...
'02j , .. ,,, k? ..6 are frequencies of the components of the input potential (fundamen-
tal frequencies). The number of fundamental frequencies participating in the forma-
tion of the frequency of the distortion product ?ii, can be different and is deter-
mined both by the order and kind of product and by the total number of fundamental
frequencies. To the primary products correspond coefficients with frequency compo-
25
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
nents of a = b = c ... = z = ... = 1; in the case of secondary products at least one
of the coefficients before the frequencies oil, '%2, 1)3, ??? ..., which'form the
frequency of the primary product of the same order, is greater than unity, and at
least one of them is equal to zero (the harmonics of the fundamental frequencies are
? a particular case of secondary distortion products). It can be shown that during an
unlimited increase in the number of fundamental frequencies, the summary power of
the secondary products tends toward zero. Thus, the characteristic criterion of the
distortion spectrum of multifrequency oscillations is the concentration of the dis-
tortion power in the primary products of nonlinearity.
BY this is determined the means of forming the frequencies of the distortion
products (frequencies of the type '11 ? '12 ? .. ? i 'en) and consequently their frequen-
cyy spectrum, and also the correlation with the fundamental frequencies. By this, at
a given characteristic of nonlinearity and at given amplitudes of the fundamental
,frequencies is also determined the character of the distribution of the power of
?
the distortion products in the frequency range. Therefore, in selecting the number
of frequencies n of the test spectrum, the main criterion must be the ratio of the
Order of
Iiat.io of the Poi?rer of ' Primary
Products of
Products of Distortions to the
Distortions
Power of Secondary Products
Number of Fundamental Frequencies
n= 5
n- 10,
2
16.0
36.0
3
7.8
21.0
it
3.38
13.1
5
1.16
8.3
power of the primary distortion
products to the power of secon-
products. It is essential to
note that this ratio does not
depend on the character of non-
Linearity of the system.
Calculations made under con-
dition of equality of the volt-
age amplitudes of the fundamental frequencies led to the results shown in the Table.
It can thus be assumed that, already at n = 10, the overwhelming portion of the
? power of nonlinear distortions is concentrated in the primary products of nonlinear-
ity and, consequently, the distortion spectrum is a spectrum characteristic for mul-
26
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
tifrequency oscillations.
In selecting the edges of the test-frequency spectrum one can proceed from:
1) filling in with frequencies the entire operating frequency range (or its greater
part); 2) filling in with frequencies only a certain section of the operating range.
?
The first alternative has the advantage that it permits simultaneously taking
into account the effect of nonlinearity over the entire frequency range.
The second alternative, in which the high-frequency section of the range is
filled in with frequencies and the section of lower frequencies in which masking of
distortions has no effect is left unfilled, will probably permit taking into account
the case of greatest differentiability of distortions and, as is shown below, leads
to a simpler circuit solution. As far as the voltage amplitudes of the frequency
components of the test spectrum are concerned, they can be identical ur variable in
accordance with the probable distribution of the levels of the frequency trans-
mission.
The concrete selection of the edges of the test spectrum, of its frequency com-
ponents, and their amplitudes requires an accumulation of pertinent experimental
data.
It should be noted that, knowing the dependence R = ,f(b), where b is the level
of-the input potential, one can judge the average dynamic level of distortions. For
a
example, a frequently encountered form is the dependence Rd = Rd max + 2 b, where
Uin
b = 20 lg , Rd max corresponds to b = 0, and a is a constant positive mag-
Uin max
nitude. It can be shown that the average dynamic distortion level is also equal to
the static distortion level when the dynamic transmission level is average, i.e., at
corresponding to the mean of its dynamic range (bd mean = -O.5Bd).
the level bd means
Circuits for measuring the static levels of nonlinear distortions can be de-
signed with application of compensation means, filtration, or blanking of the funda-
? fundamental-frequency voltages at the output of the measuring unit.
? A shortcoming of the compensation circuit in static measurements is the com-
27
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 plexity of its tuning (compensation realized in several frequency channels).
If such equipment must be employed for measuring various devices, then the
overregulation of compensation channels may greatly complicate the operation.
In the case of a partial filling of the frequency range of the measuring unit
with the test spectrum, the filtration means for fundamental frequencies can be used,
which makes it possible to realize an extremely simple measuring circuit (Fig-3).
In this circuit, the low-frequency fil-
ter F filters out the spectrum of test
a) b)
! ` I frequencies 't'l - ?,J2, together with the
products of nonlinear distortions which
b
are in the frequency band of wl - wv and
Z
are masked by this spectrum (Fig-4).
a b
LG
Therefore, at the input a of the loga-
Fig.3 rithmic logometer LG only the nonmasked
a) Oscillator of test spectrum; b) Meas- products of distortions, grouped in the
i
ur
ng unit; c) Load band,ok -col, are admitted. The input b
of the logometer is supplied with the
entire spectrum of fundamental frequen-
cies and distortions (which
as is knowm
,
,
does not lead to substantial errors).
In case it is not impossible to assume
that the distortion products in the frequency band ,12 - ov are masked by the
fundamental-frequency- spectrum, the low-frequency filter must be replaced by a band-
rejection filter in the frequency band tt~l - (112. The nonlinear inertial loop Z in-
troduces attenuation which takes into account the correction in volume necessary in
the channel of distortions in view of the low frequencies and the low levels of non-
? masked products of distortions.
A further improvement of the static methods of measurement can be obtained-by
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
?
providing the possibility of visual observation of the distortion spectrum. For
this purpose, a spectroscope (Bibl.5) can be used, whose circuit structure is shorn
in a dotted frame in Fig-5. With partial filling of the range by the test spectrum,
d)
E
k)
al~
r~ f
h
t ____
h) E:,)J
I _j III
-----j~ -------------------
Fig. 5
0 a) Measuring unit; b) Amplifier; c) Mixer; d) Frequency-band amplifier; e) Loga-
iogometer; f) Detector; g) Circuit trigger; h) riultivibrator? i) Auxiliary
rithmic T
oscillator; j) Logarithmic device; k) Oscillator of test spectrum; 1) Frequency
modulator; m) Deflection oscillator
S
exclusion of the fundamental frequencies can be obtained simply by restricting the
limits of frequency variation by using a deflection oscillator. At the same time,
the indicated limits must correspond to the frequency band occupied by the spectrum
of the nonmasked distortion products (for example, the frequency band li-n
Fig.!;). With solid filling of the range by test spectra, a trigger circuit can be
used, fed by the test-spectrum oscillator through an auxiliary reproduction channel,
and blanking the primary reproduction channel of the spectroscope for all frequen-
cies of this spectrums In view of the large number of distortion products, another
means is also possible, based on the application of a device of the multivibrator
29
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
type, which periodically unblanks the reproduction channel for short intervals of
time, sufficient for operation of the spectroscope, through intervals of the order
B
of 100 cps. The mentioned multivibrator can also be used for blanking the reproduc-
tion channel relative to the fundamental frequencie6. In this case, it is essential
that the frequency intervals of unblanking the reproduction channel contain no fun-
damental frequencies.
1. French,N. and Steinberg,J. - Factors Governing the Intelligibility of Speech
Sounds. J.Acoustical Soc.Amer; Vol.19,No.6 (1947)
2. VOltf,V.M. - Dynamic Method of Investigating Nonlinear Distortions. Radiotekh-
nika, No.2 (1953)
3. Koteltnikov,V.A. - About the Action on,Nonlinear Resistances of Sums of Sinusoi-
dal Potentials. Scientific-Technical Compendium, No.11 Svyazttekhisdat
(1936)
h. Wass,A.A. - A Table of Intermodulation Products. JIEE, Vol.95,Part III,No.33
(191,8)
5. Khlytchiyev,S.M. - Problems of Designing Instruments for Visual Observation of
Spectra of Electrical Oscillations. Radiotekhnika,;No.3 (1954).
It should be noted that, in static measurements, the method by V.M.Voltf
- (Bibl.2) can also be used; the method is improved in that the measurements of the
distortion products in separate narrow frequency bands are made in se''auence and auto-
matically with sufficiently high speed.
In conclusion I consider it my duty to express gratitude to V.M.Voltf and to
V.K.Iofe for valuable counsels.
BIBLIOGRAPHY
?
Article received by the 'Edito`rs 11 June 1956
STAT
30
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
0
?
A survey is given of the state of color television in the
USA, France, England, and Holland on the basis of data collected
during a visit to these countries by a group of CCIR delegates.
In March - April 1956, about a hundred delegates of the Eleventh Study Commis-
sion of the CCIR, including six delegates of the USSR, traveled to the USA, France,
England and Holland with the object of studying color television. Demonstrations
were arranged for the delegates, showing the operation of actual and projLcted sys-
tems of color television; also shown were various experiments for substantiating the
principles of these systems and experiments in measuring the quality of the color
picture, accompanied by reports, etc.
Color television in these countries is in various stages of development: in
the USA for example, a system of color television known as NTSC system has been stan-
dardized and is in operation for more than two years; but in the remaining countries
experimental work is still in progress in the study of original systems of color
television or in the adaptation of the NTSC system to the television standards in
effect in these countries. Since the USA at present holds a leading position in the
field of color television and is the only country in which color TV broadcasting has
been realized, the visit to the USA and an inspection of the NTSC system and various
equipment for its operation were of special interest.
USA
2
THE STATE OF COLOR TELEVISION ABROAD
by
A.K.Kustaryev
At the present time black-white television is most developed in the USA. At
the time the CCIR delegation visited the USA, the operating television stations num-
31
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
bered 475 and the black-white TV receivers about 140 million. Actually, all the most
important regions of the country are provided with several programs of TV broadcast
and about 75% of all families have TV sets. Television programs can be transmitted
t
f i
t
T
k
is
o er-
n
wor
he ne
over the entire country by a network of radio-relay lines.
urban lines comprises 109,000 channel-km and the network of local lines,
68,000 channel-km.
Color television as yet plays a still negligible role, but it is thought that,
because of the imminent saturation of the country with black-white TV sets, the share
of color television in the total volume of TV set sales will increase rapidly. From
1954 to 1955, the sale of color TV sets increased from 5000 to 35,000 units. It was.
expected that, in 1956, the sales will amount to 150,000 units and by 1964 may reach
9 million units. An obstacle to wide installation of color television was the high
cost of color receivers. A color receiver with a 53-cm screen costs $ 700 and higher,
whereas black-white receivers cost from 4100 to $300. As is known, the most import-
ant part of a color receiver is the tricolor picture tube. These tubes are still
extremely complex and expensive, but there is a trend toward reduction in their
price as a consequence of improvement in the technology of'their manufacture and the
adjusting of mass production. Thus, from the beginning of 1955 to the present time
the price of a tricolor RCA kinescope with shadow mask was reduced front X175 to r85.
It is thought that in the near future the prices of color receivers can be expected
to drop to $500 and even to $300, which will boost sales.
For the production of color programs, 23 studios (in March 1956) have been e-
quipped and 70 more have been supplied with equipment for the transmission of color,
motion pictures or slides. It is planned that an additional 19 new stations will be
equipped by 1956 for color programs. The production of color programs is as yet I
still negligible. Thus, in the NBC network in 1955 only 40 hours of color programs,
a month were transmitted. It was planned to double this figure by 1956.
Adopted as the standard system of color television in USA is the NTSC system,
J2
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
being a simultaneous compatible system, with transmission of color information in
one color subcarrier with interlacing of spectra of brightness and color signals. A
brief description of the basic principles of the NTSC system was given in the jour-
nal "Electrosvyazt" for June 1956 (Bibl.4); we will therefore not describe it here.
The NTSC system was the result of three years of work by the 30 largest American
radio-engineering firms, whose efforts were combined by the National Television Sys-
tem Committee (NTSC) and directed toward the creation of a single standard system,
approved at the end of 1953 by the Federal Communications Commission. The work on
creating a standard system of color television included investigations of the prin-
ciples selecting the main parameters of the system, establishing tentative specifi-
cations for the system, manufacturing the equipment according to these specifica-
tions and finally, extensive field tests with the system, including the use of trans-
mitters and receivers under conditions of wireless broadcasting, after which the
final standard was developed.
During the stay in USA from 5 to 22 March, the CCIR delegates attended demon-
t strations of color television organized by the National Television Committee and by
various firms. The delegates visited scientific research laboratories, studios and
transmitters of color television, and also radio-relay lines for the transmission of
color television. They observed the operation of color TV receivers with various
tricolor kinescopes and were shown the manufacture of tricolor kinescopes with shad-
ow mask, at the RCA factories in Lancaster.
The visit to color TV studios left the delegates with the impression that the
work on creating color television programs has ceased to bear an experimental char-
acter, and that in this field much experience has already been accumulated.
In comparison with black-and-white'TV studios, the work in color studios is
more complex and expensive. Here more service personnel are required, great care is
necessary in preparing and adjusting the camera, the choice of illumination, and so
? forth.
STAT
33
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Stiffer demands, in comparison with black-and-white television, are made on
transmitters for color broadcasting and on radio-relay lines that transmit color
programs. The most important additional demands as to radio-relay lines in the
transmission of color television are the requirement for a smooth frequency charac-
teristic in the range of the color subcarrier (3.58 megacycles) and the requirement
for minimal distortions of the differential phase, i.e., distortions of the phase of
the signal of the color subcarrier depending on the magnitude of the brightness sig-
nal. The delegates had the possibility of comparing, with the original, color pic-
tures after transmission over interurbanradio-relay lines; in all cases the quality
of the pictures was rated as good.
In the color receivers which were shown during demonstrations, various types of
tricolor kinescopes were used. The operating principles of two types, namely, of
the color kinescope with shadow mask and of the Lawrence tube (chromatron), were ex
amined in the above-mentioned article (Bibl.4.). We note that the tricolor three-ray
RCA kinescope with shadow mask and the 53-cm screen of the 21AXP22 type'is the only.
color tube produced at present for sale. Its shortcomings are the difficulty of
providing good convergence of the three rays over the entire screen and the large
loss of brightness, caused by the mask blocking a large part of the electrons, so
that only 15% reach the screen.
In the tube with post-acceleration,, developed by the General Electric as well
as in the chromatron, a line screen of three-color phosphors is used, together with
a wire grid in front of them for focusing the rays. In contrast to the chromatron,
however, the tube with post-acceleration has three rays. All the wires of the grid
are connected here and are impressed here with one and the same potential of 5 kilo-
volts. Three guns are arranged in one plane, perpendicular to the strips of the
? screen and at a slight mutual slant, so as to have each ray strike the appropriate
phosphor strip. The advantage of this type of tube over the tube with a shadow mask
consists in that about 85% of the electrons reach the screen.
34
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
The Philco Corporation has developed a tricolor kinescope with an indicating or
tracking ray. This kinescope has one electron gun and a screen consisting of verti-
cal phosphor strips. The electron ray created by the gun is resolved into two parts
which consist of a working and a tracking ray. On the screen, the tra.r4ng ray in-
tersects strips of some secondary-emission substance and creates a signal containing
information on the position of the working ray at any given instant of time. This
signal, after proper processing, is mixed with the signal supplied in the working
ray for getting the color picture. The summary signal modulates the working ray in
amplitude and phase so that the latter alternately excites the color phosphor in the
proportion necessary for true reproduction of the color picture. The current of the
tracking ray is low and gives no noticeable strobing of the picture. The advantage
of the tracker tube over other types is the simplicity of design of the tube itself.
But the control circuits outside the tube are fairly complex. Focusing of the ray
on the screen in a small spot is also complex. The latter is necessary to prevent
overlapping of several phosphor strips which would cause dis-saturation of colors.
In the laboratories of the Hazeltine Co., a series of interesting experiments
were shown, confirming the standards used as basis of the NTSC system. The series
of experiments was devoted to selection of the components of a full-color signal and
widths of the assigned bands. The color pictures obtained with the NTSC signal, are
comparable to pictures obtained with three simultaneous color signals at a bandwidth
of 1} me each. By means of color rings of various sizes, which were viewed at vari-
ous distances, the properties of color vision, used as basis in selecting the sig-
nals I and Q, were demonstrated. Comparisons were also made of color pictures dur-
ing variation in the bandwidths of the signals I and Q. The experiments in rating
the effect on the color picture quality of the accuracy of maintenance of the ampli-
tude and phase values of the color subcarrier showed that, for retention of good
quality of the color picture, the phase variations must not exceed t 50 and the am-
plitude variations, ?1.5 db. In a series of experiments on the action of noise and
55
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
and interference on color and black-and-white pictures, it was demonstrated that
methods of constant radiance and frequency interlacing are highly favorable. Exper-
iments on the interference resistance of color synchronization in the NTSC system
showed that color synchronization is preserved even when the signal-to-noise ratio
is insufficient to get a good picture. On the whole, the experiments confirmed the
validity of the reason in the choice of the NTSC system.
The demonstration in France was organized by the Public Service of Radio and
?
0
Television Broadcasting. The delegates were shown three experimental systems of
color television, developed in France: the system of the line-alternating type
known as "Anre de France" and developed by the RBV Society; the system of point-
alternating type known as the system with "dual communication" developed by the lab-
oratories of Electronics and Physics; and the simultaneous code system of Valance.
A11 three systems are still in the experimental stage and have not yet been devel-
oped to their final form. Therefore, the purpose of the demonstration was only to
show these systems and not to recommend any kind of definite system or standard.
The Anre de France system is a line-alternating system with respect to red and
green components of the picture. The full-color picture (one frame) is put together
from two fields of scanning, each of which contains 109 lines and is transmitted in
the course of 1/50 of a second. The lines in the two fields of one frame are not
interlaced but are superimposed, i.e., progressive scanning takes place. The frame
of the picture thus consists of 818 lines instead of 819 lines according to the
French standard for black-and-white television. The picture during the first field
contains 205 red lines, alternating with 204 green lines. During the second field
of each frame, 205 green and 2014. red lines are transmitted, which are superimposed,
respectively, on the red and green lines of the first field. The blue signals exist
continuously. They are transmitted in a subcarrier and create a blue component of
3'6 STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
the picture in each field. Thus, a picture is produced at the receiving end. In the
transmitter to the alternately transmitted red and green signals, a blue signal is
added in corresponding proportion. This is done for improving the compatibility
since, in the reception of a color transmission on a black-and-white receiver, the
? picture is created by signals of all three primary colors in each line. The red-blue
and green-blue components are transmitted with the 7-mc band, while the blue compon-
ent is transmitted in the 8.147-mc subcarrier with double sideband modulation at a
total 2-mc signal bandwidth of the modulated subcarrier. The amplitude of the sub-
carrier is maximally equal to 150 of the amplitude of the red or green signal. In
the color receiver, after detection or.the subcarrier, the separated blue signal is
fed to the corresponding gun of the tricolor receiving tube, and is also fed with
reversed sign to the red-blue and green.-blue components. Thus restored, the red and
green signals are fed to their guns of the tube. The separation of the red and green
signals in the receiver is accomplished by the asymmetry introduced into the line-
synchronizing pulses.
In the Anre de France system, in consequence of rejection of interlaced scan-
ning, the vertical definition is diminished by half. The horizontal definition is
also diminished because of the reduction of the frequency band of the video signal
from 10.4 me (the standard of French black-and-white television) to 7 me. A corre-
sponding reduction in definition occurs during reception of color transmissions in a
black-and-white receiver. With few saturated colors and with saturated blue, the
s
system operates like a system with 409 lines and 50 frames per second with progres-
sive scanning, but on transmission of saturated red and green colors, it operates
like a system with 409 lines and 25 frames a second with interlaced scanning. The
visibility of the color subcarrier on the screen of the black-and-white receiver is
reduced by selecting the subcarrier frequency as an odd multiple of half the line
frequency. In the reception of black-and-white transmissions on color receivers a
? vertical'definition is preserved of 819 lines with interlaced scanning.
STAT
37
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
The system with "dual communication", similar to the Anre de France system, is
a system of alternating type for red and green colors; however, in contrast to the
latter, this system is of the point-alternating type. The main feature of this sys-
tem is the use of one subcarrier with its seemingly double amplitude modulation.
This is achieved in the following manner: Two series of sinusoidal pulses with a
repitition rate equal to the frequency of the subcarrier, are amplitude-modulated by
red and green signals. Then, the series modulated by red signals is shifted 180? in
phase and, in reversed polarity, is combined with the series modulated by the green
signals. As a result, a signal is obtained with a subcarrier frequency whose posi-
tive envelope contains information on the green and whose negative envelope gives
data on the red. During such transmission, large cross interferences occur between
the red and green signals, which can be considerably reduced by transmitting [in-
stead of red (R) and green (G)) signals of the type R + eG and G +e R, where a is a
small magnitude. This correction of signals must be introduced after correction of ,
gamma, since it presupposes a linear system. The blue signal is transmitted, as it
is also in the Anre de France system, on a separate subcarrier outside the spectrum
of video frequency. The full-color picture is composed of four fields, during which
line-and-point interlacing takes place, i.e., the number of full-color pictures per
second amounts to 12.5. The green and red signals are separated in the receiver by
dual detection; in one of them the positive half-waves of the red-green subcarrier
are reinserted and in the other.-the negative half-waves. The chief defects of this
system are, first, the presence of two closely spaced subcarriers, the blue subcar-
rier being close to the sound carrier, and, second, that the green and red signals
can each be used only to one-half the total modulation depth. The system is not
wholly compatible, since the picture in the black-and-white receiver will be created
only by the green signal. The reception of black-and-white transmissions by color
receivers can be accomplished normally, with deflection of filtration circuits and
dual detection. But the system is designed for operation at new high-frequency
38
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
ranges in new black-and-white compatible receivers. These receivers will be similar
to color receivers with the exception of the tubes; in these, the three-color signals
will be used for composing the brightness signal which is then fed to the tube cre-
ating the normal black-and-white picture. Of course, such receivers will be more
? expensive than the ordinary black-and-white receivers.
The demonstrated color pictures still had many defects such as color fringe,
insufficient definition, etc., which can be attributed to the fact that the work on
the systems is still incomplete.
Apart from the demonstration of the systems, the delegates were shown a series
of interesting experiments on determining the quality of the color picture in the
two systems described, by means of special test charts. The prepared charts were
designed for determining such qualitative indices of the color-television channel as
linearity of the transmission of amplitude, cross distortions between the three-
color channels, fidelity of the color reproduction, and picture definition. Also
shown were experiments in the study of picture flicker in the three separate primary
? components. The experiments were made at frequencies of 33.3 and 50 cps and showed,
in passing, that flicker practically does not depend on the blue primary, and that
it decays to zero in the white and yellow, when the red and green primaries are in
the antiphase.
The operating principle of the third system, i.e., the Valance system, was
briefly described in another paper (Bibl.4).
The present state of television in England is characterized by a rather broad
development of black-and-white TV broadcasting. The BBC television network, using
five frequency channels in a range to 100 megacycles, was by the end of the summer
S 1956 to have supplied 97% of the population with television broadcasts. The country
W."--has 6 million TV receivers, i.e., about 40% of all families have receivers. SiJSTAT
39
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
September 1955, a commercial television organization also transmits a television pro-
gram. Owing to the presence of a large network of black-and-white TV receivers
England, Just as the USA, also is interested in introducing a compatible system of
color television.
The first experiments in creating a system of color television were directed
toward developing a system of the NTSC type, adapted to the English television stan-
dard (/.05 lines, bandwidth of video signal 3 megacycles with a channel width of
5 megacycles). This work was begun about two years ago and was carried out jointly
by the BBC, the General P.O. Department,, and the radio industry. The BBC investi-
gated in detail the NTSC system and the problems involved in its conversion to the
British standard, and in October l955.began to conduct experimental transmissions of
color television from the telecenter in Alexander Palace. The radio industry devel-
oped studio and radio relay equipment, transmitters and other apparatus for stand-
ards at 405 and 625 lines, and also some experimental types of color receivers in
405 lines. The General P.O. Department examined problems associated with the trans-
mission of color TV programs by cable and radio-relay lines, and with the frequency
distribution when color television is introduced.
It should be noted that, under the English standard, it is a great deal more
difficult to use combinations of the spectra of the brightness and coloration sig-
nals than it is under European or American standards, because of the more narrow band
of the picture signal (3 me against 5 and 4.2 mc, respectively). In England there
are advocates for a transition to a black-and-white television standard higher in
definition (which would facilitate the combination of spectra) and of the use, for
color television, of higher frequency bands, where it would be possible to employ
separate transmission of the brightness and coloration signals. These questions are
as yet undecided just as is also the question as to the expediency of adopting a sys-
tem of the NTSC type under a standard at 405 lines.
The demonstrations of color television in England included a showing of a STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
s
tem of the NTSC type with standards at 405 and 625 lines, a showing of a series of
experiments, in particular, on the effect of various types of interference on black-
and-white and color receivers, and a showing of various equipment, in particular, a
cclor studio and a radio-relay line for transmission of color television. In the ma-
jority of color receivers, tricolor RCA kinescopes with shadow mask and 53-cm screen
were used.
In the demonstration of the BBC system of the NTSC type, remodeled for the stan-
dard at 405 lines, the following alternates of signal transmission were shown:
transmission of three-color signals in three separate channels with a bandwidth of
3 me for each; transmission of the brightness (Y) and coloration -, (I and-Q)-signals-
in separate channels with a width of 3.1 and 0.4 me, respectively, i.e., with an
overall bandwidth of 4.4 me; transmission of the same signals with a reduction in
the bandwidth for the signal I to 0.4 me; and, finally, transmission by the NTSC
type, i.e., transmission of three signals in a common video band of 3 me. The fre-
quenc3* of the color subcarrier, in this case, was about 2.66 me and the bands for
the signals I and Q were 1 and 0.4 me, respectively.
The fl41 Cu. demonstrated several alternates of transmission, with conversion of
the NTSC system to 625 lines: broad-band system with transmission of the signals Y,
R - Y and B - Y in three channels of 5 me each; system with average bandwidth, with
transmission of the signal Y in a band of 5 me; and signals I and Q with bands of 1.5
and 0.5 me, respectively, during quadrature modulation of the subcarrier arranged
outside the spectrum of the brightness signal; and, finally, a narrow-band system
with a channel bandwidth of 7 me at a subcarrier frequency of about 4.4 me.
- The EKI Co. also demonstrated equipment for conversion of field-alternating
signals into simultaneous signals by means of optical-electronic transcription of
signals. This device, analogous to what is known as the ffchromacoder+" developed in
America by the Columbia Broadcasting Company, was conceived with the object of a-
___-..., +.hrnA transmitting
STAT .
41
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
tubes (the difficulty of exact recording of three pictures, large dimensions and
weight of the camera, complex controls, etc). When the converting device is-availa-
ble in the studio, small and simple cameras can be used of the field-alternating type
with rotating disks with light filters. The three alternating light signals from
such a camera are fed to three kinescopes, each of which reproduces the-picture in
one of the primary colors. In front of the screen of each kinescope stands a trans-
mitting tube, scanning the respective picture. From the outputs of the three trans-
mitting tubes simultaneous color signals are taken, which can then be processed in
the usual manner. In such conversion of signals, all difficulties of recording are
era operated on a standard at 405 lines, while the transcription was made at a stan-`
V
transferred from the camera to the transcription unit, which can be one or several
cameras. The control of the picture from each camera can be accomplished by moni-
tors of the field-alternating type. The transcription unit can be used.for conver-
sion of the standards of scanning. Thus in the device demonstrated, the color cam-
dard of 625 lines. The transcription unit had automatic balancing of variations of
the sensitivity of transmitting tubes according to the photocathode surface in each
of the converter channels.
Holland
nate of the NTSC system, the frequency of the subcarrier was about 4.!3 mc, and
dam and the television installation at Rostndal, and acquaintance with the research
laboratories of the Phillips Co.
The Phillips Co. demonstrated two alternates of the NTSC system at 625 lines
and a system developed by the firm, with two color subcarriers. In the first alter-
The demonstrations in Holland included visits to the PTT laboratories at Leyden-
subcarrier with a frequency of about 4.1 me was modulated by the color signals R - Y
of 1.3 and 0.5 mc, respectively. In the second alternate of the NTSC system, the
transmission of color was accomplished by means of the signals I and Q with bands
STAT
4'2
Declassified in Part - Sanitized Copy
Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
and B - Y with bandwidths of 0.9 me each, in transmission with two sidebands. The
transmission of coloration signals in equal bands is known to eliminate the causes
for the rise of cross interference, but leads to a narrower frequency of the subcar-
rier which makes it more noticeable.
In the system with two color subcarriers, the brightness signal is transmitted
in the capacity of fundamental signal, while the red and blue color signals are
transmitted in the two subcarriers. The frequencies of the subcarriers in t1ie sys-
tem at 625 lines were selected as approximately 3.59 and 4.61 mc. The main difficul-
ty in this system consists in reducing to a minimum the effect of interference from
the subcarriers on the screen. This interference is created by each carrier sepa-
rately as well as by the beat interference between them, which is especially harmful
since the beat frequency is lower than the frequency of each individual subcarrier.
To reduce the conspicuousness, the frequency of the lower subcarrier is taken equal
to an odd multiple of half the line frequency, and moreover, its phase is shifted al-
ternately by ?90? at the beginning of each field. The frequency of the upper sub-
carrier is a multiple of the line frequency, and its phase is shifted by ?1801 at
the beginning of each field. Such a selection of subcarriers reduces the conspicu-
ousness both of the subcarriers themselves and of the beat interference between them.
A transmission is used in subcarriers with suppression of the upper sideband. The
width of the red video signal amounts to 2 me and that of the blue signal to 1 mc.
The green signal is derived in the receiver by means of a matrix.
Demonstrations on a large projection screen permitted a comparison of the three
systems: transmission of each color in a 7-mc band; a system of the NTSC type with
a subcarrier of 4.43 me; and a system with two subcarriers. The difference between
the first system and the remainder was negligible, even at a short distance from the
-screen. This can be partially explained by defects of the projection system, since
the line structure of the picture was not noticeable even at a short distance.
During their visit to the color television studio, the delegates had an oppor-
I_
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
tunity to see two types of three-color cameras - in image orthicons and in vidicons.?
The demonstrations were ended with a showing of tricolor receiving tubes of the
PICA type with shadow mask and tubes of the Phillips Co. patterned on the type of the
above-described General Electric and Philco tubes.
In conclusion, it should be noted that the best and most developed system at
present is the NTSC system. The other original systems described in-the present sur-
vey, have not as yet been developed to such a stage that they might be compared with
the NTSC system. For the time being, no definite advantages of any of these systems
over the NTSC system can be detected.
BIBLIOGIbAPI IY
1. - Color Television. Study Trip of the Eleventh Commission of the CCIR. Bulletin
de U.E.R., Vol.VII,No.37 (1956) pp.1i77 - 503
2. Stier - Demonstration of Color Television before the CCIR. Nachrichtentechnik,
Vol.6, No-5 (1956) pp. 193 - 195
3. - Colour Television in the U.S.A. J.Brit. IRE, Vol.16, No.) (1956)
I}. Kustarev,A.K. - Systems of Color Television, Elektrosvyazf No.6 (1956), pp.23-32
Article received by the Editors 17 October, 1956.
STAT
44
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
TECHNICAL CALCULATIONS OF ERRORS OF ATTENUATORS
by
D .D .Voyeykov
A method is described of computing the errors of attenuators,
operating with constant input voltage. The final calculation re-
suits are reduced to Tables and Graphs which permit determining
the error of an attenuator as a function of the tolerances of its
elements and the magnitude of the attenuation introduced.
._:Introduction
In the design of attenuators, one of the main problems is providing the required
accuracy of the transmission factor. The method of approximate estimate of the er-
rors of attenuators, which is prevalent at present
7C-417 Oj (Bibl.i), leads to excessively rigid tolerances in re-,
sistances or to lowering of the accuracy of the atten-'
U, u uator transmission factor. The method proposed in the
36~ z 3 present article, while not inferior in simplicity to
U,
that previously adopted, permits, at the same time, a
U BUt derivation of mathematically substantiated results.
4? U q C Ur t 5
Analysis of Errors of Attenuator Sections
Ail D
D T ? For simplification of the analysis, attenuators
~~ - will be examined that operate at constant input volt-
Fig.l
`%~-- age rather than at the constant electromotive force of
? a generator. Such attenuators are used in the Russian signal generators ZG-11,
?1. ZG-12, I-101, and others. As shown below, the derived results can, with sufficient
45.
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Table, 1
Circuit Type and
Formula of Transmission Factor
Relationship be-
Number in Fig.1
ul
tween Sections,
T
when Rr
u2
A??C -rctg 2 -
T + 1
Pi-network
1 + B + B
r
(No.1)
. TI- I
B-rsh9=r 2T
A=B-rtg2
T-network
AB + A (r + C) :LC (B + r)
T- I
-1 F +-I
(No.2)
Cr
1 t '.2T
C'rshe -1 V-1
Bridged
B = C'- r
T-network
A(BC-r-BD+Br+CD)
- BD -}- CD)
BC
AD
1 +
A=r(T--1)
(No-3)
1
+
r (
D=r T-I
e
B=D-rtg 2 =
T- I'
Bridged
ABC+ABD -I-ACD 4- ACr+ADr+BCD+BCr-1- BDr
r T t I I.
(No-4)
r (AC - BD) --
~I
A-C=rcth2
T -(` l
rT-1
A=C=r cth 0 T -,
0-network
(No-5)
1 +(B+D)(- -f C
B=D- 2 rshO=
/
T=- i
=r 4T
1
A=B=D=E= 2 rtg 2 =
H-network
(N
6)
(A _}.. C + D) (B -f C -}- E -}- r) C
-
I T - I
r T -{- 1
o.
r
Cr
1 _2T
C-rs-h~=r~-1
46
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
accuracy, be applied to any cases of operation of attenuators, if the errors of the
sections of the latter do not exceed 0.5%.
In the general case, the circuit of any attenuator consists of a series of very
? simple sections pictured in Fig.l. The errors of the individual sections can be de-
t termined by the differentiation of the transmission factor.
We introduce the following symbols:
ul, u2 = input and output voltages, respectively, of the attenuator;
R. r = input and output resistances, respectively, of the attenuator;
At B, C, ..., M = resistances of attenuators according to Fig.l;
T = f (A, B, CO ..., M) = transm4ssion factor;
N = attenuation in decibels;
6 = attenuation in nepers.
It is obvious that
Differentiating these formulas, we derive the particular errors
Applying" Kirchhoff Is laws, we derive the formulas for the transmission factor,
given in Table 1.
N
T==u, , T=102', N=8,6811.
U2
BT,1 OTs ()Td aTnt
dA ' dB ' VC ' " '' dM
and throughthem the total relative error of the transmission factor of a section
'
?T"+-T"+`~T ATAf
50_1and the probable relative error of the transmission factor of a section
( A 12 + (!,'q + (_jC)2?... -{-(ATM)s
(1)
STAT
47
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
The cumbersome expressions for c and a are substantially simplified for the
case R = r, and in such form are they represented in Table 2.
Circuit Type and
Error Equations
Number in Fig.1
(T- 1)2 2T T -F- I l
(bc +
bs +
i b
-'
Pi-network
T - -
T-
2T=
r J
(No.1)
`1 7''
+ (T -I-,,2 b
-
2
bi
1)? 6c I_
_ (T
`T
I)2
)
)
r
2T=
V
T-network
21'1'(1'+ 1) ['-'T2bA +(T2+1)bB+2T(T-1)bc4 (T+ 1)2brJt
(No.2)
T- I 2 2 2 2
2T, T -}- 1) Y4T4b A -f- (T'2 + 1)2h n -I- 4T2 (T - 1)26% -[-- (T + 1)4b r
Bridged
21'2 [(T2 - I) b,j i- (T - I) bn + (T - 1)9bD -f? (1'2 - I) b,J
T-network
(No-3)
c = (T2- 1)262 -[- (T -1)46 2 + (T - 1)4b) -f- (T2 - 1)2b2
Bridged
T2 - I
E _ ST2 - [(T -{- 1) b,1 + (T - 1) b ti +(T (T + I) b c + (T - I) b, + 4b,J
(No-4)
s
Ts-
6T2I 1/(T -f- 1)2b2 -I-(T - 1)2b=;f +(T+ 1)=bC-}- (T - 1)2b',+ 16br
0-network
E=T.)T,1 [TbB+TbD f(T-1)bc+(T-f-1)brJ
(No-5)
T2T21 YT=b!1 +T2hp+(T-1)2bc+(T-f-1)=br
H-network
4T2(T?I) [2T-(bA+bD)+(T'+1)(bn+bc)+
(No.6)
+ 4T (T - 1) be + 2(T + I')2brI
T-1
41'=(T-f-ix
4b2 -{-41'1 2 2 161'2(T -1)2b 2 4(1' + 1)4b'a
~4TA bD+(T=+1)-(b 2' s+ b E) -I- C'f
The Table contains the following specifications:
q- i""'The magnltuaea DA, "BO DC, eeep DM can mean vatiner Lne Lgivrancva-JUL-~rsv,~cvrr.
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
r ---lsponding resistances or the relative erroijs of resistances. If bA, bB, bC, ???s
Most
s
=~tare expressed in percentages, then 9 and o are also expressed in percentages.
tical tolerances. After assuming
id
en
frequently, all resistances of a section have
bC = .., bM b; the derived formulas can be considerably simplified
b A = a
?- A = LT~
20_.~
(4)
STAT
Table 3
Circuit Type and
Number in Fig.l
Pi-network
(No.l.)
T-network
(No.2)
Bridged
T-network
(No.3 )
Error Equations Maxiimim
Errors
2b(I- T);
((4-) /1,5+271,
2
E=b(1- T?1)(3+ TS);
Tr-1 2,5-T +3T2+Ts+T,;
c=b (I- 2
2b (1- T I;
C1 T1V 1+T'
Bridged --~ p = b `1 - T,, (T -+- 1);
(No-4) /
aTb 4 f --L)1/1+ T Tz
0-network
(No-5)
H-network
(No.o
E=2b 1-T);-
a-b(1- T )/1--0,5TS;
T+I)( T2
C 1 T -; 1) I 1,87 -}- T -t- 2,75 TS + T3 + 0,375 T4
emiix = 26
Omar = 1,22b
E max
OMAK ow 1,586
E m.K = 2b
i max == 2b
Amax
E mar = 36
a max = 1,366
For any section circuit, the data of Table 3 can be represented in the form:
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
. 1
TrnTTTMA
TTTT-TTme
TTfl=
I LL.
-TOILM
lug -44RU _r A
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
This should be taken into account during calculations of the errors of complex cir-
cuits of attenuators. IMioreover,'this circumstance permits applying the graph of
Fig.2 for computing attenuators operating at constant electromotive force of a gen-
erator, when the tolerances on elements of the sections are not more than 0.5%.
Conclusions
1. The errors of the transmission factor of attenuator sections can-be unilat-
erally determined according to the cited Table 2, Table 3, and the graph in Fig-2.
2. The errors of the transmission factor of complex circuits of attenuators can
be determined on the basis of the errors of sections according to eqs.(6) and (7).
The author expresses his profound gratitude to N.N.Solovtyev, who read the pres-
ent work in manuscript form and made a number of valuable comments which were taken
into account when preparing the material for publication.
1. Solovtyev;N.N. - Foundations of Measurement Technology in Wire Communications,
Part I. Gosenergoizdat, Moscow (1955)
Article received by the Editors 17 August 1956.
51
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
PULSE DISTORTIONS IN VOICE-FREQUENCY TELEGRAPH CHANNELS WITH PHASE
MODULATION DURING THE ACTION OF INTERFERENCE
by
A.M.Zingerenko
Distortions of the pulse period in PM channels during harmonic,
fluctuation, and pulse interference are determined. It is demon-
strated that, with respect to distortions due to fluctuation and
pulse interferences, the PM channel, while having considerable
advantage over the AN channel, is inferior to the FM channel.
The system of transmitting telegraph pulses by variation of the initial phase of
the carrier oscillation by in has not yet emerged from the stage of experimental
tests. The difficulty of realizing such a system is knorm to consist of the need to
have, at the receiving end, a generator of the carrier frequency, which is synchron-
ized in phase with the oscillator of the transmitter. With random variation of the
phase of the received signals or the carriers, negative reproduction of the received
signals is observed in the receiver.
Despite the mentioned difficulties, attempts have been made to realize voice
frequency telegraphy with phase modulation (PM). In this connection, it-is expedi-
ent to compare the stability of the given system of voice-frequency telegraphy with
the stability of AM and FM systems with respect to pulse-duration distortions, dur-
ing the action of interference of various types.
l
Transient Conditions during Phase Keying
The transient current at the output of a filter, during variation of the'ini
STAT'l.
52
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
;tial phase of the source of a unit electromotive force by t n, can be represented by
.,the equation
i (t) =-= sin w,t -- 2i l (1),
t the filter output when
t
a
?is the transient curren
a force with a frequency W is switched on.
(1)
a unit electromotive
In the general case, the current il(t) can be represented in the form
ii (J)==g1(1)sinw11-}-g2(1)cos wet, (2)
'where gl(t) and g2(t) are, respectively, the cophasal and quadrature components of
?._ the current il(t).
Substituting the value il(t) in eq.(1), we derive
i t= 1- 2 sin w t 2g2 (1) cos w t. (3 )
In accordance with eq.(3),
We represent i(t) in the form of an oscillation with transient amplitude and in-
Fig.l
ar, and the transmission-factor
i (t) = A (t) sin [wit + d (1)].
A (1) = V [ l - 2g1 (1)]2 +4g2 (1),
HI1)_ -arctg I ?g~(1)
(4)
(5)
(6)
As is known, if the filter pass band is rel-
atively narrow, the phase characteristic is line-
characteristic is symmetric relative to c,11, then
92(t) 0,
(fILF
f 5111 wl
1 (1) A LF (l11) - - W. dw.
53
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
, itial phase
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
ALF(w) is the amplitude-frequency characteristic of the equivalent LF filter
(Fig.l).
9(!) -4-, (8)
Consequently, with a relatively narrow pass band and a symmetrical arrangement
of (,j 1 in the pass band, the oscillation phase will be
set instantly. But the oscillation envelope at the in-
stant of phase variation will assume a zero value
(Fig.2), i.e., t = 0 necessitates that A(t) = 0. Shown
I
in Fig,2a and b are voltage oscillograms during the ab-'
Bence and presence of the filter (pass band, 170 cps).;;
The stee
f th
b
pness o
e
uild-up of the envelope cant
be determined from the derivative A(t) according to
time, when t -y 0. Besides,
Fig.2
Srer phi =2
where f11 - fi is the equivalent filter pass band.
With sudden switching on of alternating electromotive force the relative steep-,
?ss of the amplitude build-up is
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
As is logical with PM, the relative steepness of build-up of the pulse amplitude
will be twice as great as with AM.
Distortions during Harmonic Interference
During transmission of signals by variations in the initial phase, the voltage
of the signal can be represented by the equation
uC=Ac (I) Sill [(()1)/-I-f4(t)]- -A,,(t)[sill wotCos e(t)-[-cos w0/ sill E)(1)],
where Ac(t) is the signal envelope at the output of the receiver filter;
0 (t) is variation of the initial phase during transmission of the signal.
Assume that in the pass band of the receiver a harmonic interference acts, with
a voltage of
un = U? Sin [ui0 -I i2?) t T,] = U? [C(.s (lilt -[ p) sin w0t-[- sin (Q j + (?,I) cos (1)()t],
where ftn is the difference in angular frequencies of the signal and interference;
To is the initial phase of interference.
The total voltage of signal and interference is
u = Ur -' un -- [A, (t) cos 8 (t) -4-- U, cos (fiat -[- rfn)] sin tit / 1--
-}- [f1, (t) sin 0 (t) -[- U? sin (12?t -[- p0)] cos w0t.
We represent this voltage in the form
sin H (t) Sill (S2?t
IF (t) = arctg - -
c(is H (1) + K;, CON (wnt + fo)
Declassified in Part - Sanitized Copy
Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
As is known, the voltage of the signal at the output of the phase detector will
be equal to
?
1'-= '1 (1) Si n 11, (1).
Within the limits of a pulse of a given sign. .B (t) is equal to 0 or it. Then,
Ii' (1) = .- nrctg
R? sln (!,~~t 4u)
I I I ~cns(~~,,/ TO
The greatest value (t) will be at sin (n nt + (P o) = 31. In this case, at Kn < 1.
Consequently, when Kn < 1 the interference cannot cause variations in the pulse
sign.
The greatest variation in the amplitude A(t) will be at cos(Unt + p o) = ?1. In
/1 (/)=-= lc(1)(1 fln)=,1r(/) +-K;,
0.
Thus, the pulse-duration distortions in the channel with PM will be determined
Gust as in the channel with AM) by the greatest increase of the signal envelope
AA MAX = Un.
We now determine the magnitude of the pulse-duration distortion. In accordance
-.0 - ?S (j- -
S
rel PM c rel PAt
11
- STAT
this case,
with Fig-3, the variation of pulse duration is
56
(12)
2K
2 U
n
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
The,pulse-duration distortions are equal to
i
S ,,n B ? 1000/0.
' PM ~? % = ~
rtl Phl
Substituting the value Srel FM, we derive
'PM = -fI1Kn1 B? 100%
1
If the increase of the filter attenuation 0 b),
with variation of the interference frequency is consid-
ered, then
t
Fig-3
'PM =
her c-eb
fit!-f! B?100%.
(14)
Atf n = 0, when the interference frequency coincides with the average frequency
the filter pass band, the distortionsiwill be maximal (Ab = 0) and will be deter
Fig-4
- .a) Maximum Distortions in %; b) Channel
width, 170 cps; Speed of telegraphing
fl,60 bauds; c) Measurement; d) Calculation
mined by eq. (13 ).
Plotted in Fig.4 are the experimental
and calculated curves for the dependence of
maximum distortions on the difference of
the levels of the signal and the interfer-
ence A p = in-1 . Figure 5a and b shows
Kn
oscillograms of the voltage at the output
of the phase detector in the absence of in-
terference and in the presence of interfer-
ence, with a level lower than the signal
level by 0.6 nep.
It is obvious that the experimental
data-agree well'with the calculated values,
lr--As shown previously, in the case of phase modulation, the steepness of pulse build-
57
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
up at equal channel-width will be twice as great as in the-W case of amplitude modula-
tion. Therefore-, the distortions of pulse duration in a'channel with PM, at equal
channel width and equal level of interference, will be twice lower than inlthe chan-,
nel with AM, i.e.,
Approximately the same correlation of maximum distortions occurs for channels
'qWI `maxdmum distortions due to harmonic interference, for various kinds of modulation -I
!with FM and AN. Consequently, the follow~ng comparative data can be compiled for thi
IPM ?- IPM _ 0'5 ! AM'
Distortions during Fluctuation Interference
The voltage of the fluctuation interference at the filter output can be:.repre-
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
sented in the form
where wo is the average frequency of the filter pass band,
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Here Um denotes the"'amplitude of components of the continuous interference
g
spectrum (identical at all ilq);
wq is the angular frequency-of these components;
is the amplitude-frequency characteristic of the transmission
Qq=W q_wo
factor, offset in the axis of frequencies to the left by w o (Fig.6)v
The signal. voltage, within the limits of a pulse of given sign, reads
tic = A, (t) sin wot. (20)
In accordance with eqs.(17) and (20), the resulting voltage at the output of the
r receiver will be
A(Q) A(w.,)
u _ [Ac (t)+U!(1)1 sin wot+ U2(t) cos wot
We represent this voltage in the form
u=A(t)sin[wot+tlt(t)). (21)
The envelope of interest to us is equal to
where
al envelope, during the action of a 2 lue-
, the increment of the sign
Conse entlyuation interference, is equal to
A(1) Ai(l)-}-U~(t)?
a ro
STAT
.:1
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
&.'1(t) = U1(t).
vl~ _n we derive t
ence, and substituting Kn for Kn = tt
Ab Using eq.(13), which determines pulse-duration distortions in harmonic. interfer=
,.,~ ,. ?wtp ?~ ~l c111 rl~
I , - - - I u; (`) B . 10000,
rM - Y sc'cd (ill fl)
0 where Ucd is the effective value of the signal voltage.
Since U2(t) = Un(t), where Un(t) is the mean-square effective value of the in-
The function U1(t) is a random quantity, subject to the normal law of distribu-
tion from a center equal to zero. Consequently, the magnitude I will also be a ran-
dom quantity, with the same law of distribution. At the same time, the probability
distribution of distortions is wholly determined by the root-mean-square value of
distortions, i.e., by the value
(24)
(25)
terference voltage, then
We denote this ratio by
Then,
Al I 2
f -f\
(
60
(27)
- STAT
(26)
The quantity Un(t) Ucd determines the ratio of the RMS values of the voltages
V u" (I) B- 100
Jp`1 Y ~Ucd(11,-f1)
- of the fluctuation interference to the signal.
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
The density of probability of distortions is equal to
?
12
2o`pnt
(29)
It is evident that the comparative data for the RMS distortions during fluctua-I
1u = I
--!tion interference in channels with PM and AM will be the same as for the maximum dis
--;tortions during harmonic interference. Therefore,
1='.'_ I
aAN = 2zp%1*
(30)
l3r
We compare the derived result with the RMS distortions in the channel with FM-
--'As our investigations (Bibl.2) showed, at a linear characteristic of the discrimina-;
tor, the magnitude of the RMS distortions can be represented by the formula
U NIX - 1+ i (1n P -I- 1) 28,100?/0,
~ni- V G(fll--fi V \ P-1
where p = fII - fI
2f
Af is the frequency deviation;
Kn is the ratio of effective values of the voltages of interference to signal.
At optimum value of the factor p = 1.4, we have
Fm - 0,657 fIT Knil B. 100 % .
(32)
"`- Supplementary investigations showed that the real discriminator under its opti-
mum parameters, without much changing the relative steepness of the pulse build-up,
4'--I
es a reduction of the effective value of interference, and consequently, also a
ensur
;
?11owering of the aFM. Besides, the value aFM will be approximately equal to
i 1' (33)
61
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
and a, we derive
c~M
Combining the equations for
FPM 1 1A.
Qfn, u,a9 Y
We thus derive the following comparative data for 04S distortions
1 t7 ,
AM P.m
12 i
(35)
14-1 As is evident, with respect to distortion resistance during fluctuation inter
1` ;ference, the channel with PM (while having substantial advantages over the channel
{
113 __ (with AM)- is inferior to the channel with FM.
20-11
'241
A comparison of the resistance of the voice-frequency telegraph channel to fluc4
tuation interference at various types of modula- !
tion will be more explicit if a check is made on
the possibility of the appearance of distortions-'
exceeding a value of II = 6 AM.
The probability of distortions, exceeding a
given value at any type modulation,
Fig-7
00 '
.) -_.
P(I ~ 1 ~)= cYy~) r JI
in Fig.7, the quantity P is defined by the hatched area).
We designate x. Then, dI = 6 dx and
o. Y=
P (X- X{) V 2T f e (I .
m .Yf
c c
I
t '
62
We then use a known asymptotic expression
is equal to
(36)
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
PAMnoAM~.
11
P(j~ii)_-Y'~~
b) for the channel with FM
of the appearance of distortions which exceed
By determining the Probability
exceed the value Il
the RMS distortions in the channel with AM (GAM) n times, i.e.,
,
-,1
!.
, we derive :
naAM
n a) for the channel with AM
E?J-~ c) for the channel with PM
2
13 A,
36-1
time, we
e
' FM (j= n?A-M)= V2. "''AM
PAM
P PMM
2
1 ('AM
PAM = `AA1 Cy ?FM
`FAt
PPAM `PM ,. FM
' FA1 `FM
Substituting the previously 'cited values d
63 ?
11n 1. 1
- aAA1 11
?pM
6FM, and QAM we derive :
(39)
(40)
(41)
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
f'Aa( = 2,8 (30,625)?'.
~' FAQ
"AM = 2 (4,482),',
f' P.U r
PM - 1,4 (6,753)?'
~'FAI
Cited below are the values of the examined ratios for various n:
n
I ~':1A1~~'PA1
I 1? phi/PFAi
PAMIPFAI
1
-
9,15
--
2
400
2,91 .10
2, 12.1011
3
1,434.10
4,08.10
6,22.1013
4
5,12-1010
2,65.1013
1,43-1024 .
We note that the cited ratios of probabilities also determine the ratio of the
average number of distortion surges per unit time, exceeding the given value aAN.
-Distortions during Pulse Interference
As previously shown (Bibl.3), the maximum distortions
of telegraphic ,pulse dura
_.tion during pulse interference in channels with AM and FM, are determinediby the fol
_-,lowing equations:
,- i
s I
gnu(:sAM= 4U U '3.100%,
lpIsa AAl
3,5 (46)
f rwrxPM =
Since, with phase modulation, the steepness of the pulse build-up is,twice as
(43)
(44)
great as with AM, then
Ipdu PAi = 0,5 Ipa.AM
(47)
Thus, the comparative data for maximum distortions from pulse interference of;.F(-~?'
64
STAT t
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
-_lequal magnitude, at various types of modulation can be represented in the form
1 puIieAM = 2 IpulsePnt = ib /pulse FM ? lw~
As`+ia evident, the channel with PM, while being twice a'a etable as the channel~1
with AM, relative to pulse interference, is considerably inferior to the channel
with FM.
tW
14,_ ;Conclusions
Thus, the comparative data of distortions during the action of various interfer-
13 .__.I
ences in channels of voice-frequency telegraphy with PM, in comparison to channels
j
2O
with AM and FM, are represented in the following manners
a) For maximum distortions during harmonic interference
.i ?i _
IAM.2IPM21FM'
b) For RMS distortions during fluctuation interference
GAM = 20 PM = 2,81FM
c) For maxb!um distortions during pulse interference
IAM - 2 IPM=3,5 1FM.
BIBLIOGRAPHY
4L 11. Bunimovich,V.I. - Fluctuation Processes in Radio Receivers. Sovetskoye Radio
46_i (1951)
4r__2. Zingerenko,A.M. - Optimum Conditions of Reception of Telegraph Pulses during the
t
.4 Action of Fluctuation Noises in a System with Frequency Modulation. Trans-
actions VKIAS imeni S.M.Budenny, No.41, Leningrad
? 3 Zingerenlco,A.M. - Distortions of Telegraph Pulses in Voice-Frequency Telegraph
.
Channels during Action of Pulse-Type Interference. Elektrosvyaz', No.6(1956)
STAT
65
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
J."1
2-. 4
I
"r.
n S~ _I
Article received by the Ed.itora ?7- Janua1'Y 1956?
66
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
!PROBLEMS OF COMPUTING-CIRCUITS OF T-FILTERS WITH OVERIAP -
by
I.I.Petrov
It is shown that T-filters. with overlap are a variation of
M-derived circuits and can be designed by using the same conver-
sion factor M. At the same time, all considerations remain valid
concerning arrangement of the attenuation bump in the usual M-
S derived filter.
20-1
M-derived circuits are an.extremely widespread type of circuits of electric fil-
-ters. These circuits are convenient in that, while varying the magnitude of the con-
-
ate the attenuation bump of the
t
l
oc
o
?version factor M from 0 to 1, it is possible
--'filter in the interval of frequencies which should be suppressed as much as possible:
:= 8
t
o pos-
The so-called circuits of T-filters with overlap (Fig.l) are also known
sess properties of M-derived circuits. To the present,
time, however, T-filters with overlap are computed
Z,
either by means of converting them to equivalent cir-
Z=V cuits (Bibl.l) or by means of determining the resis-
tances of open-circuit conditions and the short-circuit
Fig.1
of networks of chosen configuration.
We will examine the usual conversion of a half-section of a filter circuit with
- !constant K (the schematic in Fig.2a, where the resistances of the series leg Z1 and
i
Hof the parallel leg Z2 are opposite in sign of reactance and where the condition
= K2 is observed) into a series-derived half-section of the M type.
Z
Z
2
1
r;?
The condition of conversion is to keep the characteristic resistance of the cir-
?
'cuit on the side of the terminals 1 - 2 constant while simultaneously varying the
"resistances of the series and the parallel legs.
67
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
This condition is satisfied in the schematic shown in Fig.2b. The resistance of~
{
the open-circuit conditions in the converted network is
Z~I2 MZI + 1 ;114= ZI + M = T.1 M !,
a; Z, 3 6) ZI P,
Zp Z2 _z 4 Zp
2 4
MZ, a)
3 0
Zp I ZJ4.
2- M 4 '
Fig.2
while the resistance of short-circuit is
2012 = MZ1'
Thus, the characteristic resistance on the
side of the terminals 1 - 2 of the M-derived net-
work is
Z' + Z2 M ZI = Z12'
Z12M M
However, the circuit pictured in Fig.2c* also satisfies the same condition of
? i
keeping the character tic resistance constant. }
Here, the resistance of open-circuit conditions is
Zi+Z,
Zao12= Al '
while the resistance of short-circuit is
*Strictly speaking, the schematic in Fig.2c permits only a determination of 2012
'and Z.,22, by means of which (on the basis of the theorem of bisection) the-parameters
of a section of a T-filter with overlap can be found. We can, however, formally as
Isume that also schematic in Fig.2c can be connected from the right to a matched load'
the second half-section with a load Z12. In this case its input resistance on the
side of the terminals 1'- 2 will be equal to the characteristic. Therefore, for con-
venience of discussion, we retain this inaccuracy and will relate the idea of char-,
acteri.stic resistance Z = Z0Z ?both to the circuit of a section of a T-filter with
overlap and to the practically unrealizable circuit of ahalf-section.
.68
"STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
M Z1
_ Z11-M'M - AIZ,
T01'- - Al ZI
Z1 I - M -t- Al
-1 i
f th
id
e o
e
Consequently, the characteristic resistance of the circuit on the s
to the characteris-
ual
mains e
it
i
di
q
re
rcu
ng c
just as in the prece
I t o -terminals 1 - 2,,1 r)__ 4-4^ ^ resistance of a filter with constant K.
14._~ This circuit cannot be realized in the form ofa half-section, but two half-
16-1aections of it give a real, physically realizable circuit of a section of a T-filter,
18 _Mith overlap (Fig.2d).
20__
The?conversion of a half-section of the filter K into a parallel-derived half-
2? --section of the M type is done under the condition of keeping the characteristic re- i
21---isistance of ? a. half-section of the K type constant on the side of the terminals 3 - 4,'
20-at simultaneous variation of the resistances of the parallel and series legs of the
?23---half-section.
- ~I
ri
The condition is satisfied in the schematic pictured in Fig-3a. This is the
a#~~ Ri .-T'1 b
I f~ yi Z-2.- Z, M ~ -' z" ~Z3,
r
.2- 1 .44
11 0
(1)
2MZ, 40
Z ~ ;fz Z M ~: ..._Z~ bl/a C Lq ~
4 ? A 2M ?--Z--+
_--; Fig-3 Fig-4
R ~~.I I
-circuit of alparallel-derived half-section of the M type. It is easy to detect that'
--the condition formulated above is satisfied also in the schematic of Fig-3b. The
0._-
. network in Fig-3b is also unrealizable in the form of a half-section, but two half-
sections form the circuit of a T-filter with overlap (Fig-3c), which differs from the
,.~precedipg circuit of such a filter in that the "overlapping" resistance is opposite
69
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
__-tin sign of reactance to the resistance of the series leg.
.) 1 1
The developed circuits of T-filters with overlap, of the type of Figs.2d and 3c,
i_..I
_ are a consequence of M-derived conversion; their elements can be computed with a
,given factor M. and their properties by attenuation in the cutoff band'are the isms
as in the sections of usual M-derived circuits, since they have identical expressions
for the transmission constant.
9~_1
As mentioned earlier, the circuit of a T-filter with overlap can, in certain
-_4._
_Icases, be developed by means of equivalent conversion of an M-derived circuit. Thus,
-t-the M.-tyrte.-cM-be converted -
13-I
"0_1
n 7 1
--I
into the circuit of a T-filter with:
overlap by the method of conversion,
of the resistances connected accord=
ing to a triangle circuit, into an
y
equivalent connection according to a
star circuit (Fig.4a), which was
shown by T.Ye.Shi (Bibl.l). Acting',
in a similar manner it is possible
to convert also the circuit of a
series-derived section M into the
circuit of a T-filter with overlap,;
Fig. 5
?
0
after converting the star-connected resistances into an equivalent connection accord-
sists of a parallel or series connection of elements of unlike reactances.
Zing to a triangle circuit (Fig.4b). ?
However, such conversions are convenient for filters of lower and higher fre-
quencies, where reactances of identical sign enter the triangle or star connection
and are inconvenient for the band filter, where Z1 and Z2 are complex and each con-
The proposed method of computing T-networks with overlap makes it possible to
calculate them simply in all cases, since it reduces to conventional M-derived'
70
STAT
. ~Ar
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
0
'networks.
Shown in:Fig.5 are circuits of initial half-sections of filters with constant K
of lower and higher frequencies, and of a band-pass filter, and also corresponding
IC I ,
M-derivved sections of T-filters with overlap, with specifications for computed magni-
C,
NIL, M2
~_'M'Z C,
&N MI
L M Zp
x
1`2
,
gyp- M'
:I LZ
~f-M~f/~f M
f~
z
b-
tudes of individual elements. Not quite
clear is the application of the indicated
method to the calculation of those T-
circuits-with-bverlap,-w'ttich are V1-
derived with two coefficients: Ml and M2.
As is known, in this case during the pro-
cess of it-conversion of band filters, un-
like factors are taken for elements with
_tic resistance and the frequency of the edges of the pass band of the filter remain
1
--!invariable.
In application to a series-derived circuit of a half-section of a band-pass fil-
ter, these conditions lead to the form
Zft.
2~ Hl L
Fig.6
c) 2--L, NC,
F--
M1L1 L1
different sign of reactance, but under
the condition that the input characteris-'
Z.. and Z0, shown in Fig.6a. We-ap-
ply them in the same manner to the cir-
cuit of a T-filter with overlap.
Then the circuit of the half-section
will look as pictured in Fig.6b, and the,
entire section will assume the form shown
Z1
iCr L12M
M, LZ ~MM t-M~ 1
Fig-7
-'-in Fig.6c. Here fl and
The resistances
7-1
f2 are the edges of the pass band.
of the open-circuit conditions and the short-circuit of the
;'parallel-derived network with two
coefficients Ml and M2 on the side of the termi-
" pals 3 - 4 are shown in Fig.7a. For
.2M, G
M
2X1
deriving such Zoo and Zo on the side of the ter-
71?
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
'urinals 3 - 4 in the T-network with overlap, it is necessary to have the circuit of a
half-section,aaa pictured in Fig.7b.
The complete circuit of a section is shown in Fig?7c.
Just as in the usual M-derived filters, the coefficients Ml and M2 in the fil-
where
1 -- fi~1 fi 1 1 i
g fig
- 2 ) 1 Y 1 t2 / ( t2
1, f \ 200
In conclusion I would like to express the hope that the proposed method of com-,
puting T-filters with overlap will permit more frequent use of this type as a kind
of M-derived filter and will increase the possibility of varying the magnitudes of
the elements when designing a filter according'to a given attenuation characteristic.
.BIBLIOGRAPHY
tors in Figs.6c and 7c can be computed as a function of the desired arrangement of
frequencies of an infinitely large attenuation flm and f2 ( according to the following
formulas
?". 1. Shi,T.Ye. - Quadripoles and Electric Filters. Svyazftekhizdat (1934)
Declassified in Part - Sanitized Copy
Article received by the Editors 20 January 1956.
Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
USE OF NONCONTACT EI 3( TS IN DIAL OFFICE CONTROL CIRCUITS
by
V.N.Roginskiy
In the present article'aircuit diagrams for the use of noncon-
tact elements in the control circuits of ATS (automatic telephone
14 systems) are given, and the principles of designing separate auto-
matic office centers on their basis are discussed,
_Introduction
0
2?_.~ Automatic telephone offices which started their development in the Eighteen-
-,Nineties of the last century, completed the first stage of their development in the
20---'Nineteen-Twenties of this century. By that time, a large number of various dial of-
--fice systems had appeared, based on selectors - electromechanical devices with rely-,
-;tivel,y large moving parts and connection times of the order of tenths of a second. i
-Both selectors and electromagnetic relays also found application in auxiliary cir-
,cuits. However, the great organic defects of selectors were soon manifest: mainly
36--ithe poor quality of contact and the frequent failure due to the presence of moving
=; t-- and wiping parts and large inertia.
`f'" One of the possible ways of eliminating the shortcomings of selectors -as the
-1
,1-replacement of selectors by law-inertia mechanisms - relays which are reliable in
-operation and possess good contact. The use of relays for creating systems of
4--switching speech circuits, however, results in a very large number of relays per of-'
-fice, since the number of relays increases approximately proportional to the square
of the office capacity. The relay offices, especially offices of large capacity,
,proved considerably more expensive and required larger premises than the dial offices
with selectors. But for small offices, the use of relays is profitable, since it
0-1-11
73
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
n
permits creating equipment which does not require constant attendance. In control
circuits, the relays successfully replaced selectors in many cases.
At the same time, the development of relay circuits led to the creation of a new
.switching device, the crossbar connector (Bibl.l) which is a low-inertia device with
light-weight moving parts, little movement, low operating time (order of millisec-
-onds), and good contact of the relay type*. Beginning from the Nineteen-Thirties,
I?,
dial offices with crossbar connectors came into increasingly wider use,-and at pres-
'The use of such automatic offices permitted an approximate tenfold reduction in ser-+
'ent the majority of foreign firms have gone over or are going over to such systems.
vice personnel at dial offices.
iAt the same time, research was begun 'on the possibility of creating inertialess!
automatic telephone office systems, in which moving parts and mechanical contacts
would be wholly absent. The extensive research work carried out in many countries of
the world, showed that there is at present'no fully reliable element which might re-`j
In control circuits, however, where different demands are made on contact than
place,mechanical contact in a speech circuit.
in speech circuits, nonmechanical elements have been used increasingly since the
Nineteen-Thirties. The use of valves and gas-discharge tubes in relay circuits made
American ttuniversaltt dial office and in the modernized "Rotary-7-E" system (B3,bl.2).
In the Nineteen-Fifties an electronic director (Bib1.3) was created in England?
which accomplishes recording of a six-digit number and sets up in exchange the first
t hip circuits permitted a simplification of selector sets, As in the
c
it possible to simplify certain schematics. The introduction of electronic or ionic,
g
tubes ui ma
ber) are transmitted without change. The two-year operation showed good service
_ _ _ -, 1 s _ _ -0 4-..4 .. A- ---+^" SAmaurha f. l M r In I Belgium. the So-called ttmechanoelec-
two signs (office code) to seven pulse trains. The last four signs (subscriber num-
? 3*Such a system is sometimes called a crossbar system.
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
the relay (P) to operate, breaking the starting circuit.
between the electrodes (1) and (2) is equal to 70 - 85 v, while it is more than 200 v
'between the central electrode (3) and the outside electrodes (1) or (2).; After igni-i
s
contains a generator of 450-cycle cur-
having 12 outputs with the current-
rent,
phase shift of neighboring outputs
at 30?. Thus, there are 12 different
phases ((Pl, T2..., cp32), fed to the con
tacts of the registers and the selectors.
i
The director for setting up corre-
spondence contains a circuit with two
tion, burning is maintainedlby a poten-
tial of 85 v.
For setting up connection in the
director through one of the speaking
wires, the starting circuit,is closed
the selector clutch-magnet E, and the
selector brushes are set insmotion.
I.
of
So long as the phases of the volt-
ages in the transformers Tr' and Tr2 doh
not coincide, the alternating current
the battery
three-electrode ionic tubes of special design (Fig.2), in which the ignition voltage]
Fig.2
a) Director; b) Selector; c) Brush
ignites the tube L1 and the potential of the point A will be close to the voltage ofj
At the same time, the potential of the point A is lowered, the tube L2 is ignited
(Bib1.2 ),,, multiphase. control, is_ usid_ ,for set-_
new system 7-E ("Rotary+)
ting up correspondence. The dial'office
Fig.1
a) Director; b) Selector
(+150 v). However, when a free output of the sought direction is found,;
the phases of both transformers become identical and the tube Ll is extinguished.
76
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
and the relay W is tripped, breaking the starting circuit. The elements of the cir-
cuit_are ao selected that the operating time of the relay_ W, from the_ moment of clog-',
-?`-ing the circuit by the selector pilot brush, amounts to about 2 milliseconds.
In both cited cases the use of electronics in the director made it possible to;
1
i
6-
nd to
Icreate bY simple means an electrical subdivision of the field of selection a-10 ~ ~'
simplify the selector circuit to the maximum. In particular, the function of the
..jtest has passed from selector to director. In the 7-E system, the group selectors
1: Z
14 -
-have no relay at all.
16
' -(Registering the Number
20-
- From the moment the automatic telephone system developed to its present status,:
22
the transmission of the number from the subscriber to the dial office has been accom4
24
i
n
--plashed in, the form of separate pulse trains, with the number or pulses an a tra
corresponding to the dialed number.
28
For transmission of the number, the equipment contains a rather complex device,
1
0_7 the dial. The pulse method of transmitting the number limits the length of the sub-
31 -iscriber line, and the dialing itself of the number proceeds comparatively slowly.
-71
Therefore, attempts have repeatedly been made to replace the dial by keys and accom-;
plish the transmission of the number by combinations of frequencies, coded pulses,
and other methods; this problem however has not yet
rrent strength
c
i
ti
i
-v
,
u
n
ons
a
ar
--found a practical solution. It should be noted that the possibility of transmitting,
42~
-----the number by voice is not a fantasy.
1 Modern mechanical and crossbar (and also certain step-by-step) automatic tele-
46
--phone exchanges require registration of the number being transmitted. Noncontact
~,II-,elements are.Ive convenient for creating registering devices, and at present many
1W --i
CA Provide for registration and fixation of the number transmitted by means of a dial.
joss, the cv.ai orifice ai.reci or muau cuiit.alii a L-vjSioLPvi uv.i__ ??.?`+ ---
77
STAT
Declassified in Part - Sanitized Copy
Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
For fixation of separate digits of the number one can-have either; several regia
eT 1
---tars (similar to what occurs in the director of the mechanical automatic telephone
MW -
-.,offices of the "Krasnaya Zarya" - Red Dawn - factory), or have one register and sev-!
_,oral simpler fixators (similar to the register relay). i
lo-I - I
The majority of such circuits used at present are designed in semiconductor and?
multicathode gas-discharge tubes (Bibl.8), trochotrons - multianode tubes with elec-,
_1tron-ray,(Bibl.10) and so forth - have not as yet been fully successful. Recently,
gas-discharge diodes and triodes. The use of electron tubes was rejected because of
{
their wastefulness (high current consumption and short life).
Attempts to create and apply various multipositional tubes: electron-ray tubes,
1
Fig-3
a) To other counting chains
- for the purposes indicated above, wide use is being made of magnetic elements with
'rectangular hysteresis loops, and of capacitors.
reg-
We will examine several examples. Pictured in Fig-3 is the circuit of ths
,
78
STATw
Declassified in Part - Sanitized Copy
Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
-- responding circuit of division of separate" number digits (distributor). 1- ~.I
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
~ part of the director with ionic tubes (Bib1.7)? The counting chain for each
istering p
-,digit consists of the tubes N1, N2, ... The analogous chain or tubes D1,
plays the role of distributor, while tube V acts as a train relay.
_
i so that the point di has a posi-
In the state of rest, the tube D1 is burning,
? -?',tive potential and the valve Bl will be open. The remaining valves B2, B3, ?.. Will! (
1 I_ 'be closed. The tubes N1, N2, ... do not. burn, the voltage in the firing electrode of
pulse at the input, the tube N1 is ignited; because of the voltage drop in the resis-
__tance r1 this causes the voltage in the firing
_ the end of the pulse, the tube N1 remains
2 0 ~.
-,the tube 112 is ignited and in consequence
resistance R and the action of the charge
--,the tube N1 is extinguished. At the next
Fig-4
During a long interval between pulse
electrode of the tube N2 to rise. At
in the ignited state. At the next pulse,
of the increase of the voltage drop in thei
of the capacitor in the cathode circuit,
pulse, the tube N3 is ignited and the
tube N2 extinguished, etc*
with the arrival of the first pulse, the
tube V is opened, causing the voltage in
its anode to drop.. In the intervals be-
tween pulses, the tube V does not have time
to be closed, since sufficient voltage is
retained in the grid circuit of the capaci-
tor.
trains,
capacitor
the grid
circuit
-:--of the tube V is discharged and it is closed. Owing to this the voltage at the
- point a rises, which forces the tube D2 to ignite and the tube D1 to extinguish.
--This causes the valve B1 to close, and B2 to open. The pulses of the next train will
-`now arrive at the second counting chain, etc. The dialed number is thus fixed by the
tubes in the counting chains. A recording device of
-burning of the corresponding
I --such type was applied in an experimental director in England (Bibl.3) and in an ex-
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
the first tube being-somewhat higher than in the others. On arrival of the positive
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
perimental thyratron director operating in'Leningrad (Bibl.4) since 1955:
The schematic of the counting chain (with.gas-discharge tubes-and ~lve'5_is
shown in Fig-4. At a negative pulse, one of the tubes is extinguished and the next
ignited, due to the charge of the capacitorrybetween the tubes.
The register, whose schematic is shown in Fig.5, is designed on a different
principle (Bibl.9). The circuit con-
T+M
A#,- I
Fig.5
a) To retainers; b) Pulses of dialing;
c) Initial pulse
i
;sists of four triggers of two ionic re-
lays (gas-discharge tubes) X and Y in i
each. In each pair, normally one of the
tubes is burning. With the, arrival- of a
positive'pulse in the conduator.connec
ted with the firing electrodes of both
tubes, the second tube is ignited and
the first extinguished. Thus, if in the
state of rest the tubes X are burning,
the arrival of the first pulse in the--i
first pair of tubes ignites the tube Yl,
and extinguishes the tube X1. On arriv-
al of further pulses, the tube Y1 will
ignite after odd pulses-and the tube Xl'
after even pulses. At the same time, at
the instant of ignition of-the tube X1,{
,
the voltage rise in the cathode of this!
;tube and thecharge,of the capacitor
cause the gas-discharge two-electron
-=tube S~ to ignite briefly and the pulse to be transmitted by the second pair of the':
tubes X2 - Y2. After the fourth pulse, the trigger X3 - Y3 is brought into action,
etc. Thus, to each digit corresponds a combination of ignited tubes Y. -In all,,.? :;
80
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
0
---:such a register can count 15 pulses.
2 -
The retainer consists of four two-electrode gas-discharge tubes. At the and of
-ithe pulse train, the tube of the first retainer is connected for a short time to
-!tor Cl which, at every pulse, receives across the contact u of the pulse relay a
Fig.6
the Y tubes of the register, across the distribu-,
tor., At the same time, the corresponding tubes
are ignited in the retainer. After that, the
"initial pulse" is given in the register, causing
all X tubes to ignite; the register is then ready'
for reception of the new train.
A similar device serves also for transmission of retained pulses.
Figure 6 shows-. the .. si,.mplest_ ircuit of metering pulses by means of the capaci-
-,booster charge from the capacitor C2 or lower capacity. With proper selection of the
A)
Fig. 7
d)
I
I a) Translator; b) Register; c) Retainers of register; d) Retainer of translator
,circuit parameters, the voltages U in the capacitor Cl will be approximately propor-
tional to the number of received pulses. Such a circuit can, for example, replace
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
the selectors in the director of a multipotential system, which was discussed above.1
There are apparently good prospects for the use in telephony of magnetic cores
with rectangular hysteresis loops (Bibl.11,12), which have found wide application in'
?
-,computers and certain other fields. These elements make it possible to create very
compact and economical circuits.
In certain cases the translation of a retained number becomes the function of
-the register. Such a recount can be accomplished by the means employed-for-counting:
and fixation. The simplest circuit of translating a three-digit number to a two--
digit one is shown in Fig-7 (Bibl.7). For each translated number, the translator
contains a thyratron (T) in which the firing electrode is connected across valves
with.the corresponding outputs of the register retainers, and the cathode is connect-
ed with the firing electrodes of the retainers of translated digits across gas-
discharge tubes (L). The drawing shows only one circuit for the translation of the
numbers 11F2 to 23. The thyratron T is unlocked only in the case in which the num-
ber 142 is retained. In consequence of the rise in cathode potential, the;gas-
i
discharge tubes L are ignited, the digits 2 and 3 will be fixed in the retainer of
the translator.
Inasmuch as the translation occurs within a very brief interval of time, the
translation circuit can be used for several registers connected in series. In the
electronic director, already mentioned above (Bibl.3), a translator common to all
registers is employed, making up to 48 translations in 600 milliseconds.
Noncontact Elements in Relay Circuits and Selectors
The use of noncontact elements, and primarily of valves, in relay circuits has
in many cases permitted considerable simplification of the circuits. Thus, M.A.Gav-!
riiov (Bibl.13) showed that the use of rectifier elements affords an opportunity to
- reduce the number of contacts in the relay circuit-to one change-over contact for
each contact of the relay, and gave the theoretical method of the synthesis of such
STAT
82
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
circuits.
We willidiscuss several circuit diagrams for the application of nonlinear ele-
ments in separate automatic office centers.
Represented in Fig.8 is a circuit of conversion of dialed pulses, transmitted by
Fig.8
__,break of a DC circuit, into pulses of alternating current. So long as the DC circuit
in the contact nn of the dial is closed, the point A has a negative potential rela-
tive to the ground (on account of the battery B1), so that the valve V1 is open and
the valve V2 is closed. When the circuit in the contact nn is broken, the voltage at
the point A is increased because of the voltage of the battery B1, and the valve V1
is closed while V2 is opened. At the same time, alternating current from the genera-
for G is supplied to the winding of the transformer Tr2. Such an arrangement is
tz-used, for example, in certain systems of voice-frequency dialing.
Figure 9 gives a circuit of crossing from a two wire subscriber line to a four-
wire circuit in which the number and the call are transmitted by a current of ultra--
-sonic frequency. The transformer Tr2 and the valves V1 and V2 serve for the conver-
sion of dialed pulses into high-frequency current. The call current of high frequen-
cy, arriving from the side of the four-wire line, is separated by the high-frequency
!receiving. filter and fires the thyratron (L), through which alternating current with
a frequency of 25 cps begins to flow, Through the transformer Tr3, this current en-
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
;tars the subscriber line.
The use of rectifier elements and a thyratron in a dial office.of the 7-E type
("Rotary") (Bibl.2) made it possible in a new way to solve the circuit (Fig.1O) of
Fig.9
a) To subscriber; b) Balance; c) High-frequency filter - transmission;~d) Low-
frequency filter - transmission; e) Low-frequency filter - reception; f) High-
frequency filter - reception; g) Channel
the subscriber-line equipment and the triggering of the finder switches (N.). The
subscriber-line equipment contains two valves V1 and V2 and a set of three resis-
tances R1, R2, and R3.
From this device, a potentiometer feeds a counter voltage of
the order of -36 v. When the subscriber circuit is open, the voltage at point A is
equal to -48 v, and both valves are closed. When the subscriber circuit is closed,
?~ __ i .... ,..._?4"" +hn vn1vA V, in the subscriber
---f
.STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
The valve V1 connects the subscriber-line equipment with the triggering device
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
dine equipment and the valves V3 and V4 in'the common device. An alternating current
iith a frequency of 450 cps (voltage, about 2.5 v) will start flowing through the
I
Fig.10
-'valves V3 arid V4. At the firing electrode'of the thyratron L, this current creates a
-corresponding potential, the thyratron is ignited and begins to pass ripple current,
created by superimposing the battery voltage with an alternating voltage of 80 v of a
- frequency of 50 cycles. This trips the relay P, closing the trigger circuits of the
finder switches (IV). When the brushes of any finder switch contact the line of the,
calling subscriber through the valve V2, the test relay in the switch IV operates
and the finder stops. In addition, the potential of point A is again lowered to -48v
tr
due to voltage being fed from IV across the resistance of 240 ohms. This causes the
valves V1, V3'and V4 to close, and the tube L to extinguish at the instant of a neg-
ative half-wave of alternating current.
.t i
Such a system, at a 10,000 subscriber office, made it possible to reduce the
=--total number of relays by approximately 27%. Structurally, the subscriber-line e-
quipment comprises a small mounting plate 55 x 50 X 8.5 mm, installed in the main
r
I switchboard. jThe resistances R1 and R2 of 15 k-ohm each, permanently connected to
the speech circuit, have practically no effect on the quality of the call, since the
85
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
t
?
attenuation introduced by them amounts to 0.01 - 0.02 neper in all.
In the mechanoelectronic telephone switching system (Bibl.~) the directing
transmission inside the office
is accomplished
Fig.ll
a) Line circuit; b) On 5 lines; c) On
25 lines; d) On 100 lines; e) Selector.
pulses Na follow each
pulses Nc follow with
by means of pulses shifted in time.
Since a detailed discussion of the
entire control circuit would go be-
yond the scope of this article, we
will examine only the circuit gener-
s Ir ra distinguishing pulse which
serves for determining the number of
a calling subscriber in a hundred
(Fig.11). At the office there is a
common pulse generator which creates
14 pulse trains of varied duration
and shifted in time, as shown in
Fig.12. The pulses of trains Na
and NC have a duration of 0.2 msec;
other with an interval of 4 pulses, i.e., after 1 cosec; and
an interval of 3 pulses, i.e., after 0.8 msec. The pulses
TL
NQ n n n n n n n n n
Nc,
NCs
NCI
Fig.12
Nb, r--L-
N8,
Nh
Nb.
Nbs
N,,r n n n n n n n
N47 1Ln n n n n n
Ny n n n n n n n
N4. n n n__L n n n
me, n n n n n n n
of train Nb have a duration of 1 msec.
As long as the subscriber circuit is open,
the potential of point A (Fig.11) will be -48 v
and all valves connected to this point, will be
closed. The resistances of the circuit and the
voltages are so selected that the grid of the ca-'
thode follower L has a trigger voltage only in
the cash when the subscriber circuit is closed
(the potential of point A rises approximately to -16 v) and positive pulses are fed
86
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
Fig.14
particular in the already mentioned experimental
automatic switching system of the Bell Laborator-'
ies (Bibl.6). A combination of the characteris-
tics of valves and gas-discharge tubes permits the
selection of a free line in the wanted direction
through several selection steps.
BIBLIOGRAPHY
1. Kharkevich,A..D. - Development Trends in the Technology of Automatic Telephony.
Elektrosvyazt, No-3 (1956) 4
2. Haffter,H.W. - The 7-E Communication System. Bull.Ass.Suisse Electr.}Vol.44,
No.11 (1953)
3. Heron,K.M., Baker,H., and Benson,D.L. - An Experimental Electronic Director.
POEEJ, Vol.44, No.3 (1951); Vol.44, No.4 (1952)
4. Sobolev,O.A., B,ykova,A.S., and Ponomareva,Ye.M. - Experimental Thyratron Register.
-i Scientific-Technical Compendium NII,MRTP, No.2 (8), Leningrad (1956)
5. Kruithof,J. and Hertog,M. - Meclianoelectronic Telephone Switching System, Electri
Communic. Vol.31, No.2 (1954) [Translated in book ttNew Systems of Automatic
Telephone Officestt. Svyaztizdat (1956)]
6. Malthaner,W.A. and Vaughan,H.E. - An Electronically Controlled Automatic Switch-
ing System. BSTJ, Vol-31, No-3 (1952)
- Introduction to Electronic Automatic Telephone Exchanges. POEEJ,
T
H
Flowers
.
.
,
7.
N
4 (1951)
0)
V
l
o.
o
;
.43,
Vol.43, No.2 (195
? 8. Steinbuch,K. - Electronic Switching Systems for Communications Technique. FTZ)
88
the voltage feed will ignite only one of the gas-discharge tubes and trip'only one of
the relays 0. This property of gas-discharge tubes is used for trunk hunting, in
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
N - - No.B, (1952)
9? -.Bray,F:H., _Ridler,D.S. and Walsh,W.A.G'. - Application_ of_ Gas-Filled Tubes for
Storage and Sending. Electr. Coununic., Vol.26, No.1 (191x9)
? ^ 10. Elinson M.I. and Bernashevskiy,G.A. - Application of Electron-Beam Devices with
. i--~ Plane Electron Ray in Automation and Telemechanics. Avtomatika i Telemekhani-
ka, Vol.13, No.2 (1955)
--i11. Wang,A. - Magnetic Delay-Line Storage., Proc.IRE, Vol-39, No.4, (1951)
14--1 1 -12 . Karnaugh,M. - Pulse-Switching Circuits'Using Magnetic Cores. Proc.IRE, Vol-43.,
1,5,1
?n E
No-5 (1955)
Gavrilov,M.A. - Relay Circuits with Valve Grids. Avtomatika i Telemekhanika,
Vol.16,'No.4 (1955)
Article `received by the Editors 6 January 1956.
89
STAT,
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
BALANCING CIRCUITS OF COMMON BATTERY TELEPHONE SETS WITH._.
SEMICONDUCTOR AMPLIFIERS
by
A.S.Sadovskiy
All kinds of variations of balancing circuits of cordon -
battery telephone sets with transistors are discussed, formulas
are given for design of optimum parameters, and the most ration-
al circuit of such a set is selected. Experimentally derived
frequency characteristics of thensidetone of the set are cited.
Equivalent Circuits of Common Battery Telephone Sets with Transistors
In the selection, from a number of possible variations, the rational circuit of
a telephone set with transistors and also in determining the basic correlations for
computing its parameters, it is convenient to substitute the set circuits by equiva-
lent circuits of transmission and reception. An example of such a substitution for
one of the variations of balancing circuits was cited in a preceding article by the
author (see Bibl.l, Figs.2b, 2c, 3a and 4).
We will examine circuits of sets connected into a subscriber line of limit
-length, so that the resistance Rd (see Bibl.l, ig.3) which regulates the incoming
constant current of set feed, can be assumed as equal to zero. With feed through the
balance line Zb, the latter must contain the parallel connection of the resistance Rb
-and the capacitance Cb (see Bibl.l, Fig-3a); therefore, we will call this line a '
parallel balance line. To prevent direct current from flowing through the telephone,
a capacitor CT is connected in series with the latter.
On the basis of the foregoing, a series of possible variations of balancing cir-
cuits of telephone sets with electromagnetic microphones and single-cycle transis-
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
tors can, in transmission and reception, be represented by the equivalent cir-
cuits 1 - 6 (Fig.l). Here Rg is the output resistance of the transistor, which plays;
the role of internal resistance of the equivalent generator with the emf En, which
' replaces the amplifier.
11
The derivation of the basic correlations and design formulas are cited below for
-'circuit 1.
(1)
-? rating formulas are given in Table 1.
For all circuits of Fig.l we assume an ideal transformer and a voltage ratio
.-equal to
hq i
(2)
The sidetone-reduction condition of the set at one frequency is the equation:
`~ - Zb=MZ1=MIZ1Ielyt,
---where Z1 is the input resistance of the line, equal to the wave impedance;
i.
P1 is its angle;
Zb is the total resistance of the balance line;
M is the numerical factor which determines the voltage ratios of the differen-
tial transformer of the set.
Elements of the parallel balance line are determined from the following corre-
For the remaining circuits (Fig.l, circuits 2 - 6) the similarly found
l.ations :
R _ A? I Z11 and 1
b- cos Cb- wA j Z1 sin T1.
(3 )
In the circuit fed across the balance line Zb, the active resistance Rb of the
balance line determines the input resistance of the set to the direct current rin
(here we disregard the active resistances of the transformer windings II and III);
I therefore the magnitude Rb a rin cannot be taken arbitrarily. In subscriber lines
of limit length, we have rin < 300 ohms according to existing standards; however, it
91
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
?
?
circ.2
a)
Fig.1
a) Transmission; b) Reception
92
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
to -
lie rational to increase it in sets with transistors.
(Basic Correlations and Optimum Parameters of Circuit 1
(4)
sidetone
means that the attenuation of the
effect in transmission must be beta This leads to the condition
ZbX1_112(1+61+112)
zi 121
t.:
). Under this condition, the input resistance of the circuit Zn at the points n - n
?''_ 'during transmission is determined by the expression:
Z n = N .- (1 -I- 111 -}- 122) (1I- 111)
Zz 1 ?112
(5)
U I L (6)
Uh = FT '
where UIi and Un are voltages developed during transmission in the circuit la (Fig.l),
at the points 1 - 1 and n - no
Complete sidetone-reduction at one frequency
The operating attenuations during transmission btr and during reception bre can
" - be found in the following form:
A' - and
R, -1 p1 ZII 1 z
(1 .+ 122) VT i e 2 + (1 -1- nt) (1 l- nt -I- 122)
2 (1 + n j -{-122)
n1 (1 -}- 122)
(PT-71 Z~. -I rT "I
Z .
1
T- 2
?122(1 + 12 1
ZA )' L1' c
where ZT,is the total resistance of the telephone;
(7)
(g)
STAT
93
;
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
the set with transistors, in comparison with the usual set with carbon microphone, is
determined (Bibl.l) by the expression
T is its angle.
T
The greatest allowable increase of the undistorted level* of transmission po ofi
PO =Jn -ti I-Jn t ,t,
%JII! I'll
if, (1 -}- n1)
(9)
Here U11 and U1 are voltages of speaking currents developed in the line during
transmission at the points 1 - 1, respectively, of the set with
transistors (circuit la, Fig.l) and of the conventional set with
0-
carbon microphone; '
Uok is the voltage feed in the collector of the transistor.
The condition for obtaining the lowest possible operating attenuation of the cir-
'cuit transmission btr is equality of the internal resistance of the generator Rg and'
of the input resistance of this quadripolel ZnI or matching during transmission, i.e.,
R lZ"t.-N.
IZ71 Z1I
(10)
Taking into account the value N from eq.(5) and substituting it in eq.(7), will
r(I -}- fit) (I -f- rt,) t?r I (U)
CO
I -f- nt ? tt, -
give
The optimum magnitude of the modulus of total resistance of the telephone IZTI,'
which gives the least attenuation bre, and consequently, also the maximum of currents
.in the telephone during reception, can be found if the derivative
'ing 1ZTI to be variable, is equated to zero. Then,
of bre, assum-
? *Here, distortions are meant that arise when the alternating voltage at the output
? of a triode exceeds the direct voltage feed of the triode collector.
STAT
94
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
I' TIopt 112(1 -I-n1) T.
Z I 1110 "}- 112)
we find the value of the op-
Substituting in eq
(8) the derived value IZTl
.
opt~
erating attenuation during reception
b = I n -}- ,t,) (1 + n2) cos - T 't L
re- 111112 - -,.
The total value of the operating attenuation of the set during reception and
transmission bset characterizes the circuit of the set as a whole. This value is
20--equal to
bstt = 61r -P bre =1 n
+111 cns - cos fir- L -j-b .
11, 2) I I 1~b - r1 1" (14)
As is evident from the derivation, the first component of the set attenuation bn
--'depends on the magnitude of the voltage ratios nl and n2 of the transformer and their
correlations. The second component bpp depends on the magnitudes of the angles P1
and TT.
By selecting the optimum ratio of nl to n2, the minimum value of the first at-
tenuation component bn can be obtained. For determining this ratio, a derivative
of bn according to n2 should be taken and equated to zero. Then,
n 1 -}- 1
1-
and, consequently
I n I ! - ~'' l ~L
transmission, we get
95?
(13)
(15)
(16)
Substituting the derived value n2 in eq.(5) of the input resistance Zn during
STAT
V
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
and since JZnI must be equal to or, at any rate, approximate to hg,,we will always
Considering the optimum correlation (15) between nl and n2, and also eq.(18), we
I-
E
loo,
0
0
1
I
0.05 O
f 0.2 0.
5
1 2
5
10
2O
II v
of the plus or the minus sign.
Here, M and T each can have
Plotted in Fig.2 are curves for the dependences of M? and T? on N. constructed
according to eqs.(19) and (21).
From Fig.2 it is evident that always N < 1, and M+ < 1 when N < 0.5, and M, >
1
when N > 0.5. Since the transistor of the set has an output resistance of-Rg?-Z-
~
? the input resistance rin of the set to direct current.
have N 0.5 in such sets. The quantity M determines the value Rb which deetermines:' ',
When operating in actual automatic telephone networks, the input resi-stance,,rin
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
find the condition of sidetone-
reduction, i.e., M from eq.(1~),
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
of the set with transistors must be approximate to the input resistance of existing.
!sets (300 ohms). However, for increasing the voltage feeding the triode of the set,j
it is desirable to increase the input resistance to a magnitude of the order of 400
?---Ito 600 ohms. For deriving such values rin, the quantity M , which has the least of
-!two possible values, should be selected.
The greatest allowable increase of the undistorted level of transmission pot
-;with optimum parameters in comparison with the usual sets, can be found from eq.(9):.
1-1 -
Y Un 1 1 < U?x
p0 In U I V?2N I In I Ut V N (22)
In the case of supplying the set across the balance line, a direct feed cur-
-rent Io in flows through the transformer which, when short-circuited, provides for
24-1
-?-Ithe operation of the telephone office instruments, which also determines the magni-
!tude of this current. The current Io in creating the ampere-turns of the constant
0--itransformer field, determines the dimensions of the core of the latter.
Figure 3 shows the passage of the feed current for circuit 1 of the set (Fig.!),
which consists of two components: Iob, connecting a-
cross the resistance Rb of the balance line Zb;
and Iok, feeding a semiconductor triode. The latter
creates an inverse magnetic fields so that the ampere-
turns AW of the constant excitation are equal to
AW-10 b(w2-{--w3)--10JOI.
For comparison with other circuit variations, we relate this magnitude to the
number of turns of the least winding (in circuit 1 of Fig.l, this is w3).
e ?
Using eq.(l), we find
area = =1?b (I -}- n )-1?x nl .
'97
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
We determine aw at optimum parameters of the set circuit, for which we use
aw=lob VT --- Ins VI_
MT .
(23)__..
The quantity aw can have two values, determined by the selection of the plus or;
egs.(15), (18), (19), and (21); then,
I " minus sign in finding M and T.
Selection of a Set Circuit with Feed through the Balance Line
I8_? We will discuss the circuits 2 - 6 in Fig.l. Similarly to what was done above
_I
-)---,for circuit 1, analogous design formulas ,can be derived for circuits 2 - 6, at opti- ;
,mum parameters of these circuits. From these, the conclusion can be drawn that the
magnitude of the minimum total operating attenuation of reception and transmis-
sion bset, determined according to eq.(16), and also the magnitudes M. T, po,?and aw;
determined according to egs.(19), (21), (22), and (23) are identical for all cir-
-cuits 1 - 6.
The mathematical formulas nl, n2, and Ull for the circuits are given in
Un
Table 1.
of Circuit
-}- I
/ I ,+
fit - 1
2 :3 4 5
I
T U
I -rtt
?Y2N-2I
2(nt+1)
- it, + 2
nt + I
2nt + I
1 '4 n t
n, +, I
nt - I
98
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
r:, 1
`.1 On the basis of the quantity M, the balance-line resistance Rb is calculated,
which in case of set being fed across the balance line is found from eq.(3). Conse-,
quently, Rb also has an identical magnitude for all circuits 1 - 6.
. ~ in
f,
? Thus, with an ideal transformer all the circuits are equivalent.
ever, under actual conditions, the resistance of the winding Wl is suffi-
~y'
How
- ciently large (300 - 400 ohms), since it must have a large number of turns (order
which is determined by the tendency to approximate IZn) to the prac-
of 2000 - 4000),
tically high value Rg. Consequently' in circuits 5 and 6' the resistance of the
- winding wl causes a considerable increase of the set input resistance to direct cur-
--;rent. For its reduction, the resistance RR would have to be reduced.
2O_ the total cur-
-rent in circuits 5 and 6, in contrast to circuits 1 - 4,
?y '
-'rent Io in, feeding the set, passes through the winding w1. The minimum magnitude of
this current is equal to 19 - 20 ma so that the voltage drop in it amounts to 6 - 8 v.
--;This reduces the direct voltage feed Uok ='IO inRb in the collector to an impermis-
?. VT-! 1
---sibly low value. Thus, circuits 5 and 6 are not suitable for work with transistors
W
- when the feed is through the balance line.
Comparing circuits 1 and 2 with circuits 3 and 4, it will be noted that, in the
-latter, the collector (i.e., Rg) receives less voltage feed Uok because of the volt-
- age drop in the windings w2 and w3, created by the current Io in (order of 0.4
3.-1_.
to 0.6 v). It should be kept in mind that Io in is considerably larger than ok'
-which flaws in these circuits through the winding wl.
Circuits 1 and 2 are, therefore, preferable from the viewpoint of obtaining the
-,--'highest voltage on the collector of the transistor.
A comparison of circuits 1 and 2 shows that the transformer of circuit 1 re-
- quires somewhat less copper in the winding. Consequently, circuit 1 should be con-
sidered the most rational.
99
-'V
STAT,
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
;Frequency Characteristics of the Set during Transmission -
7 S k
The elements of the set circuit were designed on the assumption that it oper-
I
?
ates on a cable line with a conductor diameter of d = 0,5 mm, in which, at a,fre-
?- quency of f = 1000 cps, the modulus of characteristic impedance would be Z1 =
= 775 ohms, and the angle qpl = -43?40t. We assume rin - Rb = 600 ohms, which se-
'cures a current strength Io in = 19 - 20 ma in subscriber lines of more than 6.3 km I
length, sufficient for automatic telephone office operation. Equation (3) yields
the quantity M = 0.565.
From the curves of Fig.2 it is obvious that M_
rive a sufficiently high value of N = IZnI > 0.5s
IUl
ing of the circuit of the set with a triode.
From egs.(18) and (15) we find nl = -7.15 and n2 = 0.75. In testing'a`set as-
Then,
has to be used in order to de-
as is required for optmwn match'
eq.(19), considering the minus
sign, it is possible to find - V N = -4.34 and IZnI = 14.6 k-ohm.
sembled according to circuit 1 of Fig.l,(Fig.4), a semiconductor triode of the P1D
--type was used. On the basis of its parameters, measured under operating conditions
according to knoom formulas, its output resistance was calculated, which, with an
electromagnetic microphone at its input, exceeded 100 k-ohm. Thus, the resistance
of the load at the amplifier output IZnI is considerably less than Rg.
y k
Complete matching, i.e., IZnI = Rg, cannot be achieved, since IZnI cannot be
made an arbitrary magnitude; consequently, the attenuation of transmissioh'btr will
not be minimal, and the amplification furnished by the amplifier will belless than
'th matchin
wi g?
At the same time, a reduction of the load resistance IZnI at the triode,,outputJ.
in comparison with its output resistance Rg, increases 'the maxi.mum permissible un-,,,
distorted sidetone of the set because of the reduction in the voltage ratio n1. J
? We note that for the carbon microphone of the conventional telephone set, ac=
cording to the latest technical requirements, a factor of nonlinear distortions of
100
~.
K_=
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
To obtain the required condition of feed of the triode and stabilization of am-
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
_-as high as 15% at 20 b is permitted. It is known that substantially less nonlinear
distortions can-be-.achieved-in semiconductor amplifiers.__ _
22
will result, irrespective of the length of
24--,
voltage drop in Rb constant.and thus keeps
26-1 1
order of 10 v: Thus, the voltage at the operating point will be
respective of the length of the subscriber line.
i
-the resistances R1, R2, and R3 are used. In order that the reduction in amplifica-11
; _
A* f plifier operation on replacement of'the;triode or'a rise in the ambient temperature
IWK
the transformer w1 with respect to the
winding of w2. In the circuit of Fig.4,
the beginning and end of the transformer'
windings are marked by the letters n
and k, and their cutting-in corresponds
to the derived signs n1 and n2.
On cut-in, the sets establish an Rd
of such magnitude that 10 in ? 19 - 20 ma
the subscriber line. This maintains the
the voltage in the collector Uok to.the i
0 S o n M o D s s o 9-9 6l 7W M M M !fa $X M JAS =eW yA%I
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
---Jficiently low resistance to alternating current in the-microphone circuit.
The calculation of the operating conditions of the amplifier and of the magni-
tudes R1, R2, and R3 can be based on methods given in the special literature (Bibl.2).
Figure 5 shows the frequency dependence of the transmission factor of the cir-
Icuit according to the voltage Kans where (Fig-4)
Ut1
"nn - _U '
trr_
0---rand the resistances are Rl = 1 k-ohm; R', =,5 k-ohm; and R3 = 90 k-ohm.
The frequency characteristics* are shown in Fig.6,:
24 1. Sidetones of the set with differential electromagnetic microphone from a no-,
;battery set TAK ("Ship Telephone Set"). The average sidetone of the set in the linel
2()
'o--{ in the range of 300 - 3500 cps is equal to 51.5 mv/b. The irregularity is_ 50 -dbA*.-
2. Sidetones of the set with differential electromagnetic microphone from an
r---iAmerican no-battery set FE-108, with somewhat improved sidetone characteristic.
regularity is 33.4 db.
3. Sidetones of the carbon microphone AWZ of "Albiswerk" (Switzerland). The
-.;average sidetone of the microphone is equal to 22.4 mv/b. The irregularity is
tion will not be excessive at low frequendies, the capacitance of the emitter cir-
cuit. is.: taken sufficiently large (30 110-: A capacitance _ of-..10.4f_ must.-provide a- suf
The average sidetone of the set in the same range is equal to 30 mv/b. The ir-
is 9.85 db. ?
4. Sidetones of the carbon microphone of the Danish set M-52.
The average side-
tone of the microphone is equal to 35.8 mv/b. The irregularity is 14.8 db.
*In the measurement of the characteristics, A.Ya.Markovich participated.
**The measurement was made with mouthpiece` present.
?
4HRAn improvement of the characteristics was achieved by M.F.Kopp, Ye.A.Samoylenko,
and V.M.Romantsov by changing microphone design and arranging suitable air cells.
STAT
102
Declassified in Part - Sanitized Cop
y Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
When comparing the sidetone curves
of Fig.6, it should be kept in mind that the-
1 .1
~ I_curves 1 and"2 are cited for the set as
_ phone.
?
a whole, while 3 and 4 are for the micro-
The width of the spectrWn reproduced by the set with amplifier and the frequen-
I
,cy characteristic of its sidetone are determined by the characteristic of the elec-
P
12t
250 3lV 400 500 600 71 1 BLKI. 7 12V0 -5a? 1&V 2250 2730 3300 40~ ryrln
Fig. 6
-,tromagnetic microphone used, which has a resonance character. Such a characteristic
4,9
is-explained by the tendency to get as great a sidetone as possible in the no-battery
TAR and EE-106 sets.
.
A much,higher amplification can be obtained by using a dual amplifier. The
401 Jacteristic its necessary slope. This naturally requires the development of a suita-
crophone, expanding the resultant frequency spectra, and giving the frequency char-
4- I ble electromagnetic microphone, which is fully feasible. This makes it possible to
latter permits lowering the demands made on the sidetone,of the electromagnetic mi-
.F.,__obtain a telephone set with a preseribed, most rational electroacoustic characteris-
t
itic (Bibl.3).
4 - _; The use of small-dimension electrolytic capacitors as well as the small size of
;0_.-.;semiconductor triodes and resistances permit placing the entire amplifier inside a
,-'microtelephone tube of normal dimension.
r
103
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
:1. Sadovskiy,A.S. - Telephone Sets with Semiconductor Amplifiers. Elektrosvyazt,
No.1 (1956)
BIBLIOGRAPHY
?_. 2. Fedorovich,V.N. and E1'snits,A.G. - Optinu Characteristics of Electroacoustic
I ?
r Converters of Telephone Sets. Scientific-Technical Compendium "Telephone
Acoustics", Nos.1 - 2, 3 - 4, State Union NII MRTP (Scientific Research In-
stitute Ministry of Radiotechnic Industry), (1955)
Declassified in Part - Sanitized Copy Approved for Release 2013/06/04 : CIA-RDP81-01043ROO2000220002-9
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
I Y CAtCUTATION OF ACTIVE RESISTANCES OF CONDUCTORS OF TUBULAR_ FORM
4LI
G.P.Delektorskiy
The problems of calculating the active resistance of tubular
conductors are examined; an analysis is given of the optimwn cor-
relations of tube thicknesses.
At certain frequencies, conductors of tubular form (hollow)ware found more eco-
.nomical than solid conductors (or those twisted from a number of bare copper wires).
Besides, the economy is a result not only of the lower consumption of copper but
'22-4
_~ also of the reduction of active resistance.
24-A
.__J tubular form and show the frequency limits of their application.
For calculating the resistance to direct current, the following formula is ap-
.-_ plicable :
R - Ohn~c-n,
U - rat (2r - l)
06 __ where Ro is the resistance to direct current, in ohm/cm;
.i (1
-121
We will examine formulas for calculating the active resistance of conductors of,
r is the outside radius of the tubular conductor, in cm;
t is the tube thickness, in cm; 1
a is the specific electric conductivity, in
ohm - cm
However, this formula can be successfully used for calculating active resis-
'' "s-1 tio of tube thickness to
Th
e ra
stances of conductors at low frequencies (Bibl.l).
!the equivalent depth of penetration 0 or the product of the factor of eddy cur-
--'rents k times the tube thickness t determines the frequency limit of the app ca
?-`tion of eq.(1). The indicated ratio must not be greater than 1.
in these cases the equivalent depth of penetration is equal to or
-- '- Actually,
105
Declassified in Part - Sanitized Copy
Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
STAT
 Declassified in Part - Sanitized Copy Approved for Release 2013/06/04: CIA-RDP81-01043R002000220002-9
', - be distributed uni
rester than the tube thickness and the current can be assumed to
2 -1
formly over the cross section. Consequently, the applicability of eq.(1) will be
t
...determined by the inequation
0