JPRS ID: 9103 WEST EUROPE REPORT

• Document Type: CREST
• Collection: JPRS
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82-00850R000200050040-0
• Release Decision: RIF
• Original Classification: U
• Document Page Count: 72
• Document Creation Date: November 1, 2016
• Sequence Number: 40
• Content Type: REPORTS

APPROVE~ FOR RELEASE: 2007/02/08: CIA-R~P82-00850R000200050040-0
~
ELt~ EL ~ > >

21 FEBRURRY 1988 C FOUO 3r80 ) ~ OF 1
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
I~OR ~H'FICIAI. USE ONLY
JPRS L~8938
21 February ; 980
USSR Re ort -
p
~LECTRUNICS AND ELECTRICAL ~NGINEERIiVG
(FOUO 3/80)
~
~ F~IS FOREIGN BROADCAST INFORMATION SERVICE
FOR OFF[C(AL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
~
NOTE
~ JPRS publications contain information primarily from foreign
newspapers, periodicals and books, but also from news agency
transmissions and broadcasts. Materials from foreign-language
sources are translated, those from English-language sources
are transcribed or reprinted, with the original phrasii~g and
other characteristics retained.
Headlines, editorial reports, and material enclosed in brackets
are supplied by JPRS. Processing indicators such as [Text]
' or [Excergt] in the first line of each item, or following the
last line of a brief, indicate how the original information was -
processed. Whe:e no processing indicator is given, the infor-
matior. was summarized or extracted.
Unfamiliar names rendered phonetically or transliterated are
_ enclosed in pare*~theses. Words or names preceded by a ques- -
tion mark and enclosed in parentheses were not clear in the
original but have beez~ supplied as appropria~te in context. ~
Other unattributed parenthetical notes within the body of an ~
item origitiate with the source. Times within items are as
- given by source.
The contents of this publication in no way represent the poli-
cies, views or attitudes of the U.S. Government.
For further information on report content
call (703) 351-2938 (economic); 346~3
(political, sociological, military); 2726
(life sciences); 2725 (physical sciences).
COPYRIGHT LAWS AND REGUI.~ITIONS GOVERNING OWNERSHIP OF
MATERIALS REPRODUCED HEREIN REQUIRE THAT DISSEi+1INATION
OF THIS PUBLICATION EE RESTRICTED FOR OFFICIAL USE ONLY.
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY -
JPRS L/8938
21 February 1980
USSR REPORT
. ELECTRONICS AHD ELECTRICAL ENGINEERII~G
(FOUC 3/80)
This serial publication contains articles, abstracts of articles and news
, items �rom USSR scientific and technical journals on the specific subjects
reflected in th~ table of contents.
Photoduplications of foreign-language sources may be obtained from the
Photoduplication Service, Library of Congress, Washington, D. C. 20540. .
Requests should provide adequate identification both as to the source and
the individual article(s) desired.
CONTENTS PasE
Communications; Communication Equipment; Data Transmission
and Processing ..........................�............o.............. 1
Formation of Linear Frequency Modulated Signals and
Processing of Them in~an Acoustic Surface Wave Filter 1
Increasing the Effectiveness of the Reception of Signals
in a Complex Noise Situation by Aggregate Repetition 5
Components and Circuit Elements, Including Filters and Band Lines 6.
Approximate Calculation of Parameters of Band Filters
Utilizing Acoustic Surface Waves 6
Experimental Investigation of Narrowband Matched Filters
for Acoustic Surface Waves 15
~
Algorithm for Calculation of Shielded Band Lines 20
Electromagnetic Wave Propagation 29 -
Scattering of Modes of the Whispering Gallery Type by _
Ionospheric Inhomogeneities Along a Geomagnetic Field 29
Instruments, Measuring Devices and Testers; Methods of Measuring 42
Acousto-Optical Fourier Special Processor 42
Power Systems 46
Problems of Power Engineering at its Present Stage of
Development 46
- a- [III - USSR - 21E S&T FOUO]
FOR OFFICIAL U5E ONLY
~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
Page
CONTENTS (Continued)
.
Publications 55
Handbook on Acoustics SS
Handbook on Components of Microstrip Equipment 58
Signal Reception and Evaluation of Quality 61
Ultralong Range Propagation of Short Radio Waves 64
The Development of Research in Thermoelectricity in the
USSR 66
- b -
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOIt OFFICIAL ""E ONLY
Communications; Communication Equipment;
Data T~a~~smission and Processing
UDC 621.376.32.621.396.662
~ FORMATION OF LINEAR FREQUENCY MODULATED SIGNALS AND PROCESSIfJG OF THEM IN
AN ACOUSTIC SURFACE WAVE FILTER
Kiev IZVESTIYA WZOV SSSR, RADIOELEKTRONIKA in Russian Vol 22 No 9, 1979
pp 73-75 manuscript received 3 Jul 78
[Article by L.A. Belov, S.S. Karinskiy and V.G. Komarov] -
[Text] Modern requirements for accuracy in the formation of LChM [linear
frequency modulated] signals are very high. For the purpose of producing
at the output of the processing system a compressed signal with a side lobe
level of -35 to -40 dB, the permissible deviations in the characteristics
of the shaping and compression equipment in relation to phase are on the
order of a few degrees, and in relation to amplitude on the order of a few
percent [1]. '
For the purpose of satisfying these requirements, in active shapers it is
necessary to employ systems for automatic tuning of the initial frequency
and the rate of frequency modulation (AP LChM [2,3]), and the power level
[4]. Modern processing units are based on the utilization of acoustic -
surface waves [5,6]. It is possible to rate the quality of shaping in re-
lation to the instantaneous error signal of the AP LChM system's phase de-
tector. Effective in addition is a rating in terms of the level of side
lobes at the output of the combined shaping and compression system. This
report is devoted to the development of such a system.
The structural diagram of an experimental mockup is shown in fig 1. A
controlled oscillator, UG, employing a mitron, is encompassed by a ring
for automatic phase tuning of the initial ~requency, FA:~Ch, including
a mixer, Sm2, ~ilter, F2, and switch, K].2. The initial frequency, f~ ,
is a multiple o.� the ~requency of the quaxtz oscillator, Y:G: �0 -'~lf k'
The reference delay line, LZ, with a time period ok T, mixer, Sml, -
phase detector, FD, ~ilter, ~J., and switch, K11, form the AP LChM ring [2],
by means of which the modulation rate, c~/T , ig tuned to a value of
fk/n2T, where w and T are the ~xequency deviation and the length o~
tI~e LChM signal and n2 ~s the ~re~uency division factor of the KG.
Switch K12 is closed in the FAPCh mode, and switch K11 in the LChM mode,
being repeated at a rate of fk/n2n3 . The frequency division factor,
1
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY -
- m, is equal. to the number o~ tuning points during the LChkI period. ~or
the purpose of enabling astaticis~ of the FA~Ch system, ~ilter F2 was
selected as a two-channel one of the K+ K1/p kind wheze p= d/dt .
Filter F1 was selected to be of a si~iQar typ~itude detectorrAD andmfilter
keeping with the recommendations o~ [4]. Amp
F3, i.n the form of an inertial element, form a feedback loop for the auto-
matic power tuning circuit which ad~usts the control voltage of the mitron.
1) ~-3 # AQ . 2)
CuM ~ � Bnrx AUM
N
~ ~ ~ y~
/!3
yc3 f Kn l~ l ~A4 CM 1 H
a J r
3) ~ Kr 4~ ~
rnH ~ Ae~ yMN cM2
~ f , f /n f '4en ! ~ f )
lfn
r ~~G m2 - U~~ n c1 Bax
~ ~ ~ ~ c+cvmun
~ ~--T-~
Figure l.
Key: 4. Multi lication
1. Summation P
2. LChM output 5. Compression output
3. Division
At work in the output of mixer Sm2 is the LChM voltage, whose initial fre-
quency, f, is determined by the amplitude of the square pulse supplied
to the UGnvia attenuator a. This makes it possible in the simplest way
to convert the spectrum of the LChM ~rom the SHF band to the radio band,
in which the compression filter operates.
The compression circuit coresists o~ amplifier Usl, high-~x'equency switt~h
K13, by means o� ahich the required portion of the total LChI~ signal is
gated, dispersing compression ~iltex FS ~nd output ampli~ier Us2. The
control pulse for K13 is sh~ped in gate generator GS, t~iggered in syn-
chronism with the G~N [constant potential generator],
~ The acoustic sur~ace wa~e (A~Y) compression ~iltez is designed .fro~n a
y-section quartz crystal acoustiG line [5], Zt contains one non--equi- ~
distant opposing-pin-type transducer and another transducer with apodized
electrodes, perfo�rming the role o~ a weighting filter with a cosine-type
y- envelope and a pedestal of about O.OS. Bofih transducers are fabricated
by the photolithographic method from silver 0.1 mm thick.
2 c
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
, The dispersing ,~ilzex [5] w~s designed on the assum~tion tha~ the numbez of
pa~.rs of electrodes in it would be xestxicted to I~ < wT /k , where k
is the electromechanical coupling coefficient, so that distortions as the
result of diffraction and ref lections are slight. Distance k, ~rom the
fLrst electrode to the n-th is ,found froiq the relationship R n= vt ,
where v is the velocity of APV's and t represents the r~ots ofnthe "
equation for the quadratic ~unction of the phase, f t -1- wt /2T = n/2 ,
where n= 1, 2, 3, , n n n s
Measurements of an APV filter with an initial frequency f o� about 20
MHz, a deviation o� 7 MHz and a compression coefficient ofn120 demonstrated
that deviations in phase were not greater than 0.1 radians, and irregular-
ity in relation to amplitude was less than fihree percent. When to the
input of a dispersing filter with weighted processing wa.s supplied the in-
verted response from a similar filter without weighted processing, com- "
pressed pulses were produced (fig 2) with a sid~ lobe level close to the
calculated, -35 ~to -40 dB.
Figure 2.
~ Filters wit'~ weighted processing which utilize APV's were employed in a
combined system for shaping and compressing an LChM signal accarding to
the arrangement shown in fig 1. Investigations demonstrated good effect- -
iveness in the operation o~ a11 three automatic tuning rings. Pickup of
the UG's frequency with re~exence to the quaxtz standard took place in the
initial frequency dif~erence xange o~ 7 to ~i MHz, and the transient process
in the FAPCh system lasted no longex than 2 to 3 us. Constancy o~ the
LChM rate was t~qnitoxed in texuls o~ the ch.ange in phase o~ oscillations of
the di~fere*~ce ~requency in the output of the ~D. zt was shown that the
LChM rate vaxies by not moxe than 0~1, pezcent. The cotqpressed pulse at
the output o~ Us2 is similax to that shown ~.n ~ig 2, with a side lobe
level o~ about ,26 dB,
Thus, an investigation o~ a co~bined system ,~or shaping and compxessing an
LChM signal has demonstrated good shapfng quality and the high effective-
ness of compression filters utilizing APV's with weighted processing.
3
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
Bibliogxaphy
1. Kuk, Ch. and E?ern~el'd, 1~. "Radiolokatsi,oznyye s:Zgnaly" [Radar Sig-
nals], Moscow, Sovetskoye Radio, 1971.
2. Belov, L.A. and Kochemasov, V.N. "Automatic-Tuning of the Rate of
Linear Frequcncy Modulation," RADTOTE~ZKA, 1977, 32, No 8, pp 30-34.
3. Belov, L.A., Barabanov, V.B. and Kochemasov, V.N. "Discrete System
for Automatic Tuning of the Frequency Modulation Principle," ELEKTRO-
SVYAZ', 1973, No 5.
4. Kochemasov, V.N. and Belov, L.A. "Applications of LChM Signals and
Methods of Shaping Them," ZARUBEZHNAXA REIDIOELEKTRONIKA, 1975, No 8.
5. Rechitskiy, V.Z. and ~ingur, Ye.K. "Signal Generators Utilizing PAV's," '
ZARUBEZHNAYA RADIOELEKTRONIKA, 1978, No 3, pp 95-108.
6. Karinskiy, S.S. "Ustroystva obrabotki signalov na ul'trazvukovykh
poverkhnostnykh volnakh" [Signal Processing Equipment Utilizing Ultra-
sonic Surface Waves], Moscow, Sovetskoye Radio, 1975.
COPYRIGHT: IZVESTIYA WZOV SSSR, RADTOELEKTRONIKA, 1979
[53-8831J .
CSO: 1860
8831
4
FCR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
_ FOR OFFICIAL USE ONLY
USSR UDC 621.396.62.391.8:621.512.54
INCREASING THE EFFECTIVENESS OF THE RECEpTION OF SIGNALS IN A COMPLEX NOISE
SITUATION BY AGGREGATE REPETITION
ELEKTROPROM-ST I PRI30ROSTROYENE in Bulgarian Vol 13 No 5, 1978 pp 208-210
RANGELOV, KAMEN V1.
[From REFERATIVNYY ZHURNAL RADIOTEKHN":::A No 1, 1979 Abstract No 1D1 from
the Resume]
[Text] The pos:.ibility is demonstrated of increasing the effectiveness of
reception of signals by a~;gregate repetition and by the use of the structur-
al properties of signals and noise. A quantitative criteria is developed
and used for evaluation of the effectiveness of reception.
[ -6508]
5
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OF~ICIAL USE ONLY
Components and Circuit Elements, Including
Filters and Band Lines
UDC 534.86
APPROXIMATE CALCULATT~N OF PA~tAMETIItS OF BAND FILTERS UTILIZING ACOUSTIC
SL'RFACE WAVES
Kiev IZVESTIYA WZOV SSSR, RADIOELEKTRONIKA in Russian Vol 22 No 9, 1979
pp 56-60 manuscript received 28 Jun 78, af ter revision 14 Nov 78
[Article by V.A. Videnko, I.M. Grankin, Ye.A. Nelin and V.P. Pogrebnyak~
[Text] Approximate equations ~~re obtained for calculating the parameters of
band filters utilizing acoustic surface waves (PAV's). On the basis of a
comparison of the values of parameters calculated by the approximate equa-
tions and by a precise model for analyzing filters utilizing PAV's, it is
demonstrated that the error of the approximate equations is totally accept-
able in engi:~eering calculations.
The familiar methods of synthesizing and analyzing filters utilizing acous-
tic surface waves (PAV's) have made it possible to develop filters satis-
fying high technical requirements [1]. But these methods are complex and
requ~re considerable expenditures of computer resources.
Approximate equations are required for a quick estimate of the parameters
= of band filters utilizing PAV's in an engineering calculation. These equa-
= tions can be arrived at on the bas3s of a model of equivalent circuits of
a filter utilizing PAV's [2] with some simplifications. We will discuss a
filter formed by apodized and non-apodized transducers, since this design
is used most frequently. The results arrived at can be utilized also for
calculating other designs. The simplif ications are based firstly on re-
placing with a sj.nc x function the apodization function of an apodized
transducer synthesized by different methods and, secondly, by taking in'co
account only the main and first side lobes of the sinc x function. We
wili.indicate the justi�ication for tifiis approach.
The apodization function is equal to an inverse Fourier transform of the
- r~quired amplitude-frequency characteristic (AChKh) placed at the origin.
The sinc x apodization function is characteristic of a rectangular band
AChKh. Here x= Oft , where ~f is the width of the passband. Of course,
a rectangular AChKh is physically unrealizable. In this case this is
associated with the fact that the dimensions of tlie transducer are finite.
6
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
For the purpose of realizing a band AChKh approximating a rectangular one,
in the sir.~,lest niethod of synthesis function sinc x must be ~�;eighted in
keep ing with the law of one of the weighting functions in [3]. If function
sinc x is multiplied by a rectangular weighting function, then the AChKh
in this case will be most rectangular, but the levels of non-uniformity in
- tli~ pr~f?;sbr~ncl nncl ~~1de l~~l~en procluceel titet~ nrc ua~ecept~tbl~ Cur lhe m~~c~rlly
of applications. A function weighted in keeping with ~ny other law will -
differ irom function sinc x to a greater extent, the worse the rectangu-
larity of the AChKh. In weighting, reduced to the greatest extent are the
values of functlon sinc x in long-range lobes, and to the least, values
in the main lobe. It is obvious that in synthesizing an anodized trans-
ducer by other methods the anodization function will differ from funetion
sinc x in keeping with the same rules. The maximum value of function
sinc x in the first side lobe equals 0.2, and in the second, 0.13. With `
- a weighted function these values are even lower. Therefore we will take
into account only the main and first side lobe of function sine x.
The parameters of an a~odized transducer are calculated in the following
manner [2]. An apodized transducer is conventionally divided by horizon-
tal lines into bands of identical width. The bands represent non-an~dized
~ transducers. The parameters of a r~on-apodized transducer are easily cal-
culated by the familiar equations. The parameters of an apodized trans-
ducer are found by summing the parameters o� bands.
The static capacitance of a non-apodized transducer [21 equals
Co = [nC,/2 (N - 1), ~ 1 ~
where W is the converter's aperture; C/2 is the capacitance of two
electrodes per unit of their overlap; ands N is the number of electrodes.
The static capacitance of an zpodized transducer eqvals
m
Cp = ~'Ca~2 ~Nt - 1~
~=i
where m is the number of bands; N is the number of electrodes in the
i-th band; ~W is the bandwidth. T~e accuracy of the calculation in-
creases wi~i~ a reduction in ~W . Product (Nj - 1)~W represents the sum
of the amount of overlap of electrodes in the band.
m
~~NI_' 1~~I~
~=i
is the sum of overlaps of ::iectrodes of m bands. If ~W approaches
zero, then
~
~ (1Vi --1)OW
i
_ 7
- FOR OFFICIAL USE ONLY
I
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-00850R040240050040-0
FOR OFFICIAL USE ONLY
will approach the sum of overlaps of the converter's electrodes,
2N-1
v ~n'
n~l
where 2N is the number of electrodes and W is the size of the n-th .
overlap of electrodes. Taking into account t~ie sy~tunetry of the structure
" of an apodized transducer with regard to the central electrode overlap,
~ N
_ Co = C. W/2 -1- ~ [~n �
n-s ~2~
Let us consider the area beneath the apodization function curve, W(t) :
T~2
s = 2I~W(t) ~dt ,
0
where T is the length of the transducer's pulse characteristic. Function
W(t) varies insignificantly between the axes of symmetry of neighboring
electrodes, i.e., within a range of variation of t by an amo~nt of
, 1/(2f~) , where f0 is the central frequency. Let us assume that function
W(t) does not vary within these limits. In th~s case W(t) will be a
step function:
' w _ 0 ~ t < 4fo ' .
2n 3 2n - 1
_ W ~n� 4 fo ~ t < 4fo ' n - 2+ . . . , N.
Then
1/4/e 3/4/e (2N-1)141~ 1 W N
~
S 2 ~4 ~ dt [~a ~ � � � ~N ~ dl ~ 2 -f - n �
1/4/0 (2N-31/4fe tt=2
Comparing this expression with (2), we get
T/Y
- co ~ c,sfo = 2c~o f i c~ cn I dc.
o -
s
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02/48: CIA-RDP82-00850R000200054440-0
FOR OFFICIAL USE ONLY
Let us substitute as the apodiaation function the function W sinc ~ft ,
~ limited to the two lobes,
~ rn
2CsQ~fO ~J I S~� ir I!1 ~1L0 f t~.
~o no ~n c
0
We find the value of
~
- ~I sinx I~
x
0
_ by employing the function �
s
Si (x) - sin z dx~
~ x
0
- the values of wY,ich have been tabulated,
'1a
J I sin x I~=Si (rc)-}-(Si (n)--Si(2n))= 2, 3.
z
0
Thus, -
Co 1,5C,[l~f o/~f� .
(3) -
Since 6f is the width of the passband of a rectangular band AChKh, then
as ~f it is necessary to substitute a value equal to the sum of the width
of the passband and the width of the transition band of the AChKh of an
apodized transducer. For an AChKh with high rectangularity, the width of
the passband is severalfold greater than the width of the transmission band;
therefore it is possible to disregard the latter in this case.
The active component of the input admitt~nce of2a non-apodized transducer
at the central frequency [2] is G= 2k C f WN , where k is the elec-
tromechanical coupling factor. The activescomponent of the input admitt- _
ance of an apodized transducer, taking into account its symmetry with re-
- gard to the central overlap of electrodes an~ themhorizontal line connect-
ing th~ mi2dle of the overlaps,l~g ~Ga = 16k Csf~ Ni2~W . If ~W 0,
then ~_1Ni~W will approach W~ N(y)dy . In c~~~uIating G' we will -
take into account only the main lobe of ~unction sinc x. Weadetermine
N2(y) from the equation
9
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
1 sin nef/2fo1~ (9) '
y - 2 ~~f/2foN ~l+
) ' .
Let us designate ~r~f/2f~ = M. Let us employ the expansion of function
sin x into a series and limit ourselves to three of its terms,
x x' z� sin z~, 1` x' + z`
sinx~ ll 31 + 5! x~ 6 120' -
Substituting x= MN(y) , we get
N* (y) - 20/Mz1Vz (y) -}-120/Mz (1- 2y) = 0.
The solution to this equation is
NZ = io - 2 ~ ~1~ i2y i -
M _
Taking into account that permissible values of y> 1/12 , we get
lo~ 2Y5Vi2y - i ~ _
t~) dy N~ Mz - 18M= .
o ~,~s �
(4)
Thus,
~ ~'ie ~ yi~zC.t~~ ve/~~Z.
In figs la and lb are given cal~ulated dependences of the static capaci-
tance and active component of admittance of an aPodized transducer synthe-
sized by the iteration optimization method in [3], on the relative width
of the passband of its AChKh. Curves are plotted from equations (3) and
(4) and by the dots are indicated the upper and lower limits of variations
in values of parameters arrived at from the equivalent circuit model in
[2J, with a change in number of electrodes frrnn 60 to 200. The width of
the transmission band equals two percent, the level of irregularity in the
- passband 0.5 dB, the aperture is 5 mm and the central frequency is 30 MHz.
Obviously, the approximate equations produce an error totally acceptable
for engineering calculations.
10 ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
C~,n~ 1} 2) CQI061~bH;Ga�1061/OM
/0 � 3
= 8
2
~ ~ 40 �
/
~2 ?0 a~ 40 af,X ~2 10 10~~ 40 af, %
- Figure 1. =
x.ey :
1. pF 2. Ga�106 , 1/S2
i)en,86
eo
io
,~o
ao
~ 2 2~ 40 af ~ ~
- Figure 2.
~ Key:
l. Induced losses, dB
Let us note that the equations for calculating C and G of a non-
apodized transducer differ from the respective eq~ations for an apodized,
by the fact that in place of magnitude N in these equations is included
magnitude 2f~/~f , and the constant factors differ slightly. (In (1)
_ difference N- 1 ti N, since N� 1). Usually the number of electrodes
of a non-apodized transducer is determined on the basis of the condition
that the passband of its AChKh ~or the 3 dB level will equal that required
for the filter, 0,~' : N= 1.77~0/~f' . With high rectangularity of the
AChKh, N= 1.77~~/~f , i.e., the values of parameters for an apodized and
non-apodized transducer dif~er insignificantly.
According to [2] the value of the voltage transmission coef~icient of a
_ filter at the central frequency can be found frovt the equation
- 11
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
1
(Gs -I- Go /woCo) ~Ge -I- Gx -I- J'u~oCo) Y21 I
K=I ,
- Y21Go ~5~
where G~ is the intrinsic conductance of the oscillator; G is the con-
ductance of the load; Y21. is the direct transfer admittancenat the central
frequency; and the primes indicate the parameters of a non-apodized trans-
ducer.
It is easy to demonstiate that
N
Gs~ ~ ~ Yu ~ = 4k2C~f
a1V' ~ -f - ~ 1~c+1 ~n ~
~s
~There L is the number of the lobe of the apodization function. Reasoning
similarly as in calculating C~ , it is possible to demonstrate that
N
2-}- ~ I- 1)~+~ll'~n ~ 21� ~ W lt) d1.
�~1 u
Substituting under the integral sign function W sin~ Oft , limited to two
lobes, we get
sn -
GZt,:. 4kzC,fo~~b"2fo/~Of ~ S~n~ efrdcnefi~=s,6k2C~oW~'fo/Ot�
. o
If N' ti 1.77�0/~f , then G21 = 0.7Ga .
Usually fulfilled are the conditions
Go � Ge, ~o � Wo~o, G~ � ~e, Gd Wo~o ~ 6~
(a severalfold excess is rgquired). ~aking this equation into account,
(5) can be simplified, K ti Gn/G21 . According to [2] the value of induced
losses of the filter at the central frequency equals _
- 12
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAI~ USE ONLY
BI7 = l Olg 4GN 1(2, i. e., BII 201g 2Gzt .
V Go~w
- If G= Gn and N' ti 1.77f~/p~ , then, substituting the value af G21 , -
we ge~
f a
BIl 201g 12, 7RokzC,f o[t~ ( e
f l~
\ /
(7)
where R~ = 1/Gp .
In fig 2 are shown calculated dependences of induced losses of filters with
apodized transducers, the dependences of whose parameters are shown in figs
la and lb. The internal resistance of the oscillator and the load resis-
_ tance equal 75 S2. The curves correspond to (7) and the dots and triangles
indicate the values of induced losses arrived at from the model of equiva-
lent circuits for the quartz of the ST section and the lithium niobate of
the YZ section, respectively. In changing the number of electrodes from
. 60 to 200, the values of induced losses change over the range of 1 to 2 dB.
The conditions in (6) are fulfilled for quartz over the entire range of
variation of the relative width of the passband, and for lithium niobate
- only starting with 10 percent.
In conclusion, let us mention two important conclusions which result from
an analysis of equation (7). When the relative width of the passband is
increased d-fold, the induced losses (VP's) are increased by 40 log d dB.
The difference between induced losses for a filter fabricated with sub-
strates made out of different materials is
~BII - -201g kzC, .
k~Cs~
Substituting in this equation the parameters of the quartz of the ST sec- ~
tion and the lithium niobate o,f the YZ section, we get ~yP = 47 dB .
~ Bib~.iography
1. Kheys and Khartmann. "Equipment Utilizing Suxface Acoustic Waves for
- Communfcations Equipment," TITER, 1976, 64, No 5, n 98.
2. Matthaei, C.L., et al. "Simplifications for the Analysis of Interdigi-
tal Surface-W~ve Devices," IEEL~ TRANS., 1975, SU-22, N 2, p 105. `
13
FOR OFFICIAL USE ONLY
a
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
3. Rabiner, L. and Gold, B. '~Teoriya i primeneniye tsi~rovoy obrabotici
signalov" [Theory and Appllcation o~ Digital Processing o~ Signals],
Moscow, Mir, 1978.
COPYRIGHT: IZVESTIYA WZOV SSSR, RADIOELEKTRONIKA, 1~79 _
1 [53-8831]
CSO: 1860
8831
14
FOR OFFiCIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
UDC 621.396.966 -
EXPERIMENTAL INVESTIGATION 0~ NARROWBAND MATCHED ~II.TERS FOR ACOUSTIC
SURFACE WAVES
Kiev IZVESTIYA WZOV SSSR, RADIOELEKTRONIKA in Russian Vol 22 No 9, 1979
pp 71-73 manuscript received 22 Mar 78, after revision 27 Dec 78
- [Article by A.Ye. Znamenskiy and Ye.S. Muratov]
[Text] The transmission function (PF) of an opposing-pin-type transducer
(VShP) with a constant aperture governing the PF of a filter matched with
a square radio pulse (SF) is [1]:
H (w) - Ho { sin [(c~o - s/2I/((~o - ~u~ T/2] } exp (ic~n/2),
(1)
where Ho is a constant factor; w~ and w are the mean and instantaneous -
angular frequency, respectively, T= 1/R is the length of the radio pulse;
and R is the rate of transmission of binary information.
Let us transform expression (1) into the form
H(w)'_' Noi {sin[(uio - u~) s/2N~"~~~I~~o - tu) T/2N+1~ } X
N '
X exp (i~ut/2N}~) n I1 exp (i~n/2k)),
k=1
~2~
where H01 is a constant factor and N= 1,2,3,... is a whole number.
The approximation sign in equation (2) is associated wi~h the discrete
structure of a VShP, in which can be contained only a whol.e number o,f pins
spaced from one another in ze~,ation to the delay by a value o,f 1/2f
But an analy~is o~ equation (2) shoWS that the accuracy o,~ the approxima-
tion is sufficientlq high (not worse .~han 0.03 R) ~or ~he 1ow transmission
rate of R a 300 kbauds and inter~qediate ~requencies in the 10 to 100 I~IIiz
band, discussed in this article, w1:th 1ow values of N, which completely
_ satisfies the requirements for the para~meters of an SF [2].
15
FOR OFFICIAL i1SE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
It f�I I~~wr~ I'rom ~,c~uH[ion (2) !'hr~t tr ~tA poFRt.hle xo syntheAi�r.c~ required
function H(wj bv means o.~ a cascade connecti,on o~ the S~ for a trans-
mission rate of 2NR and a bank o~ N comb ~ilters (G~~s). Examples of
this type of implementation ~or a filter for acoustic surface waves (PAV's)
for N= 1 and N= 2 are given in ~igs la and 1b, respectively. Here
two-section VShP's i'orm an amplitude-~requency characteristic (AChKh) of
the comb type, and non-an~dized equidistant 9ShP's, an AChKh of an SF for
a transmission rate of 2 B. The letter M in fig 1 designates the me--
talized coating between sections of the GF. Let us note that the method
of implementation shown in fig lb makes it possible to shorten almost two-
fold the lengthwise dimensions o~ the acoustic line, as compared with the
method of implementation in fig la, which is especiallq important for low
information transmission rates of R= 5 to 25 kbauds . A distinctive
feature of narrowband PAV filters is the need to fine tune the mid-band
frequency, fn, deviations of which from the calculated value are caused
chiefly by tTie discrepancy in the speed of PAV's from one model of an acous-
tic line to another (on the order of + 3�10 2 percent for synthetic quartz),
and by the deviation in the rate of propagation of PAV's in metalized sec-
tions of the VShP as the result of a technologicaliy related discrepancy
in parameters of the metalized structure [3]:
(V - VM)/t~ = Yh~N, ( 3 )
where V and V are the velocities of PAV's on ths free and metalized
surfaces, respec~ively; h is the thickness ef the metalization; and Y
is a factor depending on the materials of the metalization and acoustic
line.
Fine tuning in relation to frequency Of0 can be performed, e.g., by -
partial removal of the metalization, M, between sections of the GF (fig
1) :
eto = co~9~v) fa ~ 4 ~
where ~V is thE equivalent change in the velocity of PAV's between
sections of the GF or. account of removal of a section of inetalization;
_ V - VM - (k2/~) v, ~ 5 )
where ~ic is the electromechanical. coupl.ing cve~�icient o~ the material
of the ~~.coustic line.
By increasing the nu~qbex of degrees of ex~ansion o~ N in equation (2),
it is possible to improve the accuracy of xealization o~ the transmission
function of a matched filter (S~) in tuning to Physically this
means that the band of the middle lobe o~ the ACh~ is expanded in the
16
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
case of a high-speed SF in~ide of which the combs o,~ the entire bank o,f
GF's are tuned simultaneously. Calculations have demonstrated that for the
two-element structure (N = 2) in fig lb, in particular, it is possible to
tune the mid-band frequency of within the range of 8+ 40 percent of
R while satisPying certain requirements for the parameters of an SF [2J.
T/1
T/2 T/? ~ ~
Bx
B, ~I \ \ ~ Beix I B~~t
~ ~t--~
_ a~ 2) ~/y r/4
Figure 1.
Key:
1. In 2. Out
Let us give the results of the development of an SF for an OFT [relative
phase telegraphy] demodulator for an inforn~ation transmission rate of
R= 272 kbauds and with f= 70 + 0.006 MHz wit'~ a required accuracy
~ in realization of the passb~nd of F= 0.885R = 240 kHz , of nat worse
than + 10 kHz, and with a deviation in the level of the first side lobe
from the calculated of not greater than + 2 dB. Since the possible fre-
_ quency variance resulting from deviation in velocity in the quartz wafers
is + 3�lOf 21 KHz , or less than 10 percent o~ R, then the decision
was made to construct an SF with a single-element structure according to
~ the type in fig la. As the material for the acoustic line was chosen an
ST-section quartz wafer; the dimensions of the acoustic line were 32 X
X 10 X 2 mm; and the metalized coating was chemically deposited silver
with a thickness of t= 0.08 u.
For the purpose of ineasuring the mid-band ~zequencies of the twa transducers -
(a GF and SF with a 2F band), in the center of the wafer was placed a
broadband measuring VShP. The measured values of the mid-band frequencies
equaled f~ = 70.014 MHz for the SF and f= 69.990 MHz ~or the GF.
The amount of inetal which it was necessary ~o clean away, ~L , was ~eter-
mined on the basis o~ equations (3~ to (5), Whexe ~y = 0.52 and k=
= 0.0012 ; the calculated value o.� ~L ~ 1,30 u, coxre3ponding to a tuning
slope o~ 13 u/kHz.
- Cleaning away the metal with a diamond cuttex by means o~ a UIM-21 micro-
scope, with a precision o~ 3 u, Which corresponds to a method precision _
of not worse than 0.25 kHz at this frequenr_y, xesulted in a mid-band fre-
quency for the GF of 70.001 MHz. The discr~pancy from the calculated value
17
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY ~
is related to the appx'oximate nature o~ the equ~tions used, (3) to (5),
and to measurement errors. The resultii~g AChKh o~ the S~ has a deviation
in the level of the first side lobes from the theoretical in the range of
- + 1.5 dB; the passband fox the -3 dB level equa?~ed 232 kHz. Tr~e experimen-
tal results, thus, agree sufficiently well with the calculated. For com-
parison let us note that the method of fine t~sning fn discussed in [2],
by etching away the VShP~s metalization deptltwise, taking into account
technological errors in its imple~entation, is less accurate than the me-
thod discussed here by an order of magnitude (the accuracy of etching
away *_he metal film is 10 to 20 percent of the thickness of the coating).
In [4] the fine tuning of f is accomplished by spraying a dielectric
f ilm onto the VShP. An eval~ation of the accuracy of this method, made
by means of graphs obtained, shows that it is less accurate than the one
discussed here by one to two orders o~ magnitude.
Another method of electrically fine tuning the mid-band frequency ibetween
plemented by the connection of a phase inverter at an angle of ~
sections sections of the GF. The tuning then equals:
~to = (~l2nn) fa ( g )
where n is the number of wavelengths, a= V/f~ , confined between the
centers of sections. of the GF.
For the purpose of testing electrical frequency tuning, an SF was used for _
f~ = 12.003 MHz and R= 25.6 kbauds , synthesized according to the
arrangement in fig lb by the cascade connection of two elements in an
acoustic line consisting of an ST section of quartz measuring 100 X 20 X
X 5 mm. After measurement of the mid-band frequencies in each element of _
the SF, they turned out to equal f~ = 12.003 MHz .
Since the GF of the upper element basically determines the AChIQ1 of the SF,
then we will tune the mid-band frequency of precisely this GF. In each
section of the GF of the upper element are contained 25 pairs of electrodes;
the distance between the centers of GF sections equals 62 mm; the capaci-
ti~ance of the sections is C= 10 pF with an aperture of 4 mm; and the
spacing of electrodes is every 0.131 mm (between centers). The phase shift
in the current passing through sPparate sections of the G~''s transducer
was made possible in the experiment by installing in one section a series
active variable resistance, r. A sezies variable capacitor, C, in
another section sexved the pu~pose o~ equalizing the amplitudes gf signals
transmitted by sections in the introduction o~ the phase shi~t,
~ ~p .^r larc tR (r�2a1C)),
where C is'the static capacitance of the transducer.
18
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ~'NLY
_ Not taken into account in equation (7) is the series e�~ectiye zadiation
resistance. ~aking into account the ~act that the ~ o~ sectiona o~ the
GF transducer is sufficiently high, this value can be di.sregarded.
In an experiment, in changing r from ze~o to r= 1/(2nfQC) , which
_ corresponds to ~_~r/4 , tuning~'of the mid-band f~equency of the GF's
effective peak was achieved at f= 6 kHz or 0.13 kHz/degree, which has
been confirmed by calculation according to equation (6). The deviation in
the shape of the SF's AChKh thereby, within the frequency di�ference range
- of + 2R, does not exceed the limits of the requirements for garameters of
an SF [2]. An evaluation with equations (6) and (7) demonstrates that re-
placing in the tuned SF the controllable resistor and capacitor with fixed
ones with a prec~sion of + 5 percent and with temperaturP coefficients on
the order of 10 /degree makes it possible to achieve tuning accuracy of
not worse than 0.6 kHz in the temperature ran~;e of + 50�C with a maximum
frequency difference of ~f < 6 kHz . The precision o~ the method is im-
proved considerably by increasing the accuracy of the rating of the re-
sistor installed in thP phase inverter and by lowering its TKR.
In experiments conducted with more complicated 0 to 180� phase inverters,
tuning of ~fp = 18 kHz was achieved for the same speed of R= 25 kbauds .
Therefore, wifh such low transmission rates, R, the electrical method of
fine tuning is feasible to use for relatively not too high intermediate
frequencies of f< 50 MHz , at which the difference in speeds in acoustic
lines does not ex~eed the limits of the tuning range.
Thus, both methods described here of fine tuning the mid-band frequency of
an SF are simple and practical and ensure high accuracy, which has been con-
firmed by experiments.
Bib 1 io graphy
1. Hartmann, C.S. "Impulse Model Design of Acousti~ Surface Wave Filters,"
IEEE TRANS., 1973, MTT-2.?., N 4, p 162-175.
" 2. Hayld, W.H. "Precision Na~�rowband Surface Wave Bandpass Filters" in
"IEEE Ultrasonics Symposii~m, Proc., 1974," pp 429-432.
3. Karinskiy, S.S. "Ustroystva obrabotki signalov na ul'trazvukovykh
poverkhnostnykh volnakh" [Signal ~'rocessing Equi~ment Utilizing Ultra-
sonic Sur~ace WavesJ, I~oscow, Soyetskoye Radio, 1975.
4. Nishika,wa, K. ".~n I~proved Surface ~coustic Wave ~ilter ~or a PCM
- Timing Tank" in "ZEEE Ultrasonics Symposium, Proc., 1974," pp 164-167.
COPYRIGHT: IZVESTIYA WZOV SSSR, RA,DzOELEKTRONZKA, 1979
[53-8831]
CSO: 1860
8831
19
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPR~VED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
- 1~OIt O1~rIC[AI. USL CEN1.Y
UDC 621.372.04.75
ALGORITHM FOR CALCULATION OF SHIELDED BAND LINES
Kiev IZVESTIYA WZOV SSSR, RADIOELEKTRONIKA in Russian Vol 22 No 9, 1979
pp 23-28 manuscript received 30 Jun 78, after revision 4 Dec 78
[Article by S.N. Arzhanov, S.A. Markova, S.B. Rayevskiy and V.Ya.
- Smorgonskiy]
[Text] An algorithm is suggested for calculating in a quasi-static approx-
imation the unit-length parameters of shielded band lines. The results are
given of a numerical investigation of the algorithm suggested, and esti-
mates are also made of its accuracy and of the agreement of the solutions
arrived at.
The wide employment in practice of shielded band and microband lines has
made topical the problem of creating general (basic) aZgorithms for cal-
culating their key characteristics, distinguished by uni.versality, simpli-
city of the structural layout, good speed of response and high accuracy.
In this study is discussed a rectangular shielded band line, shown in fig
la, which can be regarded as the basic model for the calculation of more
complex band lines. Mention should be made of the fact that a particular
variant of it--a rectangular coaxial line--has also found independent appli-
cation in attenuators, frequency multipliers, and in microwave gas spectro-
- scopy--as absorbing Stark elements, in different oscillatory systems em-
ploying solid-state electronic devices and the like.
Let us divide each quadrant of the cross section of a microband line (fig
la) into three regions, I, II and III. The general solution to the Laplace
equation (the calculation is made in a quasi-T-wave approximation) satis-
fying the condition ET/S 2= 0 (S1 2 represents the surfaces of the
- line's conductors) we wri~~ in the foi~n:
. ~ ~ .
- IIZ~ Uo b'ds y-}- ~\Am ch d"t x-}- A~, sh ~ xl X
1
X sin nm (62 - y) e r~~:~
d, ,
- 20
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
~
~z = Uo b' - y �2 X ~ Cm sin ~ (bx - y? Sh dm (aZ - x)
dz d,
~i
~
~ D sin nn (a, - x) sh ~n (b e r~~: .
n d~ d! ' - y) ,
~ '
I1Zs =~o az d~ x+~\Bn Ch ~L+t 9-~- g sh d y~ X
s
~i
nn _ro~l~�z
X sin a(a2 - z) e ~
i
where IIe , 1Ie and IIe3 are electrical Hertz vectors in regions I, II
and III ~(~ig la~J, resp ec~3vely; d~ = a2 - al ; d2 ='02 - bl ; and ee is
the effective value of the dielec ric con.etant in the Iine.
ay ~
_ 0 ~
e , F~ _ ~
f01
I
�O1 0~ Ot E
z a) b) ~A
~
Ea �o~ -
A B
~
a~
Figure l.
From boundary conditions with x= a and y= b , util~zing the property
of orthogonality of eigenfunctions o~ regions I and III in intervals of
yE[b2,b1] and xE[a2,alJ , we get a system of linear algebraic equations
in terms of amplitude factors: ~
21
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-00850R040240050040-0
FOR OFFICIAL USE ONLY
_ m
AmC~i -f - Ama~~ e ~ (Bn Ba J'`"" _ ~ Um!
. a, ~
BnF'f BnF'~ ~ ~ ~Am Arr~ K~ - ~ Un!
?nal
~1~
where
Am = Am ~h ~ a~: A~ = A;n sh d a!; Bn = Bn Ch ~ bt; ~
9
B~ = Bn sh d~ b,; as = th ~ a~ ~ cth ~ a!; ai = cth d~
e2 cth ~m a. t}i ~ b-{- Ea cth nn b~ ~l = Eth nn b! +
- -f- ~ da i = d~ i ~ d! s ~
~ cth d! bi; .l; _ m C0 ~ sh ~ (b,
sh ~ d=
~
~ ~ 2 nn cos nm cos nn
d
- y) S~ X ~ (bz - y) d9 ~ - nm = nn ~ '
+
a
K,,,,, _ 2 m cos nn Sh n!n ~a~ - x) sin (a2 - x) dx = .
ds n Sh ~ d1 da ds
s
o,
2 d cos nn cos nm
2
- ~ i
- 2 2
- ldml d1 /
22
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-00850R040240050040-0
' FOR OFFICIAL USE ONLY
b '
Umi n~1dz ~(y - b~ sin d~ (y- b~ dy =~nm~2 d cos nm;
b!
a
2U� nn 2Uo
Uni n~1~ ~(x - a~ sin d! (x - a~) dx =(nn)2 d! cos nn.
a,
Systems of equations from boundary conditions in other quadrants are
= arrived at from (1) by the substitutions:
For the second quadrant:
ct~ u2, E~ --s~ E2, e3 --Y e3, BA F'i, B~ Fi, Am
Am;
For the third quadrant:
� � . .
a~~ uz~ ~ ~s ~3~ Bn -s~ F~, B~ - - F~~
~ A"'-' Ek, Am Ek, b2 6z, E~ --a e~;
For the fourth quadrant:
b;-~ bz+ e~ Ez ~ e~ et~ B~ - Bn~ A,~ --y Ek; _
Am --Y E,t. ( 21
_ We solve the general system of equations by the reduction method upon con-
dition of equality of individual amplitude factors in adjacent. (in relation
to quadrants) partial regions, which corresponds to fulfillment of boundary
conditions for tangential components of the electromagnetic field at the
limits nf x= 0 and y= 0. -
Solutions of the system (amplitude f actors) are used for finding unit- -
length parameters of the line under consideration (ca~acitance and wave
impedance), the equations for the calculation of which are arrived at on
the basis of the relationships:
C=I-eJ E�ds~ U; ~ � s
~ S e C'
(3)
23
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
where s is the surface of a unit of length of the 11.ne's outer conductor;
E is the electric f ield component normal to S; and U i~~ the difference
in potential between conductors.
The equation for calculating unit-length capacitance characteristic of the -
- first quadrar~t of the line's cross section (fig la), arrived a~ oy integra-
tion in (3) in terms of the surface of the outer conductor, tias the form:
~ Ci = Cii -1- C!z -I- Cia'f
~
C!, = E~ ~ A~, th ~ a~ Am cth ~ as - ~ , ~
shnma
d i
s
~i
/ ~
C~~ = e: 1 f dt dz
l-~- (Am ~1m} ~th d d~ - ~~nm _
2~ dz d~ ~ a sh d d~
~
~
-f- ~(Bn -f- B~) ~th nn d2 cos nn. ~
~
l...~ d1 Sh dn dx ~ .
~i
~
Cf3 = E3 d~ -I- Bn th d! b~ Bn cth d1 b~ - nn .
~ sh -d- B, .
~
~i
(4) -
Similar in appearance are the equations for calculating the unit-length
capacitance of the remaining quadrants (C2 , C and C4). They are
arrived at from (4) by substitutions in (Z). ~u~nation of these capaci-
tances gives the total capacitance of the line. The equation for the wave
impedance characteristic of the first quadrant will be
~ Ei ei~ + ~ ELZ ~ N ~g . .
, 1~~, l` ~
~ z~,~
- FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
~ The values of W2 , W3 and W4 for r.he remaining quadrants are detertnined
by means of (2). The total wave iinpedance of the 11ne is computed Chus:
1/W = 1/W1 + 1/W2 + 1/W3 + 1/W4 .
For the purpose of analyzing the agreement of solutions and the accuracy
provided by this algorithm, detailed numerical investigations were made
for the case of a uniformly filled line with a symmetrically arranged
inner conductor. Here the systems of equations for individual quadrants
proved to be identical and in calculations was employed only system (1),
where A~*=Bn=O , e1= e2= e3=1.0 , and m=n .
The results of an investigation of the agreement of solutions are given in
f ig 2, where curve C corresponds to a calculation of capacitance in terms
of a unit-length charge on the surface of the inner conductor and curve C,,
in terms of a charge on the surface of the line's outer conductor. For `
the purpose of analyzing the accuracy of the algorithm, the calculated .
values of W were compared with the results of [1,2J arrived at by means
of the method of conformal transformations (these results can be regarded
as precise), as well as with values arrived at by other methods [3,4].
The dependence of the integral characteristic (C/4t) of the line on the
number, N, of the approximation (fig 2) testifies to the good uniform
agreement of solutions; the results arrived at by means o~ this algorithm
practically do not change when N> 10 . It is obvious from this figure
that in integrating for the surface of an outer conductor not containing -
singular points, the unit-length capacitance rapidly approaches (curve C2)
the precise value of [1] even in the case of a very thin inner conductor.
The second dependence (curve C1) shows considerably worse agreement of the
results with the precise values.
E CZ 1'-QO/ 4E ~ o~= 0,l E �'-O,B
0,4 O,B r ~
C,
_ oz o,y ~ y
~
~ .f ~ /5 N ~ 5 /0 /S N ~ 5 /0 /5 N
Figure 2.
A detailed~ comparison (tables 1 and 2) of the results arrived at by means
. of the algorithm suggested with the precise values [1,2] demonstrates their
good agreement with any geometrical parameters of the conductQrs.
In table 1 are given also the results arrived at in [3] by the net-point -
method by using 60 horizontal divisions in a net. From comparing theae ~
with the precise values, we see that even in such a high approximation the
25
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
net-p~~lnC methucl l.e conRiderably inf.erior in accuracy to the nlgorlthm
suggested. In the same table are given the results arrived at in a 20-th
approximation based on the procedure in [4]. The approximation number
agrees with that in which the authors' results presented in the table were
arrived at.
Table 1.
- 4a
~;~OM T04N0 R~?~ nax- A?p~~ g, ~pln~, ~ M no ~Al~l~
�i/as Bowmon "d~ anro- ~ oo rpa~
jKOe o~ e~er~Axtce [4]
PHTM ~
_ 273,39
0,01 270.69 260,70 3,69 - _ I~~~ 1,57
p~( 132,65 132.1 0,41 -
p,2 91,11 90,94 0,19 92,~ 0.98 L05,58 i~~ 8
p,3 ~,87 66.81 0,09 67,5 0,94 73,68
1.37 58.08 16.58
p,4 49,82 49.79 0.06 50,5
0,5 36,S1 36,80 0,03 37,0 0,52 45~82 24~48
Key:
1. SZ , precise 4. SZ , according to procedure in [4)
2. S~ , our algorithm
- 3. St , graphs in [ 3]
Table 2. -
OM AaN� Oe~
al/b= Q~/6: bi/bt H~t~ anro- ~2~
~ PNTM
- ~~.~.~5~ C~yJ.~y'4 ( V~'! I Yy~~ 5~~~
0,20398 0,79t68 0,4 59,87 60,0
0,06666 0,60036 0,4 67,90 68,0
, Key:
1. St , our algorithm 2. S2 , in [2]
rrom table 1 it is obvious in addition that the results of study [4] differ
considerably from the precise values, which can be explained by a certain
incorrectness in formulation of the problem: In describing the field in the
side region, the term of the series was not taken into account which pro-
~~ides for a fixed difference in potential between the line's conductors,
and the unit-length capacitance was calculated in terms of the charge in
the line's inner conductor. The data from [1,2] presented above were
26
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
arrived at with precise equations, but it must be mentioned that for the
purpose of practical use these equations are rather cumbersome (they con-
tain elliptic integrals of the first kind), they do not make it possible
efficiently to calculate the distribution of the field in terms of the
line's transverse section, and are suitable only for the calculation of
rectangular lines with symmetrically arranged conductors.
The algorithm suggested is free of these disadvantages and ensures high
precision of the solution (tables 1 and 2) with a relatively not too great
number of computations. Calculation of one variant of a line on a BESM-4M
computer (the program is written in the algorithmic language Fortran-20)
takes on the order of 90 s.
Based on the algorithm suggested, it is possible to construct calculation
_ equations for more complex (fig lb, c and d) band lines in a quasi-T appro-
ximation. If it is taken into account that along axis of symmetry AB in
fig lb, c and d the conditions of an electrical wall are fulfilled (with
excitation of odd parity), the line can be regarded as two individual lines,
the procedure for the calculation of which is presented above. In a simi-
lar manner are calculated the unit-length parameters of the lines in f ig
lb, c and d with even-parity excitation. In this case Hertz vectors in
partial regions of the f irst quadrant of the line's cross section (fig la)
are c~rritten in the form:
~
nZ? = Uo b2~ \Am Ch d x-}- A;n sh ~ x~ X
8
m=1
~tm _r~uYe9?u . .
X sin d(bz - y) e ,
s
~
n2z = Uo b2 d2 9-{- ~ Cm sin ~'n (bi - y) Sh d - x) -I-
a ~
M=t
~
~ Dn Co5 2d (a, - x) sh 2d (b~ _ y) e-c~~ z,
~ '
~
~
_ ~s = Vo -I- ~\Bn ch ~n y-{- B' sh nn yl X
2d! n ~i ~
X cos 2d~ (az - x) e roY-y~ :
m= 1,2,....R= 1, 3,5,....
- 27
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
After writing boundary conditions with x= al and y= bl , it is possible
to arrive at equations similar to (1). Also constructed is a system of
equations for the fourth quadrant. The total system of equations is arrived
at by a combination of the system for the first and fourth quadrants with
the boundary condition when x= a2 conforming to a magnetic walli and for
the second and third quadrants witFi boundary conditions for x=-a2 con-
forming to an electrical wal1.
Bibliography
1. Bowman, F. "Introduction to Elliptic Functions," Dover, New York, 1961.
2. Biblet, H.J. The Exact Dimensions of a Family of Rectangular Coaxial
Lines with Given Impedance," IEEE TRANS., 1972, MTT-20, N 8, pp 538-541.
3. Metcalf, W.S. Characteristic Impedance of Rectangular Transmission
Lines," PIEE, 1965, 112, N 11, pp 2033-2039.
4. Br~ckelmann, W. "Wellentyren auf der Streifenleitung mit rechteckigem
Schirm" [Wave Types in a Strip Conductor urith a Rectangular Shield],
- AEU, 1967, 21, No 12, pp 641-648.
COPYRIGHT: IZVESTIYA WZOV SSSR, RADIOELEKTRONIKA, 1979
(53-8831]
CSO: 1$60
8831
28
FOR OFFICIAL USE ONLY .
~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
Electromagnetic Wave Propagation
UDC: 621.371.25
SCATTERING OF MODES OF THE WHISPERING GALLERY TYPE BY IONOSPHERIC
INHOMOGENEITIES ALONG A GEOMAGNETIC FIELD
Moscow IZVESTIYA VYSSHIKH UCHEBNYKH ZAVEDF:NIY: RADIOFIZIKA in Russian No 10
- Vol 22 1979 pp 1195-1204
[Article by A. F. Belenov and Yu. V. Chugunov]
[Text] This article examines scattering of modes of the
whispering gallery type by inhomogeneities extended along
a geomagnetic field. Coefficients of mode excitation are
calculated with scattering of one mode into another and
with scattering of a plane wave by an extended inhomogene-
ity (analog of "angular" scattering in a geometrical-
optics approximation). Problems of mode selection in
such scattering are discussed.
As we know [1-4], ionospheric waveguides (IVK) can be excited due to the
inhomogeneous structure of the dielectric constant of the ionospheric "
plasm:~. t~any studies have been dedicated to this question. For example,
the aiithors of [2] developed adiabatic theory of capture of radio waves
in IVK due to refraction on regular horizontal ionospheric discontinuities,
while the authors of [4] solve a problem in approximation of geometrical
optics on excitation of IVK as a resu3t of scattering of radio waves on
random inhomogeneities of the ionosphere; the authors of [3] discuss the
results of a waveguide approach to the problem of the effect of in-
homogeneities of ionospheric plasma on the spectrum of natural modes of
1VK.
_ In this article we shall examine the problem of excitation oF whispering
gallery type modes in IVK [6]. Principal attention is focused on waveguide
excitation by scattering on inhomogeneities ext?nded along a geomagnetic ~
field. It is demonstrated ho4~ anisotropy of inhomogeneities influences -
_ conditions of most optimal "washing down" of IVK. At the same time we
shall note that scattering of radio waves by isotropic inhomogeneities in
IVK was examined in [9].
1. ~Je shall represent an IVK containing inhomngeneities in the form of
- a homogeneous and inhomogeneous part. The model of the homogeneous portion
29
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200050040-0

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200050040-0
FOR OFFICIAL USE ONLY
of the waveguide here was selected in the form of a cylindrical cavity with
i=1 when r< R1 and 4=~ 1.i - -l ~ ~ E~ ~ J ~z, ~I.~~cl_,
r ~
where
.~t, / !z~_ i~(~~ ~~i1=',
integration is performed on the volume occupied by the inhomogeneity. Ex-
pressions (2), (3) and (S) make it possible to calculate coefficients of
excitation D+Z of waveguide modes with sc:attering by an inhomogeneity
located in a waveguide channel.
2. Let us examine the case where an incident wave is one of the natural
modes of an ionospheric waveguide, that is, E~,~ =ES.
Utilizing Langer-(rok) uniform asymptotic behavior [7] for J.-t ('C) when
y�l, Y.` r, from (1) ,(3) ,(5) under condition LXSin a (1/3) (k0-2R~)1~3
~ae obtain (LX characteristic scale of an inhomogeneity across a
geomagnetic field, q ~eomagnetic field angle of inctination3
U.~ ~-Q~,