RADAR

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-01043R001800200001-5
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 188
• Document Creation Date: January 4, 2017
• Document Release Date: March 14, 2013
• Sequence Number: 1
• Publication Date: January 1, 1956
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
2: 2PT.T. VT
.-1ADA.T.
? AO
3ection 621. Antenna. :adiation Patterns
Antennas employed in radio stations are classified by type of transmission and
reception as directional or non-directional.
1:on-directional antennas emit electromagnetic energy uniformly in all direc-
tions and receive it regardless of the direction from which the incoming radio waves
arrive.
On the other hand, the magnitude of the e m
. .f. induced in a directional receiv-
ing antenna in a given position and constant intensity of the swept field is highly
dependent upon the direction of the received wave, i.e., on the orientation of the
receiving antenna relative to the transmitting station. A directional receiving
antenna amplifies the reception of waves arriving within a narrow directional angle,
and weakens reception of waves coming from other directions.
Likewise, when directional transmitting antenna is in a given pcisitionl_the in-.
tensity and strength oT the electromagnetic field which it emits depends, at points
equidistant from the antenna, on the direction. A directional transmitting antenna
concentrates the emission into a narrow beam in a single direction and weakens the
radiation in other directions.
The application of the so-called reciprocity theorem is very helpful in c..ntenna
engineering. This theorem states that the nature of the directional activity of an
STAT
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
antenna is absolutely identical, whether it is used for transmission or reception.
The directivity of an antenna is conveniently depicted by a radiation pattern.
The radiation pattern or characteristic of an antenna is the ratio of the emitted
antenna field to the direction toward points of reception at identical distances from
the antenna.
distinction must be made between the radiation pattern for the field inten-
sity E (the e.m.f., 7, induced in a receiving antenna) and that for the radiated
power P (picked up by the receiving antenna).
know from the theory of the electromagnetic field that the power
of a field
is proportional to the square of its intensity. Conseouently, as the direction
chances, the power of a field changes very much less than its intensity. The pattern
of an antenna is always much more distinct relative to the field power than to the
intensity.
The directivity of an antenna is complex in nature and manifests itself differ-
ently in the horizontal and vertical planes. Therefore, its pattern in the horizon-
tal plane (along the azimuth T) is different from that in
the elevation 0).
For the purposes of
tenna radiation patterns
the vertical plane
(along
generalization from these concepts, we shall speak of an-
at -a-given-angle a, (instea.a-OT T'Or 0)-mlatiVea:magni-
tude A,-of interest to us (instead of the intensity El or 'power P).
A pattern of this type may be expressed analytically both in absolute scale,
in the form of the equation AF( a), and in relative scale, in the form of the equa-
tion =F1(a). Diagrams of radiation patterns plotted on absolute or relative
max
scales are geometrically similar, since they differ only in scale, and merge into
2n ienticrl.:1 3ca103 :1.70 rose: accnrdin:ly. The term radiation pat-
nThC!P
'-1 1,_.csorto Ji a n,dar 3:3ter, o.7 coor,linates.
In an:- '7,i-1m radiation pattern, t:e 71arnitucle o: fL or 07
Amax
A
may be deter-
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
t"I
'mined for the given angle a.
In plotting a radiation pattern, its center 0 is chosen, and the line from which
the angles a are to be measured off, is indicated. In order to determine the magni-
tude A or
'A by means of the antenna pattern in a direction in which we are in-
Amax
terested, a radius-vector is drawn from the Icenter of the diagram at the desired
angle a. The length of the vector determinels A or A for the angle, a of interest
Amax
a)
b
Line
from
Amax which
angles
are
measured
Fig.6-1 - Antenna Radiation Pattern in Polar Coordinates
a) In absolute scale; b) In relative scale, with beam width y.
to us. The ends of all the radius-vectors are defined by the outline of the pattern,
and they begin at the starting point of the pattern.
Hote-that the-definition-of.the-ioatteam is lierY ofteh-determined by the -sOL.
called beam width, by which is meant the angle detormined by those points in the
pattern where the radiated power is equal to one-half of the maximum power in the
pattern (Fig.6-1).
The radiation pattern of a sending antenna may be determined experimentally as
follows. First, points are selected around the station antenna CI at uniform dis-
tances therefrom (say, one km. ) in accordance with the wanted angles a (Fig.6-2).
A special indicator is used to measure the magnitude of A at all these points suc-
cessively. The experimental data (a and A) are tabulated, and this compilation is
used to plot the pattern on an absolute scale A=F(a). To plot the pattern in rela- STAT
: i;?1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy
77.2
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
tive scale A we find the maximum radius-vector A on the prior pattern. The
Amax I ma:,
magnitude A in the Table for all angles a is divided by A .
max
It should be noted that when a directional transmitting antenna is rotated, this
motion proceeds in the direction of maximum emission. Thus, the rotation of a direc-
Indicator of
magnitude of A
?
- Eethod of Determining Fig.6-3 - Directional Antenna
Radiation Pattern Pattern
tional antenna is equivalent to the rotation of its radiation pattern in the opposite
direction.
In non-directional antennas, the magnitude of A is not dependent upon the
angle a. The radiation pattern of such an antenna is a circle (Fig. 6-3).
--Br-plotting -the-vertical pattern of -an antenna, it is necessary to take- into?
account the effect of the ground, consisting of the reflection by the earth's sur-
face of radio waves radiated by the antenna toward the ground. Let us look at
Fig.6-4a. At the point Ts, the direct antenna ray and the reflected ray are super-
imposed. The resultant field at the target will depend upon the difference in
course of the direct ray ATs and the reflected ray AOTs.
The reflected field may be replaced approximately by a mirror image of'the an-
tenna AZO' constituting an identical antenna, located below ground at a depth equal
to the height of the real antenna above ground (Fig.6-4a). The currents in the hori-
zontal portions of the antenna and in its mirror reflection must be regarded as flow-
4
STAT
Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
ing in the same direction, while the currents in the vertical portions of the anten-
na and its mirror reflection are to be regarded as flowing in opposite directions
(Fig.64.b). The ground itself is entirely disregarded.
In certain directions, the direct ray and the specular reflection ray are in
a)
c)
Direct ray b)
Reflected ray
Fig. 6-1;. - Effect of Ground on Vertical Antenna Pattern
phase on arrival at the point of intersection, causing an'approximate doubling of
.E-h-e?gain-Of-the-resultant-field, In_certain other directions the (131,ct ray and the
specular reflection ray arrive at the point of intersection in counterphase, so that
the resultant field is zero. Ground reflection gives the vertical pattern a multi-
lobed appearance (Fig. 6-4c), the number of lobes in the vertical diagram being equal,
as shown by the antenna theory, to the number of half-waves represented by the ele-
vation of the antenna above the ground. Thus, if the antenna is elevated above
X
ground to a height equivalent to 3 7i, where A; is the working wavelength of the sta-
tion, the vertical pattern of the antenna will have three lobes.
The basic requirements for the antenna system of a radar :-Ly be formulated as
follows:
5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
,
a) it must provide a radiation pattern of the desired shape (and sharpness), as-
-.suring the required concentration of radiation in the direction of the target, the
given accuracy of measurement of the angular coordinates of the target in direction
--tfinding, and the required bearing discrimination along the angular coordinates;
b) its design must be as simple as possible, mechanically strong, small and
-- light-weight, handy in terms of rapid assembly and disassembly. These are the funda-
mental factors in determining the time required to mount the station for action from
- its transport set-Up, and vice versa. The design of the antenna system must assure
__the.required method and speed of sweep scanning by the electromagnetic ray, corre-
sponding to the given type of radar.
2
_ Section 6-2. The Half-Wave Dipole
The basic component of most radar antennas is the half-wave dipole consisting
of two identical metal tubes, usually silverplated, one-quarter the length of the
--broadcast wave. The total length of the half-wave vibrator is X
Insofar as distribution of the current along its length is concerned, the half-
- wave dipole is similar to a quarter-wave segment of a two-wire line open at the end.
The site of the current loop and the voltage node is at the center of the dipole
_
-----whild-the-VPitige-laop and the current node are at the ends (Fig. 6-5a and b).
Electromagnetic oscillations around wires 1 and 2 of the quarter-wave segment
are generated in antiphase to each other and compensate each other completely, so
that radiation by this segment is quite insignificant. To do away with this compen-
sation, wires 1 and 2 are rotated and placed in a straight line. A quarter-wave
segment so rotated is converted to a dialf-wave dipole (Fig.6-5b). In both halves of
the dipole the .currents move in the same direction, an entire half-wave of current
or voltage distribution being laid out along its entire length. The lines of forces
of the electrical field of a half-;wave dipole are directed primarily along its wires,
while the lines of force of the magnetic field surrounding the wire in concentric
6
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
circles, are normal to the mires.
The plane passing through the center of a half-wave dipole perpendicular to its
wires is generally termed the eauatorial plane.
The intensity of the electrical and magnetic fields of a dipole is dependent
upon distance. At a given distance from the axis of the dipole, the electrical field
a) b)
Fig.6-5 - Conversion of a Quarter-:jave Open-End Line (a)
to an Ordinary Half-Wave Dipole (b)
in all directions in the plane of the center-line is of identical intensity Ecp
(7ig.6-6a). The same may be said of the intensity of the magnetic field Hcp, and the
radiated power density, i.e., the Umov-Poynting vector modulus Pq, which is determined
by the product dr-flieThifeetiTE.1.111e8--or-Eip-ard--11-cF----Thu.s,--a-lralfzwave-clipoi-e---ra&.i------
ates an electromagnetic field equally in all directions in the center-line plane
perpendicular to the dipole axis. This means that in the center-line plane, the
half-wave dipole has no directivity either in field intensity or radiated power.
The center-line plane pattern of a half-wave dipole is a circle (Fig.6-7a).
Any plane passing through the dipole axis is called a meridional plane. The
radiation of a half-wave dipole in a meridional plane is already clearly directive
in nature, as it is dependent upon the angle S between a line perpendicular to the
dipole axis at its center and the direction which interests us (Fig.6-7b). In slur-
plified form, this may be clarified by the following example (Fig.6-6b). The inten-
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
sity of the. electrical field E0 determining the radiated power in the direCtion of
interest, is perpendicular to the Umov-Poyntinglvector Pe and is equal only to a pro-
jection of the vector E at the given point in a direction perpendicular to that of
interest to us.
The same may be said of the intensity of the magnetic field Il(). Figure 6-6b
a) E
b)
Lines of magnetic force
Lines of electrical force
Lines of electrical force Lines of magnetic force
_
Fiz...6,4 Dipole in the Equatorial (a)
and'lleridional Planes (b)
makes it possible to establish the cosinusoidal relationship between the angle 0 of
the electrical and magnetic components of the field of a half-wave dipole radiating
in the direction of interest to us:
EG = E cos 0
u?= cos 0
The maximum intensity of the electrical and magnetic field is created in the
direction perpendicular to the axis of the dipole (0 = 0). On the other hand, in
the direction of the dipole axis in which 0 = 9001 the field is absent. The field
(6-1)
8
STAT
(1 9
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
intensity pattern of a half-wave dipole in the meridional plane resembles a figure
eight and is shown in Fig.6-7b as the fine outline.
At the given point, the Umov-Poynting vector is proportional to the square of
the intensity of the electrical or magnetic
field at the same point. Therefore the
radiation pattern of the radiated power of
a half-wave dipole in the meridional plane
(heavy line, Fig.6-7b) is sharper than the
pattern discussed above, as it is subject
a.)
b
?????11.....
11111111?1...
to the square of the cosine:
Pe = P cos2O
(6-2)
. Be it noted that the considerations
we have set forth lead to a clear concept
of the directivity of the half-wave dipole.
However, these considerations are not to
be considered wholly rigid. In actuality,
the electromagnetic field created by a
half-wave dipole at the point of interest
to us is the-Mmaf.tsullA74.4mpoRitirin o?
f.
fields induced by all of its components of
Fig.6-7 Half-Wave Dipole in the
infinitely short length, acting as elemen-
Equatori4 (a) and ileridional (b)
tary independent sources of radiation. In
Planes
view of the ineouality in the distribution
of current along the length of the dipole, the elementary fields at the given point
differ from each other in magnitude and also differ in phase due to the difference
in distanceslfrom the various components at the given point. Therefore, the actual
directivity of a half-wave dipole is somewhat less than would be indicated by a sim-
ple cosinusoidal equation in the meridional plane, and the pattern emerging is slight-
STAT
9
_.?
npciassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
34, flattened. On the same principle, a half-wave dipole reveals a barely noticeable*
directivity in the equatorial plane as well. ::owever, for all practical purposes
this may be left out of consideration to simplify-the reasoning.
row let us turn to the question of the resistance of a half-wave dipole to radi-
ation. 2esistance to radiation by an antenna RE is that conditional active antenna
resistance at which the current feeding the antenna I develops the power PE radiated
by the antenna in the form of radio waves.
According to this definition, the resistance to radiation is equal to the ratio
of the total radiated antenna power to the square of the true magnitude of the cur-
rent at the point where the antenna is fed by the generator oscillator:
PE
RE - (6-3)
p.
As shown in the theory and confirmed in practice, the resistance to radiation
by an isolated half-wave dipole is :E = 73.2 ohms regardless of the length of the
411 operating wave, so long as the dipole is one-half its length and regardless of the
manner in which the dipole is positioned, provided that only a single dipole is being
used. The resistance to radiation of the half-wave dipole determines the active
component of its total input resistance loading tne
A major shortcoming of the ordinary half-wave dipole is its low resistance to
radiation and therefore its low resistance to the load on the supply feeder. In
order to attain a high traveling-wave ratio in the supply feeder, its wave resistance
should also be of the order of 73 ohms. Ilanufacture of a feeder of such low resis-
tance to transmit power of the order of hundreds of kilowatts is quite difficult.
In order to obtain low P = l'1
4 ,,
1 , it is necessary to make sure that the linear in-
ductance L1 is low and that linear capacitance C1 is high. Both demand the use of
thick wires closely spaced, and this in turn limits the permissible working voltage
on the feeder and the power it delivers (see Section 4-7).
folded dipoles instead of the ordinary kind has two ot-TAIrs:
10
Declassified in Part- Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-016-43R001800200001-5
11
STAT
Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
(!s
a) to increase the radiated power with no change in the current in the feeder
line, i.e., to increase the resistance to radiation;
b) to make it possible to use feeder lines of higher resistance without the
need for resistance transformers.
The folded dipole may be regarded, in terms of longitudinal distribution of
X
current, as a two-wireline .-7) long, folded into a loop and shorted at the end
(Fig.6-8). Due to the equality of the currents at the upper and lower edges of the
b)
c)
Fig.6-8 - Folded Dipole
a), Shorted half-wave line; b) Same line, converted to a folded dipole;
c) Folded dipole as parallel connection of two elementary half-wave dipoles.
loop, the folded dipole consists of a system of two elementary half-wave dipoles
connected in parallel and combined into one. A folded dipole is therefore sometimes
called a double doublet or a "bivibrator".
The fields excited by the upper and lower elementary dipoles constituting the
folded dipole are superimposed. The voltage of the resultant field of the folded
dipole will at any point be twice the field intensity of the elementary dipole, i.e.,
Declassified in Part - Sanitized Copy
e
???
vs
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
C_
'it will be 2E1. Therefore the power density and the total radiated power of a folded
dipole P2 are proportional to the square of the voltage (2E1)2 and will come to 413z1.
i.e., four times the power radiated by an ordinary half-wave dipole. Given the same
amount of current in the feeder, we airive at the following expression for the radia-
-Lion resistance of a folded dipole:
Pi
PE
It= = 4 - 4Ryi (6-4)
12 12
1
This means that a folded dipole burdens the feeder with an active component of input
impedance Rin = 4 X 73.2 = 300 ohms, or four times that of an elementary half-wave
, dipole while on supply by a high-resistance feeder (o = 300 ohms) it does not re-
-
_
2 quire that a go-and-return resistance transformer be cut in, as it performs that
_
' function itself. For this reason a folded dipole is also sometimes called a loop
dipole.
The radiation pattern of a folded dipole does not differ in any way from that
of an ordinary half-wave dipole, which we have already examined.
.Section 6-3. The Functioning of Half-Wave Dipoles in the Antennas of
slab
Very-Short.4lave Systems
The major shortcoming of the single half-wave dipole is its relatively broad
radiation pattern, inadequate for the precision of measurement of angular coordi-
': _mates and the degree of concentration of radiation toward the target usually re-
quired. The power gain of a half-wave dipole is only 1.64. A half-wave dipole is
therefore to be considered a very primitive antenna and renders the use of single-
unit installations impractical.
The behavior of a half-wave dipole changes very much for the better in the
presence of other dipoles arranged in proper fashion.
Let us examine a system consisting of two-co-linear half-wave dipoles and at
their centers separated by the distance d. If the feed to the dipoles is cophasal,
12
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
346
Declassified in Part - Sanitized Copy Ap?roved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'i.e., if the phase of the current in both is identical, the electromagnetic fields
induced around them will also be in phase. In accordance with the equation for the
electrical component of the field of a radiated radio wave, then at r distance from
the dipole, the following instantaneous voltage will exist:
e(r) = Em( r) \ sin (wt
2 TE
2
The phase of the oscillations lags in proportion to the distance from the source
of radiation, while the amplitude Em(r) declines as the first power of the
It may be reckoned that at a sufficiently distant point in space the electromagnetic
waves may be taken as being emitted. from the antennas in parallel fashion, so that
in the general case, the phase difference in the plane shOwn in Fig.6-9 will depend
distance.
on the angle 0;
Ar = d sin (6-5)
The difference in the path of the rays will cause, at the point in which we are
interested, a difference in the phases of the fields radiated by the antennas:
2n Ar - _ 2E- d sin
x
(6-6)
Therefore, at the point in which we are interested, the antenna fields form a geo-
metric progression, the result of the superimposition depending upon the angle 0 so
tnat in difierbration-s-th-&-figUe's for the same distance from the radiating
system will be different.,
The difference in the amplitudes of the fields may be ignored in view of the
small size of Ar relative to the distance from r to the point of interest. In ac--
cordance with eTI.(6-5) and (6-6), Ar = 0 and AT =. 0 in the direction perpendicular
to the axial line of the system (0 = 0), so that the shaping of the fields is sim-
plified to an arithmetical progression, and the resultant field will be twice as
powerful as that of a single dipole. In any direction other than the perpendicular,
some difference in path and phase difference appears. The resultant field will not
be twice that of a unit antenna, but somewhat less, as the progression will be geo-
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'metrical and not arithmetical. Thus, a system of two cophased antennas clearly pro-
duces a more elongated pattern than does one from a single half-wave antenna.
In certain directions, the difference in path Ar . - X
2 and the phase differ-
ence is AT -
Fig.6-9 - Two Antennas in Colinear Array
2n X
X 2
In such directions the antenna fields reach the point of interest in antiphase,
and superimposition results in complete compensation, so that the resultant field
of the system is zero. The angles 0 in these directions determine the angular boun-
daries of the lobes of the system's pattern. The width of the lobe will depend upon
-the-distance d bu ween the-antenna-enters; anA.I/ron the_length of the working
wave X. In the direction of the axialaine the field of the system will always be
zero, since none of the antennas radiates in this direction.
Figure 6-10a shows the radiation pattern of a system of two cophased antennas
X
? at the distance -7) between centers. When a large number of antennas is used, the
pattern becomes even more elongated.
Another diagram results when the phases of the feed to a two-antenna system
are in opposition and the currents in the antennas each show a 1800 displacement of
phase relative to the other.. The antenna fields will also be excited in antiphase. ?
In this situation there will be added to the phase difference in any direction a
constant compensating phase difference it, determined by the fact that the antESTAT
14
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
*oscillations are in opposing phase. Therefore, in the direction perpendicular to the
axial line (e = 0), the resultant of the field will be not at its maximum, but equal
to zero, since the total difference in phase will equal n. As previously, the field
System of
two cophased
antennas
a)
Unit
antenna
Ez.
b)
Pig.6-10 - Two-Antenna Patterns
a) With cophasal feed; b) With feed in opposing phases
Unit
antenna
System of
two antennas
in opposing
phase
will equal zero in the direction of the axial line of the system as well, as none
of the antennas radiates in that direction.
The maximum resultant field will be in those directions in which the difference
in the path of the electromagnetic fields of the antennas compensates for the oppo-
sition of their phases as driven. :filen compared to the precedin- case, the patter;
of the system doubles sylwetrically relative to the direction perpendicular to the
axis. The angle of inclination of the lobes of the pattern from that direction, and
the width of the lobes, depends upon the distance d between the antennas, and t.he
STAT
15
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
li?
wr.,.velengt! X.
74rure 6-10b shows the pattern when two antennas, the centers of which are ;X)
apart, are fed in opposite phase.
3ection 1"ulti-Antenna Cophasal Arrays
3tations broadcasting in the very short wave band employ multi-unit cophased ar-
rays, constituting systems of half-wave dipoles arranged in tiers. The dipoles in
each tier are al-ranged in a single ial line.
7igure 6-11a presents a cophased antenna consisting of four tiers of four di-
poles each. The antenna is fed through a branched system connected to the transmit-
To assure a high traveling-wave ratio matching quarter-wave feeder transformers
AT-1, AT-2 and AT-3 are inserted at the junction points of the feeder system.
It is necessary for the dipole feeds to be in phase in order to prevent the
appearance of e=ess lobes in the pattern, as we saw them in the case of the feeding
of a system of two dipoles in opposing phase. The traveling wave appearing in the
feeder system at the ends of each half-wave feeder segment presents instantaneous
phases differing by 180?. As shown by Fig.'6-11P, the length of each feeder segment
X
between tiers is -T. Therefore, in order to assure cophasality, the terminals of
every other tier are crossed on connection with the feeder.
As we have seen in the case of two cophased dipoles, the maximum radiation of
a cophased antenna occurs in two diametrically opposite directions normal to the
plane in which the dipoles are located.
An increase in the number of dipoles in each tier narrows the pattern in the
horizontal plane, i.e., along the azimuth. The patterns of all tiers are identical
and cophasal7 In space they are arithmetically superposed, amplifying the resultant
field of the system in the horizontal plane by an identical factor in each direction,
this factor being proportional to the number of tiers. Consequently, an increase in
the number of tiers does not affect the shape of the pattern in the horizontal plane.
16
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Ap
?
roved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
On the other hand, an increase in the number of tiers compresses the pattern in
the vertical plane, i.e., along the tilt, due to the existence in this plane of a
to transmitter
a)
......M.IM???????????11.1.1?1????????????????
Induction bridges
Fig.6-11 - Cophased Antenna:
a) Active array; b) Parasitic reflector
'difference in the paths of the rays of the dipqles in different tiers of a vertical
series. Just as the number of tiers does not affect the shape of the pattern in the
horizontal plane, an increase in the number of vertical series, I. e., of the number
I
of dipoles in each tier, does not sharpen the pattern in the vertical
Figure 6-11a shows the pattern in the horizontal plane, while Fig.6-12 illus-
STAT
17
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
trates a multi-dipole antenna. The width of the main horizontal lobe in the pattern
is about 150. The main lobe in the vertical pattern is of approximately the same
width.
When antennas with large numbers of dipoles are used, an even sharper pattern
may be attained, but in this case the antennas themselves become overlarge both in
height and width.
It is obvious that antennas having this type of horizontal pattern cannot be
used in radar. The radiation of the antenna is concentrated irrationally in two dia-
Fig.6-12 - Directivity Pattern of the Exciter of a Dalti-Unit
Cophased Array Along the Azimuth T (in Horizontal Plane)
opposite directions. In reception withih. an. antenna it is probable
_
that there -will be an error of 180? in azimuthal direction finding. Therefore, elimi-
nation of the second major lobe in the horizontal pattern is most important in de-
signing the antenna system. The solution employed in radar has been developed from
the work done at the Nizhniy Novogorod Radio Laboratory by E.A.3onch-Bruevich,
V.V.Tatarinov and A.A.PistolTkors, who developed the world's first directional short-
wave antennas.
Parallel to the plane of the ophased dipoles, which we term the exciter, and
at a distance of -X-' an identical system of dipoles is setup which is called the
parasitic reflector (Fig.6-11b).
The reflector dipoles Fe not driven from the transmitter. The current they
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
fr.
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
contain is induced by the electromagnetic field of the exciter. The reflector merely
tunes into the working wavelength by mans of induction bridges. The functioning of
an antenna array with a reflector is readily understood with the aid of a vector dia-
gram (Fig.6-131).
The magnetic field of the exciter A, designated as -17.aa in the vicinity of the
7
a)
b)
c)
Fig.6-13 A Directional Antenna Array Consisting of An
Exciter and a Parasitic :eflector 2L distant:
4
a) nelative Placement; b) Vector diagram; 0 Directivity diagram along azimuth
exciter, is in phase with the current T, in the exciter. By -traversing a distance
X
equal to -47 , this field, before driving the reflector, undergoes a 90o change in ?
phase in the direction of lag (-477,
az)*
On excitation by this field, current Yz is in-
duced in the reflector. By Lenz' Law, Lhe direction of these currents is such that
19
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
?
?
j
the field ;17izz they create, is, in the vicinity of the reflector, opposite in direc-
tion (antiphasal) to the field Taz exciting the currentin the reflector.
As the vectors Tizz and Tz coincide in phase, like the vectors Tiaa and -fa, the
currents in the reflector lag by 270? from the currents -fa in the exciter or, what
amounts to the same thing, are 900 ahead of the; currents. The fields Taz and lfzz
cancel each other out, being opposite in Phase.! Therefore, the array doesl not radi-
ate behind the reflector, and the second major lobe of the horizontal pattern is cut
out by the reflector.
The rield.z lags by 90? in traversing the distance of Xto the exciter, and,
- I
on reaching it, is in pLase with the field Hz ullich it amplifies by 1005. Thus, the
parasitic element reflects the energy radiated by the exciter, concentrating the to-
tal radiation of the system in the direction of the exciter alone. In so doing, the
reflector makes the exciter more directional, narrowing the field intensity pattern
by about 505 and reducing the radiated power density by 3A, approximate1.
In practice, complete cancellation of the field behind the reflector does not
occur because the currents induced in the reflector dipoles are 10 to 155 weaker
than those in the exciter dipoles. Therefore, !a. portion of the energy seeps past
the reflector and the pattern presents a small secondary horizontal lobe, although
it is, however, incapable of affecting the work of the station. In practice, the
reflector is so designed, by the use of induction bridges or other components, as to
minimize the field behind the reflector.
The width of the primary horizontal beam of a multi-unit cophased array may be
evaluated with the aid of the following equation:
yo = 51 x
(6-7)
in which b is the width of the antenna,
X is the operating wavelength.
Equation (6-7) is adequate for an antenna with a beam whose width does not ex-
ceed 25-30?.
20
STAT
_
- 1)e:classified in Part- Sanitized Copy Approved for Release2013/03/14 : CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
Often a flat metal grid, also -1 from the exciter, is used as reflector instead
4
of a parasitic element (Fig.6-1/..). This type of antenna system is termed a curtain
antenna.
The reflector also sharpens the vertical pattern of the system, eliminating ra-
diation in vertical directions behind the reflector, and elongating the radiation in
Fig. 6-14 - Antenna Array with :Tire-Grid P.eflector
the vertical directions in front of the exciter. As already noted in Section 6-1,
the ground has a pronounced effect upon the shape of the vertical pattern. Ground
reflection gives the vertical pattern a multi-lobular character, complicates the
finding of targets at certain elevations, particularly law-flying targets at con-
siderable distance from the location station.
Section 6-5. Director Antennas
Another type of directional antenna, widely used in stations operating in the
very short wave and microwave ranges, is the director antenna.
Director antennas are made up of wave ducts. Each wave duct consists of a sys-
tem of dipoles, the centers of which are placed along the axial line of the
duct (Fig.6-15). The transmitter drives a single half-wave exciter A of the usual
folded type. The remaining dipoles are parasitic, their currents being set up by
21
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5 _
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
'the electromagnetic field of the exciter.
he phase of the electromagnetic field of the exciter at a parasitic element is,
determined by the distance separating them. The e m f
. . . induced in the parasitic elel
ment coincides in phase with that of the electrical field radiated from the exciter:,
The current set up in the paraditic element will differ in phase from that of the
driving e.m.f. by an angle determined by the reactance factor in the input impedancei
C.
aaflector b)
Directors
7,xciter Support
- wave Duct (a) and its Pattern (b)
of the parasitir; element.
2?e r:agnitude and nature of the input impedance reactance of a dipole depends
upon its length, measured in fractions of the operating wavelength. If the length
of the parasitic element is a little over he current in the element will lead
X
the induced therein. Placed at a distance a little less than from the
4
exciter, the elongated parasitic element behaves like a half-wave parasitically-
excited dipole ? from the exciter, or, in other words, functions as n mirror.
::at is why the elongated parasitic dipole is called a reflector.
It may be shown by similar reasoning that the current induced in a somewhat
shorter dipnle Lill la:_; in phase behind the current In the exciter. .^,s a result,
the field induced by the shorter dipole is virtually in phase with the field radiated
by the e;:citer, and arlpli'ies its field. Thus, the shorter parasitic element guides
e field of the e::citer along the axis of the duct, and is therefore called a direc-STAT
0 0
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'tor. The superimposition of the dipole fields causes the wave duct pattern to have
a width of 40 to 500 becoming narrower as the number of dipoles in the duct increases.
The axis of the pattern coincides with the axis of the duct (Fig:6-15b).
A wave duct usually consists of one reflector and four or five directors in ad-
dition to the exciter. iro exact method yet exists for analyzing/a wave duct of any
given number of dipoles. zany problems in this connection are resolved by experi-
ment. The usual wave-duct design provides for a reflector length of 0.52 to 0.65X,
a director length of 0.3 to 0.46X, a distance of 0.15 to 0.20X from reflector to ex-
citer, and a space of 0.2 to 0.25X between adjacent directors.
1
Whatever is the length of the parasitic element, the point of induction of e.m.f.
is located in its center. Therefore, wave duct dipoles are fastened firmly at their
center points to an axial metal tube, and are not insulated therefrom, except for the
exciter.
Single wave ducts are rarely used in radiolocation. Director antennas usually
4
consist of 2 to /-..unit wave ducts, two to .a tier. The exciters In a given tier must
be driven in phase so as not to produce a pattern with two lobes in the horizontal
plane.
11
An increase in the number of ducts within a tier narrows the horizontal pattern.
T!-,e various tiers may be driven either in phase or counterwise. Switching tiers
from cophased driving to antiphased raises the vertical pattern to an angle above
the horizon. In circular scanning sets, this is employed to determine the elevation
of the target, as a signal of identical strength, when feed is switched, is receiv0
only at a specific elevation. The height of the target may be determined by the ele-
vation and the known range.
Section 6-6. Antennas with Parabolic neflectors
The shorter the working wave, particularly as one enters the very short wave
band, the more the methods of obtaining precision in direction approximate those used
STAT
23
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Cop Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
*in optics. From optics we know that if a punctuate source of light be placed at the.
focus of a parabolic reflector, the reflector will rearrange the light rays into a
parallel beam directed along the focal axis of the reflector.
In radiolocation systems, the ?punctuate? source of radiation is usually a half-
wave dipole placed at the focus of a parabolic reflector. The phases of all rays re-
ceiving.parallel reflection attain identical values in any plane perpendicular to the
focal axis of the reflector. Consequently, the shape of the radio wave radiated by
a) Parabolic reflector
Energy reflected from
paraboloid
A
focus
b)
Energy directly radiated
by exciter
Fig.6-16 - Antenna with Parabolic Reflector:
a) Geometry of radiation; b) Antenna radiation pattern.
such an antenna is flat. Only those rays driven by the exciter directly along the
focal axis, and not reflected, will be of different phase and will deform the wave
front somewhat. To eliminate this undesirable phenomenon, the exciter is surrounded
with a hemispherical screen to prevent direct radiation along the focal axis.
Figure 6-16 illustrates this type of antenna with a reflector in the form of a
paraboloid of rotation. The body
of the paraboloid may either be made of bent metal
sheeting (solid or perforated) or of wire grid.
Despite the fact that all parallel rays are of identical phase in the plane of
the front, their intensity will differ at various distances from the axis. This is
explained by the fact that the individual points of reflection of the paraboloid are
at different distances from the exciter. The intensity of the rays declines with
2/;.
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
r'.istance fron. the focal axis. The concentration of the most intensive rays around
the focal axis narrows the antenna radiation pattern and weakens the lateral lobes
a) c) d)
F1c.6-l7 - T-pes of Parabolic Reflectors
a) Parabolic c;-linder; b) Full paraboloid; c) Truncated paraboloid;
d) Lemon-segirnt paraboloid
of radiation. The axis of the pattern coincides with the focal axis of the reflec-
tor.
:hien the reflector has the shape of a paraboloid of rotation, the width of the
- Typical Exciters
:rain b'ean (the flare of the pattern) may be calculated on the radiated power at half I
? the marimun level (Fig.6-16b) by means of the following approximated formula:
25
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
4--
?
?
70 --
D
(6-8)
in which D is the diameter of the largest cross-section of the paraboloid;
A is the operating wavelength (in the same unit of measurement as D).
Equation (6-8) shows that the size of the reflector has to be 10 to 20 Ain
order for the pattern to be sharpened. Use of the microwave band makes it possible
to narrow the beam to tenths of a degree and to attain a power gain factor of more
than 4.000 with paraboloids of appropriate size. An advantage of antennas with para-
bolic reflection is that the effect of the ground may be cut out almost completely
due to the high directivity, assuming of course, that the sweep is not directly
along the surface of the land or sea.
Sometimes 'truncated paraboloids, fruit-segment pa'raboloids or parabolic cylin-
ders ray be used. All these types of reflectors are illustrated in Fig.6-17.
The parabolic cylinder has a cophased system of exciters along the focal line
of the cylinder, and is used when the need is for a beam narrow along the flare of
the reflector but wide along its length. When a parabolic cylinder is emplaced
vertically, the result is a beam narrow in the horizontal plane (along the azimuth)
and broad in the vertical plane (along the elevation). ghen a parabolic cylinder is,
in horizontal position, the result is a beam narrow along the elevation and wide
along the azimuth.
In conclusion, let us note that in antennas of this type, the primary radia-
ting element nay be either an exciting dipole, a horn (Section 6-8) and systems
thereof, or single-row slot antennas (Section 6-7).
Figure 6-18 illustrates typical exciters.
Section 6-7. Slot Antennas
Slot antennas are the invention of the Soviet scientist M.S.Neyman. A basic
slot antenna consists of a slot in a driven cavity resonator through which electro-
26
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
? r01^Trt
?
?
... magnetic energy from the resonator is leaked. The slot must be cut in such a mariner
that the electrical lines of force in the region of the resonator's electromagnetic
a) Surface of cavity resonator
? Distribution Distribution of
Slot of intensity, c)
. current, I,
El along length along half-wave
of slot dipole
b)
Fig.6-19 - Basic Slot Antenna
a) Notching of slot; b) Radiation pattern; c) Comparison with half-wave dipole
field are perpendicular to the length of the slot.
In order to attain effective directional radiation, the length of the slot
X
must be half a wavelength in open space (-2) , and the center of the slot must coin-
cide with the point of maximum intensity of the electric field, in which case its
distribution along the length of the slot (Fi.q.6-19a) will be analogous to the dis-
Fig.6-20 - Multi-Slot Single-Row Antenna
111 tribution of the current alohg the length of a half-wave dipole (Fig.6-19c).
For this reason, a slot of this type will behave like an equivalent half-wave
dipole. However, unlike a half-wave dipole, the electrical field of which is paral-
27
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
lel to its axis, that of a slot is transverse to the axis. Therefore, if a half-wave
eipole has directivity only in the plane travorsin, its axis, a slot on the other
Lan' ; is directional only in the plane perpendicular to that or the slot (Fig.6-19b).
Ur-her directivity is provided by a sin:-,1e-row multi-slot antenna, in which the
fields of all th.e slots are hept in phase by making the distance between slots equal
X
to a '-alf-lrave (TT) (Fig.6-20). Such an antenna behaves a single-tier cophased
arra7 of half-wave dipoles, but yields its maximum radiation in a direction perpen-
dicular to the plane of the slot. A single-row slot antenna is suited to utiliza-
tion as the al:citer of a parabolic cylinder reflector, being placed along its focal
line with the slots facing inward.
Let it be noted that the type of slots we have discussed may also be cut into
a waveguide.
Section 6-E% Horn Antennas
A waveguide through which electroragnetic energy is channelled, open at the
end, may also be used as a directional source of radiation.
As already stated in Section 5-8, the ratio of the electric field intensity to
the magnetic field intensity may be regarded as natural impedance not only in terms
of the cross-section of a waveguide tube, but relative to the open space into which
the enerr7 flows from the end of the waveguide.
If the natural impedance of a-waveguide at given frequency is determined by the
type of channeled wave, and by its length A in the waveguide (see eqs.5-15 and 5-16),
then the open space offers the following resistance to the electromagnetic wave,
which here necessarily acquires TEE structure:
Z = ? =
0 H
110
In view of the fact that the ragnetic permeability and dielectric constant of
28
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
'air are, respectively, of the following magnitudes
- 1 -9
= 4TE x 10-7 [Him] and so - x 10 Wm]
we obtain
Z0 = 377 ohm.
The most effective radiation from the end of a waveguide is that occurring
while there is no reflection from the end of the waveguide. In order to bring the
natural impedance of the waveguide into accord with that of open space, the cross-
section of the terminal section of the waveguide is gradually expanded in such
fashion as to permit this segment to play the role of a matching transformer. A
portion of a waveguide designed for matching in this fashion is termed an electro-
magnetic horn.
a) b)
Matching bridge
c) d)
From
oscillator
Resonator
chamber
Exciter
Horn
Fig.6-21 - Horn Antennas
a) Sectoral horn; b) Pyramidal horn; c) Conical horn; d) Driving by means
of resonator chamber
The basic types of horns are illustrated in Fig.6-21. A horn provides maximum
radiation in the direction of its geometrical axis. The radiation pattern of a
te by widening the flare and increasing the length, as this
29
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
creates optimum relationships between flare and horn length. A properly-designed
horn may be considered to provide cophasality of the electromagnetic field at all
points in the horn aperture, resulting in radiation of a wave with a flat front.
A horn antenna is of rather high directivity. However, its pattern includes
not only the main lobe but side lobes, the presence of which is due to the fact that
the electromagnetic oscillations at various points of the aperture are not perfectly
in phase. Horn antennas find their rain use in radiolocation as exciters of para-
bolic reflectors. The exciter may contain either one or a system of horns along the
1 focal line of the parabolic cylinder.
Let it be noted that not only waveguides may end in horns, but cavity resona-
tors driven in the usual way by the excitation of waveguides, (Fig.6-21d).
?
Section 6-9. Dielectric Antennas
Electromagnetic energy via:- be channeled not only in waveguides of air in metal
walls, but in dielectric rods surrounded by air. In the latter instance, the outer
surface of the rod will also serve as the boundary layer between two media with
different electrical characteristics. Therefore, the mechanism for the propagation
of radio waves along a dielectric rod will be identical in principle with the mech-
anism of propagation in a metal-walled waveguide..The only difference is that con-
plete internal reflection from theboundarysurface between the media is capable of
occurring in.the given instance only at supercritical frequencies. At subcritical
frequencies, the radio-wave energy is reflected from the contact surface only in
part, and is therefore distributed along-the dielectric waveguide, while in part it
penetrates through the boundary layer into the surrounding space, being refracted
in the process. Dielectric antennas take advantage of the refraction of radio waves
on passage into the air from the dielectric:
A dielectric antenna (Fig.6-22)-consists of a cone with a rounded end, made of
a weakly absorbilt dielectric (usually polystyrene). The conical shape of the radi-
_
30
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'ating dielectric provides the smoothest possible transition from the greater natural
impedance of the rod to the lesser natural impedance of open space. This weakens.
reflection from the rod cone and provides radiation of good effect.
The directivity of the dielectric antenna is governed by the phase difference
a) Netal chamber
b)
Cone of weakly absorb-
tive dielectric
Exciter Cross section of cone
Fig.6-22 - Dielectric Antenna
a) Design; b) Radiation pattern
of the electromagnetic fields radiated by different points at the surface of the
dielectric rod, and by the difference in path of these fields to the target point
in space. The maximum radiation is in the direction of the geometrical axis of the
rod.
A rod of considerable length and large diameter is needed to render the pattern
more acute. Let us note that as the length of the rod is increased there is a si-
multaneous improvement in the sharpness of the major lobe but also an increase in
the prominence of the side lobes.
Added directivity is attained by means of polyrod antennas set up side by side
at intervals equal to a whole number of waves so as to assure driving in phase, with
the use of a branched system of coaxial lines.
Section 6-10. Lens Antennas
The principle of operation of the electromagnetic lens is based on the fact
that the phase velocity of radio wave propagation in a dielectric is greater by a
31
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'factor of than its velocity in a cavity. In antenna engineering, lenses are
used to attain cophasality of the field at all points in the antenna output aper-
ture, thus simplifying the problem of assurinz a proper radiation pattern with an
exciter not very deeply positioned.
Thus, if a short horn is used in a horn antenna, it is difficult to attain a
narrow pattern even when the flare is wide. The difficulty arises from the fact that
under these conditions the wave front at the horn outlet will be spherical and not
flat, and the phase of the oscillation will differ from point to point at the outlet
aperture, due to the fact that at the edges it is farther from the exciting wave-
guide than at the middle. The result of the variation in phase of the oscillations
Horn
a)
b)
Fig.6-23 - Lens Antennas
Dielectric lens
a) Oielectric lens with components of equal width and rear profile forming
special pattern; b) Metal lens of waveguide elements of equal length, but
differing in width as designed
radiated by various points at the outlet aperture is to cause the radiation pattern
to reveal a deep 7alley along the 7,eometrical axis of the horn.
To eliminate the axial trough a built-up dielectric lens is inserted in the
outlet aperture of the horn,(Fig.6-23a). The rear profile of this assembly is so
32
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
=MM.
STAT
Declassified in Part - Sanitized Cop Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
designed that the outside rays travelling the longest path in the horn flare are
accelerated in the dielectric exactly to the degree necessary to reach the outlet
aperture in phase with the middle rays travellinfr the shortest path.
In view of the hirh cost of the dielectric lens, a metal lens is often used
instead. In the latter case, the assembly of dielectric elements is replaced by one
of waveguide components (Fig.6-23b). As we have seen in Chapter V, the phase velo-
city of a wave in a waveguide is always greater than the phase velocity of air.
Therefore, an assembly of waveguide components with rear profile of special design
is equivalent to a dielectric lens.
The H10 is the type of wave most generally employed. The phase velocity of
such a wave in a waveguide is the greater, the less the distance between the walls.
Therefore, the length of all the waveguide components of the lens be regarded
as identical, but the components used at the extremes of the horn flare are narrower
than those at its center (Fig.6-23b). In this case, the extreme rays in the horn
flare will be accelerated more than those in the center. The oscillations at all
points in the outlet aperture will be in phase, and the purpose of the electromag-
netic lens will thus be fulfilled.
Use of a lens permits reduction by a factor of 20 to 25 in the length of the
horn required to attain the necessary acuteness of the antenna radiation pattern.
Section 6-11. Operating a Transmitter and a Receiver on a Single Antenna
As already stated in Section 3-2, the pulse method permits a single antenna to
be used in a radar for sending and receiving. The reflected pulses are received in
the intervals between the transmitted pulses, when the transmitter is not working.
The need for only one antenna considerably reduces the weight and size of a
radar and simplifies its high-freqUency channel.
Switching the antenna from receive to transmit is done automatically by means
of an antenna transmit-receive switch, performing two main functions:
33
in Part - Saniti7ed Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
STAl
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
a) connecting the antenna to the transmitter output during the period in which ?
the uhf pulse is generated, with simultaneous shorting of the receiver input, and
switching it off the antenna with the purpose of protection from the effects of the
powerful outgoing pulse;
b) unlocking the receiver input and connecting it to the antenna for the entire
interval between the outgoing pulses, while cutting out the sender output.
In order for these functions to be performed properly, the antenna transmit-
receive switch should fill the following requirements;
1. The duration of "transmit-receive" switching should be tenths or hundredths
of the duration of the operating pulse, meaning that switching should be without
lag and virtually instantaneous. Any significant delay in switching the receiver
off the antenna relative to the moment at which oscillation of the outgoing pulse
begins, has the effect, to begin with, of endangering the receiver input, thus re-
quiring supplementary protection, while, secondly, it has a negative effect upon the
111 channeling of. the energy from the transmitter into the antenna in view of the fact
?
that some of it is diverted into the receiver channel.
On the other hand, any considerable delay in connecting the receiver in terms
of the morbnt at which pulse oscillation ends, enlarges the small "unsamaned" area,
while lagging switching of the transmitter off the antenna worsens the channelling
of energy from the antenna into the receiver due to diversion into the output cir-
cuits of the transmitter.
2. Switching the antenna from receive to transmit and vice versa should not
result in a mismatching of vhf channel components.
3. The cOmrrmtator details performing the switching of the antenna should them-
selves consume minimum energy, so as not to induce any significant reduction in the
efficiency of the uhf channel of the station.
In stations operating in the very short wave range, switching of the antenna
from receive to transmit is performed as a rule by means of quarter-wave and half-
STAT
-
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
wave feeder segments, various types of gas-discharge arresters being used as the
commutator devices. If a voltage equal to or exceeding the ignition voltage be ap-
plied to the electrodes of the gas-discharge arrester, the ionization processes
developing almost instantaneously Cause the arrester to break down. The resistance
of the broken-down arrester is very small and may be regarded for all practical
purposes as a shorting of the points between which the arrester is cut in. On the
other hand, if the voltage applied to the arrester should drop sharply by compari-
a) Transmitter
Transmitter Receiver
Receiver
Main line
of uhf
channel
ON.
b) Transmitter
Receiver
Fig.6-24 - Equivalent Circuit of Transmit-Receive Switch
a) For transmit operation; b) For receive operation
son to its ignition voltage, the arrester will be extinguished. The resistance of
the 'arrester when not broken down is very high and may be taken to be infinite.
The receive-to-transmit switching Circuits used in the antennas of various
stations differ, but they may often be-reduced to the equivalent circuit shown in
Fig.6-24. At the instant when generation of the transmitter pulse in the branched
system of the uhf channel begins,,a traveling voltage wave develops which simultan-
STAT
35
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
eously breaks down both arresters.
The input resistance of the quarter-wave line
connected at points dd is equal to infinity when arrester GP-1 is broken down, and
this is the same as switching this
segment out of the uhf channel (Fig.6-24a).
Therefore, during the period when the pulse is being generated the transmitter
is connected to the main line of the channel to the antenna. At the same time, the
input resistance of the half-wave
rester GP-2 is broken down, which
line connected to points nn is zero when the ar-
is equivalent to shorting the receiver input. If
the receiver input at points nn is shorted, the input resistance at points aa on the
quarter-wave line to the receiver is equal to infinity, which is equivalent to dis-
connecting that line (and
1
points
pulse.
Thus, during the generation of the outgoing pulse, only the transmitter is con-
nected to the antenna.
At the instant when generation is concluded, the wave of voltage traveling from
the transmitter damps down and both arresters are temporarily extinguished. Now the
input resistance of the quarter-wave line, already open at the end, will be equal to
zero, which means short-circuiting at points dd (Fig.6-210)) and disconnecting the
transmitter line from points aa on the main line. At the same time, the line im-
pedance of the half-wave' line at points nn on a half-wave line from an arrester that
aa on the main uhf
consequently, the receiver with the shorted input) from
channel, for the entire generation period of the outgoing
has not been broken down, GP-2, will be infinite, which is the same as saying that
there is no half-wave line, that the receiver input has been switched off and con-
nected to points aa on the main.
Therefore, during the interval between outgoing pulses, when the arresters are
not broken down, it is only the receiver, ready to receive reflected signals, that
is connected to the antenna.
A fundamental inadequacy
acts directly on the receiver input which is therefore directly subject to the ef-
STAT
of the circuit in Fig.6-24 is that the arrester, GP-2,
36
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
-fect of the traveling-wave voltage peak breaking down the arrester. After breakdown,'
however, as a result of which the voltage drop at the arrester electrodes GP-2 be-
comes small, the receiver input will be protected against the effects of the trans-
mitter impulse. An additional arrester is usually provided at the input circuit of
the receiver for the protection arAinst the breakdown peak.
At shorter wavelengths, starting:with the decimeter range, in which the first
component of the receiver input is usually a crystal mixer very sensitive to over-
loads, protection of the receiver becomes particularly important. In this connection
a)
to antenna
Arrester
to transmitter
b)
Glass vacuum
to receiver tube
to antenna
channel
main
Hollow bars
Copper disc
to receiver
Copper disc
Igniter
Fig.6-25 - Tuned Transformers in Receiver Protection Circuits
a) Coaxial-cylindrical cavity; b) Cavity resonator with vacuum arrester
it is important not only that there be low voltage on the disrupted arrester, but
that puncture be easier, i.e., that it break down before the voltage traveling a-
long the line from transmitter to antenna attain its full pulse magnitude. With this
object in mind, the line to the receiver input is connected via a cavity resonator
acting as a tuned voltage transformer.
In the microwave band a coaxial-cylindrical cavity is used to perform the func-
tion of tuned transformer, (Fig.6-25a). It consists of a coaxial half-wave line
shorted at ding voltage wave builds up in this type of resonator, .STAT
37
a .5.)
nariaccifipn in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
?
?
?
the minima being at the shorted ends, and the maxima in the middle where
is emplaced. Thus, the arrester is in a loop where the voltage is always
in the line from the main to the resonator. The resonator behaves like a
step-up transformer. The arrester is disrupted before the traveling-wave
the arrester
higher than
voltage
voltage in
the line attains breakdown magnitude. Puncture of the arrester changes the parame-
ters of the resonator, taking it out of resonance, with the result that the input
resistance of the resonator for the receiver line connected close to the shorted
end, falls sharply to a very low level. In addition, the receiver line remains in
the voltage node. This provides good receiver input protection.
On reception of signals reflected from the target after generation of the
transmitter pulse has ceased, the arrester is not disrupted, and the coaxial-
cylindrical cavity now behaves like a half-wave resistance transformer matching the
receiver line characteristic to that of the line from resonator to main. This as-
sures a high traveling-wave characteristic when the energy of the reflected signal
is channeled from antenna to receiver.
At the beginning of the very short wave band, when use of coaxial lines is
still possible, a
(Fig.6-25b). This
properties than a
tion with coaxial
The arrester
hollow-space cavity resonator is used as tuned transformer
device has a higher Q-factor and consequently higher resonant
coaxial-cylindrical cavity. Magnetic loops are used for connec-
branch lines.
is designed as a separate, replaceable component enclosed in a
gas-filled glass tube. This type of arrester is often termed a gas-discharge tube
or soft rhumbatron. Some gas-discharge tubes have an igniter as well, connected to
an auxiliary constant-;voltage low-power source of 700 to 1000 volts. The purpose of
the igniter is to provide the first-stage ionization necessary to reduce the arres-
ter breakdown time to approximately 0.01 microsecond, and to provide rapid recovery
in the interval between outgoing pulses, in readiness for disruption by the next
transmitter pulse.
STAT
3 LI.;
- neclassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
The switch cavity resonator can be connected with the receiver by a coaxial line
and, on the main-line end, with the waveguide by a slot.
Many waveguide transmitter-receiver antenna switch systems are available. Most
of these use twin-T waveguide connections, consisting of four waveguides in a single
junction (Fig.6 -26). Three lie in the plane of the wide edge, and the fourth is per-
pendicular thereto. The TEE) wave from the waveguide A can only reach waveguides B
and CI since the direction of the electrical lines of force in all these waveguides
is identical: parallel to the narrow edge. This wave cannot enter the waveguide D1
a)
?
b)
C)
Fig.6-26 - Twin-T Connection
a) General view; b) Energy moving simultaneously from waveguides B and C enters
waveguide D; c) Energy fails to enter waveguide D, passing it by
since the electrical lines of force penetrating into the cross-section of the wave-
guide D pass along it, which does not conform with the structure of TEE:). However,
the TE10 wave cannot turn from A to D at a right angle, since the electrical lines
of force in D would be parallel to its wide edge, which in waveguide D is equivalent
to a conversion of TEE) to TEol wave. The critical length of the wave TE01, unlike
that of TE10, is determined by the narrow side rather than by the broad side of the
waveguide D.
39
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
-
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
co
Co
The dimensions of the double T-section connection are designed so that the
oscillator wave driving the waveguide system will be supercritical for the TE mode.
01
This makes it impossible for the TEc1 wave to be propagated in the waveguide D. Thus,
no diversion of energy from waveguide A to'waveruide D can occur.
For the same reason, a wave coming in along waveguide D cannot penetrate wave-
guide A. However, it can enter waveguides B and C, as the turning of the wave in the
plane of the electrical field unaccompanied by changes of the corresponding cross-
sectional dimensions of the waveguide, is not accompanied by transformation of the
TE10 wave into a TE01 wave.
A wave entering the node along waveguide B ma:- enter waveguides C or A, as the
electric lines of force in all these waveguides run along the narrow side.
A wave intersecting the branch in waveguide D is incapable of entering it on
spread from waveguide D to waveguide C or vice versa only if on both sides of this
branch the electric lines in waveguides B and C are in opposing direct:ions (the
electric fields differing 180? in phase). This is necessary if the electric lines
To receiver
Ballast load
Arrester
To antenna
Arrester
?
Fig.'-27 - Employment of Twin T-Section Connection in Waveguide Channel of
Station in Microwave Band
are to turn 90?, and is depicted in Fig.6-26b.
Similarly, waves moving toward the node simultaneously along waveguides
1.0
B and STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
.t
'may branch into waveguides D or A only if they are in opposite phase on reaching the
opposite side of that branch. However, if they are identical in phase on attaining
opposite sides of the branch, their fields, turned by 90?, will, at the cross-section
of the branch, be directed toward opposing sides and will cancel each other out
(Fig.6-26b). In this case,, no branching of the energy is possible.
Let us examine the circuit in Fij.6-27 as an example of the utilization of a
double T-section connection in a waver5uide system for a transmit-receive switch.
'Then the transmitter is in operation, the waveguide wave 4 which it radiates bran-
ches alonc! waveguides 2 and 3 and breaks down the arresters. Waves 6 and 5, reflec-
ted by the arresters, traverse the full routes 2Land 2L2 which differ by on their
4 2
return to waveguide 1, and consequently reach waveguide 1 in antiphase. Therefore,
the energy from the transmitter (waveguide 4) freely penetrates the antenna (wave-
guide 1). The insignificant leakage of waves past the arresters is directed through
waveguides 7 and 8, following identical paths from waveguide 4 to waveguides 9 and
10, and are in phase on arrival there. These leakage waves are not able to penetrate
into the receiver (waveguide 10). They branch only into waveguide 9, the end of
which is terminated at matched ballast impedance, where complete absorption of the
energy of the leakage waves occurs. When the arresters are not disrupted, the wave
of signal received from waveguide 1 is unable to penetrate waveguide 4 (the trans-
mitter), but branches into waveguides 2 and 3. The lines of force of the electric
field make a 900 turn in opposite directions on entry into these waveguides from
waveguide 1. Consequently, on leaving waveguide 1, the electric fields in waveguides
2 and 3 differ in phase by .180?. Continuing along waveguides 8 and 7, they enter
waveguide 10 (the, receiver) in opposing phase, and are therefore able to penetrate
it freely.
The circuit just described provides highly satisfactory protection for the re-
ceiver, and may be used for very powerful stations in the microwave band, in view
of the employment of waveguides.
41
1)e:classified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
1W
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
'Section 6-12. Operation of Transmitter and Receiver on Separate Antennas
In certain types of radiolocation stations, the transmitter and the receiver
are connected to their own antenna array. In this design there is usually no need
for a "transmit-receive" switch.
To weaken the reciprocal distortions of radiation patterns, the receiving and
transmitting components of a station and their antennas sometimes have to be placed
tens of meters apart. This increases the space requirement for a station.
Another major weakness of the use of separate antennas for transmitter and re-
ceiver is the bulkiness of the stations and the extreme weight involved, particu-
larly as a single. transmitting and a single receiving antenna are not always ade-
quate. In distant warning stations, for example, effective following of targets
(particularly low-flying targets) at long range requires a narrower radiation pat-
tern than is the case with short-range stations. This requires either the mainten-
ance of two antennas (one for long-range, and one for short) or switchover to the
necessary radiation pattern on switchover from one range to the other.
On the other hand, in order to combine rapid detection and accurate bearings,
certain types of stations conduct their searches in two stages: coarse-range search
- - - - - -
with a wide-beam antenna, and the taking of bearings and target following wtTi
other; narrow-beam antenna. In this arrangement, simultaneous azimuthal and tilt
direction finding are sometimes conducted in yet another way. One antenna, the beam
of which is narrow enough in the vertical plane and wide enough in the horizontal
to facilitate the azimuth search, is used for elevation determination. Azimuthal
direction finding is conducted by means of another antenna whose beam is, on the
other hand, narrow enough in the horizontal plane and, to facilitate determination
of tilt, wide enough in the vertical.
42
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
r
CHAPTER VII
METHODS OF TAKING BEARINGS ON TARGETS AND OF SWEEP SCANNING
Section 7-1. Methods of Taking Bearings on Targets
Methods of taking bearings are used in radiolocation to determine the angular
coordinates of targets. The measured elevation or azimuth is calculated on a spe-
cific antenna position such as to guarantee that the reflected signal will be of
normal magnitude corresponding to the direction to the target.
Three principle methods of direction finding are used in radiolocation:
a) Bearings by maximum signal strength;
b) Bearings by the zone of equal signal strength;
c) The determination of bearing by phase.
In the majority of cases, a single transmit-receive antenna is used, automat-
ically switching to transmission and to reception at the pulse repetition rate.
Search by electromagnetic beam and target bearing determination is performed
by moving the antenna. The close relationship between the method of direction find-
ing used at a given station and the method whereby the electromagnetic beam travels
through space (scanning) and the type of indicator used is particularly clear when
a single antenna is used.
Section 7-2. Taking Bearings by Maximum Signal Strength
The radiation pattern, e = f(a) of the receiver antenna describes the relation-
1.3
S TAT
?1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
'ship of the voltage e at the input of the receiver attached to the given antenna,
to the angle between the base line for angle measurement and the direction of the
reflected signal. The length of the vector radius of the pattern in the direction of
the incoming signal determines the signal voltage at the receiver input for the
a) Antenna b) Rotation of
rotation sweep trace
Base
line
for
reading ?
angles
Erna.x
Direction of
reflected wave
Fig. 7-1 - Taking Bearing by Maximum Signal Strength
a) Position of rotating pattern at moment bearing is taken; b) Indicator image
at the moment when bearing is taken.
given direction. When the angular position of the antenna is changed, its pattern
,
_ ..
- revolves arOuhd-THe center poifir-tFig:7-la). .111-Crinbtiori
strength, the the bearing angle (an) is read at the moment when the direction of the
incoming reflected signal coincides with that of the maximum vector radius E of
the pattern, as this means that the signal reflected to the receiver input will be
at its raximum. The function of the angular coordinates indicator in this situation
is to fix the moment when the reflected signal is at its maximum and to permit the
reading of the corresponding maximum angle of the antenna. As already indicated in
Section 3-21 the sweep trace on the indicator screen is caused to rotate about its
center (Fig.7-lb), so as to imitate precisely the angular positions of the antenna.
In this casq, the reflected signal determines the brightness of the pip. In reading
144
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
bearings by maximum signa?. strength, the pip is at its brightest at the moment when
the bearing is taken.
' The accuracy with which angular coordinates are measured depends upon the
sensitivity of the indicator to change in reflected signal strength and on the
sharpness of the pattern of the receiving antenna for taking bearings. Figure 7-1a'
shows that at angles approaching that of maximum strength in the pattern, changes
in the vector radius are quite small. In the vicinity of the maximum, the sharpness
of the pattern is at a minimum. Reading bearings by maximum signal strength is not'
a highly accurate method, but it does facilitate the discovery of weak signals a-
gainst receiver background noise, and is used in distant-warning and circular-
scanning stations in which the problem is to read the bearing on a distant target,
but where high accuracy is not required.
Section 7-3. Determining Bearings by Zone of Equal Signal Strength
Let us assume that we have two identical antennas (for example, two wave
a)
Target to right
b)
Bearing on target
Indicator
c)
Target to left
Displacement of signal
to right
Fig.7-2 - Taking Bearing by Zone of Equal Signal Intensity
ducts) so arranged that the axes of their patterns intersect at a given angle. This STAT
45
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Cop Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
? half a degree, which is far superior to that attainable by the method of maximum
signal strength.
Taking bearings by the equal-signal-zone method makes use of the voltage dif-
ferential between el and e2, which changes far more sharply than these two voltages
themselves. This permits the utilization of antennas with wide patterns and, con-
sequently, of small size. However, the range is somewhat less than that attainable
in range-finding by maximum signal strength, as the signal is below its maximwn in
the direction of the equal-signal-strength zone.
Section 7-L. Taking Bearings by Phase
Another method permitting the-taking of high-accuracy bearings on the azimuth,
for example, is the phase method. It
has found application in stations in
LB cB PB
the very short wave band.
In the phase method, the antenna
FY array consists of three co-linear hor-
izontal half-wave dipoles (FiE.7-3),
AP
to receiver
Fig.7-3 - Schematic of Phase Method of
Taking Bearings
LB) Left dipole; CB) Middle dipole;
PB) Right dipole; FY) Phasing device;
AP) Antenna switch
the middle dipole, by an angle of q) =
d being the distance between centers.
- -
Let the wave reflected by the
target impinge from the right, causing
its rays to form an angle e to a line
perpendiculnr to the axial line of the
dipoles, resulting in a difference in
the paths of the rays to the extreme
and middle dipoles Ar = d sin 0. The
e.m.f. L, induced in the left dipole,
will lag in phase behind that, Ec, in
2rt
A r and the e.m.f. En induced in the
X
47
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
right dipole will lead Ec by the same angle cp (see the vector diagram in Fig.7-).a). ?
The e.m.f.s of the outer dipoles are delivered along lines of equal length to
the output of the phasing device FY, and the e.m.f. of the middle dipole to the in-
put of the antenna switch AP. The phase differences between the m?_ e?__ f
? at the
phasing device input and at the antenna switch input remain the same as they were at
the dipoles themselves.
The phasing device FY rotates the phase of the e.m:f. of one of the outside
dipoles (say, the right dipole) by 1800, and then superimposes it upon that of the
other outside dipole (the left). It constitutes a segment of a fixed two-wire line
-LlonF (Fir.7-5). Sliding contacts permit the energy to enter the antenna-switch
2 "
feeder. The sliding contacts are placed on a bridge in such a position OW that
the path of the wave traveling from one of the outside dipoles to the sliding con-
tacts is longer by 71 than that of the wave from the other outside contact, which
is equivalent to obtaining a 1800 phase difference between these waves at the out-
a)
b) c)
Not on target, antenna On target
must be turned to left
Not on target, antenn:a
must be turned to right
Fig.7-4 - Vector Diagrams of Phase Method
put of the phasing device (points aa). Thanks to this phase shift, the resultant
48
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Ap?roved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
'e.m.f. EF at the output of the phasing device, now lags by 900 behind the e.m.f. E?
of the middle dipole (Fig.7-4a). As a result of the passage through the feeder
(see Fig.7-3), this e.m.f. now lags in phase by another 900, and enters the antenna
Phasing bridge
To antenna switch
Fig.7-5 - Phasing Device (Bridge)
switch input AP (already designated as Et, for ease of reference) in antiphase to
the e.m.f. Tc.
In position I, the antenna switch superposes E and E, while in position II it
subtracts them. The resultant signal emitted from the antenna switch output attains
the receiver input as follows: as Ec - E, at position I, and as Ec + EF at posi-
tion II. Therefore, if the signal comes in from the right, the indicator screen
_
will show it_to_11,22maller.yLpion I (due to .f.'act that e.m.f.s E'c and 9, are
in opposite phase) than in position II. If the wave reflected by the target has
arrived from a direction' to the left of the perpendicular to the axial line of the
dipoles, the e.m.f. E..;, would change its polarity, and the relationship between the
amplitudes of the signals on the indicator screens would be the opposite (Fig.7-40.
In order to take a bearing, the antenna array is rotated in the horizontal
plane until the signals on the indicator screen are equal in both positions of the
antenna switch. This requires that the direction of the reflected wave be perpen-
dicular to the axial line of the dipoles. When this occurs there will be no differ-
ence in the paths of the rays, the e.m.f.s of all three dipoles will coincide in
.m.f. EF of the phasing device will be zero (Fig.7-4b). In STAT
?
4.9
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'will cause the patterns to overlap (Fig.7-2). An antenna switch AP in constant oper-
ation automatically alternates the receiver between the two antennas. When the an-
tennas and the direction of the reflected target beam are each in a given position,
the receiver input voltages el and e2 will differ in accordance with whether the
antenna switch is at position I or II. Joint rotation of the antennas causes rota-
tion of both pattern outlines (the solid line and broken-line), with a corresponding
change in the el and e2 voltages. There will be a change, accordingly, in the rela-
tionship between the amplitudes El and E2 of the signals on the indicator screen.
When the antennas are in a position in which the direction of the reflected beam is
that shown by line OA, the voltage on the receiver input will be equal when the
switch is in either position (el = e2). At that instant, the amplitudes of the
signals El and E2 on the indicator screen will also be equal. The direction to OA is
termed the equal-strength zone, and the angle between that 'zone and the direction
of the reflected ray gives us the bearing of the target.
Thus, in taking bearings by the zone of equal signal strength, the angle being
measured is read at the moment when the signals
are equal in amplitude at both positions of the
to facilitate comparison
switch is in each of the
of the amplitudes, the
El and E2 on the indicator screen
antenna switch (Fig.7-2b). In order
signal received when the antenna
two positions is automatically shifted to and reproduced
independently on a --particular side of the screen (on the left for position I and
the right for position II), but both are seen simultaneously because of the speed
of switching and afterglow on the indicator tube. It should be noted that in the
given case the indicator tube is not a reading device but merely an indicator for
the instant when the bearing is correCt, providing information as to when a reading
of angular coordinates should be taken. This reading is taken on a dial, the pointer
posi-
of which is connected to the antennas by automatic devices, and controls their
tion.
In practice, the equal-signal-zone method readily provides accuracy to within
46
?
narinccifipri in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
both positions of the antenna switOh, only the e.m.f. will reach the receiver in-
put: When the antenna is in this position, an azimuth bearing has been obtained. The
azimuth of the target may be read off at this moment on the dial of the indicator
showing the position of the antenna. The indicator tube serves only to reveal the
BB
CB
HB
Goniometer
AP To receiver
?
Fig.7-6 - Schematic of Goniometric Method of Determining the Angle of Elevation
BB) Upper dipole; 11B) Lower dipole; CB) Middle dipole; AP) Antenna switch
moment when the bearing has been obtained, as in taking bearings by equal-signal-
strength zone.
Thus, the phase method of taking bearings comes down to equalization of the
phases of thee.mfsinduced in the antenna dipoles. This is why it Is called the
phase method.
11-sedond example of the employment of the phase method of direction finding in
stations in the meter band is measurement of elevation by an ultrashort wave
_
goniometer (Fig.7-6). The upper aftd lower dipoles of the antenna system employed
X X
for this purpose are at different altitudes above the earth (say at 24and 3--). The
2
vertical patterns of these dipoles will differ. As already indicated in Section 6-1,
the number of lobes in the vertical pattern is equal to the number of half-waves
that can fit into the distance between the dipole and the ground. In our example,
the vertical pattern of the upper dipole BB has three lobes, and that of the lower
dipole NB has two.
The goniometer consists of a s3i-stem of three coils, two fixed (stators) and
one movinF (rotor), capable of rotating within the stator coils. The upper and lower
50
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
309
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'dipoles are connected to the two mutually perpendicular stator coils by means of
two-wire feeders. The rotor coil of the goniometer is connected to the antenna
switch by means of the feeder.
Due to the difference in elevation above the ground at any tilt of the target,
the e.m.f.s induced in the upper and lower dipoles, and the currents in the stator
coils of the goniometer, connected therewith, will differ in amplitude and phase.
The fields of the stator coils Induce voltages at the ends of the rotor coil which
differ in amplitude and phase, in the general case.
However, at any elevationl-it is possible to find a specific position of the
goniometer rotor coil for which the voltages induced at its ends by the fields of
the two stator coils will be equal in amplitude and opposite in phase. In this
case, the resultant voltage at the ends of the rotor coil will be zero.
The input of the antenna switch AP receives not only the resultant voltage of
the goniometer rotor coil, but the e.m.f. induced in the middle dipole CB. When the
antenna switch is thrown from one position to the other, the e.m.f. of the middle
dipole is either added to or subtracted from the goniometer voltage. If the position
of the rotor coil has not been properly adjusted, the signals on the elevation in-
dicator screen will differ in value as between antenna switch positions I and II
. (as in Fig.7-4a and c).
The rotor coil is turned until the signals are equal when the switch is in
both positions, indicating a bearing on the target. The elevation at this moment is
read off on the goniometer dial (scale).
--------
Section 7-5. Methods of Sweeping the Electromagnetic Beam
Sweeping, or scanning, is the name given to moving the directional electro-
magnetic beam of the transmitting antenna for consecutive radiation of surrounding
space.
There are quite a large number Of methods of scanning, but all are variantsTAT
51
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
the four rain types. These four main types are:
a) circular scanning;
b) screw scanning;
c) conical scanning;
d) spiral scanning.
As already noted in Section .7-1, the method of scanning used, is coordinated
with the method of taking bearings employed in the given station, and the type of
indicator of angular coordinates. On the other hand, the rate at which the electro-
magnetic beam is scanned, its width, and the pulse repetition rate, must always be
coordinated.
Section 7-6. Circular Scanning
In circular scanning, the electromagnetic beam and the antenna are rotated in
Horizontal plane
Antenna
Fig.7-7 - Circular Scanning
a horizontal plane around a vertical axis with an angular velocity of co (T) radian
per second (Fig.7-7).
If the width of the horizontal pattern is bm radian, the time required for
b(m)
circular sweep of the beam across the entire width is '' second. A located
(c))
target is within the range of the beam during that period. But in order to obtain
a satisfactory target pip on the indicator screen, a minimum number of pulses n de-
of indicator, must be directed at the target during this period.
STAT
52
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Cop
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Therefore, when circular scanning is used, it is necessary to consider, in de- '
termining pulse frequency, that it must exceed the minimum required by that type of
scanning;
w(T)
Fu min.= [pulses/sec]
n
b(cp)
Circular scanning is used in distant warning stations and in 3600 search sta-
tions employing direction finding by maximum signal to measure the target azimuth.
The indicator image obtained by circular scanning is that which may be seen in
Fig.7-l. The rate of rotation in circular scanning is usually small, constituting
L-6 rpm., and the pulse rate does not exceed 500 per second.
A special type of circular scanning is sector scanning, when rotation of the
a)
b) Target pip
Fig.7-8 - Sector Scanning for Elevation
antenna is replaced by motion to and fro within a limited sector controlled by the
electromagnetic beam.
Combination of sector scanning and the taking of bearings on maximum signal
strength may be employed to measure the elevation of the target by means of a spe-
cial antenna operated by a transmit-receive switch. The antenna and its beam are
rocked to and fro within a specific interval of tilt, say from 0 to 45? (Fig.7-8a).
The rotatory motion of the scanning trace on the elevation indicator screen may, in
the given case, conveniently be replaced by parallel up-and-down motion (Fig.7-8b).
STAT
Each anpsular position of the antenna corresponds to a specific position of the scan-
53
y Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5 ..)53
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
fling trace on the elevation indicator screen. The brightness of the pip on the
screen is at its maximum when the eleVation is on target.
Equation (7-1) remains valid when sector scanning is used. When the elevation
is to be measured, it is necessary to write into the equation the beam width be
for the elevation, and the angular velocity we of antenna motion in the vertical
plane.
Section 7-7. Screw Scanning
In screw scanning, the electromagnetic beam emitted by the antenna is periodi-
cally rotated through a given azimuthal sector, with simultaneous but considerably
slower change in tilt. By its simultan-
eous participation in these two types of
motion, the beam describes a distorted
screw-like line in space (Fig.7-9a).
Screw scanning is also usually com-
bined with direction finding by maximum
impulse. In screw scanning, the indicator
permits simultaneous measurement of two
-V--
angular coordinages, the azimuth (P, and
the elevation 0 (Fig.7-9b). The vertical
motion of the scanning trace is timed
Fig.7-9 - Screw Scanning
with the vertical rocking of the antenna.
On the other hand, the motion of the spots on the screen along the sweep trace to
right and left is timed with the rocking of the antenna along the azimuth. In this
case, the reflected signal also governs the brightness of the pip, which is at its
greatest when the radar is on target.
In order to be subject to monitoring, every point in the search zone must fall '
within the electromagnetic beam at least on each of two successive rotations of the
STAT,
51.
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Cop Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'antenna on the aximuth. If the sector covered by the elevation sweep is b(e) radian,
and the vertical width of the antenna diagram is Y(0) radian, our condition of
double coverage will require two turns of the antenna on the azimuth for observation
of a sector. Therefore, when the azimuth angular velocity of the beam is W(p) radi-
ans per second, and the azimuthal sector of observation is T radian, the following
relationship between the angular velocities of the beam on the azimuth and in eleva-
tion must be adhered to:
_ b(o)w(p)
6)(T)
(7-2)
The minimal pulse frequency required by the indicator is related to the hori-
zontal width b(T) of the pattern, and to the azimuthal velocity wm by eq.(7-1),
as in circular scanning.
When employed in ground-based and naval aircraft search stations, circulAr
scanning follows a pattern of 3600 rotation along the azimuth (T = 2n), and a 900
elevation sector. It is also used in aircraft-based sets for searching a sector a-
bove and below the aircraft's horizon, and over a 1800 sweep in the azimuthal plane.
The main shortcoming of screw scanning is its inadequacy for directly tracking
fast-moving targets.
--------
The second appearance of the reflected signal on the indicator screen takes*
place after about ten seconds have elapsed. This is a period sufficient for the
target to be lost during uninterrupted observation of the entire search zone.
Section 7-8. Conical Scanning
If the exciter of a paraboloid be moved relative to the focus, the direction
of the electromagnetic beam.will vary from the focal axis of the paraboloid. When
the exciter is shifted downward, the beam rises and when it is shifted upward the
beam dips. When it is moved to the right, the beam is moved to the left, and when
the exciter is moved to the left, the beam will deviate to the right from the a:EMAT
55
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
? of the paraboloid. And if an excitei. dipole positioned assymmetrically relative to
the focus of a stationary paraboloid is rotated around the focus, the electromagnet-
ic beam will describe a cone. The axis of the cone will coincide with the axis of
c'
the paraboloid. This shifting of the beam in space is termed conical scanning
(Fig.7-10).
The angle between the generatrices of the cone, and the width of the electro-
Antenna
Beam axis
Beam axis
Fig.7-10 - Conical Scanning
Target
Beam trajectory
magnetic beam, are such that the positions of the rotating lobe of the pattern in
an7 longitudinal section of the cone will overlap. If a single antenna is used for
Mismatch signal
detector and
amplifier
Receiver
Antenna
Mismatch signal -
for, azimuth
Lc:nor control of rotation in azimuth
Azimuth voltage Azimuth reference
, control circuit voltage
Basic voltage
Mismatch signal oscillator
for elevation 0 Elevation voltage
control circuit Elevation reference
voltage
Motor control of rocking in elevation
Fig.7-11 - Automatic Tracking Circuit
transmission and reception, then, when the axis of the scanning cone is not in line
with the target, the energy of the initial radiation, as well as the energy reAT:F
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?bed from the target, will differ when the beam is in different positions .
'Tie of the\5
?,
the cone will the amplitude of the reflected signal from the target always be the
same, regardless of the position of the beam. Therefore, the position of the ads
of the paraboloid is easily noted and employed for taking both azimuth and elevation
bearings. The bearing is accurate, to, within a few hundredths of a degree.
It is not difficult to see, by comparing Figs.7-10 and 7-2, that for purposes
of taking bearings, conical scanning is a further development of the equal-strength-
of-signal-zone method. Conical scanning is used in the microwave-band sets used for
gun-laying with automatic target tracking, and also in certain types of aircraft
radars.
Use of conical scanning for automatic tracking in azimuth and tilt planes may
readily be understood on the basis of the following example.
If the position of the antenna is such that the axis of the scanning cone does
not entirely coincide with the direction of the target, the reflected signal ob-
tained at the receiver output (Fig.7-11) on rotation of the exciter dipole will be
modulated sinusoidally with a frequency equal to the number of revolutions of the
dipole per second. At the detector output, a mismatch signal will form from tha_
_
modulated reflected signal, and two signals of error will then be amplified. The
amplitude of the error signal for the, azimuth will be greater at a greater devia-
tion of the axis of the cone from the tarEet along the azimuth, while the amplitude
of the elevation mismatch signal will be greater at a greater deviation of the axis
of the cone from the target in elevation. If the axis of the cone is directly.on
target, the mismatch signals are zero.
An oscillator to produce sinusoidal reference voltages is mounted on the axis
of rotation of the dipole. Its frequency is also equal to the number of rotations
of the dipole ,per second. The circuit for the controlling azimuth voltage as af-
- ?
H .0-11
b bile azimuth mismatch signal and the azimuth reference volt-
57
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
/6
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
age is developed to accord with the sign of the voltage controlling the motor caus-
ing the antenna to revolve in the azimuthal plane. Likewise, the circuit for the
controlling elevation voltage as affected simultaneously by the elevation mismatch
signal and the azimuth reference voltage is developed to accord with the sign of
the voltage controlling the motor causing the antenna to rock in a vertical path.
The controlling voltage circuits act on the motors so as to cause the antenna al-
ways to take a position in which the axis of the cone is virtually coincident with
the direction to the target, and the mismatch signals are approximately zero. The
azimuth and elevation are read by means of pointers always on target.
In determining the pulse frequency to be used in conical scanning it is nec-
essary to remember that it must always exceed a minimum.
w?
Fu min
n
Here b is the width of the antenna radiation pattern, in radians;
w is the angular velocity of rotation of the dipole;
sp
n is the minimum number of pulses needed by the given type of indicator to
illuminate the target as long as it is in the zone of the antenna beam.
(7-3)
? . Section 7-9. Spiral Scanning
automatic-tx,a4kings, spiral s'...4,41,gra--3,2sed in combfLnati,-,n
with conical scanning, the same antenna being used for both.
If conical scanning is to be used for obtaining bearings and automatic track-
ing, the preceding spiral scanning is used for very quick location of targets with-
in the portion of 'a sphere bounded in elevation and azimuth. In spiral scanning,.
not only does an exciter placed assymetrically relative to the focus undergo rota-
tion around that focus, but the paraboloid itself rocks in the vertical plane. If
the paraboloid undergoes no change in azimuthal position, the trajectory of a point
on the circle described by the beam due to the rotation of the dipole, changes into
as
iral.because of the ',eking of the paraboloid. The coils of the spiral lie on
58
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
the surface of a sphere of radius R (Fig.7-12).
The ratio between the rate of .rotation of the dipole and the rate at which the
paraboloid rocks must be set in such a wanner that any point in the space being
searched falls within the beam in not less than two successive coils of the spiral.
It can be shown that the required ratio of angular velocities is as follows:
boa
sp
Wrock 4n
in which 'rock is the angular velocity at which-the paraboloid rocks in the verti-
cal plane;
is the angular velocity of rotation of the exciter;
sp
and b is the beam width.
In spiral scanning, the pulse frequency must exceed the following minimum
Fi min =
sp
13
sin ?2 [impulses/second]
(7-5)
where n is the minimum number of pulses for the given indicator;
0 is the sector through which the paraboloid rocks in the vertical plane.
The prime advantage of spiral
Beam trajectory
Fig.7-12 - Spiral Scanning
imum signal strength permits
scanning over the other types is the reduction
in scanning time to a period bf some-
times -riot than-one second,
Figure 7-12 illustrates the trajec-
tory of the electromagnetic beam in
spiral scanning when there is no change
in the position of the paraboloid in the
azimuthal plane. However, a combination
of spiral scanning with circular rotation
of the antenna in the plane of the azi-
muth, and determining of bearing by max-
the .use of a circular scanning indicator and reading
STAT
59
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
? the azimuth of the target directly on the screen, as with standard circular scan-
ning. ',Then this is the case, the indicator screen shows an image identical to that
obtained in circular scanning (see Fig.3-4b). The elevation of the target is mea-
sured Only on switching from spiral to conical scanning.
Section 7-10. Search With the Cosecant-Squared Beam
In designing surface spotter radars, it is quite logical to desire that the
signals reflected from aircraft at a given altitude, regardless of range, be of
approxirately identical amplitude. Solution of this problem is possible only if the
antenna array enp1o7ed provides a specific unequal distribution of energy in the
flare of the electromagnetic beam.
The inclined ranges R of points at unknown elevation h are proportional to the
cosecant of the elevation 0. From Fi-,..7-1311 we find that, in reality:
= = h cosec 0 (7-6)
sin
Allowing for the fact that the field intensity declines in proportion to the
first power of the range, the following expression ray be written for the field in-
tensity of the outgoing beam at a point at altitude h:
Eh =
g = hE
Ii cosec 0
(7-7)
_ ? A
in which kE is a coefficient of prcpol.crium171.,.6y-,qpending upon-t-hs-x,m,diated power
of the antenna and allowing for its directivity in terms of field intensity.
In order that the amplitude of the signals reflected from aircraft flying at
altitude h and at different distances be identical, it is necessary in the first
place that the outgoing beam be of identical field intensity at all points on that
line. The required distribution of radiation may be attained by means of a special
antenna, the field intensity of which varies with the cosecant:
E = Emax cosec 0
where E is the field intensity in the direction of maximum radiation.
60
STAT
0
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
_
From this we obtain:
lnax
Therefore, the factor kE employed in eq.(7-7) will also be proportional to the
magnitude of cosec e when this type of antenna is used, i.e., kE = k cosec 0, where
k is a constant. Taking this into consideration, we obtain the following from
eq. (7-7):
= cosec
k cosec 8
Eh = = const
h cosec 0 - h
Thus, when a cosecant-pattern antenna is used, the field intensity of the out-
going beam does not depend upon the elevation and is constant at all points in a
Fig.7-13 - Search With the Cosecant-Squared Beam
line of equal altitude.
The return paths from points on a line of equal altitude are naturally in
agreement with the same expression (7-6). Examination of the cosecant shape of the
pattern when the antenna is used for reception readily shows that the signal re-
flected from a given target at all points on a course of equal altitude and regard-
less of range R will reach the receiver input at the same amplitude.
Thus, solution of this problem is provided by a directional antenna, the field
intensity of which is proportional to the cosecant of the elevation (Fig.7-13c),
and whose radiated power is proportional to the square of the cosecant of the el(sTAT
4
61
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R00180-0200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
tion. This t-pc of a-Itenna is called a cosecant-snuared antenna. It rarliates the
lar-er part of its ener-z- i:arallel to the surface of the earth, and the smaller part
Incli:icd1)eal, Vertical beam
Horizon line
- V-beam Antenna rattern
?
at an-les thereto. The antenna reflector in this case is rade of two parabolic parts
of different focal len.,Iths, the lower
portion being thrust somewhat forward
?
(Fir7.7-13a).
A cosecant-squared beam ma:r be obL
tained not only by the proper design of
reflector components, but also by the
use of a complex exciter of special de-
sign. This exciter has to provide the
dephasing of the energy radiated by var-
ious components at the surface of the
normal parabolic reflector, as required
Fig.7-15 - Use of a V-beam Pattern to De- for a cosecant beam.
termine the Elevation
Section 7-11. V-beam Search
P a) Moment of spotting by vertical beam;
For search of surrounding space and
b) Moment of spotting by inclined beam
dimensions,
measurement of coordinates in all three
a special transmit-receive antenna with a V pattern is highly useful.
STAT
62
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
..?11
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
This antenna consists of two rigidly attached parts rotatinr joint1:: in the azirruth-.
al plane. The beams of both parts of the antenna arc quite narrow in the horizontal
plane and quite broad in the vertical. The bean of one part of the antenna is al-
wa7s in the vertical position, and the beam of the other
is at /50 to the horizon.
As a result, the resultant beam of the antenna is in the shape of the Latin letter
V (Fit7.7-14).
In tal:ing a bearing determination of maximum signal strength, each target
is spotted twice as the antenna rotates: first by the vertical beam and then b:7 the
inclined beam.
Spotting the target with the vertical beam makes it possible to reasure its
range on the incline R and its azimuth p. This moment of discovery is shown in
Fig. 7-15a. The triangle OK.: gives the elevation of the target:
= sinh ? (7-8)
where h is the altitude of the target;
D = OK is the horizontal range (projection of the inclined range onto the
horizontal plane).
The moment at which the inclined beam then spots the target corresponds to the
measurement of another azimuthal magnitude T AT, where AT is the angle of rota-
?-- ----?------.-,
I
tion of the antenna between the moments at which the target is spotted by the ver-
tical and horizontal beams. Due to the fact that the axial plane of the inclined
slope forms an angle of 45? with the earth, we find the following from the triangle
ZtOK (Fig.7-15b):
A T = sinh ?D
comparing eqs.(7-8) and (7-9), we find that
= Acp
(7-9)
Thus, taking an elevation bearing by means of a V-beam antenna comes down to
reading the difference between the azimuths of the target at the instants, when it is
63
STAT
??????.,
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
f_ _ _
spotted by the vertical and horizontal beans of the antenna.
' Section 7-12. Advantages and Disadvantages of Pencil-Beam Antennas for
Search and Bearings
The degree of directivity of receiver antennas is determined, in radiolocation,
primarily by the need for precision in the measurement of the angular coordinates
of a target in direction finding.
Higher directivity is also quite useful in increasing the effective radius of
an antenna. The sharper the directivity of a transmitting antenna, the more strongly
it concentrates electromagnetic radiation in the direction of the target, and the
stronger is the field of primary radiation of the target. On the other hand, like
an optical lens
which focuses incident light rays of a single point, a directional
receiving antenna collects the energy of the electromagnetic rays reflected by the
target at the moment it is spotted, and thereby amplifies the e.m.f. emitted by the
antenna. Thus, a most important problem in the design of radar stations is that of
increasing the directivity of the antennas.
On the other hand, extreme narrowing of the directional beam makes it difficult
to find narrower the electromagnetic beam of the transmitting antenna,
the smaller the number of targets it is capable of revealing simultaneously, par-
ticulPrly at short ranges. Narrowing of the beam makes it necessary to increase the
speed with which it is swept, and this often creates serious engineering problems.
To obtain a very highly directional beam is difficult in itself, as narrowing of the
pattern is attained by making the antenna still more complicated and by increasing
its linear dimensions relative to the operating wavelength.
Radiation of the target with the direct pulse is possible only during the time
interval in which the electromagnetic wave moving through space follows the target
with its full width. To increase the probability of spotting when scanning with an
electromagnetic beam at high speect, it is necessary either to considerably increase
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
.????
3
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
the pulse frequenc, which, .as we have seen, limits the range of the station, or to.
increase the pulse power of the trnsmitter.
Thus, the advantages provided by pencil-beam antennas are high accuracy of the
bearings obtained, and a clear improvement in the range of a station of given power,
111 while the disadvantages lie in the difficulty of finding the target and the rela-
tively larzer linear dimensions of the antenna. The decision as to type of antenna
radiation pattern and array is determined by the tactical and technical specifica-
tions for the particular locator station: military purpose, radius of effective-
ness and operating wavelength.
?
?
65
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
CHAPTER X
RADAR RECEIVERS
?
Section 10.1 Purpose, and Specifications to be Met
The purpose of a radar receiver is to amplify the radio echo pulses received
by the antenna, to eliminate noises, and to convert the radio pulses into video
pulses, i.e., DC pulses. '
The receiver unavoidably amplifies not only the useful signal but the chaotic
interfering voltages called noises. The echo signal will not be visible on the in-
dicator screen if the signal, amplitude is lower than the noise level. Therefore,
the first and most important specification for a receiver is that it produce mini-
mum set noise.
The problem of receiving the echo signal is complicated as the range to target
increases, in view of the marked weakening of the amplitude. The power of the
signals emitted by very distant targets does not exceed a fraction of a micromicro-
watt, corresponding to an antenna emf-of the order of a few microvolts. Normal
operation of indicators requires the application of power of the order of one watt,
corresponding to an output voltage of several volts or even several tens of volts.
Therefore, in addition to producing minimal set noise, a radar receiver must be
highly sensitive and have very great gain, of the order of several millionfold.
This is the second requirement.
In order not to introduce additional errors into the range measurements, the
receiver must ensure minimum distortion of the echo-pulse envelope and exceedingly
66
STAT
Declassified in Part- Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
--lbrief transit times from input to output. This is the third requirement.
1-1
_a_ For certainty of reception of echo signals, and to attain high gain, the re-
-'
ceiver must tune easily to the operating frequency of the transmitter, ensuring
6 '
Aft maxim= sensitivity there, and must readily follow any fluctuations in the trans -
--qmitter frequency. This is the fourth requirement.
Despite marked attenuation in the transmit-receive switch, the energy of a
- powerful transmitter will nevertheless leak into the receiver to some degree. There-
; ?
?
- fore, the receiver circuit must contain its own protection from overloads by the
transmitter signal, and also from the effects of powerful outside interference.
13__
--
20 This is the fifth requirement.
_-
All other requirements to be met by radar receivers are dictated by these
e,
1 five major requirements and derive directly from them.
4 --
- Section 10-2. Receiver Set Noises
'
The theory of this most important question has been worked out in detail in
- the works of V.I.Siforov and A.A.Kolosov, and also by V.A.Kotellnikov and others.
Set noises arise from phenomena of accidental and .random character and giving
--rise to chaotic, fluctuating voltages and currents in the receiver circuits.
, --Fluctuation voltages and.aurrents-show the most extreme variations in frequency and
- phase, due to their accidental nature. Noises may therefore be represented by a
t--- continuous series of a very large number of harmonics, always eMbracing the entire
receiver bandwidth, regardless of how great it may be, and regardless of the fre-
- quency to which the receiver is tuned. Noise energy consists of the total energies
of all the noise harmonics and therefore increases in proportion to the widening
of the receiver bandwidth.
Oscillating-Circuit Noises
The electrons in any conductor are in constant random thermal agitation. The
electron in random agitation may be. regarded as an elementary random current
67
? ?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5 -
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
creating an elementary voltage against the resistance of the conductor even when no
external source of electricity is present. The resultant noise voltage is composed
of the totality of such elementary random voltages. The total noise power in a
circuit is greater, the higher the temperature and the wider the pass-band of the
circuit.
The noise power in a circuit is:
9
LI;;
Pn = 4kT A F =
(10-1)
where T is the absolute temperature in degrees Kelvin;
k = 1.38 x 10-23 joules per degree denotes a scale factor;
AF is the pass-band, in cycles;
Re is the active component of the equivalent resistance of the circuit,
in ohms.
From this we obtain the square of the noise voltage, in volts
or
2
lin = 4kT 11 eA F
Un = 2 VkT R F
(10-2)
If Re is expressed in k-ohms and A F in kilocycles, we will have, at room
temperature (T = 290?K)
Un (mkv) = 155-1F
?
(10-3)
Thus, if the pass-band is A F = 1 megacycle and Re = 1000 ohms, the noise
voltage will be
1 ,/ 31.6
Un = yl x 103 =
8 4 mkv
68
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
2,
?
?
?
The noise voltage applied to the antenna, due to thermal agitation of the
charged particles of the environment, may be calculated by an equation analogous
to eq.(10-2):
2
4kTAI12A F
Un A
where TA is average antenna temperature; and
RE is the radiation resistance of the antenna.
Receiver-Tube Noises
Tube noises are determined primarily by the shot effect due to the random
thermal inconstancy of cathode emission. The shot effect causes a fluctuation of
the plate current which, in turn, results in a fluctuation voltage Una in the plate
load resistance.
The space charge in the tube, absorbing the shocks from the electron stream
moving from cathode to plate, modifies the fluctuation and reduces the noise. On
the other hand, the presence of positive ions in the tube, due to inadequate hard-
ness of the vacuum tends to increase noise. This may be explained by the fact that
the positive ions settling on the cathode and becoming neutralized there, complicate
the escape of electrons and impair the emission. As a result of the quantitative
fluctuation in emission, the neutralizing positive ions produce further fluctua-
tions in the emission currents and increase the noise.
Fluctuations in the plate current increase with its level. As a result, a
larger number of electrons undergo fluctuation. The existence of positively-
charged grids absorbing part of the electron stream moving toward the plate, in-
creases the noise, since the inconstancy of the emission current aggravates the
inconstancy of its distribution between the electrodes. Pentodes make more noise
than triodes.
The fluctuation noise voltages Una, discriminated across the resistance of
69
STAT
'
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
the plate load, may be replaced by an equivalent noise voltage Unc cut into the
grid circuit. Obviously
Una
Unc =
where K is the gain of the stage using the given tube in the given operation.
This substitution makes it possible to regard the tube as noiseless but as
having in its grid circuit a noise-voltage oscillator Unc (Fig.10-1). The most
convenient method is to substitute the voltage source Unc by an equivalent noise
resistance, R. The noise voltage across this resistance can be calculated in the
usual manner (by eqs.10-2 or 10-3) and taken as the Unc voltage.
? Two equations for the noise resistance of an amplifie4
for triodes
2.5 - 3
Rn (kilohms) =
tube are known:
(10-4)
(10-5)
for pentodes
Ia
Rn (kilohms) =
S (ma/v)
205 I(g)
+ 30
Ia I(g)
S
where Ia and I(g) are, respectively, the plate and screen currents, in milliam-
peres.
As indicated in eq.(10-4), the noise resistance and the noise both increase
as the transconductance declines, since a decline in transconductance is equiva-
lent to a reduction in the dynamic gain of the cascade K. When the tube is used
as a converter the noise resistance increases 3 to 5-fold, since the conversion
transconductance is less than the transconductance of the same tube when used as an
amplifier.
The considerable time required for the electron transit increases the noise,
since the fluctuations in the electron stream are supplemented by fluctuations in
the currents applied in the electron circuits. Tubes make more noise in the
STAT
ultra-short-wave band. Noise reduction. requires a reduction in the electron transit
70
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
time.
Section 10-3. Noise Figure and Sensitivit7
It is convenient to calculate the noises in all stages of the receiver, with
allowance for the pain of these stages, in terms of the grid circuit of the first
Ii
nc
Fig.10-1 - Noise-Voltage Oscil-
lator Unc =
Una
in the Tube
Grid Circuit, used as Voltage
across the Resistance Rn
ponding noise voltages, we may write:
tube. The power of the total (discriminated)
noises R in the grid circuit of the first
Zn
tube is composed of the powers of the lumped
noises in the following receiver components:
P in the antenna Pn in the input circuit,
nA
Pnt of the first tube Pn tbs for all the
other tubes, i.e.,
P =P +P . +P + P
E n nA n nt n tbs
Considering that the power of the tubes
is proportional to the square of the corres-
Uzn = 2 2 2 - 2 + u2
Unt n tbs
UnA Un in
From this we find that the total-noise voltage effective in the grid circuit
of the first tube, is equal to:
U = u2 4. "2 4. "2 4. "2
nA un in 'nt un tbs
The receiver noises have come to be characterized by the noise figure (noise
factor), which is usually determined by means of a signal generator.
The noise figure gives data on the factor by which the power of the signal P
cA
and of the noises P delivered to the receiverby the signal generator, are
nA
greater than the signal-to-noise ratio, P
-c cut/Pn out at the receiver detector
_
_
- 71
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
input. Thus, the noise factor is
N=
PCI'
PnA
Pc out
Pn out
(10-6)
In determining the noise factor, the internal impedance of the signal gener-
ator, substituting for the antenna, must be equal to the input impedance of the
receiver.
From the determination of the noise factor in eq.(10-6) it follows that the
power of the signal reaching the receiver input from the antenna is
c out
Pn out
PcA-= PnA N
(10-7)
The signal-to-noise ratio at the receiver detector input, necessary for satis-
factory reception of signals, i.e., the quantity D =out y will wi be denoted
c
Pn out
below as the discrimination factor.
By analogy to eq.(10-1), the antenna noise power is:
PnA = 4kTAA Fpb
Consequently, the required minimum signal power at the receiver input will be
(10-9)
Ppb min = 4k FAA FpbN D
This power is defined as the sensitivity of a radar receiver.
Equation (10-9) shows that, at a given pass band A Fpb and a known discrimina-
tion factor DI the sensitivity of the receiver is governed by its noiselfactor N.
The magnitude of D depends upon the shape, length, and frequency of the pulse
and also on the indicator and receiver parameters. The smallest possible discrim-
ination factor is obtained when D is unity.
72
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Section 10-4. Input Impedance of a Tube in Ultrashort?Wave Operation
Among the various works of Soviet scientists that have played a pioneering
role in the development of ultrashort ?wave tube engineering, particular credit is
due to the work of Prof. V.I.Siforov who, working on the basis of the general
theory of active linear electric networks, developed by E.V.Zelyakh, V.I.Kovalenkov,
and V.I.Siforov, envolved an orderly theory of amplifier tube function in the
ultrashort ?wave band.
In accordance with this theory, the current within an electron tube is, in the
general case, not equal to that in the corresponding external circuit. I
Two types of current must be distinguished for an electron tube:
2 .1
a) Conduction current ik, serving to propel the electric charges in the
space between the various tube electrodes; and
b) Displacement current or capacitance current ic, generated by a charge
in the voltage on the electrodes, i.e., by the presence of a varying
electric field concentrated in. the capacitance between the electrode's.
The total current within the tube is equal to the sum of the condtiction current
and the displacement current.
The electrons moving in the space between the tube electrodes create their
own electric field. Under the influence of this field, electric charges are in?
duced at the surfaces of the electrodes, while in the wire connecting the electrodes
between which the electrons flow, current is induced regardless of whether the
electrodes do or do not intercept a portion of the electron stream. If the magni?
tude or velocity of the moving charge varies with time, the induced current also
changes.
A displacement current also arises between the electrodes when they are under
alternating voltage.. Therefore, the total current in the wire connecting the ?
electrodes is equal to the sum of the induced and displacement currents. Thus, the
appearance of induced currents in the external circuits of tube electrodes is due
73
STAT
Declassified in Part- Sanitized CopyApprovedforRelease2013/03/14 CIA-RDP81- 677770777077333377711111111111
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
to the inductive effects of the electron flow within the tube.
Specifically, the electrons passing through the grid turns on their way from
cathode to plate, induce a positive charge in the grid. This charge increases as
the electrons approach the grid and declines as they retreat into the distance. An
increase in the positive charge induced is equivalent to the appearance of an in-
duced grid current it, while a decline in this charge is equivalent to the appear-
ance of an induced current i". these currents are opposite in direction (Fig.10-2).
Therefore, the resultant induced current in the external grid circuit is
_ in
P g g
The magnitude of the induced current ig is significantly dependent on the
electron transit time from cathode to plate.
With long and short waves, the transit time of electrons is negligible by
comparison with the duration of the oscillating voltages acting on the tube elec-
trodes. It may be considered that, in the given case, the phase of the alternating
grid voltage does not have a chance to change within the transit time. The veloc-
ity of the electrons approaching and receding the grid will be determined by the
same phase of grid voltage. The increase in the positive charge on the grid, in-
duced by the approaching electrons, is compensated by its decrease under the in-
fluence of the departing electrons. The currents it and i" will thus be equal
and no resultant current will be induced in the grid circuit. The flow of elec-
trons will show virtually no inductive effect in the long- and short-wave bands.
The electron cathode-to-plate transit time, which is of the order of 10-9 sec
in ordinary tubes; becomes equal to the period of oscillation, in the ultrashort-
wave band. Thus, at a 30-cm wavelength, the period of oscillation T is 10-9 sec.
In any case, on ultrashort-wave, during the electron transit time, constituting a
considerable proportion 2n of the oscillation period, the phase of the alternating
voltage on the grid is able to change by an angle TT . Therefore, the velocity of
STAT
74
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
the electrons passing through the grid and that of the electrons that have just
been emitted by the cathode, will be determined by different grid-voltage instan-
taneous values and will not be equal. The in-
Fig.10 -2 - Currents Induced in the
Grid Circuit due to the Inductive
Effects of the Electron Flow:
it, the current Induced by Elec-
trons Approaching the Grid;
i" Current Induced by Electrons
g,
Moving Away from the Grid
input impedance, let us consider the
fier tube.
duced currents it and i" will differ. The
g -
resultant induced current i appears in the
external grid circuit. Its first harmonic,
i.e., the component of the frequency of the
alternaiing voltage on the grid, will be 'con-
siderable.
Thus, the effect of transit time due to
inductive action of the electron stream in
source of con-
the ultra short-wave band
is a
siderable grid currents, even when the nega-
tive displacement on the grid is such that
grid, currents would be incapable of appearing
in the long and short waves.
In order to investigate the question of
pentode as the most important type of ampli-
The Input impedance of a tube is the ratio of the total input voltage to
the total input current I , i.e.,
in
,
L. =
in
'in
U?
in
(10-10)
Let us examine Fig.10-3a. The applied current required by the pentode from
the source Of-high-frequency voltage branches in part at the cathode through the
interelectrode capacitance (current Y ) and in part through the screen grid cir-
cuit across the capacitance C(g)g (current I(g)g)? Therefore
75
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
1..
lin =
Uin
Igf I(g) g
The quantities entering here are complex, since the presence of reactance
parameters will cause the currents Ye and I(g)g to be out of phase with the voltage
Fig.10-3 ? Circuits and Vector Diagram of Pentode Used
in the Ultrp.short ?Wave Band
at the input. In addition, these currents would be purely capacitive if it were
not for the effect of the electron transit time and the influence of the active
leaks of interelectrode capacitance. Inductivity in the tube leads would be lack?
ing. The current if, at inductance IT, creates a voltage Um., which leads the
current by 90?. The total input voltage Tlin is equal to the geometric sum of the
voltage between grid and cathode, and the voltage on the inductance of the cathode
lead. The current Igf2 branching through the capacitance Cf at the cathode, is
g
governed by the voltage IT'gf between the control grid and the cathode, and leads
that voltage by 900, being in phase with the voltage ULf. The capacitance C(g)g
?
is inserted between the control grid and the ground. Therefore, the current
is governed by the total input voltage 17in and leads it by 90?. The total input
current lin is equal to the geometric sum of the currents l'gr and If ? It leads
g)
the input voltage 1.Tin by less than . This means that the input impedance of STAT
9
76
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
the pentode is not purely capacitive but complex in nature.
The complex input impedance of-the pentode may be conceived of as a parallel
comectioneftheinputcapacitanceC.andtheresistanceR..Excluding the
In
R
t
i
th
l t i
effectofelectrontransittimeltheresistance. is equal e resistance Rin nL'
poverned by the effect of the inductance Lf of the cathode lead:
Cin = Cgf + C(g) g
1
Pin L = 9
W`CgfLfSf
(10-12)
where w is the anFular frequency of the input voltapes; and
Sf is the cathode current transconductance.
Due to the inductive effect of electron flow in the circuits of the control
and screen Frids, induction currents are penerated which govern the active input
impedance:
Pin T
kSW2T2
1
(10-13)
1
where k is a factor whose value is approximately ,0 for flat?electrode tubes;
T
is the transit time of the electrons between cathode and the control
prid; and
S is the transconductance of the total current from cathode to plate and
to screen grid.
The resistance is inserted parallel to the input capacitance and, consequently,
parallel to the resistance RinL (Fig.10-3c). Therefore, the total active input
resistance can be found from the equation:
1 = 1 1
= w2CgfLfSf + kSw2T2
Pin Pin L Pin T
or
1
Pin = 0
(CoLfSf + kST2)
77
(10-14)
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
-
?
?
Two conclusions can be drawn from eq.(10-1L):
a) In order to increase the input impedance, the tube must be so designed as
to provide minimum inductance of the cathode input and minimum electron transit
time;
b) The resistance of a given tube is inversely proportional to the square of
the voltage frequency on the control grid.
7
:.re may, therefore, write
Bin (megohms) =
9
f- (megacycles)
(10-15)
The magnitudes of the K factor for the most important types of amplifier
tubes are shown in Table 10-1.
Table 10-1
Type of
Tube
K,
megohms/mc2
Type of
Tube
K, 0
megohms/Mc'
6F9
3.0
2K2M
65
6P3S
6.5
6S1Zh
160
6Zh4
7
6S1P
160
6Zh8
20
6ZhlZh
200
6Zh3P
37
Section 10-5. A Tube for an Ultrashort-Wave Amplifier
An increase in transit time, combined with a reduction in wavelength results
in a reduction in R. .
In
An important consequence of the reductior,1 in input impedance is an increase in
the power required by the grid circuit:
At a iven volta e U the power of the grid circuit increases with decreasing
i STAT
78 1 i
t
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
- input impedance as the wavelength decreases. The power gain of the cascade then is
_ pout
(10-17)
Ak_._.; decreasing in the ultrashort -wave band with decreasing wavelength, due to a rise
ay_
- in P even when the power at the output Pout is stable.
in
Thus, in the ultrashort -wave band, the tube itself consumes a relatively large
portion of its own capacity for amplification.
In addition to a worsening of the amplifying properties of the tube, the gain
- of a cascade is further reduced in the ultrashort-wave range since the law input
16
impedance as well as the input capacitance of the tube in the next stage by-pass the
__ plate charge of the preceding stage to such an extent as to make its real level
_ quite small. Along with the reduction in resistance of the plate charge when the
- internal impedance of the tube is high, the output power declines considerably.
Thus, in the ultrashort -wave band, amplification encounters major difficulties
- due to the effect of the electron transit time and the influence of the inductances
of the tube leads and the interelectrode capacitances.
a) b)
Fig.10-4 - Ultrashort -Wave Receiver Tubes, Including Low-
Power Amplifiers
a - 6ZhlZh acorn pentode; b - 6K1P bantam tube
79
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
C.
?
?
?
One measure to facilitate an increase in input impedance is a provision in the
design, for low inductance at the cathode input. This is attained by shortening
the cathode pins, enlarging their diameter, and increasing their spacing. In cer-
tain types of amplifier tubes for the ultrashort -wave band, the cathode lead takes
the form of several (usually four) pins in parallel. Similarly, the inductance of
the plate lead is reduced because the plate is also provided with a lead by means
of two spaced conductors.
The reduction in interelectrode capacitances can be obtained either by reducing
the surface areas of the electrodes or by increasing their spacing. Small elec-
trodes compel the tube to work under more severe thermal conditions, since the amount
of heat the electrodes are to dissipate declines. In the long run, this also
?
limits the oscillating output power. However, ultrashort -wave receiving tubes,
including low-power amplifiers, usually operate as voltage rather than as power
amplifiers. Therefore, the small dimensions of the electrodes in this case are
entirely in order and constitute a standard design characteristic of ultrashort -
wave receiving tubes, including low-power amplifiers. As far as increasing the
electrode spacing is concerned, this procedure causes an increase in the electron
transit time and is not an acceptable method when the feed voltages at the elec-
trodes are low.
Figure 10-4 shows the most characteristic designs of ultrashort -wave receiver
tubes.
Section 10-6. Special Features of Radar Receivers
The specific features of radar receivers are determined by the fact that they
work in pulses and operate in the radar band of working waves.
The first group of special features of radar receivers is governed by the
very wide band required to pass pulsesignals reflected from a target, i.e., radio
pulses (before detection) and video pulses (after detection). Let us explain this
80
STAT
r ?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
4
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
in terms of an ideal rectangular pulse.
In accordance with FourierTs theorem, well-known in electrical engineering,
a periodically repeated video pulse voltage, like all periodic voltages of nonsinu-
soidal shape, can be considered as the result of the superimposition of an infinite
U.
imp
2U. K
imp
a)
Constant
Component 211. K
imp
Frequency of
component harmonics
Fig.10-5 - Spectrum of Periodically Repeated Rectangular
Video Pulse
a - Amplitude; b - Energy
series of constantly acting components.
When the starting point of the time scale has been properly selected, as
illustrated in Fig.10-5, the components of a rectangular video pulse will include
constant cosinusoidal components, invariable in magnitude and polarity, called
harmonics.. The frequency of the harmonics exceeds, by a factor that is an inte-
ger n, the pulse repetition rate F, where n is the number of harmonics.
A mathematical analysis of the case presented can be given in the form of the
following Fourier series:
n=
Um n=o
u (t) = +
2
n= 1
81
Umn cos n (21t Fu) t
?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in. Part- Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
or, in developed form,
Um n 0
u (t) = + Um cos (2 Fu) t Um 2 cos 2 ( 2 n Fu) t
2
4- Um 3 cos 3 ( 2 n Fu) t + . + Umn cos n (2n Fu) t
The relationship between the instantaneous values of the harmonic components
and their phases is such that, beyond the limits within which the video pulse can
be sustained, the total voltage of all the components equals zero.
The amplitudes of the harmonics with various order numbers are not identical,
although they are proportional to a single magnitude, the pulse value U of the
voltage. Thus, a doubling of U. causes the amplitudes of all the harmonics to
imp
double simultaneously. The amplitude of the harmonic depends on its order number,
the length of the video pulse, 'Cu, and the repetition rate Fu.
The following is the theoretical general equation for the amplitude of the
harmonic of order number n, for a video pulse of ideal rectangular shape:
sin (nick)
Umn = 2 U impk
(nn k)
(10-18)
where k - T- = T Fis the duty cycle.
The direct component of the video pulse is equal to half the amplitude of a
harmonic whose order number is zero (n = 0), i.e., of a harmonic whose frequency
is equal to zero. We know from the. theory of limits that the ratio of the sine
of an angle to the magnitude of that angle tends toward unity if the angle tends
toward zero.
Within the limits of our case .(n = 0) we obtain:
Um n=0 =2U. k
imp
As a result, the direct component of a periodically repeated video pulse
becomes
82
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
U0
or
2 U impk
2
U0 = Uimpk
(10-19)
Equation (10-18) shows that, depending on the number of the harmonic, its
amplitude may be either positive or negative, since the amplitude is governed by
the function:
sin (n k)
(nick)
The amplitude spectrum of the voltage of a periodically repeated video pulse
is shown in Fig.10 -5a. The adjacent harmonics of the spectrum are separated in
frequency by F.
The amplitudes of certain frequencies in the spectrum are zero (i.e., certain
harmonics do not exist). These are called zero frequencies. In order for the
amplitude of a harmonic to equal zero, we must have sin (nick) = 0. This is the
case when nick = c, 2n 3m ???, etc., which corresponds to the condition n = 1,
22 3, etc. From this we find the numbers of the harmonics of zero amplitude:
where
no
1 2 3
=
k=
Tu
Tu
Consequently, the zero frequencies of the spectrum are:
f01 = Fun01 =
FTu
f 02 = Fun02 =
Tu
83
(10-2o)
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
f03Fun03 7-3
The Same may be said of the current spectrum of a repeated rectangular video
pulse.
Each harmonic is Of an energy proportional - as we know from electrical en-
gineering - to the square of its amplitude. Therefore, the energ7distribution of
the harmonics through the spectrum of the video pulse (Fig.10-5b) is more clearly
defined than the amplitude distribution.
Now let us proceed to the spectrum of ideal rectangular, periodic radio
pulses obtained by modulating high-frequency oscillations with rectangular video
pulses.
Modulation by video pulses is a special case of amplitude modulation, in which
changes occur only in the amplitude of the oscillations, where they rise from zero
to the pulse value and then decline again to zero.
He know from radio engineering that, in sine wave amplitude modification by
several frequencies at once, the two sidebands - an upper and lower, each separated
from the carrier by the magnitude of the modulating frequency - of each of the
modulAting frequencies are added to the spectrum of the amplitude-modulated oscil-
lations.
/
The total frequency band of the-amplitude-modulated oscirations is equal to
the total width of the upper and lower sidebands, i.e., to a doubling of the high-
est modulating frequency.
Video pulse modulation is equivalent to simultaneous amplitude modulation by
all harmonic components of the video pulse spectrum, each of these components adding
its upper and lower sidebands to the radio-pulse spectrum (Fig.10 -6). Therefore,
Gil regardless of the harmonic number) the frequency band occupied by the radio pulse
is twice as wide as the frequency band occupied by the modulating video pulse.
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
(0
Co
A comparison of Figs.10-5b and 10-6 shows it to be quite obvious that, in the
video pulse spectrum, the very lowest frequency is equal to the repetition rate
is located in the low (audio) frequency band, and also has a direct component.
However, in the radio pulse spectrum, as in that of any HF amplitude-modulated
oscillation, low frequencies are entirely absent.
In order for periodic pulses to pass through an electric circuit without sig-
nificant distortions, it is sufficient to transmit only that portion of their spec-
trum that is highest in energy. We see from Figs. 10-5b and 10-6 that the maximum
energy is found in the frequency band limited (as far as periodic ideal rectangular
video pulses are concerned) by the first zero frequency, which is
1
A Fv = ?
(10-21)
while in the case of periodic ideal rectangular radio pulses, the maximum energy
is located in the frequency band limited by twice the first zero frequency, or
2
F = ?
P u
(10-22)
where Tu is the pulse length.
Equations (10-21) and (10-22) show that the pass band .of the electric circuit
must be wider, the shorter the pulse.
In deciding on the pass band of a radar receiver, the prime consideration is
the military purpose of the station it is servicing and the requirements to be met.
In all cases where it is desired to pick up all possible signals, no matter
how weak (as in distant-warning or circular-scanning stations), the band width of
the receiver in the high and intermediate frequencies is selected so as to provide
maximum signal-to-noise ratio, which gives a maximum receiver response. As already
discussed in Section 10-2, noises increase in power in direct proportion to the band.
width. However, starting at a given band width, the signal power increases much
more slowly than the noise power. This makes it necessary to find an optimum band,
85
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
0
*providing maximum signal-to-noise ratio. Therefore, significant distortions in
pulse shape cannot be avoided, and the band width for the high and intermediate
frequencies must be selected at less than that determined by eq.(10-22).
?
Fig.10-6 - Energy Spectrum of a Periodic Rectangular
Radio Pulse
As shown in practice, the best signal-to-noise ratio exists in the radio
receiver of high and intermediate-frequency bandwidth described by:
1
-AF = ---
pn T
,
(10-23)
? If the total maximum -expectecr_frequency drift of the transmitter and local
oscillator is represented by a quantity Afx, a stable reception of signals will
require the bandwidth to be expanded to
1
Fpn = + Afx
(10-24)
When automatic local oscillator frequency trim is provided, the magnitude
of Af is set at twiee the error of the automatic trim.
cec
The bandwidth of the video frequency cascades in the receiver is set at one-half
the width of the high-frequency and intermediate-frequency cascades which, in
86
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
0
*distant-warning stations, is
A Fvideo freq =
(10-25)
? and has been obtained from eq.(10 -23).
If the pulse beat is in microseconds, the bandwidth is found directly in
?
?
megacycles. Thus, if the length of the pulse T is 1 microsecond, the receiver
of a distant-warning station equippecrwith automatic frequency trim with an error
of 0.25 megacycles must have a pass band in HF and IF of
1
F = - + 2 x O. 25 = .1. 5 me
pn
while, in the video band, it should be
Fvid freq.
1
7 0. 5 mc
Where high-accuracy stations are concerned (such as a set used for gun-laying),
.the pass band must be determined from the conditions required for obtaining the
desired length of the leading front of the pulse Tv with allowance for the fre-
quency drift Afx of the transmitter and the local oscillator. The usual solution
in this situation is to decide on an HF and IF pass band for the set in accordance
with the following
.0.7 - 0.9
Fpn
tf
Correspondingly, the pass band in the video frequency must be
A Fvideo freq. =
0.35 - 0.45
f
(10-26)
(10-27)
The smaller the TV the larger must be the coefficient in eqs.(10-26) and
and (10-27). For example, in the case of pulses in which the front length Tf
is 0.2 microsecond, and the error of automatic frequency trim is 0.25 mc, the re-
ceiver of a gun-laying set must have a pass band for HF and IF, according to
87
?
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
eq.(10-26), of
0.0
A F = + 2X0.25 = 5 mc
Pn 0.2
When the front length is shorter, the band required is even greater.
Thus, the first group of special features of radar receivers results in the
requirement of a very broad pass band, exceeding tens of megacycles in high-accuracy
stations. The need to provide a wide pass band also affects the choice of the in-
termediate frequency (see Section 10-11), which, as distinct from the situation
with receivers for radio broadcasting (Fnp = 465 kc) is taken as 15 to 100 mc.
The second group of special features of radar receivers stems from the oper-
ating conditions of receiver tubes, including low-power amplifiers when used in the
radar band. As already indicated in Sections 10-4 and 10-5, it is expressed primar-
ily by a reduction in gain per stage. While, in the meter and decimeter wave-
lengths, amplification in accordance with the frequency of received signal presents
no serious difficulties, the centimeter band requires the use of traveling-wave
tubes (see Section 10-9) or else the idea of high-frequency amplification must be
discarded completely in view of the practical impossibility of satisfactorily
matching the plate load of the tube to its resistance. The various assemblies of
a receiver, working on the frequency of the received signal, do not always employ
oscillatory circuits with lumped parameters. Instead it becomes necessary to re-
sort to segments of long lines and coaxial-cylindrical cavities or cavity resona-
tors.
It is also necessary to take into account the effects of interelectrode capaci-
tances and inductances of tube leads. As regards atmospheric and industrial
noises, their level is considerably reduced in the radar band due to the fact that
the amplitudes of noise harmonics decrease continuously as their number increases,
becoming quite small in the band in which signals receivable by radar are located.
This is one of the most important factors requiring that the set noises of radar
88
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
S.
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
1
?
sets be reduced as far as possible, thus limiting its ultimate sensitivity.
Section 10-7. Block Diagram of the Radar Receiver
Radar receivers may display a great variety of circuit designs, including
straight amplification networks. However, the majority of radar sets are of the
1
?
atennna Converter
Input HF
circuit amplifier
Mixer
IF
amplifier
Local
oscillator Detector
To indicator VideO
amplifier
Fig.10-7 - Basic Block Diagram of a Superheterodyne
Radar Receiver
superheterodyne type.
1
Figure 10-7 presents a simplified typical block diagram of a superheterodyne
radar receiver. It shows only the most essential assemblies:
1) The input circuit whose basic purpose is to match the receiver input
to the HF channel of the antenna;
2) The BF amplifier, tuned to the frequency of the received echo signal;
3) The frequency converter, consisting in turn of a mixer and ailocal
oscillator and providing conversion from the frequency of the incoming
signal to an intermediate frequency more suitable for amplification;
4) The IF amplifier;
5) The detector converting IF radio pulses into video pulses;
6) The video-frequency amplifier, amplifying the video pulses to the level
89
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 :?CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
required for becoming visible on the indicator.
On wavelengths longer than 30 cm, radar receivers usually contain One or two
HF amplifier stages, five to seven IF amplifier stages, and two to three video-
amplifier stages. The reason so many stages are required lies in the difficulty of
obtaining high grain, due primarily to the use of very short waves and secondly to
the broad band over which amplification has to be provided.
The diagram shown in Fig.10-7 is generalized and therefore cannot correspond
to the actual electric circuits of all types of radar receivers. Thus, in the
centimeter band, for example, the receiver input often begins directly at the con-
verter (the UHF stares are lacking). On the other hand, it is not Uncommon for
the design of a iadar receiver to include auxiliary components lacking in the gen-
eralized block diagram of Fig.10-7. Such components include systems for automatic
frequency trim, automatic volume-control circuits, noise-suppression systems, etc.
Section 10-8: Input Circuits
The main problem with radar receiver input circuits is that of providing maXimum
signal-to-noise ratio at the grid of the first tube in the receiver. The solution
of this problem consists in obtaining the maximum input signal at the eid of the
first tube in the receiver and in matching the resistances. Therefore, the input
circuit is to be regarded as a resistance transformer matching the input impedance
of the HF channel of the set to the input impedance of the first tube in the re-
ceiver.
?
In stations operating in the decimeter and shorter wave bands, the input cir-
cuit is an integral portion of the antenna transmit-receive switch, previously
discussed in Section 6-11.
Figure 10-8 illustrates a typical input circuit of a 10-cm receiver. Matching
111 of resistances here takes place at several points of primary importance:
a) Matching the waveguide of the main line of the waveguide channel to the
coaxial line leading to the crystal mixer is accomplished by means of a cavity
90
STAT
1-;
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5 -
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
resonator, transforming the resistance when the arrester is not broken down;
b) Matching the local oscillator output so as to enable it to function on a
load resistance of optimum magnitude, is accomplished by means of a tuning plunger;
Cavity resonator with
receiver input
discharger
Coaxial line Crystal
Waveguide to mixer mixer
from main Feeder output
line of of signal to
waveguide Local IF amplifier
channel oscillator
Local o output
oscillator
output tuning
plunger
Fig.10 -8 - Typical Input Circuit for 10 -Centimeter Receiver
c) Matching the input resistance of the first tube of the IF frequency trans-
former with the line characteristic of the output feeder from the crystal mixer.
A reflex klystron serves as local oscillator in the centimeter wavelengths
(see Section 10-15), while a low-power lighthouse tube, combining in its design
the tube proper and coaxial-cylindrical cavities in the form of plate and cathode
circuits, is used in the decimeter band (Lighthouse tubes are described in Sec-
tion 11-9). Coaxial-cylindrical cavities are also used as antenna channel matching
transformers.
The line characteristic orthe feeder from the crystal mixer is matched to the
input impedance of the first tube of the amplifier by means of a quarter-wave co-
axial line (Fig.10-9) with predetermined line characteristic, constituting a
feeder transformer (for intermediate frequency). The line characteristic? of this
transformer has to be Pt = VP
iRinl, where Rinl is the input impedance of the first
tube for intermediate frequency. '
91
STAT
Declassified in Pari - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
For one-meter and longer waves, amplification at the frequency of the received
signal is now universal practice. The local oscillator and the frequency converter
(mixer) are taken out of the antenna input channel and become an integral part of
the receiver set, which thus starts with the HF amplifier (Fig.10 -10). A Class II
oscillatory circuit acting as a transformer is inserted at the input of the first
UHF tube, and is tuned to the frequency of the received signal and connected to the
antenna feeder so that the equivalent resistance of the circuit between the connec-
tion point of the feeder and the ground is equal to the feeder line characteristic.
The conditions for matching are:
2
(PLP)
e =
= pf
1
where p = is the circuit coupling factor;
is the line characteristic of the circuit;
f is the line characteristic of the feeder; and
is the loss resistance of the circuit
Feeder from crystal mixer
Quarter-wave coaxial
matching line
First Tube of
IF amplifier
Fig.10-9 - Matching to Input of the First Tube of IF Amp-
lifier in Receivers for Decimeter Band
Section 10-9. High-Frequency Amplifiers
We know that the gain of a stage is:
92
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
IL Re
K Um out
Um in Umg
In turn, the amplitude of the plate current of an amplifier tube is
'ma = Sd Umg
in which sd is the dynamic trans conductance under the operating conditions of
amplification:
Re is the equivalent resistance of the plate load;
Umg is the amplitude of swing.
From antenna
First cascade of HF
amplifier
R.
in
Fip.10-10 - Typical Input Circuit of Receiver for
Very Short Waves
In the radar band, the resistance to the plate load is so strongly shunted
by the input resistance of the next stage that Re may be taken to approximate Rin.
The equation for the gain of the HF amplifier stage will then be
K = SR.
in
From this we arrive at the conclusion that an increase in the gain of the HFampli-
fier stage requires tubes wip high transconductance and high input impedance.
Bantam penfodes are used for HF amplification on waves of the order of several
STAT
93
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
meters in length.
A typical HF amplification circuit, using a 6Zh1P circuit, is shown in
Fig .10-11.
This circuit employs parallel feed of the tube plates, a resistance Rba being
used in place of an HF blocking choke. The tube-filament feed circuits are blocked
by capacitances and small chokes, blocking the path of HF currents. Sometimes
chokes are omitted from the filament circuit, being replaced by a coiling of the
filament wires into a tight spiral.
Innut
signal CbH
Rba
Lbfl
Fig.10-11 - HF Amplifier Using a 6Zh1P Pentode
In order to reduce the shunting effect of the input impedance of the next
stage on the plate circuit of the preceding 'stage, the circuit is only partially
connected to the grid circuit of the next tube. This causes the circuit to become
a matching resistance transformer. The effective load on the amplifier tube plate
circuit, and consequently the gain of a stage wired in this manner, depend on the
coupling factor of the circuit:-
W2 U9
VY1 u 1
where W2 is the number of turns of the circuit coil coupled to the grid circuit
of the next tube;
VI is the full number of turns in the circuit coil;
94
Declassified inPart - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
u2 and ul are, respectively, the voltage delivered to the grid of the following
tube, and the total voltage in the circuit.
At very low values of m, the shunting effect of the input impedance is negli-
gible, and a comparatively high voltage is developed in the circuit, of which a
very negligible fraction is utilized so that the pain of the stage is low. At m val-
ues close to unity, the circuit voltage is delivered almost entirely to the grid
circuit of the following tube, but its input impedance shunts the circuit excess-
ively so that the gain of the stage is again low.
Obviously there must be some optimum value at which the pain is at a maximum.
It can be demonstrated that
m - ?
opt w
( W2
1) opt
where Rin is the input resistanxe of the next stage;
Re is the equivalent resistance of the plate circuit network.
This yields the following as maximum? amplification factor:
1
Kmax = Fairia
(10-28)
In the very short wave band it is difficult to obtain an Re of more
than 5 x 103 ohms, The for-, if thc-i..puu impedance of-the amplifier pentode,
. _ . .
operatingatthefrequencyofthesignal,isaboutR.--350 ohms and the trans-
in
conductance S is 5 ma/v, the maximum gain of the stage will be, in accordance
.with eq.(10-2),
5 x 10-3 ./
Kmax = 2 V5 x 10-3X 350 '-;-; 3.3
Thus, the gain per cascade is quite small, even in the very short wave band.
Shortening of the wave makes it necessary to dispense with pentode amplifica-
tion, because of the increasing noise level, and to use triodes instead. However,
the usual triode coupling employing a grounded-cathode circuit, cannot be used in
STAT
95
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
view of the high possibility of self-oscillation due to back-coupling of the plate
circuit with that of the grid across the plate-grid capacitance and the inductance
of the cathode input.
sack-coupling between the plate and grid circuits in the 1-1.5 m wavelengths
is reduced by means of grounded-grid triode HF amplifiers (Fir.10-12), invented
by M.A.9onch-Bruyevich. The amplifying voltage is supplied to a portion of the
cathode circuit and is connected between grid and ground, as usual, while the out-
put voltage is taken off a circuit cut in between the plate and the grounded grid.
Thus, the grid and plate circuits-are separated by a shield in the form of the
grounded grid, eliminating feed-back between these circuits and greatly weakening
the possibility of self-excitation.
In constructing the equivalent circuit of an amplifier employing a triode and
grounded grid, it must be borne in mind that the voltage applied to the grid is
effective twice: as amplified by a nactor of 11 and as directly cut into the
Fip.10-12 - Triode Amplifier with Grounded Grid
circuit of the series-coupled plate and grid circuits (Fir.10-12b).
In accordance with the equivalent network, the plate current is
(P. 4- 1) 1-jg
la
Ri na
96
STAT
A
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
and the rain per stage is
Uout + 1) Ba Ra
K=
ni Ra Ri (1 4. a
n
Ri
Uin
In view of the fact that
we obtain:
4 >> 1, while Ili >> Ba
K S a (10-29)
The input impedance of a stage with grounded grid
!Jill
. Ug . Ri + Ra , Ri + Ra
Rin --
"?? + 1 11
Ia
'in
When R1 ? Ra we obtain
1
Bin =
1(10-30)
To increase the pain, incomplete coupling of the circuit either on the input
or the output side is used. Here again there is an optimum coupling factor:
mopt =
T.%
na
which, in accordance with eq.(10-28,), pi.ovides the maximum amplification factor:
1
Kmax = S IFTRa
(10-31)
The theory of amplification by grounded-grid triodes-has been worked out in
detail by V.I.Siforov, and A.A.Kolosov.
Because of their very wide frequency: range, the circuits of HF amplifiers for
radar receivers, as distinct from broadcast-band receivers, have fixed tuning, and .
the transmitter is not trimmed when frequency drift occurs. Moreover, sets oper-
ating close to the limit of the very short wave band do not even contain capacitors
97
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
in the form of coils. The capacitances of the circuits are provided by the stray
capacitances of the network.
To amplify oscillations when the frequency of the received signal is of the
. order of 3000 mc and more (microwave band), a special category of vacuum tubes is
used: the traveling-wave (TW) tube. The diagram of this type of tube is shown in
Fig.10-13.
Anode of elec-
tron gun Waveguide
input
Cathode
The primary components of a
TH tube are:
Waveguide 1) electron gun;
output
Coil 2) spiral;
3) focusing coil;
4) collector.
The electron gun consists of a
heated cathode creating a beam of
Focusing coil
Heater Collector
?
electrodes, and a plate accelera:ting
Fig.10-13 - Diagram of the Traveling-
the motion of the beam. Through an
Wave Amplifier Tube
aperture in the narrow portion of
the anode, the beam travels by inertia along a closely coiled wire spiral and is
picked up by the collector, with the same potential as the anode (+1000 to 1500v).
A long focusing coil occupying the neck of the tube serves to compress the beam,
in the direction of the axis of the spiral.
The signal being amplified enters through the input waveguide and induces a
traveling electromagnetic wave, propagating along the coils of the spiral at the
velocity of light. The longitudinal component of this velocity, which is the speed
at which the wave moves along the axis of the spiral, is smaller than the velocity
of light by the same factor by which the spacing of the windings of the spiral are
smaller than the length of one of its turns.
The traveling electromagnetic wave acts on the electron beam. If the velocity
98
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
0
?
?
1. ?
of the electron beam is somewhat greater than the longitudinal velocity of the
electromagnetic wave, thebeam is retarded and yields energy to the electromagnetic
wave. As a result, the voltage of the field induced in the outgoing waveguide
is 10 to 15 times as great as in the incoming.
Amplification with a TW tube ensures a relatively low level of set noise and,
in addition, does not set up a tuned amplifier. A Ti! tube, like a waveguide, func-
tions over a very wide band and does not require tuning to the frequency of the
incoming signal.
Section 10-10. Frequency Converters in Radar Receivers
The problem of frequency conversion in the radar receiver lies in transforming
a radio pulse at the frequency of the incoming echo signal to a pulse of a lower,
intermediate, frequency, useful for further amplification relative to the noise
level and for obtaining a band wide enough to provide a permissible distortion of
the shape of the echo signal.
As in all superheterodynes, the radar receiver includes a mixer actuated by
the frequency voltage of the incoming echo signal fs, and the frequency voltage fh
of the heterodyne, a low-power local oscillator with direct drive, also constituting
a standard component-Of all suporitet7aredyne receiverB.
According to the classic theory of frequency conversion, developed by
Prof.V.I.Siforov, the effect of two different 'frequencies, fs and fh, on a voltage
mixer causes significant changes in its parameters, of such a nature that the
transconfluctance of the mixer acquires a significant component S representing
an intermediate frequency fif = fs - fh. If the mixer is loaded with a load tuned
to the frequency f this load can produce an IF voltage across a resistance R.
Like all stages, the mixer has a given gain factor in intermediate frequency:
Kif = SifRa
99
(10-32)
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
As the conversion transconductance is only a portion of thb total transconduc-
tance S of the mixer characteristic, the gain of a tube when used as a mixer is
always less than that of the same tube when used as an amplifier.
Both the incoming signal frequency and the local oscillator frequency exceed
IF
the frequencies of ordinary radio broadcast receivers. Therefore, the noise levels
and input impedances of ordinary converter and mixer tubes (pentagrids, hexodes,
etc.) make them entirely unsuited to use in radar receivers.
In very short wave receivers, the same types of tubes used for amplification
are used for conversion, i.e., special pentodes or triodes. However, while in
broadcast-band receivers, the incoming signal voltage and the local oscillator
voltage are received at different tube grids to reduce the mutual effects of signal
and local oscillator circuits, in radar receivers the object of noise reduction and
increase of amplification is served by connecting the converter, the signal voltage
and the local oscillator voltage to the same grid of the.mixer tube or, in other
words, single-grid conversion is employed.
Figure 10-1h illustrates one diagram for a single-grid mixer for waves at the
end of the very short wave band. It also shows a local oscillator assembled around
a separate tube Lh in the Hartley arrangement. The input circuit of the mixer tube
is tuned to the freoency oC the signal and is, as a consequence, considerably mis-
tuned relative to the.local oscillator. The need for an input circuit capacitance
is met by the input capacitance of the mixer tube, the capacitance of the wiring
and the capacitance C of the circuit coil L. In terms of design, the circuit is
executed as a coil with a brass or carbonyl trimming core. Incomplete coupling
of the input circuit permits matching to the plate load of the previous stage, the
HF amplifier. The tuning of the mixer input circuit is' fixed as is that of all
the HF circuits. When frequency drift of the incoming :signal occurs, only the
local oscillator circuit is tuned, by means of its capacitor CH' to provide con-
stant intermediate frequency.
100
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
The automatic grid bias of the mixer created by the cathode current,across the
resistance R 1 is set close to the-mixer-tube grid cutoff voltage so as to operate
mx
to HF *amplifier to IF amplifier
mx
Ch
Fig.10-14 - Mixer and Local Oscillator of Very Short Wave Receiver
In tne range of maximum vda.Lat,--i-orrin totaI-mixtube-transconductaace,
Local oscillators for the micrgyaVe band are usually assembled around light-
house tubes, while the mixer tube is a diode which, while providing no amplification
has considerably more input impedance and creates less noise than a triode or
pentode.
Figure 10-15 shows a typical diode mixer diagram. The signal and local oscil-
lator voltages are impressed in series onto the mixer-tube circuit, as with pentode
or triode mixers. The IF voltage is taken from a circuit wired as a coil Lir To .
provide the most favorable operating-conditions for the mixer diode and to obtain
maximum conversion effect a supply source E is cut in.
mx
101
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
C,
At wavelengths of less than 50 cm (particularly in the microwave band), crystal
detectors are used as mixers. Th 6 coupling diagram of a crystal mixer has already
Signal fs
Local oscillator
voltage fh
To IF amplifier
4MK fif
LiF
Fig.104.5 - Typical Diagram of Diode Mixer
been illustrated in Fig.10-8. The operation of the crystal detector is based on
the fact that the contact between the conductor and the semiconductor crystal
a) b)
cr
Germanium
Silicon
'Fig.10-16 - A Crystal Detector
a - Design; b - Performance curve
?
conducts in one direction only. Silicon or germanium crystals are used.
The electric characteristic of a crystal dectector. is sufficiently linear
(A/ (Fig.10-16b). The slope of the curve in the region of positive voltages determines
the direct resistance of the contact, while the slope in the region of negative
102
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
voltages governs the back resistance of the contact. The normal direct resistance
of silicon crystals is in the 50 to 500-ohm range, and the back resistance in the
1000 to 20,000-ohm range. In view of the presence of a design capacitance of the
order of 1 R?f, shunting the contact layer, the input impedance of the crystals
is complex at high frequencies and has a capacitive reactance. Its active component
of the order of 20 to 60 ohms (usually, 50 ohms). The output impedance of the
crystal at intermediate frequencies is also complex (capacitive reactance), the
active component being of the order of 200 to 600 ohms (usually 300 ohms).
In design (Fig.10-16a), the crystal detector is built into a ceramic body
with brass terminals (5) and (6), the former being connected to the inner wire of
the outgoing coaxial line to the IF amplifier and the second to the inner wire of
a coaxial line, simultaneously affected by the signal and heterodyne frequency
fields (Fig.10-8).
The crystal (1) is soldered to the adjusting screw (4). A tungsten catwhisker
(2), fastened to the lead terminal (6), serves as contact conductor. The closeness
of contact between crystal and wire is controlled by the screw (4). To prevent
the parts from moving, the internal space in the ceramic body (3) is filled with
wax.
The advantages of the crystal mixer relative to the diode are as follows:
1) Considerably lower input capacitance, significantly simplifying the
design of both the receiver input circuit and that of the intermediate-
frequency amplifier;
2) The linearity of the curve at low voltages, ensuring effective oper-
ation when the input signal is weak;
3) The reduced noise level, due to absence of an emitter cathode.
The main drawback of the crystal mixer is its sensitivity to overloads, due
to the small contact area between the wire and the crystal.
103
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
*Section 10-11. Determination of Intermediate Frequency
In determining the intermediate frequency of a radar receiver one must accept
a compromise solution that will satisfy, to the best possible degree, a number of
contradictory requirements.
1. The intermediate frequency shoilld not be in the band of powerful ultrashort-
wave broadcasting stations, to y2event significant interference with reception of
echo signals.
2. The intermediate frequency must be low enough to permit weakening of the
noise level of the IF amplifier and to ensure minimal detuning of the circuits due
to parameter spread when tubes are changed since, at the higher frequencies, the
?
interelectrode capacitances of the tubes comprise a greater portion of the'total
circuit capacitance than is the case at lower frequencies.
3. The intetmediate frequency must be sufficiently low to ensure a high gain
per stage, permitting a reduction in the total number of IF amplifier stages.
4. The intermediate frequency must be sufficiently high to weaken interference
by stations in the image channel. The image channel operates at a frequency that,
as distinct from the frequency of the echo signal, is not higher but lower than the
local-oscillator frequency by the value of the intermediate frequency. Therefore,
tnrough-tne mixer, the mirror station also yields an intermediate
frequency and is consequently magnified in the intermediate frequency amplifier.
The higher the intermediate frequency, the greater will be the difference of the
frequency of the incoming signal from that of the mirror station, and the greater
will be the degree to which the input circuit of the receiver is out of tune from
the station in the mirror channel.
5. The intermediate frequency must be sufficiently high to prevent its voltage
from coinciding with the pass and of the video amplifier and thus be amplified.
This condition is satisfied if the frequency is five times greater than that of
the video amplifier pass band, i.e" it must meet th'e condition:
104
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R0018on9nnnn1_c
STA
1 r
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
f > 5AF
if - video freq
where AFvideo freq is the pass band of the video amplifier.
6. The upper limit of the intermediate frequency is determined by the resist-
ance of the stage to direct drive due to feedback across the plate control-grid
capacitance. As demonstrated by Prof.V.I.Siforov, the stable grin factor of a stage
is
./ S
Kstable I 0.42 V
C ag
An increase in the intermediate frequency wir is accompanied by a decline in
the factor Kstable and requires an increase in the number of stages in the
IF amplifier.
The frequencies from 15 to 100 mc represent a compromise solution of the prob-
lem. The intermediate frequencies most frequently used in radar receivers are
three: 15, 30, and 60 mc.
Section 10-12. Intermediate-Frequency Amplifiers
The most commonly used intermediate frequencies in radar receivers are 30
or 60 mc__Alt.hollgh these freclac.. already exi _nto the ultrasbort,wave band
(5 and 10 m), electron transit time does not yet make itself felt and there is no
significant drop in tube input impedance. The special difficulties involved in
operating on the ultrashort-wave band are not encountered in amplifying intermediate
frequencies.
Most of the total amplification in the receiver, which is of the order of
several millionf old, occurs in the intermediate-frequency amplifier. It is import-
ant to note in this connection that IF amplifier circuits are determined for the
most part not by the operating wave of the receiver but by the specifications for
frequency range, i.e., for quality of reproduction of the pulse envelope, and by
STAT
105
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
the total gain factor in the intermediate frequency. It should be added that, if
amplification in the intermediate frequency precedes HF amplification, the IF amp-
lifier stages need meet no special requirements as far as noise is concerned, since
set noise under these conditions is determined by the HF amplifier stages, the
mixer and the local oscillator. In the case where the IF amplifier is the first
amplifier stage of the receiver, the noise it creates requires very serious atten-
tion, a fact that is naturally reflected in the IF initial stage amplifier diagrams.
Thus, the decisive factor in determining the IF amplifier diagram is the
IF pass band required.
As we know from radio engineering, the equivalent resistance of a class I
parallel circuit is:
or
p2 p
ne = = X P
r r
I
r
L
wherep=.-is the characteristic impedance of the circuit;
C
d = I is the attenuation of the circuit.
We also know that the pass band of the circuit is
A F = dl'
o
(10-33)
where fo is the resonant frequency of the circuit. From this we derive
Substitutinr eq.(10-34) in eq.(10-33), we obtain:
Finally, considering that p =
foP
Re = A F
(10-34)
1 and wo = 2nf0' we derive the equation for STAT
0C
106
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
' frequency range:
:the equivalent resistance the circuit reqUires in order to provide the required
:1
(10-35)7_
Re
2W,AFC
'where C is the capacitance of the circuit.
From this it follows that, in order to obtain the greatest possible Re (i.e.,
maximum gain per stage) at a given AF, it is necessary to reduce the capacitance
of the circuit, whose minimum value is limited by the stray capacitance at the
stage output, of the order of 15-25 Rt.if. For example, to provide a band AF = 6 mc
the circuit required must have an equivalent resistance of
1
Re =1000 ohms
2n x 6 x 106'x 25 x 10-1' '
In order to obtain such low Re values, the IF circuits have to be designed as
coils, usually with an inductance of several microhenries, shuntable by a resistance
of the order of 2-4 ohms, in order to pass the required band of frequencies.
To obtain satisfactory gain per stage at a law resistance Re to plate load,
it is necessary to use amplifier pentodes of high transconductance (7-12 ma/v) and
employ a large number of intermediate frequency amplifier stages (from L. to 9).
Amplifier networks for the intermediate frequency may be classified in two
main groups: single-circuit and twin-circuit.
In single-circuit networks (Fig.10-17a), the circuit coil Ic is coupled to
the grid circuit of the following stage and shunted IV the resistance Rsh, coupled
to the plate circuit of the preceding stage. Should the total gain in the inter-
mediate frequency be comparitivel:i small, the circuits of all IF amplifier stages
are tuned to the middle of the IF pass band. However, when high IF amplification
is required over a relatively wide band, single-circuit systems with symmetrically
detuned pairs of stages, one tuned to a frequency below the midfrequency and the
other to a frequency above this. Finally, when the band used is very wide, exceed-
107
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
'in' 15 mc, networks employing several cascades of three stapes are used. The first
stare in the cascade of three is tuned to a frequency below the midfrequency, the
a)
from
preceding
stage
Rsh
Lc
b)
Fir.10-17 - Single-Circuit IP Amplifier
a - Diagram; b - Resorance characteristic of
a system of three detnned stages
fif
Total resonance
curve of amplifier
second only to the midfrequenc7 itself, ard the third to a frequency somewhat above
it.
Figure 10-171, depicts the resonance curves of three cascades staggered in
the marner irdicated. The total resorance curve of the three cascades is found by
multiplication of the ordirates of the resonance characteristics of these stages,
since in a multistage amplifier tl-e total rain is the product of the rairs of the
individval stages. The total specified bard of IF amplification is obtained by
choice of the parameters of the circuits of irdividnal stages and of the degree of
stagr-er. In this manner, irdividnal stare pass bands are obtained that are suf-
ficiertl'r narrow to provide good amplification per stare.
If there has bee- no pre-amplification ahead of the IF amplifier (as in the
ricrowive receivers of rur-laTtirr sets), the networks of the first stares of the
110 Amplifier are Zfected someW,Lt. In view of the fact that it is necessary to
r(!duh.e the ,-oise level to a minimm, the first two stages use triodes instead of
Ty:rtode3. The first triode is coupled on a grounded cathode network, since this
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
provides for less noise. In order to reduce or eliminate self-excitation due to
grid-plate capacitance, the plate load of the first stage is intentionally reduced
at the expense of its gain. This isdone by connecting a second triode in the
grounded grid coupling. It should be remembered that the input impedance of a cas-
cade with grounded grid is virtually independent of the frequency. If the impedance
is fairly large at 200-300 mc, then at 30 to 60 me, on the other hand, it becomes
small relative to a network with grounded cathode. The result of the strong
shunting action of the input impedance of the second stage causes the first-stage
gain to decline, so that self-excitation becomes more difficult.
Figure 10-18 illustrates such a grounded cathode - grounded grid network.
From the mixer output, the signal is impressed on the first tube grid across the
input transformers Ls and L 0 matching the output impedance of the mixer to the
amp
input impedance of the IF amplifier. The diagram also shows a device used for
from mixer to next IF amplifier stage
L L
s amp
Ch1 Ch2
crystal current
Fig.10-18 - Grounded Cathode - Grounded Grid Network
controlling the direct component of the crystal-mixer current. The filters C1,
Ch1 and C2 protect the device from the effects of intermediate frequency. The
choke Ch2 protects the HF component from contact to ground.
Two-circuit diagrams of IF amplification (Fig.10-19) include, in the plate
circuit of each stage, a band filter of two inductively-coupled circuits, shunted
-109
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
by the resistances Rsh a and Rsh g? respectively, to provide the necessary pass
band without significantly reducing the gain per stage. The coupling between the
circuits is usually set at the critical, when the resonance characteristic still has
only one maximum. Less frequently, this characteristic is set at above the critical
level in order to obtain a two-peak filter resonance characteristic.
Where two-circuit diaprans are used, pairs of IF. amplifier stages, symmetri-
cally staggered relative to the middle of the pass band, are used.
Section 10-13. Detection
The purpose of the detector in a radar receiver is to convert the IF radio
pulses into the video pulses required, after further amplification, for direct
reading of the target coordinates on indi-
Rsh a
sh E cator screens.
In general, detectors have to satisfy
the following conditions in order to meet
this requirement:
from preceding a) Undistorted detection of IF sig-
stage of IF
amplifier nals, i.e., undistorted repro-
duction of the envelope of the
_
amplified and converted radio
Fig.10-19 - Twin-Circuit IF Amp-
pulse reflected from the target;
b) Provision of maximum transfer
constant for the amplitude of the detected pulse;
c) Provision of minimum penetration of the IF voltage into the video amp-
lifier grid of the first stage;
d) 7o reduction in the signal-to-noise ratio established at its input by
the preceding stages of the receiver.
1n principle, all types of detection known to radio engineering may be employed
in radar. However, diode detectors are preferred, since they provide the broadest STAT
lifier
Declassified in Part - Sanitized Cop
110
y Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
dynamic range or, in other words, the widest band of detectable radio pulse ampli-
tudes resulting from differences in the target range.
a)
b)
Ch1 u. Cf
in out
square-
law detection
Cho
Cin
video amplifier first stage
uin
linear
detection
Fig-10-20 - Diode Detection
a - Typical network; b - Detection characteristic
Figure 10-20b illustrates the dector characteristic of a diode detector.
Depending on the swing of the detectable signal, passing from the IF ampli-
fier -output to the detector-input, detection ic classified as sq,,PrP-71=4f-,--whir-h
embraces the lower curvilinear (approximately square-law) portion of the detector
curve, and as linear detection, embracing the upper (linear) portion of the detec-
tor characteristic.
In square-law detection, the amplitude of the pulse at the detector output is
approximately proportional to the square of the amplitude of the IF pulse at the
detector input. Therefore, it is desirable that detection of weak signals be in
the square-law band (and the IF amplifier is designed accordingly) in order to
arrive at the signal-to-noise ratio that is so important in this connection.
111
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
In linear detection, the pulse amplitude at the detector output is approxi?
mately proportional to the first power of the IF 1,u1se amplitude at the detector
input. Therefore, when hookups providing for noise suppression are used, it is
better that detection occur in the linear interval.
Figure 10-20a presents a standard diode detector diagram.
The choke Ch1 is designed to prevent short?circuiting of the IF current of the
IF amplifier final stage. The Cf filter and Ch are primarily designed to ground
the IF component at the detector output and secondly to block the path of IF cur?
rents across the load resistance Rn and beyond that to the video channel of the
receiver. The filter choke Ch2 has a natural frequency equal to the intermediate
frequency and behaves like an antiresonant parallel circuit connected in series or,
in other words, like an IF wavetrap. Thanks to this entire system of IF filters,
only the video frequency component of the detected radio pulse spectrum or, in
other words, the video pulse voltage is discriminated at the detector load resist?
ance Rn.
In the majority of cases, negatively polarized video pulses are discriminated
by the load resistance. This permits limitation of the amplitude of the echo
signals in the first stage video amplifier, with the object of eliminating the
possility of overload video-ampllfier stages.
The analysis of 61e detector parameters consists in determining the load re?
sistance. This is based on:
a) The equivalent load capacitance
Ce = Cf + C +C.
out In
where Cf is the capacitance of the IF output filter;
is the output capacitance of the detector diode;
out
C. is the input capacitance of the first stage of the video transformer;
In
b) the required video channel pass band AFvid f.
112
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
The detector load resistance, like the equivalent resistance of the IF ampli-
fier circuit, is determined from the equation:
1
n 5- 2 n Fvid iCe
(10-36)
The transfer function of diode detectors in radar receivers is in the range
of 0.1 to 0.7.
Section 10-14. Video Amplifiers
The purpose of video amplifiers in radar receivers is to amplify detected
signals to a level adequate for normal indicator operation. It should be mentioned
that this level is not always determined by the operating voltage required for
normal tube operation. Very often the indicator contains its own supplementary
video amplifier. In this case, the receiver has to amplify the video pulse to a
level adequate for normal operation of the auxiliary video amplifier of the indi-
cator (approximately to 5 or 8 v). Sometimes the indicator is as much as some tens
of meters distant from the receiver, and the outgoing video signal is transmitted
by coaxial cable. In this case, the receiver video amplifier has to work on a low-
resistance matched load Rn = pf, where pi, is the line characteristic of the con-
--rt-ng feeder.
In the vast majority of cases, video amplifiers are designed for amplification
across resistances. Figure 10-21a presents a generalized diagram of such an amp-
lifier. As we know, the frequency response curve of such an amplifier (Fig.10-21b)
has two steep slopes located
a) in the low-frequency band of the amplified signal, due to the very high
resistance of the separating capacitor C at low frequencies;
b) in the upper frequencies, due to the output capacitance of the stage and
the dynamic input capacitance of the following stage which, at high frequencies,
is forcefully shunted by the resistance Ra of the video-amplifier plate load.
The pass band of resistance amplifier does not, in accordance with the diagram STAT
113
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
in Fig.10-21a, exceed
1
(10-37)
where Ce = C + C + Cin 2 denotes the equivalent capacitance of the
out 1
wiring
diagram.
a)
b)
+ C . .
outi wiring
C.
1n2
out
steep slope in low steep slope in high
frequencies frequencies
Fig.10-21 - Resistance Amplifier
a - Generalized circuit; b - Frequency response
One of the major requirements to be tet-by a video amplifier is that the
build-up and decay time of the video pulse be brief. The briefer that
period, the higher will be the *accuracy of ranging. If no special steps are
taken to meet this requirement, the presence of even minimal stray capacitances
make it necessary in the conventional resistance amplifier design, to reduce the
resistance to plate load, i.e., to widen the pass band in accordance with
eq.(10-37); this results on a reduction in the video amplifier gain per stage.
However, the requirement that the gain be increased necessitates, on the other
hand, an increase in plate-load resistance, and this, in accordance with
eq.(10-37), leads to a reduCtion in the pass band and is accompanied by an increase
114
STA
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
in the rise and decay time of the output pulse.
In order for an adequately uniform frequency response to be obtained through-
out the video amplifier pass band, determined as previously described (see
Section 10-6) or, in other words, in order to assure low build-up and decay time
in the face of high plate load resistance, the video amplifier diagram should in-
clude components for correcting the frequency response in the high as well as in
the law frequencies.
Since the lowest frequency in the spectrum of the repeated video pulse is
determined by the repetition rate, and the steep response slope at the low fre-
quencies usually becomes apparent only in the interval below 200 cycles, it is not
a)
comp
b) with compensation
without compensation
Fig.10-22 - Video Amplifier with HF Compensation
a - Diagram; b - Frequency response
necessary to provide low-frequency
ceeds 200 cycles.
correction when the pulse repetition rate ex-
Thus, when the frequencies are sufficiently high, frequency response correction
need be repeated only in the high frequencies. A video amplifier meeting this
(
1.1-5
STAT
- Declassified in Part- Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
requirement would have the diagi.am shown in Fig.10-22a, and the frequency response
shown in Fig.10-22b. The simplest type of compensating circuit is used here, in
the form of an inductance L coupled in series to a resistance Ra of the amplifier
corn
plate load. The compensating inductance and the equivalent capacitance Ce =
= Cout
C wiring + Cin 2 form an oscillating circuit of high attenuation, pro-
viding a smooth rise in frequency response at the resonart frequency of this circuit.
The amount of inductance is so selected as to yield resonance at a higher
frequency than fe elope, at which a steep slope below the 0.7 level begins
(Fig.10-22b). This compensation widens the pass band relative to the cutoff fre-
quencyfe slope by a factor of V7F, where the inductance, subject to correction,
can be calculated from the equation
= 0.25 RaCe
?
comp
The diagram of a video amplifier with compensation only at low frequencies
111 (to simplify the drawing) is shown in Fig.10-23a. The compensating low-frequency
circuit consists of a resistance, R and a capacitance Ccut into the plate
comp comp'
circuit. In the low-frequency range, the resistance of the caipacitor C is of
comp
adequate magnitude and does not shunt the resistance R
comp
Corwequexily, at low fi it may he co.,,Id,A-t..717-that a suppleme.rtgxv--
/
resistance Rcomp is connected in series with the resistance R. As a result of the
increase in the total resistance of the plate load of the amplifier tube, there is
an increase in the pain factor at low frequencies, with the result that the steep
slope in the amplifier frequency response at low frequencies is eliminated. At
intermediate and high frequencies, the resistance R comp is so strongly shunted by
the capacitance Ccomp that it has no significant effect on the gain per stage.
3atisfactory compensation of the frequency response in the low-frequency band is
411 usually obtained if R 5Ra and; in addition, if the folloiling condition is
comp
satisfied:
L1.6
STAT
;ZI
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
RaCcomp = RgCe
(10-38)
The input resistance of the indicator circuits contains a large capacitance
factor, strongly shunting' the load resistance of the final stage. It is, therefore,
a)
video pulse
to be
amplified
Ftcomp
with compensation
without compensation
200 cycles
Fig.10 -23 - Video Amplifier with Low-Frequency Compensation
-----------
a - Diagram; b - Frequency response
desirable that the load resistance of this stage be low. In this case, the build-
up and decay time of the output pulse will be small and the pass band will be of
adequate width even without any compensation of frequency response. This is of
particular importance when the video amplifier is working on a connecting coaxial
cable some tens of meters in length, whicn, for purposes of matching, is loaded
at the indicator side with a resistance equal to the characteristic impedance. No
significant shunting of so small a"load resistance Rk = p of the indicator input
ven if the frequency band is quite broad.
117
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
For effective operation against a low load resistance, the output stages of
the video amplifier are usually designed in the form of the cathode follower shown
in Fig.10-24a (in the case of a long cable) or in Fig.10-24b (when a short cable
functions as a small shunting capacitance). In the cathode?follower network, the
load is small and is coupled into the cathode circuit in any manner, i.e., between
cathode and ground. The output voltage Uout, separated by the resistance Rk, is
reversedinthepridcircuitcountertotheinputvoltageU,
.which signifies the
In
presence of a negative (antiphase) couple in the cathode follower. Due to the
negative feedback, reducing the resultant grid voltage, the cathode follower is
unable to function as a voltage amplifier. In fact, if we neglect the grid current
relative to the plate current, it may be considered that the output voltage is
=
Uout iaRk
The plate current is determined by the transconductance S and by the resultant
111 grid voltage which, in view of the presence of negative feedback, will equal
so that -
and
or
?
u = u. u
S in out
ia = Su
ug = cuin ?
= lout) 5k
(1 + s = uin s Rk
uout
From this we find the gain factor of the cathode follower, as
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
K=
uout
uin
Silk
1 Silk
(10-39)
Equation (10-39) shows that the gain of a cathode follower is always somewhat
less than unity in voltage. Here we note that, in the circuit of the cathode fol-
lower, the output voltage differs from that in the usual type of amplification in
a)
b)
c)
to video amplifier
indicator grid
uin
to video amplifier
indicator grid
Fig. l0-'24 - -Cathode Follower
a - With long cable; b - With short cable; c - Equiva-
lent network
that it is in phase with the imput 'voltage and "follows" any change in this voltage;
this is the source of the name "cathode follower".
Equation (10-39) may be converted to a clearer form if we consider that
S =Ri As a result of this substitution, we readily obtain the equation
nk
K =lL
B.
n
-P.
STAT
Bk
(1p-40)
- 119
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
The formula for a conventional amplifier is
na
(10-h1)
A comparison of eqs.(10-40) and (10-41) shows that the cathode follower
can be replaced by an equivalent standard amplifier with load resistance Rk and the
following total internal tube impedance:
or
(10-42)
As indicated by the equivalent circuit (Fig.10-21c), the resistance RI shunts
the cathode resistance R so that the equivalent output resistance of the cathode
follower at the points becomes
R!P
. 1 'k
(lo-42a)
Thus, the cathode follower presents a very low output impedance, not in any
case exceeding R! = 1 which may readily be matched to the line characteristic of
S'
-
the cable. However, the input f-osl-stsce e-f the cathode_folinwer may be very high
if it is operated with-very low grid currents. As the input and output voltages are
of the same order of magnitude in pathode follower circuits, a greater input re-
sistance signifies a relatively smaller input power, and a low output resistance
a relatively high ovtput power. Thus, the cathode follower functions as a power
amplifier, simultaneously serving to match low-resistance load with high load
resistance of the tube in the stage before the final stage.
Section 10-15. Automatic Frequency Trim
Should the frequency of the transformer or the local oscillator be changed,
120
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
the latter must be aligned so that the frequency difference fs ? fh between signal
and local oscillator will always equal the intermediate frequency fif of the re?
ceiver (30 mc). If the intermediate frequency changes due to transmitter or hetero?
dyne frequency drift, the receiver will be detuned in the intermediate frequency and
reception of echo signals will be interfered with. Therefore, automatic local?
oscillator frequency control (AFC) systems are widely used in radar receivers.
?
Within the band it is designed to align, the AFC system should provide very
rapid trim within the intermediate frequency even when sudden transmitter skipping
has occurred, and should compensate for all instabilities in the frequency both
of the transmitter and of the local oscillator itself. It should not be possible
for the receiver to tune to the image frequency (see Section 10-11). The image
frequency is weakened adequately ahead of the mixer input by resonance amplification
of the reflected signal frequency in the HF amplifier stages.
The simplest AFC block diagram is that shown in Fig.10-25. The AFC mixer, like
the receiver main channel mixer, yields an IF voltage amplified by one or two stages
of the IF amplifier, when the frequency voltages of the specially weakened trans?
mitter and local oscillator signals are applied. The amplifier output voltage acts
on the discriminator. The discriminator yields a voltage whose magnitude and sign
depend on the magnitude and sign of IF detuning. Guided by the control network,
it then acts on the local oscillator frequency in such a fashion that the IF re?
ceiver detuning is rapidly diminished. Some amplification in the intermediate
frequency is necessary if the output voltage of the automatic frequency control
mixer (particularly a crystal mixer), comprising 0.2 to 0.5 volt, is to be brought
up to the level of a few volts, required for the discriminator diodes to function
in linear detection.
The Discriminator
Discriminators are often designed in accordance with the diagram in Fig.10-26,
I
rinciple than that of an FR discriminator. The LiCi
121
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
1
?
and L2C2 circuits are tuned strictly to the intermediate frequency fif, selected
for the receiver (usually, 30 mc). The connection between the circuits is quite
loose, so that the resonance frequencies of the circuits are determined only by
Weakened out-
going signal
of transmitter
AFC mixer IF amplifier Discriminator
of AFC system
local oscillator
local frequency control
oscillator system
Fig.10-25 - Simplest Type of AFC Diagram
their own parameters. Both circuits have high attenuation. The resonance curves
of the circuits are?wide enough, and the voltage in the circuits is practically
constant regardless of chances in the intermediate frequency within the automatic
tuning band.
The dicriminator provides automatically the amount and sign of the automatic
trim voltage impressed on the control network by the local oscinator.
Inductive coupling in the coil. L2 causes induction in the first circuit of
some emf T of intermediate frequency. Regardless of the magnitude of this fre-
quency, the voltage E2 at the capacitor C2 lags behind the current 12, in the
second circuit by exactly 900. Because of the presence of adequate capacitances C
and'C? the cathodes of the diodes D1 and D2 are at zero potential in the inter-
4
mediate frequency.. The lower end, of the coil L1 in the first circuit, is also of
zero potential. Therefore, the/voltage El on this coil is connected across the
?
capacitor Co (whose resistance can be neglected) to the midpoint of the coil L2.
?
Therefore, at intermediate frequency, the diode D1 is under a voltage
122
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
3
STAT
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
I)
of = -
1 1 E2
. while the diode D2 is under a voltage of U1 = El - F2 .
2 2
The output voltage of the discriminator is the difference between the voltages
created by the diode currents at the resistances R3 and R4, as a result of the
detection of IF pulses:
Uout = U3 - U4
Each intermediate frequency is of normal magnitude (30 mc), while the current 12
coincides in phase with the emf 'E and the voltage El (see the vector diagram in
Fig.10-27b). The voltage "E2 at the capacitor C2 lags behind the voltage EI by
from IF amplifier
R-F choke
Fig.10-26 - Discriminator
Uout =U 3 - U
.exactly 90?. The voltages U1 and U2' impressed on the diodes are equal in magni-
tude. The rectified voltages TI3 and IT compensate each other completely, and the
_
output voltage of the discriminator thus becomes II'out = 0. Therefore, when the
magnitude of fif is normal, the discriminator does not affect the heterodyne fre-
quency-control circuit.
At intermediate frequencies above normal (30 mc), the second circuit shows an
inductive reactance. The current Y lags behind the emf E by a given angle pl
while the voltage 72 now lags behind the voltage El not by 90?, but by 90? + T
(see vector diagram in Fir.10-27a). Therefore, the absolute magnitude of the vol-
i
1' As a consequence U < U3.
$ h The output voltage
_
tage U2 is greater than that of U
123
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
4
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
of the discriminator is negative in sign.
' Conversely, at a lower-than7normal intermediate frequency the second circuit
will have a capacitive reactance. The current 12 ther leads the em! T by an
angle T, as a result of which the voltage 72 laps behind the voltage ;. by 90? - T
(see vector diagram in Fig.10-27c). Therefore, 172 is smaller than U1 in absolute
magnitude. Correspondinrly,L -< U3' so that the discriminator output voltage be-
comes positive in sign.
The dependence of the output voltage of the discriminator on the intermediate
a)
b) c)
Fir.10-27 - Discriminator Vector Diagram
a -fif > 30 mc; b - fir = 30 mc; c - fir 4 30 mc which is
the normalfif for a receiver
frequency at its input, called the discriminator response, is shown in Fig.10-28.
At an increase in variation from midfrequency in either direction, the dis-
criminator output voltage shows a linear increase at first; however, when detuning
is very pronounced due to the resonance properties of the circuits, it again drops
to zero. The discriminator response characteristic therefore displays two maximums,
the linear segment between which determines the operating band for automatic
tuning.
The nature of the discriminator output voltage depends on the relative magni-
tude of the time constants 3c3 and R4 C4' relative to the IF pulse repetition period.
12/L
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
When the time constants R3C3 and RitCis. are low, the discriminator output voltage is
pulselike, but when the time constants are greater it virtually retains its magnitude
out
?
in the intervals between pulses, i.e., it
becomes practically constant.
Control of Local Oscillator Frequency
As previously noted, the reflex: kly-
stron is used at the end of the decimeter
band and in the centimeter band.
As it is not possible here to make a
detailed check on the klystron theory in
teristic
general and that of the reflex klystron in
particular, let us merely cite the major physical processes occurring in the reflex
klystron.
Let us imagine a constricted cavity, subject to the action of an electron
stream passing through apertures (a grid) in the constricted portion (Fig.10-29).
Let us assume that free HF oscillations are excited in some form within the resona-
tor. During those fractions of a period when there is a positive charge on the
grid (1) and a negative on the grid (2), the electron stream penetrating into the
space between the grids is decelerated by the HF field of the resonator. The oscil-
lating energy of the cavity is supplemented by a reduction in the energy of the
decelerated electron stream. Conversely, during those portions of a period when
there is a negative charge on the grid (1) and a positive on the grid (2), the elec-
tron stream in the space between the grids undergoes acceleration. The increase in
the energy of the stream due to acceleration of the electrons results from a reduc-
Fig.10 -28 - Discriminator Charac-
,tion in the oscillating energy of the cavity. Thus, the decelerated electron stream
yields energy to the cavity, while the accelerated stream takes up energy from the
cavity.
If the density of the stream is constant, the oscillating energy of the cavity
STAT
125
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
will not be supplemented in the final analysis, since the energy of the electron
stream absorbed by the cavity in the deceleration half-period, is completely returned
to the electron stream in the acceleration
half-period. To have the electron stream
compensate the losses in oscillating ener-
gy in the oscillator walls, the density of
the electron stream must be greater during
deceleration than during acceleration. In
that case, the oscillations in the cavity
will not be damped.
Fig.10-29 - Passage of Electron Flow
Figure 10-30a gives the basic diagram
Through the Grid of a Constricted
of the reflex klystron. Let us assume that
Cavity, Within Which Oscillations of
free oscillations are excited in the cavity
Constant Velocity are Sustained,
by a shock, and that the energy of these
E)erepresent the decelerating field
oscillations has to be supplemented. The
of the cavity; 4- - represent the ac-
electrons passing across the alternating
celerating field thereof
field between the grids of the .excited
cavity in various fractions of the oscillation period and thus undergoing various
changes in velocity, enter the powerful retarding field of the reflector electrode.
The reflector electrode is beyond the cavity and is approximately 100 v negative
with respect to the cathode. The retarding field is generated by the potential dif-
ference between cavity and reflector, which is about 300 or 400 v. Electrons enter-
ing this field lose velocity, come to a standstill, and then return to the cavity.
Electrons accelerated in the field between the cavity grids penetrate closer to the
electrode and therefore return to the reflector simultaneously with the electrons
retarded in the field between the cavit-7 grids; they are thus stopped and returned
(1/ at a greater distance from the reflector. Consequently, the electrons returning to
the cavit7 in a reflex klystron form clusters or bunches.
126
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
The frequency generated may be varied, within certain limits, by changing the
voltage on the reflector electrode. In actuality, if the voltage on the reflector
a)
cathode
cavity
c) reflector electrode
coarse tuning cavity
reflector
electrode flexible
diaphragm
output
reflected
b)negative
voltage generated frequency
Fig.10-30 ? Reflex Klystron
a ? Basic diagram; b ? Ratio of frequency
coaxial output
line
cathode
oscillated
to voltage at reflector; c ? A typical design
is more negative, the bunches of electrons will return to the cavity even more
rapidly. On the other hand, if the voltage on the reflector electrode is less
negative, the electron hunches will return to tha-cL':.ity after a longer pc-T&s---
elapsed. Each bunch of returning electrons transmits an energy shock to the oscil?
lations in the cavity and compels a change in its frequency. The result is a pic?
ture similar to the change in the oscillation period of a physical lighthouse, due
to external shocks not in time with its swing. An increase in the amount of nega?
tive voltage on the reflector increases the generated frequency, while a reduction
in the magnitude of the negative voltage on the reflector, on the other hand, re?
duces the generated frequency. This results in an approximately linear relation?
ship between the generated frequency and the voltage on the reflegt_? (Fig.10-30b).
Figure 10-30c presents the design of the reflex klystron.
klystron local oscillator in an AFC system, discrimina-rAT
127
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
with high time constants are used, ard frequency control is generally carried out
by means of a direct-current amplifier reversing the sign of the discriminator out-
put voltage and raising this voltage to a level adequate for control of the klystron
reflector electrode.
1,1e control voltage tapped from the amplifier is added to the
direct negative voltage or. the reflector electrode from the feed.
The direct voltage is selected so as to provide the midfrequency selected for
the receiver at the normal frequency of the trarsmitter. If, for some reason, the
transmitter frequenc- has ircreased or the heterod:rne frequency has declined, causing
ar increase ir the intermediate frevercy, the positive discriminator output voltage,
amplified and reversed in sign (see '7ig.10-28) will raise the nerative voltage at
the reflector and increase the frequency of the local oscillator. Conversely, if
the transmitter frequency has been reduced or the local oscillator frequency has
increased, resrlting in a reductior in the intermediate frequency, the negative
output voltage of the discriminator, amplified and reversed in sign, will reduce the
411 negative voltage on the reflector electrode and thus reduce the frequency of the
local oscillator. In both cases, tie normal midfrequency will be restored rather
rapidly, ard at the end of tle compe,-sation, tIle discriminator output voltag,e,
rech-ced to zero, will r_Qtal.r.n_ to_the reflector _the_initial _no r_a:tive pot e
power feed.
If a triode local oscillator is involved, the control circuit includes reed-
back wit': a 900 phase shift, krown as reactive feedback, the principle of which was
discovered by a Soviet scientist, r1.7.7raude. Control of the local oscillator fre-
T-ercy is obtaired.by means of a reacta-ce tube, whose plate-cathode circuit is
coupled parallel to the local-oscillator circuit, and serves as an aligning react-
ance resista-ce.
Figilre 10-31 sows a control system alonr these lines. Here L2 is the local
oscillator trbe, and L1 is the control reactance tube. The capacitors C2 and C3
protect the ti he grid circuits from direct voltage, while the choke L protects the
12P.
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
*feed source from high frequency. Therefore, in examining the operating principle of
the circuit, these I4F components will be disregarded. The oscillating potential
from the local oscillator circuit LhCh is impressed directly on the plate of the
tube Ll, and on the grid through the phase?shifting circuit R000. When Ro >> 1
w,C
o
the currert I, branching into this circuit, may be taken to.be in phase with the
local oscillator voltage Ta at the plate while the voltage U.rf at the grid, taken
off capacitor Do, lags hehind the current To and the plate voltage by 900. In the
to the
discriminator
Fir.10-31 ? Schematic Diagram for Control of the Frequency
of the Local Oscillator b7 Means of Reactance Tubes
---? ? ---_-_-??_-?
pentodes, the plate voltage is controlled practically by the grid voltage alone,
and is much less dependent on the plate voltage. Therefore, the plate current of
the tube. LI virtually coincides in phase with the grid voltage and lags behind the
plate voltage by- 900
.
In view of these phase ratios, the resistance in the plate?cathode gap gas a
reactive and, in the riven case, an inductive character, i.e.,
The plate voltage is
q
Ii
a
2laf = = itdgLaf (10443)
Ia
determined by the product of the grid voltage TT and the
gf
reactance tube. In turn, the grid voltage is the voltage
129
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
drop from the current 10 Ua due to resistance 1 . Taking this into con-
no jwgCo
sideration, eq.(10-43) readily yields the equivalent trimming inductance of the
plate-cathode gap of the reactance tube:
R0C0
Laf =
In automatic trimming, this inductance is regulated by changing the transcon-
ductance 3 of the reactance tube, under the influence of the discriminator output
voltage. When the intermediate frequency is above the midfrequency, the negative
potential or the discriminator shifts the quiescent point of the reactance tube into
a region of lower transconductance. The trimming inductance rises, and the total
inductance of the local oscillator circuit declines, while the frequency of the
local oscillator increases. When the intermediate frequency is below the midfre-
quenc:,-, the local-oscillator frequency drops. In both cases, the midfrequency is
restored.
Section 10-1(). Systems of Protection from Interference
rot only echo signals from the target but all kinds of interference signals,
pulsed and otherwise, reach the receiver input. Interference signals sometimes
?
dangerously overload the receiver stages and make -it.--mpossible- to- the tan7____
get s.-rals on the indicator screen. Therefore, protection from interference is one
of the most important circuiting problems in designing radar receivers.
1. 7oises overloadirE the IF amplifier are composed of long signals of high
power: reflections from local objects, clouds or, for example, powerful undamped
oscillations from ar interferiri= source. Protection from noises that overload
17 amplifiers is obtained by instantaneous automatic volume control systems.
Figure 10-32 depicts ore such system. ;then the K key is placed in the posi-
tion (1), the IF amplifier (tube Li) functions without this system. For maximum
amplification of the signal from the target, without interference, the grid bias
130
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
- of the tube L1 is designed so that theinitial quiescent point will fall into the
-:area of maximum transconductance. However, if the t',4get signal arrives simultan?
eously- with powerful interference, then, superimposing itself on the noise, it will
be in the interval of the upper bend of the characteristic curve of tube LI., where
transconductance is zero, and will thus be cut off. In this manner, a power inter?
ference, by overloading the IF amplifier, may cause the target signal to disappear
at the IF amplifier output.
An entirely different picture results if, in the presence of interference, the
amplification by an IF amplifier stage is controlled by the instantaneous automatic
volume control system. This is arranged by placing the key K in the position (2).
Li to next IF.amplifier
stage
Cl
from preceding IF
amplifier stage
C2
choke
Ri
R2
L2
Fig.10-32 ? An Instantaneous Automatic Volume Control System
The output signal of.the IF amplifier stage is delivered from the plate of the
tube Li to the grid of the tube L2, operated on plate detection in the lower bend
of the characteristic curve. The initial grid bias required by the tube L2 for
this type of operation is provided by the divider RL, R5, and R6; the grounding of
the resistance R2 with a cathode potential negative to ,ground is equivalent to
the presence of adequate positive voltage at the plate, relative to the cathode.
The grid bias of the tube Li, controlling the Eain of the IF amplifier, is tapped
STAT
131
'1.
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
from the resistance Re. This is created the plate current of the tube L2 and is
zero wher that tube is off.
The ircrease ir control bias lags somelhat relative to the instant at which
the output si-nal of the 17 amplifier appears. This lap is due to the intrinsic
time (71? R2)C, vf-ich is of the same order as the transmitter pulse length.
Therefore, tie pure tarr-et sirnal is inadequate for creatirr a large control bias,
arc), its amplification is rot sirnificartly reduced. However, a powerful noise of
loLrldurtior, amplified by the tube L1, rapid1y. increases the control bias approx-
inatelr to the le-rel of its own amplitude, and again shifts the quiescent point of
the tube L1 into the irterval of trarsconductarce. The tarret signal, 'super-
Liposod or t'-e roise after t'e instant 1.4.-:en it was set by the control bias, ro
lorrer falls into tie rerior of the upper bend of the curve, but is normally ampli-
video frequercy
detector amplifier
754-.10-33 - '3ireuit or Thneratior of a 3hort-Time Constant
fied, vf 41e t'e :oiso is si,ppressed to a considerable derree, ti-us climinatinr over-
loadi.r of the 17 amplifier.
2. -r ingtartlreous automatic: volure cortrol system, eliminatinr 17 amplifier
o7-rlo_ld does rot protect the video arTlifier. The result of detectiol of noises
-rori t'e I7 anpli:ier.is to prodycc a high rerative voltare or the load
resistroo of the letector. "his ma- either turr off the first video amplifier
rr- 4.irl - or
t off
tar ,et sirral superimposed o- the noise. The result,
132
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
0
here too, is that arrival of the target signal at the receiver output again becomes
impossible.
To prevent the video amplifier from being cut off by powerful interference, a
low-time constant circuit (RC, Fig.10-33) is introduced.
C
When there is no prolonged interference, the key K is placed in the position(1),
and the detector output is directly coupled to the video amplifier. Should inter-
ference appear, the key K is placed in the position (2), and the detector output is
corrected to the video amplifier via the law time-constant circuit, acting as a
differential circuit. When the time constant R CE is somewhat smaller than the
E
transmitter pulse length, the charging and discharging of the capacitor Cr, of
brief duration, take place accordingly at the instants when the broad negative
pulses of the detected noises start and end. The charging current at the beginning
of interference creates a brief negative potential surge at the R resistance,
while the discharge current at the end of interference creates a brief positive
voltage surge, which is impressed on the gi'id of the first stage of the video amp-
lifier. In the interval between these processes, thanks to the absence of current
in the resistance R? the noises do not create cut-off bias at the video amplifier,
and the target signal superimposed on the noise at the detector output traverses the
video amplifier normally if it appears in the interval between surges.
3. A shortcoming of the low constant-time circuit lies in the fact that: al-
though it assists the video amplifier to sustair the target signal in the presence
of detected interference of long duration, it creates additional voltage surges
which may confuse the operator who might interpret them as additional targets. To
eliminate this drawback, systems Cor selecting signals by length are used. These
systems pass only signals whose length is equal or nearly equal to that of the
transmitter pulse. All other signals fail to pass. When the target signal is
superimposed or a prolonged interference noise, only the target signal is passed,
the noise being suppressed. A shortcoming of circuits Cor selection of signals
133
neclassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
by length is the suppression of grouped targets longer than the signal of a single
target.
4. The power of echos from local objects and clouds near the target declines
sharply with increase in range. Itis therefore desirable that, in addition to
connection of the instantaneous automatic volume control, there be an increase in
receiver amplification with range, i.e., in proportion to lap time, so as to facili?
tate reception of weak signals from distant targets. A sharp reduction in gain for
Cl
initiation
of pulse
+300 V
L
Fig.10-34 ? Time?Gain Cnntrol Circuit
Gain control
voltage
_150V
short ranges reduces the overload on the receiver by powerful signals near the
target and assists in suppressing powerful echos from nearby local objects and
clouds.
Automatic increase in rain with increasing lag time is provided by timed
volume?control networks.
Figure 10-34 presents a simplified version of such a circuit. The circuit
parameters are so selected that the negative plate potential of the tube LI, taken
off the divider R6' REP is significantly smaller than the negative potential of its
cathode taken off the divider R3' R' .R5' This is equivalent to the presence of a
4
positive voltage between plate and cathode of the tube Li. In its initial stage,
do.?
134
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
(.
the tube Li is shut off by the high negative grid bias, while the tube L2 remains
in operation, since its cathode has a greater negative potential than the grid.
A brief positive pulse is impressed on the grid of the tube Li, simultaneously
with the starting of the transmitter, gating it for the full duration. The plate
current pulse thus set up in the tube Lip passing through the low-capacitance capac-
itor C3' charges it to a negative voltage whose magnitude is regulated by the po-
tentiometer Rh in such a fashion as not to cause the tube L2 to be cut off.
At the instant when the starting pulse ends, tube Li is again ungated and the
capacitor C3 begins to discharge across the resistances R7 and R8. The tube L2
functions as a cathode follower, and the output potential on its cathode follows
the negative voltage on the grid from the discharging capacitor. This is added to
the negative bias taken off the potentiometer Rto control the volume. As the
capacitor C3 discharges after the triggering pulse comes to an end, the automatic
negative bias on the IF amplifier declines along with the output voltage of the
cathode follower, causing the receiver volume to rise with increasing range.
135
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
C7APTER XI
oscruiAT01/3- FOR RADAR TRAT'S'ITTERS
Section 11-1. Tactical and Enrineerinr Data on the Radar Transmitter and
Specifications to be :,et
The object of the radar transmitter is to renerate periodic U77 current pulses
of specific power, Ienrth, and shape.
flenerall7 speakinE, a radar transmitter includes three major components:
oscillator, modulator, and' power suppl:r.
The oscillator portion of the transmitter consists of a UP? current oscillator,
feeding the transmittinE artenna. Dependinr on its tactical purpose and the tzrpe
of radar, oscillators operate in the frequency ranges from 100 mc (3 m)
to 10,000 mc (3 cm), and as a rule on fixed wavelengths.
The modulator portion of the transmitter controls its pulse operation. Pulse
,
modulation is a specific characteristic of radar transmitters, distinEuishinE
them from radio wavelenth transmitters.
The supply source provides the transmitter with rectified voltage and feeds
the filament circuit.
Due to the pulsed mode of operation, the major tactical and engineering speci?
fications for radar transmitters areas follows:
a) Frequency f of the oscillations to be produced (operating wavelenrth);
b) Length of the R?F' pulse T' and its shape;
c) Train rate F ;
D
136
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
d) R-F pulse energy W;
e) Transmitter pulse power Pp and its average power P a1,during
period.
The major specifications to be met by radar transmitters can be formulated as
follows;
1. Provision of the required pulse power. Depending on the purpose of the
station, the transmitter pulse power can range from a few watts to several megawatts.
HiEh pulse power combined with relatively low average power is a most important
special feature of radar transmitters, while the inverse of the duty cycle (of the
order of 1000) makes the transmitter and the radar as a whole possible, in terms of
power.
2. The electric circuit of the oscillator must be as simple as
requirement is dictated prinarily by the difficulty and limitations
tion of high power on ultrashort wavelengths, resulting in very low
.Therefore, radar transmitter oscillators are usually single-cascade
the repetition
possible. This
of amplifica-
efficiency.
networks. reed-
less to say, a powerful self-exciting oscillator does not ensure the high stability
of frequercy provided by low-power radio broadcasting or television transmitters.
The reduced frequency stability of radar transmitters requires provision of auto-
matic tuning trim and of a pass bandwith an adequate safety factor, as discussed
above in Section 10-6.
3. The oscillator frequency must be subject to smooth change, in tuning to a
fixed operating wavelength, to provide the most desirable operating conditions for
the station.
4. The R-F pulse produced must be stable in shape, amplitude, and length, and
the pulse frequency must also be stable.
5. The transmitter must have a reliable control system, blocking and signaling, ?
guaranteeing protection of operating personnel from injury by high voltage, and also
ensuring simplicity of control by the transmitter and protection of its most
137
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
important details from overloads and early wear.
6. The wiring must be of adequate mechanical strength and well protected from
shock, so as to guarantee vibration Stability in operation and transportation.
7. The transmitter must be light and of small dimensions, so as to render the
set mobile.
Section 11-2. Basic Diagram of Radar Transmitter
In order for a powerful R-F pulse to be generated, a pulse of corresponding
power must be delivered from the power supply. However, it would be exceedingly
unwise to use a source with a power equal to the input pulse power. Such a source
would be extremely bulky and could be used only to a very limited percentage of its
capacity.
Therefore, an element for storing the energy of a low-power source is an essen-
tial component of the modulator portion of a radar transmitter. A capacitance or
inductance may be used as storage element. Even it the stored energy used in the,
oscillator within a microsecond is negligible and does not exceed 0.6 joule, the
pulse power delivered will be P = 0.6 = 600 kw; if the efficiency
? is 0.3,
P 1 10-6 T1
the oscillating pulse obtained will be P =Po p = 600 0.3 = 180 kw.
p
Figure 11-1 shows the basic diagram of a radar receiver.
A switch automatically couples the storage element to the power supply during
the interval between pulses and to the oscillator during the time of pulse genera?
tion. The switching rate is equal to the pulse train rate.
Various methods are used for rendering the switching automatic. In the general
case, the switch is controlled by the master oscillator off:, the set, which is not a
direct component of the transmitter circuit. Although the control oscillator deter-
mines the transmitter train rate, it does not determine the shape and length of the
0 pulse. The length and shape of the radio pulse generated depend on the rate and
nature of the accumulated oscillator energy and are determined, consequently, by
138
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
(,
*the parameters of the storage element and the UHF oscillAtor rather than by the
parameters of the control oscillator.
In certain types of stations, the control oscillator is an integral component
of the transmitter. In such a case, the modulator consists of three parts: the
storage element, the switch, and the control oscillator, and is used simultaneously
for coordinating the functioning of all the other assemblies of the radar. The
pulse modulator is often termed a keyer, as its function is that of periodic coupling
and uncoupling of the oscillator.
storage
element
power source switch UHF
oscillator
control
oscillator
Fig.11-1 - Block Diarram of Radar Transmitter
to transmitting
antenna
The control oscillator proper, which is an integral component of the trans-
mitter network, can be based on any-type of oscillation or shaping of brief trigger
pulses with steep leading edges(by Means of a blocking oscillator, multivibrator
with differentiator, limiter and amplifier, thyratron networks, etc.).
The block diagram in Fig.11-1 is generalized and is not intended to correspond
to the actual circuits of the various types of radar transmitters. SometiLes, for
example, it is possible to dispense with the control oscillator by coupling the
storage element to the generating oscillator by means of a rotating discharger
(Section 12-11), whose rotor is keyed to the shaft of a motor of a particular rpm
STAT
1
139
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
of adequate stability. Finally, in certain low-power transmitters it is possible
to do without the storage unit and even without the keyer at all, the oscillator
tube being gated and unrated automatically in the grid circuit of the UHF oscillator
itself, with parameters specifically selected for the purpose.
On the other hand, certain types of radar transmitters have block diagrams
considerably more complex than those illustrated in Fig.11-1.
Section 11-3. Conditions Governing Self-Excitation or Tube Oscillator
As known from the general course on transmitters, the conditions for the self-
excitation of a vacuum-tube oscillator can be reduced essentially to two:
1) The R-F voltages at the oscillator tube plate and grid must be opposite in
phase. To satisfy this condition, there must be a phase balance in the components
of the oscillating system of the generating oscillator;
2) The excitation voltage delivered from the plate to the grid circuit by
feedback must be adequate for providing variations in plate current of a magnitude
sufficient to compensate the losses in the oscillating system, when oscillations
of the given frequency are generated. In order for this condition to be satisfied,
the amplitudes of the alternating voltages at the plate and the grid must be bal-
anced.
Phase Balance
The circuit of an: iracizum-tube oscillator can be reduced to an equivalent
tapped oscillator with self-excitation (Fig.11-2).
In accordance with the rule for tapped network designs, familiar from radio
engineering, the reactance Xrf between grid and cathode and the reactance Xaf be-
tween plate and cathode, must be identical in nature (both inductive or both capaci-
tive). It is only on this condition that the current I, successively flowing around
the equivalent reactance elements of the oscillating systeml'can keep the R-F vol-
tages at plate and 3.rid opposite in phase.
140
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-010431i001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Since all elements of the oscillating system are incapable of having reactions *
of identical nature to the oscillator frequency, the reactance Xag between plate
and grid must be opposite in sign to the equivalent resistance X between grid and
Ef
cathode. The total reactance of the left
branch of the oscillating system has to be equal
and opposite to that of the right brahch. There-
fore, the abolsute magnitude of the reactance
between plate and grid must be greater than the
absolute magnitude of the reactance between
grid and cathode.
Fig.11-2 - Equivalent Tapped
"etwork of Oscillator with
Self-Excitation
Amplitude 9alance
The amplitude of the first harmonic of the
plate oscillating voltage is
Uma = Ima _7,e
where 'ma is the amplitude of the first harmonic of the oscillator plate current;
e is the equivalent resistance of the oscillating system of the generator,
at the oscillated frequency.
The value of Uma is dependent not only on Ze but also on the excitation vol-
tage Umg0 delivered to the grid by feedback from the oscillator plate circuit. This
relationship, expressed by the function Uma = F 1(Umee), is called the oscillating
characteristic (curve 1 in Fig.11-3).
The oscillating characteristic permits determining the plate oscillating vol-
tage, if the voltage excitation amplitude at the tube grid is known. The ascending
branch of the oscillating characteristic corresponds to operation of the oscillator
4 at less than peak voltage, when the grid current does not distort the shape of the
plate-current pulse and its first harmonic increases with increasing U . The
mg
STAT
141
- Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
_
descending branch corresponds to operation at excess voltage, when the appearance
of considerable grid current produces a plate?current pulse dip, with the result
that its first harmonic and the plate oscillating voltage decrease with
increasing U . It should be mentioned that, the greater the Zes
mg
the oscillation characteristic of the oscillator.
feedback curve
operation at operation at excess
incomplete potential potential
the higher will be
Oscillation
characteristic
7ig.11-3 ? Oscillation Responses and Feedback Curves.
Regions of operation at incomplete and excess potentials
illustrated for the oscillation response.1
On the other hand, the excitation voltage U, delivered to the grid iv nega?
tive feedback, depends on the feedback factor and the oscillating voltage on the
plate. Therefore, a very
specific relationship exists between the oscillating vol?
tage at the plate and the excitation voltage describirF feedback. The ratio
describing the function Una = 72(Umr' k) is called the feedback curve.
In accordance with defirition, the feedback factor is
Umg
k =
from which it follows that
Uma
1
U=
ma k mg
142
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
2
?
?
?
When k is constant the feedback curlie is a straight line (line 2 in Fig.11,3),
and its angle to the Umg axis determines tan a
1
The.stronger_the_coupling,
- i.e., the greater the value of k, the smaller will be the angle of slope of the
-,feedback curve.
The feedback curve permits determinifig the amplitude of the excitation voltage
--at a given k by means of the known oscillating voltage at the plate U. The balance
- between the amplitudes Uma and Umg determines the point A, the intersection between
- the feedback curve, and the oscillation curve. Actually, it is only this point thati
- can determine the stability of the oscillations. If the plate oscillating voltage :
- corresponds to the point A, the excitation voltage impressed on the grid by the
- feedback is of the same amplitude U at which the amplitude of the oscillating
voltage at the plate corresponds to the point A. In pther words, the coordinates
of the point A represent a combined graphic solution af the equation for oscillation
response and for feedback curve.
Thus, the conditions of oscillation of a vacuum-tube oscillator, having a
given oscillation response, are determined by the feedback factor.
The feedback factor is not at all inmaterial, as far as provision of oscillator
self-excitation is concerned. For example, given a value of k1 corresponding to
the feedback curve 3, oscillations cannot arise in an oscillator whose oscillation
response is shown as curve 1. The fact is that, regardless of the excitation vol-
tage, the oscillation voltage at the plate in the given case excited in accordance
with the oscillation curve (point B) will be smaller than the magnitude lima
(point B1) required according to the feedback curve for delivery of the initial
amplitude Umg to the grid. Therefore, the excitation voltage drops to a magnitude
corresponding to the point Vt, and this in turn reduces Uma to correspondence with
the point V. This induces a further drop in Umg and results in a rapid consecutive
decrease in oscillations to zero. In order to excite the oscillator, it is neces-
sary either to increase the feedback so as to cause its curve to slope less and
143
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
to intersect the oscillation characteristic or, without var7irE the feedback, to
increase Z. to a derree that compels the oscillatinE; response (curve 4) to pass
a'-ove the feedback curve until the poirt of intersection is reached.
411 3ectio- 1141. The Mectron Tube as U7P Oscillator
?
operation, M:e electron tizbe loses rot orl- its amplif7inv but also its
oscillating properties.
In the first place, t: e redrctio- in input impedance due to the inductive ef-
!'ect eIectro. strea- (see 3ectior 10-5) results ir ar increase ir the power
required to be delivered 1J-- feedback from the plate to the -rid circuit.
In t:.e secord place, t'-e ef"ect of the electron tra-sit time is to cause the
e plate curfe' t to la. frid voltare ar angle Ti. In the plate
circuit, a compensafin7 lead -o,st he provided so as to carse the alternatirv
vol-
.a-es at t'.0 plate a-d 'rid to i.e in opposite phase (see point 1 of the conditions
Cor self-e:xitatio - ) It is ther necessar - to operate with starrered loads, and
e a' gq1ar ar:r-tent T e o the plate-load impeda cc rrst arree Ldth the
"quatio- (p T In sta-rered load oporatior, t..e oscillated power separated
in t -e plate circrit ttrl1 - .lecreases.
-
1
P = - U I cos Te
-o ma ma
e -trrive ')eco-es s or Ler, t:-e i---Lr ar-rments T ane increase. a
rreq-e- c w qols rot e;:co-r1 1000 mc for powerfrl oscillator triodes, To
.0co--..e3 700.
t is poi- t, e oscillating power ir the plate circrit becomes
eq' o zero, a. d is : o capa'de of fr nctiorinr as an oscillator.
T.- opera' - - wit' plate c' me t cut() C", 1portio. oC the electrons do rot
s-lcc-ei rec i ' e rri I riP,ri t' o period o" t' o posiJdv^ half-wave ir
ri volta-e, a- retnr, o Irri, 4,ho re-ativo -wa ,
. Lr1:1-:r is if ic L: tl: ro-hoo:-; e cat odo li'c.
0 ir' to , -lc .1
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
rrid voltare, it yields energy to the grid circuit. On the other hand, when the
electron is accelerated by positive plate voltage, on moving from grid to plate, it
takes up energy from the grid circuit. The increase in electron velocity raises the
dissipated power at the plate. Thus, the power required by the grid circuit is
separated at the plate in the form of supplementary losses involved in heating it.
The oscillating system of a I= generator is significantly affected by the
Fir.11-4 - Triode with Short-Circuited Plate and Grid, Used
as Oscilaatinr System--
interelectrode capacitances of the tube and the inductance of its leads, determining
the maximum oscillation frequency: the critical (wave) frequency. The critical
frequency coincides with the resonant frequency of the tube, corresponding to
short-circuiting of plate and grid.
It follows from Fir.11-4 that the critical frequency is .
vfrere
wcrit
= w
1
? V(La Lg) Ce
Ce = C C Caf
g
ag ' cgf + Caf
Tn desirrin!-, a TT oscillator tithe it is necessary to provide minimum inter-
electrode capacitance (to raise the critical frequency) and minimal electron
145
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
_
transit time (to increase the input impedance and raise the oscillator output power).
In UHF continuous oscillation tubes, particularly those of high power, reduc-
tion in interelectrode capacitances cannot be obtained by reducing electrode surface
?
areas, because of the need of dissipating high thermal losses. It is, therefore,
necessary to proceed differently, namely to increase the distance between electrodes.
In this connection, reduction in electron transit time to a minimum, when the inter-
electrode distances are large, creates an increase in plate voltage to tens of
kilovolts, which in turn constitutes a reason why the tube vacuums must be harder.
Virtually all oscillator tubes, used in the radar band are triodes, in order
to eliminate the harmful effect of inductance when screen grids and capacitances
.C(r)r, are introduced between the screen and control grids, serving to increase
the total capacitance of the oscillatinp s:rstem and to reduce the critical fre-
quency.
Section 11-5. Special Design Features of Radar Transmitter Oscillators
A resolution of the problem of generating the shortest possible waves.inevitably
leads to the difficulties encountered in obtaining power sufficient for radar pur-
poses. The critical frequency of an oscillator tube cannot be used in radar trans-
mitters. The need for coupling to the antenna, for regulation of the oscillated
frequency and determination of the most favorable opei-ating conditions, makes it
necessary to introduce the external elements of oscillating systems into electrode
circuits. These elements take the-form either of lumped inductances and capaci-
tances, or of (artificial) segments of long lines, properly selected, or finally of
coaxial-cylindrical cavities. Favery additional reactance element added serves in
all cases to reduce the oscillated frequency.
Oscillators with external artificial lines make it possible to obtain adequate
power at frequencies up to A00 mc, approximately. The use of lines increases the
dimensions of the oscillator.
At frequencies below 150 mc, the lines required he-
that lumped inductances consisting of one or two coils
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
"
are used instead.
The lighthouse type of tube (see Section 11-9) is used to realize the idea of
obtaining an oscillator of small size by combining oscillating circuits, tuning
elements, and the oscillator tube itself into a single unit (see Section 11-9).
Tubes of this type permit operation at frequencies of as high as 3000 mc (10 cm),
but the oscillating power they develop provides a pulse of not over one kilowatt.
Consequently, lighthouse tubes do not meet the requirements of powerful radar trans-
mitters. For operating at wavelengths of the order of 0.5 m, tubes have been de-
signed (see Fig.11-12) containing oscillating lines and four pulse triodes in
joint operation, providing a pulse power of up to 200 kw.
Oscillation of frequencies in excess of 1000 mc in powerful transmitters is
performed exclusively by magnetrons.
Let us next discuss the problem of frequency stabilization.
The fact that the powerful oscillator circuits used in radar are single-
cascade designs, prohibits usual methods of frequency stabilization by means of
tourmaline or quartz plates in the master oscillator circuits and subsequent multiple
frequency Fain, as is dore in television transmitters for example.
We know from radio engineering that the stability of an oscillator frequency is
higher, the better the quality of the oscillating system. Where multicascade trans-
mitters are concerned, this permits attainment of adequately high frequency stabili-
zation, without the use of quartz crystals. However, in the case of radar, when a
powerful self-oscillator operates directly on the antenna, the use of a high-quality
oscillating system that will provide good stability is out of the question.
The low-power exciter of a multistage transmitter is intentionally unloaded
so that the so-called buffer stage coupled behind it can operate without grid cur-
rent. In such a ca'se, although the efficiency of the self-oscillator is very low,
this will naturally not affect the power balance of the transmitter.
The quality of the oscillating system of such an unloaded self-oscillator is
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
'determined only by the' resistance of its own losses, and may be made high enough t-
provide sufficient stability.
The efficiency of a powerful radar self-oscillator, operatin:-, on an antenna, is
410 determined to a considerable decree by the efficiency of its oscillating circuit:
r.
in
?
nk
r r.
in
where r is tl-e resistance of the losses of tle circuit itself;
? is the resistance introduced irto the circuit from the antenna.
ran
In order to reach radar transmitter efficiencies efforts are being made
to reduce the resistance of the losses in the circuit proper, i.e., to increase the
quality of the unloaded circuit.
The quality of a transmitter circuit, normally loaded onto an antenna, is de-
termined f13ndamentall7 by the resistance introduced, meaning it will be low. In
other words, the high inherent quality of the oscillating system of a radar trans-
ritter does not determine the stability of the frequency, an increase in which re-
quires other measures, primarily that the transmitter be supplied with stable vol-
tages.
On the other hand, ti-e presence of a system of automatic alignment and of a
broad reception channel, rerders the question of frequency stability less pressing.
The major porblems,of radar transmitter oscillator desiEr resolve, then, to
the rollowin-:
1) Ens-irance of operation at the specified wavelength with adequate pulse
power;
2) 7n3urance of maximum quality of the oscillating circuit, so as to yield
the best possible efficienc:T;
3) Possibility of selecting the most desirable oscillation pattern, pro-
an oscillated pulse of satisfactory shape, high efficiency,
quired power, and adequate stability;
re-
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy A proved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
4) Possibility of smooth adjustment bf the oscillated frequency within the
given tolerance and without significant changes in the oscillPtor oper-
ating conditions.
Section 11-6. Practical Circuits for Tube Oscillators in the Meter Wavelength
In the majority of cases, vacuum-tube oscillators in the meter wavelength,
operate on symmetrical output. Therefore, they are designed as a rule as single-
stage push-pull networks. The discussion here will be limited to an examination
of networks with artificial lines, since networks with lumped inductances were
characteristic of earlier radar equipment operating on longer wave lengths and
presently not of interest. In the second place, they may be equated with the same
equivalent circuits of push-pull oscillators, to be examined below, as may artificial-
line networks.
Oscillator with Artificial Lines in Plate, Grid, and Cathode Circuits
The network Of such an oscillator (Fig.11-5a) can be converted into a general-
ized equivalent circuit, Where the elimination of specific elements makes it possible
to obtain equivalent circuits of all subsequent vacuum tube oscillators in the
meter wave band.
A segment of an end-shorted two-wire line between the tube plates has a length
of somewhat less than-i-- and is therefore equivalent to the coupling of an induc-
tance between the plates. The magnitude of this inductance can be adjusted smoothly
with the aid of a 'shorting bar (1). The same is true of the line segments between
the grids and between the plates.
The input inductances of the_artificial lines, taken together with the induc-
tances of the tube leads, form the equivalent inductances 2La, 2L and 2L of the
E ?
oscillator system, coupled, respectively, between the plates, grids, and cathodes.
In addition, the interelectrode capacitances Car, Cal., and C , enter into the
gi
oscillating system.
149
in Part - Sanitized CODV Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
STA1
?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
If the interelectrode capacitances are considered as coupled into the external
circuits of the electrodes, the equivalent circuit in Fiv.11-5b will be obtained,
plotted for the first harmonic of the plate current.
b)
c)
Fig.11-5 - Oscillator with Artificial Lines in Plate,
rrid, and Cathode Circuits
a ?Oscillator schematic; b - Equivalent circuit of
oscillator; c - Equivalent circuit of ore arm
The tube plates are coupled to the ends of the equivalent inductance 2La, with
equal and opposite a-7 potentials. The same holds true for tie tube 1-:rids and
cathodes. Corsequent17, the midpoints oftlie irterelectrode inductances in the push-
pull system are automatically maintained at zero potential in the R-F band.
To simplif- t,--.6-analysis-it is *cc0.7venient to subdivide the equivalent circuit
of a push-pull oscillator irto two completely identical arms. The equivalent.cir-
cuit of the arm sl,owr in 'ii.11-5c.makes it clear that the inductarces La, Lr
ard 1,4", co,Ipled respectivel- to plate, rrid, and cathode at the point 0, are in
star connection.
The t:leory and araly-sis of MF tube venerators has 'leer developed in the works
of a run'Der of Soviet scientists, primarily by C.A.Zaytlerk and S.:..Drobov.
150
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
The startinr point of this theory is conversion of the inductance star into an
equivalent inductance triangle (Fip.11-6). In connection with the method of con-
verting a star into a triangular connection, familiar from electrical enrineerinp,
one side of the equivalent inductance triangle is equal to a fraction, where the
numerator expresses the sum of the paired products of the irductarces of different
rays of the star, and the denominator expresses the inductance of a ray of the star
opposite to the side off the equivalent trianple in which we are interested. This
rule yields a system of three formulas:
L=
ar
Lalsg + La Lf+ LgLf
Lf
L = LaLg LaLf + LgLf
L -
E
La
LaLg-+ LaLf LgLf
La
(11-2)
Substitution of the star by an equivalent triangle makes it possible to convert
the equivalent circuit of the arm to the form shown in Fig.11-7, which is clearer
and more cdrivenient for purposes of analysis and design. Three important conclusions
can be derived from such a circuit.
1. The oscillating system of the generating oscillator consists of three paral-
lel circuits connected ir series, and has several resonant frequencies. Therefore,
oscillators of this type tend to produce several frequencies simultaneously, causing
0
jumps from one operatirr frequency to another, which often interferes with the
smoothness in variation of the oscillated frequency when the oscillator is retuned.
At the oscillated frequency, the natural frequencies of all three circuits of the
oscillating system of the oscillator aresignificant in effect, and depend on all
the reactance parameters of the circhit.
2. Taken separately, none of the three oscillating system circuits are capable
of being tuned to the frequency being, generated, since otherwise the equivalent
rprzi -rani, 1 A 1,
active, while the oscillating system of an oscillator
- 151
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
requires. that the circuits he equivalent to tile correspordinr reactances.
3. The feedback between the rrid and plate circuits of meter-wave oscillators
7ir.11-(1 - Equivalent Star (a) and Triangle b) or Inductances
in :sm. Circuit
a)
?
) .d)
to antenna
?
to artenna
? pickup loop
c) e)
,to antenna to antenna
ag.11-7 - Snmilements to retwork of Oscillator with Artificial
Lines in Plate, rrid, ard Cathode Circuits
a - Oscillatirr system of oscillator arm, as series connection
of three oscillatirgocircuits; b - Tapped network with in-
ductive f'eadback; c - Tapped network with capacitive feed-
back; d - Conductive coliplinp to antenna; c Inductive
coupling: to antenna
is complex ir nature. This qu'artity is determined by the relationship of the
152
STAT
-
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
equivalent resistances or the circuits IcfCri, and LarCaf. The equivalent induc-
tances L and Laf$ enterinp into these circuits correspordinp to the system (11-2)
depend on the ratio of inductances between electrodes of the same sipn. The capac-
itances Cf
and Car, strictly speaking, are determined by the corresponding inter-
electrode tube capacitances and by the capacitances of the external circuit wiring,
connected to the electrodes. Adjustment of feedback. may have a significant effect
on the operating frequency of the_rererator.
The corditions required for self-excitation are by no means met by any and all
combinations of equivalent circuit reactances. Oscillation is possible only in a
relationship of oscillating system parameters that permits operation either by means
of an elemertar- tapped circuit with inductive feedback (Fir.11-7b) or by means of
an elementary tapped circuit with capacitive feedback (Fig.11-7c).
Thus, regardless of the unusual design of the push-pull meter-band oscillator
as compared to pus:.-pull oscillators of long and short waves, the equivalent net-
work of an oscillator ma-r, in the long run, be reduced to the usual tapped system of
self-excitation.
As the cathodes of oscillator tubes are DC grounded, coupling to the load
(the antenna) is effected, for reasons or safety for the operating personnel, using
a two-wire feeder connected to the contacts IT, which may be moved along the cathode
line (Fir.11-7d). This type of connection is equivalent to conductive coupling, the
-
connection with tile load increasnng as the sliding contacts approach the cathodes.
Low-power transmitters sometimes use magnetic coupling to the antenna feeder,
h:f means of a loop placed in the field between the plate or cathode lines (Fig.11-7e)
The tube filament circuits of push-pull oscillators (Fig.11-a) also display
certain peculiarities. The negligible available capacitive reactance of the capaci-
tor Cbt protects the filament of the corresponding tube from further heating by UHF
current. Similar protection of the-filament transformer is obtained R-F blocking
chokes L?, made for use in the meter band from spirals of silvered bare copper
153
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
wire.
An examination of other meter-band network circuits shows that it is possible
to do without special blockinr chokes in the filament circuit by using the shieldin-
properties oC the cathode line within whose tube filament wires are placed. As far
as the leak resistance R is concerned, it is usually not shunted by capacitors in
radar transmitter oscillators, sirce the capAcitance of the Frid circuit wiring is
adequate for the purpose.
W6
Artiricial-Line Oscillators in Plate and Grid Circuits
In the circuit of this oscillator (Fir.11-8a), as distinct from the one pre-
viously discussed, the artificial line between the cathodes is lackinr. In place
of the line, parasitic inductances of the cathode leads of the tube can be assumed
to be cut in. herefore, the equivalent circuit of one arm of the oscillator
c)
FiE.11-8 - Oscillator with Artificial Lines in Plate and Grid Circuits
a - Vetwork; h - Equivalent circuit; c - Single arm circuit
may, in this case as well, be reduced to 7ir.11-5c. If the inductances of the
151:
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
0
jcathode leads are small enough, an oscillator of the given type may operate on an
equivalent tapped circuit with inductive feedback (Fig.11-8b and c).
It is important to note that for operating by means of an inductive feedback
circuit, the parallel circuits CEfLg and CafLa must have natural resonance frequen-
'cies higher than the oscillated frequency. Otherwise they will have no inductive
reaction to the oscillated frequency. The equivalent inductance Lai. must be larger
than the equivalent inductance LEI, since it governs the inductance of the right arm
of the oscillating system in Fig.11-8c. The magnitude of the inductance Laf basic-
ally determines the resonance frequency of the oscillating system of the oacillator.
Therefore, the basic element in control of the oscillated frequency in an oscillator
of this type must be considered to be the plate line. In this case, feedback is
regulated by means of the grid line.
Oscillators of this type are usually employed for work in the 100-170 mc fre-
quency band. At higher frequencies, the use of such an oscillator is not particu-
larly desirable, since the frequency oscillated is smaller than the natural fre-
quencies of the grid and plate circuit networks, i.e., the potentialities inherent
in the oscillating system are utilized poorly in shorting the wavelength.
Oscillator with Artificial Lines in Grid and Cathode Circuits
In this type of oscillator (Fip.11-9a), the tube anodes are connected by solid
wires of the least possible length, usually in the form of tube holders firmly
gripping the plate. The plates may be considered to be HF grounded to the frame
across the capacitance of this holder.
Although, strictly speaking, the inductances of the plate leads are not equal
to zero (the result being the circuit in Fig.11-5b), they may be ignored for pur-
poses of a first approximation, and it may be assumed, for simplicity in analysis,
that the plates are shorted. The result is that the diagram is reduced to the
equivalent push-pull retwork of Fig.11-9b and to the equivalent circuit of the
arm 11-9c.
155
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
Since it is only the capacitance Cgf that is coupled between the grid and the
tube cathode in the oscillating system of the oscillator, conditions of self-
.
excitation car exist only if the circuit LfCaf has only a capacitive reactance to
a) b)
c)
d)
Fig.11-9 - Oscillator with Artificial Lines in rrid
and Cathode Circuits
a - Oscillator diagram; b - Equivalent push-pull
circuit; c -,Equivalent circuit of arm; d - Hart-
self-excitation circuit
the generated frequencT, while there must be only an inductive reactance in the
circuit CL ?. Consequently, an oscillator of this type can function only in a
ag g
basic tapped system with capacitive feedback (Fig.11-9d).
It is importart to note that, for operation in a capacitive feedback circuit,
the parallel circuit LfC must have a resonance frequency lower than the generated
-af
frequency, since otherwise it will not have a capacitive reaction to the renerated
frequency. The circuit L,C,r, on the other hand, must have a resonant frequency
156
STAT
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5 ?
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
higher than that generated, to prevent an inductive response. In this connection,
it need only be recalled that a parallel circuit at a frequency higher than the
resonant has a capacitive response, whileat lower than resonant frequency its
response is inductive.
Feedback in the given circuit is determined by the ratio of the capacitance C
gf
to the equivalent capacitance Cafe, and is therefore chiefly regulated, like the
oscillAting power, by the inductance of the cathode line. However, the length of
the oscillated wave is determined primarily by the equivalent inductance Lag, since
the capacitance Cgf is invariable and the equivalent capacitance Cafe need only be
varied within ziery narrow limits, in view of its strong effect on feedback and oscil?
lation power in tuning. Th, in an oscillator of the given type, the oscillated
frequency is primarily regulated by the inductance of the grid line, and the oscil?
lating power by the inductance of the cathode line.
As a rule, the capacitance Caf is lower than the capacitance Cgf. The beedback
Cafe C f
factor k ? hasalimitingvaluandincreaseswithincreasingL?On the
Cgf Cgf
other hand, the lower the inductance Lg, the weaker will be its shunting by the
capacitance Cag and the higher will be the oscillated frequency and the resultant
control on any change in the inductance Lg. Therefore, L is always smaller than Lf,
or, in other words,. the grid line is always shorter than the cathode line.
The oscillator and the load (antenna) are usually coupled by means of a feeder
connected to contacts that slide along the cathode line.
For R?F blocking of the filament circuits, the tube filaments in this oscillator
network are shunted by the blocking capacitors CbH. The cathode line tube, elec?
trically coupled to the chassis and therefore to the grounded midpoint of the fila?
ment transformer, serves as one heater wire. The other passes within the cathode?
line tube, and is therefore shielded from the high?frequency field between the
cathode?line tubes. When such shielding exists, blocking chokes in the filament
circuit become unnecessary.
157.
411.
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
An oscillator with artificial lines in the grid and cathode circuits had defi?
nite advantages over one having lines in the plate and grid circuits. Primarily,
within the oscillator shown in Fig.11-9a, the cathode line is only under R?F voltage
when the DC current is of zero potential. This permits direct coupling of the an?
tenna feeder to the sliding contacts in the cathode line and facilitates smoothness
in regulating the antenna coupling. Secondly, regulation of the oscillated fre?
quency and power is here considerably simpler than in oscillators with lines in the
plate and grid circuits. Thirdly, in this oscillator only the short?circuited
plates which have no tuning elements, are under high direct voltage. This increases
the protection from high voltage in tuning to the operating wavelength, for the
operator engaged in tuning.
Section 11-7. Triode Pulse Oscillators for the Meter Band
Triodes specially designed for pulsing are used as oscillators in tube?type
radar transmitters. In pulse operation, the power loss at the plate is related
to the average oscillating power rather than to the pulsing. In this situation,
the electrodes of the tubes, for the duration of the pulse, can withstand powers
hundreds of times as great as in continuous oscillation, since they cool down during
relatively long pause between pulses and heat up only during the brief period of
pulse oscillation.
Therefore, electrodes in pulse triodes are rather small; this makes it possible
to have low interelectrode capacitances and lead inductances. This, in turn, permits
oscillation of shorter waves. Thus, triode pulse oscillators are relatively small
in size, despite their high pulse powers.
The pulse power obtainable from a tube is determined by the expression
- pulse = gE a le
where Ea is the plate voltage;
Ie is the cathodic emission current;
158
(11-3)
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
q is the tube utilization factor, usually 0.15-0.20.
In order to provide high pulse power, a tube must have the highest possible
emission current, and operate at a plate voltage up to several tens of kilovolts.
111 .The maximum plate voltage for this typeof tube is limited by the distance between
the electrodes, the quality of the vacuum, and the inverse of the duty cycle. An
increase in plate voltage is also useful for reducing the electron transit time.
Let us estimate the order of magnitude of pulse eMission. For example, to
provide a pulse power Ppulse = 80-kw, the tube must, in accordance with eq.(11-3),
?
provide
be:
pulse 80 X 103
EaIe =
0.20 = 400 X10 3
When fed with a plate voltage Ea of 8 kw, the pulse emission of the tube must
?
_ 400 x 103
Ie = 50 amp
103
Thus, triode pulse osciliptors have to provide enormous pulse emission, ranging
up to 50 amperes and more. Emission at that level is obtained by the use of car-
bide or oxide cathodes with enlarged surface area. Oxide cathodes are preferable
since, to provide a given emission, they require less filament source power than do
carbide cathodes and still less than those of tungsten.
In view of the fact that, in order to reduce interelectrode capacitances, the
tube plate and grid are made as small as possible, the design of a triode pulse
oscillator must provide for adequate cooling of the plate and even of the grid
(usually by forced air circulation). In the operation of power tube pulse oscilla-
tors, it must be borne in mind that the presence of a powerful cathode capable of
producing dangerous overheating of the adjacent grid, requires forced cooling even
if the tube is only under filament voltage.
Figure 11-10 illustrates the design of a GI-17 pulse oscillator power triode.
159
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
3.
This tube uses a cylindrical heated oxide cathode, providing a pulse emission
of If = 70 amp. The diameter of the external cathode surface is about 18 mm,
measured on the oxide layer. The filament leads of the heater are made of molyb _
denum or tantalum and have high thermal stability. The portion of the plate within
the tube is of nickel, but the external portion terminates in a large, finned.sur-
face of brass. The finned surface meets the requirement for forced air cooling.
In addition, the inductances of the individual ribs are connected in parallel,
making the total inductance of the plate lead equal to zero, for all practical pur-
poses.
To reduce the effective inductance, the grid is not spiral, as is the case in
long and short-wave tube oscillators, but consists of short parallel wires, con-
nected at their ends by means of rings. The portion of the grid outside the tube
grid
filament lead glass bulb
copper cathode
plate air-cooling
jacket
nickel cylinder
mounting cylinder
Fig.11-10 - Design of a GI-17 Triode Pulse Oscillator
terminates in a threaded copper boss to which a silvered brass terminal with air
channels to cool the grid is screwed.
*des a 5-mm annular space between cathode and plate, the inter-
160
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
?
electrode capacitances being Csa = 8 ?. p. f, Cot. = 11 ? f, and Car = 2 1-1. f.
The electric parameters of the GI-17 triode are as follows:
Uf = 63 v; I= 75 amp; Pf = 39 watts; Ie = 70 amp; Ea = 8000 v;
= 150 watts; S = 45 ma/v; ? = 15; X trans = 0.5 m
A Doppler
Section 11-8. Designing Tube Oscillators for Ultrashort Wavelengths
As distinct from long? and short?wave oscillators, the design of ultrashort
wave oscillators is a product of the close coordination of the electrical and design
units both as to circuiting and design. The .choice of the circuit inevitably has an
influence on the choice of the oscillator tubes, determining the principle followed
in the engineering design of the oscillator, while the latter in turn is reflected
?
in further refining of the circuitry.
The following considerations must be taken as the starting point in deciding
on the type of tube:
1) Generation of the required oscillating power;
2) Operation at the required wavelength (satisfying the condition that .
X > X )*
trans '
3) Minimum heating power;
4) Desirability of selecting a tube with the lowest possible plate voltage,
simplifying the high?voltage rectiTier for plate feed;
5) Design provision for the oscillator to be employed in the proper
manner;
6) Possibility of use in the desired method of modulation.
After the oscillator circuit and type of oscillator tube have been selected,
the oscillAtor can be analyzed. An oscillator analysis involves two items:
a) design of oscillator operating conditions;
b) design of oscillator oscillating system.
161
nprdaccified in Part - Sanitized COPY Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
The main object of the former is to determine the equivalent resistance of the
oscillating system to make sure that the given oscillnting power will be obtained
in the selected schedule, with satisfactory efficiency; determination of the feed-
back factor and determination of the resistance of the grid self-bias circuit.
The second portion of the analysis has to do with the parameters of the oscil-
lator oscillating system, which provides tuning to the riven wavelength, making it
possible to obtain the feedback factor and equivalent resistance Ze, specified under
the first portion of the design, at the operating frequency.
Let us trace the major stages in the design of an ultrashort -wave push-pull
oscillator by using a numerical example. .Let us suppose that we have to design an
oscillator on the following data:
90 kw; Tu = 2.5 microsec; Fu = 400 pulses/sec; X = 1.5 m
pulse =
In the ultrashort -wave band, 10 to 15% of the oscillated power is lost in the
oscirator tube itself. Therefore, the designed oscillating power of the oscilla-
?
tor must be increased accordingly to Ptpulse = 1.15 x 90 = 104 kw.
When an oscillator is in push-pull connection, only one'arm is analyzed. Each
oscillator tube must provide one-half the calculated oscillating power of the
PT
oscillntor pulse - 52 kw. In the given case, we may use the GI-17 tube whose
2
parameters have been given in Section 11-7.
Analysis of Oscillator Operating Conditions
The reasons why surge conditions must not be permitted in ultrashort-wave
oscillators (despite the possibility of obtaining higher efficiency) lie not only
in the Pronounced reduction in useful oscillating power yielded to the oscillator
when operated under conditions of overvoltage, but also in the appearance of sig-
nificant grid currents which reduce the oscillator frequency stability. Therefore,
it is useful to operate an ultrashort-wave oscillator at the critical level, which
is characterized by maximum useful power and adequate efficiency, or under condi- STAT
162
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
?
tions of slight undervoltage close to the critical. In fact, the oscillating char-
acteristic (Fig.11-3) shows that it is specifically at the break, corresponding to
the critical schedule, at which the amplitude of swing is at its greatest.
An analysis of the operating conditions of an ultrashort?wave oscillator in
terms of critical operation, or that of slight undervoltage, may be performed in
the following sequence.
1. Let us select a plate voltage equivalent or close to the rated value for the
given tube:
Ea = e000 v
Uma
2. Let us describe the utilization factor of the plate voltage E =
Ea
When a triode oscillator is operated under conditions close to the critical, we
have g = 0.8 to 0.9. Therefore, we take E = 0.85.
3. The amplitude of the oscillating voltage at the plate is
= E Ea
ma
U = 0.85 x 8000 = 6800 v
ma
L. The phase angle of internal impedance is
T? To
e
360 r (mm)
X (m) VETT7,3
where r is the distance between plate and cathode;
cf)
360 X 5
Te- 13.3?
1.5 4E55
5. The amplitude of the first harmonic of the plate current is
2PTpulse 1
'ma = m
,,ma cos Te
2-x 52 X103
'ma == 15.8 amp
6800 X 0.973
163
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
6. Let us determine the angle of cut?off 0 of the plate current. For an os?
cillated triode, it is desirable, from the viewpoint of good efficiency, that
0 = 80 ? 90?. We will therefore assume that 8 = 90?.
7. Now let us use the Tables by Academician A.I.Berg to find the factor for
the first plate current harmonic, a = 0.5, the factor for the constant component
1
ao = 0.319, and the factor for application of internal impedance of the tube a = 2.
8. The amplitude of the plate current pulse is
'ma
1a max
15.8
la max .
= 31.6 amp
05
9. The constant component of the plate current is
I ao = (Ito i a max
= 0.319 x 31.6 = 10. 2 amp
I ao
10. Let us calculate the critical utilization factor of the plate voltage gcr
which ensures operation in the critical conditions at a given plate current pulse:
ia max
g cr = -1 ?
Cr
.a
where S is the transconductance of. the curve for critical operating conditions
cr
and represents the geometric locus of points determining the limit of the interval
for the critical schedule (Fig.11-11).
For triode oscillators, it is usual that
and, in our case,
Scr = 0.8S
Scr = 0.8 x 45 = 36 ma/v
164
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Consequently,
gcr = 1-
31. 6
36 X 10-3X 8000
= 0.89
11. A comparison of g cr with the value of = 0.85, which we had assumed,
?
shows that our oscillator operates at less than full voltage (E <
cry' but close
to the critical. Two conditions may still apply here:
a) g = gcr' corresponding to the critical schedule;
b) E > E , corresponding to overload conditions. To avoid overloading it
cr
would be necessary to assume a new value for g , somewhat smaller than g cr2 and to
repeat the analysis from the start, so as to be sure of a critical or slightly
overloaded set of conditions.
12. The equivalent resistance of an oscillating system at the oscillated
frequency is
Ze
Uma
'ma
6800
Ze = 15.8 = 430 ohms
13. The static ihGernal impedance of the tube is:
15
R- = =
1 S --
= 335 ohms
The total internal impedance of the first harmonic of the plate current in
operation at a cut-off angle of 5 = 90? is
=
?11! ,
R! = 2 X 335 = 670 ohms
1
14. From the equation for the tube Oscillator we find the swing of the excit-
ation voltage:
_
1
Umg = 171 Ima Ze)
165
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
umg _2_ 15.8 (670 + 430) = 1150
15
15. The feedback factor is
Umd
k=
Umg
- 1150
k= = O. 16
6800
Let us determine the grid self-bias voltage permitting operation at the cut-off
?
?K?
Fig.11-11 - Critical Schedule Curve
angle selected
Eg = D (Ea - Eao) - (Umg - DUmg) cos 0
when A is 900, we have:
Eg = D (Ea - Eao)
The Ea voltage, termed the reduced
plate voltage, determines the plate vol-
tage at which the idealized static charac-
teristic ia = f(Ug) passes through the
origin of coordinates. For the GI-17 tube
we have Eao 500 v, found on the basis of
the tube curves. Consequently,
1 1
E = - - (Ea - Eao) = - (8000 - 500) = - 500 a
15
17. The mean average oscillating power is
?
pulse
Pmean
166
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
?
1
1
= 1000
T F
u u 2.5 x 10-6 x 400
while for a single tube we have
18. The input power is
Pulse value
Mean value
52 x 103
Pmean . 1000 = 52 watts
o pulse = I aoE a
Po pulse = 10.2 x 80000 82 x 103 watts
P?
o imp = 82 x 103
Po mean
103
= 82 watts
19. The mean losses per plate are
pa mean = po mean ? mean
meana mean = 82 ? 52 = 30 watts
. The permissive losses at the plate of the tube GI-17 are 150 w, so that
Pa mean < pa mean Doppler
20. The efficiency of the oscillator is
Pay 52
n = 82 = 0.64
Po av
21. Since the circuit is push?pull and the bias voltage is created by the
direct component of the grid current of two tubes, the resistance of the grid bias
circuit has to be determined from the equation
167
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Rg =
Eg
2 'go
The magnitude of this resistance is determined by the choice of the oscillator
schedule, so. that there is no need for a particularly exact analysis on the basis
of the grid-current curve.
For purposes of a rough estimate it may be considered that the grid current
is 0.15 to 0.20 of the plate current. Therefore,
Igo = O. 151Iao= 0. 15 >c 10. 2 = 1.52 amp
500
n=
1.52 2 165 ohms
g
For convenience in adjusting the oscillator schedule, a resistance consisting of
four 50-ohm resistances connected in series, with step-by-step switching, is used.
Note: If the problem were to obtain from the oscillator the largest possible
oscillating power, it would, be necessary to assume the maximum plate current pulse
as
_
where Ie is the cathode Toulse
The following are then determined:
E = 1 -
cr Ser Ea
0.8 'e
Uma = E
cr a
'ma = c611a max
Iao = ry
wo a max
--IIU
cos
2 ma ma Pe
pulse
168
for the cut-off angle selected;
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
?
?
?
Further analysis may be carried outin the manner indicated above. At the end
of this analysis, the density of the, schedule is checked by comparing the magnitude
of the nominal plate voltage and the maximum grid voltage
"a min = Ea - lima; ug max = Eg Ug
whileinthecaseof. u < conditions of overloading exist, and if
ua min max'
ua min < u g max'
the situation is one of underloading. Where ua min = ue max, oper-
ation is in the critical schedule..
Analysis of the Oscillating System of an Oscillator
For purposes of analysis of the oscillating system of an oscillator it is
necessary to draft an equivalent circuit and convert it to an elementary tapped
self-excitation system. In our case, the tapped circuit of an oscillator arm is
illustrated in FiE.11-10d. The distribution of the parameters of the oscillatinE
system over its various circuits has to serve for tuning the oscillator to the
required operating frequency, and also for obtaining the required Ze and feedback
factor k, determined by the foregoing analysis of the oscillator schedule.
1. In accordance with the tapped circuit (Fig.11-10d), the expression for the
feQdba_Qk_factor
yields
k=
Uma Xafe Cgf
Umg Xgf . Cafe
Cafe = k Cgf
Cafe = 0.16 x 11 = 1.76 1111 f
2. From the equation of conductances.of the circuit (Fir.11-10c)
1 1 1 Caf Cafe 1
Or
X cafe X Caf XL f 532X 532X X Lf
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
169
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
we find one?half the reactance component of the cathode line input impedance
X Lf =
532 X
Cal ? Cafe
532 x 1.5
x =.2 ? 1.76 = 3300 ohms
One?half the input inductance of the cathode line will be
X Lf X
L=
1 1880
3300 x 1.5
. Lf = = 2.62 ph
1880
3, The natural frequency of the circuit CafLf is
fat' =
0.62 x 2
The resultant frequency is smaller than the oscillated frequency (f = 200 mc),
which is as expected.
4. The total capacitance of the oscillating system is:
faf=
159
159
= 70 megacycles
C=
Cgf Cafe
= 1.76
C
11+ L76
1.52 IL Lf
5. Ignoring the phase angle of the equivalent resistance Tel which may be
corrected in tuning, we find the total inductance of the generator oscillating
system, permitting operation at the given. wavelength:
2.53 x 104
L = Lage =
f2c
2.53 x 104
L = Lege = = 0.415 p. h
.2002 x 1.52
170
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Ap?roved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
6. From the equation for conductances in the circuit
1 1 1 1 1 1
or = +
XL XL Xc XL Xc
aga ag ag
g age
we find, first, the conductance and then one-half the reactance component of the
input resistance of the grid line
1. X Cag
+
X Lg _ 1880 Lege 532X
1 1.5 8
. + Z 0.012.
XL 1880 x 0.415 532 x 1.5
g
1
X = = 83 ohms
XL 0.012
One-half the inpot impedance of the grid line is
XL X
L
g 1880
83 X 1.5
L=
1280 = 0.067 p h
159
fag =
- 0.067
= 220 megacycles
The resultant frequency is higher than that oscillated (F = 200 mc), which is
as should be.
After the parameters of the generator oscillAting system are defined, the
design analysis of the oscillating lines can be started.
As leaned from practical experience, it is necessary, in line design, to
strive for the optimum ratio
where D is the
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
diameter.
In order to ensure short?circuiting of the tube plates, and in view of the
dimenslons of the GI-17 tube, D has to be 60 mm. In order for the ducts not to be
too thick, we set d as 12 mm. As a result, the line characteristic becomes:
2D
p = 276 /g d
2X 60
P = 276 / = 276 ohmis
g 12
Knowing the magnitude of the input inductance of the line LBx, the required length
of the shorted portion can be determined. Considering that
we obtain
or
???
2 n
Xin = o tan?k--
/
/ = ? tan-
2 rt
X ( cm) _ Xin) ?
I (cm ) = tan ?
360
From our analysis of the parameters of the oscillating system, we have
the following:
Input impedance of the cathode line
Xin f = 2X = 2 X 3300 = 6600 ohms
Input impedance of grid lines
Xin g = 2XL = 2 X 83 = 166 ohms
Consequently, the length of the shorted portion of the cathode line becomes
150_1 6600
tan = 0. 416 (tan- 1 24)? = 36.5 cm
360 276
STAT
- 172
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
and the length of the shorted portion of the grid line will be
/g ? = 0 416 tan-12) = 0.416 (tan-1 2.3)? '.;-; 13 cm
276
In view of the complexity of analysis of'the equivalent quality of the oscilla-
ting system, it should be noted that the required equivalent resistance Ze =
Ima
is ensured when the oscillator tuning is controlled by the type of antenna coupling.
grid circuit
cathode R-F output
line
plate circuit
Fig.11-12 - Tube for N = 50 cm, Combining Four Triodes
and One Oscillating System
Section 11-9. Microwave Oscillators
In the microwave band, the interelectrode capacitances of the tubes and the
inductances of-the-leadstitute-iddrticularly significant components of the
oscillating systems of tube oscillators. It is, therefore, highly desirable that
tubes and oscillating circuits be cOmbined into single structural elements.
Figure 11-12 gives a typical design of such a tube, providing a pulse power of
about 200 kw at a wavelength of about 50 cm. In essence, such a tube is a combina-
tion of four triodes, comprising two push-pull networks operating in parallel. The
oscillating system of the oscillator forms the interelectrode capacitances of these
triodes and of four lines: the plate and grid, located within the tube enclosure,
and two external cathode lines permitting adjustment of the oscillator schedule.
For purposes of operating at frequencies up to 3000 mc (10 cm), a tube was
173
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
designed in 1940 by N.D.Devyatkov, Ye.N.Danilttsev and V.K.Khokhlov, based on disk
electrodes. These tubes became widely known under the name of lighthouse tubes.
The electrode leads from the glass enclosure of a lighthouse tube take the
form of coaxial disks (Fig.11-13a), making it possible for lighthouse tubes to be
used as removable design components of generator oscillating systems (Fig.11-13b),
composed of coaxial-cylindrical cavities. The cavities are tuned by means of plun-
gers. The disk leads are component parts of the cavities, while the plate and
b)
ground
tunable plate
plate radiator
a) cavity
output
tunable
grid
cavity
input
cathode
grid
heater
Fig.11-13 - Lighthouse Tube Fig.11-14 - GI-7B Cermet Triode
?
cathode of the tube occupy part of the oscillator volume. The electrodes of light-
house tubes are very small in surface area, so that the obtainable oscillating
power does not exceed 1 kv, and the maximum permissible plate pulse voltage is of
the order of 3 kw.
Using the lighthouse tubes as the starting point, a series of considerably more
powerful metal and ceramic tubes was developed as early as 1944. A standard tube of
this desirn is shown in Fi, .11-14. .A _comparison of Figs.11-13a and 11-14 shows
clearly that the cermet tube differs from N.D.Devyatkovts lighthouse tube only in
that the glass has been replaced by ceramic material which is a dielectric yielding
. lower losses.
174
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Certain of the cermet tubes made in our country (the GI-7B, for example) provide
a pulse power up to 20 kw at a wavelength of 11 cm. In lighthouse and cermet tubes
of conventional design (Figs.11-13a and 11-14), the grid output is in a position
midway between cathode and plate; therefore, the most conventional oscillator circuit
for such tubes is the grounded grid.
If the internal capacitances are removed from the tube, the circuit of a self-
exciting oscillator, using a grounded-grid triode, is that shown in Fig.11-15a. As
It is only the Caf capacitance that is coupled between the tube plate and cathode,
satisfaction of the conditions of self-excitation in accordance with the rule for
circuiting tapped networks require that the plate-grid circuit XkCaf at the oscil-
lated frequency be equivalent to a given capacitance CgfeP and the plate-grid cir-
cuit XCag be equivalent to an inductance L
a af.
Thus, the diagram of a self-exciting oscillator, using a triode with grounded
a)
plate
line
cathode
line
b)
Fig.11-15 - Equivalent Circuit of Self-Exciting Oscillator
Using Grounded-Grid Triode
grid, corresponds to the equivalent tapped diagram in Fig.11-15b with capacitive
Cf
feedback, the feedback factor being k - , ?
'gfe
Adjustment of feedback, and therefore of the oscillated power, is done by means
of a plate-grid cavity.
The analysis of the schedule of a microwave oscillator is essentially analogous
175 STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
to that described in Section 11-8 for the analysis of an ultrashort-wave oscillator.
The only difference is that, in the case now under consideration, the two-wire lines
are replaced by coaxial-cylindrical cavities, i.e., by sections of coaxial lines.
Where cermet triodes are concerned, the capacitance Caf is 40 to 60 times smaller
than the capacitance Cag2 so that the feedback in a circuit with grounded grid is
many time weaker than in a circuit with grounded cathode.
External feedback may be performed by means of a coaxial line (Figs.11-13b
and 11-16) of adjustable length, connecting
the grid and plate cavities, or by means
of an adjustable supplementary capacitance
between the cavities, formed by movable
wire stubs containing a grid lead common
to both (Fig.11-17a) and terminating in
small disks in one cavity and a wire ring
in the other.
Fig.11-16 - External Feedback by
Sometimes, inductive feedback using
Means of Coaxial Line of Ad-
.
special loops is employed (Fig.11-17b),
justable Length
and loops being arranged symmetrically
The end-6-Of-ed-dh-lOop are-soa.
around the perimeter of the cavity.
sides of the surface dividing the Coaxial-cylindrical cavities, and the loop wire
passes through small apertures in this surface. However, this type of feedback is
not adjustable, due to lack of access to the coupling loops.
Section 11-10. Magnetron Oscillators
Magnetron is the name riven to an electronic device in which control of the
.electron stream of the cathode is effected both by the electric and magnetic fields.
Let us imagine a diode with cylindrical electrodes (Fig.11-18), where the cavity
is subject to the effects of an external magnetic field, parallel to its axis. At
the same time, the lines of the electric field developed between plate and cathode
STAT
176
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
PI
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
under the influence of the plate voltage, are directed radially.
In the absence of a magnetic field, the electrons emitted by the cathode would
a)
b)
Fig.11-17 - Capacitive (a) and Inductive (b) Feedback by
Means of Special Loops
be accelerated equally along the radii (Fig.11-18a). Changes in plate current as a
function of the plate voltage would be subject to Langmuirls Law, as in the ordinary
diode.
magnetic
field,
a)
plate
cathode
b) direction of
electron motion
lines of
magnetic force
Fig.11-18 - Diode in a Coaxial Magnetic Field
a - Orbit of electrons in the absence of a
magnetic field; b - Forces acting on the
electron
A force equal to Hev, where e is the charge on the electron, would be brought
to bear by the magnetic field on an electron moving at constant velocity v in a
homogeneous magnetic field HI normal to the magnetic lines of force. This force,
177
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
riC1IM
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
normal to the direction of motion of the electrons, distorts its trajectory into a
circle, equalizing the centrifugal force my2 developed as the electron moves around
the circle. From the conditions for equality between these forces
mv2
Hey =
the radius of the circle described by the electron in the magnetic field can be
determined, being
(11-4)
where m is the mass of the electron.
The plate voltage transmits radial accelerations to the electron. As a result
of the acceleration the radius of the
b) electron trajectory in the magnetic field
undergoes a constant increase, as it ap-
proaches the plate. At low magnetic field
a)
c)
cr
Fl
cr
cr
Fig.11-19 - Dependence of the
Electron Orbit on the Intensity
of the Magnetic Field
??????? ????? ?
intensities, the appearance of the tra-
jectory is that shown in Fig.11-19b. At
sufficiently high intensities of the mag-
netic field, the path of the electrons
becomes curved to such an extent that
electrons of reduced velocity do not reach
the anode but, describing a closed orbit
return to the cathode. The magnetic field
begins to be controlled by the plate vol-
tage.
At a specific magnetic field intensity, called the critical intensity H , the
cr
plate current ceases completely.
epicycloid in form, return to the
All the electrons, entering closed orbits rather
cathode (Fig.11-19c). The epicycloid may be
178
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
regarded as the orbit of a point on a circle rolling on the cylindrical surface of
the cathode. If, at a critical field intensity Her' the extreme points of the
epicycloid continue to roll on the plate surface, they will stop rolling
at H > H cr (Fig.11-19b).
The critical intensity of a magnetic field depends on the intensity of the
electrical field, and increases therefore with increasing plate voltage and de?
creasing distance between plate and cathode. The greater the plate voltage, the
stronger will be the electric field, while the greater the velocity of the electron,
the more difficult will it be to swerve it from a linear path.
It can be shown that, where cylindrical electrons are concerned, the critical
intensity of the magnetic field is
5.36 V1a(v)
Her (amp/m) = a (cm) b2(11-5)
1 ?
a2
where Ua is the plate voltage, while a and b are the plate and cathode radii.
Thus, the critical intensity of the magnetic field depends on the plate vol?
tage. When H < H 1 the same plate, voltage Ua no longer permits all electrons to
cr
return to the cathode, so that a plate current appears, whose strength Ua remaining
the same, is governed by the magnitude of the intensity of the magnetic field H.
On the other hand, it is possible, at any magnetic field strength,ito select
a plate voltage Ua= Ua cr at which the given magnetic field intensity becomes
critical. All electrons, having described closed epicycloids, return to the cathode,
and the plate current becomes equal to zero. This plate voltage Ua cr Is termed
the critical.
Since eq.(11-5) corresponds to the condition of absence of plate current, it
1
remains valid if the real plate voltage is taken as equal to the critical, while
the critical magnetic field strength; on the other hand, is considered equal to the
actual intensity H.
Substituting Ua cr for Ua in eq..(11-5), and H for H and squaring both
STAT
179
1
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
parts of this equation, gives a formula for the critical plate voltage for cylindri-
cal electrodes, as follows:
b 2) 2
>t
Ua cr (v) = 3.48 10-6a2 (cm) (1-71-2- H2 (amp/m) (11-6)
When Ua > Ua Cr' this same magnetic field H is no longer in a position to re-
turn all electrons to the cathode, so that of which a plate current appears, the
strength of which, H being invariable, is governed by the magnitude of the plate
voltage.
When the magnetic and electric fields are both effective, the strength of the
plate ccarrent is a function of the magnetic field strength and of the plate voltage
la = b (11, Ua)
Plate-current characteristic curves expressing this dependence plotted in the
absence of R-F oscillations, i.e., for static conditions, are called static magne-
t, - tron characteristics or, due to their extreme steepness, clipper chaacteristics.
Figure 11-20a shows these characteristics in i a' 1a coordinates, while
Tig.11-20b shows then in ia' H coordinates.
---------- -- --In sunraarizilv-4.4e foregoing, it should be emphasized that, under static con-
_
ditions in the area of no plate current, all electrons emitted by the cathode
describe closed epicycIoid orbits and return to the cathode. Taken together, they
form a space charge which takes on the nature of an electron rotor, spinning around
the cathode. The velocity of rotation of this rotor, and its energy, depend on
the choice of H and Ua.
Although there are a number of standard magnetron types, we will discuss only
those with the multicavity resonant magnetron, the first of which in the world was
developed in 1936 by the Soviet engineers F.F.Alekseyev and D.Ye.Malyarov, on the
basis of an idea of that outstanding figure in radio engineering, M.A. Donch-
Bruyevich. This type of magnetron has been used exclusively in radar, as a powerful
STAT
180
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14 : CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
oscillator of very short and ultrashort waves.
The plate of a multicavity magnetron consists of a system of cavity resonators
in the form of longitudinal slots (chanters) whose body has slit apertures to the
a)
cri
cr2
b)
cri
cr2
Fig.11-20 - Characteristic Clipper Curves
internal cavity (Fig.1I-21a). Figure 11-21c illustrates standard types of apertures.
Each resonator has distributed inductances and capacitances and serves as a cavity
oscillating circuit.
In a multicavity magnetron it is primarily the inductive effect of the electron
rotor that is employed. This rotor, spinning around the cathode, directs charges,
varying in time, to the plate segments, and thereby excites R-F oscillations in the
resonators. The final velocity at which the rotor spins is such that the oscilla-
tions in adjacent resonators do-not coincide in phase. The phase shifts depend on
the distance between the slots and the rate of rotation of the rotor.
The R-F fields of capacity resonators are able, due to the presence of slots
which affect the DC fields created by the external magnetic system and the plate
voltage, to exert a strong influence on the configuration and energy state of the
electron rotor. If the electron rotor, passing a resonator, is retarded by the
R-F field of the slot, it will yield part of its energy to the resonator. If on
the other hand, the rotor, passing the resonator slot, is accelerated by its elec-
tromagnetic field, the rotor will absorb a portion of the resonator energy. At the
STAT
181
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
.moment of Criction shown in Fig.11-21b, the slots of resonators I, III, V, and VII
accelerate the motion of the rotor, and the slots II, IV, VI, and VIII retard its
motion.
a)
b)
rotating space
charge
c)
Fig.11-21 - Multicavity Resonator Magnetron
a - Cutaway section; b - Electron rotor;
c - Standard types of slots
The loss of oscillating energy in each cavity is compensated by the retardation
of the electron rotor which, in the final analysis, is the source of plate supply,
To make it possible for the electron rotor to yield more energy- theretarding
slots than it took up from the accelerating, it must approach more closely to the
retarding slots than to the acceleraring. The required change in the configuration
of the electron rotor is reached autpmatically. As it approaches the accelerating
slot, the electron rotor is repulsed by the negative R-F charge of the corresponding
,plate segment, and becomes indented. As it approaches the retarding slot, on the
other hand, it picks up a further R-F cl-arge on the corresponding segment of the
plate, and is pulled toward the accelerating slot.
Thus, mder_dynamic conditions, 1.e., in the presence of R-F oscillations,
the electron rotor takes on clearly:defined "peaks" facing the retarding slots,
182
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
and "dents" opposite the accelerating slots. The electrons falling into the area
of the rotor "dents" return to the cathode very rapidly. The electrons moving in
the area of the "peaks" are so affected by the R-F field as to undergo deformation
of the orbit from an epicycloid to a'spiral; after adding to the energy of a cavity
with a retarding slot, they strike the corresponding segment of the plate. Due to
this fact, a certain plate current flows when the magnetron is operating under dy-
namic conditions and the critical intensity of the magnetic field is high, the re-
sult being that there an energy denand will be made on the plate-voltage source.
The modes operative within a magnetron are defined in accordance with the
phase shift between oscillations in adjacent resonators. The most desirable are
out-of-phase oscillations of the It -mode, when oscillations in adjacent resonators
are mutually opposite in phase. If the rate of rotation of the electron rotor is
such that the transit time between adjacent slots is one-half the oscillation period
in the resonator, the rotor will undergo constant retardation, since during the
transit time from the retarding slot, the formerly accelerating slot, having under-
gone a 180? change in phase, will also have changed to a retarding slot. When the-
-regnetrob operating conditions are properly chosen, 45 to 85% of the power required
--1Y6M-the -prate- -feed -source is converted-into-R-F-oscillatin-g-puwer:--However, a -----
portion of the oscillating power is expended on heating the resonators, so that the
power yielded to the external load is only 30 to 605 of that delivered to the magne-
tron from the plate-voltage source.
A magnetron oscillating system consists of a large number of coupled cavity
circuits, the coupling being electroma-gnetic in nature and carried out through the
'space between plate and cathode, called the interaction space. In view of the
large number of coupled circuits, the oscillating system of the magnetron has a
series of resonant frequencies. The choice of H and Ua governs the mode. The fre-
ciuency of the magnetron is unstable during the increase in number of oscillations.
In pulsed operation, should the modulated pulse have a very steep edge, the magsTATin
183
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
?
may skip the specified frequency and may operate at a wavelength differing greatly
from that given, during the period Of UHF pulse oscillation. Therefore, great care
is taken in selecting the conditions of
modulation of a marnetron.
In addition to the it-Node other
modes may exist in a. magnetron. These are
parasitic oscillations, whose presence in-
creases the loss in the resonator chambers,
reduces the efficiency of the magnetron,
Fig.11-22 - Straps
and diminishes the stability with which
it -mode excitation occurs. Therefore, steps are necessary to remove the possibili-
ty of any parasitic modes being present.
With this object, magnetrons working on waves close to 10 cm length, are pro-
vided with s-called straps. These are copper rings connecting equipotential points
in resonators to operate in the it-mode. The rings are attached to every other
resonator (two rings at the top and two at the bottom - Fig.11-22a). In view of
the equipotential state, no R-F current passes throuph the straps during oscilla-
_
- For all t,I=f-piiciyit?deiftE16-Ct;e-d by the straps will
-
no longer be equipotential. Therefore, current will flow in the straps in all other
,
modes. The strap fields, interactillg with the fields of other modes in the resona-
tors, reduce the effective inductarce of the resonators for all nodes other than it .
As a result, the natural frequencies of the resonators increase considerably as
concerns parasitic oscillations, and excitation of parasitic oscillations is rendered
more difficult, since the magnetron operating conditions (H and U)are based only
on the conditions of excitation of frequencies similar to the main oscillation fre-
quency mode it
For waves close to 3 cm in length, design difficulties in the use of straps
result in their replacement by a system of nonidentical pairs of resonators
18/1.
Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
Declassified in Part - Sanitized Copy
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
(see Fig.11-22b), to eliminate excitation of parasitic oscillations. The wavelength
of the it -mode is determined by the connection of large and small resonators in
parallel. The dimensions of each are so selected that the possible parasitic oscil-
lations will be of frequencies significantly higher than the TE -mode. 1
magnetron coaxial
resonator line cavity
resonator
waveguide
to antenna
coupling loop
Fir.11-23 - Coupling of Magnetron to Antenna Line
In the it-mode, all the cavity resonatcirs of the magnetron may be regarded as
connected in parallel. The oscillating power is taken off by means of coaxial
lines coupled, when the wavelength is
about 10 cm, by magnetic loops into one of
the resonator cavities (Fig.11-23). A
- c-avity- re sonator -ac-t-ing- as _a riesistarice
transformer is used between the coaxial
line and the waveFuide to match the trans-
ition to the waveguide. At wavelengths of
Fig.11-24 - Extraction of EnerEy
less than 3 cm, the waveguide itself is
from Magnetron through Slot in
excited by means of one of the magnetron
Waveguide (3 cm wavelength)
slots, with no intervening coaxial line
(Fir.11-24).
The length of the stable wave, oscillated by the magnetron, is determined by
the configuration and dimensions of the resonator cavities, and also by their
number.
185
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
The theory of multicavit-, resonator magnetrons was largely developed by the
Soviet scientist, V.F. Yovalenko. It was he, also, who oririnated and verified in
Fig.11-25 ? General View of ilagnetron with Permanent ragnet
experiment the equations for resonator analysis.
The bandwidth of operating wavelengths for multicavity magnetrons of various
types varies ,from 50 to 1 cm, although magnetrons yield less power at the shorter'
wavelengths, due to a reduction in the dimensions of the resonator cavities. The
limiting pule powers of magnetrons are approximately the following:
-? --?_-_.--__ ?-?--? ---?-
x
P
, pulse
1 cm -
3 cm
10 cm and longer,
up to 100 kw
" " 300 kw
I I TT , '1 megawatts
Depending on the type of magnetron, the pulsed plate voltage may vary from
1 to 45 kw. The magnetic field is provided either by a permanent magnet or by an
electromagnet, the magnetron being placed between the pole pieces. The magnetic
induction of the magnet is between 60 and 15,000 gauss, the decision being made
in terms of the plate voltage and the operating wavelength. The plates of powerful
186
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
narnetrons are water-cooled.
The marnetron cathode is usually of the oxide type. The heater heads are care-
fil117 insulated from the chassis, which is rrounded. The cathode is plased nera-
tively with the equivalent oC the plate voltare. The operatinr temperature of the
cathode is determired rot so muc! b- the heater as b7 heatinr of the cathode by
hombardmert yr returrinr electrons.
arure 11-25 preserts a reneral view of a marretron.
?
187
Declassified in Part - Sanitized Copy Approved for Release 2013/03/14: CIA-RDP81-01043R001800200001-5
STAT