CONSTRUCTION OF AN ANTENNA IMPEDANCE MATCHING UNIT GIVING AN OPTIMIZED MATCH

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78-03424A000800010034-7
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 25
• Document Creation Date: December 22, 2016
• Document Release Date: February 8, 2012
• Sequence Number: 34
• Publication Date: September 1, 1959
• Content Type: MISC

Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
C P, 11 G 11 IN ?. C F'-r Z3 ~~`J 7 9
^ D:.C! L 1 VW ON ?~ / v
EXY G, YND 6 YRS BY S X9/'1 C
REASON _3 C1)
nIDFi"T!L
DOC JS7 R DATE 4 nP o _ By
OHIO COMP -s _ OPI TY1150
CRIB CLASS rAB@S _ .i V CLASS _._
JUST NUT MEY / V AUTHi HR 10.2
Proposal For The
Construction Of An Antenna Impedance
Matching Unit Giving An Optimized Match
Prepared by: I 25X1
Date: September 1, 1959
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
Table of Contents
Page No.
I. Introduction . . . . . . . . . . . . . . . . . . 1
II. Statement Of The Problem . . . . . . . . . . . . 3
III. Proposed Solution . . . . . . . . . . . . . . . . 6
IV. Manpower Requirements And Time Schedule. . . . . 9
V. Facilities . . . . . . . . . . . . . . . . . . . 12
VI. Identification Of Key Technical Personnel. . . . 15
VII. Appendix . . . . . . . . . . . . . . . . . . . . 17
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
I. Introduction
During the past year the
has been engaged in the development of an automatically tuned trans-
mitter operating in conjunction with an automatic antenna impedance matching
network. The problems associated with the design of a suitable antenna
matching network are considerable. Although the requirement for automatic
adjustment adds circuit complexity, the major difficulties arise in the de-
sign of the matching network itself regardless of whether it is for manual or
automatic adjustment.
The transmitter which is being developed during the present program
will match to the range of antenna impedances agreed upon with the Contracting
Agency. However, it is apparent that, using physically realizable components,
the power transfer efficiency of a network even under matched conditions can
still be quite low. This proposal describes a method whereby the consequences
of the losses inherent in any practical reactances may be largely avoided.
The system requires the adjustment of three variables, which in a manual sys-
tem, would make matching an extremely laborious process. However, with auto-
matic adjustment, a reasonable amount of control circuitry would enable an
optimum matched condition to be achieved. By an optimum matched condition is
meant that state of adjustment at which the antenna is matched to the trans-
mitter in such a manner that the effects of the finite values of Q associated
with the reactances are reduced to a minimum.
The proposed equipment would consist of an automatic impedance
matching unit suitable for use with a low power transmitter having a specified
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
output impedance. The unit would be provided with input, output and 12 v. dc
power supply terminals, the necessary potentials being derived from a self
contained converter.
It is appreciated that physical size is of extreme importance.
Efforts will consequently be made to keep the dimensions to a minimum. How-
ever, the size of the individual components from which the unit will be
built is a function of the amount of effort to be expended on their develop-
ment. At the present state of the art, electrically variable reactances for
operation at power levels of 10 watts are not available. It will consequently
be necessary to use motor driven components. With sufficient effort variable
reactances having very little mechanical friction could be developed. These,
in turn, would permit operation from smaller motors than are currently avail-
able. However, since a major program of motor and component development is
not being proposed, the design of the antenna matching network will have to
rely on components which are available or may be adapted for this application.
These items impose a rigorous limitation on the degree to which the equipment
may be miniaturized. At the present time a volume of approximately 170 cubic
inches is visualized.
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 I
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
II. Statement Of The Problem
In order to transfer RF energy from a transmitter to an antenna in
an efficient manner it is usually desirable to place an impedance transforma-
tion network between them. The purpose of the network is to ensure that the
transmitter always sees a load impedance which is the same as its own internal
impedance. With ideal components in the matching network maximum power will
be transferred to the antenna with such an arrangement. In any physically
realizable system, components which are less than ideal have to be used.
This is a particularly severe handicap in the case of inductances, where a
Q of 200 is considered to be very good.
As an indication of the extent to which the finite losses in the com-
ponents of a matching network become significant, the middle curve in Figure 1
shows a plot of the percentage of the available power from a transmitter which
is actually transferred to the antenna assuming a two variable u matching net-
work with a coil Q of 200 for different antenna reactances. The bottom curve
in this figure shows the amount of RF energy fed into the antenna if no
matching network is used. Figure 2 shows plots of power transferred to the
antenna, for different values of resistive antenna, in the unmatched and two
variable matched case, again with a coil of 200.
The circuit on which the calculations for Figures 1 and 2 are based
is shown in Figure 3. The following assumptions are made:
(a)
(b)
Coil Q = 200
Capacitor Q's = Co
(Not true but a reasonable assumption.)
(c)
The variable elements are adjusted for each antenna to present
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
P30X100%
P ACTUAL DELIVERED POWER
Po 100%? MAXIMUM OBTAINABLE POWER
X 100% AS A FUNCTION OF A LOAD
OF ZA ? 25 + I X AT 15 MC.
0
0
0
FIGURE I
COIL 0 ? 200
1A S
0 0
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
BOTH
MATCHED
O% L
I0
IF LOAD IS
UNMATCHED
TWO -VARIABLE - Pi
AT 15 MC
COIL 0. 200
C3. 51O??0
1
POWER DELIVERED FROM -
'TWO-VARIABLE - Pi
AT 3 MC
GRAPH OF P/Po X 1009: - ACTUAL POWER DELIVERED X 100%
MAXIMUM OBTAINABLE POWER
RA
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
ANTENNA
SYSTEM
MATCHING NETWORK
r------ ---- --1
~ QL' 200 I
FIGURE 3
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
a load of 500 + j On. to the transmitter.
The calculations, a sample of which is given in the Appendix, are for opera-
tion at 15 me except for the top curve of Figure 2 which is for 3 me operation.
The frequency is significant in that the high losses occur mainly as a result
of the high value of the fixed capacitor at the output of the network. This
large capacitor is necessary in order to be able to accommodate certain
antenna impedances at 3 me with realizable maximum to minimum ratios for the
variable elements. The losses could of course be reduced if a coil of higher
Q could be used. However an increase in coil Q by a factor of 2 does not
lead to a reduction of the losses in the network by a factor of 2. Further-
more, any improvement in coil Q above 200 would be quite marginal in an
application where small physical size is of extreme importance. It is
possible to wind relatively small coils on high frequency ferrite material
and obtain Q's as high as 100 or even 500. However as soon as such a coil
is placed in a metal case, the Q drops sharply. In applications where the
coil has to be placed close to the sides of a metal case, as for instance in
the RT21 transmitter where a total case thickness of 1 1/211 is specified,
extremely high Q's are not possible. The situation can be improved by placing
a ferrite strap around the coil but this procedure adds to the total volume
of the coil. Consequently, although under ideal conditions Q's substantially
in excess of 200 are obtainable, within the constraints of the particular
application, such high Q's are not realizable.
The situation which has been described above exists, it will be
readily appreciated, regardless of the tuning method; i.e., in both manual
and automatic systems. Fortunately a solution is available which lends
Declassified in Part- Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
itself to automatic adjustment more readily than to manual operation. A
servo system can be designed in which three variable elements are used where-
as the manual optimization of a three variable system would be extremely
tedious. As will be described in this proposal, a three variable system
largely overcomes the serious losses which can occur when a matched but non
optimized matched condition is used.
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
III. Proposed Solution
As described in the Statement Of The Problem the high losses inherent
in a conventional matching network may be attributed largely to the high value
required for the fixed capacitor at the output of the IT network. This high
capacitor value is, in turn necessary in order to make a match to some possible
loads at low frequencies. The proposed system utilizes a variable capacitor
at the output of the n network. This third capacitor is controlled by the
portion of the other two variable elements. The basic operation of the system
shown in block diagram form in Figure 4, may be described as follows.
The optimized admittance match is achieved by manually operating a
sequential three position switch. The three positions are "Off', "Tune", and
"Optimize". The desired value of input conductance is first achieved in the
"Tune" position, and then improved efficiency may be achieved by switching to
the "Optimize" position.
The manner in which the Three-Variable-Pi is driven to produce a
purely conductive admittance of value Go is as follows: The operation of the
system requires that both L2 and C3 be at their maximum positions when the
tuning cycle begins. This is accomplished automatically by means of a relay
which, when unactuated, applies signals which drive both L2 and C3 to their
maxima. When both L2 and C3 have been driven to their maxima, limit switches
then actuate the relay and the tuning cycle begins. The output of the phase
detector tends to drive C 1 to the position which cancels any phase angle
associated with the input admittance. The output of the magnitude detector
tends to drive L2 to the position which produces the desired value of input
conductance, provided that C 1 was not initially driven to its minimum position.
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
Yin ? Gp + j 0
ADMITTANCE
MAGNITUDE
DETECTOR
DC SIGNAL TO
DECREASE L2
OPTo
00
TUNE
OFF?
OPT
TUNE,
!-1,0
sf
o .. 0
0
'
MIN C3 MIN
L2 MAX C3 MAX CI
L2 MIN SW
TUNE 00 0
OFF o
I DC SIGNAL TO
I DECREASE C3
SEQUENTIAL SWITCH
CI MAX SW
ADMITTANCE
PHASE
DETECTOR
DC SIGNAL TO
INCREASE L2
DC I
SIGNAL TO I
INCREASE I
C3 fl
SCHEMATIC DIAGRAM OF SYSTEM
INPUT
CAPACI TOR
CI
INDUCTOR
L2
OUTPUT
CAPACITOR
C3
41
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
If C1 is initially at its minimum position, a limit switch is actuated. This
limit switch applies a signal which causes L2 to decrease until C1 leaves its
minimum position. At this time the control of L2 is returned to the magnitude
detector. Under certain conditions, Cl may be driven to its maximum or L2 may
be driven to its minimum before an admittance match is achieved. In these
cases limit switches which cause C3 to decrease are actuated. C3 continues
to decrease until the limit switches are released. This then completes the
tuning cycle.
When the sequential switch is moved to the "Optimize" position, a
signal which causes C3 to decrease is applied. As C3 decreases, C1 and L2
move to maintain the admittance match. C3 continues to decrease until the
limit switch at either L2 maximum, C1 minimum, or C3 minimum is actuated.
When any of these switches is actuated, C3 stops and the "Optimized" admit-
The reduction of the capacitor at the output of the n network to the
lowest value consistent with the ranges of the other two variable elements
improves the network efficiency for nearly all values of antenna impedance
contained within the 25 to 1300 + j 1000 SL impedance area originally speci-
fied. For the points where the adjustment of the output in this manner does
not lead to an optimized match, with a coil Q of 200, the resultant non opti-
mized match would still provide an efficiency in excess of 95070.
The improvement in power transfer efficiency which can be expected
as a result of adjusting for an optimized match is represented by the upper
curve in Figure 1 and by Figure 5. Figure 5 shows, for various values of
resistive antenna load, the percentage of the available transmitter power
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
a?
a
0%
10
ZA ?Ra+jO
FIGURE 5
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
MATC
HED 3 VARI
ABLE Pi
I
f ? 15 MC
COIL 0 ^
200
MATCHED 2 VARIABLE Pi
U
NMATCHED
Z
GRAPH OF P/P0 X 1009'0 ^ ACTUAL POWER DELIVERED
MAXIMUM OBTAINABLE POWER
AS A FUNCTION OF RESISTIVE LOADING AT 15 MC
X 100%
01
RA
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
which is transferred to the antenna for both a two variable network giving a
non optimized match and a three variable optimized matching network. The top
curve in Figure 1 is for an optimized match using a three variable network and
indicates, for certain values of antenna reactance, an increase in efficiency
from less than 10%o to over 80%o compared with the non optimized matched con-
dition. By going to a three variable network in order to obtain an optimized
match, the requirements placed on the maximum to minimum ratios of the com-
ponents are also eased. Consequently with the network shown in Figure 6 it is
possible to match to the complete rectangle originally specified i.e., 25 to
1300 + j 1000 ohms from 3-30 MC. It will be seen that the values required of
the variable reactances are quite realizable, if not readily available.
Declassified in Part- Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
BA
0.5-40?h
3-30 MC MATCHING NETWORK
FOR ORIGINAL Z AREA
FIGURE 6
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
N. Manpower Requirements and Time Schedule
The manpower effort required to implement the program described above
Electrical Design of System Man Weeks
ti~rl ie~
Engineering 21.
Technician 11
Mechanical Design
Engineering
Technician
Fabrication of Components
Engineering
Technician
Construction and Testing
Engineering
Technician
Total: Engineering 66
Technician 36
The proposed time schedule for implementing this program, which, it
is anticipated will be iniated on May lst, 1960 is as follows:
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
Electrical Design of System
May 1 - Oct. 2, 1960
Mechanical Design
July 10 - Oct. 2,
1960
Fabrication of Components
Oct. 2 - Dec. 25,
1960
Construction and Testing
Dec. 25 - Apr. 30, 1961
The above dates are based on a starting date of May 1st, 1960. If
the starting date is postponed, all subsequent dates shown will be postponed
by a similar amount.
A chart of engineering manpower is shown on the following page.
Declassified in Part -Sanitized. Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
Drafting
Shop
Shop
Electrical Design
Design
Mechanical
Design
Component
Fabrication
Component
Fabrication
Con- I
struction and testing
Construction
and Testing
May 1 July 10 Aug.7 Oct. 2 Dec. 25 Apr. 30
1960 1961
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
0`
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
Iq
Next 4 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
17
VII.Appendix
It has been previously stated that because of the loss resistance
associated with the coil, the losses in the two-variable Tr network become
appreciable under certain conditions. The reason for this may be seen by
examining, as an example, the two-variable Tr with resistive terminations (shown
in Figure 7). The impedance looking into the terminals a - a' may be repre-
sented as an equivalent series R - C. The n network in this equivalent form
is shown in Figure 8. If the efficiency of the Tr is defined as the ratio of
output power to input power, this efficiency,Y( , is then
_P _ I out
in ( ] + R ' )
1
1+RLR ?
(1)
Since it is desired to have the efficiency near unity, the loss resistance of
the coil should be kept small in comparison to the reflected load resistance,
R'. Unfortunately, this can not always be done in the two-variable Tr system.
The significance of this situation may be illustrated by the following calcu-
lations.
In order to match certain impedances at 3 mc, it is necessary to fix
C at a value of not less than 510 if. If, as an example, the input impedance
looking into a - a' is then evaluated at 15 me with RA = 1300-L ,
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 I
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
"ao
TWO-VARIABLE Pi WITH RESISTIVE TERMINATIONS
FIGURE 7
EQUIVALENT REPRESENTATION OF THE TWO-
VARIABLE Pi WITH RESISTIVE TERMINATIONS
FIGURE 8
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 ,-I- I
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
a 1 +
RA 4
X00+ (2trx15x10) (510x10 )
= 0.333 - j 20.6
Thus, in terms of the circuit of Figure 8,
'
R = 0.33 ohm
= 20.6
In order to see how R' compares with RL, it is necessary to find what value
of L (along with its loss resistance, RL = coL/Q) will result in a 500 ohm input.
At resonance, the circuit of Figure 8 has an input resistance of
+ R' + (wL - 2
61
R.
Solving this equation for cuL,
R' +7
R 2R' 1 Rin ER]' 2 R. -2R'
,
coL= 2 Q +OL ' + 2Q + uT +R (Ring )
(2)
Evaluating the above for Rin = 500, R' = 0.333, 1/wC' = 20.6, and Q = 200
= 500 - 0.666 + 20.6 + 500 - 0.333 2 + 500 - 0.666
2(200) 00 ] [ 2 00 (20.6) +
= 36.7
+ (0.333) (500 - 0.333)
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
 L I
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
Therefore,
-
= wL = - 36.7 -
RL 0.18359
~- -
P=out = o.643.
in l+RL/R 1 + 0.
This indicates that more than one-third of the power into the
the coil. This relatively poor efficiency is primarily due to the
of the reflected load resistance, R'. Straight forward circuit analysis shows
that R' can be made larger if C is reduced. (It is only for certain impedance
at 3 me that C must be 510 ??f.) With C decreased to 20 if,
-a 1
TM+ j (2rrx15x10) (20X1012)
= 186 - j 1115
= R' - jlc c' -
Evaluating Equation (2) in terms of these parameters,
a 500
- 2(116, + 1155 + 500 - 186 2 + 500 -2 186 + 00
_j (445)
2(20 1 MM L
= 699
caL ' 699 3.5.sa
11L =~. M
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
+ 186 (500 - 186)
 Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7
P out 1
= = 0.975
in 1 + RL/R-
Thus9 instead of losing 350/o of the input power in the coil,, the loss has been
reduced to 2.50/0 .
The conclusions which follow from analysis of the n are that Rt must
be much greater than RL if the network is to transfer power efficiently to
the load. This can be achieved by reducing the output capacitor when the
termination does not actually require a large capacitance.
Declassified in Part - Sanitized Copy Approved for Release 2012/02/08: CIA-RDP78-03424A000800010034-7