ENGN UMR0002 Technical Memorandum No. 97 Adndendum to ECOM Report 03733-F Step-Tunable VHF Power Amplifier By N. E. Abbott R. C. Rehner For U. S. ARMY ELECTRONICS COMMAND FORT MONMOUTH, NEW JERSEY COOLEY ELECTRONICS LABORATORY Deportment of Electrical Engineering The University of Michigan Contract No. DA-36-039 AMC-03733(E) Signal Corps, Department of the Army Department of the Army Project No. 1GO 21101 A04201 February 1966

HXY Op i.,,,,~~. ~~ I; " *

Technical Memorandum No. 97 STEP-TUNABLE VHF POWER AMPLIFIER by N. E. Abbott R. C. Rehner, Jr. Approved by: // x - / T. W. Butler, Jr. for COOLEY ELECTRONICS LABORATORY Department of Electrical Engineering The University of Michigan Ann Arbor, Michigan Contract No. DA-36-039 AMC-03733(E) Signal Corps, Department of the Army Department of the Army Project No. 1GO 21101 A04201 February 1966

TABLE OF CONTENTS Page LIST OF ILLUSTRATIONSi ABSTRACT iii 1. INTRODUCTION1 2. PRELIMINARY CONSIDERATIONS 2 3. SELECTION OF SWITCHING DEVICE 4 3. 1 Reed Relays 4 3. 2 Silicon Controlled Rectifiers 4 3.3 High Voltage Transistors 4 3.4 PIN Diodes 5 4. DIODE CHARGE STORAGE EFFECT6 5. DETERMINATION OF Z AND SWITCH LOCATIONS 7 6. REASON FOR ASSUMING DEFINITE CT VALUE 7 7. MULTIPLE RESONANCES OF PLATE CIRCUIT 10 8. BANDWIDTH LIMITATIONS 11 9. EXPERIMENTAL TRANSMISSION LINE TUNING 14 10. DISCUSSION OF EXPERIMENTAL RESULTS 18 11. PROPOSED AMPLIFIER CONFIGURATION 23 APPENDIX 27 REFERENCES 20 DISTRIBUTION LIST 31

LIST OF ILLUSTRATIONS Figure Title Page 1 Power amplifier configuration. 3 2 Location of ideal switches. 8 3 Locations of switches with shunt capacitances. 9 4 Inductor tuning. 13 5 Transmission line tuning. 1 3 6 Experimental transmission line. 15 7 Experimental PIN diode equivalent. 17 8 Measurement circuit for experimental transmission line. 17 9 Experimental transmission line resonance frequencies. 20 10 Experimental transmission line bandwidth versus resonant frequency. 21 11 Experimental transmission line fundamental resonant frequency. 22 12 Experimental transmission line bandwidth versus resonant frequency. 24 13 Proposed amplifier configuration. 25 14 Proposed diode mounting. 26

ABSTRACT The characteristics of a VHF power amplifier that is to be electronically tuned in the VHF band are theoretically and experimentally analyzed. This tuning method involves varying the electrical length of a transmission line by electronically short circuiting the line at different positions. PIN diodes are selected to perform this short circuiting function. The optimum characteristic impedance of the transmission line and the location of the PIN diodes along the line are determined with the aid of a computer. This memorandum also discusses problems of multiple resonance frequencies and bandwidth reduction. Experimental measurements of diode locations, multiple resonance frequencies, and bandwidth substantiate the theoretical considerations. Finally, a proposed configuration for fabricating the amplifier is presented. iii

I. INTRODUCTION This memorandum discusses design considerations for the development of an electronically tunable VHF power amplifier. The basic design objectives are: (1) Output power - 100 watts (2) Tuning range - 60 Mc to 300 Me (3) Instantaneous bandwidth - 10 Mc to 20 Mc (4) Tuning rate (from one center frequency to another center frequency) - 5 to 10 isec (5) Overlap between adjacent bands - 0. 5 Mc to 2 Mc (6) Load impedance - 50Q2 nominal with maximum VSWR of 3 (7) Plate circuit efficiency - 30 percent to 40 percent Power amplifiers covering this VHF frequency range have been developed which utilize distributed amplifier techniques. However, such distributed amplifiers generally possess low plate circuit efficiencies. A narrow bandwidth Class C amplifier is capable of providing improved efficiency characteristics. Thus, it is desirable to develop a method for rapidly tuning a narrowband power amplifier across the VHF band. Such a scheme should be able to provide effective wideband coverage without the sacrifice of efficiency. A relatively straightforward scheme for rapid tuning is selected to simplify the analysis. Considerations of similar but more complex

tuning methods can be made once the characteristics of the basic tuning problem are determined. 2. PRELIMINARY CONSIDERATIONS The configuration under consideration is shown in Fig. 1. A common cathode gain stage utilizing a 4X250B power tetrode was selected to provide the necessary gain, output power, and frequency range (Ref. 1). The voltage and current requirements of the transmission line shorting switches are of primary importance. An approximate determination of these switch requirements is given in the Appendix. A switch in the "open" position must be able to withstand peak potentials of about 1000 volts. A switch in the "closed" position will have to conduct RF currents on the order of 7 amperes RMS. 2

*0 I10. I oj.~'. - I --:31 Cb bfl 0' ^ ___ ^ I ~ 33 *S S ~l 4i -3 C) * j 03'.~d< C I-I r-i i -^ s g9 Fr~ O ii

3. SELECTION OF SWITCHING DEVICE Several types of devices were considered for performing the switching function. These were: (1) Reed relays (2) Silicon controlled rectifiers (3) High voltage transistors (4) PIN diodes 3. 1 Reed Relays Reed relays can satisfy the voltage and current requirements; and, they also have very small "on" resistances. However, the switching speeds of reed relays are much too slow to satisfy the switching requirement of 5 to 10 isec per step. 3. 2 Silicon Controlled Rectifiers High power silicon controlled rectifiers can also satisfy the voltage and current requirements at low frequencies. Unfortunately, they possess relatively large anode capacities and are difficult to turn off. A more fundamental limitation of SCR's in this application is a result of capacitive coupling from the anode to gate. With typical high power SCR's, the rise time of the anode potential must be limited to a maximum of about 500 volts/lsec to prevent the anode to gate capacitance from coupling sufficient current into the gate to turn the device on. The approximate rise time encountered in this application is 50, 000 volts/usec. Therefore, SCR's are not allowable. 3. 3 High Voltage Transistors Recent developments in the state of the art have yielded silicon transistors with collector-base breakdown voltages of several hundred volts. It would be possible to series connect two or three high voltage transistors to produce a switch capable of withstanding the 1000 volt peak potential. However, the large collector junction areas required 4

to handle the high RF current cause large collector capacitances (100 pf to 1000 pf). In addition, the low collector doping concentrations consistent with high breakdown voltages cause increased series resistances (1Q2 to 5Q). For these reasons, transistors are not suitable in this switching application. 3. 4 PIN Diodes PIN diodes seem to offer the most feasible solution to the switching problem. A source of silicon PIN diodes meeting the current and voltage requirements has been found. The devices are manufactured by the Semiconductor Division of Sylvania Electric Products, Inc., and bear a special part designation D5618. Typical device characteristics are as follows: Minimum reverse break- BV. > 1 kv at 10 Ha down voltage: Maximum average forward ImaAvg = 8 Amp current: Typical total capacitance: C z 5 pf at -50 v, 1 Mc p Typical series resistance: Rs 0. 5Q at 100 ma, 500 Mc Maximum series induc- L < 2 nh tance: smax - tance: Typical junction-to-case 0 = 8 C/W thermal resistance: Maximum operating T 175~ C junction temperature: Three diodes of this type were tested. They exhibited breakdown voltages of 820 v, 850 v, and 980 v; average capacitances of 8. 9 pf, 8. 4 pf, and 8. 9 pf; and series resistances of 0. 412, 0. 45, and 0. 452. The manufacturer has indicated that breakdown voltage and capacitance specifications can be met by selection of diodes from their existing PIN production. 5

4. DIODE CHARGE STORAGE EFFECT Experimental measurements were also made to determine the amount of forward bias current (Idd) necessary for maintaining diode conduction during both half cycles of the RF current (IRF). For example, diode = dd + IRF sin wt where IRF 712 amperes, according to the Appendix I. = total diode current. diode In general, to maintain diode conduction I > 0 diode Requiring Idd > IRF Thus, the peak diode current would be diode peak = dd + RF > 2(72 ) V 20 amperes and a PIN diode with current capability of more than 8 amperes would be required. However, at high frequencies, charge stored across the diode junction during the forward half-cycle of the RF current can contribute the carriers necessary for sustaining the current during the reverse half-cycle. This effect was measured at a frequency of 60 Mc. It was found that the charge storage effect was sufficient to sustain the reverse half-cycle RF current even without the application of the forward bias current. The series "on" resistance of the PIN diodes is reduced by a conductivity modulation effect under the influence of forward bias current. Due to this effect, a forward bias current of about 100 ma is required to reduce the series resistance to the 0. 512 value. 6

5. DETERMINATION OF Z0 AND SWITCH LOCATIONS Given the diode switch characteristics, it is possible to select an optimum characteristic impedance (Z0) for the transmission line and to determine the switch locations along the line. The choice of Z0 affects the degree to which a reverse biased diode appears as an open switch and a forward biased diode appears as a closed switch across the transmission line. The choice of Z0 also affects the magnitude of the RF current flowing through a closed switch, the length of the transmission line, and the instantaneous bandwidth of the single tuned plate circuit. In all these considerations, except that of simulating the open switch condition, it is desirable to make Z0 as large as possible. A computer solution was obtained to determine the switch locations to cover the frequency band in 20 Mc steps (Figs. 2 and 3). Figure 2 is the solution for ideal switches. In Fig. 3, an "on" diode was approximated as an ideal switch with zero series resistance; and an "off" diode was approximated as a 5 pf capacitor. The output capacitance of the tube plus the stray and trimming capacitance was assumed to be 20 pf. By using Z0 as a variable parameter, its effect on the diode locations can be readily seen. For the stated assumptions, Z0 must be limited to a maximum of about 3012 to prevent the diode locations from being too close physically for practical diode packages. 6. REASON FOR ASSUMING DEFINITE CT VALUE If the total plate circuit capacitance is other than the assumed 20 pf value, the change in center frequency corresponding to a change in equivalent inductance will vary accordingly. The relationship between the change in center frequency resulting from a change in inductance has been found to be: Af -2r2f 3 C AL 7

o 0'-4 0 d oo o 6 ci:~ ~ 0 0~ C0.s 0 oo I. o'"'0'-4 6 a a'-4 6 0 0 co 0c~a 0M d (oas/oNA) Xouonborj Ou uosoc C9 C ~ N ~ o -

0 0 6C o 6 ) 000 0.4-4 IIIIII I0 6 C) UC ts IO0 / Q C)0 CC "0 C) 6 04 Un "4- 1 1 -~-o. 044 0 s+-> II 0I ^ I o0" 0 0 0h. 0 0 0 60'3 UO(/I) n u Cos C O / /) 2 CL / / / I s~~~~~~~~~b.^r^.^^ / ^^ o~~~~~~~G - I ----- I ----- I ---------- I ----- I ----- ~ 0 0 0 0 0 0~~~~ Ifl 0 Irt 0 LO (?0 (M CM i-l T-I~~~~~~~C (oa/oA[ Aueba^:^ieusa 9Nr

where f0 = center frequency Af0 = change in center frequency CT = total plate circuit capacitance AL = change in equivalent inductance CT can be adjusted to a predetermined value with trimming capacitance so that center frequencies and changes in center frequencies correspond to calculated values. 7. MULTIPLE RESONANCES OF PLATE CIRCUIT The use of a shorted transmission line to simulate an ideal inductor introduces additional problems into the design of the plate tuning. For an ideal transmission line, the plate circuit will be resonant for those frequencies at which: 1 27mrf 2rfC Zo tan x T In general, if this transcendental equation is satisfied at a frequency f0 and the line length x is on the order of a quarter wavelength; it will be satisfied at higher frequencies approximately equal to 3f, 5f0, 7f0, etc. This multiple resonance phenomena may be especially troublesome in a Class C amplifier since the plate current pulse is rich in harmonics and only the fundamental voltage component is desired. The multiple resonances of the plate circuit would allow not only the fundamental voltage component but also all odd harmonics of it. This would distort the output waveform and could reduce both the fundamental output power and the efficiency of the amplifier. Perhaps it will be desirable to provide some tuning of the output matching network to prevent the dissipation of power in the load at the harmonic frequencies. 10

8. BANDWIDTH LIMITATIONS Another problem resulting from tuning the plate circuit with a transmission line instead of an inductor is one of reduced instantaneous bandwidths. For example, the bandwidth at a simple resonance pole is given by: f0 BW Q where 0 fO = 2- = resonance frequency 1 dB Q =2 o0Zw d- co (Ref. 2) Z = Z(o) = network impedance at 0 C = o0 resonance B = total network susceptance dB dB do- dw = slope of network suscep0 co = W0 tance at resonance For the network shown in Fig. 4, the total susceptance is given by: B = oC - coL Also at resonance w0: L = 2, Z = R 0 coC o00 From these relationships the resulting bandwidth is determined to be: BW 2rRC Now if the inductor is replaced with a lossless short circuited transmission line (Fig. 5), the expression for total network susceptance is changed. B = coC - 0 cot i- x 11

where Z0 = characteristic impedance of transmission line x = physical length of transmission line u = velocity of propagation in transmission line Now for resonance at w0, we require: 1 Oo w C = - cot - x 0 Z 0 cot - x Z C dB The expression for dB becomes dB x 2co dco = C + - csc2 ~ x'dw O uZ 1 and since 2 W0 O0 CC x- = 1 + cot2 x we can write ddB = C + XZ [1 + (Z C)2 dwo0 UZ0 00 Thus, for the case of transmission line tuning, the bandwidth may be expressed as: 1/= BW = 17 Rx RC + [1 + (Z )2 ] The above expression indicates that the bandwidth is not limited simply by the RC product as it was in the case involving the inductor tuning. In fact, in the range of interest, the denominator term Rx is larger than the RC term. As a result, bandwidths associated with transmission line tuning will be narrower than those associated with inductor tuning. 12

B -— >- RI C L Fig. 4. Inductor tuning. o — -----— CI R' CT B ----- R Co - B cot x line Z- - ot Fig. 5. Transmission line tuning. 13

9. EXPERIMENTAL TRANSMISSION LINE TUNING To verify the computer solution for the switch locations along the transmission line, and to study the multiple resonance effect and bandwidth limitation effect; an experimental stripline was constructed. In consideration of the computer solutions, an arbitrary characteristic impedance of 102 was selected. The transmission line for the experimental measurements was constructed in the form of a parallel plane stripline (Fig. 6). The Z of such a parallel plane transmission line is given by z h 2 h << w 0 w E w = width of narrower plane h = distance between planes The experimental line was constructed using two planes, one 9" wide 1it and the other 10" wide, with 4 air spacing between the planes. The line, with these dimensions, had a Z0 of about 1012. After determining the physical size of the line, a convenient method for approximating the PIN diodes was selected. As was previously mentioned, the manufacturer's specifications indicate that the diodes have total reverse bias capacity of 5 pf and a forward bias series resistance of about 0. 5Q2. These characteristics were simulated by using a switched parallel RC network as illustrated in Fig. 7. Measurements for reverse bias conditions were conducted with 5 pf capacitors connected between the top and bottom planes of the stripline. The capacitors, used to simulate the diode reverse biased conditions, were constructed using -5 inch diameter brass rods and 0. 0075 inch thick teflon sheets. The brass rods were mounted on the upper 14

o 0 0'c z co. I a)^<u~~~3 a) U. - U C^~0-l a).a~~ ~~,=~~ )a oLo~~~~~ b E\~~~ ~ -4- 0 a) U

plane and separated from the lower plane by the teflon sheets. This structure yielded capacitances of 4. 6 pf across the transmission line. The 0. 512 resistance required for simulating a forward biased diode was difficult to obtain with high frequency resistors. For that reason, the series resistance was omitted and a forward biased diode was simulated by short circuiting the transmission line planes with a wire. These shorting wires were placed in the center of the line, alongside of their corresponding capacitors. The wires were soldered directly to the lower plane; extended through the upper plane; and then, soldered to lugs on the outside surface of the upper plane (Fig. 6). Preliminary measurements indicated that such a single wire in the center of a wide transmission line did not produce an adequate short. Therefore, the shorting wires were replaced with shorting bars that could be extended across the entire width of the transmission line. Hence, the shorting bar method was used in obtaining the experimental data. In the actual stripline, it will be important to reduce the line width so that a diode will produce an adequate RF short. Eleven of these diode equivalent networks were placed at equal intervals along the line so that the longest and shortest line lengths would resonate the assumed tube output capacity of 5 pf at 70 Mc and 290 Me, respectively. Measurements were made to determine the resonance frequencies and the bandwidths as a function of the simulated "on" diode positions. The experimental measurement method is shown in Fig. 8. The results of these measurements are shown in Figs. 9, 10, 11, and 12. 16

Cp "R s Fig. 7. Experimental PIN diode equivalent. Sweep Oscillator |Z 50 Experimental Transmission Line out 502 I 5 pf. * I Diode Equivalent Network Crystal C rystal' Oscilloscope Detector Z. = 50 in Fig. 8. Measurement circuit for experimental transmission line. 17

10. DISCUSSION OF EXPERIMENTAL RESULTS The experimental results in Figs. 9 and 10 are for the 1012 transmission line with equivalent diode capacitances (5 pf) placed at the location of each switch. Figures 11 and 12 are corresponding results for the case with no equivalent diode shunt capacitances being placed down the line (i. e., ideal diode case). Comparison of the two sets of curves indicates that the effect of the 5 pf shunt capacitance at each switch location was negligible for this case of 1012 characteristic impedance. Figure 9 illustrates the multiple resonances resulting from the periodic nature of the transmission line. Higher harmonic modes were not measured because they were outside the frequency range of the test equipment. It should be noted that the modes shown exhibit an approximately odd-harmonic relationship to the fundamental resonance curve. As discussed previously, the bandwidths shown in Fig. 10 are narrower than predicted by the simple relationship: BW = 1RC For example, consider the bandwidth measured for the fundamental resonance curve at a center frequency of 290 Mc. The equivalent resistance (R = 252) and the equivalent capacitance (CT = 5 pf) predict a 2RC bandwidth of 1. 28 Gc instead of the measured 124 Mc. However, if the previously developed bandwidth expression for transmission line tuning is applied: BW = RC + R [1 + (Z0c0)2 18

where R = 250 C = 5 pf x = 0. 239 meter u = 3 x 108 meter/sec Z = 102 0 cO = 2Xr(290 Mc) = 1. 82 x 109 rad/sec Then a bandwidth of 138 Mc is predicted. This predicted bandwidth agrees reasonably well with the experimentally measured bandwidth of 124 Mc. The experimental curve of Fig. 10 also indicates an approximate four-to-one instantaneous bandwidth variation as the center frequency is tuned across the desired band. This information can be used to improve the assumption of a constant 20 Mc bandwidth which was used in the original computer solutions. 19

500 z = 10 C\ \ = 5pf Cd = 5 pf 450 400 52g Mode 4 350 0 250 \ 300 - 32g 2 \ 150 \, ^ Fundamental - Mode 100 50[ 0 0.2 0.4 0.6 0.8 1.0 1.2 Distance To Short - x - Meters Fig. 9. Experimental transmission line resonance frequencies. 20

-23E e bb " t. \| \ ~ \ C \ v \ O10 o' ~1 b.0 (, C,4D 0\ 0 0 0 0 0 I \ I I I o 0 0 ~~~~~~~~~~21 ~ ~ N 0 0 3as /D w1| / t IW!MpU

CV 1"rO ~o / I NO 0 -o / o/ | / /o - /v 3 / 4.) o o /2^~~~~~~C. /0 & o - /CO C d'-4 C/!! LaI I CNI' -4; 22 ^ ^ CO CM CM T-~ ^-1 ~ 22~~~~~~~~~~~~~~~~~~~r

11. PROPOSED AMPLIFIER CONFIGURATION The configuration proposed, at present, for realizing the wideband amplifier is shown in Fig. 13. The transmission line will be realized in enclosed stripline form. The diodes will be embedded in the stripline and biased via feed through capacitors as illustrated in Fig. 14. The use of a dielectric filled stripline with a center conductor enclosed between two ground planes allows the width of the center conductor to be reduced while maintaining a low characteristic line impedance. As a typical example, a 202 characteristic impedance can be realized with a center conductor width of 0. 6 inch if the distance between ground planes is 0. 25 inch and the relative dielectric constant is about 2. 3. Thus, effective RF shorts should be produced by diodes placed along the 0. 6 inch center conductor. The plate voltage supply may be introduced through the center conductor of the stripline, thus, eliminating the need for an RF choke at the plate. This also eliminates the self resonance problems of such a choke. The current source is necessary for supplying the 100 ma forward bias to the "on" PIN diode and Edd will be on the order dd of 5 volts. Hence, the diode bias power required will be approximately one-half watt. The electronic circuitry for performing the switching logic and driver section has not been developed and will be considered only after most other details of the amplifier have been determined. The input and output matching networks have not been developed, but it is anticipated that they will employ wideband transformers realized in the form of strip-transmission lines. 23

a) 0 Ca 0 o OJ \ \ = m \ o d o 0 \ 0 -0 a) a) o 0O IQ) Lo oL O 11 1\1 1 \ o P4o\ \ br o o o o o y ~ ) o \ \ o o e 3 \q 9- o I ooIC:tC S / 0 C 244 o o I o4 a4 "a ^ ^s. o s /zi,'i u \ qpio puS O V3 O Ln~~~~~~~~~~~~S 0 00 00 0 0 0 0 CO C~~~l r"~ O OS OO C- ^0 10'<^ OO CS| T-(~~~~~~~G I-^ r^ I -^ ro~~~~~~~~ /^I ui o pbpr NVU~~~~~~ o~2

a) Q 0 D u c Q + - oUa), I' —-I ^A r~~~~~r-~~r r-4 C ~; __ (/__ 3_ ^ I ~ I -i I- i )-| II H lI L______\ Du -— ~ LO s3 i1L2 I1 a) A~e 1' Iti 25 ll1?~~~~~ ~~' ^S ^ s L-._ _ _-_J I~~~~~F ~" 0~~2

0 - d 0 0 II ~I 0 I I 0 ^'~ I I I I 7 i i~l a I 0 I I I _ \ I I 26

APPENDIX ESTIMATE OF THE SHORTING SWITCH REQUIREMENTS For "off" voltage requirements consider the following circuit: + Ebb 2L E2 p o / ~E e RI0 1 pC RL S S e, C-T R1e if"__i n-l n -- ~i -I1 For Class C operation neglecting harmonics: ep Ebb (Ebb Ec2) sin wCt Typical design values for a 4X250B tube are: Ebb = 750 v E = 250v 27 27

Giving e = 750 + 500 sin coot For the case of S1 through Sn 1 being open and Sn closed (lowest frequency): e = 500 sin c0t and the voltage across S. is: e o O 7T e. = cos (x - x ) + S 1X v 1x n 2 7T n cos 2 - - The voltage across switch S1 is approximately: e e cos- x < e sl o u0 - o Thus, a peak to peak voltage swing of about 1000 volts appears acrossS1. Also e S > e 2 > > and Shas the most 1S sl s2 S'> n- 1 severe voltage requirement. For the current requirements: assume RL' 1K and that CT can be adjusted to give the desired bandwidth. e 2 0 P = 125 W L RL' e I R =, - 0.354 amp RMS RL RL Now at resonance: e _IRLRL' IL c= X - XL ~ Q IRT c L L 28

To determine the maximum Q to be encountered in meeting a bandwidth requirement of 20 Mc consider: fo 290 Mc BW max 20 Mc 14.5 Now IL, the maximum current flowing into the terminals of the transmission line becomes: IL Q IRL = (14. 5) (0.354) = 5.13 ampRMS The current through a shorted switch (Is) at distance x from the input terminals to the line is related to the input current by the expression: IL I = s co cos - x Now if x is less than q radians Is < 1. 4 = 7. 18 amp RMS s L 29 29

REFERENCES 1 B. F. Barton et al., Countermeasures Research, Quarterly Progress Report No. 2, Contract No. DA-36-039 AMC-03733(E), USAEMA, Cooley Electronics Laboratory, The University of Michigan, Ann Arbor, Michigan, July 1964, pp. 2-4 (SECRET). 2. D. F. Tuttle, Jr., Network Synthesis, Vol. I, Wiley, New York, 1958, p. 717. 30

UNCLASSIFIED DISTRIBUTION LIST Copy Copy No. No. 1-2 Commanding Officer 11 Headquarters U.S. Army Electronics Command USAF U. S. Army Electronics Laboratories Washington 25, D. C. Fort Monmouth, New Jersey ATTN: AFRDR ATTN: Senior Scientist Electronic Warfare Division 12 AFAL (AVWW/ECM Technology) 3 Commanding General Wright-Patterson Air Force Base U. S. Army Electronic Proving Ground Ohio 45433 Fort Huachuca, Arizona ATTN: Director 13 Commander Electronic Warfare Division Aeronautical Systems Division Wright-Patterson Air Force Base 4 Commanding General Ohio 45433 U. S. Army Materiel Command ATTN: ASAPRD Bldg. T-7 Washington 25, D. C. 14 Commander ATTN: AMCRD-DE-E Aeronautical Systems Division Wright-Patterson Air Force Base 5 Commanding General Ohio 45433 U. S. Army Material Command ATTN: ASNP Bldg. T-7 Washington 25, D. C. 15 ESD (ESTI) ATTN: AMCRD- RP-E L. G. Hanscom Field Bedford, Massachusetts 6 Commanding Officer Signal Corps Electronics 16 Commander Research Unit Rome Air Development Center 9560th USASRU Griffiss Air Force Base P. 0. Box 205 New York Mountain View, California ATTN: RAYLD 7 U. S. Atomic Energy Commission 17 Commander 1901 Constitution Avenue, N. W. Air Proving Ground Center Washington 25, D. C. Eglin Air Force Base ATTN: Chief Librarian Florida ATTN: ADJ/Technical Report 8 Director Branch Central Intelligence Agency 2430 E Street, N. W. 18 Chief of Naval Operations Washington 25, D. C. EW Systems Branch OP-35, Department of the Navy 9 U. S. Army Research Liaison Officer Washington 25, D. C. MIT- Lincoln Laboratory Lexington 73, Massachusetts 19 Chief, Bureau of Ships Code 691C 10 Commander Department of the Navy Air Force Systems Command Washington 25, D. C. Andrews Air Force Base Washington, D. C. 31 UNCLASSIFIED

UNCLASSIFIED DISTRIBUTION LIST (Cont.) Copy Copy No. No. 20 Commander 32 President Bu Naval Weapons Code RRRE-20 U. S. Army Airborne and Department of the Navy Electronics Board Washington 25, D. C. Fort Bragg, North Carolina 21 Commander 33 U. S. Army Anti-Aircraft Artillery Naval Ordnance Test Station and Guided Missile School Inyokern Fort Bliss, Texas China Lake, California ATTN: ORL ATTN: Test Director - Code 20 34 Commander 22 Commander USAF Security Service Naval Air Missile Test Center San Antonio, Texas Point Mugu, California ATTN: CLR 23 Director 35 Chief of Naval Research Department of the Navy Naval Research Laboratory Department of the Navy Countermeasures Branch ashington 25, D. C. Code 5430 ATTN: Code 427 Washington 25, D. C. 36 Commanding Officer 52d U. S. Army Security Agency 24 Director Special Operations Command Naval Research Laboratory Fort Huachuca, Arizona Washington 25, D. C. ATTN: Code 2021 37 President U. S. Army Security Agency Board 25 Director Arlington Hall Station Air University Library Arlington 12, Virginia Maxwell Air Force Base Alabama 38 The Research Analysis Corporation ATTN: CR-4987 McLean, Virginia 22101 ATTN: Document Control Officer 26 Commanding Officer-Director U. S. Navy Electronics Laboratory 39-48 Headquarters San Diego 52, California Defense Documentation Center Cameron Station 27 Commanding Officer Alexandria, Virginia U. S. Naval Ordnance Laboratory Silver Spring 19, Maryland 49 Commanding Officer U. S. Army Electronics Research 28-30 Chief and Development Laboratory U.S. Army Security Agency Fort Monmouth, New Jersey Arlington Hall Station ATTN: U. S. Marine Corps Liaison Arlington 12, Virginia 22212 Office, Code: SIGRA/SL-LNR ATTN: 2 Cpys - IADEV 1 Copy - EW Div. IATOP 50 Director Fort Monmouth Office 31 President Communications-Electronics Combat U. S. Army Defense Board Developments Agency Headquarters Building 410 Fort Bliss, Texas Fort Monmouth, New Jersey 32 UNCLASSIFIED

UNCLASSIFIED DISTRIBUTION LIST (Cont.) Copy Copy No. No. 51- 6 5 Commanding Officer 68 U. S. A. F. Project Rand U. S. Army Electronics Command The Rand Corporation U. S. Army Electronics Laboratories 1700 Main Street Fort Monmouth, New Jersey Santa Monica, California ATTN: AMS EL-RD-DR AMSEL-RD-NSR 69 Stanford Electronics Laboratories AMSEL-RD-SM Stanford University AMSEL-RD-SA Stanford, California AMS EL-RD-SEA AMSEL-RD-SEJ 70 Director AMSEL-RD-SES National Security Agency AMSEL-RD-SEE Fort George G. Meade, Maryland AMSEL-RD-ADO ATTN: RADE-1 AMS EL-RD-SR AMSEL-RD-SE 71 Bureau of Naval Weapons AMSEL-RD-ADT Representative AMSEL-RD-GFR Lockheed Missiles and Space AMSEL-RD-PRM Company AMSEL-RD-RHA P. 0. Box 504 Sunnyvale, California 66 Commanding Officer U. S. Army Signal Missile Support 72 Dr. T. W. Butler, Jr., Director Agency Cooley Electronics Laboratory White Sands Missile Range The University of Michigan White Sands, New Mexico Ann Arbor, Michigan ATTN: SIGWS-MEW 73-83 Cooley Electronics Laboratory 67 Commanding Officer The University of Michigan U. S. Naval Air Development Center Ann Arbor, Michigan Johnsville, Pennsylvania ATTN: Naval Air Development Center Library Above distribution is effected by Electronic Warfare Division, Surveillance Department, USAEL, Evans Area, Belmar, New Jersey. For further information contact Mr. I. 0. Myers, Senior Scientist, Telephone 59-61252. 33 UNCLASSIFIED

UNIVERSITY OF MICHIGAN I3 901i5 02523 0957