Technical Report No. 137 4853-12-T THE MICHIGAN AUTOMATIC DIRECTION FINDER by D. L. Mills Approved By: @ go B. F. Barton for COOLEY ELECTRONICS LABORATORY Department of Electrical Engineering The University of Michigan Ann Arbor Contract No. DA 36-039 sc-89227 Signal Corps, Department of the Army Department of the Army Project No. 3A 99-06-001-01 February 1964

TABLE OF CONTENTS List of Illustrations v Abstract ix Foreword xi 1. INTRODUCTION 1 2. MICHIGAN AUTOMATIC DIRECTION FINDER SYSTEM DESCRIPTION 1 2.1, AN/TRD-4A DF Set 2 2.2. MADFI 2 3. RECEIVING SYSTEM 7 3.1. Operation of Antennas and Goniometers 7 3.2. Operation of Receiver and Detector 8 3.3. Receiving System Instrumentation 10 4. PROCESSING SYSTEM 11 4.1. Narrowband Filter Operation 13 4.2. Narrowband Filter Instrumentation 13 4.3. Bearing Computer Operation 15 4.3.1. Model Bearing Computer 15 4.3.2. MADFI Bearing Computer 17 4.4. MADFI Bearing Computer Instrumentation 22 4.5. Synchronizing Generator 23 5. DISPLAY AND RECORDING SYSTEM 26 5.1. Cathode-Ray Display Operation 26 5.2~ Modifications and Additions to Cathode-Ray Display 27 5~3. Pen Recorder 30 5.4. Decimal Display 30 5.4.1. Count-Control Circuitry 32 5.4.2. Zero-Crossing Detector 32 5.4.3. Binary-Coded Decimal Counter 33 5.5. Paper Tape Punch 33 6. FIELD MEASUREMENTS 34 7. SUGGESTIONS FOR FURTHER RESEARCH 37 7.1o Collection of Bearing Statistics 37 7~2. MADFI System Improvements 37 8. APPENDIX-OPERATION NOTES AND CIRCUIT INFORMATION 39 9. REFERENCES 72 111

LIST OF ILLUSTRATIONS Figure Description Page 1 MADFI System Block Diagram 3 2 MADFI Physical Layout - Front View (Case Covers Removed) 5 3 Patch Panel Connections 6 4 Receiving System 6 5 Adcock Antenna Octantal Error Curves 9 6 Phase Equalization Network 12 7 AN/TRD-4A Detector Modifications 12 8 Narrowband Filter (Simplified) 14 9 Model Bearing Computer 14 10 MADFI Bearing Computer (Simplified) 18 11 MADFI Bearing Computer Waveforms 19 12 Synchronizing Generator Waveforms 24 13 Synchronizing Generator (Simplified) 25 14 Manual Sense Operation of AN/TRD-4A 25 15 Automatic Sense Operation of MADFI 28 16 AN/TRD-4A Indicator Modifications 29 17 Automatic Sense Circuit (Simplified) 29 18 Decimal Display (Simplified) 31 19 Zero-Crossing Detector (Simplified) 31 20 Paper Tape Punch 35 21 MADFI Deviation — Indicated Bearing Curves 36 22 System Logical Diagram 40 23 System Connector Diagram 41 24 Synchronizing Generator Connector Placement 42 25 Analog Case Module and Connector Placement 42 26 Digital Case Module and Connector Placement 43 27 Connector Panel Connector Placement 44 28 Patching Panel Connector Placement 44 29 Control Unit Control and Connector Placement 45 30 Analog Case Test Panel 46 31 Digital Case Test Panel 47 32 Analog Case Logic 48,49 v

Figure Page 33 Digital Case Logic - Count Control 50, 51 34 Digital Case Logic - BCD Counter 51, 52 35 Power Supply DEC 722 53 36 BCD Light Driver DEC 1671 (Digital Modules M4, M6, MB) 53 37 Inverter DEC 4102 (Digital Modules M3, M10) 54 38 Inverter DEC 4105 (Digital Module M11) 54 39 Dual Flip-Flop DEC 4209 (Digital Module M12) 55 40 4-Bit Counter DEC 4215 (Digital Modules M5, M7, Mg) 55 41 Integrating One-Shot DEC 4303 (Digital Module M13) 56 42 Pulse Generator DEC 4410 (Digital Modules M14, M15, M16) 56 43 Pulse Amplifier Module - DEC 4604 (Digital Module M2) 57 44 Indicator Driver Module DEC 4689 (Digital Module Ml) 57 45 Zero- Crossing Detector Module (Digital Module M19) 58 46 Count Multiplier Module (Digital Module M17) 59 47 Chopper Module (Analog Module M1, M2) 60 48 Comb Filter Module (Analog Module M6) 61 49 Narrowband Filter Module (Analog Module M9) 62 50 DC Amplifier Module (Analog Modules M3, M4, M7, M10; Digital Module M18) 63 51 Regulated Power Supply Module (Analog Module M5) 64 52 Auto-Sense Module (Analog Module M8) 64 53 Synchronizing Generator 65 54 Control Unit 66 55 Control Unit - Decimal Readout 67 vi

Figure Page 56 Paper Tape Punch 68 57 Paper Tape Punch - Solenoid Drivers 69 58 Paper Tape Punch - Pulse Generators 70 59 Paper Tape Punch - Power Supply 71 60 X - Y Recorder Pen Pecker Amplifier 71 vii

ABSTRACT This report describes the theoretical basis and practical instrumentation of the Michigan Automatic Direction Finder (MADFI). The MADFI is an integrated data collection, processing, and display system for determination of bearing statistics of signals in the high-frequency radio spectrum and is intended as a research tool in the study of propagation phenomena and direction-finding techniques. Signal processing includes narrowband postdetection filtering and a type of linearly-weighted time averaging which improves the bearing estimates in the presence of multipath and noise processes. The MADFI is equipped with several types of readout suitable for telemetry transmission and computer processing of bearing statistics.

FOREWORD The effort described in this report represents significant contributions by a number of individuals, each of whom has made his addition by expanding or specializing the body of theory inherited from his predecessors and leaving in turn a more profuse collection of reports, memos, and notations. Such a collection has made possible this report on the Michigan Automatic Direction Finder (MADFI). In this vein, I would like to acknowledge the invaluable assistance received through notes left by D. S. Heim and W. J. Lindsay, former Project Engineers on projects involving direction-finding research, and in particular to L. W. Orr, whose unpublished notes furnished the basis for a sizeable amount of this report. I also wish to credit R. E. Lillie, G. Ziedins, and K. E. Burkhalter, engineers who developed much of the hardware of the MADFI, and in particular, G. A. Hellwarth, whose technical advice "unstuck many an ornery circuit."

1. INTRODUCTION The effort narrated in this report is a continuation of that initiated by W. J. Lindsay and D. S. Heim and described in Reference 7. The effort is directed toward the realization of a semi-automatic direction finder (df) suitable for collection of bearing statistics of signals in the high-frequency radio spectrum. Of considerable importance is the introduction of signal processing which simulates the behavior of a trained operator in stabilizing an indication distorted by multipath processes. As a result of this effort the Michigan Automatic Direction Finder (MADFI) was constructed. The MADFI is an integrated data collection, processing, and display system for determination of bearing statistics of radio signals in the 2-30-mcs range. The system consists of a modified AN/TRD-4A Tactical Direction Finder Set, laboratory-constructed analog and digital computing-type components, and a variety of display devices. It is the purpose of this report to describe the theoretical basis and practical instrumentation of the MADFI. Since others have in some areas developed and published the theory (cf. References) responsible for the realization of some of the components, only those developments pertinent to the operation of the MADFI as a system are included. Accompanying these developments are descriptions of the AN/TRD-4A modifications and the additional laboratory-constructed equipment. Due to time restrictions on this effort the MADFI has not undergone extensive field tests. For this reason little operational experience has been gained with the system, although preliminary tests indicate a significant improvement in the MADFI performance over that of the unmodified AN/TRD-4A equipment. 2. MICHIGAN AUTOMATIC DIRECTION FINDER SYSTEM DESCRIPTION The MADFI consists of three basic subsystems: (1) The receiving system, which includes the antennas, goniometer, receiver and detector; (2) The processing system, which includes the filtering, computing, and synchronizing elements; and (3) The display and recording system, which includes the cathode-ray and digital displays, a pen recorder, and a paper punch. The three subsystems will be described in detail in Sections 3 through 5. The sections immediately following describe the organization of these subsystems as a whole. 1

2.1 AN/TRD-4A DF Set The AN/TRD-4A, around which the system is designed, is a spinning-goniometer type direction finder utilizing a four-element Adcock antenna and a cathode-ray tube display. Its antenna, classified as a medium aperture system, may be of two dimensions: one for 2-10 mcs, the other for 10-30 mcs. Techniques of the MADFI are by no means restricted to this particular df set, however, and any df set which produces a sine wave signal of amplitude and phase corresponding to the amplitude and bearing of the incident signal is suitable for MADFI-type processing. Examples of these types of df sets include doppler and other commutated-antenna systems as well as spinning-antenna systems. The twin-channel system is applicable with some slight changes. The reader is directed to pertinent publications concerning the operational characteristics of the AN/TRD-4A, including the R-390/URR Radio Receiving Set, and is presumed familiar with them (References 8 and 9). However, the casual reader will find an essential description of the operation of these components in this report. 2.2 MADFI Figure 1 will be helpful in describing the system operation of the MADFI. As seen, there are two modes of operation: (1) The manual mode, which is that of the essentially unmodified AN/TRD-4A, and (2) The automatic mode, which includes the processing and display equipment of the MADFI. The two modes are selected by a single switch, enabling rapid comparisons. The bulk of the antenna, goniometer, and receiving components of the MADFI is adopted with few changes from the corresponding AN/TRD-4A components. The two most significant changes involve the addition of a sense antenna at the center of the Adcock array and a redesigned receiver detector. These components will be described further in Section 3. The processing system includes the narrowband filter and bearing computer. These elements, which will be described in Section 4, help to reduce bearing error due to multipath processes. A synchronizing generator utilizing a code wheel coupled to the goniometer rotor is used with the bearing computer and digital display elements. Meters monitor important signal levels. The display and recording system, described in detail in Section 5, includes both analog and digital real-time displays together with corresponding analog and digital peripheral recording devices. The real-time displays consist of the AN/TRD-4A cathode-ray indicator 2

OUTPUT METERM - -l1 PAPER I ______________ ~PUNCH /CHOD~ SYNCHRONIZATION I GITAL PUNCH -HEE - GENERATOR DISPLAY L J ANTENNA IU —NSGONIOMETER / METER PEN NSMETER RECORDER RECEIVER I.NARROWBAND BEARING FULL-WAVE E RECEIVER FILTER COMPUTER RECTIFIER SENSE l AUTOMATIC C..3T SENSE ANU!VMANUAL f d\'PHASE \ AUTOMATIC NETWORK SENSE F- 1 / F....~ l72 - kcs NET W ORK I \ I, OSCILLATOR CRT HV O INDICATOR CRT BLANKING GONIOMETER Fig. 1 MsADFI Systu- Bl.ock Diagra!i

and a three-digit decimal display reading directly in degrees. An automatic sense circuit eliminates the 1800-bearing ambiguity inherent in the AN/TRD-4A. The peripheral recording devices consist of an eight-channel paper tape punch and a pen recorder. (The full-wave rectifier, modulator, 72-kcs oscillator, indicator goniometer, and cathode-ray tube elements of Figure 1 are all part of the modified AN/TRD-4A indicator.) A photograph of the MADFI equipment is shown in Figure 2. Four of the five antenna elements, (not shown) the goniometer, the receiver, and the cathode-ray display are part of the AN/TRD-4A; the rest of the components are laboratory-constructed or modified to varying degrees. Except for the antennas the entire system is contained in two equipment racks. One five-foot rack contains the modified AN/TRD-4A DF Set and the processing and real-time display equipment. A smaller three-foot rack contains the peripheral recording devices. Except for the original AN/TRD-4A components and the Moseley pen recorder, the entire system is constructed with solid-state modularized circuitry. Most of the component modules are visible in the two racks illustrated in Figure 2, the analog rack and the digital rack. Meters and test jacks are provided for system adjustment and servicing, and testing circuitry is in some instances an integral part of the equipment. The patch panel shown in Figure 2 is provided so that individual processing, display, and recording elements may be interconnected for demonstration and evaluation procedures. Its connections are shown in Figure 3. Standard phone jacks are provided with a normally-closed contact which is disconnected when a simple patch cord is inserted. These contacts are wired so that when all patch cords are removed the system is interconnected as in Figure 1. To allow the greatest flexibility when the MADFI components are interconnected in other than the normal manner shown in Figure 1, or when external devices are connected, the patch panel voltage and impedance levels are standardized. All circuits are designed to be linear to twice the operating level of 1-volt rms and to maintain an indicated bearing accuracy of ~ 1/20 under extreme conditions of circuit loading. Records made with this equipment may be processed by either analog or digital generalpurpose computers and may be used to evaluate system performance, to aid in refining processing techniques, and to study ionospheric propagation phenomena. Together with a remote-controlled df receiver, (Reference 1) and a telemetering link, the MADFI is suitable for installation at an unattended field site. 4

MADFI Photodiode Holder Equipment Meters (Part of Sync Gen) Digital Display Modified AN/Trd-4A Gonio Drive:iiiii!: piii i Peripheral........______... __ Recording Equipment Control Unit { ^ r~84~~ j I1IdMosley Model L Modified I t ~ X-Y Pen Recorder Modified AN/TDR-4A Indicator........ r R-390/URR ~ iiiiiii|i g 0 >84XPaper Tape Punch Receiver: -i!~?,ii',1"''!iii i',i'DX...'................ 11 Patch Panel Patch Cord Analog Case Digital Case {i Test Panels Connector B.......... Panel Power Supply......i. -..._ FIG. 2. MADFI PHYSICAL LAYOUT- FRONT VIEW (Case Covers Removed)

MODIFIED NARROWBAND BEARING AUTOMATIC DIGITAL AN/TRD4-A FILTER COMP SENSE DISPLAY DET. IND IN OUT IN OUT IN IN NC NC NC NC NC PATCHING ( I ) 2 0 3 0 4 0 5 ( 6 0 7 ( 8 9 10 11 12 JACKS (ALL LEVELS w z z I-V RMS) H Z -z a- HE w wD w w O~ W wH 0:O o- ) ) a- a a- aH o 5o H a. (L a. W - O M D: o z z m m _ Fig. 3 Patch Panel Connections NORTH DIRECTION OF SIGNAL ARRIVAL NORTH REF ~N eN ROTOR m ~70)~~~~ ~~~POSITION ADCOCK ANTENNA MONOPOLES GONIOMETER eg E d d w SENSE ___ eE_ ee RECEIVER BEARING AND AM o SIGNAL rw DETECTOR OUTPUT ec sENSE ep PHASE Fig. 4 Receiving System

3. RECEIVING SYSTEM The receiving system used in the MADFI is diagrammed in Figure 4. The antennas consist of a four-element Adcock array together with a fifth centrally-located element used as a sense antenna. The antenna goniometer and sense-phase network act upon the element voltages to produce an amplitude-modulated signal er which is amplified and detected by the receiver. Phase information contained in the detector output voltage ed is used in the processing and display systems to reconstruct the direction of signal arrival 8. The following analysis indicates how this is done and how the dimensions of the antenna system and adjustments of the sense-phase network affect the system accuracy. 3.1 Operation of Antennas and Goniometers The central sense element will be assumed the reference element of the antenna array. Relative to a cw signal of frequency aw arriving from bearing 0 relative to true north, the voltage induced in the central element is ec= A cos w t, (3.1) C C c while the differential voltages induced in the north-south and east-west pairs of elements are eN e = 2A sin(- cos 0) sin wt, and (3.2) eE eW= 2A sin ( Asin 0) sin w tc If the goniometer rotor angular position y is measured as in Figure 4, then the rotor voltage is e = K[(e - e) cos o + (e - e) sin ], (3.3) =[N E e) where K is the transmission coefficient of the goniometer. This voltage passes through a null where eN -eS sin ( cos ) tan o e =E e- i (3.4) eE eW sin (-d sinO C. E. Lindahl has analyzed K and other losses in the antenna system. See Reference 5. 7

which occurs twice in every revolution of the rotor. Thus the angle Sp at null may be used as 7Td an indication of the bearing 0 if -f- is sufficiently small; for in that case tan 0 A tan (?p ~ r/2). (3.5) The error AO in o = 0 + AO is a function of the wavelength and the geometry of the antenna system, and is called the octantal error. A plot of A8 as a function of d/X is shown in Fig. 5, which indicates that for values of d/X less than about 1/3 (the maximum value permitted in the MADFI) the octantal error is less than 3~ 2 The unmodified AN/TRD-4A operates in the fashion described above, the null points indicated in Eq. (3.5) becoming the tips of the propeller-like pattern displayed on the cathoderay indicator. The ~7r/2 ambiguity apparent in Eq. (3.5) is resolved in the MADFI equipment by summing the rotor voltage and a voltage of appropriate phase and amplitude derived from the sense antenna. Letting the transfer constant of the sense-phase network be ae, this voltage is ep =aA cos (wct + f). (3.6) The goniometer rotor is spinning at a frequency o. If the octantal error is neglected, then the rotor voltage is, from Eq. (3.3),,rd e = 2KA X cos (wt - ) sin o t. (3.7) 3.2 Operation of Receiver and Detector The voltage e and e are summed in the input circuitry of the receiver: P g er =aA cos (coct +3) + 2KA cos (wgt -) sin wt. (3.8) The results of this analysis will not be affected if the amplitude dependence in Eq. (3.8) is 2KA Trd dropped by setting aA = 1 and = m, so that er becomes proportional to e' =cos (w t + ) + mcos (w t - 0) sin w t. (3.9) r C g c 2C. E. Lindahl has demonstrated a aantenna design that permits an error of less than 10 for values of d/X as high as unity.. See Reference 6. 8

25 d 3 X' w 2015' -- $ 10 o a:E -2 I I i 0 io 20 30 40 50 60 70 80 90 TRUE BEARING (DEG) Fig. 5 Adcock Antenna Octantal Error Curves

Following the receiver and a conventional envelope detector, the signal becomes proportional to 2 2 0)1/2 ed= [1 + 2msin sin(wt -) m sins (Wgt - 0) (3.10) It is evident that if 3, the phase shift introduced by the sense phase network, is set equal to ir/2, Eq. (3.10) may be further simplified. In this case, the detector voltage becomes proportional to et = 1 + m sin (gt - ). (3.11) If:0 may not be adjusted to exactly 7T/2, then, since Eq. (3.10) is in the form (1 + x) /, it may be expanded in a Maclaurin series. The expansion is valid for values of a and j chosen such that Ixl < 1. The result, after some rearrangement of terms, is m 2 2 e= 1 + m sin sin (w t - 0) + cos 3sin (gt - 0) + (3.12) d g 2 g where the implicit terms are all at third and higher harmonics of sin (wgt - 0). If the second and higher harmonics of sin (cogt - 0) in Eq. (3.12) are removed by a low-pass filter following the detector, then Eq. (3.12) becomes eI = 1 + m sin f sin (wgt - 0), (3.13) which is similar in form to Eq. (3.11). If the d-c term in either Eqs. (3.11) or (3.13) is removed by a high-pass filter and the goniometer rotor position y noted at the time of a positivegoing zero crossing, then it is clear that either Eqs. (3.11) or (3.13) will yield an unambiguous bearing indication So = 0. Furthermore, this indication does not vary over a range of the sense-phase network adjustments. In the MADFI equipment, the narrowband filter included in the processing system functions as both the high- and the low-pass filters. The display system extracts the bearing indication from the resultant signal in the manner described above. 3.3 Receiving System Instrumentation The MADFI receiving system consists of the AN/TRD-4A antenna and receiver components together with a sense antenna, sense-phase network, and the redesigned receiver detector. The sense antenna, located at the center of the AN/TRD-4A Adcock array, is constructed and connected in the same fashion as the other four array elements. The other antenna and 10

goniometer components of the AN/TRD-4A remain unchanged, except that the manual sense provision of the AN/TRD-4A is disabled in the automatic or MADFI mode. In the R390/URR receiver used with the AN/TRD-4A there is a 900-phase difference between the balanced antenna input and the unbalanced antenna input which is maintained up to about 10 mcs. Accordingly, if the goniometer is connected to the balanced input and the sense antenna is connected to the unbalanced input, then no sense-phase control is required over these frequencies. Above 10 mcs, however, phase equalization is required such as is provided by the simple delay line shown in Fig. 6, which consists of switch-selected lengths of coaxial cable. The increments, selectable at 1/20 per mcs, are adequate at frequencies up to 30 mc. The R390/URR receiver is not modified in any way. Predetection bandwidths of less than 2 kcs are desirable but cannot be obtained with the receiver i-f selectivity because of the excessively nonuniform phase characteristics introduced. An outboard i-f filter having phase characteristics uniform within 10 over a 500-cps bandwidth has been constructed, but is not incorporated in the MADFI due to the critical tuning problems associated with its use (Reference 4). No attempt has been made to utilize either the automatic gain control or the envelope detector circuitry of the receiver, and output to the external detector is at the i-f frequency of 455 kcs. The detector circuitry of the AN/TRD-4A is incorporated in the indicator chassis and consists of an i-f amplifier followed by a nonlinear detector. Considerable redesigning of these circuits was necessary in order to obtain linear operation over the dynamic ranges encountered under typical operating conditions. The redesigned detector, shown in Fig. 7, includes a phase-corrected high-pass filter to attenuate low-frequency transients generated by fluctuating carrier amplitudes. The low source-impedance amplifier included is capable of a 40-v output into the processing system. 4. PROCESSING SYSTEM The heart of the MADFI is its processing system, including the narrowband postdetection filter, the bearing computer, and the synchronizing generator. The narrowband filter is used to remove modulation products and noise from the bearing signal. The bearing computer is used to improve the bearing accuracy under conditions of severe multipath distortion. 11

PHASE SHIFT - DEGREES PER MEGACYCLE 0.5 1.0 2.0 4.0 8.0 16.0 72 n 1 1 1 72 f INPUT -- - -- OUTPUT, I.1 I, I~ I I Ii II I "-._ /' i,,' L LI. I \!I I1 \ -,'Li \ —JL2 \ \./ L4 3 L 5 /L6 L2: 1'10" L3- 3 8 LENGTHS OF Fig. 6 Phase Equalization Network L4=7 44 RG /59 U L5-148 L6 =29'4" B+ 275 V 12AT 7 20/i 2-IN270 150 v WI + - OUTPUT.0 5 _ _.05.02 IM INPUT IF AMP Imh PHASE.001 lOOk -1_ 2- lN270 220R.001 100k 2-IN270 5.6 MANUAL 5.6 MODE OUTPUT IOK 00I mOOh kA100K Fi. 7 N100k

The synchronizing generator supplies special signals to the bearing computer and display circuits. Since it is believed that the processing system offers significant performance features, its theory of operation will be expanded in considerable detail below. 4.1 Narrowband Filter Operation A narrowband filter of center frequency wo is necessary following the detector for the g reasons explained in Section 3.2. The filter also removes any modulation products outside its passband which may accompany the incoming signal. Busignies and Dishal (Ref. 2) have shown that in addition a signal-to-noise performance improvement is possible through the use of filters of this type. The improvement is a function of the ratio of the predetection to the postdetection bandwidth: (signal-to-noise improvement) = nar f ban filndwidth 1/4. \ narrowband filter bandwidth/, (4.1) which corresponds to an increase in signal-to-noise performance of 5 db per decade of bandwidth narrowing. The input to the narrowband filter is the detector output voltage ed described by Eq. (3.13). After passage through the filter, the bearing signal becomes ed =A sin (wt - ), (4.2) where A represents the various multiplicative factors implicit in the development. 4.2 Narrowband Filter Instrumentation The principal requirement of the narrowband filter, other than a good selectivity characteristic, is a high degree of linearity. This requirement is necessary for proper operation of the bearing computer and display circuits under conditions of extreme fading. A linear phase characteristic is also highly desirable so that variations in goniometer spin-frequency wo will not affect the bearing accuracy. The MADFI filter, which satisfies these requirements, is diagrammed in Fig. 8. This filter does not employ iron-core inductors, thus avoiding their inherent nonlinearities. The two filter sections shown have a bandwidth of about 30 cps. Using the formula of Busignies and Dishal (4.1) and noting that the prediction bandwidth of the system is 2 kcs, we see that this order of selectivity allows an increase of about 8 db in the signal-to-noise ratio. The amplitude limiter shown between the filter sections is included as an operational convenience. Its limiting threshold is set somewhat above normal operating system level in 13

INPUT NARROWBAND AMPLITUDE NARROWBAND I IOUTPUT 0 DC I-V RMS FILTER LIMITER FILTER 1D RMS AMP IV RMS IOK IK Fig. 8 Narrowband Filter (Simplified) sin ug t GENERATOR ed 1 4 0 0eb 2bAt3 4b cos wat GENERATOR Fig. 9 Model Bearing Computer 14

order that overloading of the remainder of the processing system by occasional large peaks in signal amplitude can be avoided. 4.3 Bearing Computer Operation In the preceding analysis the tacit assumption of a stationary amplitude and phase of the incident signal has been maintained. Under conditions of severe multipath distortion, this assumption must be abandoned, since signal amplitude variations of as much as 30 db and phase variations of a full 3600 are quite common in practice. The amplitude and phase variations of the incident signal may be represented through a time dependence of the A and 0 parameters of Eq. (4.2): ed = A(t) sin [wgt - 0(t)]. (4.3) g Now it is assumed that a good estimate of the true bearing 0 may be obtained from the expected value of 0. If this is correct, then a better estimate can be made by taking A into account. If it is agreed that for a multipath process the 0's obtained with correspondingly large A's are'better' than those obtained with correspondingly small A's, then an obvious type of weighted average computation would use the expected value of the product of 0 and some monotonic function of A, say f(A). An initial choice for f(A) is a linear relation of the form a E (A sin 0) tan 0 = E (A cos 0) (4.4) This process, called linearly-weighted time averaging, is carried out in the bearing computer. No claim is made as to its accuracy, but it appears to be a good initial choice from the impressive results seen on the display devices. Since the practical details of the MADFI bearing computer somewhat obscure its intended function, the operation of an ideal bearing computer will be outlined in the following section. Following this outline the operation of the actual system implemented in the MADFI will be described. The operation of the MADFI system, however, does not depart from that of the ideal system except in those details as will be explained. 4.3.1 Model Bearing Computer The ideal bearing computer is diagrammed in Fig. 9. The input to this system is normally the output from the narrowband filter ed described by Eq. (4.3). In this system two sinusoidal generators phase-synchronized to the goniometer spin-frequency produce outputs of sin Wgo t and cos owgt. Voltage ed is applied to two linear four-quadrant multipliers as shown in Figure 9. The outputs from these multipliers at positions (2) are 15

A A e2a = A sin (w t - 0) sin w t = cos 0 + cos (2w t - 0), (4.5a) e2b = Asin (gt- 0) cos 2 t = sin 0 +A2 sin (2wgt- ). (4.5b) The two integrators are permitted to operate on these voltages for an interval of time T. If their outputs are both zero at the beginning of the integration interval, the voltages at time T are T T e3a = 1/2 | A cos 0 dt - 1/2 | A cos (2wgt - 0) dt, (4.6a) 0 0 T T e3b = -1/2 A sin 0 dt + 1/2 f A sin (2w t - 0) dt. (4.6b) 0 g Equations (4.6) cannot be integrated analytically without a knowledge of the functions A(t) and 0 (t). Nevertheless, the first term in Eq. (4.6a) has the form of a weighted-average value of cos 0 indicated by A cos 0 multiplied by T. The second term in Eq. (4.6a) will be small compared to the first term provided that T is large compared to 2ir/wg. With a goniometer spin-frequency of 30 rps, 2ir/w is only.034 sec. Since T can be made much greater than this, the error term can be made small. Similar reasoning applies to Eq. (4.6b) so that for long integration intervals, the following approximate equations are valid: T T =1/2f A cos 8 dt =- A cos 8, (4.7a) e3a = 2/2 TT e3b = -1/2 J A sin 0 dt = - A sin. (4.7b) These voltages are again multiplied by the sine and cosine generator voltages as indicated in Fig. 9, and the outputs at position (4) added to produce a voltage at position (5) given by e 5 = A sin (wgt - ), (4.8) where T 2 21/2 A- [A sin 0 + A cos 2 ] and tan 0 = A sin 0/A cos 8. If ideal integrators are used and if their inputs are both disconnected after the integration time T, the output voltage e5 would clearly be a steady-state sine wave of amplitude A and phase angle 8. 16

If it is now assumed that the variations in A and 0 are slow compared to cog, that is, that the fading process is not too rapid, then T 0 A sin 0 dt tan - T (4.9) A cos 0 dt 0 For long integration times T,this is equivalent to Eq. (4.4). 4.3.2 MADFI Bearing Computer In order to realize in practice the system just discussed and to maintain the desired accuracy the following four conditions must be met: (1) The sine and cosine generators must be of equal amplitude, have very low output distortion, and have a relative output phase of exactly 900; (2) To this is added a second problem of constructing four-quadrant multipliers which will give good accuracy over an appreciable dynamic range; (3) The integrators must be stable and not subject to output drift over long periods of time and must be started together exactly in synchronism at the beginning of the integration interval; (4) Should it be desired to vary the integration time T, integrator gains must be made to vary inversely with T so that the output multipliers will not be overloaded. These problems are pointed out for the sake of clarity that, although providing the processing desired, the simplified bearing computer described in the previous section is in fact quite impractical. However, the simple bearing computer described serves as a model for the practical system used in the MADFI. The MADFI bearing computer differs only in the realization of some of the components of Fig. 9 and in the type of integration performed. The MADFI bearing computer is shown in Fig. 10. The sine and cosine generators are replaced by square waves generated by the synchronizing generator which is, in turn, coupled to the goniometer rotor. The wave corresponding to the sine generator is the north-south (NS) signal, while the wave corresponding to the cosine generator is the east-west (EW) signal. Note that the fundamental components of these waves are in phase with the simplified generators. If the multipliers in Fig, 9 are replaced by simple choppers, or electronic switches, driven by the NS and EW generators, the waveforms at 2a and 2b are produced as shown in Fig. 11. This figure also shows the input voltage el and the two chopper-drive square waves. 17

PEN RECORDER NS 100k | 2a0 RI| I r NOTE: OUTPUT CHOPPERS INCLUDE INPUT AVV" COMPONENTS OF ADDERS. CHOPPER 27k VOLTAGES AT 4a AND 4b 2 AMP 4a ARE BEFORE ADDITION. OUTPUT 3a CHOPPER BRG COMP DRIVE EIN V RM lW SYNC PEN RECORDER IV RMSB C 5 OMP: o tEW ~( C COMB DC co NS SYNC FILTER S 30 C/S 100k AM 2b RI 22k INPUT CHOPPER 27k DC OUTPUT 3b CHOPPER 4b DRIVE EW SYNC Fig. 10 MADFI Bearing Computer (Simplified) 30 C/S

eNS 0 eEW 0 O ITw 2ir 32r e2o e2= e4a e4b e4 Fig. 11 MADFI Bearing Computer Waveforms

The input to the bearing computer is normally taken from the narrowband filter, and is given at (1) in Fig. 10 by e1 =A sin (wgt - ) (4.10) The choppers operate in such a way that the outputs (2) consist alternately of the input signal and then ground potential in synchronism with the system NS and EW signals as illustrated in Fig. 11. With the input voltage el, the chopper outputs are e2 =A sin (w t- 0) - + -E (k )sin (2k + 1) co t (4.11a) 2a g 27f 2k k+1 g k=0 2b g 2A sn2 k gt k=0 After some rearrangement and transformation, 2 1 e2 = Asin (w t - ) sin w t + A sin (w t - ) 2a 7T c g g 2 (+ ) A sin (w t - 0) sin (2k +1) wt, (4.12a) k=1 2 1 e2b A sin(w t -) cos t+ Asin (w t - 0) e2b= -rr g g 2 g 2 k + - 7 1) l A sin (cwt - 0) cos (2k + 1) co t. (4.12b) iTr / 2k+ g gb k=1 Comparing Eqs. (4.12 a, b) and (4.5 a, b) we see that the first terms of e2a and e2b are each equal to e2a and e2b of Eqs. (4.5 a, b) except for a multiplicative constant. The harmonics differ in amplitude and number, but are all at multiples of cg and attenuated significantly by the integration as in the previous case. The transfer function of the integrators in Figure 10, where R2 is typically 3-10 times R1, may be written (dropping obvious subscripts): e3 1 K -- 3\- 1 K (4.13) e2 Rl C(+R)p +a2 20

I 1 where K = 1C and a = C. In differential-equation form, R1C R2C de3 dt + ae3 = K e2' (4.14) If e3 (0) = 0, the general solution of this first-order linear differential equation is e3 (t)= 0 e2 () ea(-t) e dT. (4.15) Now e3 can be found when e2 is the output of the choppers given in Eqs. (4.12). Finally, when the first terms of Eqs. (4.12 a, b) are rearranged as in (4.5 a, b), an expression for the outputs of the integrators is developed: t t K i0 a(T-t) K 0 a(T-t) e3 = A e (t) cos 0 dT — K A e cos (2w r - 0) dT + ea(T-t) - 1 JAea(T-t) + A ea(Tt) sin (Wg T- 0) dt +2K — 1)k t ) 2 g ~rr 2k l. k=l sin (gT - 0) sin (2k + 1) w T dT, (4.16a) t t e -K A e s sin (2 T- 0)dT +K g d+eaAe-t) + A ea() sin (w g - ) dT + A e f~0 ea(T-t) k=l + cos ( T - 0) sin (2k + 1) w TdT. (4.16b) g g If we make the same assumptions as in Section 4.3.1, that the integration time T is long compared to the goniometer spin period 2 7r/w and steady-state conditions have been reached, all terms in Eqs. (4.16) containing integrands periodic in multiples of g, may be neglected and equations similar to Eqs. (4.6 a, b) result: K T a(t-T) e3a A e cos 0 dt, (4.17a) 21

e3b = -K 1 A ea(tT) sin 0 dt. (4.17b) 0 Next the integrator,outputvoltages (4.17) are multiplied by the chopper-drive waves as in Figure 10. (Note that the high-order terms contain only odd harmonics of wg.) e4a sin o A (t-T) e4 =- sin Wgt A e cos 0 dt +... (4.18a) 4 2 g e4b KCOS (t - T) e4b =-2 cos A e sin 0 dt +... (4.18b) The voltages at position (4) are linearly added, after which the high-order odd harmonic terms are removed by the comb filter. Thus the voltage at position (5) is: e5 =A sin (wgt - ) (4.19) where t12 a(t-T) and A e s in 0 dt tan - - (4.21) T A ea(t-T cos 0 dt Except for the e (t-T) factors under the integrands, Eq. (4.21) is identical to Eq. (4.8). Thus the MADFI bearing computer does, infact,perform substantially the same processing as the simple bearing computer described in Section 4.2. 4.4 MADFI Bearing Computer Instrumentation Realization of the MADFI bearing computer components indicated in Fig. 10 is for the most part straightforward. The chopper and d-c amplifiers are designed for extremely low d-c offset and drift. The d-c amplifiers have an open loop gain of over 100 db but a closed loop d-c gain in the operational circuit shown of only about 3 db. It appears feasible to do away with these operational amplifiers by replacing them with simple RC circuits. They were included in the MADFI equipment for reasons of design flexibility and convenience. 22

The time constant of the feedback components of the operational amplifiers is selectable between 1/3 and 30 seconds. The exact values depend on the multipath process fading rate and are chosen experimentally for each operational situation. The integrated'bias' in favor of a particular bearing may be reset at any time by a switch, which simultaneously shorts the feedback capacitors. The comb filter contains three zeroes positioned to provide maximum attenuation of the most significant odd harmonic components of the goniometer spin-frequency wg and minimum attenuation of the fundamental component. A problem encountered during design was the change in incremental inductance of the iron-core inductors with changes in signal levels. Bearing errors of as much as a few degrees accrued at extremes in input signal levels, but were minimized by operating the inductors well below saturation levels. 4.5 Synchronizing Generator The synchronizing generator provides the NS and EW signals to the bearing computer as required in the preceding sections. In addition it supplies special signals to the digital display, which will be described in Section 5.2. Figure 12 shows the three-output signal waveforms, which are periodic at multiples of the goniometer spin-frequency w. Channel 1 (count) pulses occur at a repetition rate of about 5.3 kcs and are used in the digital display circuits. Channel 2 (NS) and 3 (EW) signals occur at a rate of about 30 cps and are used in the digital display and bearing computer circuits. The waveforms shown in Figure 12 are generated by a photoelectric method involving a rotating code wheel coupled to the goniometer rotor. As indicated in Fig. 13, the rim of the transparent code wheel is divided into three tracks, one corresponding to each output channel. Each track consists of a series of strips of opaque material placed in the pattern indicated in Fig. 13. Three miniature lamps placed in the general area behind the strips and behind the outer web of the wheel provide illumination of three photosensitive diodes mounted outside the rim of the wheel. The output of each of the photodiode amplifiers corresponds to the presence or absence of opaque material in its respective track as the wheel revolves in the direction of the arrows and results in a square waveform as shown in Fig. 12. The circuit shown for supplying operating voltage to the lamps was chosen for mechanical convenience. The protective diodes shown prevent a multiple lamp burnout situation in event of a single lamp filament failure. In this and other digital-type circuits used throughout the equipment, waveshape and driving levels are standardized for compatibility reasons. System waveshape characteristics 23

NORTH REF NORTH REF 34 ms CHANNEL I COUNT 1 (2o PULSES) - — 3V I~~~~ CHANNEL 2 I NS (9O00PULSES) I II I~~~~~~~~~~~._! I I~ ~ ~ ~ ~ I__ I - CHANNEL 3 EW _ _ _ _ _ _ _- - V Fig. J2 Synchronizing Generator Waveforms 24~~~~~~~~~~~~~~~~~~

Z ~ CODE 80 z I WHEEL I I I CHANNEL PHOTODIODE AMPLIFIERS PHOTODIODES CHANNEL I LAMPS COUNT 1.4v ~300ma CHANNEL 2 NORTH/SOUTH NORTH REF CHANNEL 3 EAST/WEST t DIRECTION 15v OF MOTION o -I Ov 25.A 2-RE30 o GND. Fig. 13 Synchronizing Generator (Simplified) ADCOCK ANTENNA BEARING DISPLAY DEFLECTION PLATES ANTENNA DRIVE INDICATOR C GONIOMETER SHAFT GONIOMETER J AMPLiLTUDE 72 kA AND DETEC\OR|MODULATOR OSCILLATOR T R (a) (b) Fig. 14 IMnual Sense Operation of AN/TRD-4A 25

have been chosen appropriate for l-mcs operation, with zero volts corresponding to a logical "0" and -3 volts to a logical "1." 5. DISPLAY AND RECORDING SYSTEM The MADFI display and recording system displays the amplitude and bearing of the incident signal on several types of readout devices. Instantaneous bearing indications are visible on the modified AN/TRD-4A cathode-ray display as the conventional propeller-shaped pattern. As an added convenience the automatic sense circuit blanks the anomalous half of the propeller associated with the 1800~ ambiguity mentioned in Section 3.1. A permanent record of the cathode-ray display may be made by the pen recorder, which traces a polar-coordinate plot not only of the indicated bearing but of the bearing signal amplitude as well. A three-digit decimal display reading directly in degrees is included in the system as well as paper tape punch which records the decimal display indications. The decimal display and the paper tape punch are sampling-type systems which display and record bearings periodically at a preset rate. Such a system in conjunction with the processing system yields very low information rates for peripheral computer processing. 5.1 Cathode-Ray Display Operation The cathode-ray display utilizes components of the prototype AN/TRD-4A system. Its principal modifications include the addition of the automatic sense circuit and provisions to convert the MADFI bearing signal to a compatible signal appropriate for the AN/TRD-4A cathode-ray display circuits. The operation of the modified AN/TRD-4A cathode-ray display may be better understood by contrasting it with the operation of the unmodified display. Figure 14 is ablockdiagram of the unmodified AN/TRD-4A system. The cusped waveform (a) appearing at the receiver detector output is a full-wave rectified sine wave resulting from amplitude detection of a double-sideband suppressed-carrier wave. This wave, normal to the unmodified AN/TRD-4A operation, is connected to an amplitude modulator in such a fashion that its amplitude zeroes correspond to peaks in the output. The 72-kcs subcarrier frequency is used to facilitate passage of the wave through a second goniometer. The two resultant goniometer stator voltages deflect the cathode-ray tube to produce the conventional propeller pattern. 26

Figure 15 is a block diagram of the modified system with the addition of MADFI automatic sense circuitry. In this case, the receiver output is an amplitude-modulated wave with index less than unity, so that the detector output is a sine wave plus a d-c component. After removal of the d-c component by a high-pass filter, the resultant sine wave bearing signal is full-wave rectified to provide the normal cusped signal for the 72-kcs modulator. If the remainder of the circuitry were unchanged, the propeller pattern would be identical with that described above. In the MADFI system, however, a blanking signal is generated in a balanced modulator by driving this modulator with wave (d), a 900 phase-shifted version of the sine wave (a), and with the same 72-kcs subcarrier as the original modulator. This 72-kcs blanking signal, wave (e), is phased in such a manner as to blank one-half of the propeller pattern as shown. 5.2 Modifications and Additions to Cathode-Ray Display Modifications to the AN/TRD-4A cathode-ray display are shown in Fig. 16. An amplifier was added to increase the blanking signal at the grid of the cathode-ray tube. The synchronizing signal which is routed to the automatic sense circuit is derived from the grid of the 72-kcs oscillator tube in the indicator chassis. The transformer and diodes and RC network shown are components of the full-wave rectifier and its associated phase equalization network. Although the transformer is operated well below saturation, considerable pattern distortion is observed with rapidly fluctuating incident signal amplitudes when the detector is patched via the patch panel directly to the cathode-ray display. The rapidly-changing d-c component generates low-frequency signals which cause saturation of the transformer. Since this patching condition is rarely used, a solution to this problem was deferred. The automatic sense circuitry is contained in an external module located in the analog case. Its circuitry, shown in simplified form in Fig. 17, is that of a ring-diode balanced modulator preceded by a phase-shifting network and followed by a 72-kcs amplifier. Its performance is such that a discernible blanking action is visible on the cathode-ray tube even if the bearing signal is too low to form a recognizable propeller pattern. The manual operation of the AN/TRD-4A indicator is left undisturbed, including the manual-sense circuitry. The MADFI operation is identical with manual operation except that the sense operation is automatic. The relay evident in Fig. 16 is part of the manual-automatic switching circuitry and is shown in the manual mode. 27

ADCOCK ANTENNA BEARING DISPLAY DEFLECTION PLATES SENSE ANTENNA (RT ANTENNA DRIVE INDICATOR BLANKING GONIOMETER SHAFT GONIOMETER AMPLITUDE 72-kc BALANCED AND DETECTOR MODULATOR OSCILLATOR MODULATOR co HIGH PASS I / I FULL-WAVE 90-DEGREE FILTER I I RECTIFIER PHASE SHIFT (b) (a) Fig. 15 Automatic Sense Operation of MADFI

5CPIB CRT V304 +275 V CR304 10k AN/TRD-4A C326 500 MANUAL BLANKING 680k BLANKING'CONTACTS RF OSCILLATOR V309 AMP 6UJ8Az OF AUTO-MAN RELAY 6005/6AQ5W CRT BLANK J322 275V 1,7 HF SYNC J324 FROM DETECTOR 5750 (2)15Op0f A 2UTC IN67 MODULATOR V320 15V IN - 0.252L Io' - | N67 IOOk 10 IN ~ r'.' H SYNC J301 12V I RELAY COIL Fig. 16 AN/TRD-4A Indicator Modifications 900 BALANCED MODULATOR PHASE SHIFT 025k 600 72 kc CRT BLANK AUTO-SENSE Iv RMS IN IN 0- 0.25 Iv RMS 25k 6 HF SYNC 72 kc 400mv P-P Fig. 17 Automatic Sense Circuit (Simplified) 29

5.3 Pen Recorder The pen recorder used in the MADFI to record amplitude and bearing indications is a Moseley Model L X-Y Recorder. This recorder is connected to the bearing computer operational amplifier outputs so that the X-channel indicates the voltage e3a at position (3a) on Fig. 10 and the Y-channel indicates the voltage e3b at position (3b) on the same figure. Points are plotted on polar coordinate paper arranged on the pen recorder so that the Y-axis corresponds to an angle of zero degrees and the X-axis corresponds to an angle of 900. Inspection of Eq. (4.18) reveals that the amplitude A and bearing 0 of the incoming signal are represented by the radial distance and angular argument respectively of points plotted in this manner. The recorder is equipped with a'pen pecker' which causes the pen to strike the paper periodically and produce a sequence of dots rather than scrawl continuously and produce a tangle of wavy lines. It is believed that'scatter' plots produced in this manner over a time interval of a few minutes will provide considerable information as to the nature of the fading process. 5.4 Decimal Display Digital readout of bearing indications is shown by a three-digit decimal display. The display contains large, easily read incandescent projection numerals and reads directly in degrees. Convenient binary-coded-decimal outputs are provided for either a peripheral paper punch or a digital telemetering link, making unattended operation at a remote field site convenient and attractive. The circuitry used in the decimal display and shown in block diagram in Fig. 18 is basically that of an event counter which operates during that interval the goniometer shaft moves through the angular displacement yo corresponding to the incident signal bearing 0. The start-signal to the counter is generated by the synchronizing generator as the code wheel passes through an angular position established as a north reference. The stop-signal to the counter is generated at the time of each positive-going zero crossing of the bearing signal eb from the bearing computer. During the interval between the start- and stop-signals, the counter totals the number of 1/2 -pulses produced by the synchronizing generator in conjunction with a circuit called the count multiplier. Several design innovations have been included in these circuits for reasons of accuracy and reliability and will be described in the following sections. 30

COUNT COUNT SYNC 0 COUNT U T COUNT MULTIPLIER GE "'OUNITS DECIMAL - 01TENS OUTPUTS REGISTER 3 - DIGIT -~HUNDREDS TO L DECIMAL COUNT COUNTER -oUNITS BCD CONTROLBCD _OTENS OUTPUTS SET I RESET = oHUNDREDS TO PUNCH CLEAR SYNC START NS o O' GATE ~ REGISTER CLEAR TIME DISPLAY O DISPLAY ZERO TI ME D DEL AY I r CROSSING DETECTOR INHIBIT RESET DIG DISPLAY HOLD NO-SIGNAL LAMP IN IV RMS Fig. 18 Decimal Display (Simplified) 4.7k 1 -AVV — 0-12V FEEDBACK DIODE SILICON DIODES I N k D OUTPUT 3V p-p ~3V SILICON DIODES _0 + 6 6V FEEDBACK 1.8k DIODE Fig. 19 Zero-Crossing Detector (Simplified) 31

5.4.1 Count-Control Circuitry The count-control circuitry of the decimal-display consists of all the blocks in Figure 18 except the decimal counter and the zero-crossing detector. The count multiplier employs a blocking oscillator synchronized to the fourth harmonic of the 5.3-kcs count signal from the synchronizing generator. Since the count pulses occur once every 20 of rotation of the code wheel, it is apparent that the pulses produced by the count multiplier occur every 1/20 of rotation. The count-control and count-gate circuits gate the pulses produced by the count multiplier according to a start-signal produced by the synchronizing generator, a stop-signal produced by the zero-crossing detector, and an inhibit-signal produced by the display delay circuitry. The inhibit-signal is generated by the display delay circuit, which controls the length of time a count is left on display before another count cycle is initiated. The circuit, controlled by a simple RC delay, establishes an enable condition to the start gate at some time during a revolution of the goniometer shaft. The next start-pulse to appear switches the count-control circuitry to the ON state. The following stop-pulse resets not only the count-control circuit, but the display delay circuit as well, and the delay begins timing the display interval. At the end of this interval, the enable condition is again established and the cycle repeats. The display interval may be adjusted by the operator, but a count may be initiated at any time by a reset control provided for that purpose. Particular care was taken in the design of this circuitry to minimize false counts due to undesired coincidences of the various pulses. If in any case the counter is'confused,' it is reset to zero and awaits the next start-pulse. This reset condition persists for only a fractional revolution and is generally unnoticeable. 5.4.2 Zero-Crossing Detector The practice of finding the bearing estimation 0 of Eq. (4.5) by zero-crossing detection requires a highly precise zero-crossing detector. The design precision of the digital display was set to be ~ 1/20, which corresponds to three significant figures on the three-digit display. This precision corresponds to about one part in 700, which is also a voltage differential of one part in 700 since the slope of the bearing sine-wave signal at its zero crossings is unity. Precisions such as this led to the approach chosen in the MADFI and illustrated in the circuit of Fig. 19, a circuit borrowed from analog computer technology. Each of the strings of four silicon diodes in this diagram provides a convenient voltage reference. When the voltage 32

at the output exceeds this voltage the feedback diodes conduct and the device gain approaches zero, whereas when the feedback diodes are not conducting the full gain of the d-c amplifier is active. The circuit used in the MADFI includes a level detector which supplies a lock signal to the count-control circuitry in the event of insufficient bearing signal amplitude. The lock signal holds the decimal display at the last successfully completed count. 5.4.3 Binary-Coded Decimal Counter The counter indicated in Fig. 18 is a three-digit Binary-Coded Decimal (BCD) type of conventional design. A feature of note is the automatic-rounding circuit, which rounds off the units digit according to whether the number of 1/20-pulses passing the count gate is odd or even. Instead of the ~ 1~-ambiguity present with conventional counters of this type, the reading is now accurate to the nearest half-degree. The technique consists of complementing the 1/20-flip-flop of the counter with the register pulse and propagating any'carries' generated. The ten BCD outputs from the three decades-four each from the'units' and'tens' decades and two from the'hundreds' -are connected by three BCD-decimal decoding matrices to driver transistors for the decimal-display lamps. Buffered outputs are also available to drive the paper tape punch. 5.5 Paper Tape Punch The paper tape punch provides a means of saving bearing estimates produced by the system for later computer processing. With the present system the bearing is punched once for each count cycle of the decimal-display. No provision is made for recording any amplitude or time information. However, even records of this simpler type may be highly useful in evaluating various combinations of processing elements. The tape punch, affectionately called the Michigan Automatic Generator of Indigenous Confetti (MAGIC), is adapted from a commercial tape transport through the addition of laboratory-constructed solenoid drivers and timing circuits and an integral d-c power supply. The unit is completely self-contained and may be driven from any source with compatible input signals. The punch is capable of recording eight binary bits of information per line of tape at a rate of 60 lines per second. This rate is more than adequate to obtain one recording for each 33

revolution of the goniometer drive shaft should such a rapid rate be desired. Since ten channels are required to record the full three-digit number from the decimal-display, two lines of tape are required per count. At present no provision is made in the MADFI system to do this; only the low-order bits from the'units' and'tens' decades are recorded as one line of tape per count. A device to perform the desired multiplexing could be realized using, say, a multicontact pulsed relay. Also, the received amplitude information could be recorded through an analog-to-digital converter, but neither the complexity of currently available techniques nor the high cost of commercial units was considered justifiable under the present effort. The diagram of Fig. 20 shows a logical diagram of the punch as well as the tape format. The inputs are connected to a set of lines in which a level of -3 volts corresponds to a punch and a level of zero to a no-punch. The inputs are normally gated off so that no-punch action is initiated except on the receipt of a register-pulse. The -3 volt register-pulse causes a 4.5-ms pulse to be produced which then routes the inputs through the solenoid drivers to the punch solenoids in the tape transport. Following the punch action, asprocket hole is automatically punched and the tape advances one line. A button is provided to punch sprocket holes and advance the tape without punching any input information. This button is used for loading and clearing the transport and preparing leaders, as well as for producing confetti for a multitude of uses. 6. FIELD MEASUREMENTS Due to time limitations imposed on the present effort, very little quantitative information has been gathered concerning system performance. A site for the antenna array was chosen that is believed comparatively free from man-made noise. The site is a partially forested, gently rolling, hilly terrain about thirty miles northwest of Ann Arbor, Michigan. With the aid of a surveyor's theodolite, the antenna array was erected and surrounded with bearingmarker stakes every 150 at a radius of about 100 feet from its center. The target transmitter supplied with the AN/TRD-4A equipment was placed on each of these stakes in turn and the MADFI indicated-bearing recorded. Deviations between the actual and recorded bearings are summarized in Fig. 21 for frequencies of 6, 8, and 10 mcs. The deviations are evidently due in part to the influence of the shelter and, at the higher frequencies, to the octantal error. 34

MOTION --- CHANNEL o 0 0 8 0 000 7 0 O 0 00 6 0 00 0 0 1 0 0 00 2 0 0 0 0 0 0 0 0 0 0 SPROCKET 0 00000 3 0 000 4 0000 00 5 TAPE FORMAT GATE SOLENOID I DRIVER I O 2 GATE SOLENOID 2 - 2 DRIVER 2 | 3 3 I o 4 4 o " Z- 0 0 w~J 5FIVE z _ 5 5 u > o- ADDITIONAL,o I w 0 0 CHANNELS c) ~~~6 6 z > 0 7 7 a.N o —- — o? GATE SOLENOID = 8 DRIVER 8 TAPE ADVANCE SW. SPROCKET PUNCH a DRIVER ~~~~~~~PU ~NCH NORMAL lTAPE ADVANCE PULSE _ PUNCH ADVANCE ESCAPEMENT PULSA I PULSE GEN. PULSE GEN. ADVANCE 60 cps FROM POWER TRANSFORMER Fig. 20 Paper Tape Punch 35

4 10 Mc 2 0 0.7/~~~~~~~~~~~~~~~~ CD 8MC 0~~~~~~~~~~~~~~~~~~~~~~~~ z 04~~~~~ 0 6012 1024 30 6 2 04 CAD ~ ~ ~ ~ SETRIDIAE ERIG(E. uFB-2 < 6 10 00 6 Mc36 > 24 -000000'E w Ndr.00~~~~IVDIAT EA NG( G. 0 0%,ADIDvato-Idcae erigCre

Qualitative data regarding the worth of the processing system components are difficult to collect since relevant data are largely statistical in nature. It is certain that, on a subjective basis, the appearance and stability of the cathode-ray display are vastly improved over that of the unmodified AN/TRD-4A. Stable bearings on single-sideband, hand- and machinetelegraph and even kineplex transmission are readily achieved. When the bearing of a station of known coordinates was compared to that indicated by the MADFI, agreement to within a couple of degrees was common even in the presence of multipath processes that caused the unmodified AN/TRD-4A indicated bearing to vary over a range of 1800. However, to support these subjective observations quantitatively, it would be necessary to collect a large enough quantity of data for a thorough statistical analysis. 7. SUGGESTIONS FOR FURTHER RESEARCH The MADFI system is intended as a research tool. As such it has application in studies of ionospheric phenomena and direction-finding techniques. However, some of the MADFI operating characteristics may need to be refined so as to inject the least amount of error into some specialized measurements. Thus, it is suggested that further research include some considerations of equipment improvement. 7.1 Collection of Bearing Statistics Bearing statistics furnished by the MADFI equipment could well be used in studies of ionospheric phenomena related to radio transmission. Probably the most useful record of these would be the paper tape record, together with additional amplitude and time information. The bearing estimates could be read at high-speed directly from the tape by a general-purpose computer. Results of a large amount of data taken over several months could be processed in this manner to give statistical information concerning ionospheric dynamics and ray-retracing techniques. Of equal importance would be the quantitative information available concerning the evaluation of antennas, sites and direction-finding methods in general. As a system, the MADFI has operating characteristics which are deemed undesirable under certain conditions. A few of these will be discussed in the following paragraphs together with suggested improvements. Some of the problems in the MADFI are directly related to those in the original AN/TRD-4A DF Set, and many in the areas of antennas, goniometers, 37

and receiver filtering have been discussed elsewhere (cf. References). It is believed that attention therefore should be concentrated on problems of the processing and display systems of the MADFI. Throughout the MADFI effort,little attention was paid to the effects of nonuniform phase characteristics of the postdetection filtering components. The effects of this distortion are significant only in rapidly fluctuating multipath processes and in this connection may be worthy of further study. The realization of the bearing computer has demonstrated the promising features of linearly-weighted time-averaging. Virtually no quantitative testing of this particular weighting scheme was possible on this project. Examination of this and various other weighting techniques for use in the bearing computer would be desirable, especially in connection with the problem of rapid and widely varying signal amplitudes. Some system of utilizing automatic gain control-circuitry in the receiver appears particularly attractive in view of the wide dynamic range associated with some multipath fading processes. The operational amplifiers in the integrator circuit of Fig. 10 could be eliminated from the system by replacing their gain with the other active circuit components. Although the amplifiers were designed for maximum stability through the use of differential circuitry and silicon-type transistors, the small but measurable drift affects MADFI long-term accuracy. As mentioned in Section 3.2.2, the comb filter might be improved through further study. The three-zero design was used as a result of an arbitrary choice made some time ago in the history of MADFI. Unquestionably, a method of eliminating the nonlinear iron-core inductors would be most desirable to implement. An estimate of bearing error due to residual harmonics at the filter output should be studied. As mentioned above, the paper tape punch is not equipped to record the amplitude of the bearing signal. An obvious improvement in the recording system is to incorporate provisions for this. An analog-to-digital converter would be required along with a multiplex unit. Additional convenience would be gained by including a provision to record periodically the time of observation. 38

8. APPENDIX OPERATION NOTES AND CIRCUIT INFORMATION Most test points adjustments and indicators for setup and operation of the MADFI are readily accessible from the front of the equipment. Figures 22 and 23 show the logical diagram and cable interconnections, Figs. 24 through 28 show the equipment and connector placement within the racks, and Figs. 29 through 31 show the arrangement and labelling of the various controls. Figures 32 and 33 show the logical interconnection at the modules in the Analog and Digital racks. The remaining figures, 34 through 60, show the detailed circuit diagrams of the modules within the rack. After the usual procedure is followed, the AN/TRD-4A is tuned and adjusted with the MADFI in the manual code. With the receiver RF gain control at minimum and MADFI in the automatic mode, the regulated power supply should be adjusted, and all d-c amplifiers zeroed. In the case of the operational amplifiers in the bearing computer, zero adjustments are made both with the INTEG RESET Button depressed and released. The balance control in the automatic sense module is adjusted for minimum 72-kc carrier at the output and the null controls of the comb filter adjusted for minimum goniometer-spin frequency harmonics at its output. The lamp current control (on the goniometer drive chassis) is adjusted for lamp brightness consistent with good pulse shape of the count (Channel 1) output. The narrowband filter gain is found by experiment. Sense phase adjustment is made using an oscilloscope connected to the R-390/URR IF output to obtain a good amplitude-modulated wave. Using the RF gain control of the R-390/URR, the system gain is set at a point where the output meter reads one-half scale (50 Ala) allowing for the integration time of the bearing computer. The input meter should not swing upward off-scale under any fading conditions. The narrowband filter gain may be used to increase the processing system gain if necessary. Integration and count times are chosen appropriately for the signal dynamics, and the AUTO SENSE and AUTO/MAN switches are used as desired. 39

POWER SYNC GEN POWER SYNC COUNT I BCD CONTROL I COUNTER DIGITAL CASE < Cr.-z cn -- Lcr 1 DISPLAY CONTROL LL uCONTROL UNIT II I8 AUTO I NB I BRG SENSE [ FILTER I COMP ANALOG CASE CRT BLANK HF SYNC.. I RECEIVER INDICATOR MODIFI ED AN/TRD4-A Fig. 22 System Logical Diagram 40

TYPE OF CIRCUIT w Cy) z ~ TYPE OF POWER SYNC., RF AF CONTROL TYPE OF cnr,, c EQUIPMENT J401-J404 GONIO, DRIVE J401- (G) P701 PG2 PGI J4 5J405 P401 X X ww ww >U CONTROL aZ f~O Z O~ PC6 PC5 PC2w PC4PC3PCI (C) 0 ~ ~ ~ ~~Cu') u3 o o ~o) 0' J324 o ~J304 J322 - J308- W J303 INDICATOR J322~ ~ J311 H w 115V P 30 w 115V P302 J302 "'J323" J321 P301 J301 (I) T To'~~~~~~~~ - - - - - - T0-; 6 < JIO8 J106 LOCAL RECEIVER AUDIO 115V J104 o J107 (R) z z z w w z z z ( W~l L < C PPI PP6 4<0 PATCHING 0 0 / PP8PP2p3PP4pP5PP7 (p) z z H zHP >- > - w ZZ ~ O U) U) U) ~2142PL44 co Mz / ~ANALOG 0 oI Z 6 PAIO PA7PA4PA5PABPA6 PA2 (A) PAII PA3 z x PLOT Nx PLOT m DIGITAL (.D (,Du3 0 0 PD3 PD4PD5PD7..J PD4 PD5 PD7 (D) Z Z PUNCH.D~ P PD0 PXI PX2 PX3 PX5PX7PX8 PXII PXI 12 PXIO PX9 CONNECTOR PX4 PX6 (X) PS2 POWER 115V PSI (S) Fig. 23 System Connector Diagram 41

TP2 I~~~~ TP3 PGI PG2 TPI 0 0 LIII [El ~~~TPIa O SYNC GENERATOR- REAR Fig. 24 Synchronizing Generator Connector Placement o o~) 00 I,.~,, - LO d w TEST CZ & H Q L O M. W a 3 o4 o z ~ o o~ < < 70 a8 ANALOG CASE-FRONT PA6 LAI PA4 P PA9 W ~~0 PA7 P2PA5 0 2 W2 0OPA8 PA3 PAIc ANALOG CASE-REAR Fig. 25 Analog Case Module and Connector Placement

ZXD M19 DC AMP M18 COUNT MULT M17 DEC 4410 M16 iiiii DEC 4410 M15 DEC 4410 M14 0,\> DEC 4303 M13 0Q or c_ DEC 4209 M12 ~ + - -I > > DEC 4105 MII > > rn m I m DEC 4102 MIO H;o (D > 0::o z DEC 4215 M9 F" —T1 -4 o a) | L | |DEC 1671 M8 DEC 4215 M7 DEC 1671 M6 DEC 4215 M5 DEC 1671 M4 DEC 4102 M3 DEC 4604 M2 DEC 4689'MI o o olor I o0C -

SENSE Jx JX3 JXl JX7 JX9 ANT. JXII JX12 JX2 JX4 JX6 JX8 JX1 CONNECTOR PANEL-REAR Fig. 27 Connector Panel Connector Placement JPI2 JP1I JPO gJP9 J7 JG J J4 J3 J72 J6 PATCHING PANEL-REAR Fig. 28 Patching Panel Connector Placement 44

INPUT OUTPUT METER AUTO AUTO METER L MLTLMAN SENSE COUNT TIME - DIGITAL DISPLAY I INTEG TIME o O COUNT INTEG RESET RESET CONTROL UNIT- FRONT JC JC6 JC4 ( CONTROL UNIT-REAR Fig. 29 Control Unit Control and Connector Placement

( S OFF OJI OJ2 OJ3 OJ4 OJ5 OJ6 OJ7 OJ8 Si J1, NOT USED J2 9 J3 - 72 kc AUTO SENSE J4 - COMB FILTER OUTPUT J5 - EW OUTPUT CHOPPER J6 - NS OUTPUT CHOPPER J7 - REGULATED +6 VDC J8 - REGULATED -12 VDC Fig. 30 Analog Case Test Panel

ON SI S2 OFF OJI OJ2 OJ3 OJ4 OJ5 OJ6 OJ7 OJ8 Si {ON - TEST OFF - NORMAL S2 PUSH TO CLEAR Ji1 - N-S SYNC J2 -. 5~ CLCK J3 - E-W SYNC J4 - COUNT TIME J5 - 20~ CLCK SYNC J6 - DELAY M. V. DISABLE J7 - STOP PULSE J8 - TEST PATCH NOTE - FOR TEST J8 IS PATCHED TO J1 (00 OR 360~) OR J3 (900) Fig. 31 Digital Case Test Panel 47

JAIO-2 o.HF SYNC -15V JA2-6 JA6 -AUTO SENSE IN JA6 JA7 - N FILT IN NB FILT OUT JlA7 A F H K TDCAMP NOTE NB FILT REG PWR F AUTO E CERT BLANK MIO 4 M9 M5 SENS x Y x Y 0 TP3 JAII -3 + —-- 0 IV +6 JAII - 2 o 0-~ -15v 40 6 JAIO -1 ~~~ ~~~-12TP JAIO-I -12 JAIl -I mx Y X Y X Y x y X Y x Y JA3-4 X 00 JAI-l DC AMP NOTE CHOPPER DC AMP NOTE CHOPPER DC AMP NOTE COMB FILT BRG COMP M3 I MI F M4 2 M2 F M 3 M6 v OU, JA2- I N HL NIH L N K T S JA3-3 SYNC EW JA3-2 SYNC NS BRG COMP IN JA5NS NOTE I. SEE DETAIL A NOTE 2. SEE DETAIL B NOTE 3. SEE DETAIL C NOTE 4. SEE DETAIL D Fig. 52 Analog Case L~ogic

DETAIL A DETAIL B DETAIL D TP5 V TP6 V W W DC AMP Z Z CHOPPER DC AMP CHOPPR DC AMP NB FILTER M3W J MI M4 w J M2 MIO M9 KK K K K Z Z JA2-2 e JA2-4 o JA2-3 JA2-5 o co JAI-3 o JAI-2 o DETAIL C J J DC AMP Z Z COMB FILT M7 K IP M6 TP4 o Fig 32 Analog Case Logic

COUNT COUNT FLIP FLOP I P R COUNT LAMP SYNC TP5 TP7 M9 LAMP SNDRIVE JD4-31 COUNT F COUNT E S PULSE E H NVm JD6-1 MULT GEN ~MI7 [ R ~~~M17 M14 ~~~FLIP FLOP MREGISTER PULSE MI2 N PULSE S AMP JDI-II COUNT M M2 o JD4-25 JD6-4 TP2 TP2 ~~~~~~~~~~~~~~~~~~~~REC~STER SYNC NV INV JD5-I NS S PULSE F M12 M11 M JD6-2 GEN E I NS _____________________~~~~~~~~~~ —-- JD2-1 SYNC JD1 S12 JD6-3 —-o TP3 I --— a I S FLIP FLOP M10OJD2-2 M11 M12 Cn JeT1 ~~~~~~~~~~~~~~~~~~~D55-I4 -TP4D EW o JD5-5 K L-L +2 -(IN CONTROL TJD4K JDK ~F3 PULSE STOP LEVEL INV E INV NA I GA~~~~~~TP4 F S M13 M13 CLEAR LDI'-G.~" DISoNO JD5-6 (IN CONTROL UI)0 _______ SINANDTE S2 SHOWN INE NOR' FF'F;; o ~JD4-29 JD3UIg ENALoi ULSE Co E TP7 LAMP DRIVE I ~ "J43 TI~ ~ ~ ~ ~ ~ ~ ~~~~~~~ ~ ~ CLEAR F EM ~~~MII AA GIG. DISP. NOTEMI m l FII T ZXD W S(TES INTE (NOTE U.SE D)TI J03 M~~I9 INM1 NOTUE I.STP EE ADETASH NIL NRAL ~~~~~~~~~~~~~~~~~~~~~MI9 US 30 ~ ~ ~ ~ ~~~~~i. 3 Dgtlcpse!4) - Kon CnrL

DETAIL A DETAIL B R.W M M13 JD5-2 S K M9 S K M18 INTEG J o JD5-3 ZXD DC AMP DELAY D JD5-1 Y Y V J Fig. 33 Digital Case Logic - Count Control HIGHER DECADES TAPE PUNCH JDI-4 o N F INV M E DECIMAL M3 TY z DI SPLAY UNITS I ||FLIP FLOP JD4-10 JD4-Io M5 JDI-3 c O D JD4-9 L - JD4-8 INV K ___ M3JD4-7 BCD JD4-6 JDIc- FLPFLPDECODER JD4-5 JDI-2w JD4-3 INV H K S, ~~M3 TN R~~ JD4-2 JD4-I I FLIP FLOP M5 COUNT o Fig. 54 Digital Case Logic - BCD Counter 51 Fig. 34 Digital Case Logic BCD Counter

TAPE PUNCH DECIMAL JDI — 10 I O W DISPLAY | ~~~~~~~F Lu~ ~HUNDREDS!,.v I~ JD4-30 BCD T JD4-23 FLIP FLOP DECODER JD4-22 M9 JDI — 9 _, IM8 JD4-21 TZ x w N YNV M M3 T U FLIP FLOP M9 J D I-8 o A - 1S R F NV W E TENS FLIP FLOP V P o JD4-20 M7 JDI-7 o JD4 -19 o JD4-18 INV U HX o JD4-17 M3 BCD W rrl r_ ~t ~BCD JD4-16 < FLIP FLOP DECODER V - JD4-15 M7 JDI-6 c M6 -U JD4-14 T T JD4-13 INV S K S I JD4-12 M3 N P JD4-11 FLIP FLOP M M7 JDI — 5 c R L K N INV P M ~M3H J FLIP FLOP M7 F E LOWER DECADE Fig. 34 Digital Case Logic - BCD Counter 52

RS-722-2 DI IN2Q08 OUTPUT SOCKETS 3 CINCH JONES S-S30 F.P TI 4 INe08 +10V A) SI DEC N328 0 to 7 0AMPS 100-X-1016 | |vi o 1 P INPUT C S + AM Nr - _ 35,000 MFD 3,000 CI I S Is 5 5W C-%DAL I" -"ON 6I - 6,oooV DC I No, PIN I - + IOVCAI 160-4 1 I0-5 I f,, DIALITE SN D6+ C6 2 4 1020 Al 320 T H ISAS SCH EAFD S ND D43 WF S 25W 0-3 1 Sto 6.2AMPS IN3208 $%... --' to 6.5 OAuPS SUM OF OUTPUT CURRENTS ARE LIMITED BY THE FOLLOWING EQUATION 5110 + Gli5 b UNLESS OTHERWISE INDICATED RESISTORS ARE I/OW, 10% CAPACITORS ARE 1/2FW 1%THIS SCHEMATIC IS FURNISHED ONLY FOR TEST AND MAINTENANCE PURPOSES. THE Fig. 35 POWER SUPPLY DEC722 CIRCUITSARE PROPRIETARYIN NATURE AND SHOULD BE TREATED ACCORDINGLY. COPYRIGHT 1961 BY DIGITAL EQUIPMENT CORPORATION RS-1671-3 RI 4 I!! I!. R.RI R oIl 0 03 04 s 54? RR 049 52 R R2 > 64 G 9-26N224 G 82N4 N224 222NZ224 2 N 224 2 N24 1 224 1224 26N224 P 5% T U V 1 X v 2 031, + - - 1 - -I - 1 r - 05 ( W224 2 4 %t24 Nt22 2f t (@ U 4 Nt1 ( 4 e 2 4 o t 3 * sr 9 p R s T U v 1 X Y Z THIS SCHEMATIC IS FURNISHED ONLY FOR RUESITOS OrARE I/W, T, TEST AND MAINTENANCE PURPOSES. THE OOES A.RE,N7N Fig. 36 B.C.D. LIGHT DRIVER DEC1671 CIRCUITSARE PROPRIETARYINNATURE AND SHOULD BE TREATED ACCORDINGLY..L~ ~~ ~ (Digital Modules M4, t~1, MB) CO.RIG..T ST DIGITAL EQUIPMENT CORPORATION 53

RS-4102 A + IOV (A) R3 R6 R9 R2 RI5 RIB R21 R21 R24 R27 2N1305 2N1305 28)305 2NI305 2N1305 2NI305 21305 2NI305 2N1305 8N645 012 IN645 CI I R2 C2 R5 C3 R8 C4 RI1 C5 R14 C6 R17 C7 R20 C8 R23 C9 R26 D IN645 680 3, <0 680 3,000 6 3,000 3000 300 3,000 80 3,000 68 000 i% / % 50% 5% 5% 5% 5% (N640 | I F, J |, L _n0 | R 1-T _V _X -Z _MF o E 8 K M P O I ) Di 02 03 D4 05 D6 07 D08 D IN276 18276 N1276 IN276 IN276 IN276 IN276 IN276 IN276 4 RI R4 R7 RIO 813 R16 R19 R22 S 828 1500 1,00,500 50 C1 500 1,0 50000 10500 1,500 560 C-IsV UNLESS OTHERWISE INDICATED Fig. 37 INVERTER DEC4102 RESISTORS ARE 1/2W 10% CAPACITORS ARE MMFD (,., \, THIS SCHEMATIC IS FURNISHED ONLY FOR (Digital Modules vM3, M10) JTEST AND MAINTENANCE PURPOSES. THE CIRCUITSARE PROPRIETARYIN NATURE AND SHOULD BE TREATED ACCORDINGLY. COPYRIGHT 1961 BY DIGITAL EQUIPMENT CORPORATION RS-4105-5 OB t IOV(B) 82 R4 A +IOVIAI 68,000 68,000 R6 88 RIO 1/4W 1/4W 68,000 68,000 68,000 H M 1/4 W S 1/4 W W 1/4 W F J N T UN~,,~~LE~~sS OT, ~RW,~SE,INDI,, CATED:~D GND QI 02 03 04 05 2305R AN305 2 2I0 5 IS FURNISHED ONLY FOR 18645 L K P U 7 18645 MFD TMED 12 13 ICOPYRIGHT 1961 BY DIGITAL EQUIPMENT CORPORATION 1,500 1,500 1,500 DI X 5% 02 59 3 5% 560 C-15V UNLESS OTHERWISE INDICATED: ~~~~~~~~~~~~~~~~RESISTORS ARE 11/~~~~~4W, THIS SCHEMATIC IS FURNISHED ONLY FOR CAPACITORS ARE MMFD. Fig. 38 INVERTER DEC4105 TEST AND MAINTENANCE PURPOSES. THE CIRCUITSARE PROPRIETARYIN NATURE AND (Digital Module M11) SHOULD BE TREATED ACCORDINGLY. (Digital Module Mll) )~~~~COPYRIGHT 161 Y 000DIGITAL EQUIPMENT CORPORATION 54

RS-4209-10 A A SET CLAEAR 0 DI ARE 176 CLEAR I 3C000 CPCO R3, A O1 I A B".. 5% F Q2 IENT 000 COMPLEMENT 6~:~ ~ ~ ~ ~ ~ 5 OI 47 47 O PRW0 ~~~~~~~~~R49 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~SUDETACTE RDi NL FRS425- A2 0C3 470 70I I [4 [, A'd'~o" AA"d'A'T' <OI 47FO 4 IF WI~~~~~~~~~~~~~~~~~~~~~~~~ IN S~/~.-],, 30 05 3,0 O4D 30 01 5 300 0197 5%~~~~~~~~~~~~ — 5%e 5%I5 RI R ~ ~,o 64 D1 R7 15rI I BiOO Is RI O60 65000 6W300 68,000 - 65000 650070 66,00't 40 68,00 6800 000 SI - ~~~~~5 - ~~~~~3 04 95~~~1 06Ar. 07 IRDON 000 CSCR2 C03 ~ CS ] CI DI5!2 IV-~~~~~~~~~~~~~~~~~~~~~~~, L! 470 D221 D23 470 x - g I 47O' IN9SIS ISOIO ISO I 6 -OA4OV(A) R19 8+40V~~~~~~~~~~~~~~~~~~~~CIVB. R 1l5,000 00 R2 Roo 37 I'Oto R800 R& R42 R14 MOOD 15, 14) 40 00 Rao R21 R25 R2643 4000 62 0 RS2 RB o RIO RIR 619 603 HO 02 GFO 3,000 3,000 3,000 500 3000 330 3000 300 IT 5% 5% 75% 5% 5% 5% 5%D5% 02 03 04 07 08 09 012~~~~~~~~~~~~~~~~~~~~~~~~5 013 04 N175 16 184 if 5% I 4 5%.. 5%,. 4: %5o3% 5 1911/645!W/ 17:,RS50500R51 5% 2,20 750 A D7 420.01 i D12 DO 630 ~ ~ ~ ~ 9 63 8 1,500's ('Y I C!S!!, l!~ 026!,30!!! C. i.FD I! INS45 FFA FA FF 16 FF% FF8 1F F4 S C9 5 UNLESS OTHERWISE INDICATED: RESISTORS ARE I/AO, 10% CAPACITORS ARE MMFD — DIODL$ ARE IN276 TRANSISTORS ARE 2NF305 THIS SCHEMATIC IS FURNISHED ONLY FOR 2NI305 TEST AND MAINTENANCE PURPOSES. THE Fi~3 UT FZ-TO E42609 DIODES ARE 18276 Fig 40 4-BIT CO NTER DE 42CIRCUITSARE PROPRIETARYINNATURE AND (D igit:j~ ESS~~~~~~~~SHOULD BE TREATED ACCORDINGLY. (Digital Modules IM1 M..... RS-4215-4 _C, ~, T~~nNSIST~~.OIS MR 135TISSHMTC FURIHD OL C 4?1.........,,o.,:. oooL,LPFO E40 ETAD ANEAC UPSS COPTRIGHT 1961 BY DIGITAL CQOIPBCNT CORPORATION 55.001 MFD - 0MFD OOIMFD 3R7 R 14 R21 R28 DI 300 V5 36000 IDIO 3,000 1 F D1 3,000 1 5%? 5% 5% _.,,.OV(A) RI R 4 RS R R11 R15 RIB R22 R25 61%000 W O 68.00c 64000 sa8poo 6B,000 66000 600oo0 o1i 02N G33 04 - 08' 2Nrf54 M;'54 2NII754 2NIN 1754 D21 C5 C C7 ce o 10 c i 1 D22 150 150 15' 150 15'50 1150''150' CN64,13.01 R5 R9 R19 R23 R MFD:~, /I oo./,ooo:~ ~..oo/o~,o~,,,. <5,. ~..'000 - 4 i I I D~~~~~~~~~~~~~~~~~~~~~~~~~~24 R6 4,9 RIO R13 R17 R0 R24 R27 AJDI k5OO 45 ~~00 k500 Di 1,500 1,5AD1 1,500 450 j D20 I T R31 R ~ i I I ~0R37 15 VC-S ~~~~~d~~~~~~~ r9~~~~~~~~~~~~~~4 R36 R30 R38 1,500 k5ODC17 k500 19 5~~~~~~~ -15 30 "336 MO 265 330 CI 21 R31 333 ~~~~SO DZ3 1 135~~~~~~~~~~~~~~~~~~~~~~~~~~00 Ow 6Y6 O OU IK hN m T EH 6i SET ZERO ONE SET ADD ZErRO ONE ADDZERO INHIBIT ONE ADD ZERO ONE ONE OUT U ER F OUT OUT 13 OUT FF" 3 OUT *4 OUT OUT FFA F'A FFA FFA I- FF-3 FF3 R'4 W4 UNLESS OTHERWISE INDICATED: RESISTORS ARE 1/4',' " CAPACITORS AIRE MMFD 77 —--- THIS SCHEMATIC IS FURNISHED ONLY FOR DIODES ARE IN276 Fg 4B C U T D 42 TEST AND MAINTENANCE PURPOSES. THE Fig.S40 4-BIT COUNTER DEC4215ECIRCUITSARE PROPRIETARYIN NATURE AND (Digital Modules ME T SHOULD BE TREATED ACCORDINGLY..1) M7~ mg) COPYRIGHT 1961 BY DIGITAL EOUIPMENT CORUR.TIO) FFA FFA FFA FFA FFB5

,'S-4303- 4 so-A 0 3600 6 000 33,000 6,I00 UNES THRIE NIATED04 GND RSS330TRS 1 1 I / 0 N645 IN276.0022 T AU (Dg lMD M4SHOULD E TREATED ACCYORI NU NL R, RIO _ _ R+ 1KS2,20 27 I0,00 0 331 -0 9C F:26 500 C1I.,..,.,OR..0. D13 ACI<URS AR TFE / 3~T~~~~~~~~~~~~~~~~~~~~~~~~~~SOL B TRAE ACRIN645 A~~ ~ ~~~ ~ ~~~~~~~~ ~~D3 ~ ~~~ ~ ~~~~~~~~~ YI G 1961 DIITAL EQUIP' CORPoRATIO 156000 D% 17545 20,000 De 1,500 1,500.39 309 39 51/< 5% 5% 5%,%0 /I I __J~T.~~,%.~.T o C-15V UNLESS OTHERWISE INDICATED 2,200 5. D AFD 3T000 IN276 PRE DE SLD3 B ETREAIN64 TANTC 120 IN276 330~ C~2N17W I R6 O e +lOVHe) M~FD zo IN276 04 =5! R MFD " R12 UKLESS OTHERWISE IN1DIT'-_ T`' 11'r,31STORS ARE 1/2W,.'' Fig. 42 PULSE GENERATOR DEC4410 THIS SCHEMATIC IS D LF (Digital [odules. 14, IVIL5, M16) TEST AMAINTENANCE Pl~ 2SES. THE HOULD BE TR EATED ACCOR DGLY.,'~"RIGHT 1961 BY DIGITAL EQUIPV~1- CORPORATION

RS-4604-4 vb ols 0~~~~~~~~~~~~~~~~D1 22 C4 330 IN914 CIO 330 IN914 C6 330 IN914 _o-j N-j.[ o "' -IN TI -IN T3 -IN T' C3 330 0M3 ID20 0 C 0 C I 5 330 019 DEC 05 2T 6 kloi 30DI T2 YO- T203 *IN D5I +IN D'~lP 13 +IN 2'Ti ~:lr?'"'l I "~$'~ ~ I l I1 RI R30 RI3 w 45500~~ 22~ 1,r2500 2I2' I DE 25% % 5 0 4 ---— A+IOV (A) T T T __0~DIODE+ IOV(AR 1N Rl0 R24 3 14_I I'000 6,,000 8,000 59C V4W s V4W sY1/4 W TNLESTSAND MANTE CEPURPOSES IATHED A.E~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l,.2,T Fig 43 Pulse Er 06 7 DEC 09 = - ~ Amplifier Module - DEC4604 ~~~~~CIRCUITS ARE PROPRIETARYIN NATURE AND (Digital Module M2)SHUDRTEAD 2N19 2N305 2N75 27 RCOPYRIGHT23 161305 YR37 DIGITAL EQUIPMENT CORPORATION RS-4689-1 I~~~~~~~~~~~~~~~~~~~~~~~~~~w ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ A IOVIAI I I 28 CIS I 1 I 12 cis I N645I C2 RI R.01 C R21 R R 1 I R36 22,000,00 22,0 2,000 2,000 2, 000.2,120 2,000 22, MF C1 3010 030 5 F 04 05 i MF B 5?SRO 5%C 59 5% 5% 09 % 0 02 0 54 6 6 D272ORS K(C~c — DEC PoC —- L DEC W( —DEC 64 R49 T2037191 14T 4 1N9'IN19114 IN9 220 2O 220 3 V 0264 DS 016 D24 645 F H I BR L N P R26 S T 17U W 20 R40 D7 U OTERISE IOIC'TE 914 R4B94 120, 120 120 RESISTORS ARE I/W, 0% THIS SCHEMATIC IS URNISHED ONLY OR DRITES AREPRPRRINNR A TEST AND MAINTENANCE PURPOSES. THE Fig. 44 Indicator Dfiver Module - DEC4604 CIRCUITS ARE PROPRIETARYIN NATURE AND (Digital Module M2) SHOULD BE TREATED ACCORDINGLY. (Digital Module M1)............OPYRIGHT 1961 BY DIGITAL EQUIPMENT CORPORATION RS — 4689-1 5 - __ A+IOV (A) RI R 3 R 5 R R R 9 R 1 R13 R15 R 17 2,000 <2,000 (22,000 (2,000 (2,000 <2,000 (2,000 <2,000 <22,000 __ D GND 0 2 03 04 s 05 0 6 07 08 09 2N527 ) 2N527 _2N527 2N527 1 2NI127 2N527 2N 21527 1 2N527 1 2N527 DI D 2 D3 D4 D 5 D ~6 D I D S D9 IN14 IN14 IN1 I N914 ir'1914 1 914 1N914 N914 R2 R4 I I < me ROO R 12 R 14 R16 RIB I~o 1,000 I IoO I <,~ 1,000 1 1000 I 1,000 WOO 1,000 000' 000'0 F H i W L M N P R Q T U V W X y z UNLESS OTHERWISE INDIC~TTV RESISTORS ARE 112'o, 10% THIS SCHEMATIC IS FURNISHED ONLY FOR TEST AND MAINTENANCE PURPOSES. THE Fig. 44 I-ndicator Driver Module - DEC4689 CIRCUITS ARE PROPRIETARY IN NATURE AND SHOULD BE TREATED ACCORDINGLY. (Digital Module Ml) COPYRIGHT 1961 BY DIGITAL EQUIPMENT CORPORATION 5'/

NO-SIGNAL DETECTOR 2N1404 ~2 ~~NI ~808 5o.6 ~ F OUTPUT UTC IN270 k (4) 0-6 2.2k IN457 _-2100kf,, lOOk1f 2 5 k ( 18.N45 2k 2pf __,_,' X, 6v' ~[IN45 7 - ~1N270 ~~~~~~~~~~~~~~(2)~ ~ ~ ~ ~ ~ ~ ~ ~C -150 v. k 4 Ok TO o R AMPLIFIER Fig.45Zero-CrossingDet4.7kector Module (DigitalModule )MODU

,02 OUTPUT 3.3k Ik -15 4~~~~~~~~~~~~~~~~ C BLK | BLU RED 3.3k NIN404,005 2.7 k<; H LM I CA GND D CLOCK IN FREQ I.005F IN270 SYNC INPUT Ik Fig, 46 Count Multiplier Module (Digital Module M17) 59

o A + IOv Ik 2N1404 12 k 12k L CHOPPER INPUT 5.6k 2N1626 2 w m )LK INTEG. INPUT o F SYNC.,6.8k J INTEG. OUTPUT 10t O~~~~~~~k 0-l~ | [ |V TO TEST POINT /i. 2N1808-1GROUND /0 Z IN457{o z k $J g 10/25 10/25 0 H TO ADDER H 1N327A >5k / — 1 2N1808 ADJUST X +6v 1.5k NA/ V-'+- 150k _ 82 J-o y -12v _,o C -15v Fig. 47 Chopper Module (Analog Module Ml, M2) 60

__'100/15 loos Y Ik oK 90cps 150cps 210cps IJlfIOH IO IH IOH 4.7k 30cps 4 30cps. T0.65.65 0.26 0.26 0.12 0.12 P Ad6~~ V (2) 5033kf 15V 470 b~ 22k (2) 5 oW 15D 0. 35_f BRG COMP B INPUT NET Fig. 48 Comb Filter Module (Analog Module M6) 61

o F INPUT A4 +1Ov SELECTED'42k 2NI808 2N1808.5/i Ilk I N270 ( Ilk OUTPUT It l TO BASE #? I5/ilt pllk W J nR O.5/i Ilk AMR MOD. IO Ik TRIMPOT 33k _ _ 15k LIMIT LEVEL C -15v w AMPLIFIER MODULE #10 H FEEDBACK RESISTORS TO OUTPUT 1Ok oK BASE #1 Ik oD GROUND Fig. 49 Narrowband Filter Module (Analog Module M9) 62

2.2k -12 12k 12k 3.3k (LOWER) IN706 0.22 560 ---- 2 ~ 22 270 270 6. 8k 6.8kI k 270k 0.22 *.002 1Ok< 180k ~L (2)T=,-492 (2)2N1808 OUTPUT I 1900 1 I 1,, I I I N OUTPUT 2 0.47 68 6.8k 6.Bk 3.3k 1.2k 1.2k 470 -457 ~ D Fig. 50 DC Amplifier Module (Analog Modules M3, M4, M7, M10; Digital Module M18)

+ IOV oA 2N1 183 5 +6V t0 20 50 - --- IN710 -- 20 6f 1.2k 390 5 00 f -1V IN457 -15V 2NI808 2 i.N.4L5k7< ~IN716 Fig. 51 Regulated Power Supply Module (Analog Module M5) AUTO SENSE INPUT 30cps 3Vp-p 330 15V-2 J ~~~~~IN4~~~~~~~57 200u.f I 15V ~~~~~~~~~~639 k 82NI404 UCI UTC 0-12 N.02 CRT BLANK15V 8 E 72kc 400mVp-p 100Fig. 52 Auto-Sense Module (Analog Module M8) 0-1 64BLN I00 00 0 5 5. 64~~~~~~~5

TPI I k 1/2W BRN BROWN CT 47l 15 k IkIN 2 RED IN - 2NS 270 EW ORN NE. - X.PGI COUNT BROWN OR-N i E CHANNEL BLACK IN2175 4GND 10k 390 ALL R A NSSTORi.1.40V -3V + IOV TP2 Ik 1/2W RED -15V 47< 15 Q < IN k )~C /270 NS IRED / (4) CHANNEL I' L 457 IN2175 I, 390P, +10 TP$ Ik 1/2W ORN 47 15 1 5 k> k ~~IN 270 EW ORANGE CHANNEL 5fl 0W RE-30 47k~~~~~~~~~~~~~k:N2175 390 (2) < ~~~~~~~~I GND 25W 1.2K +10 03 +10V PG' 2 -15V CHAN i ALL TRANSISTORS 1.4 V 300 MA CHAN 2 2NI404 CYSTOSCOPE CHAN 3 LAMPS Fig. 53 Synchronizing Generator 65

INTEG. TIME I COUNT TIME AUTO- SENSE JC I- 2 10 (ALT. ACT) INTEG JC3-2 o -15vc RESET JC3-3 100, tol fCOJC3-I 100 _____ 80o4 JC3-4-$2T'~8 2 COUNT JCI-6 ci'2 JC-4-32 P RESET INTEG. 270p f I-IOv 100 k 1, 780?1 0 0~~~~ 0 ~~~~~~~~JC4-31 CI-4 0 ~~~~~~~~~~~~~~~~~~COUNT 4 9IET2-IfW 2 RESET r oL2 - -(~ I +.JC3-6 O 1 JC3-4 os ~~~~~~~~~~~~~JC3-5 15 v ~ —oJC 4-24 0~ ~ ~ ~ ~~~~~~~~~~210 100 k e J 150k 15k1S ~~~~~~~~~~~~I% JCI-3 0-100 0-100 AUTO- MAN (a-V+ (ALT. ACT) -15v o JC2 -5 ALL DIODES IN67 10i~~~ALL LAPS 330 o, (10 o ~~~~~~~~~~~~~~~~~~RESET i 2 -, AUTO MAN oJC2-6 JC3- 5 o Ii. or —-I — 5vo o- C c4-24 I 0 lOOk 0 I2 1%~~~~~~~ 20 JC5 6 JCI-5 10 5 AUTO-MAN ~ -15v -~~~oJC 2-5ALDIESNG 9oJC2-I AL AP 3 AUT MAN CFig. 54 Control Unit

UNITS ITE NS HUNDREDS JC4-2 o o JC4-30 JC4-5 o JC4-23 JC4 -10l0 - 0 o -o JC4-22 JC4-I J04- 21 3 4 5 3 4 5 3 4 5 JC4-3 C - JC4-4 I 6 7 8 6 7 8 6 7 8 JC4 -6 o o o o JC4-7 I 9 C 9 C 9 C JC4-9 o o o o o JC4-8 a JC4-24 -15V -, o, Fig. 55 Control Unit - Decimal Readout 67

12 - COND JONES I — BRN I I a2 20 RED 12-COND JONES / 13 ORN o3 j. 4o Y EL GND I- — GRY WH ADVANCE MAG 4 YEL GND WHT-GRY 15- GRN 6E WHT-GRN GRN 390 z a. 50 WHT-BL' BLU ADVANCE rz 60 BLU 8E WHT-BLU 0 3, L IN457 YELI m 70 VLT -24vDC BLK I BLK RED- WHT IWHT 190o6o I10o I I 60c/so I WH T-BRN I 7oBRN BLK BLK CLOCKISII IWHT I I AC YEL-WHT YEL RED 68 n I WHT _ __ 112 I 9 WHTL.L iO 0 10~ IIlRN W IWHT WHT-T - I CHASS15 I ~ 0 |PUSI00II PERFORATOR ~~~~~iPULSE zoIGND -24v I GENER-! —-OI 60 BRN TOR 2 R I R. 2 4I IPL I -z 54- — 1 VYEL IM 6 BLU 6 N -J8 I I I O 8:: GRY I oD 8 ( 19 BLK I A -A IATOR _ _2 RED-WHT I O IADV >___3 _ _ ORN-WHT -24vDC YELONBLK I N I HMRE- 14 Fig. 56 Paper Tape Punch Fig. 56 Paper Tape Punch

LU 2N 1404 8.2 k I N 270' —Jo~~~~~~~~~~~~~~~~~~~~~p 220 k (k 2N 1404 AN1183 330 8 4 840 3.3 k 2NI 183 IN270 z Z Cr~~~~~~~~~~~~~~~~~~~c 00 22k 10k 2NI140 330 2NI404 3.3 k 2N I183 IN 270 o >3 > w Cr 10 ~2N140 330 PULSE PULSE z (9 STAGES IN ALL)' J -i X~~~~~~~~~~~~~~~~~~~. LLJ 0~~~~~~~~~~~~~~~~~J( SOLENOI D DRIVERS AN D PULSE GATES Cr~~~~~~~~~~~~~~~~~~~c 876543,2 8765432 S INPUTS PUNCH SOLS. Fig. 57 Paper Tape Punch - Solenoid Drivers 69

-24v I PUNCH PULSE ADVANCE PULSE Ik 3.3k 56k 3.3k 3.3k 2Lf 3.3k 56k 3.3k 0. IN 500 500 718 2.7 k 27k 27k 1.5k 100 Ok 1 00 k 10Ok.002 lN 470 TAPE ADV. IN IN IN.2703 k 275k20 $3.3k 1 3.3k Ik 27k _ {-7}4470 6Ok t 500 CLOCK INPUT GROUND 6ocps ALL TRANSISTORS 2N1404 Fig. 58 Paper Tape Punch - Pulse Generators

POWER I AMP I I 8 I (4) 1 N 3208 S 2AMP SLO 2 5AMP (MAX) T AMSLOP 224BLO 9 24 3 | 820 o Y24 4 I,, I2k v8 115vIISV~~~~~~~~~AC 5 ~IW I W 11 5 v AC 62 PE J000 I I[ TO 50Fv 6 3+ - -- 716A 11-( I I 50vT* T3 0DOW POWER 68 UO LAMPS RE-17YELLO30 330 __ __J - - - 0+24 60 cps TO AMP Fig. 59 Paper TapePunch- Pn Power Supply -24v YELLOW [/2W 27k 470. 5v 82k 2~L0 N1183 ORANGE.DOWN 2 OLLcu 2.2k \Ik N2N1404 N1404 IN457 JR-I, TRIGGER *4 INPUT Fig. 60 X - Y Recorder Pen Pecker Amplifier 71

9. REFERENCES 1. E. M. Aupperle, T. W. Butler, Jr., and D. L. Mills, Remotely-Controlled Tactical Direction-Finding System, Cooley Electronics Laboratory Technical Report No. 124, The University of Michigan, Ann Arbor, 1961. 2. Busignies and M. Dishal, Some Relations Between Speed of Indication, Bandwidth and Signal-to-Random-Noise Ratio in Radio Navigation and Direction Finding, Proc. IRE, v. 37, pp. 473-488, 1949. 3. R. D. Carlsen, Amplitude, Phase and Time Delay Characteristics for the R390/URR AM Receiver, Cooley Electronics Laboratory Technical Memorandum No. 38, The University of Michigan, Ann Arbor, 1957o 4. C. E. Lindahl, A Linear Phase Bandpass Filter, Cooley Electronics Laboratory, Technical Memorandum No. 82, The University of Michigan, Ann Arbor, 1961. 5. C. E. Lindahl, Study of Input Circuitry of Direction Finder Set AN/TRD-4A, Cooley Electronics Laboratory Technical Memorandum No. 77, The University of Michigan, Ann Arbor, 1960. 6. C. E. Lindahl, System Study Covering an Antenna Suitable for a Spinning Goniometer Direction-Finding System, Cooley Electronics Laboratory Technical Report No. 125, The University of Michigan, Ann Arbor, 1961. 7. W. J. Lindsay and D. S. Heim, Considerations in the Automation of a SpinningGoniometer Radio Direction Finder, Cooley Electronics Laboratory Technical Report No. 85, The University of Michigan, Ann Arbor, 19610 8. Instruction Book for R-390/URR Radio Receiver, Department of the Army Technical Manual TM 11-856. 9, Instruction Book for AN/TRD-4A DF Set, Department of the Army Technical Manual TM 11-688. 72

DISTRIBUTION LIST Copy No. 1-2 Commanding Officer, U. S. Army Electronics Research and Development Laboratory, Fort Monmouth, New Jersey, ATTN: Senior Scientist, Electronic Warfare Division 3 Commanding General, U. S. Army Electronic Proving Ground, Fort Huachuca, Arizona, ATTN: Director, Electronic Warfare Department 4 Chief, Research and Development Division, Office of the Chief Signal Officer, Department of the Army, Washington 25, D. C., ATTN: SIGEB 5 Commanding Officer, Signal Corps Electronic Research Unit, 9560th USASRU, P. O. Box 205, Mountain View, California 6 U. S. Atomic Energy Commission, 1901 Constitution Avenue, N.W., Washington 25, D. C., ATTN: Chief Librarian 7 Director, Central Intelligence Agency, 2430 E Street, N.W., Washington 25, D. C., ATTN: OCD 8 U. S. Army Research Liaison Officer, MIT-Lincoln Laboratory, Lexington 73, Massachusetts 9-18 Defense Documentation Center, Cameron Station, Alexandria, Virginia 19 Commander, Air Research and Development Command, Andrews Air Force Base, Washington 25, D. C., ATTN: SCEC, Hq. 20 Directorate of Research and Development, USAF, Washington 25, D. C., ATTN: Electronic Division 21-22 Hqs., Aeronautical Systems Division, Air Force Command, WrightPatterson Air Force Base, Ohio, ATTN: WWAD 23 Hqs., Aeronautical Systems Division, Air Force Command, WrightPatterson Air Force Base, Ohio, ATTN: WCLGL-7 24 Air Force Liaison Office, Hexagon, Fort Monmouth, New Jersey 25 Commander, Air Force Cambridge Research Center, L. G. Hanscom Field, Bedford, Massachusetts, ATTN: CROTLR-2 26-27 Commander, Rome Air Development Center, Griffiss Air Force Base, New York, ATTN: RCSSLD - For retransmittal to - Ohio State University Research Foundation 28 Commander, Air Proving Ground Center, ATTN: Adj/Technical Report Branch, Eglin Air Force Base, Florida 29 Chief, Bureau of Naval Weapons, Code RRR-E, Department of the Navy, Washington 25, D. C. 30 Chief of Naval Operations, EW Systems Branch, OP-35, Department of the Navy, Washington 25, D. C. 31 Chief, Bureau of Ships, Code 691C, Department of the Navy, Washington 25, D. C. 32 Chief, Bureau of Ships, Code 684, Department of the Navy, Washington 25, D. C. 73

Copy No. 33 Chief, Bureau of Naval Weapons, Code RAAV-33, Department of the Navy, Washington 25, D. C. 34 Commander, Naval Ordnance Test Station, Inyokern, China Lake, California, ATTN: Test Director - Code 30 35 Director, Naval Research Laboratory, Countermeasures Branch, Code 5430, Washington 25, D. C. 36 Director, Naval Research Laboratory, Washington 25, D. C., ATTN: Code 2021 37 Director, Air University Library, Maxwell Air Force Base, Alabama, ATTN: CR-4987 38 Commanding Officer - Director, U. S. Naval Electronics Laboratory, San Diego 52, California 39 Office of the Chief of Ordnance, Department of the Army, Washington 25, D. C., ATTN: ORDTU 40 Commanding Officer, U. S. Naval Ordnance Laboratory, Silver Spring 19, Maryland 41-42 Chief, U. S. Army Security Agency, Arlington Hall Station, Arlington 12, Virginia, ATTN: IADEV 43 President, U. S. Army Defense Board, Headquarters, Fort Bliss, Texas 44 President, U. S. Army Airborne and Electronics Board, Fort Bragg, North Carolina 45 U. S. Army Antiaircraft Artillery and Guided Missile School, Fort Bliss, Texas 46 Commander, USAF Security Service, San Antonio, Texas, ATTN: CLR 47 Chief, Naval Research, Department of the Navy, Washington 25, D. C., ATTN: Code 931 48 Commanding Officer, 52d U. S. Army Security Agency, Special Operations Command, Fort Huachuca, Arizona 49 President, U. S. Army Security Agency Board, Arlington Hall Station, Arlington 12, Virginia 50 The Research Analysis Corporation, 6935 Arlington Rd., Bethesda 14, Maryland, ATTN: Librarian 51 Carlyle Barton Laboratory, The Johns Hopkins University, Charles and 34th Streets, Baltimore 18, Maryland 52 Stanford Electronics Laboratories, Stanford University, Stanford, California, ATTN: Applied Electronics Laboratory Document Library 53 HRB - Singer, Inc., Science Park, State College, Pennsylvania, ATTN: R. A. Evans, Manager, Technical Information Center 54 ITT Laboratories, 500 Washington Avenue, Nutley 10, New Jersey, ATTN: Mr. L. A. DeRosa, Div. R-15 Lab. 55 Director, USAF Project Rand, via Air Force Liaison Office, The Rand Corporation, 1700 Main Street, Santa Monica, California 74

Copy No. 56 Stanford Electronics Laboratories, Stanford University, Stanford, California, ATTN: Dr. R. C. Cumming 57 Stanford Research Institute, Menlo Park, California 58-59 Commanding Officer, U. S. Army Signal Missile Support Agency, White Sands Missile Range, New Mexico, ATTN: SIGWS-EW and SIGWS-FC 60 Commanding Officer, U. S. Naval Air Development Center, Johnsville, Pennsylvania, ATTN: Naval Air Development Center Library 61 Commanding Officer, U. S. Army Electronics Research and Development Laboratory, Fort Monmouth, New Jersey, ATTN: U. S. Marine Corps Liaison Office, Code AO-C 62 Director, Fort Monmouth Office, Communications-Electronics Combat Development Agency, Bldg. 410, Fort Monmouth, New Jersey 63-71 Commanding Officer, U. S. Army Electronics Research and Development Laboratory, Fort Monmouth, New Jersey ATTN: 1 Copy-Director of Research 1 Copy- Technical Documents Center ADT/E 1 Copy- Chief, Special Devices Branch, Electronic Warfare Div. 1 Copy- Chief, Advanced Techniques Branch, Electronic Warfare Div. 1 Copy- Chief, Jamming and Deception Branch, Electronic Warfare Div. 1 Copy File Unit No. 2, Mail and Records, Electronic Warfare Div. 3 Cpys Chief, Security Division (For retransmittal to - EJSM) 72 Director, National Security Agency, Fort George G. Meade, Maryland, ATTN: TE C 73 Dr. B. F. Barton, Director, Cooley Electronics Laboratory, The University of Michigan, Ann Arbor, Michigan 74-97 Cooley Electronics Laboratory Project File, The University of Michigan, Ann Arbor, Michigan 98 Project File, The University of Michigan Office of Research Administration, Ann Arbor, Michigan 99 Bureau of Naval Weapons Representative, Lockheed Missiles and Space Co., P. O. Box 504, Sunnyvale, California - For forwarding to - Lockheed Aircraft Corp. 100 Lockheed Aircraft Corp., Technical Information Center, 3251 Hanover Street, Palo Alto, California Above distribution is effected by Electronic Warfare Division, Surveillance Department, USAELRDL, Evans Area, Belmar, New Jersey. For further information contact Mr. I. O. Myers, Senior Scientist, Telephone 5961262 75

ERRATA SHEET FOR TECHNICAL REPORT NOo 137 COOLEY ELECTRONICS LABORATORY 16 Equation (4o 6a) should read T T ea = 1/2 f A cos e dt + 1/2 A cos (2gt - ) dt, 0 0 g 20 In Eqo (4~ 13) the term R2C should read RC 21 Line 10 should read + sin (w rT- 0) sin (2k + 1) ogr dt, 23 In line 15, replace "Section 5o 2'" with "Section 5o 4"o 32 In line 25, replace "Eqo (40 5)" with "Eq. (4.4)". 38 In line 19, replace "Section 3., 2. " witi "Section 4o 3o 2".