7274-3-T TH UNIVERSITY OF MICHIGAN COLLEGE OF ENGINEERING DEPARTMENT OF ELECTRKCAL ENGINEERING Radiation laboratory SIMPLIFIED MODELING TECHNIQUES FOR AVIONIC ANTENNA PATTERN SIGNATURES Interim Technical Report No. 3 15 October 1965 - 15 January 1966 W. DeHart, J. Ferris and R. Wolford ll'iB. -.— ll I February 1966 7274-3-T = RL-2150 Contract AF 33 (615)- 2606 Task 435703, Project 4357 Contract Monitor: K. W. Tomlinson AVWE Contract With: Air Force Avionics Laboratory Research and Technology Division Air Force Systems Command Wright-Patterson Air Force Base, Ohio 45433 Administered through: OFFICE OF RESEARCH ADMINISTRATION * ANN ASIII

THE UNIVERSITY OF MICHIGAN 7274-3-T TABLE OF CONTENTS Page ABSTRACT ii I INTRODUCTION 1 II SIMPLIFIED MODELING 3 2.1 Fly-By Tests 3 2.1.1 Fly-By Tests WPAFB 3 2.1.2 Fly-By Tests RADC 10 III DATA RECORDING AND REDUCTION TECHNIQUES 14 IV A SYSTEM ASPECT 15 4.1 Source Impedance Measurement 15 4.2 Complex Impedance Load 21 4.3 AN/ARC-27 Study 25 i

THE UNIVERSITY OF MICHIGAN 7274-3-T ABSTRACT To further justify the use of simplified modeling, it has been suggested that consideration be given to the use of fly-by tests (employing an instrumented aircraft) to obtain patterns of a typical airborne antenna at the fundamental, spurious and harmonic frequencies. A discussion of two techniques that may be employed to obtain fly-by tests is presented in this report. Spectrum signature data has been collected on the AN/ARC-27 transmitter employing three measurement systems and a discussion of the results is included. In addition, techniques for obtaining the output impedance of the transmitter have been further considered and a discussion of the problems is included in this report. ii

THE UNIVERSITY OF MICHIGAN 7274-3-T I. INTRODUCTION The present study is divided into three parts, each to be concerned with the determination of the spectrum signature of airborne electronic systems. The three areas being considered during this study are: 1) practical methods for obtaining the desired antenna pattern spectrum signature at a minimum cost, 2) data recording and reduction techniques, and 3) the system aspect of the problem. To obtain antenna pattern spectrum signatures consideration has been given to the use of simplified models rather than precision models or an instrumented airborne system. An extensive experimental study has been conducted investigating the feasibility of the simplified modeling technique and the results have been very encouraging. To further justify the use of the simplified model, it has been suggested that consideration should be given to the use of fly-by tests (employing an instrumented full-scale aircraft) to obtain full-scale spectrum signature data at the fundamental, spurious and harmonic frequencies to be further compared with the simplified model data. Techniques for recording spectrum signature data of antennas and for reducing the data to statistical format are also being investigated during this study. Data recording techniques have been centered around the use of digital recording systems so that data can be later reduced into a statistical format to aid in prediction and analysis. A technical report of the data recording and reduction pro 1

THE UNIVERSITY OF MICHIGAN 7274-3-T cedures is being prepared. During the system portion of the study, consideration is being given to methods for determining the output impedance characteristics of the transmitter at the fundamental, spurious and harmonic frequencies. The purpose of this study is to determine a practical method for obtaining the output impedance characteristic of the transmitter so that more accurate interference predictions can be made of airborne systems. A justification for this study is based on the inaccuracies that have been noted in the interference prediction analysis made of airborne systems. Further, the data collected will be more basic in nature and therefore be applicable to a wider range of antenna and transmitter configurations. 2

THE UNIVERSITY OF MICHIGAN 7274-3-T II. SIMPLIFIED MODELING In the second interim it was noted that the study of simplified modeling was drawing to a conclusion. To insure that all areas pertinent to obtaining airborne antenna spectrum signatures have been properly investigated, further consideration has been given to the use of a full scale aircraft. The full scale aircraft antenna study would necessitate the use of fly-by tests. From these signature data would be obtained for comparison to simplified model data. 2.1 Fly-By Tests To ensure that adequate fly-by test data would be obtained, a simplified model of the aircraft to be flown during the fly-by tests would be required, and a set (at all frequencies of interest) of three dimensional antenna pattern data collected. Further, it would be necessary for the antenna used on the simplified model to be similar in configuration and location to that used on the full scale model. To carry out the fly-by tests two facilities have been investigated and will be discussed in the following sections of the report. 2.1.1 Fly-By Tests WPAFB Consideration was at first given to the use of the Trebein facilities at Wright Field. A factor in favor of the use of the Trebein's facilities is a 40 foot parabolic reflector located there. In addition to the dish, several HF receivers and auto"Simplified Modeling Techniques for Avionic Antenna Pattern Signatures" Interim Technical Report No. 2, 15 July - 15 October, 1965, W. DeHart J. Ferris and R. Wolford, under Contract AF-33(615)-2606 3

THE UNIVERSITY OF MICHIGAN 7274-3-T matic recording systems are available. In addition to these facilities two UHF receiver types would be required similar to those manufactured by the Stoddard and Empire Device Corporations. The two UHF receivers noted above have an input sensitivity of 10 micro volts in the UHF band which is equivalent to -87 dbm into a 50 ohm load. The gain and pattern characteristics of the 40 foot parabolic reflector are shown in Table I. To obtain an insight into the transmitter power levels that may be associated with an ARC-27 transmitter, the data presented in Table II has been extracted from several reports from the Army Electronic Proving Ground, Fort Huachua, Arizona entitled Spectrum Signature Data of Radio Set AN/ARC-27. The slant range from the airborne antenna to the 40 foot parabolic receiving antenna was calculated from the geometry of the problem shown in Fig. 1. Using the above parameters and the Friis transmission formula, the general expression for the received power may be determined as shown below. PGA p. t t e 47r S where P Transmitter Power t Gt Transmitter Antenna Gain A Effective Aperture of Receiver A = k r r e k = 0.55 r = Radius of Antenna (20 ft) S = Slant Range= 11, 000 feet such that P G kr2 Ptt 4S r 4S 2 4

THE UNIVERSITY OF MICHIGAN 7274-3-T TABLE I 2 r = 40' Dish Electrical Characteristics TABLE II ARC-27 Spurious Emissions (f = 312.5 MHz) o S.N. 1167 Freq. Pwr. Pwr. (dbm) (watts) I - - I f 0 2f 0 3f 0 4f 0 5f 0 +42. 0 +13. 0 -3.0 -7. 0 -2.5 15.85.02.0004.0002.0006 S.N. 3410 Freq. Pwr. Pwr. (dbm) (watts) f +37.5 5.5 o 2f +15. 0.031 0 3f -3.0.0005 0 4f -6.5.0002 0 5f -10.0.0001 o S.N. 4589 Freq. Pwr. Pwr. (dbm) (watts) f +40.0 10.0 o 2f +15.0.031 0 3f -2.5.0006 0 4f -10.0.. 0001 0 5f -20. 0. 00001 o I. 5

THE UNIVERSITY 7274-3-' OF MICHIGAN x w s h D 2 2 1/2 S=(h +D ) - 11 (10) where h = 10, 000 feet D= 5,280 feet w = S a - 230 feet a '1.2~ where w X = 30 cos 30 265 feet FIG. 1: Geometry of Problem 6

THE UNIVERSITY OF MICHIGAN 7274-3-T and P (db) = P (db) + Gt (db) + 20 log R - (6 db + 20 log S) r t t = Pt (db) + Gt (db) -66 db To obtain the directivity of the transmitting antenna use has been made of three dimensional pattern data for the modified monopole mounted on a simplified model of the T-33 aircraft. The gain of the modified monopole was obtained by reducing its directivity by the appropriate missmatch (VSWR) loss factor. The VSWR data was collected with the modified monopole over a 4 foot circular ground plane. The directivity, VSWR and gain data are shown in Table III. From the geometry of the problem shown in Fig. 1, it can be deduced that the pattern recorded during the fly-by tests will be a conical pattern. The included half angle will be 30 as measured from the normal to the underside of the aircraft. Therefore, it has been necessary to modify the transmitting antenna gain figure to account for this 30 angle. To obtain a more realistic transmitted power level, the data of Table II was averaged. The transmitter power level and antenna gain are noted in Table IV along with the probable power received. The last column of Table IV denotes the margin of safety between the receiver sensitivity (-87 dbm) and the actual power available at the output terminals of the 40 foot reflector. From Table IV it can be seen that at the fundamental frequency there will be approximately a 58 db margin of safety between the receiver sensitivity and the power available at 7

THE UNIVERSITY OF MICHIGAN 7274-3-T TABLE III Modified Monopole Electrical Characteristics (f = 300 MHz) o Frequency VSWR Directivity Gain (Gt) f 2.6:1 6.1db 5.2db 0 f 1.8:1 7.6db 7.1db f3 5.0:1 6. 5 db 4. 0 db f4 7.0:1 6.1db 2.6db f5 10.0:1 7. 5 db 3.0db f5 TABLE IV Predicted Power Received (30 From Normal to Aircraft) r =(P +Gt)- 66 db r t t Frequency P G P A P (dbm) (db) (dbm) (dbm5 f 40 -3 -29 58.0 0 f2 14 -3.5 -55.5 31.5 f3 -3 -0.3 -69.3 17.7 f4 -8 1.3 -72.7 14.3 f -18 2.0 -82 5.0 I..... 8

THE UNIVERSITY OF MICHIGAN 7274-3-T the output of the 40 foot reflector. However, for the second through fifth harmonics, the power level will be but a few db above the minimum detectable signal of the receiver. Since it will be desirable to record patterns both at the fundamental and harmonic frequencies, it will be advantageous to use a more sensitive receiver such that a greater dynamic range will be possible at the harmonic frequencies. To ensure that the receiving system has an adequate dynamic range so that a sufficient amount of useful data will be collected, two RFI receivers will be required to collect data at the fundamental and second harmonic frequencies. In addition to the two RFI receivers, a Wide Range High Sensitivity Receiver is recommended for obtaining data at the third or fourth harmonic frequency of the AN/ARC-27 transmitter. Wide Range High Sensitivity Receivers typically have a sensitivity of -100 dbm such that a safety factor of 31 - 43 db would be available at the higher harmonics. In the above discussion, it has been assumed that the antenna on the aircraft will be flying directly through the pattern maximum of the 40 foot reflector. However, this is not probable since the RF window through which the aircraft must fly will be approximately 230 by 265 feet as may be concluded from Fig. 1. Further, we have assumed that the radiation pattern of the 40 foot reflector is triangularly shaped (polar coordinates), which if true (as we know it is not), the aircraft would have to be flown precisely along the plotted course to insure that it would pass through the radiation pattern of the 40 foot reflector each time it passed over the ground 9

THE UNIVERSITY OF MICHIGAN 7274-3-T reference point. An additional limitation is that the aircraft will be within the RF window (of the 40 foot reflector) for less than a second if we assume it is being flown 250 mph (360 feet per second). Since the radiation characteristics of the 40 foot reflector are not triangularly shaped but rather have a parabolic configuration, the RF window may be larger. For example, assuming the 10 db beamwidths (rather than the 3 db beamwidths of Table I) the aircraft will be within the RF window for approximately 3 seconds which would be more desirable. Because of the limitations associated with using the Trebein facilities consideration was next given to the use of the Rome Air Development (RADC) facilities. 2. 1. 2 Fly-Byv Tests RADC Three techniques are employed at the Rome Air Development Center (RADC) to obtain antenna patterns from an airborne antenna. The technique used normally is similar to that outlined in the Final Report prepared under Contract AF 33(615)1964. It will be recalled in the above report a reference point was located on the ground and the aircraft flown over it from several points of the compass and pattern data obtained on a point-by-point basis. A second technique employed by RADC also employs a ground reference point to have an aircraft flown across the point tangentially to it such that data is collected continuously as the aircraft flies past. Sufficient data is collected so that the pattern may be reconstructed. The third technique also requires a reference point on the ground. The aircraft is flown along a radial that extends from the ground test site through the ground 10

THE UNIVERSITY OF MICHIGAN 7274-3-T reference point. Data is collected continuously as the aircraft is flown from several miles away from the ground test site directly overhead to several miles in the opposite direction from the test site and reference point. The principal difference between the technique employed by RADC and that noted in the Final Report of AF33(615)-1964 is that RADC uses a radar system to continuously track the aircraft. An antenna pedestal is slaved to the radar tracking antenna such that the pedestal follows the movement of the radar antenna in both azimuth and elevation. Receiving antennas are mounted to the slaved antenna pedestal and track the aircraft such that data is collected continuously. Up to 8 antennas may be mounted on the pedestal at one time. However, for the purposes of our tests only 3 antennas will be required. The frequency range of this system is 100 MHz to 10 GHz. The radar system used is an MSQ auto tracker which has better than 10 accuracy. A second system that may be used is an MSQ 63, which is operational in the frequency range of 250 MHz to 40 GHz. However, when using this system it is not possible to track the aircraft (with radar) as it is using the MSQ auto tracker and, therefore, the accuracy would be less. Presently two KC-135 aircraft are instrumented to obtain radiation patterns of airborne antennas, and two C-131 aircraft are instrumented to obtain radiation patterns of ground based radar systems. For the present task, only one of the KC-135 aircraft would be required. The KC-135 ins a Military version of the 11

THE UNIVERSITY OF MICHIGAN 7274-3-T commercial Boeing 707. Each of the KC-135ts used by Rome has from 30 to 40 antennas mounted on the underside of the fuselage. The system employed to obtain antenna patterns using an instrumented aircraft requires three facilities: 1) the airborne, 2) radar, and 3) the receiving facility. The three facilities are inter-connected through radio communications to facilitate communicating between them. The procedure is for the radar system to track the aircraft and to provide ground control to ensure the aircraft is flown over the point of interest. The receiving system monitors the data as the aircraft approaches the point and if the reference point is missed by too great a distance, the operator may request the aircraft to be flown over the data point again at the completion of that test flight. The test procedure requires that the aircraft be flown over the point of interest a sufficient number of times to obtain 24 data points (15 increments). At the present time RADC has selected three points on the ground as reference points. These points have been chosen so that omnidirectional patterns may be recorded at elevation angles of 5, 80 and 120 below the local horizon of the aircraft. The aircraft is generally flown at an altitude of 35, 000 feet, such that the ground reference points are several miles from the test site. Typically the aircraft can be flown within 500 feet of the selected ground point which is considered to be adequate for obtaining antenna patterns from an instrumented aircraft. 12

THE UNIVERSITY OF MICHIGAN 7274-3-T To obtain a single antenna pattern (employing the KC-135) would require approximately five hours of flight time. Therefore, to obtain the data required for the simplified modeling study, a conservative estimate of 50 to 100 hours of flight time would be required. 13

THE UNIVERSITY OF MICHIGAN 7274-3-T III. DATA RECORDING AND REDUCTION TECHNIQUES In the previous interim, it was recommended that a technical report be prepared on the data handling techniques developed by Michigan. Preliminary work on this report is progressing and it is anticipated that the report will be completed prior to the conclusion of the present contract. No significant changes have been made since the writing of the previous interim and therefore no further discussion will be given at this time. 14

THE UNIVERSITY OF MICHIGAN 7274-3-T IV. A SYSTEM ASPECT The system study is being continued, employing the AN/ARC-27 transmitter. Additional spectrum signature data has been collected employing both a Polarad and a Hewlett-Packard spectrum analyzer as well as two Empire Device RFI meters. Three sets of data have been obtained using these instruments. It is, however, difficult to correlate the results at the present time. Additional studies have been conducted to determine the output impedance of the transmitter. Approaches being considered are Reiche diagrams, cold tests and techniques for measuring the absolute output impedance of the transmitter. 4.1 Source Impedance Measurement In the first interim report, the power transfer from a transmitter to an antenna was shown to require measurement of the internal impedance of the transmitter. Here it will be shown that, if the transmitter impedance is not a function of the termination, the impedance can be measured with a line stretcher terminated in a short circuit and a detector tuned to the frequency of interest. The measurement procedure is very similar to that for measuring a termination impedance with a slotted line. The circuit used is shown schematically in Fig. 2. The transmitter is represented by a voltage source E and a series impedance Z. 5 - 15

THE UNIVERSITY OF MICHIGAN 7274-3-T Trar s -litter z Transmission Line s Char. Imp. = R 100 I A Short Circuit Es V Detector V ____B__________________________ _______ <._ —_____? --- ——,~ FIG. 2: Measurement Circuit Neither E nor Z are a function of the termination. The transmitter ters s minals AB are connected to a transmission line of variable length, 1. A detector, tuned to the frequency of interest, is placed across the line a fixed distance, d, from the short circuit. The exact value of d is not important but should remain fixed throughout a measurement. The largest voltage for the detector is obtained when d ' X/ 4, 3A/4, etc, but the effect of detector loading is most pronounced at these points. For a measurement device, d -X/ 8 may be a good compromise. The impedance Z seen at the input of a shorted transmission line of length I and propagation constant k is given by: Zt =jX =jRR tan kI (1) 16

THE UNIVERSITY OF MICHIGAN 7274-3-T This is the impedance seen looking to the right of terminals AB of Fig. 6. The voltage V1 across terminals AB is given by: V1 =1Iijxt (2) E |IX s (3) Z + jX let Z R + jX (4) then E IX I R2S + (X + Xt ) E R tan ktl = s ' (6) 2 2 \R (X +R tan k)2 Because the transmission line is shorted at the far end, an infinite sinusoidal standing wave appears on the line such that V =V (sinkl) (7) The quantity V represents the peak value of the standing wave appearing at every odd number of quarter wavelengths from the short. The voltage, V, indicated by the detector will be some fixed fraction, a, of this peak voltage. Hence, 17

THE UNIVERSITY OF MICHIGAN 7274-3-T V = aV p (8) Combining equations 6, 7, and 8, we obtain V I sin kI a E R tan kif R s + (X+R tank VRS + (X + R tan kf)2 s S S 0 (9) from which aE S V = R COS kl + ( F- cos ki + (- cos kl + sin kt) L| J0 0 (10) In equation 10, the magnitude signs are no longer necessary; the positive sign is attached to the radical in the denominator. Now let us examine equation 10 for the condition of R =R and X = 0. s o s Equation 10 becomes V R =R s X =0 s aE s ~ Mcos 2k+ sin kt (11) = aE s (12) Thus when the source impedance matches the line impedance (Z = R ) the detector reading is independent of the line length!. This condition is equivalent to 18

THE UNIVERSITY OF MICHIGAN 7274-3-T the matched termination of a line that produces a flat line. The flat line counterpart for a "source standing wave" is demonstrated by equation 12. Now what happens when Z A R, i.e., the source impedance is not matched. 5 O For this situation, let us consider, in general, the situation viewed at terminals AB but at terminals A'B' a distance I1 from the transmitter. Let us choose 1 such that the transmitter impedance Z is transformed to a pure resistance Rs s s so that R' > R. Equation 10 then becomes s- o aE V=s (13) RI \2 S k Cos k(l-1 )k+ sin2 k(Equation 13 has a maximum value when cos k( -11) - 0 and a minimum value when sin k (t - ) = 0. To show this, differentiate the quantity under the radical 1 and set it equal to zero. d2 d [k( )] C os k(- + sin k(I ) 11 1 1(14) = -2 soSink(f -l)cosk( -1)- 2 sin k s k( -l)COSk ) (14) 1 1 = 2 [1 - sin k (l-t ) cos k(l - ) (15) Lo Equation 15 is zero when sin k (i - ) 0 (16) (16 19

THE UNIVERSITY OF MICHIGAN 7274-3-T or cosk(l-I )=0 (17) Using equation 16 in 13 we have R V =aE' o (18) mmin s R' s Using equation 17 in 13, we have V a E t (19) max s Equation 18 gives V. because R < R m5n s o Now, dividing equation 18 by equation 19 gives V R' max s (20) (20) V R min o To determine the impedance seen looking back into the transmitter terminals AB, we must enter the Smith Chart at R' V s = VSWR R V o min and go around the chart a distance equivalent to I1 in the direction "toward the load" indicated on the chart. This point on the chart represents the transmitter impedance Z. s 20

THE UNIVERSITY OF MICHIGAN 7274-3-T 4. 2 Complex Impedance Load In the previous section it was assumed that the transmitter impedance was not a function of the load. In this section consideration will be given to techniques for fabricating a laboratory load whose impedance is matched to the transmission line at f and unmatched at harmonics of f. 0 o FIG. 3: Circuit for Terminating a Coaxial Line Let Y be matched to the characteristic admittance, G, of the line, i. e., 1 o Now, if at the fundamental frequency(21) Now, if at the fundamental frequency Y = -Y 2 3 (22) the termination of the line is Yt Y1 Y2 Y3 (23) 21

THE UNIVERSITY OF MICHIGAN 7274-3-T =Y =G 1 o (24) Now what is the condition that satisfies equation 22? If Y2 and Y3 are obtained from shorted transmission lines of lengths 2 and 3 respectively, then 2u 3 Y2 = jG tan k2 Y3 jG tan k! 3 o 3 Hence, to satisfy equation 22 j G tan k2 = -jG tan ki o 2 o 3 or tan kl = - tan k3 2 3 (25) (26) (27) (28) From trigonometry, Hence tan A = -tan( r -A); if 0 <A<?r tan kl =tan(7r - kL ) 2 2 (29) (30) or from which kl = r - kl3 S +I _ = -=2 3 k 2 22 (31) (32)

THE UNIVERSITY OF MICHIGAN 7274-3-T At the desired harmonic frequency, either ~2 or 13 (or both) must be one half wavelength long for the load conductance at the junction Y1 to be effectively short circuited. Since equation 32 shows that either length A2 or 13 can be arbitrarily chosen, let 12 be assigned a value (aA)/2 at the desired harmonic frequency. If a is an integer 0 < a < n, then 2 will short circuit Yi. Thus aX 2 2 (33) in which n 2, 3, 4,... as desired. Since a X (34) n n equation 32 becomes a-= (35) 2 2n and (n - a) (36) 3 2n Figure 4 illustrates one technique for constructing this load. K- /6 -j- 1/12-4 4 S -ortCir)-i")-ortCircui ~S.ortCiruit Short Circuit / / I I \ Matched Termination FIG. 4: Load Construction 23 iding Section

THE UNIVERSITY OF MICHIGAN 7274-3-T The sliding section need slide only X/ 12 to the left of the center of the shorted section. In operation, the sliding section would be adjusted to satisfy equations 35 and 36 at the harmonic frequency of interest. An alternate load is shown in Fig. 5. Matched Termination One-half Wavelength Loop of Coax Line Double Tee Joint FIG. 5; Alternate Load The load of Fig. 5 has not been analyzed thoroughly, but appears to be automatically adjusted for all harmonic frequencies. It has two advantages over the construction of Fig. 4. First, no adjustment is necessary for the various harmonics, and second the load is easily changed (by replacing the half wavelength loop) to another fundamental frequency. 24

THE UNIVERSITY OF MICHIGAN 7274-3-T 4.3 AN/ARC-27 Study The purpose of the investigation was to scan the spectrum, locate, measure and record the spurious emissions of the transmitter under test. The measurements were performed as a closed-system test, i. e., the output power from the transmitter was fed to a medium power, 50 ohm dummy load. A signal sampler (directional coupler) was used to couple a small portion of the output to a rf receiver. Four rf receivers were employed during the testing. Initially, a Polarad Spectrum Analyzer, Model TSA-S, was employed as the receiver in the test set-up as shown in Fig. 5. The transmitter was tuned to a fundamental frequency f = 300 MHz for this test and all subsequent tests. The spectrum was scanned by the analyzer from 10 - 1500 MHz. This frequency range was chosen so as to encompass all emissions up to and including the fifth harmonic spurious emission. The signals observed on the spectrum analyzer are listed in Table V. It is to be emphasized that the only signals that appeared on the analyzer that were felt to have their origin in the transmitter output were the fundamental and harmonic frequency emissions. The origin of all the other "apparent" signals listed in the table (referred to as "tunreal") can be shown to be due to the spectrum analyzer, i. e., spurious responses of the measuring equipment itself. It is to be noted that these "unreal" responses were not image signals. Their generation within the analyzer might have been prevented by the insertion of appropriate frequency rejection networks or filters ahead of the analyzer, but these devices were 25

TABLE V Spurious Frequency Origin of Signal Lower Upper Dial Dial 71.0 - - - Unreal 78.0 | - Unreal 94.5! 350 Unreal (Both)* 106! 372 Unreal (Both) 118 396 Unreal (Both) 174 509 Unreal (Both) 197 555 Unreal (Both) 224 606 606 2 f - 259 678 Unreal (Both) 303 765 303 = fn 375 909 -909=3!f 398 955 Unreal (Both) 965 2090 Unreal (Both) 1114 2386 Unreal (Both) 1193 2548 1193 = 4f,0 1375 2910. Unreal (Both) 1388 2938 Unreal (Both) 1414 2988 Unreal (Both) 1449 3055 Unreal (Both) 1475 3110 Unreal (Both) 1490 3140 1490 = 5f _ * Either the upper or lower dial reading gives correct frequency. Determination of frequency of ~"unreal" signals not shown. A 26

THE UNIVERSITY OF MICHIGAN 7274-3-T not immediately available. Also, since there appeared to be no convenient means for measuring the relative power level of signals observed on the Polarad Analyzer, no results are given. It is therefore recommended that the above spectrum analyzer not be used for spurious emission measurements unless appropriate filters are available and the results are viewed with some suspicion. A second test was conducted with the measurement set-up shown in Fig. 5, but with a Hewlett-Packard Spectrum Analyzer, Model 851A/8551A, used as the receiver. The frequency range scanned was again 10 - 1500 MHz, with the transmitter again tuned to 300 MHz. The results of this test, shown in Table VI, indicated the presence of two non-harmonic spurious emissions at frequencies of 160 and 660 MHz in addition to the fundamental and four harmonic emissions. The relative power level of the signals observed on the analyzer were recorded and referenced to the fundamental signal. The fundamental power output was determined absolutely by means of a power meter. The coupling factor of the signal sampler and the insertion loss of the fixed attenuators were assumed flat across the frequency range of interest. It is felt the Hewlett-Packard Spectrum Analyzer yielded a convenient, first-order approximation of the spectrum produced by the transmitter. The final spurious emission measurements were conducted with more refined measuring equipment and a test set-up as shown in Fig. 6. The test procedure followed that outlined in section 5. 2. 3. 3 of MIL-STD-449 B. Two frequency 27

Power Meter FIG. 5: Initial Spurious Emission Measurement Block Diagram r ------ - - --- Transmitter_______ I T-178/ARC-27 I | Serial NQ 20015 RG-9B/ Signal DummySampler Load I. I Fixed. Shield d Attenuators Enclosure RG-9A/U Frequency Voltmeters RG-9B/U * Generators L RG-9B/U FIG. 6:: Final Spurious Emission Measurement Block Diagram 28

TABLE VI Spurious Origin Spurious Power Frequency of Output (Average) (MHz) Signal (dbm) 160 -12.0 300. +44. -600 2fo +17.0 660 -8.0 900 3f -10 0 1200 4f -8.0 1500 5fo +12.0 jo 29

THE UNIVERSITY OF MICHIGAN 7274-3-T selective voltmeters, Empire Devices Models NF-105 and NF-112, were employed as rf receivers. The results obtained with these high sensitivity meters are shown in part in Table VII. The table lists the significant spurious emissions observed between 300 and 1500 MHz, namely, four harmonic emissions and one non-harmonic emission at 658 MHz. The power output of each spurious signal was determined by duplicating the voltmeter response to the spurious signal with a signal generator substituted for the transmitter. The power output was then given by the sum of the signal generator output dial reading in dbm and the signal sampler coupling factor and attenuator insertion loss in db at the frequency of interest. Also shown in Table VII is a selected group of low power spurious emissions from 240. 0 to 285.0 MHz. This group is representative of the transmitter emission spectra that appeared to exist over much of the frequency range scanned, i e., 20 - 1500 MHz. It is not felt that the insertion of a fundamental frequency rejection network would alter the results as shown. The entire frequency range was not scanned for the low-level emissions due to a limitation of testing time and the lack of certainty as to the importance of these signals. It was evident, however, that if the importance of the low-level emissions is established, a more thorough investigation with frequency selective voltmeters will be required to determine the exact nature of the transmitter spectrum. 30

TABLE VI Spurious Origin Attenuation Signal Spurious Power Frequency of Inserted Generator Output (Average) (MHz) Signal (db) (dbm) (dbm) 240.0 __ 20 -35. 0 -15 255.0 __20 -53.0 -33 260. 0 20 -59. 0-39 278.0 20 -74.0 -54 280.0 20 -60.0 -40 285. 20 53.0 -233 300.0 f 50.0 -5.0 +45.0 600.0 2f 21.0 -10.0 +11.0 O.__________________________________ 658.0 21.0 -55.0 -34.0 900.0 3f 20. 0 -17.5 +2.5 1200.0 4f 20.5,-29.0 -8. 5 1500.0 5f 20.5 -15.0 +5.5 0 I 31