AIR FORCE REPORT NO. AEROSPACE REPORT NO.SAMSO-TR-72-178 TR-0073(3404) -1

1.,9:1 Bandwidth UHF 3 x 3 Array ofCrossed Open-Sleeve Dipoles

Prepared by H. E. KING and j. L. WONG"Electronics Research Laboratory

Laboratory Operations

72 JUN 30

A; Sy' tns Engineering Operations

THE itEROSPACE CORP)RATION

Prepared for SPACE AND MISSILE SYSTEMS ORGANIZATION:AIR FORCE SYSTEMS COMMAND

LOS ANGELES AIR FORCE STATION'Los Angeles, California

Reopodvced byNATIONAL TECHNICALINFORMATION SERVICE

u S em, of c . .... ER CESPI-0-eld VA 22151 D

APPROVED FOR PUBLIC RELEASE:

DISTRIBUTION UNLIMITED JUL RA 19

1A -.z

UNCLASSIF'IEDSecurity Clas'sification

1.ORGIA IN A VVOTY(Copt~* utor 12. RIEPORIT SECURITY CLASSIFICATION

;The Aerospace Co~rporation Ucasfe

3. NEPONRT TITLE

1. 8: 1 BANDWIDTH UHF 3;X.3 ARRAY OF CROSSED OPEN-SLEEVEDIPOLES

4 OuC RIP To Vi NCO~TE5 (Typo of rport ahd Inclusivedittoe)

S U TORS) (First nlam, middle initial, last name)

Howard E. Kin~g and Jimmy L. Wong

P EPRTDAE7a TOTAL NO. OF PAGES1 7b. NO. OF NEFS

72_____JUN _______30__ 49ý 4

Be CONTRACT ON GNANT NO. 98 ONIGINAT'3Nq REPORT NUMBER(S)

F047Oi-72-C-0073 TR.-0073(3404)-ib. PROJECIT NO.

c this OTeRNpor t NOS(Aythrumtatamy eeiic

d SAMSO-TR-7Z- 17810. CiSTNIIUUJTION STATEMENT

Approved for'pu1blic'release; distribution unlimited

It. SUPPLEmlENTAAY NOTES 12. SPONSONING MILITARY ACTIVITY

Space and Missile Systems OrganizationAir Force Systems CommandLos Angeles, California

11.' AOSTRACT

'A circularly polarized nine-element (3J X 3) planar array antenna systenihas been evaluate-! foi use on a synchronous satellite. The array elementsare mounted on' a 104-in, square reflector, and the basic radiating elementis a crossed open-sleeveý dipole witH good VSWR and pattern characteristicsover a 1. 8:1 frequency band (225 to 400 MHz). The electrical pe~rfoxmnanceof the antenna was measured using a half -scale model. It is shown that theante'nna can'provide an EOE gain (gain in the direction of the edge-of-the-earth) of greater than 14 dB3 from ..25 to 250 MHz and greater than I11dBf-om 250 to 400 MI-z, based on the results of the half-scale model mieasu~re-ments. The measured axial ratio is less than 1. 6 dB throughout the entireopt~rating lro;quency band.

Details of ifustrat;o1011 I1this documecnt May be t~e~tr

studied on microfiche

UA00FORMIU 14E UNCLAS iSIFIED ____

FACSMME)Security Classification

UNCLASSIFIEDj Securit Classification

.KEY WORDS

Open-sleeve dipoles

Planar array

Synchronous satellite antenna

Distribution Statement (Continued)

Abstract (Continued)

UNCLASSIFIEDSecurity Classification

Air Force Report No. Aerospace Report No.SAMSO-TR-72-178 TR-0073 (3404)-i

1. 8:1 BANDYW. "TH UHF 3 X 3 ARRAY OF

CROSSED l'PEN-SLEEVE DIPOLES

Prepared by

H. T. KING and J. L. WONGE1ect:'o ±ics Research Laboratory

Ltboratory Operations

72 JUN 30

Systems Engineering Operation.THE AEROSPACE CORPORATION

Prepared for

SPACE AND MISSILE SYSTEMS ORGANIZATIONAIR FORCE SYSTEMS COMMAND

LOS ANGELES AIR FORCE STATIONLos Angeles, California

Approved for public release;distribution unlimited

FOREWORD

This report is published by The Aerospace Corporation, El Segundo,

California, under Air Force Contract No. F04701-72-C-0073.

This report, which documents research carried out from July to

November 1971, was submitted 6 June 1972 to Colonel Walter W. Sanders, SK,

for review and approval.

The authors wish to thank H. F. Meyer and P. J. Parszik for their

interest and suppoit of this study. Thanks also go to D. G. Coder,

M. E. Schwartz, and C. Romero for their technical assistance, to

J. T. Shaffer and 0. L. Reid for testing the antenna, and to D. S. Chang

for the computer programming.

Approved

A. H. Silver, Director H. F. Meyer, Syst ms EngineeringElectronics Research Laboratory DirectorLaboratory Operations Tactical Satellite Communications

ProgramsSatellite Systems DivisionSystems Engineering Operations

Publication of this report does not c-. -stitute Air Force approval of

the report's findings or conclusions. It is published only for the exchange

and stimulation of ideas.

WALTER W. SANDERS, Col, USAFDeputy for Space

Communications Systems

-ii-

ABSTRACT

A circularly polarized nine-element (3 X 3) planar array antenna

system has been evaluated for use on a synchronous satellite. The array

elements are mounted on a 104-in, square reflector, and the basic radiating

element is a crossed open-sleeve dipole with good VSWR and pattern char-

acteristics over a 1. 8:1 frequency band (225 to 400 MHz). The electrical

performance of the antenna was measured using a half-scale model. It is

shown that the antenna can provide an EOE gain (gain in the direction of the

edge-of-the-earth) of greater than 14 dB from 225 to 250 MHz and greater

than Ii dB from 250 to 400 MHz, based on the results of the half-scale

model measurements. The measured axial ratio is less than 1. 6 dB

throughout the entire operating frequency band.

-iii-

CONTENTS

FOREWORD . . . . . . . . . . .. iiS,/

I. INTRODUCTION o..... . ...... ..... *.............. . .

II. DESCRIPTION OF ANTENNA SYSTEM . ......................... 3

III. RESULTS .... .. . .................................. . 9

A. Computed Patterns and Directivity ................ 9

Bo Measured Data . ........... o . .. .o .... o..............o.. ............ 6

. 1. Feed Network Characteristics ............... 162. VSw W ......... . ..... ... .. .................... 193. Radiation Patterns .......... ......... ........ 224. Directivity and Gain . ........... ....... 32

5. Axial Ratio ................... .... ....... 41

IV. SUMMARY AND CONCLJSIONS............ ,43

REFERENCES........ 45

!IpI

Preceding page blank --

FIGURES

I. Overall View of 3 X 3 Czossed Open-Sleeve DipolePlanar Array ............................... 4

2. Close-Up View of Crossed Oper.-.S!eeve DipoleAssembly ............... ................... 5

S31 Crossed, Flat Open-Sleeve Dipole Model ........... 6

4, Schematic Diagram of Array Feed Network ............... 8

5. Measured Element Patterns for Dipole-to-ReflectorSpacings of 4.31 and 5 in.. ..................... t0

6. Array Geometry ................................. 1 I

7. Computed Array Patterns for Element Spacing of16 in. and Dipole-to-Reflector Spacing of 5 in ....... 12

8. Computed Array Patterns for Element Spacing of16 in. and Dipole-to-Reflector Spacing of 4.31 in....... 13

9. Computed Array Patterns for Element Spacing of18 in. and Dipole-to-Reflector Spacing of 4. 31 in.. ...... 14

10. Computed Directivities for Dipole-to-Reflector

Spacings of 4. 31 and 5 in.. ..................... 15

1i. Feed Network Loss Characteristics ............... 17

t2. VSWR of Single Element, Elements Embedded inArray, and Complete Array for Dipole-to-ReflectorSpacing of 5 in. and Element Spacing of 16 -in ............... 20

13. VSWR of Single Element and Complete Array forDipole-to-Reflector Spacing of 4. 31 in. and ElementSpacings of 16 and 18 in ......... ........................ 21

14. Mismatch Losses for 4.31- and 5-in. Dipole-to-Reflector Spacings .......................... 22

-vi-

FIGURES (Continued)

15. Measured Radiation Patterns for Element Spacing of16 in. and Dipole-to-Reflector Spacing of 4.31 in.with Rotating Linearly Polarized and CircularlyPolarized Illumination . . . ... .................. 24

16. Measured Radiation Patterns with ExpandedAbscissa Scale, Using a Rotating Linearly

SPolarized Source .... .. .. .. .. .. .. .. .. .. .. .. .. 29~

A 17. Comparison of Measured and Computed RadiationPattern Characteristics for Element Spacing of16 in. and Dipole-to-Reflector Spacing of 5 in..... ....... ... 33

"18. Comparison of Measured and Computed RadiationPattern Characteristics for Element Spacing of16 in. and Dipole-to-Reflector Spacing of"4. 31 in .......................................... ... 35

19. Comparison of Measured and Computed RadiationPattern Characteristics for Element Spacing of18 in. and Dipole-to-Reflector Spacing of4.31 in .. .................................... 36

20. Comparison of Directivity Determined fromComputed Patterns and Measured Patterns forthe Planar Array ......... .......................... 37

21. Comparison of Measured and Predicted Gain forI Three Planar Array Configurations ....... ............... 39

22. Measured Axial Ratios at the Beam Peak fcrz. Three Planar Array Configurations ...... ................ 42

-vii-

III. INTRODUCTION

This report describes, the results of a theoretical and ex:perimental

* evaluation of a wideband UHF planar array of crossed open.s'Jeeve dipoles.

This array is a candidate antenna configuration for a communication satellite

at synchronous altitude operating in the 225 to 400 MHz (1. 8:i1 bandwidth) fre-

quency range. The: primary objective bf this study was to establish the antenna

gain at an off-axis angle of 8. 65 deg from the beam peak. This angle corre-

sponds to a line-of-sight to the edge-of-the-earth (EOE) from a synchronous

orbit satellite. The EOEIgain represents an important factor in the system

anilysib of the communication 1~i"K. I

The experimental work -,as done with a half-scale model (450 to 800 MHz),4• and all the resulti are reported with reference to the scaled dimensions. The

array. cons, sts of nine elements '(3 X 3 configuration) of crossed open-sleeve

dipoles (Rpef. 0) mounted on a 52-in. square reflector. The orthogonal dipoles

are fed in phase quadrature to acquire circular polarization. Sleeve dipole s

were'selected to attain the desired bandwidth. The dipole-to-reflector spacing

and the distance between elements were varied during the investigation. The

dipole configurations and spacings were selected to optimize the array per-

formance in the 225 to 300 MHz (450 to 600 MHz, half-scale) band without

substantial degradation of antenna gain at the upper frequencies.

The element pa~terns uised for the array pattern and directivity compu-

tations are discussed. Measured antenna gain and patterns are compared withIi •he computed results. The measured VSWR of the individual dipoles and the

array of dipoles are also included in 2he report.

IU. DESCRIPTION OF ANTENNA SYSTEM

The frequency of operation for the half-scale model is from 450 to

800 MHz. There are no strict requirements on the antenna beam shape and

sidelobe levels, but the off-axis (*8. 65 deg from beam peak) gain is to be

maximized. Circular polarization is required.

The array consists of nine crossed open-sleeve dipoles arranged in a

square grid as shown in the photograph of Fig. t. The elements are fed by

a corporate feed structure and are uniformly excited. The reflector is 52-in.

squnre and is fabricated with 1/2-in, square mesh wire screen.

To achieve wideband impedance and pattern performance, open-sleeve

dipoles (Refs. 1, 2) were used. The VSWR is less than Z. 5:1 over the 225 to

400 MHz band. Sleeve antennas generally have wider pattern bandwidth

characteristics than a conventional cylindrical (fat) dipole (Ref. 3). The

dipoles and sleeves are constructed with solid metallic surfaces, although

they could have been constructed in a wire-grid arrangement to minimize

weight (Ref. 2). A close-up view of the dipole assembly is shown in Fig. 2.

The original open-sleeve antenna reported by Barkley (Ref. 4) con-

sisted of a dipole with two closely spaced parasitic elements, the length of

the parasites (sleeves) being approximately one-half that of the center-fed

dipol,,. Experimental studies of the open-sleeve antenna (Ref. 1) revealed

that the sleeves could have a wide variety of configurations without any de-

gradation to the VSWR response. For simplicity in construction of the crossed

dipole antennas, a flat sleeve arrangement was chosen. Figures 1 and 2 sho:

the flat metal sleeve construction and the Styrofoam surpport for the sleeves.

Figure 3 shows the construction details of the crossed open-sleeve

dipoles and balun. A coaxial line was used to feed the dipole, and the antenna

structure was made to incorporate a balun. The feed line consisted of a

copper-clad, 0.141-in. diam semirigid coaxial cable. The balanced line of

the balun was also a length of the semirigid cable, but without the center

Preceding page blanko,3--

I~141

.. ... ... z. th

I ý 44

~ A~1~Nm

-L-4-

1A 5

*5 1

t, 4

0- V

rU,- eir:d

Figure~~~~~~~~, .OvrlViw o rse pZ--e'eDplPlna Arrayi

11,ý -4-

.4Lwýtv k

FLAT METAL0K OPE N SLEEVE

?STYROFOAM-2k.SLEEVE SUPPORT.ý,<_ý

Rroduced f om O

Figurev 2. Close-Up View of Crossed Open-Sleeve D~pole Assembly

-5-

FEED POINT DETAILS

DIMENSIONS ARE 0.141-in. SEMIRIGIDIN INCHES COAXIAL CABLE

0.141 DIAM. I0.875 0.362

0.50 OD

450 0.150

S0 .!4 1= 0 .375

Sx

I OPENING

5 25 -

FLAT SLEEVE 0.75

DRREFLECTOR SURFACE

Figure 3. Crossed, Flat Open-Sleeve Dipole Model

-6-

:i

conductor, and the short circuit of this line was coincident with the reflector

surface. The dipoles were screwed into the feed terminals and the sleeves

were supported by Styrofoam.

The array feed network schematic diagram is sho•n in Fig. 4. Fouir

3-way power dividers are used to feed the "Y" dipoles and another four to,

feed the "X" dipoles. The X and Y dipoles are fed in quadrature by a 90-deg

hybrid to provide circular polarization.

-7-

0

Ocn xt~CI cER -0

d0 C L L0 (X -

cr-zcnc:, CfC

cn . c-:z z'a c)= >

0

z (D U-0

0 CC)

120

-8-:

III. RESULTS

A. COMPUTED PATTERNS AND DIRECTIVITY

Measured patterns of an isolated element were used for the array pattern

computations. A single crossed open-sleeve dipole assembly mounted in the

center of a 52-in. square reflector was used to measure the element pattern.

E and H plane patterns (linear polarization) were recorded. To derive the

response for circular polarization, the RSS of the two measured field com-

ponents was calculated; i.e.,

22Ecp - + EE

where EH and EE are the H and E plane patterns, respectively. The resulting

circtularly polarized element patterns for 450, 500, 600, 700 and 800 MHz are

shown in Fig. 5 for dipole-to-reflector spacings of 4.31 and 5 in. The pat-

terns in the backlobe region (90 < 0 _< 180) are not shown, since they are

neglected in the array computations. Mutual coupling effects were also

6 neglected in the array analysis, but they will be considered in a forthcoming

companion report.

The coordinate system used for the array analysis is illustrated in

Fig. 6. The general equation for the field pattern of an N X N planar array

with uniform distribution may be written as

sE(inb)= e ( T sin e :i• sin( Nr- sin 8 cos

e Ns(-s N sin(n sin N cos

where s is the element spacing, E (0) is the measured elemente

direction of the array. The computed patterns for N 3

-9-

i

A109-RUN4

700, 800 MHzc-5

-I0 450MHz:> -- 500 MHz

600 MHza: .... 700 MHz

-15 . ........ 800 MHz

DR - 4.31 in.52 in. SQUARE REFLECTOR

20

A109-RUN 2

800 MHz

00

-5-"•70 MHz_

S-,ov \'<.0'--I

X - DR :5 in.\W

15 52 in. SQUARE REFLECTOR "II

-20 r0 20 40 60 80 90

ANGLE 8, deg

Figure 5. Measured Element Patterns for Dipole-to-ReflectorSpacings of 4.31 and 5 in.

-10-

z

0 0

if- YS

Figure 6. Array Geometry

are plotted in Figs. 7, 8, and 9 for a 5-in. dipole-to-reflector spacing and

an element spacing of 16 in. and also for a 4.31-in. dipole-to-reflector spacing

with element spacings of 16 and 18 in. Because of symmetry, patterns in the

S= 90 and 0 = 135 deg planes are identical to those of the 0 = 0 and 0 = 45 deg

planes, respectively.

The directivity was determine l by integration of the power pattern over

the visible space (0 - 0 :5 7r/2), since the radiation characteristics in the

region ir/2 < 0< 7 are not known. The directivity is expressed as

41TD = 2(2)

"fo Z(eO, ) sin 0 dO do

Figure 10 shows the peak directivity and the EOE directivity values as

functions of frequency for dipole-to-reflector spacings of 4.31 and 5 in. with

-11

!0

: \ \ .500 : + + + - - --5 -600-..o ---. -------- --- -.800 + ++

-o0 ?00:\\\\.X : 60 . 70 -•. 4 Dr-`5. 0 : 8 :I

" -- '.:-•; "" • -•,• •A - - - i • - : PATTERN IN _

- 15 0 *. .0 PL A

-25

• .0' ..

. . . . . . .. . . ...... •• ••-' ' : ..- •. -.; - -

.450 MHz,,' ! , ,500: "

-o -30 16:800 .

I _ j ATTERNIN15 )450 PLANE.-

-20----A T 8001

-25

-300 10 20 30 40 50 60 70 80 90

ANGLE, 6,deg1b) 245' PLANE

Figure 7. Computed Array Patterns for Element Spacing of 16 in.and Dipole,-to,-Reflector Spacing of 5 in.

-i..io! -•,R._, • ;• .i\L. .. -

). .. 7 0 \ ' % ; : : : ,I • •

: •}, i\ i : ; ! I- I\1L-1612-'-"". .o1• i i i • . - - - • , ,

-I1 - T --- I / , I -

0 -- 1--

* .%.•% .456 MHz IS.... .- 500 -- --4x•oo i• I\ 0o / ,o - ,+ +, +•

-5 NW 0_ ++

m~ -

0 .(a)O" PLANE

660

7 0 0 - -+ S :1- 0

SLA X •"l\ 450M•,-i5o "{-10 70000 Dr - -4.31 0

60 50 i\PATTERN IN

- -0 + PL-NEPLANE,I i

I~ ) /0 . =in

0- - 10 40 00 80 9

ANGIII V de

-25 i i ! \ o

-30 , I

(ba) 0 45 PLANE

Figure 8. Computed Array Patterns for Element Spacing of 16 in.: I and Dipole-to-Reflector Spacing of 4.31 in.

-13-

FC

[0

a - - - - __ATTERN IN

Hz 1 7i t= I

-560.0 800 +4-700 VT------S:180

A If zr

15o) 0 :0 PLANE

-I --- - :450 PLANE

-20 1A--- -- 3 C -

-25

-30-

0 i0 20 30 40 50 60 70 80 90ANGLE,O8, deg

(b) 0 :45 PLANE

Figure 9. Computed Array Patterns for Element Spacing of 18 in.and Dipole-to-Reflector Spacing of 4. '.1 a.

-14-

19 '

,8 16

o, S '14 P EAK -

16 16 14

d •DEOE16 3 x 3 ARRAY

14 N. + _ +__ 4DR 4.31 S_-

12 1

19 I8PEAK18 - ' "q•>'' "•"

,18-//; f•• " 14

16 166S=14 N.16 - ,"00-".18•1

: 14 -i4

DR:= .0 =112 S:1\ -

I I I I ! I I ,I ..400 500 600 700 800

FREQUENtCY, MHz

Fi'gure 10. Computed Directivities for Dipole-to-Reflector Spacingsof 4.31 and 5 in.

-15-

pL DO

r I

the element spacing as a parameter. At the low end of the frequency band,

the directivities are )essentially the same for the two dipole-to-reflector

spacings. At 800 MHz, the 4.31-in. spacing provides an enhanced directivity

of approximately I dB. This was expected, because the 4.31-in. spacing

corresponds to 0.29k at 800,MHz and is the position where the radiation pat-

tern beam bifurcation begins to become noticeable. At 5-in. dipole-to-

reflector spac'.g, the spacing is ;0.34% and beam splitting definitely exists,

resulting in , lower element directivity and! thus a lower array directivity.

The directivity represents the gain of an antenna with no losses, and the

actual gain of the planar array is determined by subtracting out the various losses.

The measured feed network, polarizer, and mismatch, losses are presented in

Section II-B,. The antenna ohmic, balun, and other unknown losses cannot be

measured directly and are inferred from the difference between the computed

'and measured gain values..

For the 3 X 3 array, a 16-in. elemrnt spacing provides optimum EOE

directivity (> 15. . dB') in the frequency range of 450 to 700 MHz; however, at

800 MHz, the EOE directivity drops po 13.9 and 12.9 dB for the dipole-to-

reflector spacings of 4.31 and 5 in., respectively. With an element spacing

of 18 in., the EOE gain at the low end of the frequency band may be enhanced

by about 0.5 dB but only at the e7.pense of a gain' degradation of about 1. 5 dB

at the high end' of the band. A 14-in. element spacing provides a more uniform

EOE directivity, as compared to s = 16 and 18, but the overall directivity is

lower, particularly in the more important 450 to 600 MHz band.

B. MEASURED DATA

i. FEED NETWORK CHARACTERISTICS

The measured losses of the feed network (Fig. 4) and its individual

components are shown in Fig. 1i. The circles represent data points, and

the dashed-straight line is a least squares fit to the data points with the

standard deviation, sd'

Figure 1i(a) shows the overall losses of three cascaded lengths of

RG-58C/U cable, each approximately 20-in. long, complete with TNC corl-

nectors. These connector-cable assemblies, as used in the teed network,

-16-

I% I

RG-58 C/U ClOA- RUN IS1.0- 5 f t LENGTH 00

,..- --- 0- EXPERIMENTAL.op..o.-~o-o'r'( - DATA

-AMPHENOL CATALOG) - - - STRAIGHT LINEC.. 0.5 I R 8ClU. FIT TO DATA

S0.3 'RSI 3 WAY POWER

9o2 DIVIDER, No.2751-03LU m 1j-1"

Sd0.04 dB RSI POWER0C48-RUNI2 DIVIDERS

. 0- " I ' I '0'

oiF% 0 106S~~~Sd= O.02dB_.(.,O..

S02 (>o O-o--5 o-

QHM-2-.750GS/N 233 C48-RUNII TYPICAL

m 2 , i , I ' I ' f

u~u• Sd 0.25dB _ ... - ....... SUMMATION OF

-9 INDIVIDUALI - COMPONENT LOSSES

(USING STRAIGHTz AVERAGE LOSS TO 18 LINE FIT DATA)

ANTENNA PORTS C109-RUN I'• o - ' I , I . , I , I.

400 500 600 700 800FREQUENCY, MHz

Figure 1i. Feed Network Loss Characteristics

-17-

I'

have an overall length of 5 ft. The cable lss.s for a 5-ft length of RC,-58C/U

cable, as given in the Amphenol catalog, are xiso shown for comparison.

The Radiation Systems, Inc. (RSI) 3-way power divider losses shown in

Fig. li(b) represent the average of the three output ;zrt• and the average

of eight units. The stripline units have an avrage insertion loss of less

than 0. 22 dB. The maximum deviation among the power 6ividerL, was less

than *0. 2 dB. From port-to-port for th. same unit, the deviatons were

less than *0. 1 dB. The phase deviation between each port for a single unit

was less than 0. 1 deg.

Trhe insertion loss of the polarizer is the average of the losses at the

two output ports. The compact unit with OSM connectors has an insertion

loss of less than 0.27 dB. The differenial phase from 90 deg between the

tw2 ports was less than 0.6 deg and, as an average, less than 0.2 deg over

the frequency range. The power imbalance between the two output ports is

given in Section III-B(5). The performance of the hybrid was satisfactory

at 450 MAHz, even though the unit wa.- advertised by the manufacturer for

operation in the 500 to 1000 MHz frequency range.

Figure I 1(c) shows the overall loss of the feed network from the polari-

zer input to the antenna-balun ports. The data represent the average for the

18 output ports. The maximum deviation among the ports was ±0.3 dB. The

st-.-dard deviation of the regression line is 0.25 dB. When summing up the

losses of the individual components that compri.e the feed network, the total

loss is usually represented by a straight line such as the regression line.

The deviations from a straight line are caused by the unpredictable mismatch

effects and other unknown factors. As a comparison, the regression lines of

each of the individual components were summed, with the result as shown by

the dotted line of Fig. li(d). There is a reasonable agreement with the

experýrnental regression line. The differential phases an-%ong the 18 output ports

wvere less than 5.8 deg, which is an indication of how well, the cables were cut

to identical lengths. As an average, the differential phase from 90 deg between

the X and Y dipole ports was less than t deg. The powe.: imbalance between

-18-

the X and Y Jipole balun ports was essentially the same as that of the hybrid

outputs, which will be eviden+ from the axial ratio discussion (Section IH-B(5)).

Z. VSWR

The VSWR of a single isolated dipole, the dipoles embedded in the

3 X 3 array, and the inputs to the 3-way power dividers and to the polariz'er

were measured prior to making any pattern and gain measurements. The

results for a dipole-to-reflector spacing of 5 in. and with a 16 -in. element

spacing are shown in Fig. 12. The VSWR of one isoleted dipole is less than

2.3:1. Both the X and Y dipoles of elements No. 5, 7, and 8 (see Fig. 4)

were measured at the balun input port, with the remaining elements terminated

in 50-ohm loads. These three elements were chosen for the measurements

because they have different neighboring dipole arrangements, and the VSWR

characteristics would provide an indication of mutual coupling and reflector-

edge effects. Rather than plot each curve individually, the maximum spread

in the VSWR values is shown (Fig. 12(b)).

Since the X and Y dipoles in the array were fed with separate sets of

power dividers, the input VSWR to each set (Ports A and B, respectively,

Fig. 4) was measured. The results are shown in Fig. 12(c). The VSWR

response is considered good. The VSWR measured at the input of the polarizer

with all the dipoles connected is shown by the dotted line. This VSWR is low,

as expected, since most of the reflected energy is dissipated in the fourth port

of the 90-deg hybrid.

Figure 13 shows the VSWR characteristics for a dipole-to-reflector

spacing of 4.31 in. and element spacings of 16- and 18-in. The maximum

VSWR of a single dipole is 2.8:1. It should be pointed out that the VSWR

can be significantly reduced by making the dipoles fatter and by varying the

sleeve parameters, as discussed in Ref. 1. For the purposes of thisstudy, the change in diameter was not considered necessary. The VSWR at

the input to the power divider network feeding the X and Y dipoles for element

spacings of 16 and 18 in. are shown in Figs. 13(b) and 13(c), respectively.

-19-

I I ' I I SINGLE ISOLATEDELEMENTB 109 - RUN I

0 5 in. DIPOLE-TO-2 _REFLECTOR,••ii , • • o SPACING

RANGE OF VSWR3 FORTHE X AND

8109 -RUN 2 Y DIPOLES FORELEMENTS 5, 7,AND 8 EMBEDDED

2 _IN ARRAY(.> ýy

I ++ +x7 84+ +

3 ' I ' I I ' I IP~T--- INPUT TO

B 109-RUN 4 Y DIPOLES'POWER DIVIDER

3•: 2i FEED NETWORK

--- INPUT TO"•\ Y ••L/rl, X DIPOLES'

S.POWER DIVIDERFEED NETWORK

400 500 600 700 800 •""INPUT TOFREQUENCY, MHz HYBRID

Figure 12. VSWR of Single Element, Elements Embedded in Array,and Complete Array for Dipole-to-Reflector Spacing of5 in. and Element Spacing of 16 -in.

2

' -20-

3B109 -RUN I •ISOLATED

9 -RU N l I D R 4 , 1R: 4 .3 in. ELEMENT

319 -RUN?_ 810O9-RUN 7 S =16 in. _

2

INPUT TO POWER@0.0.. 0., DIVIDER FEED NETWORK

'- Y DIPOLES3I I ' I --- XDIPOLES

B109 -RUN 8 S= 18 in. . ...... INPUT TO HYBRID2

400 500 600 700 800FREQUENCY, MHz

Figure i3. VSWR of Single Element and Complete Array for Dipole-to-Reflector Spacing of 4.31 in, and Element Spacings oft6 and 18 in.

-21-

SThe power loss due to the VSWR, which is not transmitted by the antenna

system, is referred to as the mismatch loss. The exac't mismatch loss for the

array is difficult to determine from the VSWR data. However, for purposes

of the present study, the VSWR of a single Isolated element was 'used to

determine the mismatch losses. In this manner, the effects resulting from

the interactions between the various components in the feed network and the

array are not accounted for, but from an examination of Figs. 1Z and 13 it

appears that this error is quite small. Figure 14 depicts the mismatch losses

for the dipoLe-to-reflector spacings of 4.31 and 5 in., which are used to

determine the array gains from the computed directivities.

2 *i I

SINGLE ELEMENT VSWR') D:431

400 500 600 700 800FREQUENCY, MHz

Figure 14. Mismatch Losses for 4.31- and 5-in.., Dipole-to-Reflector Spacingb

3. RADIATION PATTERNS

Radiation patterns of the 3 X 3 array were measured for an element

spacing of 16 in., with dipole-to-reflector spacings of 4.31 and 5 in., and

for an element spacing of 18 in. with a 5-in. dipole-to-reflector spacing.

A summary of the computed and measured pattern characteristics is pre-

sented in this section.

-22-

Until a circularly polarized transmitting antenna was made available

for the array radiation pattern measurements, a 6-ft parabolic reflector

with a rotating linearly polarized dipole feed was used as an illuminating

source. With the rotating linearly polarized source, the axial ratio can be

determined for any aspect angle. The pattern response to a circularly

polarized wave can then be determined by taking the RMS value of the maximum

and minimum signals for a given angle:

E 2 + E2

max mincp 2

By using this RMS technique, the measured circularly polarized response

was established, and the principal radiation characteristics such as beamwidth,

EOE correction factors, and sidelobe levels were compared with the computed

values and with the measured patterns for a circularly polarized source.

Figure 15 shows representative patterns recorded with both a rotating

linearly polarized and a circularly polarized source for 450, 500, 600, 700,

and 800 MHz. These patterns are for an element spacing of 16 in. and a dipole-

to-reflector spacing of 4.31 in. The distance between the transmitting and

receiving antennas was approximately 60 ft. Only the patterns in the 0 = 0 deg

plane (in the plane containing a row or column of elements) are shown, since

this plane has the most pertinent information for comparison with the computed

results. The 0 = 45 deg plane pattern has approximately the same beamwidth

as the 0 = 0 deg plane, but the sidelobe levels are less than -20 dB as shown

by the computed patterns and confirmed by the measured patterns. One 0 = 45

deg plane measured pattern for 450 MHz is shown in Fig. 15(b).

Figure 16 shows the measured patterns of the main lobe region (expanded

abscissa, ±30 deg) taken with the rotating linearly polarized source. The

element spacing was 16 in., and the dipole-to-reflector spacing was 5 in.The half-power beamwidth, EOE correction factor, and axial ratio can be

accurately determined by using the expanded patterns.

-23-

... .... .450M- -

L0L S --- -S 16% Dp -4.31'

I ' l -1~4 Id8

-

1. ....

i ..... . .. ...- t - -_ t i V <ijii -it- i f

I1, _dI , 552, '-14 6o

"t'I.: : : :: .. . '.... .. , .. u.I. '/I:," j .,'.,.S,

(a 5 M ~, Roaing, Linea Polarization~-" "

--,- - ... .• io " I - ,.., ..U.,' ] ' . " I.. .....,. ..: "":. 292* c " ' *.i, . ....I.O:.ZA ,thg ;... .. ....... ~. • .h :: :. :

- 1. ,:•::jj f ' '..T, .. I',-.::: I

* . • T - , . .. £-"-- .:4z5 *T',-.i,.,-,.,:o.-

-4: , : : T -. . .... ..................-.

7 T

4, . . _ ,.

(b) 450 MHz, Circular Polarization

Figure 15. (Sheet I of 5) Measured I~adiation Patterns for ElementSpacing of 16 in. and Dipol3-to-Reflector Spacing of

4. 31 in. with Rotating Linearly Polarized andCi rcula rly Pola ri zed Illumination

-24-

! -i°Am -RUNPATTI" 0S- TI .00.,-1-

I1 A -. (I*39. S0IS O i n -______-14___Idl. -I-•~-.

(c) 500 MHz, Rotating Linear Polarization

SII " , - 0I0U 1

""S4 .. 163, Do 4 31'

Am -RU I

.,, I

- I N J

I' I'

(d) 500 MHz, Circular Polarization

Figure 15. (Sheet 2 of 5) Measured Radiation Patterns for ElementSpacing of 16 in. and Dipole -to -Reflector Spacing of4. 31 in. with Rotating Linearly Polarized andCircularly Polarized Illumination

-25-

i $F 16% 01 4.311

4 ....... - -L . .~A L * __ iso.IdI

(e) 600 MHz, Rotating Linear Polarization

I. 2 F(.•,,10*

i I CORCLULA

21 * , PO i l ' 4.T1"

" " ATT[

- . . . . .. : . ... .. .

aI' t *

" - - -21-4-- -

*II 2 j

t' ' , 'I ,

(f) 600 MHz, Circular Polarization

Figure 15. (Sheet 3 of 5) Measured Radiation Patterns for Element

Spacing of 16 in. and Dipole-to-Reflector Spacing of

4.331 in. with Rotating Linearly Polarized and

Circularly Polarized Illumination

-z6-

' " - q - -E ( 0, #. O . 10 0 ° 1

- ~ r 700MM.

.__ _

PATTERN $

Pt.. i{. - I i 0- 3 1

~t1 i.1

-

-~~~~ A s.

()700 MHz, Rotating Linear Polarization

HI LI + y...-.S. ,+-++ +-:+-++! I+ - +..L f...... ... l _ - - . ... ... ,,,,,,,

• . :L ... 4... .. . I+, .... - , -,_- __ L .- . . o.._._| OZ?

T h.... -a.'•, '_____:+-+- -+t:++ - • . ,._ . ._____.__.1,--..°.

r ft

S~a- i- a

-1 .+I:.• +_ , , _..- Io _-+'... .

(g) 700 MHz, RoainguLinar Polarization

Sp .a.-.. L f M norg of

4.31i. wih oatn*Lnary oarzd n

Porclary oariz dIluiation

S-27-

......... I L7 .$ S 1 ' , 0 1 , 4 ,3 1 "

, T

- I - ,

- I T t ý t t ' °

(i) 800 MHz, Rotating Linear Polarization

lf. # . ., -O.I

17 S1 1- 16. 4.

., C1

, ,. P O L A RI ZA T IO I

4744.,.. .

'," ""S.5- J- -*am a

.73,,,. . 4 ..

* T .... _2

' . , , * .,..

\ I

(j) 800 MHz, Circular Polarization

Figure t5. (Sheet 5 of 5) Measured Radiation Patterns for ElementSpacing of 16 in. and Dipole-to-Reflector Spacing of4.31 in. with Rotating Linearly Polarized andCircularly Pola rized Illumination

-28-

I - EXmWPATR

Sel I11,4

I T -'--iu i~at 8 - - PATTERN 2

1 0.r-

10 2 *.~i 0 .

Li.. 7 _ __ 7

(a) 450 MN2E R.# 0. ISO"

I 00

(b) 500 MHz 4

_

Figure 16. (Sheet I of 3) Measured Radiation Patterns withExpanded Abscissa Scale, Using a Rotating

* Linearly Polarized Source

-29-

-174 EXMANKOFOTTLA

4 -- - Frda rt74

I 10- mm$14IEN3

(c) 600 MHz

i; I ro-oi - 2A4

Tt-

.,4

()700 MHz

Figure t6. (Sheet 2 of 3) Measured Radiation Patterns withExpanded Abscissa Scale, Using a RotatingLinearly Polarized Source

-30-

(e)60 800MI

Figure ~ ~ ~~ W EXPNDE (SetATf3)MaurdRditoEPtenswtExpande Abcis Scale, Using a Rotating5

Linearl7y Polaize Sorc

12-31d

A summary of the measured half-power beamwidths (HPBW), EOE

(0 = 8.65 deg) correction factors, angles to the first null, and sidelobe levels

is shown in Figs. 17, 18, and 19 for 3 array configurations, as follows:

Dipole-to-Reflector ElementFigure Spacing (Dr), in, Spacing (s), in.

17 5.00 16

18 4.31 16

19 4.31 18

Comparisons are also made in these figures with the computed patterns.

The measured values represent the average of several patterns. Since the

computed HPBW and EOE correction factors for the 0 = 0 and 45 deg planes

are essentially the same, only the values for the @ = 0 deg plane are plotted.

The HPBW in the 0 = 45 deg plane is approximately Io wider, while the EOE

correction factor in the 0 = 0 deg plane is slightly worse than the 0 = 45 deg

plane (differential less than 0.05 dB). The HPBW and EOE experimental

data points for both principal planes compare very well with the predicted

va'ues, as shown in the summary curves. Also, there is a good correspondnece

betveen the computed and measured first null positions in the 0 = 0 deg plane,

and the measured sid-lobe levels are comparable to the computed levels.

4. DIRECTIVITY AND GAIN

A complete set of radiation patterns was taken for an element spacing

of 16 in. with a dipole-to-reflector spacing of 5 in. and another set for s = 18

and D = 4.31 in. The purpose of these measurements was to determine ther

directivity by integration of the measured patterns. Great circle pattern

cuts were taken in the 0 = 0, 11.25, 22.5, 33.75, and 45 deg planes for

frequencies of 450, 500, 600, 700, and 800 MHz. The measured 4irectivi-

ties are compared with the computed values as shown in Fig. 20. For the

-32-

Cc 0 -- T

A109 -RUN 32 - I - &k-.t ~s-u

H"--,,,• EOE :8.6502--

.- •- � COMPUTEDo 3- .. -:0 SPLANE

Lu 8 -4 - I I I I, I L .L._8-4

(a) EOE CORRECTION MEASURED DATA30 0.. , =00 o4,o

_ 0 -. A109-RUN 3 x 0 =450

m2O-N& .* - _ 33x3 ARRAYT_ + + ++

101 S=16

(b) HALF-POWER BEAMWIDTH + ± +35 o. H --

S30 A109o-RUN 3 DR: 5

-25 O =O0 PLANE

20 -

15-400 500 600 '700 600

FREQUENCY, MHz

(c) ANGLE TO FIRSTr NULL

Figure 17, (Sheet i of 2) Comparison of Measured and CormputedRadiation Pattern Characteristics for ElementSpacing of 16 in. and Dipole-to-Reflector Spacingof 5 in.

-33-

I

0A109 -RUN 3

2nd A5 0 PLANE SIDELOBE �- ---- COMPUTED

/ / EXPERIMENTALIst SIDELOBE / Ist SLL

n -10 / -

oo/ 2nd SLL

LUJ /

So15 /LU /

o 0

c,3-20 -I'"

_Z

/ 0:• ~~-25- %-5// =450 PLANE

-// 0 Ist SIDELOBE -

-A0 I .I I400 500 600 700 800

FREQUENCY, MHz(d) FIRST AND SECOND SIDELOBE LEVELS

Figure 17. (Sheet 2 of 2) Comparison of Measured and Computed"Radiation Pattern Characteristics for ElementSpacing of 16 in. and Dipole-to-Reflector Spacingof 5 in.

-34-

4.

.435 6005/ ' I I I•I 'I o ~0o

30- A1o9-RUN? 7 x #450

25-

20-

15 '3 x 3 ARRAY

-5 + +2nid SlDELOBE• ++

y M f .• S:16

- 0

-0 DR : 4 .3 1

-15-91 COMPUTEDLU :00 PLANE

-20 A109o-RUN 7

, i , T , I , I

n 0;•" •,A109-RUN 7

CD-I1=('-'

LU4 1

I 4 , I I , I ,

400 500 600 700 000FREQUENCY, MHz

Figure 18. Comparison of Measured and Compute,I Radiation PatternCharacteristics for Element Spacing oi 16 in. and Dipole-to-Reflector Spacing of 4.31 in.

-35-

30-'.

2 " 2 5 A 0 R N 1

- *-,,.---112 NULL WIDTHS20 -.. _- )COMPUTED

cQ <z 15 W :0 PAN

•-~ =O P OL A NE l • 3 x 3 A R RAY

(a) HALF-POWER BEAMWIDTH AND ANGLE- TO FIRST NULL

o 2nd SIDELOBE A109-RUNIO

" _q -5D -- Io t- 0S=18

IstSIDELOBE

0j"d -20/ "

co

-20 -=0 0 PLANE

-251(b) SIDELOBE LEVELSS0 , I ' ' I ' I ,

* -I A109- RUN 10

400 00 nO =00 800Nc OE -2 =0O0 PLANE

CC-3

51400 500 600 700 800

FREQUENCY) MHz.

(c) EOE CORRECTION

Figure i.Comparison of Measured and Computed Radiation PatternCharacteristics for Element Spacing of 18 in. and Dipole-to-R'eflector Spacing of 4. 31 in.

-36-

20 ' I ' I

A109-RUN 10 3 18SDR :4.31

18

"16 I I I I - COMPUTED

2U ~INTEGRATIONo20 I ',, II OF MEASURED

S A9-RUN 3S :16 PATTERNS

18

16 -

I I nIii I L L ,400 500 600 700 800

FREQUENCY, MHz

Figure 20. Comparison of Directivity Determined from ComputedPatterns and Measured Patterns for the Planar Array

-37-

s -- 16, Dr = 5 case, a linearly polarized rotating dipole was used in the

pattern measurements, and the circularly polarized response was determined

by taking the RMS value at each angle. For the second set of patterns, a

circularly polarized source was used. The computed directivities are shown

in Fig. 10, as determined from the array pattern using the measured isolated

element patterns.

The predicted gains for the three array configurations tested (s = 16,

Dr = 5; s = 16, Ir = 4.31; and s = 18, Dr = 4.31) were computed by using thetheoretical directivities of Fig. 10 and subtracting the mismatch (Fig. 14)

and the measured feed network losses (Fig. if). Figure 21 shows plots of the

predicted peak gain curves as well as the measured data points. The dotted

line is the least squares fit of the measured data, with the standard deviation

designated by sdo The measured EOE gain curve is the peak gain less the

computed EOE correction factor (Figs. 17, 18, and 19).

The correspondence between the measured and predicted gain values

is good in certain portions of the band, and the maximum spread is as much

as 1. 2 dB. The average differences between the computed and measured gains

are as follows:

Difference BetweenElement Dipole-to-Reflector Computed and Measured

Spacing (s), in. Spacing (Dr), in. Gains, dB

16 5.00 0.18

16 4.31 0.18

18 4.31 0.63

As an average, the computed gain is approximately 0. 33 dB higher than the

measured gain over the operating frequency band.

The gain measurements were made by the substitution method, using

a ridged waveguide horn* as a reference antenna. Each data point represents

the average of six measurements. The procedure involves setting the trans-

mitting antenna to vertical polarization. The array is then pointed to acquire

The horn (NURAD, Inc., Model No. 7 RH) was calibrated in-house with astandard deviation of 0. 25 dB by using both the two- and three-antenna methods.

-38-

18

-DR = 5 PEAK (CALCULATED*)S : 16 .2 j - -

16 o~*. PA -~ T MEASURED GAINPEAK ATBEAM PEAK

(MEASURED) o\co ..... .. Sd :0.35dB .8 .... LEAST SQUARES

0 FIT TO MEASURED14~ GAIN DATA

EOE +C1091-RUN2,3 (CALCULATED) S=1

(a) DIPOLE -TO -REFLECTOR SPACING:5 in.,

ELEMENT SPACING = 16 in.

18 1 1 1 1DR =4.31 PEK(CALC~ULATED*') *USING MEASUREDS = 16 .. LOSSES

16 ~ ~ '~

4' PEAK (MEASURED) -

-o 0. 5 :0.2 1dB

12

C109 -RUN 4 (CALCULATED)10 1 1 1 1 1400 500 600 700 800

FREQUENCY, MHz

(b) DIPOLE -TO -REFLECTOR SPACING :4.31Iin.,ELEMENT SPACING= 16 in.

Figure 21. (Sheet i of 2) Comparison of Measured and PredictedGain for Three Planar Array Configurations

-39-

18 I I IC109-RUN 6 PEAK (CALCULATED*) o MEASURED GAIN

(MEASURED) AT BEAM PEAKSd :O.37dB

16 . ......... LEAST SQUARES0 1.. ..0 o FIT TO MEASURED

•....."\ - GAIN DATA

14 ... "' o" *USING MEASUREDFOE(CALCULATED) LOSSES

(MEASURED)

12 - DR 4.31 --... \ - 3 x 3 ARRAY

S:21DR 3 I - +-t-+T

400 500 600 700 800FREQUENCY, MHz

(c) DIPOLE-TO-REFLECTOR SPACING :4.31 in.,ELEMENT SPACING : 18 in.

Figure 21. (Sheet 2 of 2) Comparison of Measured and PredictedGain for Three Planar Array Configurations

-. 0

the maximum signal. The "head" of the antenna mount is rotated so that the

maximum or minimum of the polarization ellipse is found. Three readings

are taken as a function of range for the array at both the maximum and minimum

of the polarization ellipse. Next, the horn is physically substituted for the

array, and readings are made at the same corresponding range. The dif-

ferential readings between the array and the horn are then used to calculate

the measured gain. The correction factor from the axial ratio must also be

used to convert the gain with respect to a circularly polarized wave; thus,

the gain is referred to the RMS readings of the maximum and minimum of

the polarization ellipse. The procedure is repeated for horizontal polarization,

thus providing a total of six readings.

5. AXIAL RATIO

The axial ratio of the array may be determined from the gain measure-

ments by taking the ratio of the maximum-to-minimum of the polarization

ellipse. The average of the six readings for both the vertical and ho'-izontal

polarization measurements represents one data point in the axial ratio plots

of Fig. 22.

The dotted curve in Fig. 22 is the power imbalance between the two out-

put ports of the 90 deg hybrid. * The cross marks represent the measured

power imbalance between the X and Y dipole ports of the feed network. The

data indicate that the axial ratio is primarily determined by the power split of

the hybrid, while the feed network and dipoles cause only negligible deteriora-

tion of the axial ratio. Based on the three sets of axial ratio measurements,

the maximum deviation with respect to the power imbalance of 'he hybrid out-

puts is 0.9 dB, and, as an average, the deviation is less than 0.1 dB over the

operating frequency band.

Figure 22 depicts the axial ratio at the beam peak. The axial ratio at

EOE can be established from the expanded-abscissa patterns of Fig. 16, which

show the relative axial ratios throughout the main lobe. The increase in

axial ratio at the EOE angle over the beam peak is generally less than 0. 3 dB.

Designed for 500 to 1000 MHz operation.

-41-

MEASURFO2, IAXIAL RATIO

SDC109-RUN3 PEAK

"" I .-..... POLARIZER.I • IMBALANCE< (MERRIMAC HYBRID,

_,I QHM-2-.750G)I (C48-RUN II)

X POWER2 i IMBALANCE OF- co09 -RLu,4 FEED NETWORK

-. D A109-RUN 6 OUTPUTS (AVERAGE)oDR 4.31 (C109-RUN I)% S = 16 0

x... 0 3 x 3 ARRAY0 I "$... ... ? , I +++

+++

-o 0 C109 -RUN6oDR =4.31

0:: S =:180 I

0 0<0. 00

< , 1 60 6 o400 500 600 700 800

FREQUENCY, MHz

Figure 22. Measured Axial Ratios at the Beam Peak for ThreePlanar Array Configurations

-42-

IV. SUMMARY AND CONCLUSIONS

A half-scale model of a planar array intended for use on a synchronous

satellite was constructed and tested over the 450 to 800 MHz frequency range

(225 to 400 MHz full-scale). The array consists of nine (3 x 3 configuration)

crossed open-sle'l.re dipole elements mounted on a 52-in. square reflector.

The array elemei.'• are uniformly excited in order to maximize the antenna

gain. The basic radiating element has relatively good VSWR response and

pattern characteristics over the 1.8:1 frequency band. The VSWR of the

array looking into the power divider input ports of the X and Y dipole elements

is less than 2.8:1 and 2.3:1 for dipole-to-reflector spacings of 4.31 and 5 in.

.(half-scale), respectively. With further optimization, i.e. , by varying the

dipole and sleeve parameters, the VSWR response can be improved.

Comparisons were made between the computed and measured results,

and relatively good correlation was observed. The calculations were per-

formed using the conventional sin Nx array factor and a measured elementNsin x

pattern of a single isolated crossed dipole mounted on the center of a 52-in.

square reflector. By integration of the measured array patterns, the direc-

tivities obtained compare very well with the predicted values over the entire

irequency band. However, the measured gain as an average over the fre-

quency band was 0.33 dB lower than the predicted gain values. It is shown

that the antenna can provide an EOE gain (gain in the direction of the edge-of-

the-earth, or *8. 65 deg from the nadir) of greater than 14 dB from 225 to

250 MHz, and greater than I dB from 250 to 400 MHz, based on the results

of the half-scale model measurements and also using the component losses

associated with the scaled frequencies.

The measured on-axis axial ratio of the array is less than 1. 6 dB

throughout the entire frequency band. However, the axial ratio of the array

was found to be deterrned primarily from the power imbalance between the

two output ports of the 90-deg hybrid that was used as the polarizer; i. e., the

effects of the feed network and the dipoles did not significantly affect the axial

-43-

ratio. Based on the measurements, the maximum axial ratio deviation due to

the power imbalance of the hybrid outputs is 0. 9 dB, and, as an average, the

deviation is less t1 an 0. 1 dB. Thus, if a hybrid is designed for the frequency

band of interest, the axial ratio would be less than the values reported here.

The increase in EOE axial ratio is less than 0.3 dB from the value at beam

peak.

-44-

4 lREFERENCES

1. H. E. King and J. L. Wong, "An Experimental Study of a Balun-Fed,Open-Sleeve Dipole in Front of a Metallic Reflector, " IEEE Trans.Antennas and Propagation AP-20, 201-204 (March 1972).

2. H. E. King and J. L. Wong, 225 to 400 MHz Antenna for Spin-StabilizedSynchronous Satellites, Technical Report TR-0172(2162)-i, TheAerospace Corporation, El Segundo, Calif. (Jan. 1972).

3. E. L. Bock, J. A. Nelson, and A. Dorne, "Sleeve Antennas, " Very HighFrequency Techniques, ed. H. J. Reich, Radio Research Laboratory,Harvard University, McGraw-Hill Book Co., New York (1947) p. 123.

4. H. B. Barkley, The Open-Sleeve as a Broadband Antenna, TR No. 14,U.S. Naval Postgraduate School, Monterey, Calif. (June 1955),DDC AD 82 036.

-45-