RICHARD C. JOHNSON ANTENNA ENGINEERING HANDBOOK THIRD EDITIONAntenna Engineering Handbook THIRD EDITION Richard C. Johnson titor Georgia Institute of Technology Adanta, Georgia Henry Jasik stor otis eionISSN 1063-665X. Copyright © 1993, 1984, 1961 by McGraw-Hill, In. All rights reserved. Printedin the United Sates of America. Except aspermitted under the United States Copyright Actof 1976, no partofthispublication may be reproduced or distributed in any form or by any means, or stored in adata base or retrieval system, without the prior written permission of the publisher. 67890 DOCDOC 09876543210 ISBN 0-07-032381-X ‘The sponsoring editor for this book was Daniel A. Gonneau, the editing supervisor was Stephen M. Smith, and the production supervisor was, Pamela A. Pelion. It was se in Times Roman by Progressive Typographers. Printed and bound by R. R. Donnelley & Sons Company. Information contained in this work has been obtained by McGraw-Hill, Inc, from sources believed 1 be reliable. How= ever, neither McGraw-Hill nor is authors guarantee the accu. ‘Rey or completeness of any information published herein and ‘either McGraw-Hill noritsauthors shall beresponsibleforany rors, omissions, or damages arising out of use ofthis informa tion. ‘This work is published with the understanding that “MoGraw-Hilland its authors are supplying information but are ‘ot attempting to render engineering or other professional er- Vices. If sich servioes are required, the assistance ofan appro- ate profesional should be sought.Dedicated to Dr. Henry Jasik (1919-1977) _ ‘An internationally known authority inthe an- tenna engineering field, Dr. Henry Jasik was highly respected for his breadth of information, een insight, and depth of knowledge in both fundamentals and applications. He received the degree of Bachelor of Science in Electrical Engineering in 1938 from Newark College of Engineering and the degrees of Master and Doctor of Electrical Engineering in 1951 and 1953, respectively, from the Polytechnic Inst- tute of Brooklyn, Dr. Jasik worked n the Navy Department feom 1938 to 1939 and in the Civil Aeronautics ‘AGministrtion (CAA) from 1939 to 1944. Whilein the CAA, he worked on radio aidstoair navigation, including instrument landing systems and VHF radio ranges. From 1944 to 1946, he was an officer in the United States ‘Navy, where he worked on the development of airborne radar and commu- nications antennas atthe Naval Research Laboratories. From 1946 to 1949, Dr. Jasik was associated with Andrew Alford, Consulting Engineers, as Senior Project Engineer and as Vice-President of the Alford Manufacturing Company. From 1949 o 1952, he wasemployed by the Airborne Instruments Laboratory (AIL), where he was associated With the Special Devices Section and the Antenna Section as Assistant Supervising Engineer. In 1952, he became an independent consultant. Tin 1955, Dr. Jasik started Jask Laboratories, Inc. a completely inte- ¢zated operation forthe design, development, and production of antennas ‘Under his leadership the company grew and prospered. With design expe ‘ence from simple Yagi-Uda arrays to exotic multiple-eed systems, Jasik Laboratories became widely known throughout the electronics indusiry as ‘an unusually talented team of antenna specialists. They designed, manu- factured, and tested a wide variety of antennas, such as dipoles, homs, log periods, double-conica feeds, and many others. In 1969, Cutler-Hammer (hich later became the Eaton Corporation) acquired Jask Laboratories, and atthe time of his death, Dr. Jasik was a Vice-President in Eaton's AIL Division. Dr. Jasik was elected a Fellow of the Institute of Electrical and Blec- tronic Engineers in 1958 for his contributions to “the theory and design of VHF and microwave antennas.” He was a member of Sigma Xi and Eta Kappa Nu and was a registered profesional engineer in New York and Massachusetts, Tithe course of his distinguished career, Dr. Jasik was awarded several patents and he designed practical antennas for U.S. Navy ships and for the FM broadcast antenna which operated for many years atthe top of the Chrysler Building in New York City. Healso made many important contri-butions to the young science of radio astronomy by designing and produc- ing improved feed structures for many paraboloidal-reector radio tele- scopes. His innovative work significantly improved the tools of radio astronomers. Dr. Jasik was Editor of the First Edition of the Antenna Engineering Handbook, published by McGraw-Hill in 1961. His vision and creative work produced the most valuable reference book for antenna engineers throughout the world. This new edition is dedicated to his memory in order to convey appreciation for his outstanding and lasting contributions to the field of antenna engineering.Contents 1_ INTRODUCTION AND FUNDAMENTALS: 1. Introduction to Antennas ‘Richard C. Johnson, Georgia Institute of Technology 2. Fundamentals of Antennas Henry Jsik, AIL Division af Eaton Corporation 3. Arrays of Discrete Elements Mark T. Ma, National Insitute of Standards and Technology 2_ TYPES AND DESIGN METHODS 4 Dipoles and Monopoles Chen To Tai, The University of Michigan Stuart A. Long, Univesity of Houston 5 Loop Anteanes Glenn S. Smith, Georgia Institute of Technology 6 ‘Small Antennas Haroid A. Wheeler, Hazeline Corporation 7 Microstrip Antennas Robert E. Munson, Ball Communication Systems Division 8 Slot Antennas William F. Croswell and Mary Beth Selby, Harris Corporation 9 Slot Antenna Arrays ‘Hung Yuet Yee and Robert C. Voges, Texas Instruments Inc 10 Leaky-Wave Antennas ‘Arthur A. Oliner, Polytechnic University AL Long-Wire Antennas ‘Bamund A. Laport, RCA Corporation 13 2 a SA o1 catn 13 4 15 16 7 9 a n ‘Surface-Wave Antennas Francis J. Zucker, Rome Laboratory, Hanscom Air Force Base Helical Antennas Howard E. King and Jimmy L. Wong, The Aerospace Corporation Frequency-Independent Antennas Raymond H. DuHamel, Antenna Consultant James P. Scherer, Loral Randtron Systems ‘Hom Antennas ‘Allan W. Love, Rackwell International Lens Antennas Donald G. Bodnar, Georgia Instiure of Technolagy Reflector Antennas Kenneth S Kelleher, Consultant Geotrey Hyde, COMSAT Laboratories Electromechanical Scanning Antennas Kenneth S. Kelleher, Consullant Frequency-Scan Antennas James. Ajioka, Hughes Aircraft Company Phased Arrays ‘Raymond Tang and Richard W. Burns, Hughes Aircraft Company Conformal and Low-Profile Arrays Robert J. Mailloux, Rome Laboratory, Hanscom: Air Force Base Adaptive Antennas Leon J. Ricard, LIR, Ine ‘Methods of Polarization Synthesis Warren B. Off, W’& B Observatory Lome K. DeSize, Consultant 3_ APPLICATIONS m4 1 141 151 161 mA 181 194 204 21 24 Low-Frequency Antennas ‘Boynton G. Hagaman, Kershner, Wright and Hagaman, PC. Medium-Frequency Broadcast Anteanas John A. Lundin and Ronald D. Rackley, dv Pri, Lundin & Rackley, In High-Frequency Antennas Roald Wilensky, Technology for Communications International EF and UHF Communication Antennas Brian S. Collins, Radiation Systems International Ltd ‘TV and FM Transmitting Antennss Raymond H. DuHamel, Antenna Consultant AIR. Mahnad, Jampro Antennas The ‘TY Receiving Antennas Edward B. Joy, Georgia Institue of Technology Microwave-Relay Antennas Charles M. Knop, Andrew Corporation 2-4 281 29-131 Radiometer Antennas ‘William F. Croswell, Haris Corporation Calvin T. Swift, University of Massachusetts 32, Radar Antennas Paul E. Rawlinson, Raytheon Company Harold R. Ward, Consultant 33. Microwave Beacon Antennas Phillip N. Richardson, Texas Instruments, In. 34. Tracking Antennas Josh T. Nessmith, Georgia Insitute of Technology Willard T. Patton, General Eecric Company 35 Satellite Antennas ‘Leon J. Ricard, LIR, In 36 Earth Station Antennas ames H, Cook, Jr, Scientfe-Atlana, Ine. 37 Aircraft Antennas William P. Allen, 1c, Lockheed Aeronautical Systems Company Charles E. Ryan, i, Georgia Institute of Technology 38 Seeker Antennas James M. Schuchard, American Electronic Laboratories Dennis J. Kozakolf, Milimeter Ware Technology, Inc 39 Direction-Finding Antennas and Systems Hugh D. Kennedy and Roy B. Woolsey, Technology for Communications International 40 ECM and ESM Antennas Daniel F. Yaw; Westinghouse Defense and Electronic Systems Vernon C. Sundbere, GTE Systems 41 Radio-Telescope Antennas John D. Kraus, The Ohio State University 4_ TOPICS ASSOCIATED WITH ANTENNAS. 42 Transmission Lines and Waveguides Roderic V. Lowman, Techni-Ques, Ine 48 Impedance Matching and Brosdbanding David F. Bowman, Antenna Consultant 44 Radomes Gene K. Huddleston, Martin Mareta Corporation Harold L. Basset, Georgia Instinde of Technology 45. Microwave Propagation Geottey Hyde, COMSAT Laboratories 46 Materials and Design Data Donald G. Bodnar, Georgia Institute of Technology Index follows Chap. 46 314 34 34 23 431Contributors James S. Ajoka,* Manager, Electromagnetics Laboratories, Hughes Aircraft Com- ‘pany, Fullerton, California. (Chap. 19) William P. Allen, Jr, Consultant, Atlanta, Georgia (Chap. 37) Harold L. Bassett Chief, Modeling and Simulation Division, Georgia Tech Re- search Institute, Atlanta, Georgia. (Chap. 44) Donald G. Bodnar, Principal Research Engineer, Georgia Tech Research Institute, ‘Atlanta, Georgia (Chaps. 16 and 46) David F. Bowman, Antenna Consultant, Williamstown, Massachusetts, (Chap. 43) Richard W. Burs, Manages, Microwave Techniques Department, Array Antenna Laboratory, Hughes Aircraft Company, Fullerton, California. (Chap. 20) ‘Brian S. Collins, Radiation Systems International Ltd., CSA Antenna Systems Divi- sion, Rochester, England. (Chap. 27) James H. Cook. Jr, Principal Engineer, Scentic-Atlanta, Inc, Atlanta, Georgia. (Chap. 36) William F. Croswel Senior Scientist, Hartis Corporation, Melbourne, Florida. (Chaps. 8 and 31) Lome K. DeSize, Consultant, Littleton, Colorado. (Chap. 23) ‘Raymond H. DuHamel, Antenna Consiltant, Los Altos Hills, California. (Chaps. 14 and 28) Boyton G, Hoguman, Keser, Wight & Hagaran, PC. Abound Veiia, (Chap. 24) Gene K. Huddleston, Director, Radar Engineering, Martin Marietta Electronic Sys- tems, Orlando, Florida. (Chap. 44) Geoftey Hyde, Senior Staff Scientist, COMSAT Laboratories, Clarksburg, Mary- land. (Chaps. 17 and 45) Henry Jask,+ Vice-President and Director of Antenna Systems Division, AIL Divi- sion, Eaton Corporation, Deer Park, Long Island, New York. (Chap. 2) Richard C. Johnson, Consultant, Dallas, Texas. (Chap. 1) Edward B. Joy, Professor, School of Electrical Engineering, Georgia Institute of Technology, Atlanta, Georgia. (Chap. 29) Kenneth S, Kelleher, Consultant, Alexandria, Virginia, (Chaps. 17 and 18) ‘Hugh D. Kennedy, Consultant, Bortola, California. (Chap. 39) + Deceased.Howard E. King.* Director, Antenna Development Office, Antennas and Propaga- tion Department, Communications Systems Subdivision, The Aerospace Corpo- ration, Los Angeles, California. (Chap. 13) (Charles M. Knop, Director, Antenna Research, Andrew Corporation, Orland Park, Minos. (Chap. 30) Dennis J. Kozakoff, President, Millimeter Wave Technology, Inc, Atlanta, Georgia (Chap. 38) John D. Kraus:* Director, Ohio State University Radio Observatory, Columbus, Ohio. (Chap. 41) Edmund A. Lapor,,t Corporate Director of Communications, RCA Corporation, Princeton, New Jersey. (Chap. 11) ‘Stuart A. Long, Professor, Department of Electrical Engineering, University of Houston, Houston, Texas. (Chap. 4) Allan W. Love, Consultant, Corona del Mar, California. (Chap. 15) Roderic V. Lowman, President, Techni-Quest, Inc, Greenlawn, New York. (Chap. 42) John A. Lundin, du Treil, Lundin & Rackiey, In, A subsidiary of A.D. Ring, P.C., ‘Washington, D.C. (Chap. 25) ‘Mark T. Ma, Senior Research Engineer, Electromagnetic Fields Division, National Institute of Standards and Technology, Boulder, Colorado. (Chap. 3) Ali R. Mabrad, Sci-Tech Consultants, Carmichael, California, (Chap. 28) ‘Rober’ J Mailloux, Senior Scientist: Antennas, Rome Laboratory Electromagnetics and Reliability Directorate, Hanscom Air Force Base, Massachusetts. (Chap. 21) ‘Robert E. Munson, Manager, Advanced Antenna Programs, Ball Communication ‘Systems Division, Boulder, Colorado, (Chap. 7) Josh T. Nessmith, Principal Research Scientist, Georgia Tech Research Institute, Atlanta, Georgia. (Chap. 34) Warren B. Offat, Owner, W & B Observatory, Cloudcrof, New Mexico. (Chap. 23) ‘rihur A. liner * Professor of Electrophysics, Polytechnic University, Brooklyn, New York. (Chap. 10) Willard 7. 74.0% Staff Scientist, General Electric Corporation, Moorestown, New Jersey. (Chap, 34) ‘Ronald D. Rackley, du Teil, Lundin & Rackley, Inc., A subsidiary of A. D. Ring, P. C., Washington, D.C, (Chap. 25) Paul E. Rawlinson, Manager, Radar Systems Laboratory, Equipment Division, Raytheon Company, Sudbury, Massachusetts. (Chap. 32) Leon J. Ricardi, President, LIR, Inc., El Segundo, California, (Chaps. 22 and 35) Philip N, Richardson, Texas Instruments, Inc. Dallas, Texas, (Chap. 33) Charles E. Ryan, Jr,* Principal Research Engineer, Georgia Tech Research Insti tute, Atlanta, Georgia. (Chap. 37) James P. Scherer, Senior Vice President of Technology, Loral Randtron Systems, ‘Menlo Park, California. (Chap. 14) James M. Schuchardt, Director, Antenna Division, American Electronic Laborato- ries, Lansdale, Pennsylvania. (Chap. 38) Mary Beth Selby, Senior Engineer, Haris Corporation, Melbourne, Flori. (Chap. 8) Retired. 1 Decent.Glenn S. Smith, Professor, School of Electrical Engineering, Georgia Institute of Technology, Atlanta, Georgia. (Chap. 5) Vernon C. Sundberg,* Section Head, GTE Systems, Mountain View, California, (Chap. 40) Calvin T. Swif, Professor, Department of Electrical and Computer Engineering, University of Massachusetts, Amherst, Massachusetts, (Chap. 31) Chen To Tai Professor of Electrical Engineering, University of Michigan, An ‘Arbor, Michigan. (Chap. 4) ‘Raymond Tang, Manager, Array Antenna Laboratory, Communications and Radar Division, Hughes Aircraft Compeny, Fullerton, California. (Chap. 20) Robert C. Voges. TI Feliow, Texas Instruments, Inc., Dallas, Texas. (Chap. 9) Harold R. Ward, Consultant, HRW Consulting, Bedford, Massachusetts. (Chap, 32) Harold . Wheeler® Chief Scientist, Hazetine Corporation, Greenlawn, New York, (Chap. 6) Ronald Wilensky, Vice President, Technology for Communications International, Mountain View, California. (Chap. 26) Jimmy L, Wong, Senior Engineering Specialist, Antenna Development Office, An- ‘ennas and Propagation Department, Communications Systems Subdivision, The Aerospace Corporation, Los Angeles, California. (Chap. 13) Roy B. Woolsey, Vice President, Engineering, Technology for Communications International, Sunnydale, California. (Chap. 39) Daniel F. Yaw, Senior Advisory Engineer, Electronic Warfare Systems Engineering, Westinghouse Defense and Electronic Systems, Baltimore, Maryland. (Chap. 40} ‘Hung Yuet Yee, Seniot Member of Technical Staf, Texas Instruments, fnc., Dallas, ‘Texas. (Chap. 9) Francis J. Zucker, Physicist, Antenna Section, Rome Laboratory, Hanscom Air Force Base, Massachusetts. (Chap. 12)Preface THIRD EDITION. 1t has been more than three decades since Henry Jasik envisioned and edited the Fist Edition of the Antenna Engineering Handbook. During this time, many significant and. far-reaching advances have been made inthe field of antenna technology. ‘The Second Edition was published more than two decades after the First Edition, and it required many major revisions and additions. For example, it contained 15 new chapters to cover new subject areas, Work on this Third Edition of the Handbook began 10 years after starting work on the Second Edition. Ten years is along time given today’s rapid advancementsin antenna technology, so an update is welcome and beneficial. ‘This new edition again contains four major parts: Introduction and Fundamentals Part | presents basic comments, defines parame- ters, and discusses fundamentals that are common to most antennas. ‘Types and Design Methods Part 2 presents the primary astenna types and design methods that currently are in use. Emphasis is on succinct descriptions, design data, and references. Applications Part 3 discusses major applications of antennas. Emphasis is on how ‘antennas are employed to meet electronic system requirements. Design methods which are unique to the applications are presented. ‘Topics Associated with Antennas Part 4 deals with topics that are closely related to antenna design The topics ae covered sucinay, but more detailed information canbe found in the references. ‘Thanks are extended to the many publishers who have granted permission to use ‘material from their publications. As was done in the previous editions, we have tried to credit all sources of information by relerences, any omissions are due to oversight rather than intent,‘The work of many outstanding engineers who reviewed and updated the individual ‘chapters has made this Third Edition ofthe Handbook possible. Their magnificent efforts enabled this project to be completed according tothe inital schedule. Ttis impossible to acknowledge all the other individuals who have made contribu- tions to this book, so thank them all asa group. Again, extend special thanks to Maurice 'W. Long and J. Searcy Hollis, who introduced me tothe fascinating world of antennas. ‘Their keen insight and guidance helped me, in many cases tose” the electrical currents flowing on an antenna. Finally, I want to acknowledge the lasting contributions of Henry Jasik, who con ‘ceived and edited the First Edition ofthis Handbook. When I took the assignment to edit the Second Edition, I considered several alternate ways to organize the subject matter; ‘however, in the end, I concluded that Henrys outline was better than all the others. This ‘Handbook is dedicated to the memory of Henry Jask. His influence is clearly present in this Third Edition, and it probably will continue through future editions. RICHARD C. JOHNSONAntenna Engineering HandbookPART 7 Introduction and FundamentalsChapter 1 Introduction to Antennas Richard C. Johnson Georgia Institute of TechnologyFUNCTIONS AND TYPES ‘An antenna usually is defined as the structure associated with the region of transition between a guided wave and a “free-space” wave, or vice versa. The adjective free- space is in quotation marks because in practice there always is some interaction with the surroundings. On transmission, an antenna accepts energy from a transmission line and radiates it into space, and on reception, an antenna gathers energy from an inci- dent wave and sends it down a transmission line. ‘When discussing an antenna, one usually describes its properties as a transmit- ting antenna. From the reciprocity theorem, however, we know that the directional pattern of a receiving antenna is identical with its directional pattern as a transmitting, antenna, provided nonlinear or unilateral devices (such as some ferrite devices) are ‘not employed. Thus, no distinction needs to be made between the transmitting and receiving functions of an antenna in the analysis of radiation characteristic. It should be pointed out, however, that the reciprocity theorem does not imply that antenna ‘current distributions are the same on transmission as they are on reception. ‘A large variety of antennas have been developed to date; they range from simple structures such as monopoles and dipoles to complex structures such as phased arrays. The particular type of antenna selected for a certain application depends upon the system requirements (both electrical and mechanical) and, toa lesser extent, upon the experience of the antenna enginee®. BASIC CONCEPTS AND DEFINITIONS: Consider an antenna which is ocated at the origin of a spherical coordinate system as strated in Fig. 1-1. Suppose that we are making observations on a spherical shell having a very large radius r. ‘Assume that the antenna is transmitting, and let . Po = power accepted by antenna, watts P, = power radiated by antenna, watts 1 = radiation efficiency, unitless anrenna Cocarion FFIG, 1-1 An antenna ina spherical coordinate ‘ystem. “4‘The above quantities are related as follows: 2 ane (41) Let 6 £(0.9) = radiation intensity, watts/str Note that since was assumed to be very large, is independent ofr. Ths indepen- dence of risa characteristic ofthe far-field region. The total power radiated from the antenna is P= ff a0.) sin asco (12) and the average radiation intensity is Be . a = Be (43) Lat : 240) = circtivity, unites Directivity is a measure of the ability of an antenna to concentrate radiated power in 4 particular direction, and itis related to the radiation intensity as follows 260 (1-4) ‘The directivity ofan antenna isthe ratio ofthe achieved radiation intensity in a Particular direction to that of an isotropic antenna, In practice, one usually is inter ‘sted primarily in the peak directivity of the main lobe. Thus, if one says that an antenna has a directivity of 100, iti assumed tbat 100 i the peak directivity of the main lobe. Let . G(6,8) = gain, unitless ‘Lie gain of an antenna is related to the directivity and power radiation intensity as follows: 66) = 1000) 196.8) (1-5) *"P/4e and from Eq. (1-1), G04) = ae (+6) aus, the gain is @ measure of the ability to concentrate in a particular direction the Power accepted by the antenna. Note that if one has a lossless antenna the directivity and the gain are identical. Let _ PQO.d) = power density, watts/square meter108) +7; POg) = =F fed Substituting Eq. (1-6) into Eq, (1-7) yields Po , POO.) = G00) Pos (8) The factor Po/4ar? represents the power density that would result if the power accepted by the antenna were radiated by a lossless isotropic antes Let . Ad0.8) = effective area, square meters It is easier to visualize the concept of effective area wlten one considers a receiving antenna; itis a measure of the effective absorption area presented by an antenna to fan incident plane wave, The effective area is related! to gain and wavelength as follows: (09) = © G00) (1-9) Many high-gain antennas such as horns, reflectors, and lenses are sid to be aper- ture-type antennas. The aperture usually is taken to be that portion of a plane surface ‘eat the antenna, perpendicular to the direction of maximum radiation, through which ‘most of the radiation flows. Let . ‘ta antenna efficiency of an aperturetype antenna, unitless . A = physical area of antenna’s aperture, square meters Then, uaS (1-10) has been called aperture efficiency aperture antennas, we see from Eqs. (1-9) and (I-10) that an Game (411) ‘The term ng actualy is the product of several factors, such as, Me = TTT © (1-12) ‘The term mis radiation efficiency as defined in Eq. (1-1). The term 7, is aperture illumination efficiency (or antenna illumination efficiency), which is @ measure of how well the aperture is utilized for collimating the radiated energy i is the ratio of the directivity that is obtained to the standard directivity. The standard directivity isIntroduction to Antennas 4-7 ‘obtained when the aperture is excited with a uniform, equiphate distribution. (Such a distribution yields the highest directivity ofall equiphase excitations.) For plana fpertures in which A > A the standard directivity is 44/2, with radiation confined toa half space. ‘The other factors, nym. include all other effects that reduce the gain of the antenna, Examples are spillover loses in reflector of lens antennas, phase- Ric (1-19) where kis beamwidth constant 2 is the wavelength, and Dis the aperture dimension i che pfane of the pattern, ‘Mast antenna engineers seem to use a valve of k = 70°. This is adequate for ‘most rough estimates; however, more accurate estimates must take into account the fact that the value of & depends upon the aperture illumination function. Generally speaking (but not always), illumination functions that yield lower sidelobes result in a larger value of k Komen reported on the variation of the beamwidth constant for teflector-type ‘antennas. Froat computed patterns for various edge illuminations, he determined that k= 1.052387 + s5.9486 (1-20) where 1 is the absolute value of edge illumination (Including space attenuation) in decibels and & is in degrees. (In practice, one normally would calculate & to only a few significant figures.) By applying Eqs. (1-19) and (1-20) to measured data from several antennas, Komen concluded that the relationship between beamwidth and edge illumination holds regardiess of frequency, reflector size, reflector type, or feed ‘ype. The beamwidth constant and approximate sidelobe level versus edge illumination for reflector type antennas are illustrated in Fig. 1-9. The beamwidth constants for antennas having one of several special aperture distributions are illustrated in Figs. 46-7 and 46-9 in Chap. 46. ‘A convenient rule of thumb for predicting gain (of a relatively lossless antenna) s K one (4-21) Where K is @ unitless constant and 0, and 6; are the 3-dB beamwidths (in degrees) in the two orthogonal principal planes. The correct value of K fot an actual antenna depends on the antenna efficiency. A popular value used by many antenna engineers '8 30.000, but many other values are in ase. For example, Stutzman and Thiele! sug. eS! 2 value of 26,000, and Stegen"’ suggests a value of 35,000. One should subtract Stout 1 €B from the estimated gain for amennas with cosecantaquared shaped 7,1618 biroduetion and Fundamentals 85 80 75 jelobe Levels SLL (4B) 70 Beamwidth Constant, k(*% 65 Theoretical 60 -25 -20«--15 -10— = 0 Edge illumination | (48) Fi. 1-9. Bearwidth constant and approximate fst sidelobe love versus edge ituination {eiudng spece etensator) for paraboloiareector antennas. Tho SLL curve isso for plane-wave feds and dashed for feeds having parebote primary pattems (cB versus ang) REFERENCES: 1S. Sire, Mlerowave Antenna Theory and Design, MeGrawAHill Book Company, New ‘York, 1949, 266 2.14 2 IEEE Standard Test Procedures for Amennat, EEE Std, 149-1979, aatitateof Electrical ‘and Electronics Enginoers, New York, 1979, sec. 11. 3. ILS. Hollis etal, Techniques of Microwave Antenna Measurements, John Wiley & Sons, Tne, New York, 1984 4 RC. Hansen etal, “IEEE Test Procedures for Antennas; Number 149 (Revision of 48 IRE 252) January 1965,” [EEE Trans. Antennas Propagat, vol. AP-I3, May 1965, pp. 437-466. 5 R.C, Hansen, “Aperture Theory," in R.C. Hansen (ed), Microwave Scanning Antennas, ‘Academie Press, Inc, New York, 1964, pp. 31-32. 6 Silver, op. cit. sc. 69. 7 HLT. Fri, “A Note on a Simple Transmission Formula," IRE Proc, May 1946, pp.254~ 256, 8 M.J. Komen, “Use Simple Equations to Calculate Beam Wik 1881, pp. 61-63 ‘Microwaves, Decemberbhroducton to Antennas 1-47 9 EM. T. Jones, “Paraboloid Reflecior and Hyperboloid Leas Antennas,” IRE Trans Antennas Propagat, vol. AP-2, July 1954, pp. 119-127. 10 W.L, Stuteman and G. A. Thiele, Antenna Theory and Design. John Wiley & Sons, Ie, New York, 1981, p. 397, 4AM R.J,Stegen, “The Gain-Beamwidth Product of an Antenna,” JEEE Traw. antennas Prop- ‘gat, vol AP-12, July 1968, pp. 505-296.Chapter 2 Fundamentals of Antennas * Henry Jasik AIL Division of Eaton Corporation 21 Radiation from Blectric Curent Elements 22 2-2 Radiation from Magnetic Current Elements 25 2-3 Antennas above Perfect Ground 27 2:4 Radiation from Apertures 2-8 ‘The Equivalence Principle 2-8 Appliations ofthe Equivalence Principle 2-10 2S Impedance Properties of Antennas 2-11 Self-Impedence 21 Mutual Impedance 2-13, 26 _Directvty Patterns from Continoous Line Sources 2-14 Line-Source Distributions 2-15 27 Patterns from Atea Distebutions 2-19, Rectangular Apertures 2-19 CCireular Apertures 2-19, Elliptical Apertures 2-21 2-8 Effects of Phase Errors on Line Sources 221 29 Effect of Random Errors on Gain and Sidelobes 2.25, Diserete-Element Arrays 227 Continuous Apertures 2-31 Periodic Errors in Aperture Iumination 2-32 2-10 Methods of Shaping Patterns 2-34 Methods Useful for Linear Arrays 234 “Methods Usetal for Reflector Type Antennas 2.38 Gain Limitations for an Aperture of Specified Size 2-39 Scale Models of Antennas. 2-40 “This chaper was writen by the ae Henry Jak forthe rt eon; it as ben shortened and edited hu by Ricard C. Johnson fr fcsion fn thi eon.2-1. RADIATION FROM ELECTRIC CURRENT ELEMENTS 00 ‘One of the types of radiators frequently used in antenna practice is some form of thin wire arranged in a linear configuration, Ifthe current distribution on such a wire i, Known or can be assumed with a reasonable degree of accuracy, then the radiati pattern and the radiated power can be computed. This computation is based on the Integration ofthe effects due to each differential element ofthe current along the wie. It is therefore of interest to set down the complete expressions for the fields at any distance due to a differential element of Current oriented along the z axis as shown in Fig. 2-1. The rms electric and magnetic feld components are given as follows: e,= costae (Be - po] noe ppaz(L-A-1 cd aoe [ Gr GF ae [ 1d | - n= jE tae | > sind em en Tag lar” BAF Epa H,= Hy 0 where / dz = moment of diferential current element (Fis given in rms amperes and aris given in meters.) Gistance, m, to observation point p= mr = wavelength, m vat is given in volts per meter Hisgiven in amperes per meter ‘A time factor of e has been omitted, since forall the cases in which we are interested itis asumed that we have a sinusoidally time-varying current of constant frequency. FIG. 2-1. Coordinate system for an ecto dipole.Fundamentals of Antennas 2-3 For most problems of interest itis only necessary to know the components in the {at field ie., when r is very much greater than the wavelength. Under these condi- tions, the field components are simply given by , 3081 de ‘These expressions apply only for a very short element of current having a con- stant value along its length. However, they may readily be used to determine the field from any wire having a known current distribution by integrating the fieid due to each of the differential current elements along the length of the antenna. Taking into account the variation of current and the phase differential due to the varying distance. from the observation point to each current element, the general expression for the field of any cuctent distribution becomes GOrsing [1 core 2) dz mo _ ome aap whore both I(2) and 7(2) are now functions of 2 and the integration takes place along. the length of the antenna from —€/2 to +£/2. For very short antennas, the above expression can be simplified to 60x sin @ x vi Toke where Jo = current at center of antenna L, = effective length of antenna defined as wn 1a) dz The effective length is of interest in determining the open-ircuit voltage at the ter- tinal ofa receiving antenna, It is also used on occasion to indicate the effectiveness of a trarsmitting antenna For a short toploaded linear antenna which has uniform current distribution as shown in Fig. 2-2a, te effective length is simply equal to the physical length. For a short antenna which is much les than a half wave long, as shown in Fig. 228, the curren distribution is essentaly triangular and its effective length is one-half ofits Physical length, For antennas with an overall length greater than about a quarter wavelength, the Yarlaton of the phase term cannot be neglecied and the integral must be evaluated by taking this term into account. The method will be very briefly illustrated for the case of thin half-wave radiator which can be assumed to have & sinusoidal current2-4 Introguction and Fundamentals UP @ wo FIG. 2-2 Curent istbution ona shortlin: FIG. 2-3 Coordinal far entamna. (2) With top loading. (6) tion from a hait-wave dips. Without top loading istribution so that 1(2) is given by fy cos 82. The geometry for finding 2) in Fig. 23, from which its readily seen that Ha) =r 26088 “The field for a half-wave dipole is then given by which reduces to ‘The relativecadiation pattern for a half-wave antenna is shown in sold tines ia Fig. 2-4, For comparison purposes, the relative-radiation pattern ofa very short dipole js shown in dotted lines. The patterns shown are those in a plane which contains the ‘axis for the antenna. ‘The pattern in the plane perpendicular to the antenna is perfectly citoular because of symmetry. There are a umber of other properties for the half-wave dipole which are of considerable interest, such as radiation Tsistance, gain, and input impedance. These properties are discussed in Chap. 4. ‘The method of computing radiation =~ —— patterns for thia linear radiators is basic regardless of the length or complexity of shape. As a matter of interest, the follow- ing formula gives the radiated field from a ‘centered thin wire of arbitrary length ¢ with an astumed sinusoidal current distribution: ed SS — HALe-WivE DIPOLE SHORT DIPOLE p= 2 FIG. 24. Radiation patterns. Y 71 willbe noted that the radiated feld perpendicular tothe antenna continues to increase asthe length i increased until the overall length is about 1X wavelengths. Beyond this point the field stars falling of. When the overall fength is two wave- lengths, the eld is zero normal to the axis of the antenna. For this length the radia- tion patera has broken up into two major lobes which are directed off the normal to the antenna, For a stil longer length the sntcana pattcrn will continue to break up into large number of lobes whose positions depend on overall length of the antenna (Chaps. and 11). 2-2 RADIATION FROM MAGNETIC. CURRENT ELEMENTS: ‘Another basic radiator which is frequently used in antenna practice is a magnetic current element. Although magnetic currents do not exist in nature, a number of com figurations produce fields identical with those which would be produced by a fictitious ‘magnetic current. For instance, a circular loop carrying electric current whose diam- ‘ae is vecy small in terms of wavelengths will produce fields which are equivalent to those of short magnetic dipole. The fields far any distance are given by the following expressions: E, = 308? dm ik -&] sin 0 e-# ir mor gip] re Gr? BrP =f am[ td ten Sono B= B= HO where the coordinate system is as shovn in Fig. 2-5, and dim is defined as the differ- ‘ential magnetic dipole moment. For a smalleliameter loop, the magnetic moment of the loop is equal to the electric eustent / lowing through the loop times its area 4. For the far field, when 7 is very much greater than the wavelength, the field components reduce 10 CN res ‘sin 0 ei ee inde It will be noted that the field expressions for the magnetic current element are ‘almost exactly analogous to those forthe electric current element except for the inter- ‘change of electric and magnetic quantities. The radiation of a short magnetic dipole ora smallaiameter loop is also a doughmist pattern, as inthe case ofan electric dipole. For a small oop the radiation pattern inthe plane of s loop is perfectly circular, while2-6 introduction and Fundamentals x te wo FIG, 2-8 Coordinate systems for a mag FIG. 2-6 (a) Thin slot in ground plane atic dipole. (2) Complementary dipole. the pattern in the plane through the axis of the loop isa figure of eight whose ampli- tude is proportional to sin @. The expressions given are accurate for loop diameters Which are considerably less than one-tenth wavelength. As a matter of fact, for very “Small loops, the radiation pattern does not depend on the exact shape of the 100p, ‘which may be square, rectangular, or some other shape, provided the overall circum ference remains much less than a quarter wavelength. For loops whose diameter is ofthe order of the wavelength, the radiation pattern ‘can deviate considerably from the doughnut form, depending on the nature of the carrent distribution along the loop and diameter of the loop. These considerations are ‘treated in Chap. 5. "Another antenna whose radiation characteristics are essentially similar to those cof a magnetic dipole consists of a very thin slot in an infinitely large metallic ground plane, as shown in Fig. 2-62. For this type of antenna, the electric field is applied Peross the narrow dimension of the slot. Ic is possible to show that the field radiated by this slot is exactly the same as would be radiated by a fictitious magnetic dipole witha magnetic current distribution M whichis numerically equal to the distribution Walectrie voltage V across the slot. Thus the radiation pattern ofa thin rectangular Slo is identical with the radiation pattern of the complementary electric dipole which sold just fl the slot. as shown in Fig, 2-66. The only difference between the two {ypesof radiator isthe fact that the electric and magnetic quantities have bees inter~ cpenged. This complementary relationship has been treated in considerable length in technical literature and will not be elaborated on in this section. Tis pertinent to point outa few precautions concerning the application of these complementary relationships, For one thing, the complementary relationship is based oo the assumption that only electric field exists within the slot end that no magnetic Geld is present. Ths is true only for vanishingly thin slots. For slots whose width is appreciable in comparison with their length, the above assumption is no longer valid ah the radiation pattern will be somewhat modified as compared with the case of the ‘ery thin slot, Another assumption on which the complementary relationship is based 1b that the size of the conducting ground plane is infinitely large. This, of course, is lpever true in practice, and for nite ground planes the size of the ground plane may exert a large infivence on the radiation pattern, particularly at angles close to theFundamentals of Antennas 2-7 plane of the sheet, Even for sheets which are fairly large in terms of wavelength, there are some minor modifications of radiation pattern. ‘It might also be mentioned that the complementary relationship holds only when ‘the sot i cut in a large fat ground sheet. For sats cut in. circular cylinders, the pattern ‘can be considerably different from that predicted by the complementary relationship, particularly in the plane perpendicular to the axis ofthe cylinder. However, use of the Complementary relationship is still very useful to engineers as an intuitive guide, and while in many cases the results will not be exact, they will certainly give a first approx- imation to the actual radiation pattern 2:3_ ANTENNAS ABOVE PERFECT GROUND “The characteristics of antennas operating near ground level will be modified by the effect of ground reflections. This is particularly true of antennas operating at frequen- cies below 30 MHz when the height of the antenna above ground may be less than ‘one or two wavelengths. For airborne antennas or for large-apesture, nartaw-beam antennas, in which the main beam is elevated upward at leat several beamwidths, the ground may play relatively small role. In all cases, the ground will play a part in the propagation between transmitter and receiver, and to compute its effect the charac- ‘ersties of the ground and the geometry of the propagation path, as well as the pattern characteristics of the receiving and sransmitting antennas, must be known. ‘For antennas at relatively small heights above ground, the ground is basic part of the antenna system and will affect not only the radiation pattern of the antenna, but also its impedance properties. To obtain a first-order estimate of the effect of ground, imple image theory can be applied to the case of a perfectly conducting ground sur- face. It is well known from electromagnetic theory thatthe tangential field On. perfect ‘conductor must be zero and that the electric field must be normal to the conducting surface. To satisfy this requirement, a conducting charge above a conducting plane will induce a charge distribution oa the plane exactly equivalent to that which would be produced by an equal charge of the opposite sign at the same distance below the plane, as shown in Fig. 2-7a. Since current is a movement of charge, itis readily possible to deduce the direction of the images for vertical, horizontal, and inclined wires above ground, as shown in Fig. 2-7b-d. The image for any other configuration ‘can readily be determined by using the rule that vertical components are in the same direction while horizontal components are in opposite directions. ae ANTENNA spouno ~ ace «a 9 te we FIG. 2-7 images above @ pertectly conducting ground plan ‘charge. (2) Vertical wire. (2) Horizontal wire. (2) Inclined wire. (4) Point2-8 invadscton and Fundamentals The field in any direction above the ground plane can easily be determined by replacing the ground plane by the image and computing the resulting field due to the antenna plus its image. This is valid above the ground plane, since for a perfect con- ‘ductor the field below the ground plane is zero. ‘Although the above method is rigorously true only for antennas above highly ‘conducting ground, it does give good results in many cases of interest. The same tech- nique can be used when the ground has arbitrary values of conductivity and dielectric constant by assuming an image current which is related to the antenna current by the ratio of the complex-reflection coefficient for the appropriate angle of incidence. ‘The effect of the ground on the impedance of an antenna can also be determined by image theory. For instance, the input impedance of an antenna above perfectly ‘conducting graund is simply the input impedance of the antenna in free space plus the ‘mutual impedance due to the image antenna, For arbitrary ground, the same sort of relation is still true except that the mutual impedance due to the image must be mu lied by the complex-groundsreflection coefficient for normal incidence. ‘A special case of considerable interest occure when one end of the antenna above ground is exactly one-half of the input impedance of the antenna plus its image when driven in free space. For example, the input impedance of a quarter-wave dipole above ‘round is exactly one-half that of a half-wave dipole in free space. 2:4 RADIATION FROM APERTURES. ‘The computation of rediation patterns for linear-wire antennas is relatively simple if the current distribution on the wire is known. The current distribution is not usually known exactly except for a few special cases. However, physical intuition or experi ‘mental measurement can often provide a reasonable approximation to the current dis- tribution, and for many engineering purposes a sufficiently accurate result can be ‘obtained. In theory, of course, an exact result can be derived from a boundary-value solution if the nature of the exciting sources is known, From « practical point of view, the amount of {aor involved in obtaining numerical results is excessive even for those cases in which the geometry is relatively simple and a rigorous solution can be expressed in terms of a series of tabulated functions. Its therefore necessary in many situations to be able to compute the radiation pattern by alternative methods based (ona reasonable assumption of the nature of the electromagnetic fields existing in the vicinity of an antenna structure. ‘The Equivalence Principle One powerful technique for simplifying this type of computation makes use of an ‘equivalence principle given by Schelkunoff. Briefly stated, this principle supposes that a given distribution of electric and magnetic fields exists on a closed surface ‘drawn about the antenna structure. These fields are then canceled by plucing a sui able distribution of efectric and magnetic current sheets on the closed surface so that the fields inside the closed surface are zero, The radiation is then computed from the clectric and magnetic current sheets and, except for a difference in sign, is identical to the radiation which would have been produced by the original sources inside the closed surface, Tt will be noted that this principle is essentially a more rigorous for-Fundamentals of Antennas 2-9) Trane cena iazeens’ Principle. If the fis onthe closed surface are nown exactly, then the cenuting computation is also exact, The degree of approximation which cc be obtained by this technique depends only on how accurately the Selde sense aan closed surface may be estimated shen the electric and magnetic fld strength are respectively given by F and i, then the equivalent densities are given by Electric current density JeaxHn ‘Magnetic current density M= -aXxE yiere both J and 77 are expressed in amperes per meter and M and E are expressed in volts per meter, Fete ae gos Product has been used to show thatthe electric current i per- ppendicular to direction of ‘eld vector. f the intial ‘De way of visualizing the effect of a small portion ofthe wavefront isto consider the physical equivalent ofthe electric and magnetic current sheets, The tlavine son to TS MAGNETIC DIPOLE oR BLEctiNe math tobe" wo te oaSlE FIG, 2-9 Radiation pattems in plane of " sectric curent loop. («) —— Pattern of tan rg QUTEN Shestflatons. (2) Insc eurent oop. Panera Sins fol, (2) Reaauing elette Ie dpa. (8) Cardo patton ans eee ciecrengnate curants, (¢) Orentation of bination of elctre ourert leer sna Nacory ‘“lctricdpole and electric ourentioop. dipole2-10 Inroduetion and Fundamentals 2, The electric dipole is oriented in the direction of the eiectsic field E, while the elec 2 geurrent loop is located in the plane defined by the electric field £ and the direction of propagation n, as shown in Fig. 28c. Tree look at the radiation pattern in the plane of the loop, the component due to the electric dipole wll have a cos 0 type of variation while the component due to the electric current loop will have 2 cit- ‘cular pattern as shown in Fig. 29a. For portion of wavefront in free space in Which the E and HY fields are related as JI 120rH he restive amplitudes ofthe two om- A the a areata ation wo pattern of the combination is given by (1 Bicas@) of a cardioid pattern as shown in Fig. 296. “Although the concept of the electric and magnetic current sheets is most use- fal for derivation purposes. and_ for taining a physical picture of radiation from apertures, its possible o compute the reradiated field from 3n_ aperture dliyectl in terms ofthe tangential electric and magnetic eld components in the apertue itself. With reference to Fig. 210, the Sperturs les in the xy plane with feld components ES° and #59, At a distant point Pr the radiation components ofthe electric field are given by wee con 8) 82 dx dy 7 LES) + 1208H 6) Dr dx dy eo" & FIG. 2-10 Coordinates for current sheet ‘computation. hs eee et Be a dy ‘The above is true for an aperture in which EY? and Hi take on arbitrary values. ‘An important special case is that in which the field components are the same as exist ina free-space plane wave. For this case BD = 12001? and the distant fields are now given by Ey = + iE cos 8 + 1208") Ey = -JE9 + 0s) SSE de dy em an 4 enn tte dye™ £, = JEU + 028) SP ae dy Applications of the Equivalence Principle ‘One typical application of these results isto the problem of determining the radiation pattems from an electromagnetic hotn’ The tangential fields at the aperture of such Bihorn can be approximated by assuming that the fields atthe aperture plane are theFundamentals of Anieonas 2-14 same as would be obtained ifthe guiding surfaces of the horn were extended to infin- ity. Although net exact, this assumption is quite good when both linear dimensions of the aperture are greater than one or two wavelengths. After this astumption of the cleciric and magnetic fields at the aperture, the far-field relationships are then for- ‘mulated by taking the fields across the aperture and integrating over the aperture, taking into account the amplitude and phase variations of the field incident on the aperture and the phase differential due to the varying distance from the observation point to each area element of the aperture. ‘The above method gives results which are good 10 2 high degree of approxima- tion, particularly for large-aperture horns. The equivalence principle can also be used ‘with small-aperture horns, but caution must be exercised with regard to the assump- tions of the relative values of electric and magnetic field strengths. In small horns, it isnot true that the aperture fields are those which would exist if the waveguide were extended to infinity. For instance, in a waveguide radiator whose E-plane width is small, the field at the aperture will be predominantly electric and the magnetic field ‘may be quite small. For this condition, the equivalent current sheet will consist mainly ‘of magnetic current, and for a small aperture this will have a substantial effect on the E pattern, particularly in the E plane. For those cases in which itis possible to measure the standing waves in the waveguide leading to the aperture, itis possible to-estimate the relative values of electric and magnetic field strength at the aperture and to make a first-order correction to the radiation-pattern computation. ‘The equivalence principle can also be used to determine the radiation from the ‘open end of a coaxial line. For this case, the field at the aperture is predominantly electric, with only a small component of magnetic field. The field variation in the aper- ture is assumed to be that of the dominant mode in the coaxial ine. The problem is readily formulated in cylindrical coordinates.‘ ‘Another problem which can be solved by using the equivalent-current-sheet ‘method is the case of seflection from a conducting sheet such as paraboloidal reflector ‘or a plane reflector a8 employed in microwave relay applications. To formulate this ‘problem, the tangential fields that would have existed ifthe reflector were not are determined by using only the magnetic field term for the aperture field and inte- tating over the surface, taking into account variations of the aperture field amplitude land phase and the phase differential to the observation point. 2:5_IMPEDANCE PROPERTIES OF ANTENNAS Self-tmpedance ‘Ve impor impedance ofan antenna is a characteristic of considerable intrest o eng ‘eer since they afe concerned with the problem of supplying the astenaa with the ‘maximum amount of transmitter power available or in abstracting the maximum amount of received energy available from the antenna. Except forthe simplest types of antenna configuration, the theoretical computation ofthe input impedance is an extremely arduous task, and for a large number of antenna types it is usually easiest to make a direct experimental measurement of the input impedance. However, for linear antennas which are relatively small in siz, i is posible to make some reason- ably good estimates as tothe magnitude ofthe input resistance. Its also possible to ‘assess, with a reasonable degree of accuracy, the mutual impedance between linear212 brocucton and Fundamentals radiators for the purpose of estimating the input impedance of an individual element in an array of radiators. In most practical cases, the input impedance of even a simple antenna is affected toa considerable degree by the terminal conditions at the point where the transmission ling feeds the radiator. For most accurate results itis therefore necessary to measure the various impedances involved and to use the cafculated values primarily as a guide during the design procedure. For very short wire dipoles, the radiation resistance is a quantity which is closely allied with the resistive component of the input impedance. The radiation resistance is normally defined as the ratio of the total power radiated by an antenna divided by the square of the effective antenna current referred to a specified point. For short antennas this isa useful quantity because it enables one to estimate the overall radia- tion efficiency ofthe antenna by separating the radiation component of the input resis- tance from tke loss resistance due to the ground system or the loss resistance due to the impedance-matching elements To compute the radiation resistance, it is necessary to know the radiation field pattern ofthe antenna in terms of the current flowing at the point 10 which the radia- tom resistance is referred. The total radiated power is then computed by integrating the total power density passing through a sphere surrounding the antenna. This com- putation will be carried through very briefly for the case of a very short dipole having. fan effective length Ly and carrying a current Jy The electric feld intensity for this is given in Sec. 2-1 as ont 1B) = Eee sin 6 ‘The power density in the far field is given by the Poynting vector, which is equal to 3/120n, where the electric feld is given in rms volts per meter and the power de is expressed in watts per square meter, Integrating over a large sphere surrounding the antenna, we then obtain pa [Eat sino as where Ris # radiation resistance in ohms. ft will be noted that the important variable is the ratio ofthe effective length L, 10 the wavelength and that the larger this ratio the greater the radiation resistance. Thus for a short antenna of physical length ¢ which is topoaded so as to give a uniform current distribution, the effective length sill be equal to che physical fength and the radiation resistance will be 80x222/X?, For ‘an antenna with no top loading, the current distribution will be triangular (Sec. 2-1), and the effective length will Be equal to one-half of the physical length, 30 that the radiation resistance will be 20x223/2?, It ean be seen that the radiation resistance for Che uniform-curtent case is thus equal 14 times that obtained inthe case of triangularFuncamentais of Antennas 2-43 current distribution, despite the fact thatthe radiation pattern and directivity are the same for both antennas since thei length is small with respect to the wavelength, ‘The problem of computing the input impedance of a dipale of finite diameter whose length is ofthe order of a haif wavelength is one that has been considered ac sreat length by a number of writers. The subject is treated in detail in Chap. 4 where ‘a considerable amount of measured data is presented. For the case of a very thin half-wave dipole, the input resistance may be com: ‘uted by assuming a sinusoidal current distribution and integrating the total pooey radiated over the surface of a large sphere, in the same fashion as just described for the short current clement. Using the far field for the hat-wave dipole as given in Sec. 241 and performing the appropriate operations, the input retistaace of the hall-oseg thin dipote is found to be 73.1 9. The reactive component of the input impedance annot be determined by the far-sield method sines the reactance is governed primar, ily by the electromagnetic fields in the vicinity of the antenna itself The input tose: tance is also a function of the relative diameter ofthe dipate and ofthe terminal com, ditions at the driving point. Mutual Impedance ina array of antennas, the drving-point impedance of an individual clement may differ considerably from its seltAmpedance because ofthe effect of mutual coupling with other elements of the array.* In a multielement array, the relations betwees he currents and voltages are given by Wom hZut h2ate os + hin W=hZat hla t+ + 12y Va = Nkig + InZag ++ Lag Where V4 = impressed voltage at nth element 1, = current flowing in nth element Zu = selimpedance of mth element 2m = Znn = mutual impedance between mth and nth elements The driving-point impedance for element 1, for instance, is found from the ratio impressed voltage to the current and is obtained from the above equations 2s of eh2-14 nroduetion and Fundamentals ‘An alternative method for accurately controlling the current distribution in cer- tain types of arrays is to use a transmission-line distribution system which forces the required current to flow in an antenna element regardless of the effect of mutual ‘impedance. For instance, the constant-current properties of a quarter-wave line are such that the current in @ load at the end of a quarter-wave line i equal tothe driving voltage divided by the characteristic impedance of the quarter-wave line regardless of the load impedance, This property is also true fora fine whose length is an odd number Of quarter wavelengths. Thus, for example, in order to feed an array of four dipole elements with exactly equal currents regardless of mutual coupling, the length of transmission line from the dipole to the junction would be an odd number of quarter wavelengths. By making use of the constant-voltage properties of a half-wavelength transmission line, it is possible to build up a distribution system to feed a large number of antenna elements by means of combinations of half-wave and quarter-wave lines, It is worth mentioning that although the uniform half-wave line behaves as a voltage transformer with a transformation ratio of {, itis possible to obtain other transfor- ‘mation ratios by constructing the half-wave line of two quarter-wave sections of dif- {eting characteristic impedances. ‘In many situations itis not possible to sidestep the effects of mutual coupling and itis necessary to have a reasonably accurate estimate of the value of mutual imped ance between antenna elements. It is possible to calculate the mutual impedance for very thin dipoles, and the results are given in Chap. 4 for several cases of interest. Although the finite diameter of @ dipole does have some effect on the magnitude of the mutual impedance, the effect is @ second-order one and for many computations may be neglected. This is not true for self-impedance, whose value is very definitely a function ofthe dipole diameter. lements other than the simple dipole or slot radiator, litle theo- retical work is available on the magnitude of mutual-coupling effects and itis neces- sary to use experimental methods for determining the mutual impedance. Even in the case of dipole element, it is frequently desirable to measure the mutual impedance, particularly for a dipole whose diameter is not small compared with its length. Several experimental methods are available. When the antenna elements are identical and reasonably small physically, one simple method is to measure the input impedance when the element is isolated and then to repeat the measurement when a ‘ground plane is placed near the element to simulate the effect of an image. The dif ference between the two impedance measurements is the mutual impedance for a dis- tance corresponding to the distance between the driven clement and its image. An alternative method when two elements are available isto measure the input impedance when one element is isolated, and then to repeat the measurement when the second clement is in place and has short circuit across its terminals, 2-6 DIRECTIVITY PATTERNS FROM CONTINUOUS LINE SOURCES For antenna systems that have apertures which are very large in terms of wavelength, it is frequently desirable to use a continuous type of aperture distribution because of the relative simplicity, as compared with a discreteclement type of array which requires a large number of driven elements, For instance, a common form of large-Fundamentals of Antennas 2.45, socrags antenna is @ paraboloidal reflector illuminsted by « point-source feed. To ree Ag eTuze ofthe order of 100 wavelengths in diameter by a discreteciemere uray mon fequire more than 5000 individual radiating elements, cachof which mart 'be fed with current of the correct amplitude and phase For se1y large apertures itis apparent that a reflector type of antenna is consid- erably simpler than discrete-element array. In addition, the reflector ean be ace to operat over # wide range of frequency simply by changing the feed, while snace ezete-lement array can operate efciently only over a small bandwidane of 15 to 40 percent at most. Where very low sess na suite some improvement because of the elimination of blocking by tre feed and ts supports, but even for this type of structure special techniques oe we ‘sary t0 achieve sidelobe levels lower than 30 dB down, gtoush the continuous aperture type of antenna has practca imitations with [epand to very low sidelobe performance, it neverthelesis widely used becawe of ot rene fimplicity, For this reason, itis desirable to know what can be copected Fane pees of eal distribution functions since these in effect place an wipes lose dace oeatial Performance ofa continuous antenna. Also, foc very lng’ rays ot here elements, the continuous distribution may be used to obi ee sre Approximation to the element ex Line-Source Distributions Fee Soures, the curent distributions considered tobe function of oly a single Fane ene ectivity patter E(u) resulting from » given dstbution e singe ‘eltted tothe distribution by« five Fourier transform? as given bela nin! | pasete ate where fix) = relative shape of field distribution ver aperture asa function of x W = (re/A) sing 4 = overall length of aperture @ = angle measured from norteal to aperture X = normalized distance along aperture —1 Fue) + 50) 3 e0( aT where the symbols have the same meaning as used eavier inthis section, except for ¢, which is defined as the correlation interval. The effects of random errors on the side- lobes and gain of a circular paraboloid have been computed by Ruze. Some of his results forthe case ofa cosine-quared illumination are shown in Figs. 2-18 and 2-19. i. HW 8 & a [Se fe rt 8 AVERAGE CLOSE-IN SIDE LOBE LEVEL-db 7 BEe ncBtcron heen °° MG. 2418 Sputoussacinten of prado 7 oid with cogine-squared Mumination. Cor- AE/F (ravine) ‘elation Interval o = a8 ‘The relationships for the loss in gain have also been worked out by Ruze, For small errors, simplified formulas have been obtained for the limiting cases of small and large correlation intervals, a follows: G2) -39ee £ GaP wget and Exi-F when > 1 ‘Some results for the loss in gain are given in Fig. 2-20. Periodic Ertors in Aperture lilumination [In addition to random errors introduced in the munufscturing process, certain types of antennas may also introduce periodic errors due o the particular technique used in fabricating the antenna. For instance, in certain types of reflectors using the bulkheadFundamentals of Antonnas 2.98 ——— ° a 02 0R & 2 so ERROR BMS REFLECTOR enor o— “0 — 30, 4g 80 DEGREES FIG. 2-19 Average system pattern tor ciculer paras bolold with cosine-squared iikuminstion. Comation Interval c = » calculated tor reflector 24) in diameter soak eaMork type of construction, mechanical stresses st up during fabrication are Thin ate surface has periodic eror at roughly equal interals along the ofan ‘This results ina periodic phase error along the aperture, sig, ect of a sinusoidal phase error on the sidelobes has been treated in Ret, witch points out that a sinutcidal phase error will produce wo eae Goce Jus amplitude relative to the main-beam amplcude i cqual to one tall ere peak, seca of any type of phase error onthe antenna gun has been treated by Spencer wha shows that the fractional loss in ‘gain is equal to the mean square phase ‘tor rom the least-square plane-wave approsimation fo he phase front Shee pat2-34 Introduction and Fundamentals edie phaser would not alter the dreston of he unperturbed phase ron th Siigin mul toe cea square vas of te poh pas enor ee ate eee se tn ae teee cra mete ae ace oe on aug merecg een napus. Another type of dribton err plied Loses elitr) repelled y is a periodic amplitude error.” The side- wn Schvir thesplode mae sss neon sey ea ott 1 eta ceat te cama pcs ‘petal leeron cher ae oe inane will appent or etch Sis recat oi enenar ome Pe areeereyers toe a use rp, te eters emcee sera haves ve ree heal ee muessus renee eer ane ms of eps amps al to the amplitude ofthe constant term. AS [| atest pecete rine cnor ibe Loss OF GAIN ab periodic amplitude error will produce two eee reaiaeasese aoe TA] Parsee came 5 . tance of mA/é rad, where m is the num- eek eae ease afta emnaeteaaae ee ata co ates el nl ae fp 2:10_METHODS OF SHAPING PATTERNS: ‘A variety of radar system applications require that the radiation pattern of the antenna be shaped to meet certain operational requirements. One common require- ‘ment for the vertical-plane pattern js the cosecant-squared pattern, which produces 2 ‘uniform ground return echo from an airborne radar antenna. There are, of course, other shapes of interest, but the basic principles of obtaining shaped beams are the same for a wide variety of shapes. Methods Usetul for Linear Arrays One of the techniques useful with slot arrays or dipole arrays is to determine what aperture distribution will produce the desired radiation pattern and then design the aay feed system so as (0 achieve the required aperture distribution. ‘There are several methods in use for determining the form of the aperture dis- tribution. Historically, the first method is based on the Fourier approximation to a ‘given function. For an equally spaced array of radiating elements, the radiated field ‘due to the array cam be represented by a finite trigonometric series with a direct rela- tionthip between the current in each element and the coeficients of the trigonometricFundamentals of Antennas 26g saiicnagayeatited tasasion pattern is analyzed by standard Foutier methods, th Seemiicnts ofthe Fourier series then determine the current amplitudes and phocy Tone sesents The accuracy with which the finite series approximates he deg {Raction depends on how many terms are used, that is, on how lage ay apecene Mache ned number of terms, it well known from the theory of Four occ i best in the sense thatthe mean square deviation is lease 1 is also a property of the Fourier approximation that at a point of discontinuity the Face amahghon# value that is the average ofthe value exsting on slr sae ren discontinuity; ie, HLE@ + 0) + £@ — 0) ib the partettod of approximating the desied radiation pattern is one that asbles the pattern to be specified exacuy at a fixed number of points Th method, ukich was independently proposed ty Woodward and by Leviason particularly perern ig n tat it gives the designer some physical insight into the way i ahich ee, Tacoma tized and in addition allows some contol ofthe pattern ay pms discontinuity, isl, the method consists of superposng a series of uniformamplitud, lin frokeced a tabutions across the aperture in such away tat the sum of oe pace Produced by each of the distributions adds up to the desired pattern: Tite spomure distribution js then found by adding up, ia proper phase, the indi {form disee?butions, Toimplity matters, the folowing discussion wil deal with the case of «contin: iscrete-element array fora aminuts from the discusion As ponied out in Sec. 26, the radiate sane, fora uniform in-phase illumination is given ty Fe) = St x where sing aoe batter wll passthrough zero whenever sin 6 = nd/e, n= 1, 2,3, 4 ete. and Ae Pater has its maximum when $ = 0. For a case inwhich = Tox, she pattern Plotted against sin ¢ will appear as in Fig. 2-21 It will be noted that when the pattern is plotted 8, the zeros of the hic ose lly spaced except for the two zeros on either side ofthe sala Sears apes aig Wo Spaces. The width of one spaces equal toh/2, the weap oe Aperture width in wavelengths, sty form illumination with a linear phase variation will have a feld pater fiven by sin(u — 89 (9) = 4 8) 9 write Bis one-half ofthe al phase variation across the aperture &. The maximum ‘his pattern occurs when sin, = Ok bd 22-98 Intoducton and Fundamentals SUBD yeabang 110% Beso FIG. 2-21 Radiation pattern of unity umingted aperturé {OX fong. (8) 8, = 0. (0) By = 8x, If the phase variation is chosen so that 8, = mx, then the pattern wil ‘m/, oF m spaces when plotted against sin ¢. A typical pattern for £ 3e is showa in Fig. 2-216, By combining a number of patterns shifted by integral numbers of spaces, it is possible to synthesize a pattern which can be uniquely specified at 2m + 1 points for fan aperture which is m wavelengths in extent. The total radiation pattermis given by Sle) , 8m) role rey EG) and the corresponding normalized aperture distribution is given by Sia) = SO Ge oe Gein where ~1s.x< 41. ‘As an illustration of this method, let us consider a problem in which i! ie desired to specify that the pattern have a cosecant shape over the range in sin ¢ from 0.1 to 1.0 and that it have no radia- tion for negative values of sin 4. For an aperture length of 10 wavelengths, sia is divided into spacesone-tenth wide from =I to +1 and ordinates erected at each division with a height equal to £(@). For the problem at hand, £(4) is proportional 10 ese & A plot of the £(@) diagram is shown in Fig, 2-22, where at each diy sion an individual pattern is to be speci- — fied with a maximum value equal to cach ordinate. The summation of the individ- AX BIOEOSUe Gs Gs OY OOO Ising FIG, 2-22 (4) pattern for cosecant uel patterns is shown in Fig, 2-23, while the resulting total aperture distribution, ‘consisting of the sum of each of the indi-sine Fundamentals of Antemas 2-87 FIG. 2-23 Synthesis of cosecant pattem for &(0) = 0. sae ae ae te SST SEIT 4 0° : DisTRauFion ooo NER APE rs) FIG. 2.24 Aperture distribution or Fig. 2- 28, (a) Amplitude detributon, (b) Phase Sitibvton, vidual aperture distributions, is shown in Fig. 224. For the case shown, £(0) has been chosen to equal zero, and it will be noted ‘that the resulting pattern has a fair amount of ripple in it. If the value of (0) is chosen to be 0.8, then the result- {ng pattern shown in Fig. 2-25 is consid erably smoother although the aperture distribution is somewhat more peaked, as seen from Fig. 2-26 ‘This particular method of synthesis is quite useful because of the Rexibility in arriving at the inal radiation pattern and the designer's ability co adjust che theo- retical pattern by graphical procedures, Although the example discussed has used ‘nly real values for C, tosinopity the cal- culations, there is nothing in the proce dure which prevents the use of complex Values for Cy. It is quite possible that a itdicious choice of the phase angles for the C's might have given as effective a ‘contr over the ripple as did the chang- ing the value of E(0). One note of caution should be added. Although beams can be added for values of |sin 6) > 1 in order to controtte100 £10) =08 0 sae 3s FIG. 2-25 Synthesis of cosecant pattem for E(0) = 0.8 2:88 lnvoducton and Fundamentals 10 os: os: 5 ighly accurate. Uwually the general trend of impedance characteristics is determined from the seale model but final impedance-matching work can be completed only on the fullsize antensey REFERENCES —ee eeeeeeSeseseseseseF 1 SLA Schellanoll, Electromagnetic Waves, D. Van Nostrand Company, tne, New York, 1983, 2 S.A. Schelanolf and H, T, Fis, Antennas: They and Practice, Soha Wiley & Sons, Inc, New Yark, £98 3 Schelkunoff, Electromagnetic Waves, p. 360. «S.A Sctelunof “Some Equivalence Theorems of Electromagnets and Ther Apples: tion ta Radiation Problems,” Bell Syst Tech J, vol. 18, 1936, pp. 92-112 5 B.S. Canter, “Circuit Relations in Radiating Systems and Appleton to Antenna Prob: ® RC Steneer and P.M; Autn,“Tabies and Methods of Celelation for Line Source.” MIT Rad. Lab. Rep. 7632, March 1946; see uso Rep, 7621, ? 2 Ramsay “Fourit Tranafors in Acial Theory.” Marcon! Rev, vo, 9, 1946, p. 139: ol 10, 1947, pp. 17, 41 81,157,2-42 nvodueton and Fundamentals 10 G.J. van der Maas, “A Simplified Calculation for Dolph-Tehebyackeft Arrays.” J. App. Phys, vol. 28, January 1984, pp 121-124 MT. T.Taylor, “Design of LineSource Aatenaas (or Narrow Beamwidth and Low Side lobes," IRE Trans. Antennas Propagat. vol, AP-3, January 1955, pp. 16-28; ge also RJ. Spellmire, “Tables of Taylor Aperture Distributions,” Hughes Aircraft Co. Tech. Mem. SB1, Culver City, Calif, October 1958 12 T.T. Tayler, “One Parameter Family of Line Sources Producing Modified Sin u/xu Pat- terns," Hughes Aircraft Co, Tech. Mem. 324, Culver City, Calif, September 1953, 43. E Jahnke and F, Emde, Tabler of Function, Dover Publications, Inc, New York, 1943, pm. 14 British Assocarion Mathematical Tables, vols. VI and X, Cambridge Univesity Press, London, 1950 and 1952. “The Effects of Phase Errors on Simple Aperture Iluminatins.” Proc. Conf: Centimetrc Aerials for Marine Navigational Radar, June 15-16, 1950, U. M. Stationery Office, London, 1952. 16 J. Ruze, “Physical Limitations on Antennas,” MIT Res. Lab, Electron. Tech. Rep. 248, ‘Cambridge, Mass, October 1952; see also “The Effect of Aperture Erors on the Antenne Radiation Patter,” Supplementa al Nuovo Cimento, vo. 9, , 3, 1952, pp. 364-380. 17 LL Bail and M. J. Ehrlich, “Factors Affecting the Performance of Linear Arrays,” [RE Proc. vol. 41, February 1983, pp. 235-241. HLF O'Neill and L. L. Balin, "Further Ect of Manufacturing Tolerances on the Per: formance of Linear Shust Slot Arrays.” IRE Trans. Antennae Propagat, vol. AP-4, December 1982, pp. 93-102. 19 N-I Korman, E.B. Herman, and J. R. Ford, “Analysis of Microwave Antenna Sielobes,” RCA Rev. vol, 13, September 1952, pp. 323-33 20. RC. Spencer, "A Least Square Analysis ofthe Effect of Phase Errors on Antenna Gain,” Ar Force Cambridge Res. Cen. Rep. 3025, Bator, Nass, January 1989 21-5, Brown, “The Effet of a Period Variation inthe Field Intensity across « Radiating ‘Apertute." IEE Proc. (London) part SI, val. 97, November 1950, pp. 419424. 221, Wolf, “Determination ofthe Radiating System Which Will Produce @ Specified Direc- tional Characteristic," IRE Proc. vl. 25, May 1937, pp. 630-683. 23. PLM. Woodward, "A Method of Calculating the Field overa Plane Aperture Required to Produce a Given Polar Diagram,” JEB J. {Zondon), part {1 (4, vol. 93, 1947, pp. 1554— 1558, 24 TT. Taylor, 36, September 1948, p. 1135. 25 L.'S. Chu, “Physica! Limitations of Omnidirectional Antennas.” J. App. Phys. vol. 19, December 1948, p, 1163. 26 P.M. Woodward and J.D. Lawson, “The Theoretical Precision with Which an Arbitrary Radiation Pater May Be Obtained from a Sourey of Finite Size," JEE J. (London), part val. 98, September 1948, pp. 263-370, 21 N. Yara, “A Note on Super-Grin Arrays,” IRE Proc. vol. 39, September 1951, pp. \O81— ss 28 A. Block, R.G. Medhure, and S.D. Pool, “A New Approach othe Design of Superdirec- tive Aesial Array,” IEE J (London), pare 11, ro. 100, September 1953, pp. 303-314, 29 G Sinclar, “Theory of Models of Electromagnetic Systems." IRE Proc. vol 36, November 1948, pp. 1364-1370, 30, G. Sinclar, E-C. Jordan, and E. W. Vaughan, “Measurementof Aircraft Antenna Patterns ‘Uting Models," IRE Proc, vol. 38, December 1947, pp. 1451-1462. 1sChapter 3 Arrays of Discrete Elements * Mark T. Ma National Institute of Standards and Technology 31 Introduction 3.2 32 Uniform Linear Arrays 3.2 Broadside Arrays of Isotropic Elements 3.3 Broadside Arrays with Practice Antennas o¢ Blemetis 37 Ordinary End-Fize Arrays of Isotropic Elements 3.10 Ordinary End Fite Arrays of Practical Antennas as Elements 3-12 33 Yaghtida Arcays 313 34 Improved End-Fire Arrays 315 38. Linear Array with Tapered Exctations—Equal Sidelobes 3-17 ‘A New Approach for Treating Optimum Broadside and End-Fire Arrays 3.19 Opcimam Broadside Arrays When d= 2 324 Optimum End-Fire Arrays When d Is Les Than the Maximum Allowable Spacing 3-26 36 Ary of Arrays 329 snsbution ofthe Neonat of Standards and Technology wo sje eo copyright1_ INTRODUCTION A practical objective of directive communication i an improvement in recelved signal 4s measured relative tothe prevailing noise. More precisely, directivity improves the Signalso-ase ratio, This inprovement may 6e accomplished either atthe transmitter side by using an antenna that projects the transmitted wave power in the form of a concentrated beam toward the distant recsiver, or it may be accomplished at the receiver side by using similar antenna to intercept a maximum wave power. Thus, any desired signal improvements may be taken either atthe transmitter or at the receiver or be shared between tbe 20. ‘One of the common methods of obtaining directive antenna characteristics is an arrangement of several individual antennas of the same kind so spaced and phased that their individual contributions add in one preferred direction while they cancel in others. Such an arrangement iskaown as an array ofdscreteantennas, The individual antennas, called elements, in an atray may be arranged in varicus configurations such 45 straight fines, squares, ectanges, circles, elipes, aes, oF other more innov geometries. Years of experience have reduced the number of configurations thet are considered practical toa relatively few. Tmprovements in radiation characteristics by arraying may be described in a variety of ways. Sometimes the result takes the form ofa plotted pattern which shows, at a slance, he relative signa level in various directions. In other cases the result measured in terms of directivity or power gain. For purposes of this chapter direct ities or power gains are generally referenced to some accepted standard such as an isotropic source or an array element, ne of the practical arays consists of a number of individual antennas set up along straight line known as the linear array. For this particular configuration, the individual elements may be excited by currents in phase with equal or tapered ampli tudes. Under tis condition, radiation proceeds both forward and backward in direc- tions perpendicular to the lines of the array. Such an aray is said to be broadside. Tt may be made essentially unidirectional by adding a second identical array tothe rear ofthe fist and exciting the second array with appropriate phase difference. The elements ofa linear array may also be driven by currents with phase pro- sressively varying along the array axis in such a way as to make the radiation sub- stantially unidirectional, This array is referred tos an endsre array. Both the broad- Side and the end-ire arrays of different elements are presented in Secs, 3-2 through 35, Necessary formulas without detailed derivations are given together with che frst hand design curves or tables with the number of elements, inteelement spacings, amplitudes, pases, and other array characteristics asthe parameters. A special array ‘of arrays known as the curtaln antenna is briefly discussed in See. 3-6 ‘When the general principle of arrays with isotropic elements is presented, no consideration of the mutual-couping problem is needed, Fo the case involving simple practical amennas suchas the short Yagi-Uda array (a small numberof dipoles) for ‘hich methods of computing the mutual impedances between elemeats are evailabe, the mutual impedances wil be duly included, 3-2 UNIFORM LINEAR ARRAYS: When the identical elements of a linear array are equally spaced along the array axis and excited with uniformly progressive phases as shown in Fig. 3-1, the array patternArrays of Discrete Elements 948 in fre space may be considered, assuming thatthe principle of pattern multiplication applies," a8 a product of the element pattern and an array facto, EGS) = 108)SO) (3-1) ‘here £(0.) is the far-eld pattern exprested in terms of the ordinary spherical coor- inates (6,8), 4,6) is the element pattern determined by the particular antenna ele, ‘ment used in the array, and (6) is the array factor for a linear array placed ‘along the 2 axis with @ measured from the array axis, The array factor S( may be TO pstaNT PONT a 201 explicitly written as BE nner aay of acuaty oy E hex cin) (2) where J; represents the current amplitude excitation of the ith element and = kal{cos 6 ~ cos 6) (33) rit being the wave number (2/4, \ being wavelength, d the interelement spacing in wavelength, dy the desired direction of the maximum radiation, and » the nent ‘umber of elements in the array. When Z.= 16 0,1,.+-, (m— 1), uniform linear array results, yielding - = La exp inu) oa & 8 (IM) = cep Ua) Cay where the subscript u should not be confused withthe symbol defined in Eq, (3-3), Broadside Arrays of isotropic Elements wamfilition, the elements are isotropic andthe desired maximum radiation sin the Sas tide Stection, we have 6) = 1,65 = 2/2, u = kd os 8, and BO) oe ‘SU®). The basic characteristics ofthis array may be summarized as flew, F2nT Maximum Tt is clear that S.(@)au, = 1 (the total umber of elements) when @=s/loru~ 0, Nulla If we express Eq, (3:4) in a different form such as 1 = cos mu _ sin? (nu/2) T= cos usin? (u/2) dad) ‘tis easy to see that the null positions (S, = 0) are given by iy /2 = re (36a) [Sel = 8,51 008 lm = miN/ndy m= $1, 42,..., MS nd/X (346) aang fis» Ou are the angular nll postions on one side ofthe beam maximum in the forward direction and 8-4, 6 8. are null positions on the other side af8-4 Introduction and Fundamentals the beam maximum, also in the forward direction, Of course, #4, are symmetrically located with respect to the beam maximum. Its clear from Eq. (3-62) that u = tx are always possible null positions when n is even and d = ¥. The exact aumber and locations of null positions as given by Eq. (3-58) depend on m and d/d (or approxi- mately the overelt array length). A designer may have choices of different arrays to produce a null in a particular direction in order to minimize possible interferences at this direction. For example, the arrays of four elements with d = 32/8 or 34/4, of six clements with d = /4,/2.0r 30/4, etc., may be used to satisfy the purpose of having ‘a null at @ = 48.19° (or 41.81" from the main beam). The final choice among these possibilities should be made fromt other design or cast conéiderations. First-Null Beamwidth One way of showing the directive pattern of a uniform ‘broadside array is by the fist-null beanrwidth, which is defined as the angular space ‘between the first nulls on each side of the main beamn, (W), = 01-85 2(f 5 ) nd (3-74) (BW), = 200" — 4) (3-76) where 6, in Eq. (3-72) isin radians while that in Eq. (3-70) is in degrees. For very large arrays, we approximately have in radians in degrees (38) Halt-Power Beamwidth Another more commonly used beamwidth isthe angular space between che half-power points on each side of the main beam. The half-power point uy may be determined by solving up|? = Smad) ISN? = So Gayay 72/2 (3-9) which in tra, yields the corresponding 6 in accordance with 08 04 = uy/kd (3-10) ‘The half-power beamwidth then becomes ew.- a(f-a) in radians 2190" — 4) in degrees (3-11) ‘Sample results for (BW), are presented in Fig, 3-2 Sldelobee Precis locations ofthe sidelobes ofthe uniform linear broadside arrays may be found by setting the derivative of Eq (38) to 2er0, yielding? mtan (u/2) = tan (m/2) (3-12) tis clear that u = 1 are possible sidelobe positions when n is odd and d > A/2. Other solutions of Ea. (3-22) for w in degrees are presented in Table 3-1. The corre-Arrays of Discrete Elements 305 WA 88a FIG. 3-2 tHaltpower beamwicths for uniform broadside errays of isotropic laments. Sponding angular sidelobe postions in terms of can be determined by u = kd cos @ ‘ho the clement spacings known, The level ofthe ith sidelobe atu relative tothe beam maximum may be com- Buted as |Ssu)|/ S40) = |S,(u)]/n. The sidelobe levels ik decibel ie aes iluded in Table 41 Noe from Tabie 3-1 that the sidelobe positions approach u,=-(2/+ t/a fad or (21+ 1)180"/n, i= 1, 2, 3, (or halfway between nulls) and that the sidelobe levels approach |S(u)|/n = 2/(2i + Un for arrays with a very large num- brrof dements. In particular, the first sidelobe level of large uniform arrays is Ser or “1346 dB Pirwetivity |The directive gain of an array ina given direction is defined asthe ratio oF the radiation intensity in chat direction to the average intensity, = LEG)? | G6) = wer 471 £@.9)17/Mo (3-13) ‘here WH cepresentng the total power sadisted by the array is given by w= JP" J. 1001 sino ao do (344) ‘The diectivty ofthe array is the maximum value ofthe directive aun, Dm COD au, = 41E09)|2ax/ Ho (3-18)aim AA, Tore ware eo NTI OS LI wise SEI ‘lors er wINE o evel 0 O8IF ‘ 8 L 9 5 ’ ap (7n) 6 ap (fn) ep n) gp O yene1 ‘uonredd §=— “10X07 UoRIBOd eA ‘UoNIeed “1007 yunos aL puocses “Sjuouerg ooncsy jo SKeuiy oprepeorg feour] WuOHUN 20} S]ANT pus BUOHLOTT GOES bE TTAVLArrays of Discrete Elements 3-7 For the uniform linear broadside array of isotropic elements considered here, we have 0) = 1, |EO)I? = |S,P?, w= kd cos 0, and | £(0,9) dus =m. The directivity ‘becomes p=» {to (3-16) (nkd) +25 AM sin (mid) Note that when kd = pr ord = ph/2, p= 1, 2... Dy = n.Wimplies that under this condition the directivity is numerically equal to the total number of elements in the array. For other values of ka, the results are presented in Table 3-2. For a given TABLE 3-2 _Directivties for Uniform Linear Broadside Arrays of Isotropic Elements ax a % % %* % % % % 1 4 128 216311 400484558529 4.00 5 Kas 270383 $0062 5.00 ‘ 1s) 317 463-600-730 6.00 7 193 366 534700 7.00 3 220 416 610800984 8.00 9 248 468686 © 9.001107 9.00 10 274 517739 1000 1241 10.00 1 the directivity is almost linearly proportional to d/) in the range of % =< d/A = and then drops to m when d = A where the grating lobe at the same level as that of the main beam appears. For this reason, the element spacing for a fixed-beam broad- side array is normally kept at less than a full wavelength. For arrays designed with ‘beamscanning capabilities the element spacing is even smaller. However, the element spacing should not be chosen to be too small for the obvious reasons of obiaining a respectable directivity and minimizing the mutual coupling among the array elements. Thus, the actual value for the element spacing is usually a compromise of the above considerations. Broadside Arrays with Practical Antennas as Elements Broadside arrays with practical antennas may be achieved by placing the antenna clements in such a way that the maximum radiation of the element coincides with the desired broadside direction of the array. One example is to place the half-wave or short dipoles colinearly along the array axis as in the case for the wellknown Marconi- Franklin antenna.* There are, of course, many other arrangements that satisfy the ‘equirement. We consider only the specific example mentioned above. Arrays of Half-Wave or Short Dipoles The particular arrangement is shown in Fig. 33. Although the current distribution on two or more antennas in an array envi- Tonment is not the same as when the elements are isolated because of the influence of348 troduction and Fundamentals ng FIG. 3-3 An equally spaced linear array Far collinear dipoles. ‘mutual couplings, the mutual impedance of collinear dipoles is, however, relatively ‘mali, ard therefore the change in current distribution is not large. This is especially true when the element digole length is near or less than A/2.* Under the assumption ‘of no mutual coupling, the element pattern may be written as** 0s (kh cos 8) — 0s ch sin ‘where fis the half teugth of the individual dipole in wavelength. "For half wave dipoles, h= 4/4, kh x/2, the element pattern reduces to FO9) = (3-17) 2 $08) = Sa “The maximum radiation of Eq. (3-18) occurs at @ = x/2, thus satisfying the require- ‘ment, The overall array pattern is then given by m=) (3-18) cos (Zea £00) = Sag IS (3-19) where Seis the array factor given in Ea. 04). ‘The mull positions remain the same as those determined by Eq. (3-6) except for theinurodution of addtional us at 0 = Oand oving othe element pattern. This ae eal Beant alo remnin unchanged. Locations an levels of the Sit few ‘titer do pot change much ether. The half-pover beamwidth becomes naturally serene smaller then that with isotropic elements The directivity is somewhat fitter vecause the aiectivity for a ingle half-wave dipole is proximately 1.64 rather than 1.00 forthe isotopic case ‘Far very short dipoles, h =< /16, kh < /8, the element pattern may be shown approximately to be proportional to JOA) = sine (3-20) ‘Thus, the overall array pattern is proportional to £66) = sin 6 |Sul (321) ‘which, again, is not much different from Sy. The directivity in this ease can be expressed as (3-22) where We (3-23)Arraysot Discrete Elements 3-0 ‘vis clear that the directvites for d = A/2 and d = 2 are no longer equal to each other. Furthermore, the grating lobe at @ = Ofer d = kd = 2x) owing ta the array factor S, is now eliminated by the null of the element pattern. The directivity for d = Ais therefore expected to be higher than that for d = A/2 in this case, Represen- tative results for D, with n = 4, 5, and 6 are presented in Table 3 ‘Note that the directvites for various cases given in Table 3.3 are numbers, which can be converted into decibels with respect to the isotropic element (4Bi) by taking 10 log D,. The directivity can also be expressed relative to that of a single clement. Since the directivity ofa single short dipole is 1.5, the directivity of arrays ‘with short dipoles relative to a single short dipole in decibels is then given by Dy = 10 log (D,/1.3). Dicectivties for arrays with a larger aumber of shart dipoles in this latter unit are given in Fig, 3-4 TABLE 3:3 _Directivities for Uniform Broadside Arrays of Collinear Short Dipoles a a MS o% * % % % 1 4 245339 429 sm gS wa 695 3 294 405 530645733859 8.86 6 3a 487 629781951037 1077 NORMALIZE0 0, 48 Tri oor a2 as OF OF 88 OS 10 an FIO. 3-4 Oven fo broadside aay of collnger short dipole (lave to @ singe leet).3-10 ttodvetion and Fundamentals Ordinary End-Fire Arrays of Isotropic Elements When the condition 85 = 0 is imposed upon Eq. (3-3), an ordinary end-re array results. The visible range for u = kd (cos 8 — 1) corresponding to 0 =< @ = 180° is then —2kd = u << 0. The basic characteristics forthe uniform (J, = 1) ordinary end- fire array of isotropic elements (/(@) = 1] are again summarized first and then are ‘compared with those for the broadside array presented in the subsection “Broadside Arrays of isotropic Elements.” Samples of end-fire arrays with properly phased prac- tical elements or parasitic elements (short Yagi-Uda arays of small number of dipoles) are discussed in the following subsection, Beam Maximum Since the array factor in terms of u remains unchanged, the beam ‘maximum is still numerically equal to n, which occurs at u = O or @ = 0. Nulla The expression (Eq. (3-6a)} for determining the null positions sul! applies provided that m = —{, ~2,.... The exact mumber of nulls depends on m and d/A. In terms of 6, the null positions are given by (3-24) m et mm a as = 1 + Firat-Null Beamwidth The first-null beamwidth for the ordinary end-fre array is defined as (BW), = 28.) = 2e0r(1 - 3) (3-25) For very large arrays, 9_, is small, yielding x (BW), =2 fal rad (3-26) By comparing Eqs. (+26) and (38), we conclude thatthe first! beamwidth for te ordinary end-fire array is always wider than that forthe broadside array with the same n and d/>. Halt-Power Beamwidth Equation (3-9) is still valid for determining the half- ower point 1, (taking the smallest negative solution), which in turn gives cost 1 (3-27) ‘The half-power beamwidth thus becomes BW), = 2 aeor'(1 + #4) (3-28) ‘Sample results for (BW), for the uniform ordinary endfire array are given in Fig. 35, Comparing Fig. 35 with Fig, 3-2 also reveals that when the beam position ‘changes from the broadside direction (Q, = 90°) to the end-fie direction (tp = 0), the half-power beamwidth becomes brosder for an array with the same m and d/XArrays of Dscrete Elements 3:44 1189, 10] Ht power beam * a8 6 sj ° 10 « a i FIG. 35 Haltpower beamidths for unitorm ‘ordinary enc arrays of iotropicslemenne, {idelobee Equation (3.12) and the fist few sidelobes (postion and levels) given the tle 31 fr the uniform broadhide aray are also good for tena array, with, 1p Andestantng she ony tke negative solutions for ar a Reve Ga Eg, (3 jaja Fable 31, The conesponding sidelobe potions in tema a ee ee obtained ont = (Cos — 1), The element spacing d's norma sere oe N72 for ‘S cndire array because the grating lobe will appear with the paca aint Directivity The di expressed as? vity for the uniform ordinary end-ire array may be (edynt a (3-28) nikd + 5 =P comedy312 FIG, 3-8 Hall-power beamwidine for a uniform Erray ot 10 isotropic elements as a function of bear ‘maximum position. [Note thatthe second term in the denominator of Eq. (3°29) vanishes when kd = P= Jhued = ph/dcp = 1.2... Under this condition, Dy =m Fr ther values of My Leo eats of D, can be easily obtained from D, given in Table 32. For the same 1. ‘Detor a value of dis equal to Dy for 2d. For example, for n = 4, Dd & NS) =e (2S Aya), whichis 2-46 ia accordance with Table 3-2. This fact is obvious if Fas (G-16) and (3-29) are carefully examined. End-Fire Arrays of Practical 18 Elements Endre arrays of practical elements may be realized by proper arrangements of the evens elements. One example which willbe discussed here isto place short dipoles parallel to the x axis, forming an array * long the axis as shown in Fig, 3-7. ‘Another method is to excite only one antenna element and leave the other ele- zerdtie) mots unexcited (parasitic) with appro- | a priste antenna lengths and spacings to produce approximate phase disiibutions 10.2.7 An equ spaced near aray ot Prong te elements Tequired for end-fire Parallel short dipoles. fadiation, This latter arrangement is known as the Yagi-Uda array. Since the current amplitudes on the lements of 2 Yagi-Uda array are no longer the same {Ih * 1), the array cannot, strictly speaking, be considered as a uniform ‘ray, For this reason, characteristics of short Yagi-Uda arrays (with a sma) ronioes Gf dipotes) are presented ina separate section. Long Yagis treated from the viewpoint ‘of surface-wave antennas are given in Chap. 12‘nays of Discrete Elements 3-13 For the array of parallel short dipoles as shown in Fig. 3-7, the general element pattern is given by /(0,8) = (i — sint 8 cov g), The element pattern in the xz plane (6 = 0) reduces 10 (8) = cos 8, which does have the maximum radiation at the end-fre direction (0 = 0). When the elements are excited with equal amplitude land progressive phases, an end-fre array results. The overall array pattern then, neglecting the mutual-coupling effect, becomes AB) = con 8 (Sku (3:30) where u = kd (cos 0 ~ 1). Clearly, the element pattern produces @ null at @ = 1/2 which may or may not bbe a mull for S,. Since Emuy = £(0) = n, we obtain, using Eq, (3-15), the directivity (3:31) where (3-32) Representative results for D, are presented in Table 3-4. Directvitics relative to that of a single element, which are shown in Fig. 38, may be compared with those in Fig. TABLE 3-4 _Directivities for Uniform Ordinary End-Fire Arrays of Parallel Short Dipoles. ayn a % % % % 4 3m S21 689 at 5 33 626 Ba 574 6 4733999677 3-3 YAGI-UDA ARRAYS ‘The typical Yagi-Uda array consists of many parallel digoles with different lengths tnd spacings as shown in Fig. 39. Only one of the dipoles is driven. All the other clements are parasitic and may function respectively as a reflector oF asa directr. ‘Arrays of this kind were frst described in Japanese by S, Uda’ and subsequently in English by H. Yagi In general, the longest element, of the order \/? in length, is the feflesor. The director elements are always shorter than the driven element in length. ‘though there may, in principle, be many reflector, experience shows that lie is ‘sained by having more than one reflector. The reflector is usually spaced 4/4 to the3-14 lnroduction and Fundamentals rear of the driven element, Considerable gain cen be realized by adding numerous directors. The simplest Yagi-Uda has only one director, making a three-element array. Since the element patterns for each dipole are diferent, the principle of pattern multiplication certainly does not apply. The element pattern of the ith dipole in the xz plane is given by” Sip) = Se si9 = es By (3-33) here is he Bung o the ith dipole “The overall array patern may be writen & 20 = SHU exo (3:04) where 1 is the total number of dipoles in the array, do = 0, and 1,8 the maximum ‘current amplitude of the ith dipole, which may be determined by the available method? ‘The power gain may be computed by G(0xp) = 6016(0)|*/ Pa, where Py = H{fal?Ry represents the input power. Ry is the input resistance, and [ais the base ‘current of the driven element. As an exezaple, the normalized pattern with m = 3, hy = 0.26, fz = 0.252, hy = 0.232, dy = 0.25K, di = 0.45 and dipole radius of 0.001 mis shown in Fig. 3-10 to demonstrate that the Yagi-Uda is indeed designed as an cend-fire array (beam maximum at @ = 0), For this example, the normalized base currents io amperes, with the assumption of a three-term cutrent distribution? are Iu] = 0.0038334 exp (/ 69.86") In] Vz = 014546 exp (= 43.16") 38 Ia/V_ = 0.009361 exp (J 166.97") 2 We see here that J has an approximece phase lag of 113° relative to J, although the phase difference should ideally be — ‘kd, = ~90". Similarly, the phase differ- fo ence between Jy ard Jy is approximately =263" rather than —kd; = 162". ‘Nevertheless, it is because of this kind of phase distribution approximately in the ‘ight direction over the Yagi-Uda cle- ments that the end-fire pattern is pro- ea ae FIG, 9-8 Dhrectvites for ordinary end-tre arrays of parallel short dipoles (relative toa ‘single element). recta Pea 4 FIG. 3-9 A pice! Yaghie aayArrays of Discrete Elements 318 ee aye ~ 1 Saeee \AAGISSS2- FIG. 3-10 Horizontal patter fora three-element Yagi-Uda array with hy = 0.26) ha = 0.26), = 0.231, d, = O.25X. dy = OLABN, and clpte ese = 0.001 m, ‘uced. The input resistance for this exemple is approximately Ra = 50.1 0, and the Power gain is about G = $.01 or 7.0 dBi, 3 atotber example, for n = 4, hy = O26, fy = 0.25%, hy = he = 023%, dh Be ndtge O2m, ds = 0.88, and dipole radi = 0.001 m, we obiie Be 82.80, sad G = 5.92 of 7.72 dBi. ‘An analytical method, ax well x perturbational procedures, is available for maxe imizing the Yau ida gun by adjusting the dipole lengths and spacings" A ears in the order of 10 dBi can be cay designed with only @ moderate numberof eign the array. Yagi-Uda arrays with lange number of dipoles are wealad in Oban as surface-wave antennas, 9-4_ IMPROVED END-FIRE ARRAYS eeennays Tepgcrdinay end-tire array with u = kd (cos 0 — 1) has been presented in See, 32 {thas been known that an increase in directivity may be realised by increasing the Oe gene Piase Jag beyond that foram ordinary end fire by a phase quantign bor ate SPesitically it was concluded that a maximum dircetwits when may be ‘alized is given approximately by Bq = 1nd (3-35) when (3-36) For the purpose of clarity, we have designated the one considered in the subnar-3618 Intocwetion ane Fundament tion “Ordinary End-Fire Arrays of Isotropic Elements” as the ordinary end-ire and cal the case being presented here as the improved endhfire array. ‘The array factor in terms of u forthe improved uniform (/, = 1) end-fire array remains the same as that in Eq. (3-5). That is, 2 Sin® (mu/2) Ot des rerrny (3:37) provided that w= kd (os —1) 3 (3-38) whieh is herein called the improved endsie condition rom Eq. (3-38, it can be seen thatthe location of the beam maximum has been shifted to u = 3, which corresponds to @ = 0, The beam-maximum strength thus becomes sin (1/2) = 28/2) 3 (Sdeu = FG (399) hich is smaller than 1 for the ordinary end-re case. Naturaly, for a given n,§ cannot be too large in order to have a respectable (S,max [Nall and sidelobe postions are similarly shifted. More specifically the frstnull position is determined by 1 4d te oa(1- 448) (3-40) ‘The half-power point uy can be obtained from sin®(ran/2) _ si (nb/2) sin? (ul) 2ain® 72) an whic, in tun, nify 4 mt) t= cor(1 4 8548) (3-42) Again the selection ofthe smatiest negative solution from Eq, (341) is understood. Since > 0, both the frst-aull and half-power beamwidths, (BW), = 28, and (BW), = 2p, obtained respectively from Eqs (3-40) and (3-42), should be smaller than those fr the ordinary end fir, where 5 = 0. Numerical results for (BW), are showa in Fg. 31 “Although the sidelobe levels relative to the main beam ae increased slighly, the overall directivity is always increased besause ofthe improvement in beamwidth. For the improved uniform end-te array of isotropic elements, te diectvity may be com- puted 28 follows: sin? (3/2) eg C0812) >= —— OR ______ (aaa) kd + 2 5 "=" inka) cos mk + m8) ‘Sample results for D, with atypical element spacing of d = 2/4 are presented in Table ‘55, where the optimum 3 required to produce the maximum directivity can be easilyArrays of Discrete Elements 3-47 102036 a5 ay FIG, 3-11 Halt power beamwidth for improved end {iro araye with o = X/4, TABLE 3-5 _Directvities for Uniform improved End-Fire Arra Cllsotropic Elements with kd = x/2 (5 in Degrees) ° 3 345 3.90 429 474503 ae 4492 589 62 704 G30 S637 772 Bet g37 S57 6 810 1005 1036 729 crags For larger numbers of clement, the drctivties relative to those for the rae race ft aay are given in Fig. 3:12. The coresponding phases sete ny penser Woodard condition (Eq. (336)] are ale indlated ne deere Person purposes, 25 LINEAR ARRAYS WITH TAPERED EXCITATIONS —EQUAL SIDELOBES tar it lbs level eatzd from ner ares with constant ampli enc esses 2/(3x) or ~13.46 dB when mia very large, as indicared ate copie level onsidered too high fr apts aleee ee ‘imple or apered excitations are normaly sd oredusathe nese aceept-3-18 Inodyction and Fundamentals Noroazed D, 38 736 46 2 30 FIG. 3-12 Normalized dlrectvties (relative to the ordinary end fre) for uniform Improved end:fre arrays of iotropic eloments with d= 2/4. able level. One method for achieving this objective is the well-known Dolph-Che- byshev array," in which all the sidelobes are set at the same level. This array is con- sidered optimum in the sense that che frst-null beamwidth is minimum for a specified sidelobe fevel or that the sidelobe level is minimum for a specified first-null beam width, However, the above statement is true for broadside arrays only with an element spacing no less than one-half wavelength or for ordinary end-ire arrays in which the element spacing is no less than one-quarter wavelengtb."? Many others have also devised methods for optimizing the sidelobe and first-null-beamwidth relationship for speciat cases."*" All these methods are essentially based on manipulation of Che- Dbyshev polynomials. ‘ANew Approach for Treating Optimum Broadside and End-Fire Arrays ‘An alternative unified approach is to deal with the power pattern in the form of a ‘polynomial of (n — 1)th degree with real coefficients,®"* where m isthe total number of elements in the array. This approach has advantages that it not oniy applies to the array yielding equal sidelobes with 2 specialized tapered amplitude excitation but also valid for other linear arrays including the uniform array considered in Sec. 3-2. Basically, the power pattern for a linear array with equally spaced isotropic elements may be expressed as ry) = E awn (3-44) where Y= 2eosu w= kd (cos 9 — cos 4) (3-45)‘Arrays of Discrete Elements 3-19 and the coefficients 4 are real. When, in addition, the array has a symmetecal ampli tude excitation, a8 isthe casein practice, the power patter takes the following prod: oun Pi) = TT + bn)? forodd n (3-46) “ eat FO) = +2) TT 0+ 5,)? — forevenn (3-47) ‘where all b's are real, Note that both P4()) and Py) are nonnegative in the visible range —2 =< y < 2, as representing the power pattern they should be. In addition, if the mth null is visible, we require (64 <2. Ifa the nulls donot coincide, b's must also be distinct, Thus, there are at most (n ~ 1)/2 independent real nulls for odd n and m/2 indepen. b> > b> >> ene > bena> 2 for sian (361) and 2>-b >> b> A>. > ban > Yorn > 2 forevenn (3-62) 1kis convenient to express, for a given n, by (mt 2) and y (all) in terms of by first For the casein which K*is specified, this reduces the entire problem to ony one equa, tion, Pu(2)/P—2) = ? ot P2)/ Pd.) = K2 representing the desired sidelobe Jevel, from which 6, may be solved numerically. Ths should not be « dificalt task since y = —by represents the fist null. For avery large mb i clone to 2. On the ther hand, forthe case in which a desired fist nul (by) i specified, all the other ars and ys can be easily determined from Ea. (3-59) or Eq. (3-60, and the resultent sidelobe level can be obtained by computing 4(2)/P—2) or PA2)/ Py). 1k is insirctive €o demonstrate in Fig. 3-13 how 6, varies with K* and n. In {eneral, fora piven 1, , always increases with X?..For one extreme situation when the sidelobe evel is sot 10 be minus infinity in decibels or equivalently whee X2 approaches infinity, by and, in fact, al the other by’ wil approach 2 as their limits, Under this condition, the power pattern Py) will approach (y + 2)" The required amplitude excitations, according to Eq. (3-48) or Eq. (3-49), will be the expansion Ceefciens of (1 + 2)*F, which is the socalled binomial array.” Thus, the arays of {uel sidelobes discussed herein include the binomial aray as a special casein which All the sidelobes are equal to —co decibels wth respect tothe beam maximum, Cer-3:22 Invodction and Fundamentals ey sc eT a CT) by FIG. 2-13 First-eull postions and sidelobeevel relationships for arrays with ‘equal sidelobes, tainly, the accomplishment of a very low sidelobe levels at the expense of wide beam- ‘width and highly tapered amplitude excitations. The other extreme situation occurs ‘when the sidelobe levels set ata very high evel or when K? approaches unity. Under this condition,the sidelobes are at the same level as that for the main beam. The value for by may be determined from Fig. 313, and the other by's and y/s can be obtained from Eq. (3-59) or Eq. (3-60). The requited amplitude excitations computed from Eq, 6-48) or Eq. 3-49) will vanish except fr the first and last clements, exhibiting a so- called edge distribution? For the practical case, the sidelobes, beamwidths, and required amplitude excitations are somewhere between these two extremes. ‘The first-null position in # can be determined easily from b; when the element spacing d and the beam:maximam direction fy are known. That ‘ad{cos 8, — £08 86) = cos! (—by/2) (3-483) For the broadside array (Go = 90°) with d = 2/2 and the ordinary endhire array (Go = 0) with d = 2/4, the fist-null and sidelobelevel relationship shown in Fig. 3- 13 is known as the optimum in the sease defined by Dolph." Of cours, the clement spacing cannot be allowed tobe too large in order to avoid the appearance ofa grating lobe. For singlerequency operation, the maximum allowable element spacings for the ‘optimum broadside array (., equal sidelobes) are given in Table 3-6, The maximum allowable element spacings for the corresponding ordinary end-fre arrays of equal sidelobes are just one-half of those listed in Table 3-6. ‘The amplitude excitations required to produce the optimum array characteristics, sown in Fig. 3-13 for n = 3 through 8, with the condition on the element spacing presented in Table 3-7, The results are valid for ether broadside or ordi-TABLE 3-6 | Maximum Allowable Element Spacing in Wavelengths for an Optimum Broadside Array with Equal Sidelobes, 0.7438 0.5796 08179 0.6566 8601 one asta 08216 TABLE 3-7 _ Amplitude Excitations for Optimum Arrays with Equal Sidelobes K a 1 0 61 to ro 1 1 1 3 0 103916361877 1.960 Tea 1 1 ' i * 1 1 1 4 0 08791736 2a 2.669 symmetrical 14 1 1 1 5 0 0724 16092413, 3.013 9 0790 1932 340 ata symmetrical rn 1 1 1 ‘ 0 0608 1.437 2312 3.087 0681 1.850 33834975 symmetrical at 1 1 1 0 05191277 aast 3.008 1 9 0586 688 3.308 5.269 © 0610 18393788275 symmetrical roa 1 1 1 © 0482 La39 978.863 5 © ost0 11509 30975200 0 oss 1753818848 symmetrical —_—3-24 Introduction and Fundamentals “The directivity fr the array with eqatl sidelobes may also be computed in accor- dance with Eq, (3-15). In terms of y, we have 202) — = 2kd PQ) (3-64) [70 snes where W= (3-88) and yy = 2eos [hala + eae ye = 2606 [Aatt — cos 9] (3-66) For the broadside aray, & = 90°, yg = 2» Ea. (3468) should be replaced by > _ PO), wat 4 3-67 Vea” eee where ye Dees kd (3-68) “The directivity for the optimum broadside array with various sidelobe levels has been see ly cuzapated® Sample results for d = A/2 ate presented ia Fig. 314, from eee ay conclude (1) that, for a specified sidelobe level, the directivity always voice yew tye ttal number of elements inthe array and (2) that, for a siven inerairecivty does not always increase when the sidelobe level i decreased. The te te shown in Fig. 314 are, of ooUr3, also valid forthe ordinary end-e array with taal sidelobes and with d = 2/4 ‘Optimum Broadside Arrays When d < 2/2 ‘twas noted previously that the frst-nufk and sidelobe-level relationship shown in Fi. 3:33 is optimum in Dolph’ sene fr the broadside array of equa sidelobes only when 20 a) 0 Sidalade tee FIG. 3-14 Drectiities fac optimum broadside arrays of equal sidelobes when d= 2/2.Arrays of Discrete Elorents 9-26 the element spacing is no less than one-half wavelength. When d Sample results are given in Fig. 3-19 to show a designer how to choose an afray apertre to realize desired level of directivity ‘Two-dimensional phased arrays of caely spaced waveguides with thin walls in which the mutual coupling cannot be neglected are discussed in Chap. 20 and else- vwhere.™ Other two-dimensional arrays taking the form of a cicle,concensrc cies, or an ellipse are also possible. Details of these arrays may be found from the references.) REFERENCES 1 B.C, Jordan, Blecrromagneic Waves and Radiating Systems, Prentice Hall tc, Engle- wood Cliffs, NJ, 1950, 2 J.D. Kraus, Antennas, McGraw-Hill Book Company, New York, 1950. 3M. T. Ma, Theory and Application of Antenna Arrays, Jobn Wiley & Sons In., New York, 1974 4 RW, P. King, The Theory of Linear Antennas, Harvard University Press, Cambri Mass, 1956. 5H. Bach, “Direcvity of Basic Linear Arrays.” IEBE Tran Antennas Propagat, vol AP- fanuary 1970, pp. 107-110. 6 R.C.Hamsen, “Array Theory and Practice,” Microwave Scanning Antennas, vol. Yl, Kea: demic Press, In, New York, 1966, "Wireless Beam of Short Electric Waves," J IEE (Japan), no. 452, March 1926, 1p. 273-282; no. 472, November 1927, pp. 1209-1219 9, "Beam Transmission of Ultra Short Waves," IRE Proc. vl. 16, June 1928, pp. 9 RW. P. King and 7. T. Wu, “Currents, Charges, and Near Fields of Cylindrical Anten as," Radio Set, vol. 69D, March 1965, pp. 429-446, 10 C. A. Chen and D. K. Cheng, “Optimum Element Lengths for Yagi-Uda Arays.” IEEE Trane. Antennas Propagat, vol. AP-23, January 1973, pp. 8-15. M1 D.K.Cheng and C. A. Chen, “Optimum Element Spacings for Yagi-Uda Arrays.” IEEE Trans. anfennas Propagat. vl. AP-21, September 1973, pp- 615-623 MW. W. Hansen and 5. R. Woodyard, “A New Principle in Directional Anteana Design, IRE Prec, ¥3.26, March 1938, pp. 333-345. 13 TM, Maer, “Optimum Pregreasive Phase Shifts for Discrete Endre Array,” Syracuse Uni. Res. Inst. Rep. EE 492-600278, Syracuse, N.Y., February 1960, 14 CLL Delp. "A Current Distribution for Broadside Arrays Which Optimizes the Reation- ship between Ben Width and SfeLobe Level” IRE Proc vo. 3, ume 1946, 9.335 15 HJ. Riblet, “Discussion of Dolph’ Paper," JRE Proc. vl. 35, May 1947, pp. 489-492. 16 RH. DuHlamel, “Optimum Patterns for Endre Arrays,” IRE Proc, vol. 4l, May 1953, p- 652-689, 17 RL. Pritchard, “Discussion of DuHlame!'s Pept APS, January 1955, pp. 40—43, 1 MT. Ma, “A New Mathematical Approach for Linear Array Analysis and Synthesis.” PhD. dissertation, Syracuse Univesity, Syracuse, N.Y, 1961. 19 4.5. Stone, US. Patems 1,643,325 and 1.715.433. IRE Trane. Antennas Propagat. vol.‘3:34 Inroduction and Fundamentals 20. L.B. Brown and G. A. Sharp, “Techebyscheff Antenna Distribution, Beamwith, and Gain Tables,” Nav. Ord. Lab. Rep. NOLC 383, Coront, Cali, February 1958 21 A. Block, RG, Medhurst, and S, D. Pool, “A New Appreach to the Design of Super Directive Aerial Arrays.” IEE Proc. (London), vol 100, part IIT, 1953, pp. 303-314 22 M.T. Ma and D.C. Hyovalt A Table of Radiation Characteristics for Uniformly Spaced Optimum Endfire Arrays with Equal Sidelobes, NBS Monograph 95, National Bureau of Standards, Houlder, Colo, December 1965, 23, J.L. Wong and H. E. King, “Directvty ofa Uniformly Excited WX N Array of Disective Elemenis,” LEEE Trans. Antennas Propagat, vl. AP-23, May 1975, pp. AO ~404, 24 N. Amitay, V, Galindo, and C.P. Wu, Theory and Analysis of Phased Array Antennas, John Wiley & Sons, Ine., New York, 1972 25 1.D. Tillman, 5, The Theory and Design of Circular Antenna Arrays, University of Tex ‘nessee Engineering Experiment Station, Kaowille, 1966. 26 CO. Stearasand A.C. Stewart, "An Investigation of Concentric Ring Antenna with Low ‘Sidefobes” IEEE Trans. Antemas Propagat, vol. AP-13, November 1965, pp, 856-863, 27-D.K- Cheng and F. 1, Tseng, “Maximisation of Directive Gan for Cireular and Elliptical ‘Anmays," IEEE Proc, (London), vl. 114, May 1967, p. 389-594.PART 2 Types and Design MethodsChapter 4 Dipoles and Monopoles Chen To Tai ‘The University of Michigan Stuart A. Long University of Houston Equivalent Circuit of a Sleeve Dipole 4-18 ‘Open Folded Sleeve Monopole 4-20 46 Effective Height of Antennas 4.23 ‘General Formula and Its Role inthe ‘Theory of Transmitting and Receiving Antennas 423, Patternsasa Function of Length 4-7 Coupled Antennas 4.24 and Diameter 410, ‘Crcuit Relationships of Radiating 43. Biconical Dipoles 4-12 Symems 424 Impedance as Function of Length 4-8 Monopole Antennas 4-26 ‘and Cone Angle 4-12 ‘Relationship to Balanced Antennas Patterns of the Biconial Dipole 426 412 Effoct of Finite-Size Ground Plane 44° Folded Dipoles 413 ‘on Impedance and Pattern 427 Equivalent Circuit of « Folded ‘Radiation Patter of a Monopole on Dipole 4-13 ‘aCireular Ground Plane 4.27 Impedance Transformation 28 8 ‘Monopole Mounted on the Eage of Function ofthe Ratio of Sheet 430 Conductor Sizes 4-17 ‘Monopole Mounted on a 45. Steeve Dipoles 4-18 ‘Conducting Box 4-324-1_ INTRODUCTION Since the publication of the first edition ofthis handbook several sections in the original ‘chapter on linear antennas have become outdated and have been deleted inthis eition. ‘The availabilty of computer programs for finding the impedance and other characteris- tics of antennas, particularly linear antennas, makes parametric tabulation of limited ‘usage. Only some essential formulas and design data are therefore included in thischapter. For the entire subject of linear antennas, the book by R. W. P. King? remains auortative, Another book* by the same author on the tables of antenna characteristics contains the most comprehensive data on the characteristics of cylindrical antennas. Calculations on circular-loop antennas and some simple arrays are also found there. ‘Asection on the efective height of antennas is included in this chapter. The usage of this parameter in describing the transmitting and receiving characteristics oflinear anten- ‘nas and other simple structures is discussed in detail. In addition, material on the general {formulation of receiving antennas is included so that engineerscan apply the formulation {or design purposes or for estimation of the coupling effect between elements made up of ‘both linear and other types of antennas. ‘Antennas in lossy media are of great current interest. Unfortunately, the subject cannot be covered in this chapter because of limited space. The book by King and ‘Smith! on antennas in matter cam be consulted for this subject, particularly for linear antennas embedded in a lossy medium. 4-2_ CYLINDRICAL DIPOLES: Impedance aa a Function of Length and Diameter The impedance characteristics of cylindrical antennas have been investigated by many seiters. Theoretical work has mainly been confined to relatively thin antennas (length- todiameter ratio greater than 15), and the effect of the junction connecting the antenna proper andthe transmission line is usually not considered. Among various theories, the induced-emf method of computing the impedance of a cylindrical antenna based upon a sinusoidal distribution is still found to be very useful. The fo. ‘mula derived from this method i extremely simple Its, however, valid only when the half length ofa center-driven antenna is not much longer than a quarter wavelength, In practice, this isthe most useful range. To eliminate unnecessary computations, the formula has been reduced to the following form’ 2,~ net) — j[120(in— 1) oa xe - a40)] (4-1) where Z, = input impedance, 2, of a centerdriven cylindrical antenna of total length 26 and of radius @ ke = 2e(E/d) = electrical length, corresponding to é, measured in radians ‘The functions R(ke)and X(ké) are tabulated in Table 4-1 and plotted in Fig. 4-1 for the range k¢ = 1/2. For calculation purposes, these wo functions can be approximated to ‘within 0.5 by the following simple third-order polynomials: Rik) = ~0.4787 + 7.3246Kt + 0.3963(k2)F + 15.6131ke? Mkt) = -0.4456 + 17,0082k¢ — 8,6793(ke) + 9.60310ke)?>Dipoles ant Monopoies 45 TABLE 4-1 Functions A(k¢) and X(k¢) Contained in the Formula of the input impedance of 2 Conice Sores Cylindrical Antenna Ke RUKt) (ke) ke ACRE) ox) ° ° ° 098s 1501 o4 01506 = 1g10 10 19 02 07980 2302 2056 03 1821 aes 12 2393 4 3266 55613 2788 05 sam ria 1 20 06 7.563 sa rs 38.00 07 oas 10.68 a2 4246 083.99 1273 Shen te Jength of the antenna is short compared with a wavelength but stl large ‘compared with is radius, the same formula reduces to (2am = 200867 — p24)" (in — 1) (42) Reneé theories provide formulas for the computation of the impedance funelon, Rone of them, however, is simple enough to be included here. As far as numeict Trpulation is Concerned, Schelkunof's method? is relatively simpler than Halen? Hisheuld be emphasized that all these theories are formulated by using av idearneg ‘model in which the terminal condition is not considered. 8 70| 8 UKE) AND X(KE) IW OHMS Bos 38 o 05 io weanid FIG. 4-1. The intone ACHE) and x4)4-8 Types and Design Methods In pace he antennas fd by a assis ne. The complete sy- tem ma ave the spenancs shown in Fig: 42. Te efectve terminal inpoance Si the ne (olen Teed toa the a seae sntnna inpeaee) then depends et ATEN EAL SE Sly upon te length andthe lametr of xe Oe Sa ctene bate upon he tenn Sond eno an 6 the taped | Stee mui ss ea funn of the es ine Earnna of the ground plane. For a given terminal contin the ation fe impotence i) wy of a cylindrical antenna as a function of the mgt tnd the dnt ofthe incase shown th egret ‘GROUND PLANE work of Brown and Woodward." The data Soret we nage of tv af the iSeginio-damotar ati, Too uefa ses ANTENNA eee of curves are reproduced in Figs. 43 and 44, The impedance refers to a cylindri- o cal antenna driven by @ coaxial tine FIG. 42. Drivng an antenna by transmis- through a latge circular ground plane son Ines. placed on the surface of the earth. The arrangement is similar to the one sketched in Fig. 42a. The length and diagreter of the antenna are measured in degr a length of one wavelength is equivalent to 360°. If the effects due to the terminal condition and finitesize ground plane are neglected, the impedance would ‘correspond to One-half of the impedance of a centerriven antenna (Fig. 4-2c). In using these data for design purposes, one must take into consideration the actual ter- minat condition as compared with the condition specified by these two authors. Ia pésticular, the maximum value of the resistance and the resonant length of the Antenna may change considerably if the base capacitance is excessive. Effect of Terminal Conditions Many authors have attempted to determine the equivalentcircuit elements corre- sponding to different terminal conditions. Schetkunoff and Friis? have introduced the concepts of base capacitance and nearsbase capacitance to explain the shift of the impedance curve asthe terminal condition is changed. Similar interpretations have been given by King'®for a cylindrical antenna driven by a two-wire lin or by a coaxial line and by Whinnery" for a biconical antenna driven by a coaxial line. The impor- tance of the terminal condition in effecting the input impedance of the antenna is shown in Figs. 45 and 4-6. They are again reproduced from Brown and Woodward's paper. Because ofthe large variation of the effective terminal impedance of the line with changes in the geometry of the terminal junction, one must be cautious when using the theoretica’resufts based upon isolated antennas. For junctions possessing simple geometry, the static method of Schelkunoff and Fris, King, and Whinnery can be applied to estimate the shunt capacitance of the junction. The latter then ean be ‘combined with the impedance of the antenna proper to evaluate the resultant imped- ance. For intricate junctions, accurate information can be obtained only by direct measurement,Dipoles and Moncpoles 4.7 a9 AesisTaNce ~ ous g 8 3 3 300 200 G 5075 00 ase ANTENA LENGTH OEGREES reanct vertua antenna length A when a constant rato of waa ho eta A’ Die mained Hare the angih anddametar rote cna ‘wile the frequency Is changed. 73 2002S = Ma +b)" (43) Where a = m 2 major axis of eipse = minor axis of etipse4-8 Types and Design Methods 600 00] g CT 100 5 150. 17S 200 Be ANTENNA LENGTH = DEGREES FIG. 4-4 Reactance curves corresponding to the resistance curves of Fig. 4-3Dipoles and Monopoles 4-0 RESISTANCE ~OHMS 08888 8858a8 SRHSSSLSRSLSIISSLSRSIags ANTENNA LENGTH-DEGREES FIG. 4-5 Resistance asa function of antenna length A. The diameter 0 8 20.8°, ‘Curve A: The arrangement shown In Fig. 4-2b. Curve &: The arrangement of Fig, ‘4-24 with the dlameter of the outer conductor equal to 74°. The characteristic Impedance ofthe tranamission line ie 77.0 9. Curve C: The outer-conductordlam- ‘ter I 49,6", and the tranamiasion ine has & characteristic impedance of 52.50. ‘Curva 0: The diameter of the outer conductor i 33°. The characteristc impedance 's 28.3.0. Curve E: Tis curve wes obtained by tuning out the base reactance with ‘an inductive reactance of 65.0 2. o ANTENNA LENGTH ~DEGREES 220 230 240 250 2d p -20) ~29} 29} 100 te a0 69 =| 200] ~2204 ~240] 280) 200) 209] 329 ~xal FIG. 4-8 Reactance curves corresponding tothe resiatance curves of Fig. 4-5. EACTANCE -OnMS4-10 Types and Desion Methods 08, o.5| : a ° SI OE OS 04 OS 08 OF 08 OS 10 FIG. 4-7 Equivalent radius au ofa rectangle a8 & function of fhe ratio of thickness fo width 8 Fora rectangular cross section the result is pleted in Fig. 47. In the cae ofa 8° Fees er pete dental renal, When th cot section has the frm Fo a eC rl is tabulated in Table 42. The equivalent radi of 0 parallel cylinders of radius p, and P2 separated by a distance d between the centers is Ben ye 1 inp atm + bln oa + 2a ln 44) pe gpieiinn tain ttanng 44) omuas for the equivalent radius of thre cinders and an angle strip ae found in Bet i Patterns as a Function of Length and Diameter tn this subsection only the radiation pattern of center-driven ylindrical antentes 1 te eet puucdrven amennas, the Gattems depend very much upos the size of the ground plane, The subject will be discussed in Sec. 48. oo reo pater ofa centerriven cylindrical antona in general depens ‘spon fis length and thickness. The terminal condition which plays an iopostnt role wre SSninng its impedance bas» nealgible effect on the patter. For thin antennes ao gene tern obtained by assuTing a sinisidal cursentdinebutian i & god “TABLE 4-2 Equivalent Radius of Regular Polygon n 3 4 8 6 fF 24 0.5903 0.7563 09200 a frm tmber of side. rain fa cutserbed ice,Cipoles ana Monopole 414 approximation of the actual pattern. Thus, with an assumed current distribution of the form IG) = hsink(@—|z1) +82 2z-€ (45) the radiation field, expressed in spherical coordinate system, i given by = Selec [08 (KE cos 8) ~ sate Fen eR { sind (46) where 9 = (4/6)! = 120n 9 4 = angle measured from axis of dipole or z axis ‘The field pattern is obtained by evaluating the magnitude of the tm contained in the brackets of Eq. (46). Some ofthe commenly referred-o patterns are sketched in Fig. “£8. Comparing those pattern with the actual pattern ofa thin cylindrical antenna Lek LHLZSALSTBA reo FIG. 4-8 Radiation pattems of conter-driven dipoles it sinuecidal cur- rent stribution ip assumed, ‘obtained by measurement, one finds that the theoretical patterns based upon a sinus- cidal current distribution do not contain the following information: 1 The aulls between the lobes, except the natural mull in the direction of the axis, are actually not vanishing 2 The phase of the field varies continuously from lobe to lobe instead of having sudden jump of 180° between the adjacent lobes. 9 The actual patterns vary slightly with respect to the diameter of the antenna instead of being independent of the thickness. Depending upon the particular applications, some ofthe fine details may require Special attention. In most cases, the idealized patterns based upon a sinusoidal current ‘distribution give us suficent information for design purposes. ‘When the half length £ of the antennas is less than about one-tenth wavelength, Eq, (4-6) is well approximated by ale kereh* by Rai (47)4-12 Types and Design Mattiods ‘The figure-cight pattern resulting from the plot of the sine function isa characterist ‘ot only of short cylindrical antennas but also ofall small dipole-type antennas. Equa- tions (4-6) and (4-7) are also commonly used to evaluate the directivity of linear antennas. The directivity is defined as = Maximum radiation intensity a ‘average radiation intensity (48) For a short dipole, Dis equal to 1.5. The directivity of a half-wave dipole (€ = /4) is equal to 1.64, The half-wave dipole is often used as a reference antenna to describe the gain of more directive antennas, particularly arrays made of dipoles. 43 BICONICAL DIPOLES Impedance aa a Function of Length and Cone Angle ‘When the angles of a symmetrical biconical antenna (Fig. 49) are small, the input impedance ofthe antenna can be calculated by using Schelkunof"’ formula.” Some rn ae fee ‘i wate nt oe sample curves are shown in Fig. 410. While the biconical antenna is an excellent theoretical model for studying the essential property of « dipole-type antenna, small- Angle biconical antennas are seldom: ase in practice. Wide-angle biconical antennas or their derived types such as discones, however, are frequently used as broadband antennas. The broadband impedance characteristics occur when the angle of the cones, of Fig. 49, lies between 30 and 60°. The exact value of ty is not critical. Usually it is chosen so that the characteristic impedance of the biconical dipole matches as clsely as possible the characteristic impedance of te line which feeds the antenna. The characteristic impedeace of a biconical dipole as a function of the angle is plotied in Fig. 41. For a conical monopole driven against an infinitely large ground pilane, the characteristic impedance and the input impedanceof the antenna are equal to haf of the corresponding values of a dipole. Several formulas are available for ‘computing the input impedance of wide-angle biconical antennas. Actual computation ‘has been confined to a very few specific values of fo” More complete information is evailable from the experimental data obtained by Brown and Woodward." Two ‘curves are reproduced in Figs. 412 and 4-13. The case corresponding to a = O° rep- resents cylindrical antenna having @ diameter of 2.5 electrical degrees ata frequency of 500 MHz, since the feed point was kept fixed at that diameter. Patterns of the Biconical Dipole ‘The radiation patterns of biconical dipoles have been investigated theoretically by Papas and King.” Figure 4-14 shows the patterns of a 60°-fare-angle (@ = 30°) ‘conical dipole for various values of ka, where k = 2x/A and a = half length of theDinoles and Moncpoles uta, en (4) en (2) cee see MPL imoudance of small FIG. 4-106 Input impotence of sma ‘ool coil ntenas(rsetance).”” angle Nooricaletemn freency dipole which isthe same as the £ used in Fig. 49. Similar curves corresponding to {ferent values of the flare angle have been obtained. ‘experimentally by Brown and Wood ward,!* 44_ FOLDED Dipoves Equivalent Circuit of a Folded Dipole thaorde site is formed by joining two cylindrical dipoles at the ends and driving i Pit of transmission ins atthe center of ene arm ax shows in Pig. ei teciameter ofthe two arma cat be either identical rere simple analysis, mene 8 quasistatic approach, ofthe operation ofa folded dpsie ne arbitrary ‘mension has been given by Uda and ‘Mushiake."? According to their method, the400. e202 (88) 300} KIN Ons, 100 oS BG AO 50 BG TO AO GO 8 DEGREES FIG. 4-11 Characteristic impedance ofa biconical pote. g pemsauce-ovees 8 30 708 150 200-50 Soo ANTENNA LENGTH ~OEGREES. FIG. 4-12 Measured resistance curves of the conical unipole versus length in ! lcircal dagrese for various fare ances. a jREACTANCE -OHMS, -s00f 30 708 anteNna FIG. 4-13 Measured 780 LENGTH -oegReES 5 ~~ 3oo .rv08 of the conical unipale versus length in wactance ‘electric! degrees for various fare angios, FG. 4-14 i iota of the absolute values of ‘enthal anale 8 tor various vahies of ke and {he farzone electric felt as & Anction of the with a fare angle equal to €0° (45 = 30°). ans4-16 Types and Design Methods =- FIG. 4-15 Folded dipole. excitation of a folded dipole can be considered as a superposition of two modes as shown in Fig. 4-16. The impedance of the symmetrical mode, characterized by two ‘equal driving voltages, can be calculated by making use of the equivalent radius of wetloany tr hte wf] le wit B] vb g Sty ovtgg iv tt naous fm FIG. 4-16 Decomposition of the folded dipole into two Kundamental modes. two conductors as discussed in Sec. 42. The equivalence is shown in Fig. 417. The impedance function Z, is therefore the same as the impedance of a cylindrical dipole with an equivalent radius p, given by 14g Inge Ines + pean n + 2H ine) (49) ‘where the various parameters are explained in Fig. 416. The impedance of the asym- metrical mode, characterized by equal and opposite currents on the two arms, is the same as the shorted section of transmission line of length equal to &; that is, Ze CEO o say tan kt (4-10) ular af] a. J sf odP 2 VBB I VR 2 ep Vv - VA 2 isan FIG. 4-17 The equivalent representation FIG. 4-18 Equivalent cicult of a folded of the symmetrical mode in computing Z. poe.Dipoles and Monopoles — 4ea7 sphere Zo isthe characteristic impedance ofthe two-wte line. Expressed in terms of Zand Z; the input impedance of a folded dipole is given by zoti G4 Or_ m+ ap2z, 1” EF 0+ a), 423, An equivalent circuit based upon Eq (4-11) is shown in Fig. 18, For a folded dipole of length ¢ equal 10/4, Z;is very large compared with (1 + a)*Zs hence Zum + oz, (4-12) (4-11) Impedance Transformation as a Function of the Ratio of Conductor sizes [The step-up impedance ratio (1 + 2)? a8 a function of w andr has been calulated by ‘Mastinke." The diagram is reproduced in Fg. 4.19 by using the formula fora ve IMPEDANCE STEP-UP Rat] 2530°39-40-45-501~60-1-701 g. B1s1204 30 = - FIG. 4-19 Stop-up transformation chart for fldnd dipole in Fi 416. When o, and ge are smal! compared withthe value of ais given toa 8064 approximation by = 24/0) 13; on nae (#13) This formula was fist derived by Guertler 2" Avother presentation” of the transformation ratio (1 + a)? in a logarithmic scale 98a function of ox/p, and d/o given in Figs, #20 and 4°h‘Types and Design Methods 25 q 15| a & lita 20 ris 2 essa 8 8 Fe 8 ve FIG, 4-20 px/p; versus impedance transformation ratio and dos. (@ 1982 IEEE.) 4-5 SLEEVE DIPOLES Equivalent Circult of a Sleeve Dipole ‘The geometrical shape of a sleeve antenna, ot a sleeve monopole, is sketched in Fig. 42a, Ifthe image of the structure is included, then we have a sleeve dipole as shown in Fig. 422. A sleeve dipole can therefore be considered as a doubly fed antenna in which the current is a relative maximum at the center of the dipole or at the base of ‘the monopole. Antiresonances of the antenna impedance function take place when SDipoles and Monopoles an o/b =65) Iso, 35 59 49] 39 2s, °S 10} ig a 7| ¢ 5 as ema Seem ESH dare 4nasig@h cael ic ee FIG. 4-21 impedence transformation ratio versus p2/p, and d/o. (© 1982 IEEE.) i approximately equal to an odd multiple of a quarter wavelength or Z isa multiple Git half wave. A special case of interest is that in which L + S'is equal toa quarver Wavelength, Then the current distribution along the structure is approximately cost. ja soidal At resonance the input resistance of the sleeve monopole or the sleeve dipole is approximately given by Reo ne (ae2t) (414)4-20 Types and Design Methods where R denotes either the input resistance ofa resonant quarter-wave monopole or that of a half-wave dipole. The sleeve in this case plays the role of an impedance transformer. ‘Wong and King! have shown exper- —; mentally that opensieeve dipoles properly designed exhibit @ broadband voltage- 4 standingswaveratio (VSWR) response { {and unidirectional radiation patterns over early an octave bandwidth when placed 1 above a reflector. Figure #23 shows the ‘VSWR response of open-seeve dipoles for various dipole and sleeve diameters. Figure {426 shows the same wit sleeve spacing as 1 o the parameter. Open Folded Sleeve Monopole ‘The work of Uda and Mushiake! on folded dipoles can be extended to include a Joad at the unexcited arm as shown schematically in Fig. 425. The impedance of the loaded dipole is given by FIG, 4-22 The sleeve antenne. ae (Zot ZZ, + & "5 ) 2d (ass) Zt B+ a are whe 2, = feet 2, Blom! E+ a =a) + cob ot t+) Z, = input impedaice ofthe folded dipole when itis driven simultaneously by ‘4 common voltage at the base. Itis approximately equal to that of @ 4ipole with an equivalent radius p, given by Eq. (4-4). ‘The parameter a in Eq. (4-15) and the parameters y and w in the expression forthe characteristic impedance Z, of a transmission line made of wires of unequal radius are the same as those defined for Fig. 416. In the special case ia which 2, approaches infinity, Eq. (4-15) reduces to Za ts+( Z (4-16) ‘The structure then corresponds to an open folded dipole or monopole. Josephson studied ths structure both theoretically and experimentally. However, his analysis is correct only if the wires are ofthe same size. A sleeve version of an open folded mon- ‘opale was also investigated by Josephson. The two structures are shown in Figs. 4-26 and 4.27 “The open folded monopole with a displaced feed point is equivalent toa foldedFIG. 4-23 | VSWR reeponee of opemsleeve dipoles for various dipole and sleeve diameters. SLEEVE spac, s-190In (46.72 mm). 301.2910 (82.77 mm} sive wom, LS 11.88 In (391.78 ron) POU Loe, He 217 im 1582.06 mm) a0 0 PReaUENCY, ne FIG. 4-24 VSWRtora tin (18+mm-) dlameter open-sleeve dipole with aloove spasing as the parameter, 421be si Py Pa —k4 FIG. 4-25 Loaded folded aipole end monopole. FIG, 4-26 Open folded monopole. || F1G.4-27 Open falded monopole witha di placed feed point. 492Dipoles and Moncpoles 4-23 sleeve monopole. The input resistance at resonance ofthe open folded sleeve monopole is given approximately by Rak (sn #3) /(in 2} (417) where Ro denotes the resonant input resistance ofthe open folded monopole and R, is the input resistance of the folded sleeve monopole. The precise value of Ry depends on the radius of the wires and their separation. A typical value is about 10 2. The open folded sleeve monopoles made of arms with different sizes are useful models in design- ing aircraft trail antennas or fnrtype antennas on vehicles as ilustrated by Josephson, 4-6_ EFFECTIVE HEIGHT OF ANTENNAS: General Formula and its Role in the Theory of ‘Transmitting and Recelving Antennas ‘The radiation field of any antenna can always be written in the form = DUKZahe 9 Ee (418) where Zp = (so/4@)"* ho afd bh effective height of antenna 4, = input current to antenna ‘The effective height ofan antenna was originally introduced by Sinclair. Is related to the radiation vector defined by Sehelkunoft to characterize the radiation field of en antenna, ie, Ne= Lh (4-19) where N, denotes the transversal part of Schetkunofs radiation vector. The effective ‘eight isa very useful parameter in antenna engineering. For example, the open-i= cuit voltage ofa receiving antenna can be exprested as Vg Ey-b (420) where E; denotes the incident electric fel. It i also an important parameter involved in the polarization-matching factor of a receiving antenna, Inthe theory of receiving ‘antennas the receiving cros section or the effective aperture is defined by* x A= i Doe (4-21) where d = operating wavelength D = directivity of antenna 9 = poarization-matching factor Ee in Te 4 = impedance-matching factor4-24 Types and Design Methods Zt% Z, * input impedance of antenna 2, = load impedance The polarization-matching factor ‘The effective-height function of a center-driven short dipole, ifa linear current distr- bution along the dipole is assumed, is given by b= ~esin 00 (4-22) jn involves the effective-height function. where £ denotes the half length of the dipole which is assumed to be pointed in the vertical direction and 6 denotes the polar angle measured between the axis of the dipole and the direction of observation or the direction of radiation. For a Jineariy polarized incident field making a skew angle « with tte axis ofthe dipole, the polari- 2ation-matching factor is given by pacwta (423) “The effective-eight functions for other simple antenaa elements are listed below? Antenna type Effective height Shart dipole of length 2€ Halfwave dipole ‘Small loop of radius a pointed in the = direction Half-wave folded dipole 4-7_ COUPLED ANTENNAS Circuit Relationships of Radiating Systems ‘When several antennas are coupled to each other, the input voltage and input currents ‘othe antennas follow the same relationship as ordinary coupled circuits.” For a sys- tem of n antennas, the relationships are Wad yh iakBe ue (4-24) where Zy is called the self-impedance of antenna { and Z or Zy is called the mutual impedance between antenna f and antenna j. In the case of linear radiators, Carter's ‘method, or the induced-emf method based upon sinusoidal current distribution, is the simplest one to use in determining the various Z's, The arethod applies only to anten-Dipoles and Monopoles 4-25 tas shorter than a half wavelength. The seimpedance determined by this method ia the same as that given by Eq. (4-1). The formulas fr the mutual impedance of twa Parallel antennas of equal size are found in Carter's original paper or in Kraurs ‘ook: Figure 4-28 shows the mutual impedance of two parallel halfware ancenes 80) MUTUAL RESISTANCE MUTUAL REAGTANCE ws 8 RESISTANCE OR REACTANCE IN OM x FIG. 4-28 Mutual impedance between two parallel hallwave antennes placed side by side, placed side by side. Figure 4.29 shows the mutual impedance of two parallel collinear uae antennas, Mutual impedances of wo parallel antenas of anegual ore ive been investigated by several authors >">! The induced-emt meliod tes eo ‘tplied to crossed or skewed antennas™” to evaluate ther mutual impedance Retard Grrelations based upon Hall's integralequation cechnique are found te the asic Battie Bouwkamp, and Uda and Mushiake.” The lst Iwo asthon alg ee, sigpelFimpedance and mutual mpedanceof parallel antennas of unequal ioe ee Ultimately applies to the design of Yagi-Uda arrays. ot Poles separated by stance which is large compared with 9 wavelength, iceaitval impedance between the two dipoles can be calelaedby ving the asymp: totic formula dkZee¥* “anR Zam (by » by) (4-25) where k= 26/n Zo = (uo/eg)'™ R = distance between centers of dipoles In be ~ effective height of dipoles Srl i quite accurate for hll-vave dipoles with s separation barely greater than one wavelength.4-28 Types and Design Methods * { | CECE | TH | MUTUAL RESISTANCE T TI LC LEH RESISTANCE OR REACTANCE IN N Os ti (LI See4 Yi ml sta es aR FIG, 4-28 Mutual impedance between two collinear hel-wave anvennes. 4-8 _MONOPOLE ANTENNAS Relationship to Balanced Antennas When a monopole is mounted on an ideally infinite ground plane its impedance and wMitation characteristics can be deduced from that of a dipote of twice its length in Tear puce. For a basedriven monopole, it input impedance is equal to one-half that Gane center driven dipole, and the radiation pattern above the infinite ground plane ee dcntiel with the upper half of the radiation pattem of she corresponding dipole. ‘When the ground plave is of finite siz, the image theorem does not apply. ‘Several methods have been devised 10 investigate the characteristics of » mono” pole mounted ona finitesize ground plane. The frst method is due to Bolljahn.* who prnsiders the problem from the point of view of symmetrical components, The desor xition i shown in Fig 4-304, in which the ground plane s assumed to be ofthe frm zo le 4 2 1 4 NZ. ae =e 7 + F FIG. 4-08 Monopole and frite-size around plane tnd Its decomposition into two modes of excitation.Dipoles and Monopcles 4-27 ofan infinitely thin conducting disk. For the sypumetical mode of excitation the pres- ence of the disk has no effect upon the radiation ofthe two elements. The problem is therefore the same as if the twa elements were placed in free space. The antisym- ‘metrical pair of current elements excites equs currents on the top and the bottom sides of the disk. This mode is responsible for the variation of the input impedance of the antenna asa function of the disk diameter Its also responsible fr the asymmetry of the resultant radiation pattern with respect to the ground plane. Boljahn's original ‘work was developed by assuming a short monopole on a disk. The entre analysis is found in Schelinaffs book Advanced Antenna Theory.” His study of the character- ics of large ground planes was aver extended by Storer” to monopoles of arbitrary length. Effect of Finlte-Size Ground Plane on Impedance and Pattern ‘According to Storer, vba used a variational method o formulate the problem, the ‘change ofthe input impedance of a base-éven monopole erected upon a large circular round plan can be writen as « "2 4)? =z-2,-j%em | az-2-2-j Re je [Bal (4-26) where Zp = impedance of monopote referred to an infinite ground plane, @ = diameter of circular ground plane k= 2p fh height of monopole 1G) = currentdisribution function of monopole 1(0) = base current or input current ‘The function j(60/ kd)e™™, which is independent of the current distribution, is plotted in Fig. 30. 7, teal and the imaginary parts, “fe Tonction are respectively equal - “1 44)" gna ix — WO al" ora war n9/ tf 2aepam ce 20/ (t Bal tora qanee a 10) ‘wave monopole, if we assume I(z) = 1(0) cos kz, then * 1) * Lao! ‘Thus, with aground plane ofa diameter greater than 10 wavelengths, itis seen from Fig. 4306 that the variation of the resistance and fhe ectanoe ofa qurirnave monopole isless than 1, Radiation Pattern of a Monopole on a ‘Circular Ground Plani ‘While the effect ofa ground plane upan the impedance of a monopole is not very great, ‘the radiation pattern is affected considerably. The pattern of such a composite antenna ‘material contalaed fa thi subsecion econdowed rom « commoniasion from Dr, Robert Kou soumjin excluively prepared for this handbook The help of Dr. Kouvoumjan i geal = iy pep a voamyjian & griefely4-28 Types and Design Methous Lal UNVERSAL CURVE FOR THE CHANGE OF ANTENNA IMPEDANCE 'AS A FUNCTION OF THE QIAMETER OF THE GROWNO- PLANE FOR A DIPOLE ANTENNA FFIG. 4-206 Universal curve for the change of antenna impedance as a function of the lometer of the ground plane. ‘can be obtained quite accurately from the solution of the uniform geometrical theory of diffraction (GTD)." For a short monopole of length h positioned at the center of f circular disk of radius a, as shown in Fig. 4.31, the current on the monopole is assumed to be sin kth = 2) ne (RT) {the point of observation is not near the vertical axis (8 not close to 0 or =), the 1a) = IeDipoles and Monopoles 4-28 radiation field can then be expressed in the form E= (Eo + Ea + Bao (4-28) where Bp is the geometrical optics field given by ( szoloe-™® [open £08 ki }osecd Ee Jazere (429) t 1, in practice the result is valid even {for ka a8 small as 2x. When the argument ofthe transition function F(x) is greeter ‘than 10 oF so, F(x) is approximately equal to I; then the sum of Ey and By, yields a a i ra) ere Fat Fam ~E4an[n! \/ a {ue($+ $e ‘i = 8/91) ) go = 8) enateam eran {yg RMN) goa we(E-gem [1+ wae | |S os |ntheaxil ion (9 closet 0 orn) se pattern ca be found by means ofan equivalent de curent © The total diracted field in this ease given to eittenni Ent bam Fitdat)SsecT Jasin oft + Ser) ass)4-20 Types and Design Methods ‘This approximately equals Ey, and Eq given by Eas. (4-30) and (4-31) at ka sin 8 = 5 for # « x/2. Thus, the pattern calculated from Eq. (4-38) joins smoothly with that ccleulated by Eqs. (4-31) and (4-32) when ka > 5. Figure 4-32 showsa typical pattern based on Ea, (4-28), the solid line, and Ea. (4-35), asm neoweay FIG. 4-32 The radiation pattem of @ monopole above 8 circular dak having @ radius of three wavolongine. ‘the dashed line. For practical purposes the term Eq due to the doubly defracted rays is ‘uch smaller compared with Ey except in the region where @ is close to x/2. By using a hybrid moment method jointly with GTD, itis possible to determine the input impedance of a monopole centered on a perfectly conducting circular disk.*! Monopole Mounted on the Edge of a Sheet When a monopole is driven against an infinitely large conducting half sheet, the prob- lem can be formulated conveniently with the aid of the dyadic Green's function per- taining to the half sheet. By transforming the resultant series into definite integrals, Sawaya, Ishizone, and Mushiake® have been able to calculate the impedance of a ‘monopole mounted on a half sheet in several orientations. Theirresultsareshown in Figs. 433 and 4-34,Dipole and Moncpoles. 4-91 Taro r r -£ (ast In 3 y de do 0 i i hoo 3 nbicrance (ous) Escunce oss) gs i & FIG. 4-93 impedance of a horizontal (a) and of a vertical (6) monopole antenna as a func- ‘ion of ¢. (© 1981 IEEE) Ve=200 [noasig EE REN] can cof rare avers oo see NSE, eH eaters a r eo i oat} fs i» S| Boe bl Bae [oaks [essl as op a | dl tee! TEE ‘he Sor as ear om ae oo FIG. 4.34 Impedance of vertical monopole antenne on conducting half sheet as function of Pi, whore x the stance from the edge ofthe ha sheet to the base of the monopole.) Fesistace. (0) Reactance. (© 1981 IEEE.) ‘These theoretical data can be used to estimate the impedance of monopoles ‘mounted on large but finite conducting sheets. These authors also calculated the imped- ‘ance of a notch antenna cut on a half sheet. The radiation patterns of short dipoles and small loops mounted on a half sheet are found in Ref 42. Other work also bas been reported by Pozar and Newman“ and by Marin and Catedra for the impedance and radiation patterns of monopoles located near the edges of finite half sheets or disks.G,B (m—mhos) 8 12 14 18 18 20 Frequency (GHz) FG, 4-35 Admittance versus frequency of @ 6cm monopole mounted on a cubical con ‘ducting box with edge length 10 om for various focations between the canter and comer. (@ 1987 IEEE) Monopole Mounted on a Conducting Box ‘Recent investigations by Bhattacharya etal and Chu eta” have reported the imped- ance and radiation patterns ofa monopole mounted at arbitrary locations on aconducting ‘box that is comparable in size to a wavelength. Figure 4-35 shows the behavior of the admittance of a tonopole as a function of frequency as the position of the monopole is moved from the center of the box face to one comer.* REFERENCES 1 G.J. Barke and A.J. Pogeio, “Numerical Electromagnetic Code (NEC),” Tech. Doc. 116, Naval Ocean Systems Center, California, 1980. 2 RW. P. King, The Theory of Linear Antennas, Harvard University Press, Cambridge, Mass, 1956. 13. RLW.P. King, Tabler of Antenna Characteristics, Plenum Press, New York, 1971 4. RLW, P. King and G. Smith, antennat In Matter, The M.LT. Press, Cambridge, Mas. 198, ‘5. P.S, Carter, “Circuit Relations in Radiating Systems and Applications to Antenna Prob- lems, IRE Proc. vol. 20,1932, pp. 1004-1041. (6 RS. Eliott, antenna Theory and Design, Prentice Hal, Inc, Englewood Clifs, NJ 1981, pp. 301-302ty 2 4 3 Dipoles and Moncpcies 4-39 S.A. Schelkunof, Advanced Antenna Theory, John Wiley & Sons, Inc, New York, 1952. George H, Brown and O. M. Woodward, Jr, “Experimentally Determined Impedance (Characteristics of Cylindrical Antennas,” IRE Proc. vol. 3, 194S, pp. 257-262 S.A. Sebelkuroff and H. T. Pris, Antennas: Theory and Practice, ohn Wiley & Sone, Ine, New York, 1952, sec. 13.22, pp. 445-448, RW, P. King, “Antennas and Open-Wire Lines, Part 1, Theory and Summary of Met- surements," J. Appl. Phy. vl 20, 1949, pp. 832-850, “The End Correction fora Coatal Line When Diving an Anienna over a Ground Screen" JRE Trans. Antennas Propet vol, APS, no. 2, Apil 1955, p. 66 Joba R. Whinnery, “The Effect of Input Configuration to Antenna Impedance” J. Ape. Phys, vol. 21, 1980, pp. 943-956, Y. T. Lo, “A Note on the Cylinrical Antenna of Noncircular Crosssestion.” J. App. Phys, vl. 24, 1983, pp. 1338-1339. 'S, Uda and Y. Mushiake, Yagi-Uda Antenna, Maruzen Co.,Ltd, Tyo, 1984, p19. S.A, Schelkunof, “Theory of Antennas of Arbitrary Siz and Shape,” IRE Proc, vl. 29, September 1941, p. 493; se also Ref. 7, Chap. 2. Rel. 7, Chap. 2. .H. Papas and R. King, “put Impedance of Wide-Angle Conical Antennas Fed by a Coaxial Line," IRE Proc, vol. 37, November 1949, p. 1269. €.T. Tai, “Application of a Variational Principle to Biconieal Antennas." J. Appl. Phys. val 20, November 1949, p. 1076 G.H. Brown and ©. 4, Woodward, Jr, “Experimentally Determined Radiation Charac- teins of Conical and Triangular Anta.” RCA Ret, vol 13, 4, December 1952, 425, C.H. Papas and R. King, “Ratliation from Wide-Angle Conical Antenna Fed by a Coaxial Line," IRE Proc, vol. 38, November 194, 9. 1268. YY. Mashiake, “An Exact Step-Up Impedance Ratio Chart of a Folded Antenna,” JRE Trans. Antennas Propagat, vol. AP-, no. 4, October 1954, p. 163. R. Guerer, "Impedance Transformation in Folded Dipoles" J. Brit. IRE, vol. 9, Septem ber 1943, p 344. RC, Hanten, “Folded and T:Match Dipole Transformation Ratio,” IEEE Trans. Anten- nas Propagat. ol. AP-30, no. 1, January 1982. J.-L, Wong and H. E. King, “An Experimental Study of Balun-Fed Open Sleve Dipole in Front of a Metallic Reflector,” IEEE Trans. Antennat Propagat, vol. AP-20, March 1972, p. 201, See also “Design Variations and Performance Characteristic ofthe Open- ‘Sleeve Dipole.” Aerosp. Rep. TR-0073 (3404)2, Electronics Research Laboratory, The ‘Acrospace Corporation, Las Angeles, Cali Bengt Jouephson, “The Quarter-Wave Dipole," JRE Wescon Conv. Rec, part I, San Fran sco, August 1957, p. 77. George Sinclair, “The Transmission end Reception of Eliptically Polarized Waves,” [RE Proc. vol 38, 1950, 9.148. “On the Definition of the Bective Aperture of Antennas," JRE Trans. Antennas Propagat..vel. AP-9, 1961, p. 224 P.S, Carter, “Circuit Relations in Radi lems," IRE Proc, vl. 20, 1932, p. 1004 J.D. Kraus, antennas, McGrav-Hill Book Company, New York, 1950 .R. Cox, “Mutual Impedance between Vertical Antennas of Unequal Heights,” [RE Proc. vol. 38, November 1947, p. 1367 G.Boraili, “Mutual Impedance of Parallel Aerial” Wireless Eng. vol. 25, November 1948, p. 34 Madhur, "Muu Impedance of Farall Aer,” Wireless Eng, vol 28, February SI, p. 67 L Lewin, "Mutual Impedance of Wire Aerials,” Wireless Eng, vol. 28, December 1951, 382. ig Systems and Applications to Antenna Prob-“a a a RG. Medburt, “Dipole Aerials in Close Proximity,” Wireless Eng, vol 28, December 1951, p. 356 C.T.Tai, “Coupled Antennas, IRE Proc. vol 36, April 1948, p. 47. 1. Bouwiamp, “On the Theory of Coupled Antena,” Philips Res. Rep. vo. 3, June 1948, 213. J.T Boljahn, “Antennas neat Conducting Sheets of Finite Size," Univ. California Dept. ag. Rep, 163, December 1949, I.E Storer, “The Impedance of an Antenna ove & Large Circular Seren," J. Appl. Phys, vol. 12,1951, p. 1088, RG. Koujoumiian and P. H. Pathak, “A Uniform Geometrical Theory of Diftacton for an Edge in a Perfectly Conducting Surtace,” EEE Proc, vol. 62,1974 pp. (448-1461. 1R.G. Kowyoumyian, “The Geometrical Theory of Diftacon and Its Application,” in R. ‘Mitra (ed), Topics in Applled Physics vo. 3: Numerical and Arymptot Techniques in Elecrromagnetic, Springer-Verlag OHG, Bri, 1975, . 204 EF. Knot and T B.A, Senor, “Comparison of Three High Prequency Diffraction Tech- niques." IEEE Proc, vel. 62,1974, pp. 1468-1474. GA. Tiel and T. H. Newbouse, “A Hybrid Technique for Combining Moments withthe {Geomata Theary of Difco,” EE Tro. demas Propel val. AP, 1973, pp. 62-69, GE Ta, Dyadic Green's Fant inElecvomegnee Theor, INTEXT, Season, Po, wi. XK. Sawaya, T tshizone and Y. Mushiake, “A implied Expression forthe Dyadic Green's Function fora Conducting Halhect." IEEE Trans. Antemas Propage, vol. AP29, Sep- tember 1981, 7.789, D. M. Pozar and E. H, Newman, “Analysis of a Monopole Mounted Near an Edge or @ ‘Veries," IEEE Trns, Antennas Propagat, vo. AP-30, May 1982, pp. 401-408. M. Marin and M. F Catera, “A Stady of Monopole Arbitrarily Located ona Disk Using Hybrid MM/GTD Technique,” IEEE Trans. Antennas Propagat. vol. AP-35, March 1987, pp. 287-292. '. Bhattacharya, S.A. Long, and D. R. Wilton, “The Input Impedance of a Monopole ‘Antenna Mounted on a Cubical Conducting Bos," /EEE Trans. Antennas Propagat, vol ‘AP-35, July 1987, pp. 756-762, ‘A.W. C. Chu, S.A, Long, and D. R. Wilton, “The Radiation Pattern of a Monopole ‘Antenna Attached toa Conducting Box,” JEEE Trans. Antennas Propagat, vol. AP-38, December 1990, pp. 1907~1912,Chapter 5 Loop Antennas Glenn S. Smith Georgia Institute of Technology ‘51 Introduction 5:2 $2 Electrically Small Loops 5-2 Traneating Loop 5.2 Receiving Loop 5-4 Ferrte-Loaded Receiving Loop $6 $3 Electrically Large Loops 5.9 CCireular-Loop Antenna 5-9 Resonant Circular Loop 5.13 Circular Loop with Planar Reflector Coaxial Arrays of Cieular Loops § $4 Shielded-Loop Antenna 5-19 SS Additonal Topics $215-1_ INTRODUCTION ‘The single-turn loop antenna is a metallic conductor bent into the shape of a closed ccurve, such asa circle or a square, with a gap in the conductor to form the terminals. ‘A multiturn loop or col is a series connection of overlaying turns. The loop is one of the primary antenna structures; its use as a receiving antenna dates back tothe early experiments of Hertz on the propagation of electromagnetic waves.! ‘The discussion of loop antennas is conveniently divided according to electrical size, Electrically small loops, those whose total conductor length is small compared with the wavelength in free space, are the most frequently encountered in practice. For example, they are commonly used as receiving antennas with portable radios, as directional antennas for radio-wave navigation, and as probes with feld-strength ‘meters. Electrically larger loops, particularly those near resonant size (circumference ‘of loop/wavelength ~ 1), are used mainly as elements in directional arrays. ‘The following symbols are used throughout the chapter ‘= wavelength in free space atthe frequency f= o/x, where the complex har- monic timedependence exp (ja) is assumed 8 = 2x/A = propagation constant in free space T= Vials = wave impedance offre space (~ 3779) 2 = mean radius ofa circular loop of mean side length of «square lop 4 = radius of loop conductor (All results presented are for thinwire loops, a/b < y A= area of loop WV = aumber of turns 4, = length of solenoidal ait §-2_ ELECTRICALLY SMALL LOOPS ‘The axial current distribution in an electrically small loop is assumed tobe uni that i, the current has the same value fat any pont along the conductor. For single- tum loops and multiturn loops that are single-layer solenoidal cuils, measurements suggest that ths is a good aseumption provided the total length of cbe conductor (NV X circumference) is small compared with the wavelength in free space, typically 'S0.1A, and the length-to-dameter ratio forthe solenoidal col is greater than about 3(€./26 3.0) With a uniform current assumed, the electrically small loop antenna is simply analyzed asa radiating inductor? ‘Transmitting Loop ‘The electromagnetic feld of an electrically small loop antenna is the same as that of ‘a magnetic dipole with moment m = uN: a= H2(1—Z)eras en 52\somuma a4 Jers (say #606 8 (5-3) wbere the plane of the loop is normal to the polar axis of the spherical coordinate system (6, #) centered atthe loop as shown in Fig. 5-1. In the far zone ofthe lop im fr ~* co), only the leading terms in Eqs (5-1) and (5-2) are significant, and the a he , : i OSes] FIG. 5-1. Loopantenna and accompanying FIG. 5-2 Far-zone vertical-plane Feld pat ‘pherical coordinate syatem. tern ofan elecrcally small oop, field pattern for both £, and B, in the vertical plane is the simple figure eight shown in Fig. 5:2. ‘The driving-point voltage and current are related through the input impedance ‘of the loop, V = Zl, For electrically small loops, the impedance isthe series combi. nation of the reactance of the external inductance Z* with the radiation resistance and the internal impedance of the conductor Z' =! + jul!: ZeRFZ + jolt + Rt jl +L) (5-4) {In the equivalent circuit for the small loop, a lumped capacitance C is sometimes laced in paraffel with Z to account for the distributed capacitance between the sides of a single turn and between the turns ofa solenoid, as shown in Fig, 53. This capace ‘tance is omitted here, since in practice a variable capacitance is usually placed in Parallel with the loop to tune out its inductance; the capacitance of the loop simply Aecreases the value of the parallel capacitance needed. Note that a loop with a truly uniform current distribution would have ro capacitance, since from the equation Of continuity there would be no charge along the conductor of the loop. ee fa The radiation resistance of the senall loops propariona to the square of tent the product of the area and the number of turns: FIG. 63 Ezuivaont cout for input im £ » ‘dance Zo an electrically smal oop. Rey (55)5-4 Types and Design Methods For single-turn loops and solenoidal coils whose turns are not too closely spaced, the internal impedance is approximately Z' = 2 X total length of conductor (5-6) where 7 is the internal impedance per unit length of a straight conductor with the same cross section as the loop conductor." Ifthe turns of the coil are closely spaced, the proximity effect must also be included in determining Z'* The external inductance is determined from one ofthe many formulas available for the inductance of coils* For a singleturn circular loop Lt = gbtin (88/a) ~ 2] (5-7) ‘and for a single-turn square loop z= Be cn copa) ~ 0-714 (58) ‘The external inductance of a tightly wound single-layer solenoidal col of length & and a radius 6 is often approximated by Lorenz's formula for the inductance of @ circumferentially directed current sheet. Numerical results from ths formula can be pt ina form convenient for application: Lr KugNA/t (5-8) where the factor K, known as Nagaoka’s constant, is shown as a function ofthe ratio £126 (length of the col to the diameter) in Fig, 5-4. Note that, for a long coil (24/ 2b > 1), K~ 1. The use of Eq. (5-9) assumes that the turs a the col are 50 closely spaced that the winding piteh and insulation on the conductors can be ignored; if highly accurate calculations of L*are necessary, corrections for these factors are avail- able in the literature.® Receiving Loop ‘When the electrically small loop is used as a receiving antenna, the voltage developed at its open-ircuited terminals Vo is proportional to the component ofthe incident ‘magnetic flux density normal tothe plane of the loop Bt: Vag = jeNAB, (5-10) ‘where the incident field is assumed to be uniform over the area of the loop. This simple relation between Vac and Bf makes the small loop useful a a probe for measuring the ‘magnetic flux density. Ifa relation between the incident electric and magnetic fields at the center of the loop is known, Voc can be expressed in terms of the magnitude of the incident electric field £' and an effective height h,. This i the case for an incident plane wave with the wave vector k; and the orientation shown in Fig. 5-5: Voc = jwNAB" cos yy sin = hte 9)E (511) where HAUn 9) Voc! E! = j8NA cos Hsin 6, (5-12)Loop Antennas 55 LT I 10 Tl os 7 oa l ° ul ill os 1 10 "60 es FIG $4 Nagaoka’ constant Kor solenoidal cola unton ot ‘the col length tothe diameter, £4/26, : OO Me ono FIG. 5-5 Plane-wave feld incident on re- FIG. 5-8 Théverin equivalent clout for 2iving loop. the receiving loop, [ihe voltage across an arbitrary load impedance Z, connected to the terminals ofthe Ig? sith input impedance 2 is determined from the Thévenin equivalent circuit in vs. Vi = Voc2iZ + 2d) (5-13)546 Types and Design Methocs Ferrite-Loaded Recelving Loop The open-circuit voltage at the terminals of the electrically small receiving loop can be increased by filing the loop with a core of permeable material, usually a ferite. The effect of the core is to increase the magnetic ux through the area of the loop, as lustrated in Fig. 5-7 for a solenoidal coil with a cylindrical core placed in a uniform axial magnetic field. ‘The ferrite material is characterized by a complex relative initial permeability ir fio = ae — jn and a relative permittivity ¢, = e/eq The material is usually selected to have a loss. ete PT 15/1, which is small at the frequency of opera- tion, and consequently 47 is the analysis except when the power dissipated inthe coeis being calculated The dimentons of the cre ae also assumed tobe small compared with the wavelength in te ferrite Ay \/ Veo prevent internal reso. ‘ances within the core.” The open-circuit voltage for a single-turn loop at the middle of a ferrite cylinder of length 2, and radius 6 is increased by the factor ig over the value for the same Joop in fre spa Vor = jitisaAB, (5-14) COIL IN INGIOENT FIELO COLL WITH FERRITE CORE IN INCIDENT FELO FIG. 5-7 Etectofa cylindrical ferrite core on the mag- netic fux through a solenoidal col “The inl permeate the derivative 8d Late Umit ab Hs ede to zr, Dilsr oss 4 the fetes ignore here a the permitivty is assumed 1 be rel