Naval Undersea Warfare Center Division Newport, Rhode Island

Transcription

1 NUWC-NPT Techncal Report 11, June 2007 Quadrflar Helcal Antenna Array for Lne-of-Sght Communcatons Above the Ocean Surface Mchael J. Josypenko John P. Casey Stephen M. Davs Undersea Warfare Electromagnetc Systems Department A NEWPORT Naval Undersea Warfare Center Dvson Newport, Rhode sland Approved for publc release; dstrbuton s unlmted.

2 PREFACE Ths report was prepared under the NUWC ndependent Appled Research (AR) Program, Project No. G6220 1, "Multpath-Abatng Antenna System for Hgh Data Rate Communcatons wth the Undersea Grd," program manager Rchard B. Phlps (Code 01CTO). The sponsorng actvty s the Offce of Naval Research. The techncal revewer for ths report was Paul Mederos (Code 3413). The authors are grateful to Paul Mederos (Code 3413) for hs techncal gudance and to Donald H. Stenbrecher (Code 3404) for hs suggeston of ths project. The authors also wsh to thank Matthew W. Atwood (Code 3433) for hs desgn of the collar that attached the array to the buoy plate and for hs desgn of the rack that mounted the amplfers and flters to the buoy plate. The authors acknowledge Jeremy M. Smth of Anteon Corp. for hs mechancal assembly of the antenna support structures. Revewed and Approved: 25 June 2007 Gerald M. Exley Head, Undersea Warfare Electromagnetc Systems Department (D

3 REPORT DOCUMENTATON PAGE Form Approved T7OMB No Publc reportng for ths collecton of nformaton s estmated to average hour per response, ncludng the tme for revewng nstructons, searchng exstng data sources, gatherng and mantanng the data needed, and completng and revewng the collecton of nformaton. Send comments regardng ths burden estmate or any other aspect of ths collecton of nformaton, ncludng suggestons for reducng ths burden, to Washngton Headquarters Servces, Drectorate for nformaton Operatons and Reports, 1215 Jefferson Davs Hghway, Sute 1204, Arlngton, VA M402, and to the Offce of Management and Budget, Paperwork Reducton Project ( , Washngton DC AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED 25 June TTLE AND SUBTTLE 5. FUNDNG NUMBERS Quadrflar Helcal Antenna Array for Lne-of-Sght Communcatons Above the Ocean Surface 6. AUTHOR(S) Mchael J. Josypenko, John P. Casey, Stephen M. Davs 7. PERFORMNG ORGANZATON NAME(S) AND ADDRESS(ES) 8. PERFORMNG ORGANZATON REPORT NUMBER Naval Undersea Warfare Center Dvson 1176 Howell Street TR 11,820 Newport, R SPONSORNG/MONTORNG AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORNG/MONTORNG Offce of Naval Research AGENCY REPORT NUMBER 875 North Randolph Street Sute 1425 Arlngton VA SUPPLEMENTARY NOTES 12a. DSTRBUTON/AVALABLTY STATEMENT 12b. DSTRBUTON CODE Approved for publc release; dstrbuton s unlmted. 13. ABSTRACT (Maxmum 200 words) Ths report descrbes the development of a lnear array consstng of four quadrflar helcal antennas (QHAs) that s mounted on a buoy to support transmt and receve communcatons at 2.45 GHz along lne-of-sght (LOS) paths above a sea water half-space. The array elements are desgned to be mmune to the multpath nterference created by reflectons of an ncdent sgnal from the surface of the ocean. A novel feed desgn for the array s descrbed. The crcut and radaton characterstcs of each QHA element measured n free space are presented. The predcted radaton pattern of each QHA element at ts nomnal heght above a flat ocean surface s determned from the measured element pattern n free space. 14. SUBJECT TERMS 15. NUMBER OF PAGES Quadrflar Helx Multpath Cut-n Frequency VSWR 76 Pattern Front-to-Back Rato Bflar Helx Turnstle Antenna 16. PRCE CODE 17. SECURTY CLASSFCATON 18. SECURTY CLASSFCATON 19. SECURTY CLASSFCATON 20. LMTATON OF ABSTRACT OF REPORT OF THS PAGE OF ABSTRACT Unclassfed Unclassfed Unclassfed Unclassfed NSN Standard Form 298 (Rev. 2-89) Prescrbed by ANS Std. Z

4 TABLE OF CONTENTS Secton LST O F LLU STRA TO N S... LST O F TA BLES... LST OF ABBREVATONS AND ACRONYMS... Page v v 1 N T R O D U C T O N ANALYSS OF SEVERAL CANDDATE ANTENNAS Turnstle A ntenna A rray Q H A E lem ent DETERMNATON OF OPTMUM QHA ELEMENT DESGN PARAMETERS QHA ARRAY DESGN AND CONSTRUCTON MEASURED CRCUT AND RADATON CHARACTERSTCS OF THE QHA A RRA Y N FREE SPACE Effect of Conductor Wdth on the nput mpedance of a QHA Element Q HA Elem ent- Fnal Desgn Measurement of QHA Elements n the Array Confguraton Effect of Radome on the Crcut and Radaton Characterstcs of the Array Measurement of Array wth Buoy Plate and RF Electroncs Package PREDCTED RADATON CHARACTERSTCS OF THE QHA ARRAY ELEMENTS ABOVE SEA WATER SUMMARY AND CONCLUSONS R E F E R E N C E S... 73

5 LST OF LLUSTRATONS Fgure Page 1-1 Lne-of-Sght Communcatons n an Ocean Envronment Conceptual Vew of Multpath-Abatng Antenna System nstalled on a Buoy Thn-Wre Model of Eleven-Element Turnstle Antenna Array Used for GPS Applcatons as Obtaned from Counselman (Reference 6) Computed Drectve Gan Patterns (db) n the Vertcal Plane of the Counselman Turnstle Array (Fgure 2-1) n Free Space and at Varous Base Heghts Above Sea W ater: (a) M Hz and (b) M Hz Computed Drectve Gan Patterns (db) n the Vertcal Plane at MHz of the Counselman Turnstle Array (Fgure 2-1) Tlted at 20' wth Respect to the Vertcal Drecton n Free Space and at Varous Base Heghts Above Sea Water Q uadrflar H elcal A ntenna Computed Total Gan Patterns (db) n the Vertcal Plane at 2.45 GHz of a Resonant QHA (3/4 Turn, 3/4 X Long, X Radus, Ptch Angle, X Axal Length) n Free Space and at Varous Base Heghts Above Sea Water Computed Gan Patterns (db) n the Vertcal Plane at 2.45 GHz of the Resonant QHA Descrbed n Fgure 2-4 and Tlted wth Respect to the Vertcal Poston n Free Space and at Varous Base Heghts Above Sea Water: (a) Tlt Angle = 200 and (b) T lt A ngle = Thn-Wre Model of the Resonant QHA Descrbed n Fgure 2-5 Above a 36-n. D am eter Crcular G round Plane nput mpedance of One Feed Arm of the QHA Descrbed n Fgure 2-5 at 2.45 GHz as a Functon of Tlt Angle for Varous Base Heghts Above a 36-n. Dameter Crcular Ground Plane: (a) Resstance and (b) Reactance Computed Helcal Length per Turn (n Wavelengths) as a Functon of Ptch Angle of a QHA W here Cut-n Occurs QHA (Element Length = 1.25 X, Radus = X, Ptch Angle = 30', Axal Length = X) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of O n e B flar P ar QHA (Element Length = 1.25 X, Radus = X, Ptch Angle = 400, Axal Length = X) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz, and (b) Smth Chart of the Computed nput mpedance of O n e B f lar P ar... 24

6 LST OF LLUSTRATONS (Cont'd) Fgure Page 3-4 QHA (Element Length = 1.5 X, Radus = X, Ptch Angle = 300, Axal Length = X) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of One Bflar Par QHA (Element Length = 1.5 X, Radus = X, Ptch Angle = 40', Axal Length = X) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of One Bflar Par Constructed QHA: (a) Closeup Vew of Antenna and (b) Antenna Connected to Feed N etw ork Measured Crcut and Radaton Characterstcs of the Constructed QHA Shown n Fgure 3-6: (a) Gan Patterns at 2.45 GHz and (b) nput VSWR of B oth Feed A rm s Smth Charts of the Measured nput mpedances of Both Bflar Pars of the Constructed QHA Shown n Fgure 3-6: (a) Feed 1 and (b) Feed Four-Element Lnear Array of QHA Elements Descrbng (a) nstallaton on a Buoy and (b) Array Geometry wth Dmensons of Array Elements and Ther Locatons Relatve to Buoy Surface and Ar-Sea Water nterface Vertcal-Plane Vew of Feed-Cable Network for the QHA Array End Vew of Bottom Element of the QHA Array Showng Feed Cables Four-Element QHA Array Assembly QHAs of Dfferent Conductor W dths Secton of Brass Tube Assembly and Plastc Spacers Assocated wth a QHA Element QHA wth Quadrature Feed Network of Power Spltters Closeup Vew of Element Feed Regon Closeup Vew of QHA Element Showng Feeds and Spacers QHA Array wth Feed Assembly (a) QHA Array Mounted to Buoy Plate Wthout Radome, (b) QHA Array Mounted to Buoy Plate wth Radome, (c) RF Electroncs Mounted Below Buoy Plate, and (d) Q HA A rray and Buoy... 45

7 LST OF LLUSTRATONS (Cont'd) Fgure Page Planar Vew of the Geometry of a Helcal Conductor of Wdth w Wound at a Ptch Angle a Around a Crcular Cylndrcal Surface VSWR as a Functon of Frequency of a QHA Element wth Enclosed Brass Tube for V arous Conductor W dths U 5-3 Smth Chart of the Measured nput mpedance of One Bflar Par of a QHA Element (Ptch Angle = 300, Dameter = n., Axal Length = 2.71 n.) wth an Enclosed Secton of Brass Tube for Dfferent Conductor Wdths: (a) w = 0.24 n. and (b) w = n QHA Element Measured wth Brass Tube: (a) Gan Patterns n db and (b) VSW R of Both Bflar Feed Arms Smth Chart of the Measured nput mpedance of Both Bflar Feeds of the Fnal QHA Element Desgn (Ptch Angle = 30', Dameter = n., Axal Length = 2.90 n.) wth an Enclosed Secton of Brass Tube Measured Gan Patterns of QHA Array Elements n Free Space: (a) RHCP or Copolarzed Component and (b) LHCP or Cross-Polarzed Component Measured VSWR as a Functon of Frequency of One Bflar Feed of the Two Bottom QHA Array Elements n Free Space Losses Assocated wth Power Spltters, mpedance Msmatch, and Mutual Couplng for the QHA Array Measured n Free Space Measured VSWR as a Functon of Frequency for One Bflar Par of an solated QHA Enclosed Wthn G- 10 Fberglass Tubes of Varous Outer Dameters Effect of G-10 Fberglass Radomes of Varous Dameters on the Measured RHCP Gan Patterns vs Elevaton Angle n Free Space of the QHA Array Elements: (a) Element 1 and (b) Elem ent Effect of G-10 Fberglass Radomes of Varous Dameters on the Measured RHCP Gan Patterns vs Elevaton Angle n Free Space of the QHA Array Elements: (a) Elem ent 3 and (b) Elem ent Measured Gan Patterns of the QHA Array Elements wth 1.75-n. Dameter G-10 Fberglass Radome n Free Space: (a) RHCP or Copolarzed Component and (b) LHCP or Cross-Polarzed Component

8 LST OF LLUSTRATONS (Cont'd) Fgure Page 5-13 Gan Patterns vs Elevaton Angle n Free Space of the QHA Array Elements Measured wth Radome and Mounted to Buoy Plate as Shown n Fgure 4-1 lb: (a) RHCP or Copolarzed Component and (b) LHCP or Cross-Polarzed Component Sphercal Coordnate System wth Assocated Unt Vectors ( P, 0, ) Drectve Gan Patterns vs Elevaton Angle of the QHA Elements Above Sea Water as Determned from the Measured Element Patterns n Free Space Shown n F gure LST OF TABLES Table Page 3-1 Optmum Desgn Parameters of QHA Element Obtaned from Thn-Wre M odel A nalyss Parameters of Selected QHA Element Desgn Desgn Parameters of QHA Element *LST OF ABBREVATONS AND ACRONYMS GPS AR LHCP LOS LPF NEC NUWC QHA RF RHCP VSWR Global Postonng System ndependent Appled Research Left-hand crcularly polarzed Lne-of-sght Low-pass flter Numercal Electromagnetcs Code Naval Undersea Warfare Center Quadrflar helcal antenna Rado frequency Rght-hand crcularly polarzed Voltage standng wave rato v (v blank)

9 QUADRFLAR HELCAL ANTENNA ARRAY FOR LNE-OF-SGHT COMMUNCATONS ABOVE THE OCEAN SURFACE 1. NTRODUCTON Consder an antenna located above the ocean surface that s used for the transmsson and recepton of rado sgnals along lne-of-sght (LOS) paths as shown n fgure 1-1. For smplcty, a flat sea surface s assumed. Sgnal transmsson and recepton occur va two prmary paths,.e., the drect path and the sea water reflected path. Because these two paths are of dfferent lengths, the sgnals assocated wth them arrve wth dfferent phases. The sgnals from the drect and reflected paths mght arrve n phase, resultng n constructve nterference and an ncrease n total sgnal ampltude by as much as 6 db above the free-space value. Alternatvely, the two sgnals may arrve out of phase, resultng n destructve nterference and a decrease n total sgnal ampltude. The reducton n sgnal strength can be very large f the two sgnals are nearly equal n ampltude and 1800 out of phase. Because of the dynamc sea state, the dfference n length of the drect and reflected paths vares wth tme. Therefore, to reduce the probablty of ntermttent loss of sgnal, the antenna system should be desgned to be mmune to the reflected-path sgnal. VrF*t Path 3' * Fgure 1-1. Lne-of-Sght Communcatons n an Ocean Envronment

10 U The objectve of ths project s to develop a short, three- to four-element vertcally-stacked antenna array and array-sgnal processor system that s nstalled on a buoy and s capable of supportng the transmsson and recepton of nformaton at hgh data rates along LOS paths above an ar-sea water nterface. The operatng frequency chosen for ths antenna system s 2.45 GHz (reference 1) wth a narrow nstantaneous bandwdth. Over ths frequency band, the deal 3 array elements should be crcularly polarzed wth hemsphercal radaton patterns n the upper half-space. n addton, each element should have sgnfcantly less gan n the lower half-space 3 to mnmze the ampltude of the reflected sgnal. The antenna elements should not couple sgnfcantly to sea water so that the mpedance and radaton characterstcs of the array are farly mmune to buoy moton and sea state. The array must attach to a mast that extends above the buoy (fgure 1-2), have a length of no greater than 18 n., and should be able to wthstand buoy tlt up to approxmately 200 from the vertcal poston. To reduce the probablty of washover, the array s separated from the top surface of the buoy by approxmately 12 n. 3 Several antennas were nvestgated to determne f they can meet the above requrements. ANTENNAU., ARRAY MAST BUOY

11 3- compared. A detaled descrpton of the chosen antenna desgn and the measured crcut and n ths report, the results of an analyss of two canddate antenna arrays are presented and radaton characterstcs of each array element are presented. A descrpton and analyss of the entre antenna system s gven n another report by the authors (reference 1).!! 3 (4 blank)

12 - 2. ANALYSS OF SEVERAL CANDDATE ANTENNAS At the outset of ths project, several possble antenna desgns were consdered that mght 3 provde both hemsphercal coverage and mmunty wth respect to the sgnal reflected from the ocean surface. Although a gmbaled reflector antenna system would be a sutable canddate for 3 ths applcaton, the hgh cost assocated wth the gmbal system s prohbtve. Two canddate systems consdered n ths nvestgaton were a turnstle array and a quadrflar helcal 5antenna (QHA) array. Ths secton provdes a summary of these nvestgatons. 2.1 TURNSTLE ANTENNA ARRAY -- The turnstle antenna (references 2 and 3) conssts of two center-fed dpoles that are perpendcular to each other, wth ther axes ntersectng at ther mdponts. The dpoles are fed out of phase wth equal exctaton voltages, resultng n a nearly omndrectonal radaton pattern wth crcular polarzaton n the drecton perpendcular to the plane of the dpoles. Consequently, the dpole current dstrbutons are equal n magntude and 900 out of phase wth respect to each other. Counselman (reference 4) has developed several turnstle antenna array desgns for the L 3and L2 Global Postonng System (GPS) frequency bands that have a farly constant gan n the upper hemsphere and sgnfcantly less gan n the lower half-space. Counselman obtaned ths performance through a careful desgn of the array feed network and an optmum choce of element exctatons or weghts. One of Counselman's turnstle antenna array desgns, consstng 3of 1 elements (separate elements used at each GPS frequency), was analyzed through use of the Numercal Electromagnetcs Code (NEC), Verson 4.1 (reference 5). The NEC model used n 3 the analyss of ths array was obtaned from Counselman (reference 6). Fgure 2-1 llustrates a thn-wre model of an 11-element, dual-band turnstle antenna array obtaned from Counselman (reference 6). n ths desgn, fve turnstle elements are used n the 3 L1 GPS band (center frequency = MHz) and sx turnstle elements are used n the L2 GPS band (center frequency = MHz). The element locatons and weghts are gven n

13 table 2 of reference 4. n ths desgn, each turnstle antenna element conssts of four coplanar and mutually perpendcular monopoles that are fed n tme-phase quadrature, a desgn that s equvalent to the two-dpole turnstle dscussed earler. Each monopole extends horzontally from a center post, whch both mechancally supports the monopoles and contans the transmsson lnes that feed them. The center post extends approxmately n. long and ncludes four 1.0-n. wde radal blade conductors that ntersect along the array axs at rght 3 angles. The radal blades serve as ground planes for the monopole elements. n the thn-wre model of the turnstle array, the radal blades are represented by wre grds (reference 6). -Center Post A---' & Ground Planes (Wre Grd Model) Antennas -- 4 Fgure 2-1. Thn-Wre Model of Eleven-Element Turnstle Antenna Array Used for GPS Applcatons as Obtaned from Counselman (Reference 6) The turnstle array n fgure 2-1 uses separate arrays for the L and L2 GPS bands because of the dfference n the wavelengths at these frequences. The electrcal length of each array s 2k, where the physcal lengths of the Ll and L2 arrays are 0.38 m and m, respectvely. n addton, the elements n each array are unformly spaced by k/3 at ther respectve center frequences. The phase center of each turnstle element s concdent wth ts geometrcal center, and the phase centers of the two arrays are located at the same pont. 6

14 *sea Fgure 2-2 shows the computed drectve gan patterns of the Counselman turnstle array n fgure 2-1 at the L and L2 band center frequences n free space and at varous heghts h above sea water. The free-space patterns are approxmately hemsphercal overhead wth a maxmum gan n the axal drecton. n addton, the free-space patterns show very lttle radaton n the lower halfspace. Consequently, the radaton pattern n the upper half-space should change very lttle when the array s located above sea water. n fgure 2-2, the radaton patterns of the turnstle array above water are nearly dentcal to the free-space pattern except at small elevaton angles where multpath nterference of as much as 3 db s observed. The small amount of multpath nterference observed on these plots suggests that Counselman's array s mmune to multpath nterference. - Because the array s to be nstalled on a buoy that floats above a dynamc sea surface (fgure 1-1), the effect of buoy tlt on the radaton pattern must be examned. Fgure 2-3 gves the computed drectve gan patterns of the Counselman turnstle array tlted at 200 wth respect to the vertcal drecton at the L 1 band center frequency n free space and at varous heghts h above sea water. The tlt of the antenna results n a sgnfcant ncrease n the free-space gan n the lower half-space over a 20' regon extendng from 90' to 1100 and a notceable decrease n the free-space gan over the regon extendng from 2700 to 2900 as ndcated n the free-space pattern n fgure 2-3. Therefore, the Counselman array produces a large pattern front-to-back rato (.e., the rato of the total gan along the forward and backward axal drectons) only when t s orented vertcally. n fgure 2-3, the patterns over sea water show an overall reducton n gan as well as substantal multpath nterference at the lower elevaton angles, especally over 3 the regon extendng from 2700 to 2900 (opposte to the drecton of tlt) where the drect feld ampltude has sgnfcantly decreased and the reflected feld ampltude has ncreased. 7

15 (a) Free space h n h 9 n, h = 1 8 n " (b) (- Free space h = 0.1 n h-9 n. h = 18 n Fgure 2-2. Computed Drectve Gan Patterns (db) n the Vertcal Plane of the Counselman Turnstle Array (Fgure 2-1) n Free Space and at Varous Base Heghts Above Sea Water: (a) MHz and (b) MHz 8

16 Free space h = O.S n h-9n h - 18 n. 300 " Fgure 2-3. Computed Drectve Gan Patterns (db) n the Vertcal Plane at MHz of the Counselman Turnstle Array (Fgure 2-1) Tlted at 20' wth Respect to the Vertcal Drecton n Free Space and at Varous Base Heghts Above Sea Water 5 as the base heght above sea water ncreases. f the tlt angle of the array s ncreased, the n fgure 2-3, note that the null depths assocated wth the multpath nterference become deeper multpath nterference wll also ntensfy. Results smlar to those n fgure 2-3 were obtaned at *] the L2 band center frequency. The analyss of the turnstle array revealed the followng problems: - 1. The turnstle elements produce consderable radaton ncdent nto the lower half-space. Therefore, the radaton pattern of the turnstle array s very senstve to the array orentaton relatve to the ground plane. The large pattern front-to-back rato observed for the Counselman array when orented vertcally wth respect to the ocean surface s attrbuted to an optmum weghtng of the turnstle elements. Any varaton n the orentaton of the array wth respect to the sea water surface wll produce a change n the element radaton characterstcs wth 1 9

17 ncreased multpath nterference. As a result, unless the element weghts are modfed dynamcally to account for antenna movement assocated wth sea state, the turnstle array s qute susceptble to multpath nterference. 2. A dpole antenna generally couples to nearby structures that are located close to ts axs. Therefore, f the turnstle array s close to the sea surface, the elements located closest to the sea surface (.e., a fracton of a wavelength) wll couple most sgnfcantly to the sea water. Ths couplng wll affect both the mpedance and radaton characterstcs of the antenna elements, makng t dffcult to determne approprate element weghts. n concluson, the Counselman turnstle array performs poorly when placed above sea water because t s optmzed only for the case where t s n the vertcal orentaton wth respect to the ocean surface. Because the multpath-abatng antenna wll often be tlted when placed n a buoy, the turnstle array was rejected as a possble canddate desgn. 2.2 QHA ELEMENT 3 An antenna element that s better suted for the current applcaton because t can be desgned to have a radaton pattern wth a large front-to-back rato s a QHA, or volute, as shown n fgure 2-4. A QHA conssts of four dentcal helces, each spaced 900 apart and fed wth voltages of equal ampltude and n tme-phase quadrature. Equvalently, a QHA conssts of two dentcal bflar helces that are orthogonal and excted n tme-phase quadrature. The QHA behaves as an nterleaved endfre array wth a largely hemsphercal overhead, undrectonal pattern and crcular polarzaton. As a result, ths antenna s somewhat mmune to reflectons when located above a ground plane. n addton, a bflar helx s a predomnantly closed antenna, wth ts nterleaved elements hghly coupled to each other. Ths self-couplng reduces 5 ts senstvty to the exstence of nearby structures, especally those located along the helcal axs. An overvew of the crcut and radaton characterstcs of the volute wth desgn nformaton has been presented n several papers by Klgus (references 7 and 8). 10

18 , ~HELCAL. ELEMENT - DAMETER f,-4 AXAL LENGTH.4 ; /\ / t, QUADRATURE FEED - PTCH -RADUS ANGLE. T- Fgure 2-4. Quadrflar Helcal Antenna (%-turn and open-ended) Narrow, resonant QHAs n the range of 0.1 to 0.2 k n dameter and wth element lengths of less than 1 X possess large pattern front-to-back ratos on the order of 15 to 20 db. As an example, consder a 3 /4-turn, 3/4-X long open-crcuted QHA wth a radus of k. Ths U11 antenna has a ptch angle of 66.1' and an axal length of k. Ths QHA has a rather broad overhead pattern wth an mpedance bandwdth of 12.5% and a resonant mpedance of approxmately 100 ohms, whch can be easly matched to a 50-ohm transmsson lne wth a 1800 power dvder.

19 The narrow QHA descrbed above was modeled (wth NEC 4. 1) both n free space and above a sea water half-space, at base heghts h rangng from just above 0 n. to 27 n. above sea water (n 9-n. ncrements) and vertcal tlt angles rangng from 0' to 900 (n 10' ncrements). Computed patterns of the total gan of the antenna n free space and at varous heghts above sea water at a frequency of 2.45 GHz are gven n fgure 2-5. The patterns show only a small amount of multpath nterference that s most notceable at low elevaton angles and for large antenna heghts above the sea water. Fgure 2-5 shows that the maxmum loss n gan (from the free-space value) due to multpath nterference s approxmately 4 db and occurs when h = 27 n. (= 5.60 X). Overall, the narrow QHA has good mmunty wth respect to multpath nterference. Free space h =.065 n. h=9 n h= 18 n.... h = 27 n. 32o0,' ~ ~ * ~ ,,*-' 270 ' -" U Fgure 2-5. Computed Total Gan Patterns (db) n the Vertcal Plane at 2.45 GHz of a Resonant QHA (3/4 Turn, 3/4 A Long, A Radus, Ptch Angle, A Axal Length) n Free Space and at Varous Base Heghts Above Sea Water 12 3

20 = Computed radaton patterns of the total gan of the narrow QHA orented at tlt angles of 200 and 400 (wth respect to the vertcal) above sea water at 2.45 GHz are shown n fgure 2-6. A comparson of the QHA pattern n fgure 2-6a wth the turnstle array pattern n fgures 2-3 for the 20' tlt case ndcates that the QHA element possesses much greater mmunty wth respect to multpath nterference than the turnstle array. t should be noted that the turnstle array patterns n fgure 2-3 were computed at a lower frequency ( GHz) than the operatng frequency (2.45 GHz) used n the evaluaton of the QHA patterns n fgure 2-6. Therefore, the heghts above sea water plotted n fgure 2-3 are actually electrcally shorter than the correspondng heghts plotted n fgure 2-6. n partcular, h = 18 n. n fgure 2-3 s equvalent to h/2 = 2.4, and h = 18 n. and h = 27 n. n fgure 2-6 are equvalent to h/2 = 3.73 and h/2 = 5.60, respectvely. Consequently, the patterns of the tlted turnstle array n fgure 2-3 would have to be rased to greater heghts above sea water to be compared wth the QHA patterns n fgure 2-6a, resultng n addtonal multpath nterference n the turnstle array. A comparson of fgures 2-6a and 2-6b ndcates that as the tlt angle of the QHA s ncreased from 200 to 40', only a small amount of addtonal multpath nterference results. Therefore, the radaton feld of the narrow QHA should have good mmunty wth respect to multpath nterference when placed above a dynamc sea surface. The above results show that one narrow QHA has superor pattern performance above a dynamc sea surface than Counselman's fve-element turnstle array. t s antcpated that an array of QHAs wll provde both the gan requrements and the necessary pattern mmunty wth respect to multpath nterference when located above sea water. Therefore, the QHA appears to be a more sutable antenna element for ths applcaton. 1 a! * 13

21 -Free space h n. h -9 n h = Sn ; (b) Free space 0~ - 9-ln h=.065 n h 10 n

22 n addton to the radaton pattern performance of an antenna element n the presence of a dynamc sea surface, t s also mportant to consder the mpedance propertes of the antenna under these condtons. n partcular, the antenna should not couple (electromagnetcally) to the sea surface such that ts mpedance vares wth tme under varous sea-state condtons. Equvalently, the nput mpedance of each antenna element should be ndependent of tlt angle over the range of buoy moton that s antcpated. f the element mpedances vary sgnfcantly wth orentaton above the ocean surface, the array wll experence losses due to mpedance msmatch and wll create addtonal challenges n the beamformer desgn. Therefore, t s mportant that each array element has stable mpedance characterstcs n a dynamc ocean envronment. To evaluate the effect of tlt angle on the nput mpedance of the 3/4-X long QHA descrbed n fgure 2-5, consder a thn-wre model of the antenna above a 36-n. dameter crcular ground plane llustrated n fgure 2-7. The ground plane, modeled as a wre mesh, extends across the top surface of the buoy. The ground plane was added to the model to provde a more stable reflectng surface, at least for hgher elevaton angles of ncdence, resultng n possbly more stable mpedance characterstcs. Although the ground plane was ncluded n ths prelmnary analyss, t was later found to be unnecessary and was elmnated. n the analyss descrbed below, the ground plane s placed at 9 n. above sea water (correspondng to the nomnal heght of the top surface of the buoy above the sea water) and the axal heght h from the base of the helx to the crcular ground plane s vared. n addton, the QHA and ground plane were tlted from the uprght vertcal poston for angles rangng from 0 to 400. All calculatons were performed at 2.45 GHz usng NEC. 1 * 15

23 -4 Fgure 2-7. Thn- Wre Model of the Resonant QHA Descrbed n Fgure 2-5 Above a 36-n. Dameter Crcular Ground Plane Fgures 2-8a and 2-8b show plots of the nput resstance and reactance, respectvely, of one feed of the QHA as a functon of tlt angle for varous antenna base heghts h above the crcular ground plane. The plots show that the mpedance s senstve to tlt angle only when t s close to the ground plane. At a 6-n. heght (h = 1.24 X) above the ground plane, only the nput reactance shows a slght varaton wth tlt angle. For heghts of 12 n. or greater above the ground plane (.e., h > ), the mpedance shows lttle varaton. Therefore, n the desgn of a lnear array of QHAs, the bottom element should be placed at least - 2 k. above the ground plane to have neglgble couplng to the sea water. Note that ths analyss gnores the effect of element mutual couplng n the array. 16

24 (a) 30 h 0 h 6n.... h 12 n h=18 n. - -h =24 n. * E r (b) 10h= - Tlt Angle (degrees) h=6n.... h... =12 n h= 18 n. - -h = 24 n * TtAgeEdges Fue28nu meac foefe mo h H ece nfgr - tz4 ~ as 0 ucno tageo au aehhsaoea3-n mtrcrua GrudPaea etac n b ecac 17(15lak

25 U N 3. DETERMNATON OF OPTMUM QHA ELEMENT DESGN PARAMETERS n the desgn of a lnear array consstng of four QHA elements, 16 coaxal cables are requred to feed the array. t was determned that the feed cables must ft wthn the nteror of the array and travel along a drecton parallel to the helcal axs before connectng to the 3 approprate helcal element. n addton, the feed cables must ft wthn a metallc tube or condut to sheld the cables from the radatng elements. Despte ts desrable propertes that 3 nclude a radaton pattern wth a large front-to-back rato, wde beamwdth, and an acceptable bandwdth and matchable nput mpedance, the dameter (0.119 X) of the 3 /4-turn, 3 / 4 -k long open-crcuted QHA studed n the prevous secton was found to be too narrow to allow passage of the 16 feed cables. *degrade To mantan the geometrcal symmetry of each QHA element, the axs of the metallc tube must be concdent wth the array axs. The tube dameter must be made as small as possble to mantan a symmetrcal arrangement of feed cables and QHA arms. A large tube dameter wll the element radaton patterns by ncreasng the radaton along the backsde drecton. n addton, a large tube dameter wll decrease the mpedance bandwdth of the QHA. Therefore, the metallc feed tube must be made as narrow as possble to preserve the desrable crcut and radaton characterstcs observed for the QHA n the prevous secton. A large-dameter QHA generally has a wder mpedance bandwdth than a narrow QHA. A QHA wth a ptch angle less than 500 and a total element length of at least about 1.5 X operates above the "cut-n" frequency when the axal length per turn s approxmately 0.5 X. Ths cut-n frequency s defned as the frequency above whch the QHA mpedance has lttle varaton and becomes manly resstve; resultng n a low voltage standng wave rato (VSWR) and, therefore, the QHA operates as a wde-bandwdth antenna. Note that a cut-n frequency does not exst for all QHA desgns; therefore, a QHA operatng above the cut-n frequency does not have to operate at resstve resonance to match ts mpedance. The followng are some dsadvantages of a large-dameter QHA: * 19

26 1. A QHA wth a dameter greater than 0.2 X tends to produce more backsde radaton that s attrbuted to the ncreased radaton from ts wder radal feed sectons. 2. Although a QHA produces mnmum backsde radaton at the cut-n frequency, the backsde radaton ncreases wth frequency above the cut-n value. 3. The larger radal feed sectons assocated wth a large-dameter QHA wll tend to couple more to the sea water below t. t should be noted that the axal length per turn of a QHA at whch pont the cut-n frequency occurs s dependent on the helcal ptch angle. Fgure 3-1 s a plot of the computed axal length per turn as a functon of ptch angle where the cut-n frequency occurs. The plot was produced from a NEC model (reference 5) of the QHA. The plot shows that at a ptch angle of 400 the cut-n frequency occurs at an axal length per turn of 0.5 X Ptch Angle (degrees) Fgure 3-1. Computed Helcal Length per Turn (n Wavelengths) as a Functon of Ptch Angle of a QHA Where Cut-n Occurs 20

27 U A parametrc study was performed to determne the optmum desgn parameters for the QHA element. The nvestgaton was carred out wth a thn-wre model of a forward-fre, open-crcuted QHA va the NEC code. The QHA parameters that were vared were the helcal radus, element length, and ptch angle. ntally, the axal length of the QHA was vared rather than the helcal element length because of the spacng requrements between the array elements. However, because the radaton patterns are largely governed by element length, the parametrc runs were performed usng element length as a varable nstead of the axal length. The axal 3] length of the QHA was montored by vsual nspecton of the model where necessary. n the parametrc nvestgaton, the crcumference was vared from 0.1 k to 1.0 k n ncrements of 0.1 X 3(equvalent to varyng the helcal radus from X to k n ncrements of X), the helcal element length was vared from k to 2.0 X n ncrements of X, and the ptch 3 angle was vared from 100 to 900 n ncrements of 100. n all, a total of 1440 antennas were modeled. n the parametrc study, the pattern beamwdth and front-to-back rato were used to 3determne the optmum QHA element desgn for the array. Acceptable pattern front-to-back ratos, on the order of db, were found to correspond to QHA desgns wth ptch angles between 300 and 500, a helcal element length of X, and a helcal radus of k. Table 3-1 lsts four QHA desgns that fell wthn these ranges. Note that the helcal rad lsted n table 3-1 are approxmately double the sze of the QHA nvestgated n secton 2. Table 3-1. Optmum Desgn Parameters of QHA Element Obtaned from Thn- Wre Model Analyss Element Ptch Helx Helx Axal Axal Pattern Front- Length Angle Radus Radus Length Length to-back Rato W (deg.) (n.) X) (n.) (db) Cauton must be taken n usng the pattern front-to-back rato as the sole parameter for determnng the optmum QHA desgn because only the overhead and backsde (.e., axal) gans * 21

28 are nvolved. Therefore, the patterns of the total gan n free space should be studed for each 1 case. For example, a QHA wth a 60' ptch angle and helcal radus of k has a pattern front-to-back rato of 19 db, but the man overhead beam s splt nto two beams and, therefore, wll not provde sutable hemsphercal coverage. t was found that to radate a sngle overhead beam, the QHA must have a ptch angle of 40' or less. The computed free-space radaton patterns and Smth charts of the nput mpedance for each of the QHA desgns lsted n table 3-1 are shown n fgures 3-2 to 3-5. A comparson of the total gan patterns shows that fgures 3-3a and 3-5a exhbt a wder beamwdth than those n fgures 3-2a and 3-4a. Ths s a desrable feature and results from the antennas havng a larger ptch angle. However, as ndcated n table 3-1, because the axal length of the QHA ncreases wth ptch angle, the spacng between array elements must also ncrease. Too large of an element spacng wll result n the appearance of gratng lobes n the array radaton pattern. Consequently, to elmnate the exstence of gratng lobes n the radaton pattern of the array, an upper lmt must 3 be mposed on the ptch angle to keep the spacng between elements suffcently small. The maxmum allowable spacng between array elements was set at /8. 3 The normalzed mpedance loc for each QHA desgn n table 3-1 are llustrated n fgures 3-2b to 3-5b. n these Smth charts, the mpedances are normalzed by 100 ohms. The mpedance plotted corresponds to that across two helcal arms, or equvalently, one bflar par of the QHA. The mpedance plots show that the characterstc mpedance Z 0 of each QHA s approxmately 300 ohms. The characterstc mpedance of a QHA s the resstance that the mpedance locus approaches as the frequency ncreases toward nfnty, or equvalently, s the resstance that the mpedance approaches above the cut-n frequency. Because the QHAs were modeled n NEC wth thn-wre helcal elements, the characterstc mpedance depends prmarly on the wre dameter, and to a lesser extent, the ptch angle (note that Z 0 = 3 L where L s 2 C' the nductance per unt length of helcal element and C s the capactance per unt length between the two helcal elements of a bflar helx). 22

29 (a) , ' (b) Fgure 3-2. QHA (Element Length = 1.25 A, Radus = A, Ptch Angle = 300P Axal Length A n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of One Bflar Par * 23

30 U (a) U "-30 " U 270 " (b) &0 5 0, Fgure 3-3. QHA (Element Length = 1.25 A, Radus = Ptch Angle = 400, Axal Length = A) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of One Bflar Par 24

31 (a) (b) " l Fgure 3-4. QHA (Element Length= 1.5 A, Radus = A, Ptch Angle = 30', Axal Length = A) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of One Bflar Par 5 25

32 m (a) (b) 1.0 ( ,o 0.2O Fgure 3-5. QHA (Element Length = 1.5 A, Radus = A, Ptch Angle = 40, Axal Length = A) n Free Space: (a) Computed Drectve Gan Pattern (db) at 2.45 GHz and (b) Smth Chart of the Computed nput mpedance of One Bflar Par 26

33 Thn wres have large L and small C and, thus, Z o s large. Wde elements have small L and 3 large C, and thus Z 0 s small. f the helcal wres are replaced wth wde strp elements (that comprse a large percentage of the avalable surface area), Z 0 can reduce to approxmately ohms, thus allowng the QHA to be easly matched to a 1800 power spltter. Strp helcal 1 elements of suffcent wdth wll tend to mprove the convergence of the mpedance locus about Z 0 over that of the thn helcal wres n the NEC model due to wder elements wth less nductance and a lttle more bandwdth. t s also antcpated that the ntroducton of a metal tube 3placed along the QHA axs that contans the feed cables wll produce a small change n the mpedance (.e., a decrease n Z 0 because the capactance per unt length of element ncreases) ft and a mnor reducton n mpedance bandwdth. U To avod the occurrence of gratng lobes n the array radaton pattern and allow for a QHA element of suffcently long axal length, the element separaton was set to 0.75 k, equvalent to 3.61 n. at 2.45 GHz. Ths separaton permts a QHA element desgn wth a suffcently wde mpedance bandwdth and a free-space radaton pattern wth only a small amount of backsde radaton. n addton, suffcent spacng between array elements must be ncluded to reduce the nterelement mutual couplng. These constrants resulted n the selecton of a QHA element desgn wth the parameters lsted n table 3-2. The radaton pattern and normalzed mpedance locus of the selected element desgn n free space are llustrated n fgures 3-2a and 3-2b, respectvely. 3 Table 3-2. Parameters of Selected QHA Element Desgn Parameter Value Ptch Angle 300 Total Element Length 6.03 n. (1.25 k) Axal Length 2.71 n. (0.562?,) Helcal Radus n. (0.127 k) 3 A QHA element wth the desgn parameters gven above was constructed and measured to expermentally verfy the crcut and radaton characterstcs that were predcted by the NEC model. Fgure 3-6 shows photos of the antenna and assocated feed network. The helcal conductors each consst of 3-ml-thck copper tape of wdth n. that s bonded to a 3-mlthck Mylar sheet. The edges of the copper tape were cut by placng the copper-covered sheet nto a mechancal plotter and usng a damond scrbe to cut the edges. 5 27

34 (a) (b 900 PUTTR 11Z 180 SPTTERANTENN All, QUADRAT 900~~~~FE PPOEN10TSLTTR Fgure 3-6. Constructed QHA: (a) Closeup Vew of Antenna and (b) Antenna Connected to Feed Network 28a

35 The quadrature feed pont for the antenna was made by solderng together the outer conductors of a square bundle of four n. outer dameter semrgd cables, and the center conductors became the feeds for the four helcal elements. Because the center conductors of the feed cables were physcally dsplaced from each other, the QHA was not fed from a sngle pont, but nstead from ponts slghtly dsplaced (radally) from the axs, as shown n fgure 3-6. (As wll be shown n the followng secton, the concept of "radally dsplaced feed ponts" occurred once the metal tube was nserted along the axs of the QHA array. For the constructed QHA n fgure 3-6, the 5feed ponts are only slghtly dsplaced from the QHA axs, resultng n small changes n ts mpedance. Once the tube s nserted, the dsplacement of the feed ponts becomes much larger wth more notceable changes n the mpedance.) As a result, t was antcpated that the measured mpedances would not exactly match those predcted n fgure 3-2b. The mpedance plot n fgure 3-2b s based on a NEC model of the selected QHA element desgn wth thn-wre conductng elements. n contrast, the constructed antenna n fgure 3-6 has wde strp conductors and, therefore, has a smaller measured mpedance than that computed n fgure 3-2b. The desred characterstc mpedance Z 0 of the QHA element s 100 ohms as compared wth the predcted value of 300 ohms n the NEC model. Although the thn-wre model of the QHA predcted much hgher mpedances than that of the constructed QHA wth wde strp conductng elements, the computed VSWR should be close to the measured VSWR. 5n addton, a QHA wth thn-wre conductors wll produce less backsde radaton than a QHA wth thck conductors. The QHA shown n fgure 3-6 was constructed wth wde strps to obtan 3 a characterstc mpedance at the target value of 100 ohms. 3The constructon of the antenna requred lnng up the radal sectons of the helcal elements (etched on a Mylar sheet) to the correspondng crcumferental sectons and solderng the boundary between each of them. Ths process can lead to physcal asymmetres between the elements producng asymmetres n the radaton pattern, some of whch may be seen n the measured patterns. The helcal conductors deally occupy 97.5% of the avalable surface area along the helcal crcumference n order to produce an mpedance across each bflar helcal par that s close to 29

36 100 ohms. Because of constructon lmtatons n the attempt to create small slots between the conductng elements, the amount of avalable surface area occuped by the helcal conductors was reduced to 90%, resultng n a small ncrease n the characterstc mpedance. t should be recalled that a QHA conssts of two orthogonal bflar helces that are fed 900 out of phase. The feed network n fgure 3-6b provdes a quadrature feed for the antenna. The four coaxal cables at the bottom of the antenna are connected to the output of a par of 1800 power spltters that are 3 n turn connected to the output of a 90' power spltter. n the receve mode, the sgnal power from the antenna appears at the nput to the 90' power spltter. Fgure 3-7a shows plots of the measured gan as a functon of elevaton angle at 2.45 GHz for the QHA n fgure 3-6. The gan pattern plots are shown for vertcal, horzontal, and rghthand crcular polarzatons. The measured gan patterns show some asymmetry that s attrbuted 3 to asymmetres n the feed network and n the mechancal constructon of the antenna. These asymmetres are not observed n the computed patterns shown n fgure 3-2a. Note that fgure 3-2a s a plot of the total antenna gan, whch s a combnaton of both the vertcal and horzontal polarzatons. From fgure 3-7a, the measured pattern front-to-back rato of the QHA s approxmately 10 db as compared wth the computed value of 15 db observed n fgure 3-2a. The 5-dB reducton n the measured pattern front-to-back rato s attrbuted to the ncreased 3 backsde radaton caused by the antenna asymmetres and by the use of the wde helcal conductng elements. Fgure 3-7b shows plots of the measured VSWR of both bflar feeds of the constructed QHA n fgure 3-6 as a functon of frequency. t should be noted that the mpedance (that produced the VSWR plot n fgure 3-7b) was measured through one 180' power spltter and a 3 short length of cable connectng the power spltter to the antenna. The short length of cable has a small loss; thus, the measured VSWR s slghtly lower than that of the antenna. n addton, the power spltters are desgned to operate from 2 to 4 GHz; thus, the measured mpedances outsde of ths range wll probably have some error. The plots show a low VSWR of magntude less than 2.3 for frequences above 2.2 GHz. 3o 3 3

37 E (a) 90 ~~ 10.~ -- 0 RHCP L (b) 10- Feed e----[ F e 1 1 2Frequency (GHz) 5 Fgure 3-7. Measured Crcut and Radaton Characterstcs of the Constructed QHA Shown n Fgure 3-6: (a) Gan Patterns at 2.45 GHz and (b) nput VSWR of Both Feed Arms 1 3'

38 Fgure 3-8 shows Smth charts of the measured mpedance loc for both bflar feeds of the constructed QHA. The plots of the measured mpedance show that above the cut-n frequency the mpedance locus sprals nward toward a characterstc mpedance Z 0 of magntude slghtly above 100 ohms and rotated n the nductve drecton. n comparson, the computed characterstc mpedance obtaned from fgure 3-2b has a larger magntude than the measured values because of the narrow, thn-wre helcal elements used n the model. The measured nductve component of Z 0 may be attrbuted to two possble sources: (1) the slghtly dsplaced feed pont attrbuted to the separaton of the feed-cable center conductors from the antenna axs, and (2) the short length of center conductor requred to connect the cables to the radal feed sectons of the helcal elements. The measured mpedance loc n fgure 3-8 are tghter than the computed locus n fgure 3-2b 3 because the narrow conductng elements used n the model have less bandwdth than the wde-conductng strp elements used n the constructed QHA, and there s some small loss n the 5 short cables that connect a bflar helcal secton to a 1800 power spltter. n concluson, the constructed QHA n fgure 3-6 produces an acceptable radaton pattern. The VSWR of the constructed QHA at 2.45 GHz s a lttle hgh because the antenna s operatng 3 just above cut-n as can be seen from the Smth charts n fgure 3-8. The mpedance loc n fgure 3-8 are not tghtly wrapped about Z 0 n the vcnty of 2.45 GHz. To mprove the 3 mpedance match of the QHA to the power spltter, the nput mpedance can be ncreased by decreasng the helcal conductor wdth. Ths adjustment n the conductor wdth can be made to 3 ncrease the characterstc mpedance Z 0 to approxmately 200 ohms n order that the nput mpedance at 2.45 GHz s approxmately Zo/2 or 100 ohms. 3 1a 32 5 a

39 E (a) ~ _ ,.0 5 o ~~.5 0,j (b) * * Fgure 3-8. Smth Charts of the Measured nput mpedances of Both Bflar Pars of the Constructed QHA Shown n Fgure 3-6: (a) Feed 1 and (b) Feed 2 33 (34 blank)

40 -- 4. QHA ARRAY DESGN AND CONSTRUCTON The analyss descrbed n secton 3 concluded that the array element wth the optmum 3 radaton performance that meets the sze constrants for mountng on a buoy s a forward-fed, open-crcuted QHA wth a helcal dameter of n., a 300 ptch angle, an axal length of n., and a helcal conductor length of approxmately 1.25 X (6.03 n.) at the center frequency of 2.45 GHz. To smplfy the constructon of the antenna, each helcal element was fed at ts 3 crcumference nstead of ts center, thereby removng the radal sectons of conductor. Consequently, the axal length of the antenna was ncreased from 2.71 n. to 2.90 n. to mantan the same helcal element length and, thus, mantan an optmum mpedance match at 2.45 GHz. Four of these QHAs were arranged as a vertcally-orented lnear array wth an element spacng of 3.61 n., equvalent to 0.75 X at 2.45 GHz. Ths spacng was suffcently small to remove the possblty of gratng lobes and large enough to reduce sgnfcant element mutual couplng. Fgure 4-1a gves an llustraton of the QHA array nstalled on a buoy and fgure 4-1b shows the dmensons of the array elements wth ther locatons relatve to the buoy surface and ar-sea water nterface. Ths secton descrbes the mechancal constructon of the element and array desgns, ncludng the feed assembly. The measured crcut and radaton characterstcs of the 3 array are deferred untl the followng secton. Fgure 4-2 shows a vertcal-plane vew of the array feed-cable assembly. Each QHA element 3s fed by four coaxal cables, resultng n a total of 16 cables that are requred to feed the entre four-element array. The feed cables travel parallel to the array axs and are enclosed n a brass 3 tube placed nteror to the QHA elements and coaxal wth the array. The brass tube provdes electromagnetc sheldng of the feed cables from the radatng elements. The brass tube extends 3 a short dstance beyond the top and bottom array elements n order that these antennas experence approxmately the same amount of couplng to the tube as the two nteror elements. 1 a Each coaxal cable passes through a hole n the brass tube n order to feed a helcal conductor of a QHA element. These holes are drlled at a 40' angle wth respect to the vertcal drecton to allow for a more gradual bend of the feed cables and mnmze the possblty of breakng the thn center conductor of the cable. The outer conductors of the coaxal feed cables are soldered to a brass rng that s attached to the brass tube. A cross-sectonal vew of the feed ponts of the bottom QHA element s shown n fgure

41 (a) 1' " QUADRFLAR HELX ARRAY 3 BUOY (b) QHA Element ---,,-.q n. T n n. Buoy Surface t 12 n. 8.7 n. 5 Ar 4 Sea Water Fgure 4-1. Four-Element Lnear Array of QHA Elements Descrbng (a) nstallaton on a Buoy and (b) Array Geometry wth Dmensons of Array Elements and Ther Locatons Relatve to Buoy Surface and Ar-Sea Water nterface 361

42 U 3 Feed Cables Brass Tube Enclosng Dameter Of Brass Tube * _ Helx Dameter 5Helx Element 4 Cable Coaxal Cable Dameter Helx Element 3 m Spacng Between Elements U -- Helx Element -- to Allow Connecton Jl 2et TtoHelx Feed Fgure 4-2. Vertcal-Plane Vew of Feed-Cable Network for the QHA Array 37

43 Dameter of Brass Tube Contanng Coax Cables - Coaxal Cable Coaxal Cable Dameter Brass Tube Contanng Coax Cables Center Conductor 0, Outer Conductor. Helx Dameter 3 Fgure 4-3. End Vew of Bottom Element of the QHA Array Showng Feed Cables 3 n the orgnal desgn of the array, the brass tube and QHA array were to be constructed as a contnuous pece. However, ths desgn was determned to be dffcult to assemble, especally n the attempt to pull the feed cables through the tube and slde the array of QHAs over the protrudng center conductors of the feed cables. Consequently, the array was splt nto four separate peces that could be assembled together. Fgure 4-4 shows a vertcal-plane vew of the entre array assembly, whch ncludes the antenna elements bonded to a Mylar tube, feed cables, sx sectons of brass tube, and plastc 3 spacers. Because the array elements and ther assocated brass-tube sectons are dentcal, only the dmensons assocated wth the bottom antenna element are shown. The colors n fgure 4-4 correspond to the dfferent materals from whch the array assembly s constructed. The feed tube s constructed of brass nstead of copper because of ts lower thermal conductvty, whch allows localzed heatng assocated wth solderng. Each secton of tube s stepped at an end that connects to an adjacent secton of tube to provde contnuty. Adjacent sectons of brass tube are 38 1

44 U attached va four set screws. There are four radal and crcumferentally symmetrc-spaced holes placed about the tube to allow four feed cables to connect to each QHA element. Each hole was made slghtly larger than the n. outer dameter of the feed cables to permt solderng of the cable to the hole and to avod any possble rado frequency (RF) ground loops that may form. However, because t was determned that solderng the feed cables to the brass rng was suffcent, the cables were never soldered to the holes. f Fgure 4-5 shows several QHA elements of dfferent conductor wdths that were tested to determne the strp wdth w that provdes the proper characterstc mpedance Z 0 that permts an optmum mpedance match over a narrow bandwdth centered about 2.45 GHz. Each QHA n the photo was bonded to a 3-ml-thck Mylar tube. From an analyss of the measured 3 half of the avalable space along the helcal crcumference. A comparson of the measured mpedance data for each conductor wdth s gven n secton 5.1. mpedances, the optmum conductor wdth was found to be w = 0.24 n., correspondng to one- Fgure 4-6 shows the secton of brass tube that s assocated wth a gven antenna element. The tube ncludes four symmetrcally placed holes for entry of the feed cables. A brass rng, attached to the tube, s soldered to each of the four feed cables and allows each helcal wndng 3- of the QHA to be fed at the helcal radus. Two hard plastc spacers are glued to the brass tube wth a thn spacer attached to the top surface of the brass rng and a thck spacer at the upper end 3of the tube. The spacers are used to hold the Mylar tube n place and to mantan a constant helcal radus. Holes were drlled n the thck plastc spacer to reduce the amount of delectrc 5 loadng that t has on the antenna. Delectrc loadng can produce the undesrable effect of ncreasng the amount of radaton n the backsde drecton. n the orgnal desgn, the top 3 spacer was made thck to support the ends of two adjacent Mylar tubes. However, n the current desgn, because the thck spacer only supports the top of one Mylar tube (correspondng to one QHA), the thckness of ths spacer may be reduced. 1 39

45 *RE Conductors El Brass * Plastc Spacers1 EJ Mylar Tube *Feed Cable /56x 3/4.11 1/6 St Scew - 1/1 Sp/c Se/c1w634 R/56 - jx Ealen 1/ Sacer /16, /43 5/8 Fgure 4-4. Four-Element QHA Array Assembly3 403

46 U w =0.36 n. w 0.24 n. w 0.12 n. w 0.06 n. Fgure 4-5. QHAs of Dfferent Conductor Wdths CABLE ENTRY HOLEFORSLOT TO SOLDER ALONG CABLE LENGTH U Fgure 4-6. Secton of Brass Tube Assembly and Plastc Spacers Assocated wth a QHA Element Fgure 4-7 s a photo of a QHA element and ts assocated feed network of power spltters. 3 The feed network conssts of an nput 900 power spltter wth ts output ports connected to two 1800 power spltters. The four output ports of the par of 1800 power spltters provde a quadrature feed for the antenna. Note that ths feed network s dentcal to the one used for the 1 41

47 prelmnary constructed QHA shown n fgure 3-6b. The measured crcut and radaton characterstcs of ths QHA element n free space are presented n secton ANTENNA SPLTES 90 PLTTE R Fgure 4-7. QHA wth Quadrature Feed Network of Power Spltters Fgure 4-8 gves a closeup vew of the feed regon of one helcal wndng of a QHA. The photo shows that the outer jacket of the feed cable s removed to solder the outer conductor of the cable to the brass rng. The solderng of the cable outer conductor to the rng helps to establsh the outer edge of the rng as the locaton of one end of the feed pont. The other end of the feed pont s the connecton of the cable center conductor to the helcal element. Ths procedure helps to prevent any RF ground loops that may be produced that could nduce undesrable currents along the brass tube. Fgure 4-9 s a closeup vew of an nteror QHA element n the array. NPUT Fgure 4-10 s a photo of the entre QHA array assembly, whch ncludes the four antenna elements and 16 feed cables and the four element support structures, one bottom tube extenson, and one top tube extenson. The dmensons and constructon of ths array are consstent wth the drawng n fgure 4-4. The array elements are numbered sequentally from 1 to 4, where elements and 4 denote the bottom and top elements, respectvely. The 16 RF feed cables of the array are of dentcal length n order that the losses assocated wth each element are the same and that the mpedance measurements for each element are referenced to the same pont on a Smth chart

48 :D CONNECTON PONT,CONDUCTOR MYLAR SHEET OUTER- C - Fgure 4-8. Closeup Vew of Element Feed Regon M - Fgure 4-9. Closeup Vew of QHA Element Showng Feeds and Spacers! 3 Fgure QHA Array wth Feed Assembly a 43

49 Fgure 4-11 a s a photo of the QHA array mounted to the buoy plate. A Styrofoam rng spacer s attached to the top secton of the brass condut to allow for nstallaton of the radome (over the array) wthout causng damage to the helces. A 10-n. long Teflon tube, attached at the top to the lower secton of the brass condut, s desgned to ft wthn an alumnum sleeve at the bottom to permt adjustment of the heght of the array above the buoy. The alumnum sleeve s ftted wth a flange to bolt the array assembly to the buoy plate. A rack s attached to the bottom 5 of the buoy plate to mount the electroncs package requred for the at-sea demonstraton. Fgure 4-1 lb s a photo of the array wth the radome attached. The radome conssts of a thn 5 (1/16-n. thck) tube of G-10 fberglass wth an outer dameter of 1.75 n. and a crcular cap attached at the top to protect the array from exposure to the harsh sea water envronment. The 3 measured results presented n secton 5.4 show that the radome has only a mnor effect on the crcut and radaton characterstcs of each QHA element. 5 Fgure 4-11 c shows the RF electroncs assocated wth the QHA array that s mounted to the 3 lower surface of the buoy plate. To collect the data necessary to provde a frst-order estmate of performance that could be expected from the array and receve beamformer, the receved sgnal 1 envelope from each of the elements was measured as a functon of tme. The magntude of the envelope s expected to vary wth the presence of multpath sgnals. t s crtcal that the only contrbutons to the recorded sgnal are due to the sgnals of nterest, namely those assocated wth the drect and reflected paths. Because the bandwdth of the antennas s relatvely wde, the output 3 of each antenna was fed nto a bandpass flter centered at GHz wth a bandwdth of 22 MHz. The sgnal was then amplfed to brng t nto the range of the envelope detector. A low-pass flter 5 (LPF) wth cutoff frequency of 100 Hz was mounted at the bottom of the assembly for possble use n the sgnal detecton. A comparson of the measured output of the detector wth the same sgnal 3 after gong through the LPF ndcated no apprecable dfference n the output. Consequently, the LPFs were not utlzed durng testng of the array. The sgnals were dgtzed and recorded on a 3 laptop computer for later analyss. Fgure 4-1 ld s a photo of the QHA array sttng above the buoy pror to a sea test of the array n Narragansett Bay. Once the batteres and laptop computer were nstalled nsde the buoy, the array was then bolted to the buoy. The results of the sea test are presented n another report by the authors (reference 1). 3 4

50 ! Fte (a) (b) - 3one Arraw Feedaead() Low Pass Flter 3- (c) (d) H rayadbo ( _ Fgure (a) QHA Array Mounted to Buoy Plate Wthout Radome, ~(b) QHA Array Mounted to Buoy Plate wth Radome, (c) RF Electroncs -- Mounted Below Buoy Plate, and (d) QHA Array and Buoy 45 (46 blank)