(YAS!4-CE-l51866) S O L A R PCWER S A T E i L I T E RECTENWA DESIGY STUDY: DIFECTICNAL R E C E I V I N G ELEYJENTS A N C F A B A L L E L - S E R I E S C O l l B I N I N G A N A L Y S I S F i n a l Report, 3 Eeb, 1978 1 Dec. (Bensselaer Polytechnic Inst.,
-
N79-16039
63/15
COFJTRACT NAS 9-E4!3WIM (~ATIONAL A ~ ~ ~ ~ uAND r rSPACE c s ~INISTRAIION LYNDONB, JOHNSON SPACE CENIER
FINALREM D E ~ E 1978 R
TJncLas 43502
TABLE OF CONTENTS
............................ L i s t o f F i g u r e s ........................... L i s t o f mbles ............................ 1.0 Programoverview ........................
Acknowledgment
1.1 Wsk Definition 1.2 Approach m e n w i t h Underlying Asanptions 1.3 S u m a q of Key Accomplishments 2.0 More Directional Recei-ring Elements 2.1 Rectenna Alternatives Considered w ia @..erviewComparison 2.2 Principal Electrical Considerations 2.2.1 Printed Circuit Implementation Considera'dons 2.2.2 Yagi-Uda Design Features 2.2.3 &&ne Considerations 2.3 Cornprison of Ught RecteMa Element Designs 2.3.1 Designs Considered l compwiso= 2.3.2 o ~ r a l System 2.3.3 Cost Coqazison 3.0 -9mer Combining E d u t i o n 3.1 Rectifier Circuit Models 3.1.1 Computer Simlation 3.1.2 Closed Form 3.2 bad Line analysis 3.3 Rmer Combining Analysis
&s i ii iv 1
...............
8
...................
66
.
3 3.1 Concepts 3.3.2 Results
3.4 m e r Module Size Constraints 4.0
................. .............................
S v and Future Directions References.
113 U5
Acknowledgments
Numerous students a t Rensselaer contributeci t o Lhe successfrrl conpletioc of this -program.
I n particular, R. Gworek contribute6 s i g L f i c a n t l y i n the area of Yagi-Uda desiqn and cost evaluation, 3. Seyr-ur and 14. Van slyke in t h e printed circuit i n p l e n a t a t i o n f e a s i b i l i t y e-raluation and J. ?arricelli and M. Mackiw i n t h e computer s i t i o n e f f o r t . We m u l d also l i k e t o acl.sarledge tbe sugport and t e e n i c a l ir?teraction w i t h numerous JSC personnel. I n particular, the suggestions and feedback from J. S. KeUey and R. E. Dietz tkoughout *&e program and t3e come,-sat i o n s with F. Stebbins on rectenna structure deserve special recogfition. Finally, t h e assistance of M s . M. Jadlos i n prepariag t h e f i n d report 2s gratefully acknowledged. 3. J. Gutmann j .
i
M. B r r e g o
LIST 07 FIG*XtXS 2-1
Advantages and Disadvactages cf a FLgher WE, Ebre 3irectiong3 Receiving Element
..........................
2-2
iYinted Circuit €!!-Wave
Dipole Recteme E l m e x
2-3
Concept of Chebyshev Input r a t e r Design.
.
6
2-4
.............. Fhotogra2h of. Prototype Lumped Zlement IapA P l l t e r . . . . . . . . .
2-5
A Yagi-Zida Array wi'&
2-6
Xogline rwabolic Reflector Conductor ?.eo_*i?i,re3ent.s and zipole illigment f o r f i o n g i t u c n a l an2 Trzzsverse 3larizzticr-i
Directors of Constant Lecgth, 3iadLus and Spzcing
9
--16, 16 21 26
.......
kl
2-15
'-3
4
3-5
-"eckage an2 Chi? Voltage Weveforz of Ccr2cter S i r S a t i o n !.bdel a t l"Leve1
.............................
S-le
S e c t i f i e r C i r c ~ ~ Land t 3x4 Voltage! Wzvefcrzl
..........
3-6'
3-7 3-8
3-9 3-10 3-11
d
3-12
Xectifier Circuit with a Low Pass F i l t e r i n the O u t s t and Associate2 Current and Voltage Weveforns
............ *. Rectifier Circuit w i t h Input and Output Filters and t h e i r Frequency Characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . Realization of a Righ E f i c i e n c y Rectifier Circuit with Lumped Element Input and Outgut Filters . . . . . . . . . . . . . . . . . . . Realization of a E i c i e n c y Rectifier Circuit ~ 5 t ha T r a n d s s i o n Line as Output F i l t e r ......................... . Current and Voltage Wavefoms i n a FA& Efiicieccy Rec'dfier Iligh
Circuit
Output Equivalent Circuit and Load Liae C!mracteristics f c r a 3igh Efficiency Closed POD S e c t i f i e r C i r c e t
............... Loat! Line of Computer Simulation Model at 1.04 CoEpaned t o Close5 FomModel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C
3-13 had Lice of Compufer S k i L a t i o n Model as a 3 m c t i c o of b c i d e n t -~r................................. 3 - 1 ~ Ou%?ut Equivalellt Circldt ,'tarmeter of CcmDuzer S i t i c n k d p 7
...................... Conversion Z f i z i e n c y of Computer 2unction Caoacitaxe es . . . . .Xcdel . . .wi*& . . .Zem-Sas ....... as
a Fmction of Incident %wer
3-15
Sbdamt;"=ion a Z3xameter
3-16 Generzl
V-Z Characteristics 02' ac E e c t r i c d CeX and O _ s + , i m O_Der&thg
FbiIlt...............................
3-17 Series o r & d e l
Interconnection of "WC X f f e - r e ~ Z t l e c t r i c d Ce?'s
.
3-12
3-19 O u t p t 3qAvalent CLrc-&t of Rectema Zlexents %
a a,
E
tri
28
YN 0)
0
.
3 h
. d
0 Icc
a
&
k
1
4
Y
n
0
0
0
.rl
r(
m a k
m
\
fr
.
k
Ln
m
\D 0
\
t k
c
M
5 c +; C
8
ri iri
+J
29
possible, structural. considerations and diminishing returns on g a i n improve-
a n t s indicate that longer elements are not desirable. A key f a c t o r i n t h e Yagi-Uda element design f o r L e S-3s would be maxihizing gain f o r a specified F/B r a t i o . A high F/S r a t i o would eliminate t h e ground plane mesh i n the baseline design, and possibly lead t o principal In order t o eliminate the s t r u c t u r a l advantages (discussed i n Sec. 2.3.2). ground plane, a F/B r a t i o > 25 dB i s probably required, i n order t o insure high collection efficiency and l o w microwave reradiation past the rectenna.
l a t t e r specification depends upon multiple land usage. However r o t a t i o n of the RF f i e l d polarization i n the ionosphere places a laser limit on the microwave leakage ~ M c can h be obtained without a ground plane. That is, t h e perpendicular componen'; would not be affected. SigrLficantljr by t h e Yagi-Uda receiving element. If' a groune -ne mesh i s included w i t h lower F/9 r a t i o designs, the e l e c t r i c a l performance of the Yagi-Vda elem&t i s affected. The ground plane i s assumed t o be .25X behind t h e active elernent, i n p r o x h i t y t o the Yagi-Uda passive r e f l e c t o r . Since the "reflector" curreat d i s t r i b u t i o n W i l l be less s-patiiilly concentrated wit,, 8 grou.116 Flaae, the antenna p a t t e r n i s expected t o change. Eowever, such a design has not been analyzed t o date and wes beyond the scope of this i n i t i a l study. Ve have assumed that s h L l . a r perfomance i s a t d l a b l e T&*&a ground plane mesh, although slightly higher gain i s expected. W i t h a ground g l m e included, trJc a l t e r n a t i v e s were ccrsidered 8 15 dB F/B r a t i o judged t o be appropriate f o r relaxiag electricei requires e n t s on the ground ?,lane mesh and 5 d3 "3 r a t i o r e q d r f c g excellent ground plane conductivity (as required in baseline r e c t e m a w i t 2 ha?D-wave dipole receivlng elements). X i t h the proyosed s t r u c t u a l configurations (16) Of course, quantifying the
-
-
groiJnd plane mesh requirements a r e not severe. However witn (e-ected) lower cost s t r u c t u r a l design using mesh held i n tension, ('7) relaxing e l e c t r i c a l conductivity requirements would be desirable. Based upon t h i s w r k , Table 2-8 is our best estimate of Vagi-Uda receiving element e l e c t r i c a l c a p a b i l i t i e s and r e s u l t a n t reductfon.Ln number o f receiving elenents. While appreciable analytic and experimeotal work is needed t o verify these p a r t i c u l a r Ferfornance c h a r a c t e r i s t i c s
(particularly in an array environment), these parareters are s u f f i c i e n t l y acctuate t o evaluate the p o t e n t i a l of a Yagi-Uda recterma. mey have been used i n t h e d e t a i l e d rectenns work described in Section 2.3. Because of an expected s i g n i f i c a n t effect of a ground plane on Yagi-Uda elements having
a F/B ratio of only 5 dB, they are not included in our detailed design work i n Sec. 2.3. Hcwever the l a r g e r gain indicates t h i s t o be a promising area f o r further work. While our emphasis has been on maximizing gair: f o r a specified F/B r a t i o , additional electrical f a c t o r s af'fecting s u i t a S i l i t y i n the S-PS rectema have been considered. Referring t o Tables 2-6 and 2-7 the i a p u t or radiation impedance of the various designs are presented. In general the Yagi-Uda r a d i a t i o n resistance i s less * h n the half-wave dipole (nominally 73 o h ) , which is undesirable f o r h i @ KF %o E conversion efficiency. The input bpedance can be increased by using 8 folded dipole f o r the driven element. This i s easily included wiA& p i n t e d c i r c u i t hplementation and only slrghtly more dif'ficult wi*& basellne type implaentation. All Yagi-Uda designs described i n Sec. 2.3 have a folded di-le as t h e driven (active) eleznent. An additional f a c t o r affecting SPS implementatioc of a Yagi-3da r e c t e m i s the dimnsional tolerances involved. As d i r e c t o r spacings vary as l i t t l e as .01X, F/B r a t i o can Se s f g r 3 f c a n t - y reduced With the Ugh F/B r a t i o designs. 2'ort;znately ozher parameters e r e not as sensitive to dimensional w i a t i o n s , b u t F a m e t e r tolerances need t o be investigated. C e r t a i w less s e n s i t i v i t y i s yossible i f p e r f o m a x e ?aremeters ( g a i n and I/S) are "backed o f f " t o scme degree. Certzhljr parameter secsitilety nus$ be examined i n f u t u e Yagi-Uda evaluations.
2.2.3
Eogline Considerations During the i n i t i a l stages of the p r o g r a , the hogline r e c t e m a concept was emlored. However, t o avoid duplication of ongoing e f f o r t s a t 3oeing and t o a l l o w detailed evaluation of Yegi-Uda elements a& 3riDted c i r c u i t implementation the hogllne e f f o r t yas c u r t a i l e d a f t e r our interim presentat i o n a t JSC. This section CL t h e report presents mstly e l e c t r i c a l desigr? considerstions t h a t need to b e addressed and ~ u pr r e l h h s y evalilaticr, r e s d t s . tu?lile these considerations ?3a;r have beer. eid.;z"ued by sei.%, tbey
are not presented in much d e t a i l in the reports available t o
US.
(6,7)
Because of our miniml e f f o r t on the hogline and incomplete design t r e a t ment, a comparison t o the baselhe r e c t e n m o r our a l t e r n a t i v e concepts presented elsewhere in Chapter 2 i s not possible at t k i s time. In this brief discussion familiarity with the hogline antenne concept i s assumed.
In paxticular,
we will discuss three considerations:
the e f f e c t of
longitudinal versus transverse polarization, the e f f e c t of multise diffraction, and consideration of multiple rectifying elements perpenWhile these f a c t o r s me interrelated t o some extent i n the electrical design of the hogline, we f i n d it u s e m t o consider them separately f o r discussion purposes. dicular t o the f o c a l line.
The hogline, a l i n e a r i z a t i o n of the hog-horn antenna developed f o r satellite reception, i s depicted i n Figure 2-6, i n which the parabolic r e f l e c t o r surface i s enphasized. The incident power bean can be polarized either i n the longitudinal, i.e., p a r d e l t o the f o c a l l i n e of t h e parabolic cylinder, or transverse, i.e. , parallel t o the narrow dimension
(W) of the parabolic r e f l e c t o r aperture plane, direction. For emphasis every other parabolic r e f l e c t o r depicts t h e conductor requirements nftl? these incident polarizations. The incident polarizetions a f f e c t the e l e c t r i c a l d e s i g i appreciably. For example t n e corner r e f l e c t o r design at the mouth of t h e inclined plane i s polarj zation dependent. (16) w i t h tracsverse polarization, t h e e i e c t r i c f i e l d w i l l be perpendiculer t o the condtrcting surfaces i n the i n c l i x d ?lane region. Therefore, d t h proper design tbe power density cafi be nearly spatially unfforc? across the region betweer, the inclined planes. With 1ongi.tudi.d polarization t h e e l e c t r i c f i e l d is 5 a r a l l e l tu these conducting boundaries, so that a TE& mode (rather than e TZM mode) becomes a fundmental mode i n t'nis region. Therefore, u n i f c m power der?sity across t h e inclined plane region i s much more d i f f i c u l t t o obtain. This uniform power density i s p a r t i c u l a r l y i q o r t a a t i f multiple receiving elements a r e placed perpendicular t o t h e f o c a l l i n e . only approximate as the incEned plane ape-rture A
9.f~ discussion i s
wix probably be i n the
ne= f i e l d of t h e parabolic apertqne W f o r high collection efficiency. However, longitudinal polarization has 8 m r e eesirable r e f l e c t o r condllctor requirement (Fig. 2-6).
Since or&
con&x:ors
g a h e l t o the
._
I .
$32 Fig. 2-6
Hogline %raboUc Bef'lector Conductor ilequlrements and X p l e Alignment for Longitudinal and ensv verse ala-rization
ground are needed, a l i g h t weight s t r u c t u r a l design with "wires" held in
.
tension appears possible W i t h transverse pol&ization a parabolic surface must be formed, so that a conducting mesh may be desired. Naturally a mesh could a l s o be used with longitudinal polarization as well. A significant difficulty with the longitudinal polarization is DC
collection buss r e q u i r m n t s . Since the e l e c t r i c f i e l d in the inclined plane region is p a r a l l e l t o the f o c a l l i n e , the i n d i v i d u a l output terminals
must be brought through the ground plane. W i t h transverse polarization, the DC collection buss can be p a r a l l e l t o t h e f o c a l l i n e "inside" the incllned planes, as the e l e c t r i c f i e l d is perpendicular t o this l i n e . Not protru2.ng through the ground plane is considered a significant advantage. A fundamental consideration with the hogline i s diffractioK.
Besides d i f f r a c t i o n f r o m t h e top surface of the _parabolic cylinder aperture (similar t o that from the grow16 plpne i n the baseline rectenna), there i s a double d i f f r a c t i o n from the bottom surface of t h i s aperture .(same as t o p surface of the i n c l i n e d plane aperture). The l a t t e r near f i e l d , double -action i s a complicated problem t o analyze and i s a l s o polarization dependent. This i s considered t o be a signfficsnt e l e c t r i c a l design problem that needs t o be addressed i n f u r t h e r hogline f e a s i b i l i t y studies. Instead of the simple l i n e a r array of receivlng elernents along the focal l i n e &s fndicated i n Fig. 2-6 , 2oeing is proposing a p l a w array dth 10 elements placed across the fncuned plane &de(7) (probably
-
symmetrically located w i t h respect t o f o c a l l i n e ) . With tkis arrangement the f i e l d d i s t r i b u t i o n across this plane is of conceyn. L" d i f f e r e n t elements receive d i f f e r e n t powers, conversion c i r c u i t r y ineffLciencies can result. This nonuniformity e x i s t s even i n e. f i r s t order analysis of the incllned plane-cylindrical ?ar,oolic r e f l e c t o r configuration and i s axpected t o be enhanced with t h e near f i e l d double diffraction. It should be emphasized that none of the above f a c t o r s have been evaluated i n s u f f i c i e n t depth t o indicate that e f f i c i e n t power beam
reception and RF t o DC conversion would not be possible with the hogline. However, in our brief ho&+2e evaluation, they appeared as key e l e c t r i c a l design f a c t o r s that need t o be addressed i n some detaLl i n any f u r t h e r hogl i n e investigations.
Fig. 2-7A and B Baseline Half-Wave Dipole Rectenna (after Ref. 1)
34
Proposed design of Rectenna motivated by environmental proFigure 2-7A tection and c o s t considerations.
Figure &7B Physical construction of two-plane rtctenna. I V i t h the exception of covers (white teflon s l e e v e s in photograph )'this i s the same five element foreplane that was electrically tested in earlier work. Reflecting plane made from hardware cloth i s representative of what could be used in SSPS rectenna.
,
Mg. 2-7C and D Baseline KE-Wave Dipole Rectanna ( m e r Ref. ?>
35
Fig. 2-7C Core Assembly Fgbrication
t = .015
I>i?Pensions ir c e n t b e t e rs
Fig. 2-73 Foreplane A s s ~ n b l yS h i e l d
Q)
rl
.rl& f
.
bD
rl
R
37
2.3
COMPARISON OF EXGET RECTENNA ELEMENT DESIGNS In t h i s section we present design concepts f o r seven a l t e r n a t i v e rectenna
elements and compare these w i t h the baseline rectenna element.
Idtially
the eight designs are presented w i t h a brief description of each, followed by a general comparison of the designs. F i n a l l y cost estimates for the eight designa are discussed. It should be emphasized that the costing methodology developed by Raytheon (lS6) is followed, but resource limitations and relative lack of s u f f i c i e n t design refinement i n d i c a t e that o i l y teutative comparisons are possible. However mre 3.ikJ.y alternatives,
and relative advantages and disadvantages, clearly appear. W e believe the d a h presented i n this section should be used i n evaltlating f u t u r e directions of alternative SFS rectennas. 2.3.1
Designs Considered
The eight designs are grouped i n t o haf-wave dipoles, 3 element Yagi-Uda rectema elements and 6 element Yagi-Uda rectenna elements. Our designs w i t h a b r i e f description of each follows: --wave
dipoles (baseline construction and pricted circuit i=tplemen-
t a t i o n ) : The SPS r e c t e m a baselfne design as develope6 by Raytheon consists of a half-wave dipole over a cozducting ground g b n e . me gain of this dipole a quarter wavelengtk from a ground plaoe is 5 . 1 dB with respect
to an X/Z dipole r a d i a t o r (17)o r 7.25 d9 w i t h respect t o an i s o t r o g i c radiator. However, the &n of t h e baseline dipole ii t h e rectenna array 'is about 6 . 4 d9, tfie slight decrease exgLahed by m t i i couplicg amow the elements i n *de closely spaced c r a y . The two plane construction depicted i n f i g . 2-7 :A and 3) consists of t h e ground plane mesh and the foreplane, which contaiss tfie d i p l e , c i r c u i t r y and DC buRs (Fig. 2-1C). a?e foreplane shield(F5g. 2-73), &e from aluminum sheeting, form the major s t r u c t u r s l d e r and p o v i d e s environmental protection. It i s attached t o the r e c t e m a frame v e r t i c a l members and also has the ground plane folded i n t o t h e foreplane stem. 3 - z t.he ~ ground plane has an auxiliary s t r u c t u r a l ? u r p s e besides i t s function as a n antema reflector.
The printed circuit design should have c h a r a c t e r i s t i c s similar t o the baseline dipole since it a l s o i s a half-wave dipole over a conducting ground plane. However, because of the smdler conductors, c i r c u i t l o s s e s
are increased w i t h a r e s u l t a n t reduction of about 2-55 i n conversion efficiency (discussed i n Sec. 2.3.l), an i s p o r t a n t limitation. 'ibis proposed printed circuit board (XB) implementation shown i n Fig. 2-8 consists of three main parts: the ECB, the FCB socket, 8nd the DC collection buss bars. The RB hss been previously discussed and i s me rectenna frame supports the ground plane and depicted i n Fig. 2-2. the DC buss bars. The socket extensions also s ~ ~ p r *t&se ground plene, keeping it a t the one quarter wavelength distance f r o m the dipole. %e sockets are made from strong, durable plastic w i t h embedded conducting strips t o malre the connection between the X B an8 the buss bars. '5e socket c l i p s on and holds tightly to the two aluminum buss b a r s which are the major structural su-gorting member of the rectema element. Three element Yagi-Uda rectenaa d e n t s ( w i t h and w5'hut
growxi
p h n e as well as printed c i r c d t and b a d h e @e irnlezentation): Figs. 2-9 through 2-13 depict a l t e r n a t i v e w e e element Yagf-UCa elezents, based upon e l e c t r i c a l deslgn considerations presexed fz Section 2.2.2. In the -minted circuit implementation (Figs. 2-9 and 2-10] t h e E 3 socket w i t h the ground plane i s i d e n t i c a l t o the pC9 di-mle socket. Xmver wi%h the given dimensions the socket without a ground ;lane is soxnewh-~
larger, because the r e f l e c t o r can be above t h e X buss. Recall that pelletration through the ground plane i s considered .mdeslrable, so * & t
the
M: buss bar i s i n f r o n t of the r e f l e c t o r i n t h e design wLtk a &road 2lane.
P a s extra length for ';?le socket may be 8n advan--e IC3 Yagi i s greater than t h e
XB d i s l e .
since loading oc -,he
Printet! c i r c u i t board
bglementa-
t i o n is sham i n Fig. 2-ll. With baseline type implementation (Figs. 2-12 and 2-13) t i e baseEne rectenna i s modified i n a straightfcrward fashion. If 8 gz3und plane i s needed, t h e Raytheon foreplane i s modified i n t o a Yagi 5g attaching a twopiece plastic support mast for t h e director. Without a ground slane, the environmental shield i s the main sapporting member and is af'tacheci
39 i%g. 2-8
Prfnted c i r c u i t Iialf-wave pole (aU dimensions in centimeters)
1'
1
-5
Mg. 2-9
lbree Elenrpnt Frinted Circuit Yagi-Uda without
(adimensions
Ground
K
3
in ccatimeters)
Fig. 2-10 !5ree Element Rinted C i r c u i t Yagi-Uda with Ground &e
42
t3
43 I
Gd
4
C
I-
43
Fig. 2-32 Three Element Baselfne Coastsuction Yagi-Uda with Ground m
e
44
Fig. 2-13 l b e e Elemeat Baseline Constructioa Yagi-Uda d t h o u t wound &e ( a l l dimensions i n centimeters)
-1.3
1-
45
d i r e c t l y t o the rectenna frame. The shield i s similar t o the foreplane shield of the baseline dipole without the stem extended t o the ground plane. A two-piece p l a s t i c support mast encircling the shield snaps together clamping the passive elements of the Yagi t o the active plane.
Six Element Yagi-Uda Elements: With printed c i r c u i t implementation (Fig. 2-14) the RB socket i s sl?niln.r t o those already presented (Fig. 2-9) with minor changes i n structure t o support the longer Yagi-Uda elements and perhaps other changes t o accomodate a l a r g e r per element power reception (although t h i s is not expected t o be needed a t this time). The. baseline coastructed long Yagi-Uda element i s also very similar t o p r e d o u s l y described designs (Fig. 2-12) with exception of minor s t r u c t u r a l modifications. Detailed designs are given assuming no ground plane. 3ecause of the different lengths of the six element designs (see a b l e 2-7), two designs are shown i n Fig.. 2-15 f o r comparison purposes. Overall System Comparisons The l i s t of f a c t o r s considered i n our work and t o be discussed i n this section include: a S ~ ~ ~ I E XofY e l s c t r i c a l considerations (described i n d e t a i l i n Secs. 2.2.1 and 2,2.2), manufacturing, construction and maintenance considerations, environmental f a c t o r s and multiple lmd use potential. While p r i o r i t i e s a r e not given e x 2 l i c i t l y here, they have been described previocsly i n Sec. 2.1. Because the long Yagi-Vda elements ( i n *-is secCYfon called "Yagis") a r e most affected by environmental loading (snow, i c e , rafr,, v i n d and gravity) and because the long Yagis have the highest g a i n they neet! the strongest suggort and *&e most accurate antenna W i a t i n g . The short Yagis have both lower gain and l e s s environmental loading Lhus requiring l e s s accurate pointing and l e s s s t r u c t u r a l support f o r the element. NaturaUy, 2.3.2
the dipole i s the e a s i e s t t o support and point. Ground s e t t l i n g o r s h i f t i a g of the rectenna frame (or superstructure as i t is s o m e t i a s c a l l e d ) may become s i g n i f i c a n t as the antenna gaia i s fncreased, and s a t e l l i t e beam p i n t i n g accuracy must a l s o be reconsidered. As described e a x l i e r , system studies t o data indicate that higher g a h rectenna recei-ring elements can indeed be accomodated.
46 Fig. 2-14 S i x Element Printed Circuit Yagi-Uda without
.
c
c
d
.
n N 4
x
rr,
I
Q
GrounC Plase
Designs 5 and
3
given in %&le 2-7.
The i s o l a t e d dipole r a d i a t i o n pattern has an F/B r a t i o o f one (0 dB) and therefore depends on a good ground plane t o allow a high collec-
t i o n efficiency.
W i t h a s o l i d ground plane an i n f i n i t e F/B r a t i o i s
achieved, but a s o l i d ground plane i s expensive and s t r u c t u r a l l y unclesirable. A mesh ground plane can have a transmission l o s s > 30 d9 and should have
a good conducting surface t o keep l o s s e s i n the ground plane low, Some Yagis have high enmgh F/B r a t i o s (- 15 dB) t o ease the elect r i c a l c o n s t r a i r t s (e.g., lower conductivity) on the ground plane mesh. If one wanted t o keep l o s s e s i n t h e ground plane constant, t h e smaller amount of induced currents in the ground plane when using a higber F/B r a t i o a l l o w s t h e mesh material t o lave a higher resistance, a l l e v i a t i n g one constraint on the mesh materials. Some Yagis have high enough F/B r a t i o s (- 25 dB) t o consider elimination The ground plane of the ground plane (discussed i n Section 2.2.2). adds r i g i d i t y t o t h e rectenna frame and provides s t r u c t u r a l support f o r t h e baseline constructed dipole and Yagis b u t it adds weight and material. costs. It i s also susceptible t o environmental loading and a source of shading. For these reasons it mig!!t be desirable t o dispense with the ground plane where adequate support can be given t o the rectenna elernents and, of course, t h e F/S r a t i o i s high. C f course, multigle land use issues impact F/E r a t i o requirement and Faraday r o t a t i o n i n t h e ionosphere rust !x considered. The variety afforded by t h e nfdc range of gain and F/B r a t i o possible with Yagis (Table 2-8) a U o w s t%e selection of a slritable design governed by =re rigorous study of the pnysical and e l e c t r i c a l Units of t h e rectenna structtzre. The higher gain antenpas have a higher microwave power f 2 p u t to the diode. Since t h e present Pt-GaAs Schottw r e c t i f i e r s a r e capable of improved conversion efficiency a t higher power l e v e l s (2 t o 5 watts instead
of 1watt as i n the baseline rectenna), improved performance r e s u l t s . If the antenna gain i s increased so t h e r e c t i f i e r s operate above 10 m t t s , then diode r e l l a b i l l t y and/or circuit design must be reevaluaced, In E S antennas conductor l o s s e s are greater t h a n i n t h e baseline because the FCB conductors a r e thlrmer. Although The higher conductivity of copper over the aluminum (factor of 1.64) does compensate, we conclude
a
2 t o 5% conversion efficiency penalty r e F d t s with printed Iilrcuit
implementation (Sec. 2 . 2 . 1 ) . &cause the Dc b u s i s the main s t r u c t u r a l support i n the printed c i r c u i t board implementation it w i l l have a larger cross sectional. area than i t s basellne buss counterpart, rezulting i n lower CC collection losses. The basellne dipole rectenna has a per element resistance of 1.72 x while the f i v e millimeter diameter DC buss f o r the RB dipole implementation has 0.22 x loD3, per element, Indicative of w s s i b l e reductions. The higher g s i n antennas have l a r g e r e f f e c t i v e areas, thereby increasing spacing between the rectenna elements. This increased distance i s important f o r t h e baseline constructed designs where the conversion c i r c u i t r y i s positioned between the elements (two plane construction). me e x t r a room a l l o w s additional harmonic f i l t e r s t o be iacluded. I n addition, the increased spacing reduces t h e number of rows of foreplane or DC bussing by a f a c t o r between 1.3 and 2.4 with e2ther printec? c i r c u i t
or baseline m e imglemntation (depending upon nwnber of directors and
F/B r a t i o ) . Sesfdes these e l e c t r i c a l factors, manufactEing and constructing a rectenna plays a large role ir, o v e r a l l rectenrs. feasibility and cost. I n the following conparison between baseline construction and XI3 ixqlenent a t i o n the Yectenna frame (or superstructxre) is assumed t o be s i d l s and is not diccussed. Although we recognize that sugerstructwe dif'ferences my e x l s t with and without ground planes and width t i f f e r e r t type eleDent realizations, we believe element c o s t differences w 9 l cicminate. Therefore we stressed t E s evaluation. The e r e c t i c n c f t h e Fectenna frame is a d i f f e r e n t phase o f construction and i s assmed t o be independent of our e m ition. The foreplane core of the baseline c o n s t r x t i o n s designs is fabricated from two continuous aluminum s t r i p s sandwichiag it middle s t r i p containing the d i e l e c t r i c for the capacitors and t h e diodes. The foreplane shield!
structural menber i s formed around the core from three sections 01' sheet alumlrm. The dipole caps and seals are then f i t t e t completing the fore3lane. The long s t r i p of foreplane is attached t o the rectezna frane by folding t h e ground plane between the open ends of t k e foreplene s t e n a36 then r o l l i n g and indenting tfie whole assembp;.
"be fabrication of the foreplane and i t s a t t a c b n t t o t h e ground plane Bse two separate assemblies. Eiare-er, i f it were possible, W i t h the long lengths of the aluminum foreplane sheeting, t o form the foreplane shield and attach it t o the me-folded ground plane far enough behind t o prevent buckling there would be less t k spent 00 each section of t h e recteana. This eUnthttes moving and positioning completed foreplanes t o be attached t o the ground plane i n a d i f l e r e n t phase of construction. Be printed c i r c u i t b o d implementation consists of three parts:
RE, R B socket, and the DC b u s bars.
The RB's and the sockets a r e
manul'actured mst l i k e l y near the rectenna s i t e and batches a r e ' s e n t to the moving assembly factories. Rolls of the X buss bar material are brought t o ';he assembly f a c t o r i e s where they are e i t h e r notched o r mdif2c.l t o suit the assembly procedure. 'Be whole +Ad*& of one r e c t e m slat fs constructed at .the sene time. The requked n a b e r of horizontal buss bars i s unrolled, straightened, notched a t i n t e r v a l s of the aFpropriate
element spacing, an2 perhex the last stages of i t s material properties
.
t r e a b n t i s performd (e.& cold working) Attaching the buss b a r s t o the r e c t e m a ? r e is easily done by insulating f i t t i n g s . T3e 33 i s plugged i n t o 'the socket md a weather seal i s f i t t e d &t L5e iunc'don of +&e R3 and the sGcket entrance. With the buss iir&j f n ,$ace t h e socket/R3 i s then a p p d onto the buss. %e not,& c o t only @des tbe element s2acing but preveats ';fie socket from sliding alcng tine hec. m e socket shodd be able t o yrovide excellent su??ort f o r C,e X9 aqd should g i g *&e X k s s f a i r l y well.
Its square base melres C fu stable
i n all directions. Anozher function cf the socket I s the eZfect9-re zransit i o n ?ram t h e X3 contacts t o the E buss bars. This i s accom:Lished by two copper o r a l i i n u n &rips embedded i n tke ,lastic, e m s e d k'zlere it contacts the X3 and around the i n s i d e of the c l i p s *.ere the buss bars run. If keeping the diode f r o a overheating i s a problem (one that increases with a higher g a i n element) and the socket p l a s t i c does not kave a high enough thermal conductivity a heat sink can be embedded above *the diode during, socket fabrication (not needed with c w r e n t parsmeters i n 0.2 opinion).
The baselhe designs and the
RB implementation u t i l i z e both continuous
and modular construction t o varying degrees. In the baseline design 8 completed f o r e m e rolls out contiauously but i s modular beczcse it has a fi'ni€to le-&. An estfmate of the foreplane lengths was 20Q meters. (7)
In the RC implmentation the DC buss bars can be continuous f o r a large distance. !Be modular aspect n a t u r a i l y i s the socket and XB units. A pure continuous canstruction can be made stronger and, if not too complex, i a also faster. Huwever, mistakes hold up construction and, if maiatenance is ever required, pure continuous construction makes replacing parts difficult. ?me modular construction could also be d i f f i c u l t t o replace pBFts by the great numbers of - z n i t s . %s is also i t s main dfsadvantage
for arsembu e n c e large amounts of i n d i v i d u a l modules mt be handled. Bdt w e modular construction more e a s i l y e f f e c t s a_uaUW control measures and corrections of machinery nsl-ctions before making i t pa?? of "permanent" r e c t e m structure. PCB Y a g i s are e a s i 3 made and involve no more o r lit"Je e f f o r t ther mabing a FCB cS?ole. Fo? Y a g i s used over a grouot2 plane only t h e directors and & ' e dipole need t o be etched and this can be done oc tfie sane side. This. fs recomended t o i n s u e that the maxigai2 cf t h e Yagi p a t t e r c is i n "&e seme directlon as that i n which the E 3 is poiating. The ?cur d i r e c t o r s on t h e six element Y a g i can be o c e i t h e r side of t h e B j s l e because t h e travelling w w e follows t h e d i r e c t o r plaae which is p a r a l l e l t o t3e X9. :iatLt"aUj-, tke r e f l e c t o r on -T3 Yagis i s on the other side of the active ai-pole to a m i d the coriversicr? c i r c t d t r y . 3aselixe constraction Yag5s f a l l i 3 t o two categories : witk groun2 3025 a r e modifications of the Iiayttheon fore2lane plane and w',tb!iJ-,. b u t t h e Yagis wtthout t h e ground plane remove sone o f t h e dZfficiLties I n construction. The zmin difference i s that ';he Yagi fore3lane is s c p x t e d i n t h e active plane where the bulk of t h e weight 2s centered Lid n o t from the ground plane, as i n the baseline dipole, wMch poduces undesirable moments. Construction i s f a s t e r because L ? e r e is no groufid fl&qe t o f o l d i n t o t h e f o r e p k n e sten which is not done i n l i n e ~ 5 t h the cofitinuous l b d i f i c a t i o n is easiljr doze by srxippiq fabrication of t h e foreplace. on the sugport mast and tSen i n s e r t i n g the passive elezents.
Manufacturing tolerances are more important f o r Y a g i s tharr dipoles (discussed i n Sec. 2.2.2) since there are m r e d i m c s i o n a l variables (spacing, radii, and lengths of dl the antenna elements) which have t o be controlled and sone q u i t e closely, In t h i s regard the €CB antennas axe best and the b a s e l h e dipole reasonable, with three element and s i x element baseline constructed Yagis expected t o be samewhat d i f f i c u l t t o control. Y r y i t o l e r a c e s m e more c r i t i c a l a t high F/B r a t i o , as discussed i n Sec. 2.2.2.
Besides manufacture and construction, maintenance of the rectenna of a 30 yeax l i f e requirement needs t o be considered, It is expected that individual failure of elements will not accumulate to a large enough number t o require replacing the defective parts. Eowever, if no precaution i s taken t o prevent domino ef'fect failures o r multiple catastrophic failures (suck? as lightning storm damage) large sections may became inoperable,in which case it might be cost e-ffective to r e g a i r these sections. me cost of replacement is probably muck? greater than the i n i t i a l assembly cost per elemerrt, because of the special nature cf t b e repair. Mulitple land use and RFI considerations must a l s o be e d u t e d when evaluating re-pair options. A t t h i s stage all designs have similnr maintenance features, although the two plane continuous construc+doo najt be m r e difficult; Enviromer?tal factors come under two categories: s t z u c t u r d lo&ing and environmental p.otec$ion cmsiderations S t r z c t w s l loading from the weather and environmental considerations will - , t k recteoca location. I n tkis report environmenLd conditions t - n i c e l of :ior%%eestULted S%aates
.
a r e assumed. Fromloadir?g considerations, it i s most desirable t~ .have the heaviest member as the p r i m 7 supporting structure of t h e elemerks.
Ex',ensfons
should be lightweight and strong t o withstand loading. %us, long Yagis i n both W e s of construction have the mst proLlens with loading. B e r e fore, t h e socket i s =de r e l a t i v e l y stronger (or lotiger) In the XB case and the support mast is made more rigid i n the baseline case. *TB d i e l e c t r i c cui from in between t h e directors on the Yagis wLU a l l e v i a t e mcch of the loading especially on the long Yagis. C f course, a central d i e l e c t r i c s t e n of good width is l e f t tn corrnect %ne directors.
Ir, shor', ?agfs
without ground planes, the r e f l e c t o r helps cou3terdefght 3.e director,.
53
Ihe foreplane shield i n t h e baseline constructed designs iz h o r i z o n t a l f o r construction continuity but i s acted on by g r a v i t y and other environmental loads. mus environmental protection constrainSs require t h e solid cons'truction of the baseline designs. (me baseline constructed Yagis without the ground plane muld withsfand losdinq much b e t t e r because the kavy f o r e w e i s supported i n the sctive &pole plane, where most of t h e weight is, instead of at the ground plane. The dipole and the three element printed circuit Y a g i i n the socket forms the stronger designs. Ihe methods for envirormrental i s o l a t i o n should protect t h e circuitrg
from weather and keep radio frequency interference (€ET) generated by the harolonics of t h e diode A.cm escaping w h i l e allowing adequate cooliog of the diode, Easeline construction performs dl three functions reasonably w e l l , From the discussion on R B socket design weatherproofing and the heat sink are no problem. RFI, however, would not be stopped by the p l a s t i c , It could be reduced by using a high l o s s p l a s t i c i n the dipole sockets and ground plane Yagi sockets or possible i n the exposed r e f l e c t o r sockets as well. With existing d i e l e c t r i c s there would be too much loss at t h e fundamen"d f'requency, 2,45QIZ, if second and higher order harmonics were-to be suppressed.
However, i f a d i e l e c t r i c could be developed with
a low pass capability, lar loss a t the fundamental and high loss at higher frequencies, and had good s t r u c t u r a l q u a l i t i e s , r a d i a t i o n of RJ?I would be g r e a t l y reduced. Of' course, i n a l l cases h a i m n i c reradiaticn is of s p e c t r a l concern, (7 1
Because of the l a r g e amount of real estate required f o r one rectenn8 the idea of m l t i p l e land use has received a t t e n t i o n as a means of iccreasing the monetary output of each acre. Each different land use w i l l have i t s own d e f i n i t i o n of how much microwave r a d i a t i o n or sun shading i s tolerable. But i n ~paqycases t h e amount of microwave power d o w e d t o pass through t o the ground will probably be t h e same f o r ecological reasons (or for
.
favorable biological consequences) However, some economical considerations aad enough leeway i n the safew standards may ?rove othe&ise, In any case, if shadowing i s important the methods used t o achieve c e r t a i n mic,-owave i n t e n s i t g l e v e l s w i l l a f f e c t the amount of sheding (e.g. d;** A erent
mesh dimensions i n the ground plane o r going t o long RB Yagis with high F/B r a t i o s ) .
54
Shading is a d i f f i c u l t factor t o compare a t & a s stage of
OUT
design.
F o r t w t e l y , a good g~oundplane with 33 dB t r a n s d s s i o n loss has about transparency. (I8) me foreplane shield has a sizable amount of shading and fortunately is horizontal in an east-west orieztation. Recall that the rectenna f a c e s sou* t m a the satellite i n maz zero incuriation orbit. I n the R E implementation, since the DC b u s m s east-west for
constructional ease, the RB's must be aligned north-south t o amid inducing t h e varying currents in the buss. North-south 33's w i U have more shading. I n general the antenna designs wfiich contribute the mst
shading per element have the l e a s t number of elements. To f i r s t order *&ese may cancel but &&eless bussing involved favors the more directional elements. Clear
[email protected] fs used for the PCB socket 5ut since the X B i s b u r i e d deep i n the socket there is about 5% of the socket stZll opsque. The development of a clear d i e l e c t r i c f o r XB applica';ion would LLev'Late * a s problem if it became considerable enough to j u s t f B the resemck am? development costs. This would find most use i r , the 10% Yagis which cannot a f f o r d t o have too mch d i e l e c t r i c cut away between the directors. 2.3.3
Cost Com_oarison
Alt3ough ap-mecfable design work is needed t o evaluate the a??roLzaticns and judgements discussed throughout Chapter 2 , it is desirable t o Xovide as eccurate cost estirnate es pcjssfble t o serve as a gside i n p r o e r l y d l o c a t l x g f u k r e resources. Th,is section Cescrfbes tne cost met3odologr and r e s u l t s kqA& t h e eight rectenna elements described fr? Section 2.3.2. m e cost ardysis followed the percect estimates 3roviderl by ?aj-tfieon f o r t h e baseline rectema. (7) For SaseUne t y p cocstructed rectennas With Yagi elenents, the r e s z l t s are expected t o be of siaccuracy. ,Although more Yagi s*ructi;-ral design is needed, a d e t i o n a l material and construction costs a r e not expected t o be too sfgnificent. However, with printed c i r c u i t implenentation cost estimates a r e more a 2 p r c x i a t e . I n p a r t i c u l a r the socket costs and Dc buss b a r sizi,lg requirements are d i f f i c u l t to quantif"y at this time, as structtlral loading requirenents
have not bee2 designed. n u s our r e s u l t s s3ould be considered pre2,iclinay.r only, with e d C t i o o s l work certainly necessary.
55
However, even with these appropriate qualifications, we believe t h e results fndicate which designs are most worthy of f u r t h e r investigation, In p a r t i c u l a r , t h e expected strong c o r r e l a t i o n between rectenna element d i r e c t i o m l l t y and rectenna cost is clearly indicated. The rectenna element density used for the cost estimates a,re shown in Table 2-9 for the half-wave dipole and various Yagi configurations. The gain values have been discussed i n Sec. 2.2.2 and are equal (except for round off differences) t o t h e extrapolated performance given i n Table 2-8. Other values i o tkis key t a b l e can be calculated from geometric considerations, assuming elements are placed i n a triangular g r i d as i n . . the baseline, The resultaat costs obtained are presented i n Taole 2-10. 'Be trend toward lower cost with increased rectenna element gain i s c l e a r l y apparent. The comparison between baseline construction and printed c i r c u i t imFlement a t i o n i s less apparent. The printed c i r c u i t esthates are based upon less detailed design, but these r e s u l t s do not i n d i c a t e a s u b s t a n t i a l reduction w i t h priated c i r c u i t implernentatioo. O d y i f socket and X buss bar cost can be reduced w i l l a b g e cost advantage result. These w5U -pobably be possible only wit3 c s r e f u l s t r u c t u r a l designs requiring l e s s material usage and l o w cost. manufacturing. Cost estimates for t h e printed s i r c u i t board, XB sockets and r e l a t e d Iy: buss h a r are given i n Table 2-U. t k o u s h 2-13 respectively. Note tkt r a w material c o s t s OXQ are included, as manufactitring compleXi,C,- and related c o s t s a r e assmed low. Zowever desigz refhemezts $0 redirced cost, such as eliminatior, of gold f l a s h i n g i n non contect areas, reduci-.; d i e l e c t r i c t f i i c h e s s below 1/15", and reducing buss b a r diameter, are not included The feasibility of tbe two l a t t e r , c r f t i c a l f a c t o r s depends on s t r u c t u r a l considerations beyond the scope of the present p o g r a m Fromthese cost e s t i s t e s it is apparent that the E buss b a r , ground plane where needed and GaAs diodes dominate the rectenca e1emer.t c o s t with printed c i r c u i t implementation. However t h e socket desigr? and c o s t i n p a r t i c u l a r Reeds additional work ane t h e E buss b a r s t r x c t i i a l support warrants f u r t h e r study. t j i t h the baseline constructed r e c t e n m s the l a t e s t I,ay%heon(? 1 estimates were used, and additions to fora t:?e k'agI-Xa elexer.'-,z
56
!Table 2-9
Rectenna Element Density Used i n Cost Estimates
3 element
Half-wave
Mpole
ground plane
Yagi without ground plane
with
*
Gain with respect t o isotropic (a)
4.5
1002
8.4
Gain r a t i o w i t h respect t o i s o t r o p i c
4.4
10.4
6.8
52
192
6 element Yagi wjthout ground plane smaller s i z e larg& s i z e
1l.n
12.7
. 12.7
18.4
123
150
218
81
67
46
Effective Area, Ae 2
(cm /element) Element Density (NO. of elements/m )
(eoo)*
Density Reduction Factor
1
2-37
1.56
2.87
zemer,t Spacing (cm) (on trian&a.r g r i d )
7 -8
u*9
9.7
13.2
15.9
Raws of X 3uss/m
14.9
9.7
11.9
8.E
7.3
Ae = GX
2
Ln
---
G
is the gain r a t i o w i t 5 respect t o isotroFic.
A~ is the e f f e c t i v e m e a of the rec2enna element.
The baseline had all other parameters derived from tne elenent spacing, the reverse of what is done here f o r the r e s t of the elenaents.
*
used fn Raytheon estimates.
57
A.
FCB Implementation 2
(costs are given i n $/m )
3 elemert Yagi
Half-wave
Dipole
elem. Element Density (7) 192
with ground plane
without ground plane
81
123
6 element YaRi without ground plane (averwe size) 57
Socket,
Dc buss bar
1.55
2.78
1.81
2-23
1.91
1.91 -
.oo -
$5 85
$4.35
$3e 7 7
$2.51
$1.92
$ -81
$1.23
$ 957
$7.77
$5 -16
$5 -00
$3 08
PCB ( l e s s diode)
Ground
plane
cost/!TI2 Diodes at $.01 each* TOM Cost/m2
3.
Baseline TSrpe Construct* 2 (costs are gi :en i n $/m )
I'I-wave Dipole
(e) 192
zlemer?t Density elem m" Foreglane Core Aluminum Shield/ Structural Member Yagi-Uda Additions
-00
3 elemeD%Yagi with ground plane
without ground p h n e
6 element Y q i +,thou';
ground plane
81
57
$3.13
@.%?
$1.09
2.14
1.40
64
.00
30
.?6
1.91 -
.00 -
$7.18
$5 008
$2 49
Diodes a t $.01 each
1.92
.81
.57
T o t a l Cost/m2
$9.10
Ground Ehne
1-91
2
Cost/m
*
$5.06 $5 89 * Large c o s t uncertainty with probability of lower c o s t pac:ka,-e x i t h E 3 inpleEentatio2 Table 2-10
Overall C o s t E s t h a t e s
!Table 2-llA Cost Estimate f o r P r i n t e d Circult Bards lislf-wave Mpole 2
Area (in. )
2
.015
3
element Yagi without ground plane ground plaae
6 element Yagi
5.54
6.25
15.6
volume (in. 3 ) Weight (lb.)
5
of Total c o s t of 1. d i e l e c t r i c s
2. copper
3- gold
NOTES:
58.4
70.3
1.2
0.9
1.1
0.9
40.4
28.8
32 -2
24.9
74.2
66.7
1. Diode cost not included.
2. Copper is SmBU percentage of cost ( a s s d n g copper etched from P.C. boards and reused at nc cost). Therefore more &rea on the boasd coQd be used to reduce losses i f no f l l e f f e c t s fro= chazging impedence levels
3. Gold flashing may
o n l y be needed a t the D.C.
contacts.
The d i e l e c t r i c
coating may be adequate t o protect copger from corrosion.
4.
(long)
length =1.25X
with
Biggest cost f a c t o r i s the d i e l e c t r i c (+&chess of 1/16" is assumed). Some cutaway i s possible between the d i r e c t o r s of Yagis, b u t not a t t h e base since this region fits i n t o socket. Cutting away of dielect r i c is not considered in t h e calculations so Yagi costs would be less. Dielectric is assmed t o be recyclable.
59
Table 2-UB Detailed Cost Estimate for Half-wave Dlpole with Printed Circuit &pleaentation
area:
(1 cm. x 2.5) + (1.5 x 7) = 13 an2 = 2.015 in. (2.015 he2) (1/16 in.) = . ~ i6n O 3
volume: copper
2
-0
( - 5 m i l s thick): (2.44 i n . + 1 in.)(.O50)
gola flash
--
0
(250 A ) :
( 2 . b in. + 1 in.)
(.@O)
x ( . 5 x 10-3in.)
(loo6) =
=
-6
8.6
.15 x 10 in.
x 10-5 in. 3
3
4 (.050 x .050 x .C25) = .25 x 10-3 in. 3
capacitors:
two g0U bond wires (1 mil.): 2 ( ~ ( . 0 0 1 ) ~ / 4(.050) ) = 7.85 x 10-8 in. 3 dielectric coating:
(2.015 x .005) = 10.1 x 10-3 i n . 3
Weight Calculations = (specific gravity) (density of water)
dielectrics coDper
-- 3 (62.14
-- 8.9
(126+ 10.1
-L
.25) 10"
(volume)
= 1.46 x 1 0's
15.
(62.4) (1/12)3 (8.6 x 10-5in. 3) = 2.8 x loo5 l b s .
cost -
dielectrics (polyethylene :k/lb.)
gold ($200/ounce )
5E.i
$7.4
1.57
10-5
-5.12 $1.27
of total cost
1.2
L0.4
10-3
or 0.l27k (per E3 w i t h o c t &iode)
60
Table 2-12 Cost Estimate f o r FCB Socket . materia
- &kin
(an a c e t a l r e s i n )
: specific gravity
1.43 (acetal p l a s t i c ) ; e s t h t e d . c o s t
: $.15/lb.
(0.04914) ($.15) = $0.00737
or
.737#
! b o 1 4 2 " aluminum conductor s t r i p s , 20 m i l s thick; specific gravity = 2.7; cost $.80/lb.
v o l i i = (2) (1.5) (l/8) (.020) = C.0075
inO3
weight = 0.0073l 15.
cost = $0.000585
or
.059#
resistance = 1.33 x Weather Seal:
silicone rubber; for s o f t commercial rubber 69 lb./ft.';
type 1 (w/G.P) TOTAL COST NOTES: -
.4v#/socket
cost $l.OO/lb.
type 2(w/o G.?.)
913k/socket
1. S e a l may be considerably smaller i f o n l y sealed where c i r c a i t r y had t o pass t h r ~ 3 hentrance s l o t of t h e socket. 2. Aluminum conductors could be made I E g e r t o lesse:: losses. 3. Heat sink, i f needed, i s not considered.
61
Table 2-13
Cost Estimte for DC Buss &s
for Printed Circuit Implementation
Two aluminum rods w i t h 5 x 3 diameter.
volume = 2(n) (.25)
2
(7.77 cm spacing as i n half-wave
(7.77) = 3.05
dipole rectenna)
weight = 0.0185 lb. = 8.3 grams. cost = (0.01815 I b . ) ($.80/lb.) = $0.0145 or 1.454
-
resistance = 2.24 x 10g4~/elementspacing
Half-wave Dipole
Element Density 2 ( elem. /m )
1%
Element Spacing (cm. )
7.77
Cost/element of
1.456
3
element Yagi
$2.79
e;z;negagi
without
ground plaae
ground plane
specifications )
123
57
9.67
14.5
1.816
2a
61
u.9 2 *23k
Dc buss 2 Cost/m of rectenna
6
with
$ 1.81
$2.23
$1 55
6
62
were included. Table 2-14 summaxfzes data provided by Raytheon while Table 2-15 thrvugh 2-16 provided detailed information f o r t h e three eleaents Yagi designs. Six element designs c o s t s were obtained i n a similar fashion (although d e t a i l e d calculations not presented) As w i t h the printed c i r c u i t elanent costs, the strong dependancy 02 cost per u n i t area upon element gain (or density) is apparent. Thus we concluee that 'LEER3 IS A LARGE REXTENNA COST SAVING H)SSIEXE BY FURTHER CONSIDERATION OF MXtE D I # E C T I O W RECEIVING EIEMEXTS. I n addition, 3-t appears that c o s t s are sixnilax f o r p i n t e d ' c i r c u i t and baseline type construction. Altnough these cost e s t q s t e s are derived from a d i f f e r e n t basis, we conclude that the 2 t o 4% s a c r i f i c e in conversion efficiency with printed c i r c u i t im$ementation does not appear t o be overcome by a l a r g e cost advantage. That i s , -1el;lentation
baseliae m e construction appeass preferred over p r i a t e d c i r c u i t inolemen-
t a t i o n at this time. However a s t r u c t u r a l design i s considered worthwhile t o s p e c i e more rigorously DC buss b a r and ICB socket reqcirexnents. For example, i: the DC buss bar cost can be reduced by a f a c t o r of 2 (by :tducing diameter below 5mm o r developing an appropriate composite material), E 3 implementation would be 25% below t h e baseline type construction cost.
If this cost reduction was obtained, efftciency-cost tradeoffs would need t o be considered i n more detaiL. Another key f a c t o r that could influence the tradecff c s h e e i printed c i r c d t and baseline type construction is diode package COST. A large uncertainty i n projecting rectenna cost i s the diode estimate. %en glassed . a non-negligible cost ccmpared t o the diode s t a n d alone packages wUhave diode chip i n the q u a n t i t i e s needed.
XB iaplementation i s expected t o be
l e s s expensive since t h e chip could be d i r e c t l y bended i n place. While more work i s needed t o evaluate cost f a c t o r s i n t!-e quantities needed, t N s advantage of X B implementation could be significant.
63
Table 2-14 Cost Estimate f o r Baselin2 Constructed
Half-nave Dipole (1% elements/m2)
Foreplane Core
Quantity/m2
Cost $/m2
- aluminum in common buss bar
418
- aluminun i n dipole actennas
142 g.
.26
-
768 units
.
g.
and microwave c i r c u i t
common assembly pins a d capacitors mamde from AL02
- enclosure ca?s f o r dipole
LO
384 units
85
384 anits
.85
made frornA2.0,
C
-
seals f o r dipole caps
- steel
TOTAL
$3 1 3
TOTAL
$7.18'
G o a n d plane
Diodes a t $.01 ea.
64
Cost Estimate for Baseline Constructed 3 Element
Table 2-15
Yagi w i t h Ground Flane (81 element/n2)
-
Quantitlv/n2
Foreplaae Core almdnum i n comaon buss bar and microwave circuit
273 a.
$ -50
(factor = 907
- duninurn i n dipole antenna
cost $/xu2
= .05) .11
58 8 .
-
(factor = 81 = -405) 200
- common assembly pins and c capacitors made from A102
324 d t s (factor = -405)
- made enclosure caps for dipoles *om A102 - seals for dipole
.16
162 u a f t s (factor = .US)
35
zps
790 g.
(factor =
.$>
81 units
7c
0-/
at . o .232#/eies;.
- s t e e l ground plane
2k23 g.
mTAL
- Diodes at $.O1
ea.
l.9L $5.08
65
Table 2-16
Cost Estimate for Easeline Constructed 3 Element 2
Yagi without Ground Plane Cost (123 element/m ) Foreplnne Core
wtity/m
- aluminum i n common buss bar and microwave circuit
3% e* (factor
- alumiaum in dipole ante-
2
cost
$/El2
$ -62
- r9 u . 9 = ,S@) 87.3
g.
.16
= -615) (factor = 3 200
- comm assembly pins and
capacitors made from AL02
- encloswe
caps for dipole made
from A102
246 units (factor = .615)
*53
- s e a l s for Si-pole czps Additions
-&knyun
shield/strcctcra member
- a l d m n passfve elements (factor = 2 ( 3 ) = 1.23)
- Delrf n suppor - Diodes at $.Gl
cI
masts
ea.
I23
Un2t.s
33
66
3.0’ FUWER COMBININ3 EVALUA!IIOI? Most of the rectellna development e f f o r t t o date has properly emphasized high RF t o DC conversion efficiency. (1-3) As a result experimental arrangements have been somewhat idealized as far as SPS operation i s concerned. For example flat rectenna a r r a y s have been t e s t e d rather than serrated rectennas and DC combining networks have been arranged to insure, as m c h a s possible, that devices to be combined operate a t the same power level. In t h i s part 09 our program, we emphasized an evaluation of the power conbining i n e f l i ciencies expected when m81)3r RF t o M: conversion c i r c u i t s (- 10,000 t o l,OOO,OOO) share a common load. This consideration, i n h e r e n t i n SPS operation, had act bee2 e d u a t e d previously. While the power variations due to d i f f r a c t i o n from a serrated and the power variations due to the power be=
t ~ g e r ‘ ~have ’ been quantified t o some degree, *%he impact With DC load sharing on o v e d RF to Dc conversion efficiency had not been e v a l i t e d . I n additior, the relationship between incident pmer variations, conversion circuitry
operation and Dc load sharing had beendelineated as a sowce of S?S power output fluctuation, which i n t u r n sects power g r i d i n t e r f a c e retyireaents. (19) In t h i s task we made a good f L r s 2 order evaluation of the r e s u l t a n t ?owe? combidng ine?ficfency, while developing a methodology which can be appliee d i r e c t l y i n future mrk. The p r i x i p a l concept of our evaluation i s t o a t i l i z e a 3 m e r de-pendezt output equivalent c i r c u i t of the conversion c i r c z i t q , obtaimble k y it&.?ng the IcaC iapeciance a t each ?F‘ -power level, i.e. a load l i n e a r d y s i s . 3 e q l o y %histechnique a c i r c u i t model of the r e c t e m element conversfoE circyiitr;. i s needed. I n Section 3.1 t m r e c t i f f e r c i r c u i t m d e l s (a c c q x t e r S-El a t i o n model an2 a closed f o r n model} developed i n & a s prcgrtm a r e ?resented, followed by a discussion of the load l i n e analysis in Section 3.2. These two sections contain a detailed treatment of +&e basics for t h e pwer
.
combining aWysis I n Section 3.3 the parer combining analysis methodology and r e s u l t s obtained with bot9 computer simulation and closed f ~ m models a r e presected. These r e s u l t s focus on evaluaticn of s e r i e s versus 2 a r a U e l conbining inefficiency w i t h numerous r e c t f f i e r s sharing a cornon load. X ~ C G U S
67 d i s t r i b u t i o n s in power density axe presented. In Section 3.4 the e f f e c t of these results on power module s i z e co&Lraints and S2S recter-mt design &e presented.
3.1
RECTIFIER CmCUIT MODELS
I n t h i s section two independent c i r c u i t models of the iW t o DC con~(,+Ap r i n c i p a l results obtained version c i r c u i t r y are preseated, d o r w wlth these models. A v a l i d -del of the coaversion c i r c u i t r y is necesssFy in order to obtain a reasorably precise output equivalent c i r c u i t f o r t h e load l i n e aaalysis. However a c c w a t e modeling a non-linear c i r c u i t can be a consuming and resource absorbing underta,l~Zrq.(~~~) Our approach focused on d e v e l o p n t of a detailed computer simulation model using a general non-linear c i r c u i t ,program (Spice 2 ) and a parallel develo-pment of a closed f c m n c i r c u i t model. The detailed computer s i m l a t i o n m o d e l contains 30 device and c i r c d t p a m e t e r s and closely represents an a c t u a l rectif'ying c i r c u i t elemest. It differs frcm previously developed ccnputer s i m l a t i o n models of the rectenna elenent (3'8) i n that a geJera-' -rcuit program i s =sed r a t h e r than developing i n t i i v i d d z e d code f o r the p a r t i c u l a r c i r c x i t . ~ I L S resources needed i n development of the model a r e reduced appreciably, a'; the expense of less e f f i c i e n t operatioc f o r the g a r t i c u l w circqdt. Attention was focused on obtaining baseline CJ-X efffciency gerfc-rzlance w i t h Q q i c a l packaged diode characteristics. The r e s u l t s e r e ?resented in Section 3.1.1. i n addition closed fom models of the conversion c i r c - d t r j .*re investigated. Ini+dally "be -pur?oses cf t h i s work - a s t o a i d 35 antierstanding r e c t s i e r operation and t o develop prelimimry cutgut e q d v a l e n t c i r c u i t models f o r &&e power combining analysis, thereby allaJ;,ng a first order evaluation of our main task. :%ever t h e clcsed fcrzl a W j s i s results i n an output equivalent c i r c u i t i n e x x d l e n t agreenent wi*A the canputer sinnllation m d e l r e s u l t s , indicating a a t n i l i d r e s u l t s CEC -be obtained with this simpler model. The model is presented i n Section 3.1.2. 3.1.1
Computer S h d a t i o n Model The b a s e u n e type r e c t i f i e r computer s i r n r l t i o r z =del
fs Ze2icted i n
Fig. 3-1. i.?itidly we selected diode p a r m e t e r s 2nd asscciete2 m u z t i n E
68
P
c
a 0 0
n
n
0
s.1 F
n
N
8
k y s2o "a cu
33 d
II
II
rn
rl
u
3
0
u
2
-o_.
0 1
l-l
n Ln
r" url
I 4
G
H
E k r-zz c
0 4
n
II
II
II
c
ORIGINAL PA(:!. 'i OF PWR QUAI.! I'
4
E
'
p a r d t i c parameters. These were selected t o be comparable t o that of the baseline r e c t i f i e r with the present mounting configuration''' and were held constant throughout our investigation. Also selected was a 75 ohm antenna resistance, comparable to an i s o l a t e d X/2 dipole receiving element. For sfnrplfci- it was decided t o keep s 75 ohm impedance level throughout the circuit. Although higher efficiency may be possible by impedance transforming to a higher d u e , detailed efFiciency optimization was not within the scope of our e f f o r t . Wi+& a 75 ohm impedance level selected, 8 ?ive stage lumped f i l t e r was used a t the input and two s t a g e sm0cthin.g f i l t e r a t the output.
The
input fflter i s described i n d e t a i l i n Sec. 2.2.1 and & ' out3ut is desigDed similarly.
Obviously the outgut filter has a l m r frequewy cutoff,
with a value depending upon e. tradeoff betwees inauctence required a f i l t e r r e j e c t i o n a t 2.45 Qiz. me output f3lter has a 3 d3 ripple, cutoff frequency of 612 I43 and -provides 30 d9 of fundamntd. irequency rejectLon -when operated between 75 ohm izcpedances. I n i t i a l l y results obtaine6 W L t h the model i n d k a t e d low efficiency u n t i l i n p J t and output transinission X n e s were added between tbe m u t e d diode and filters. me trammission l i D e c c ~ t r o the l phase of r e f l e c t e d signals and i s p a r t i c u l a r l y b p o r t a s t a t the input. I n nany sfrm2lntions, a five sectior? G C network was used t o replace the transmission, X n e , *a reduce program running time without s a c r i f i c i n g overail clrcvdt perfonaace. Wi*& t h i s m d e l , the incident ;owe= is -mried by changcg t!x value o f t h e a q d i t u d e of - . I t a g e source. As a r e s e t , +&e I?O"?inea- c S r c d t pezfonnsnce changes, - m i n c i - e due t o &%e Zode trirz-on voltage of apgroximately 0.8 vclts. A p l o t of conversicn efficiency versus incident m e r is shown i n Fig. 3-2, i n which no c i r c d t o r Ciaie -parameters na-ve been varied. This decrease i n ef'f'iciency ~ 5 t hdecreasing power i s greater tbar? experf?nentdly obtained h e n c i r c u i t is r e o g t h i z e d at eacn ,per l e v e l , 3 )
as described previously. ( 5 ) Because of resource constraints, optinization a t each power l e v e l was not performed. Ifowever we believe that t h e parer combining inefk'iciency i s r e l a t i v e l y insensit2je t o t U s f c r t h e r c 2 t h i z a tion.
70
\ \
\ \ \
.
Besides t h e conversion efficiency one can obtain u s e m voltage and current waveforms as w e l l as Fourier *.nalysis of these waveforms. For example Figures 3-3 and 3-4 indicates packaged diode and diode chip wave*
forms of voltage and current respectively a t a 1~ power lev&. The chip reverse voltage exceeds slightly the peak d t a g e amplitude of the source ( 2 9 compared t o peak amplitude of the open circuited RF source voltage of 25V); w h i l e the chip current (conduction p l u s displacement) exceeds 400 mA during the middle of the 180' conduction angle (compsred t o peak amputude of the shorted circuited RF source current of 333 mA). !Be waveforms w e obviously rich i n harmonic content, partly a t t r i b u t e d t o resonances from packaged diode p a r a s i t i c s . Fourier snalysis of three voltage waveforms, namely t h e output voltage, the diode package voltage and the diode cbip voltage are presented i n Table 3-1, along w i t h the input current.
The effectiveness of the smoothing fflter is apparent as the Dc voltage i s sjmilar w h i l e t h e fundamental component of the output voltage compared t o that of the mounted package voltage i s reduced by a f a c t o r of 2 6 ( 1 2 . 7 9 ~t o .497V) o r 28 dB, with harmonics reduced s t i l l further (e.& second harmonic 40 dB). Also the three values of DC voltage a r e sUghtly d i f f e r e n t , due t o numerical appro-tions i n the program and f i n i t e program running t i m e . The approtdmation, as w e l l as the effect of t h e input f i l t e r , can be seen from the input current, which contains harmonics 50 dB beLm t h e fundamental. !IUS
i s about the precision of tfie program as used by us, as can be seen by the erroneous f i n i t e value of E input current (should be zero). It should be mentioned that a d d i t i o n a l investig&tions possible with t h i s program such as efficiency optimization a t n o m i d p a r e r and lower powers, efYect of gackage and mount p a r a s i t i c s , s e n s i t i v i t y of performance t o diode and c i r c u i t pasameters and harmonic performance evaluations were beyond the scope of our program. Emphasis was ?laced upon p r i n t e d c i r c u i t board implementation evaluation (Sec. 2.2.1) and developicg a model that provided efficiency c h a r a c t e r i s t i c s s h K L a r t o that demonstrated e-erimentally (for the power combining analysis). As a r e s u l t of these and s i d 2 . a ~consiierstions, we coocluded that +be computer sirmllRtion model developed results i.n performance c h q r a c t e r i s t f c s sizclLlar to t h e Itaytheon baseline. This m d e l was use? to d e t e d n e the
72
d d
I
I
I
73
W
a
8
f
4-
cu
b c
s,
Table 3-1 (.
nquter Shnalation Output
fclr
Baselbe '4rpe Rectifier (&weer. In =
-
1.04
74
W)
'Ourier Coqponents of !transient Response of Output Voltage DC Conponent = -8.052D 00 (volts) HarM>XliC Frequency Fourier Normalize d Base Normalized Component (E) Component NO.. (m3) Rase (Ded 1.OOOOOO -U4.98 0.000 4. W O ! . 1 *. 2.450D 09 5 658~02 2 4.gOOD 09 0.113826 9.055 123.963 7.350D 09 1.14'7D-02 0 023079 61.376 176.284 3
.
4
9.8aOD 09
3.n4D-03
0.00'7472
0.004274 2o U 5 D - 0 3 1 . 2 2 a 10 2.U5D-03 - 6 l.b'[OD 10 0.004255 0.005237 7 1.715D 10 260~03 0.001295 1.96OD 10 6.4380-04 8 7 756~-04 2.205D 10 0.mi560 9 'ourier Components of Transient Response of Mounted hckage Voltage 5
.
.
HarmodC
MO 1 2
3 4 5 6 7 8 9
Frequency
Fourier Coqpnent
Normellzed Component
09 og
1.279D 01 5 . 6 m -00
1.oo0Ooo
09
2.533.00 l.075C 00 Ll62D 00 1.233 00 2.458D 00
(m
2.4500 4.gooD 7.35m 9.800D
09
1.22513 io 1.470D 10 1.715~ io 1.960D 10 2.205D 10
0.U3533 0-198223 0.084ogg 0.090878 O.O963?9
-15.882 -47 665 23.977 2.376 14.474 23.437
99.026 67.243 138.886 117.285 129.382 138.346
- DC Component = -8.1943 Base
(Bg)
59.228 -173.104 -U479 152 082
. 101.2LI:
Nomdized -se (be> 0.000 -232.332
-171.707 4.854
.
k 015
-206.359 177.288 118-059 0.192248 4 57-02 0.003575 -U400 -70 628 4 473-01 0.034980 -1jl.388 -190.616 'ourier Conponents of Transient riesponse of c%p Voltage E Component = -8.269 00 (volts) Ra3lnoIliC
Frequency
Mo 1 2
(E> 2.4503) 09 4.gOOD 09 7.350D 09 9.800D 09
.
3 4 5 6
..
Fourier Coqmnent 1.3$D 01
6.437D
-147,130
.
-
No,+.melized Cmponeat 1.000000
00
0.460942
1.6233 00
0.116224
Phase (De61
.
51 526 -173.i.199 -112.665
.
NomeUzed Base (Be> 0.OOo -225 025
. -78 .881 49.204
-164.191
2.093D 00 -27.55k 0.19868 1225D 10 1.7493 00 0.l25243 100.730 1,470D 10 7 46oD-01 -145.123 -196.649 O.O53U 7 1.725D 10 4 952% 01 0 035u1 -5.93 -57.40 8 1.960D 10 3 836~01 0.027&~ 159.988 106 462 9 2.205D 10 1.35OD-01 0.00g670 -119.93 -170.859 'ouzier Components of Psnsient .Response Source Current Dc ConFonent = -1.318D-04 (Amps)
.
HgsmoniC NO
1 2
3 4 5 6 7 8 9
.. .
Frequency
(E>
2.450D 4.goOD 7.350D 9.8OOD
09
09 09
09
1.22513 io 1.47013 10 1.715D 10 1.?6OD 10 2.205D 10
.
-
F'ourier Conponent
Normalized Conponent
1.588001 4.186~44 1.699-05 1 084~-05 4.-06 9.712D-07 4.553-06 7.6510-07 4.034006
1.000000 0 002635 0 OoOlO7 0 000068
.
.. .
o.000030
. c moocg 0.
.
Pase
(as1
179 579 33 824 71:.036
.
-12 -489 107 372
Nordzed -%se (Deg)
0.000
.
-145 856 -105.643 -1% 0168 -72 307
0 000006
5.270
0.000029
-66.767
-174.410 -248 4 6
1711.106
-5 573
77.587
-102.093
WOO25
.
00 (volt
75
equivalent circuit as a f’unction o f power l e v e l -,n Section 3.2.2 and the power combining analysis of Sections 3.3.2.1. HCJ. 3.4. ( 11.;i!:i>
Y P O
with x,
y
and
r inkm
PD in */an2
10
n (u
6
3 W
k
8
3
1
L
b
1 1
I
I J 1 2 -3 4 . 5 Xcrizoctal Distaace-x (-b) Figure 3-22 Fbver Beam Taper 'Jsed in R m r CcmSiFlng Evaluation (Horizontal Distance is Parallel to E Combining 3uss or East-jlest while
I
-
h
110
3:1 t a p e r while near the middle of t h e r e c t e n m combining i s r e s t r i c t e d Starting . .-. from the edge of aqy y value, combining i s ' p e r f o r m d f'rom 1 to 3mW/cm2, 3 t o W/cm2, 9 t o 18mW/cm2 and 18nW/cm2 t o the center, as presented i n the second column of Table 3-5. !!!he range mrhs are obtained from Figure 3022. Using the element spacing and knawn pawer levels the number of elements in a row and power output per row are easily calculated. l i n e a r i z i n g the taper over the ranges indicated one can use "r;he power combining efTiciency data presented in Figure 3-21, as the power distributdon over aay combining now i s uniform. Note that the power loss (power combining inefficiency t i m e s power output per raw) i s nearly the same i n power densitg ranges from 3 t o 18W/cm2 and i s lower neaz the center of t h e rectenna (where parer taper i s s m a l l ) and at the edge (where power collected i s s m a l l ) with o m assumptions. The power combining inefficiency averaged over the rectenna is about i n t h i s case. However, the parer output per row i s only 1 t o 19 KW. L-1 M i pwer module are desired, the rectenna slat width gets prohibitively l a r g e a t the edges of the rectenna (= 5Om). I n addition t h e 1MM parer module would probably operate at -1 kV and 1kA. Conductor l o s s e s a r e expected t o be i n t o l e r a b l e at such high current l e v e l s wi"h the baseline elements. That i s , an a d d i t i o n a l conductor buss would be required t o sccomodate such high currents. it is c l e a r that even with more modest power module s i z e s of 100 iCW, numerous columns must be ccolbined, so that dfffractian induced variations wiU. cause an increased power taper. While nct included in our q u a n t i t a t i v e evaluation such a f a c t o r must be considered i o f u r t h e r studies of row combining. For combining in concentric e l l i p t i c a l rings, the ,mer combining t o a 2 : l taper range.
1.a
inefficiency i s appreciably reduced f o r any sized mdule, at an expense of rectenna slat interconnection and decreased rectenna modularity. In f a c t , f o r any conceivable power module size, the power combining inefficiency due t o power density dffferences becomes negligible if coinplete concentric r i n g combining is used. For example, i f 50 MW modules a r e considered, . the rectenna is divided i n t o 100 concentric rings of average width of 50 P.
m
5 $8 c u r l
t
cu
0 .
4 w 2 0
.
. .
= P I 4P-
8 8C c 8N L8n 0 z
d
;IKi-lL PAGE P W R QUALI'J
2
.
Q\
m I
'4
N
d
d
Q
m . I
scu
m h
d o
N
tm
More precisely the width of the r i n g varies from 260 m a t the center of the rectenna, t o 41 m a t a 1 lOn radius, t o 29 m a t 2 kn~radius, t o 28 m at 3 h, t o 67 m a t 4 km, t o 260 m a t t h e 5 Inn rectenna edge with the '
The outermost ring would have a m e r combining befficiency of only O.&, w i t h a rapid decrease toward the center. Ihe power loss over the recteMa would be below O.l$. IVaturslly the effect would have t o be reevaluated if pie sections were considered instead of complete r i n g s . W e conclude that the power combining inefficiency i s of considerable importance d t h row combining and will affect rectenna modularity and construction technique (continuous construction less desirable than " b i l l taper i n Figure 3-22.
board" Wpe construction and assembly). Ring combining is certainly preferred from the viewpoint of power combining i n e m c i e n c y due t o power beam taper but propagation induced &+ations must also be considered. In addition, conductor inauced l o s s e s must be considered, from the rectenna element level t o the power g r i d interface. It i s c l e a r that a major e f f o r t is needed to properly consider t h e multitude of f a c t o r s involved i n this area of rectenna design.
4.0 SUMMARY AND F U m DIREETIONS The l i s t of key accomplishments of OUT program have been summarized i n Section 1.3 cf our report and described i n d e t a i l i n Sections 2.0 and 3.0. lhese accomplishments can be summarized concisely as follows: Delineation of Desirable Characteristics f o r Rectenna Receiving Element w i t h Comparison of Viable Alternatives (Sec. 2.1).
-
Design and Analytical Evaluation of R i n t e d Circuit Implementation (Sec. 2.2.1). F i r s t Order E l e c t r i c a l Design of Yagi-Uda Elements with Delineation of "!fadeoff of Gain, F/B Ratio and Size (Sec. 2.2.2). F i r s t Order Design of Eight Rectenna Elements with Comparison of Advantages and Disadvantages (Sec. 2.3.1
and 2.3.2). Preliminary Cost Analysis Indicating That Directional Receiving Elements Inwer O v e r a l l . Rectenna Costs with Either Baseline me o r Printed Circuit kplementation (Sec. 2.3.3). POWER COMBINING EVALUATPION Implementation of Computer Model of Baseline m e Conversion Circuitry Using General Furpose Non-Linear Program (Sec. 3.1.1).
.
Closed Form Models of Conversion Circuitry Developed (Sec. 3.i.2). Development of Methodology of Power Combining Inefficiency Evaluation (Sec. 3.2 and 3.3.1). Power Combining Inefficiencies f o r Series and Ferallel Combining Evaluated with Comparison of Results with Computer Simulation and Closed Form Models (Sec. 3.3).
Impact of Rwer Combining Inefficiencies on m e r Module Sizing and DC
Buss Network Evaluated (Sec. 3.4). Although the program has been geared toward obtaining r e s u l t s useful f o r current SPS system definition, there are a number of extensions of this program which s b u l d be considered a t t h i s time. These a r e outlined as follows :
114
DmCTlOmAL RECEIVING ELEMENT EXIZNSIOIQS B e Yaef-Uda receiolng element analysis indicates that appreciable r e c t e m c o s t sawings caa be expected. More design i s needed t o a r r i v e
at more rigorously obtslined e l e c t r i c a l parameters, thus replacing the first order paraneters @veri i n Wble 2-8 and discussed i n Sec. 2.2.2. Consideration of a l t e r n a t i v e Yagi-Uda designs besides the three and six element designs evaluated t o date. Structural design is needed t o evaluate the mechantcal advantages and disadvsntages of printed c i r c u i t implementation, ae w e l l as delineate t h e sdvantages of eliminating the ground plane mesh as considered feasible w i t h Yagt-uda dements with high F/B r a t i o . Development of a nma'l.7 array t e s t structure f o r experimental evaluation of t h e Yagi-Uda rectenna element. Further consideration of a l t e r n a t i v e antenna elements.
Eackf'ire arrays should be considered as an a l t e r n a t i v e t o the higher gain Y a g i - U d a arrays.
POWER COMBINING EVALUATION EXT!3I?SIONS Use of closed form model results i n evaluation of a l t e r n a t i v e X buss networking a l t e r n a t i v e s (instead of row based or complete r i n g based dehigns ) Use of computer simulation model t o optimize devlce c h a r a c t e r i s t i c s f o r
low, medium and high power levels, including performance and r e l i a b i l i t y factors
.
Extension of computer simulation model and modification of closed form analytical models t o evaluate metbods of reducing harmonic reradiation. Use of closed farm and computer simulation models in evaluating e f f e c t s
of t r a n s i e n t loads Use of closed form device and c i r c u i t These extensions innovative e l e c t r i c a l
on the spectrum of rectenna element reradiation. and computer simulation models t o evaluate e?fects of parameter toierances. a r e needed t o reduce projected rectenna costs with and structural. design, t o properly i n t e r f a c e the rec-
tenna with the power grid and t o evaluate an important environmental impact
of the SPS (reradiation from rectenna).
1. W. C. Brown, "Electronic and Mechanical Improvement of t h e Receiving Terminal of a Free-Space Microwave hwer Transmission System", MA%-CR-135194, August 1977
.
2.
R. D. Dickinson, '%valuation of' a Microwave High-Power ReceptionConversion Arrey f o r Wireless Pater Transmission", JPL Tech. Memo 33-741, September 1975.
3. W. C. Brown, "Free-Space Microwave Power Transmission Study Ilhase If1 and F i n a l Report", l?ASA-CR-14&51,
September 1975.
-
4.
Lyndon B. Johnson Space Center NASA, "Initial Technical, Environmental and Economic Evaluation of Space Solar Paser Concepts", JSC 11568, August 1976.
5.
R. J. Gu-M, "Microwave System Studies Affecting SPS Rectenna Performance", JSC Internal Report, August 1977.
6. Boeing Aerospace Cornpaw, "Solas Power S a t e U t e , System Definition
-
Study Part I1 VOX. IV rncrowave hwer Transmission Systems", mASA Contract NASg-15196, December 1977.
7. B e i n g , General Z l e c t r i c and Raytheon, "SPS System Evaluation -?base 111 Study acument", :me 1978, preliminary corn. 8. J. 3. Nahas, "Simulation and Experimental Studies of YAcrowave-t0-E Energy Conversion Systems", NASA Grant No. NSG-3070, (undated but approximately January 1976).
9.
J. Appelbaum, J. Bany and A. Br%unstein, "Array hwer Output of
Non-Identical E l e c t r i c a l Cells", 12th An??ual Intersociety Energy Conversion Engineering Conference (ECEIC ) , 1977, pp. 16% 1692. 10
.
-
S. Uda and Y. Mushiake, Yagi-Uda Antenna, Sasak.2 lUblisU& Co.
, 1954.
u.
2 . W. P. King, R.
12.
I. L. Morris, "Optimization of the Yagi Array", Doctoral Thesis, Harvard University, April 1965.
13
G. L. Matthaei, L. Young and E. M. T. Jones, !ttcrowave Filters,
B. Mack and S. S. Sandler, Arrays of Cylindrical Dipoles, Cambridge U n i v e r s i t y Press, 1968.
Impedance Matching Networks and Coupling Structures, McGrawH i l l , 1964.
14. D. A. Daly, S. P. Knight, M. Caulton and E. Ekholdt, "Lumped Elements i n Microwave Integrated Circuits", I E E Tram. Xcrowave Theory and Techniques, M!lT-15, December 1967, pp. 7l.3 721.
-
15. . C. A. Chen and D. K. Cheng, "Optbum Elemnt Lengths f o r Yagi-Uda A r r a y s " , fEEE Trans. Antennas and Propagation, AP-23, January 1975, pp. 8-15.
16. R. W. Klopfenstein, "Comer Reflector Antennas with Arbitrary Mpole Orientation and Apex Angle", IRE Trans. on Antennas and Propagation, AP-5, July, 1957, pp. 295 306.
-
17. J. D. Kraus, Antennas, McGraw-Hill, 1950. 18. E. A. WoMf, Antenna Analysis, John Wiley, 1966. 19. R. J. Gutmann, "Outaut Power Variations with Solar Power Satellites", Journal of Solar Energy, 21, November 1978, pp. 323-330.
-
20.
J. Appelbaun, J. Sany and A. Braunstein, "Array -%er Output of NonIdentical E l e c t r i c a l Cells", 12th Annual IECEC Conference Digest, 1967, pp. 1686 1692.
-
A S , JSC. 21. Conversation with R. H. Dietz, M
