A compact dual polarized GHz rectenna for high voltage by benbenzhou

VIEWS: 255 PAGES: 25

More Info
									I   . .

             A compact, dual-polarized 8.5 1 GHz rectenna
             for high voltage (50 V) actuator applications
                                Larry W. Epp, Abdur R. Khan, Hugh K. Smith, and R. Peter Smith

                                                                       paneldetermines the         total power an individual     panel
    Abstract-- This paper describes a dual-polarized rectenna          receives.
    capable producing
           of        a         50 V output voltage for driving                   additionally
                                                                                  By       separating individual
    mechanicalactuators.      This work demonstrates a circuit                 panels segmented,
                                                                       rectenna across   a         actuatordriven
    topology that allows the output of multiple rectenna elements      reflector, multiple rectenna panels can provide for power
    to be combined in series to step-up the output voltage from 18     distributionwithout  physical        or
                                                                                                     wiring interconnects.             If
    diodes. In this work, an independent rectifying circuit is used
                                                                       each rectenna panel is sized to provide suitable voltage to
    for each of two orthogonal polarizations to    minimize the
    rectenna size. This helps maximize the output voltage so that a    driveanactuator,       this providesa spatial distribution of
    9 element array can contain 18 diodes. By proper independent       power toeachactuator.           Ideally, theindividualrectenna
    combination,the output voltage is doubled over thesingle           panels on each actuator provide a source of control signals
              case,       producing an output combining the            for each actuator. For example, in addition        to DC power,
    maximum voltage output capable from 18 individual diodes.          each        could remotely
                                                                           actuator be                         by
                                                                                                   controlled proper
    Such panels are being explored for use on the Next Generation      modulation of the incident microwave beam.
    Space       (NGST)                       wiring
                                   to eliminate between                           The development of rectenna     a involves
    actuators and providefor true mechanicalisolation.                              the
                                                                       maximizing efficiency                          the
                                                                                                          of capturing microwave
                                                                       energy by the antenna, and maximizing the efficiency the    of
    Index                      actuators, free      space   power      rectification process.In this application,therectification
    combining andrectification
                                                                       efficiency involves a trade-off between developing a DC
                                                                       power conditioning system that provides proper terminating
                                                                       impedance to the     rectifying         and
                                                                                                       circuits, generatingthe
              A rectenna is an antenna that captures and converts      propervoltageneeded          in theparticularapplication.     The
    RF or microwave power to DC power. It is useful as the             rectenna development work here wasinitiated as part of the
    receiving terminal of a power transmission system where            development of microwave driven smart material actuators
    DCpowerneeds         tobedelivered     to aload,throughfree        for control of Next Generation Space Telescope (NGST) by
    space, where physical transmission lines are not feasible. It      Prof. Sang Choi of NASA Langley [l]. It was desired that
    is also useful in applications where DC power needs to be          the rectenna panels produce an output voltage of 50 V, for
    distributed to a large number of load elements in an array.        minimal panel size, while operating at a frequency of 8.51
    The power distribution is achieved by the distributed nature       GHz.
    of microwave energy in space, eliminating the need         for a
    large number of physical interconnects to individual load                         11. ANTENNA
                                                                                                ELEMENT DESIGN
              The goal of a           rectenna
                               suitable       is         to convert
                                                                       dipoleantennas      [2] - [6] and received  single
                                                                                                                   a     linear
    microwave energy into DC, essentially the opposite process
                                                                       polarization.A thin-film, printed-circuit  dipolerectenna
    of moderngridamplifierdesigns,whichconvertDC                  to
                                                                       design was proposed initially by Brown [3] and had simple
    microwave energy. Analogous to grid amplifiers, a rectenna
                                                                       DC removal. This single polarization design minimized the
    can use the distributed nature     of the microwave power to
                                                                       thermal path between the diodes and the outer surface. See
            the from elements,
    combinepower many               which                        are
                                                                       Fig. 1. But the printed capacitors of these thin-film designs
    spatially separated by the element spacing      of the array or
                                                                       meantheyarenot        as readilytunable as Brown’s original
    panel.Thereforetheeffectivearea          of theentirerectenna
                                                                       ‘bar-type’ rectenna element [3]. In practice, the ‘bar-type’
                                                                       rectennas have been demonstratedby Dickinson in the 1975
    This work was supported by the Jet Propulsion Laboratory,          JPLGoldstonemicrowavetransmissiondemonstrationto
    California Institute of Technology,underacontractwith              have efficiencies as highas 82.5% [7] - [8].
    the National Aeronautics and Space Administration.                           For dual polarization the extension of the thin film
    Larry W. Epp, Abdur R. Khan, and R. P. Smith are with the          dipoledesignquicklyencounterssignificantobstacles            to
                 Laboratory, Institute
    Jet Propulsion        California                          of       successful implementation. A separate layer, one layer for
    Technology, Pasadena, CA91 109 USA.                                   polarization,
                                                                       each          is
                                                                                     required                for eachpolarization.
    Hugh K. Smith was with the Jet Propulsion Laboratory. He           Electromagnetic            between DC
                                                                                        interaction      the collection
    is now with Motorola, Cellular Subscriber Sector Research          lines,parallel to thedipole on theorthogonallayer,can
    Laboratory, Harvard, IL 60033.
    compromise the rectenna performance. Having one dipole                increasinglypopular in rectennadevelopment [IO] -[12],
    layer“buried”beneaththeotheralsohasseriousthermal                     [ 141 - [ 161 at 2.45 GHz and 35 GHz.
    problems.                                                                        Aperture coupled microstrip patch antennas were
              Inaddition,thefabricationprocess            of thedual      first introduced by Pozar in 1985 [17]. In December 1986
    linear dipole design can be quite challenging since diodes            Pozar presented a reciprocity-based method for the analysis
    and capacitors            may be  potentiallyburiedbetween            of aperture coupled microstrip patch antennas [ 181. Other
    layers of foamand/orpolyimidefilm[6].              To avoid this,     researchershaveofferedsolutions        to thesingleaperture
    Alden and Ohno [9] developed a single foreplane design.               coupledmicrostrippatchantennausingmodalorcavity
    But the dipole designs, in general, suffer from the situation         methods [19] - [20]. More, recently, Rostan, Wiesbeck, et.
    that the feed circuitry and the antenna (dipole) are tightly          al. have  employed        methods
                                                                                              Pozar’s                in designingdual
    coupled, in the sense that thetwinleadtransmission             line   polarized aperture coupled microstrip patch antennas. These
    impedance relates to the dipole impedance; therefore posing           patch antennas, used for synthetic aperture radar (SAR) and
    aconstraint on thedesign.Alsoinherentwithadipole                      satellite reception antennas [21] - [24] were the basis for
    designisthefactthefeedcircuitrywasexposedtothe                        the design used here.
    outside world and therefore parasitic radiation from these
    feed lines including harmonic radiation (generated by the               B. Construction Details of the Aperture Coupled
    diodes) could bean issue.                                               Microstrip Patch Antenna
              For dual-polarizationneeds,microstrip
                                         a          patch                           The microstrip  patchantennas constructed used
    designcouldpotentiallyalleviatemany              of theproblems
    mentioned   above. advantage
                      The                    of dual-polarizationis       Rohacell@ 5 1 (e= 1.07), as the foam spacer, RT/duroid@
             it           doubles receive
                                 the       powerper                       5880 (0.020inchesthick,         = 2.2, Yi ouncecopper) as the
    element area, and makes the rectenna capable of receiving             microwave substrate, and Sheldahl’s Novaclad@ G2200 (50
    either dual linear or a single circular polarization. The latter      pm thick, = 3.3, 1 ounce copper), which is a copper clad
    ability to receive circular polarization makes the rectenna           polyimide film, on which the microstrip patch antenna was
         more         for
    panel suitable applications      such               ascircling
                                                          a               etched.Method of Moments         solutions
                                                                                                                   indicate        that for
    airplane platform [21, [ lo].                                         every 0.001 inch difference in patch thickness a 10 MHz
              A microstrip patch which is aperture coupledrelies          change in resonant frequency will occur.
    on an aperture, or coupling slot, to couple electromagnetic                     To insure accurate thickness of the foam used, a
    energy from a feedcircuit to the microstrip patch antenna        as   method was developed to pre-compress the thickness of the
    is shown in Fig. 2. In this configuration, the antenna and the        Rohacell@ to 35 mils by using a compression fixture and         an
    microstrip feed circuit are divorced from one another. This
                                                                          oven. A pieceof 40 mil thick Rohacell@ foam is placed a       in
    alsoallowsthediode         circuitry to beisolatedbehind        the
                                                                          compression        which foamthen
                                                                                       fixture    the is preloaded
          plane         the
    ground containingcoupling     apertures.                      This
                                                                          (placed in compression). The entire assembly is then placed
    groundplane,whichseparatesthepatchantennaand                    the
                                                                          inside an oven that is heated above 375” F, the temperature
    feed circuit, protects the feed circuitry from the incident     RF
    energy.Asaresult,theincident            RF energy will notbe                   the
                                                                          at which Rohacell@ its loses compressive  strength.
    coupled to the DC lines that collect the output power.                Using 35 mil shims sets thefinal thickness.
              The ground    plane prevents harmonics,
                                 also        the                                    Typical solder reflow temperatures are above 400”
    whicharegeneratedbythediodes,               fromradiatingback         F and the Rohacell@ begins to expand at 375” F. Therefore,
    towards the incident wave. A microstrip line filter is used to        if the Rohacell@ is bonded to the RT/duroid@prior to the
    prevent radiation back through the aperture feed. Due to the                                                      to
                                                                          solder reflow process and then exposed the solder reflow,
    potential interference of generated harmonics, this concept           the foam severely distorts, destroying the rectenna panel.
    of separating the receiving element and rectifying circuitry          An alternativeis to usearoomtemperaturevulcanizing
    to preventharmonicradiationhasalsobeenexplored                   by
                                                                          silicone rubber           such Dow
                                                                                           adhesive, as Corning@          3140
    otherresearchers[11 J usinga circular patchandsingle
                                                                          RTV. Using the 3140 RTV Coating allows for components
          Other            have
    diode. researchers explored use    the                       of a
    Frequency Selective Surface [ 121.                                         wave
                                                                          to be soldered               to the            then
                                                                                                             RT/duroid@, the
                                                                          RT/duroid@ becan bonded           to theRohacell foam.       The
      A. Antenna                                                          actual fabrication process uses a Maher bar (grooved bar)       to
               As shown in Fig. 2, the patch is supported above               uniform
                                                                          applya                coat of adhesive;          insuring
    the ground plane by a lightweight foam support. This foam             repeatability in the fabrication process.
    has a relative dielectric constant of 1.07, thereby helping to
    avoid any surface wave modes that may limit the microstrip              C. TRL Standards for Patch Measurement
    patcharrayperformance.         In particular,thisconfiguration                    input
                                                                                   The impedance            of the         patch
    wasfirstdiscussed        by Ziircher in his SSFIP(StripSlot           antennaswasmeasuredonaHP8510Cnetworkanalyzer
    Foam        Patch)
         Inverted concept                  in 1988 Just
                                                    [13]. as              usinga TRL (through-reflect-line)   calibrationprocedure
    important, the ground plane provides a good thermal sink                  The     10C
                                                                          [25]. HP85 is                             with
                                                                                                     first calibrated the         TRL
    for the diodes with a via connection       to the ground plane.       standard, after which, both ports of the antenna test fixture
    Recently, patch have
                    antennasbecome                                        are connected to the network analyzer. The TRL calibration
accounts for the loss of the 3.81 cm (1.5 in.) of microstrip        resultsactuallyrepresent the bestthat could be achieved
line and SMA    connector    feeding the aperture. Similar          with a given
                                                                               device.However,                       is
                                                                                                          the inverse true     - a
calibration standards will be used to determine the line and        properly designed microstrip circuit, especially     with extra
connector loss        effective
                 in the            area measurements   that         components (e.g., chip caps,etc.) should certainly do as
follow.                                                             well as the empirically tuned results presented here. It was
                                                                    also found that optimization of efficiency using non-linear
  D.PatchMeasurement                                                simulations was time consuming, and of course inaccurate
           The goal of the   microstrip      patch design is to     until diode models were extracted.
present a low return loss (S11, S22) at 8.51 GHz. It is also
desired to have isolation between the two ports (a low S21,            B. DiodeMeasurement
S12) of Fig. 2. The scattering parameter data and finalpatch                  For the purposes of this effort         readily
dimensions are shown in Fig. 3 . The measured return loss in        available in a packaged format with large breakdown were
Fig. 3 is shown to exceed       18 dB (only port one is shown       chosen. The diode used is the commercially available MIA-
due to symmetry) at8.5 1 GHz.                                       Com 40401 Schottky diode in package model 213. For the
           The measured resonance frequency, for optimum            purposes here the rectification efficiency is defined to be
return loss, was 8.565 GHz.The simulations          indicatea       (PDC/(Pinc-Pref), where PDC is the DC output power, Pinc
worst case coupling between the two ports greater than -20          istheincident     RF power, and Pref is the reflected RF
dB,whileactualmeasurements           were slightlygreaterthan       power). Overall efficiency is definedby ( P D c / P ~ ~ ~ ) .
this value. Measurements         gave a worst case coupling                   The DUT (device under test) in Fig. 4 consists of a
between ports of -18.5 dB (at 8.07 GHz) for all frequencies         microstrip test fixture with the diode connected in a shunt-
below 10 GHz. To correct for the foam compression, the              to-ground configuration. The DUT could more specifically
relative           increased
                 was                               ratio
                                              by the           of   be considered to be the diode on the fixture. The test fixture
compressed to uncompressed height to & = 1.2.                       included a ground via next to the diode location. All power
           Fig. 3 also shows the predicted gain and scattering      measurements were made by manually tuning the input and
parameters up to 18 GHz. allows simulated
                                This     the                        output microstrip lines with small gold-coated silicon chips
parameters to      be shown  at    the harmonic
                                     first                of the    in an efforttoprovidesuitableharmonicterminations.            In
operatingfrequencyat17.02          GHz.The measuredresults          addition a Maury microwave tuner at the output contributed
stop at 16 GHz       wherethemicrostriplinebecomesover-             to tuning as well as blocking the RF signal.
moded. (Note the increase in measurement error for S1 1, as                   Measurementresults at thedesignfrequency            of
compared to the predicted value, at 16 GHz.) The predicted          8.51 GHZ indicated a maximum overall efficiency of 66%
patch gainat 8.51 GHz is 8.4 dB. At thefirstharmonic                with 65 mW of DC output power, and over 100 mW DC
frequency the gain is less than -17 dB, indicating that the         output power for a lower efficiency of 62%. This diode was
harmonic         is
         radiation                        back
                             not radiated through the               found to exhibit  a     higher output voltage     and higher
coupling        and       the source.
         aperture towards signal                              The   efficiency than asimilar diode by anothermanufacturer.
additional low pass filter, discussed below, provides further       Note that by choosing commercially available diodes, these
suppression of unwanted harmonic radiation.                         diodes were not optimized for maximum efficiency in this
                                                                    application, and therefore higher efficiencies are certainly
   111. DIODE MEASUREMENT AND MODELLING                             possible. The output voltage increased from 3.2 V to 4.1 V
                                                                    at the lower efficiency of 62%, indicative of the trade-off
  A. Diode Background                                               between maximum output voltage and efficiency expected.
                                                                    The large output voltage of 4.1V would allow for maximum
         Most rectennadesigns have followedBrown in
                                                                    output voltage if a suitable combination method could be
usingasingle     diode in aclampingcircuitconfiguration
rather than a traditional multiple-diode rectifying circuit. At
microwavefrequencies,theserectennacircuits          are highly
                                                                      C. DiodeModel
nonlinear and difficult to design based upon purely analytic
        Commercially harmonic
equations.          available    balance                                      The diodeis the mostcriticalcomponent         in the
         are               but necessarily limited by the                  element.
                                                                    rectenna               All aspects of performance of the
accuracy, of the diode model at     large      Therefore
                                         signal.                    rectennadependprimarilyon          the diode parameters. The
previous works that have developed equivalent circuits of                resistance, example,
                                                                    series         for                limits
                                                                                               directly    efficiency
the diode have also validated with experimental results [4,         through 12R loss. The junctioncapacitance,together with
12, 261.                                                            packagecapacitance       and lead inductance,   affects   how
          Since optimallytuning a rectifierrequiresideal                   currents through
                                                                    harmonic oscillate                            the diode. The
tuning atharmonics       as wellas     the fundamental tone,        breakdown voltage limits the power handling capability of
parameters of any diode model must be known the       at            eachrectifyingcircuit.     These parametersalsoaffect      the
harmonics as well. Experimental large signal measurements           match of the circuit. Since the diode, as a power-rectifying
provide a method of extracting and verifying diode models           element,mustoperate       in alargesignalenvironment,      the
and of searching for maximum efficiency. Since the tuning           diode model needs to be valid for a wide range of biasing.
done here is purely empirical, there is no guarantee that the       Finally,since high efficiencyrequiresa         large amount of
harmonics to be produced, the model also needs to be valid        higher than expected, approximately 50%. This difference
over a wide frequency range.                                      was due to a discrepancy between the placement of the via
          The diode model used for patch rectenna element         holes in the circuit schematic and the physical layout. The
simulationsis    shown in Fig. 5. The capacitor and the           movement of the via holes turned out to be significant.
inductor were added as the packagecapacitance and lead                     With minimal movement of existing components,
inductance.Valuesfordiodeparameters          such asgrading       and therefore minimal      board re-design,simulation    and
coefficient (M), transittime (m),energy gap (EG), and             measurement  showed      that additional tuning stubs could
saturation-current temperature exponent (XTI) were chosen         reduce the unwanted reflections. The trade-off was that the
tosimulateaSchottkydiode.Otherparameters,such                as   expectedoverallefficiency     was expected to fall to 60%.
         current     (IS), series        (RS),
                                resistance junction               Fig. 8 shows the unit cell with the additional tuning stubs,
capacitance (CJO), and current at breakdown voltage (IBV)         show shaded,referred to here as Prototype 2.
werechosen so thatthe DC performance of the model
agreed with measured data.                                          C: Analysis and Measurement Comparisons
          The simulationdatafor      the DCperformanceis                    Several cells
                                                                                    unit         of this version (Prototype 2)
shown in Fig 6. The voltage, current, and resistancedata          were fabricated and measured. Figs.9 and 10show the
were taken from the IV curve, and the total capacitance was       measured       for
                                                                           results 3separate             boards. Notethat    the
measured by resonating the diode with a series inductor. As       rectification efficiency of all circuits remained close to the
can be seen, the simulation data agrees      well    with the           overall
                                                                  design efficiency              of 60%, with high overall
measured data shown in Fig. 7. The breakdown voltage was          efficiency when reflected  power         was minimized. The
selected to be slightly higher since it was possible to select    average output voltage of the H circuits was 4.14 V at an
diodes, with breakdown around 9SV, for use.                       averageoverallefficiency      of 57.7% when usinga load
                                                                  resistance of 325 Q . This             result the
                                                                                                desirable meets
              I v . RECTENNA
                           CIRCUIT DESIGN                         maximum output voltage fromthe diode measurements.
                                                                            The average output voltage of the Vpolcircuits was
  A. Circuit Design                                               lower, 3.84 V, at an average efficiency of 49.8% with a
          The patchrectennaismosteasilyimplemented                load resistance of 325 Q. The trade-off of the higher output
usingstandard microstrip
                            Fig.          5 shows the basic       voltageforefficiencyis      indicated by the loweraverage
rectenna circuit topology. In order to economize the use of       output voltage of 3.56 V for the higher average efficiency,
real estate, an attempt was made to use a minimum number          53.7%, when a load of 250 Q is used. Higher sensitivity to
of stubs the filterlmatching
         for input               section.          The line       variations in component assembly may have contributed to
lengthswerealso      made as shortas possible in order to         the lower averageoverallefficiencyfor        the VP1 circuits.
minimize the RF losses. The corresponding      microstrip                         ,
                                                                  The fact that V circuit of board 2 performed as well as the
implementation of Vpol andH,I circuits, for each of the two       HPI circuit of board 2 indicates this to be due to variations
patch apertures, is shown in Fig. 8. Note the stepped line        in component assembly.
impedance and chip capacitorperformsthefinalsteps           of
low filteringmatching.
           and                            The circuits were                        v.   PANEL MEASUREMENT
optimized for an incident power of lOOmW, the expected
reflection is 3 to 4 mW, and the expected overall efficiency        A. Rectenna panel: 3 by 3 elements
was 65%.
                                                                            Rectenna consisting
                                                                                   panels                   of a 3 by 3
                                                                  arrangement of unit cells were fabricated     using      the
  B. Measurement of Unit Cells                                    Prototype 2 circuitry. The averageoutputvoltage       of the
         Measurements of the unit cells show in Fig. 8 were       circuits was 4 V, requiringtheseriesconnection        of 13
first completed without the microstrip patch.     Fig. 4 shows    circuits to reach the design goal of 50 V. Since each patch
the test setup used to test the unitcells. The HP8671B            providestwocircuits,      one for VP1 and one for HP1, the
frequency generator and the Hughes TWTA provided the              minimumnumber of cells required is 7patchelements.
incident power. Both the incident and reflected power were        Choosing theminimalsquarearraycontaining          at least7
simultaneouslymonitoredusingtheNarda             lOdB couplers    patches leads to an array of 3 elements by 3 elements, for 9
and the HP438 power meter. A short section of waveguide           total patch elements.
served as a DC block in 'the coaxial lines to prevent anyDC                 The effectivearea of asinglepatch      elementis
loading on the input port of the DUT. The DUT consisted           given by,
of a single unit cell with a circuit for H and a circuit for
VP1 in the same layout, as they would appear in an array.                               A,=G -                              (1)
The input microstrip lines, however, were extended to       the                              4n
edge of the substrate to coaxial connectors. DC output was        Where Gp is the gain of the patch, or 8.4 dB at 8.51 GHz.
picked off the circuit by simplysolderingasingle-strand           The effective area for a single' patch, before placement    in
wire directly to the DC bus line. A decade box was used as        the array, is therefore 6.8 cm2.Inorderforthearray          to
the load.                                                         absorb all the incident power, it is necessary that the unit
          Initial measurements showed that, in particular, the    cell area be less than 6.8 cm2. For rectangularspacing,a
reflected power fromthe rectifying      circuitry   was much
minimalcell-to-cellspacing        of 2.62 cm is thenrequired.                 To properly compute the maximum effective area
Accordingly, a more dense cell-to-cell spacing of 1.97 cm         of each patch in the array configuration, a rectenna panel
was chosen to further shrink the overall panel size while            built
                                                                  was where      microstrip leadinga
                                                                                 a        line      to                         SMA
still leaving sufficient room for,the panel circuitry.            connector replaced each rectenna circuitry as shown in Fig.
                                                                   13. Each port was individually tuned until the return loss
 B. Calibration of Measurement Chamber                            for all patches exceeded 21 dB with all other ports matched.
          In order to ensure that the rectenna panel could be     By symmetry, only one polarization was measured and the
         measuredfar custom-built
efficiently      in field,
                 the a                                            panel         for orthogonal
                                                                        rotated the            polarization.                The line
measurement chamber was designed to allow quick access            losses and connector losses where removed           by calibration
to the panel. Fig. 11 shows the measurement chamber with          standards,asdiscussedpreviously,foreach                of thetwo
therectennapanelholderabovethetransmithorn.The                    microstrip feed line configurations. To ensure accuracy of
         gain used illuminate rectenna
standard horn to                the                        is a   the           area
                                                                     effective measurements the   for tightly packed
Narda 640 Standard Gain Horn. The gain of the Narda 640           array, the coupling between ports was measured and found
@ 8.51 GHz (frequency of the incident microwave energy)           to be  less than 17.5dBforallports.Thislowmutual
is 15.1dB.Fig. 12 showsthefrontside            of therectenna     couplingforthisdenselypackedarraycanbeprimarily
panel in the chamber.                                             contributed to the useof the low dielectric foam superstrate.
          The Friistransmissionequation       isemployedto                    The results of the effective area measurements are
verify the powerdensity the containing
                       on plane          the                      shown in Fig. 13. The transmitted polarization was a single
rectenna. An identical Narda 640 standard gain horn is used       linearpolarizationatanangle           of 45 degrees, in order to
as the receive horn at the locationof the rectenna panel. To              both
                                                                  excite patch          equally.
                                                                                    ports                  The effective area was
calculate the power received by this receive horn the Friis       averaged between the two orthogonal ports          of each patch.
transmission equation is used and the system calibration is       Notethattheeffectivearea            of thecenterpatchclosely
checked and measurement chamber accuracy verified.         The    matches the physical unit cell area, as expected. Note also
measured gain of the standard gain horn is 15.03 dB, for an       that the effective areaof the corner patches is slightly larger
error of 0.07 dB.                                                 than the unit cell area, since these patches are on the outside
                                                                  of the array.
  C. Effective Area Measurement of the 9 Array Elements                       In
                                                                           order                        the
                                                                  discrepancies in thepanelmeasurementsthatfollow,the
         During the diode and unit cell measurements, the         cornerandcenter-edgeeffectiveareameasurementswere
overallefficiencywasdefined       to by thetotal DC output        averagedusingasymmetryargument.               The totaleffective
power at the load in proportion to'Gincident power. Thus          area of the 3 by 3 panel was then found         to be 1 cm more
any power that is reflected by the circuitry is not converted     than the physical area of the panel, or 3% greater than the
into DC power decreases overall
                and        the     efficiency.                    physical area. Data from Port 8 was thrown out, since due
        in panel
Likewisethe measurement,           efficiency
                             overall                              to symmetry conditions it was evident that the proximity of
should be defined such that any reflected power from the          the SMA     connector Port
                                                                                       from                  5 was        the
panel lowers the overall efficiency. An over-simplified and       measurement.
erroneousmethod of computingoverallefficiencyfor            a
rectennapanelis     touse thepowerdensitystriking         the
                                                                    D. Panel Results
rectenna          from Friis
                      the                        equation
multiplied by thephysical panel surface area to compute the                 To maximizetheoutputvoltage           of the panel,a
"received" power.                                                 seriescombination of allindividualrectennacircuitswas
         But more correctly the received power is given by,       desired. To makethispossible,thegroundplane(which
                                                                  contains coupling    apertures)       each
                                                                                                 around individual
                                                                  patchwas DC isolatedbelow each patch. To ensure RF
                                                                  connection,a thin layer of copper-coatedpolyimidewas
where A represents the maximum effective
        ''                                         area of the    used to ensure capacitive coupling. Opposing installation      of
panel as the sum of the maximum effective area         of each      diodes each
                                                                  the     for                of thepatchesorthogonalports
patch in the array configuration. Thus it can be seen that for    correctly,seeFig.14,allowed         for aseriesoutputvoltage
                                                                  for eachpatch. It was found thatin order to properly reverse
rectenna panel applications it is desirable, and possible, for
the maximumeffectivearea         of thepaneltoexceed       the       all
                                                                  bias diodes                                   that
                                                                                      when power is applied additional
physical area of the panel. And, that the maximum effective               betweenisolated planes
                                                                  resistance the ground was
                                                                           And additional
                                                                  desirable. the                  allowed
                                                                                         resistance      each
                patch be
area of each must      measured                  in the array
                                                                         ground to
                                                                  isolated     plane        when
                                                                                    discharge power                             was
configuration matched            i.e.
                        conditions,with                     no
reflection, when mutual coupling is present.                      removed, protecting the diodes.
                                                                            Fig. 15 shows the measured results of the rectenna
           If the overall panel efficiency is then defined as,
                                                                  panelwhenloadedforoptimumoverallefficiency                 with a
                                                                     resistance       of 5400 R. The overall              is
                                                                  calculated via (3). The load resistance of the entire panel is,
                                                                                         the cell
                                                                  in general, 18 times unit resistance.                The series
any power reflected from the rectenna panelwill be provide
                                                                  connection of the circuits on the panel involved connecting
a decrease in the overall efficiency.
slightly different, not identical, circuits. Therefore a starting                  Raytheon Co., Wallops Flight.Center, NASA Report No. CR 156866,
load resistance of approximately 18 times 325 Q or 5850 Q                          Contract No. NAS6-3006, July 1980.
                                                                                      W. C.Brown, “Rectenna technology program: ultra light 2.45
was indeed close to giving optimal overall panel efficiency.                       GHz rectenna and 20 GHzrectenna,” Raytheon Co., NASA Lewis
          The overall panel efficiency exceeds       52% over a                    Research Center, NASA Report No. CR 179558, Contract No.
largeregion of inputpowers,withapeak               of 53% at a                     NAS3-22764, March 1987.
receivepower of 38.8      mW/cm2. It alsoshows that the                               Tae-Whan Yo0 and Kai Chang, “Theoretical and experimental
                                                                                   development of 10 and 35 GHz rectennas,” IEEE Transactions on
desired output voltage of 50 V can be achieved for an input                        Microwave Theory and Techniques, vol. 40, no. 6, pp. 1259 - 1266,
power density of 25.2 mW/cm2. For the Narda 640 standard                           June 1992.
gainhorn,thisrequiresatransmitpower             of 13.6 W at a                        Michael Tran and Cam Nguyen, “A new rectenna circuit using a
distance of 37.1 cm to provide 50V of output power. Note                           bow-tie antenna for theconversion of microwave power to DC
                                                                                   power,” Microwave and Optical Technology Letters, vol. 6, No. 11,
that to provide maximum efficiency, the panel requires not                         pp. 655 - 656, September 5, 1993.
only sufficient loading, but sufficient input power to place                          J. 0. McSpadden, “Theoretical and experimental study of 2.45
thediodes in anefficientregion         of operation.Thepeak                        GHz rectifying antennas,” Master of Science Thesis Submitted to
overall efficiency
      panel                   is 4% than average
                                    less the                                       Texas A&M University, December 1993.
                                                                                      R. M. Dickinson, “Microwave Transmission System for Space
efficiency of the H, unit cell measurements and 1% less                            Power,” Raumfahrtjorschung, Heft 5, 1976.
than the average V unit cell efficiency (measured without                             R. M. Dickinson, “Performance of a high-power, 2.38 GHz
the patch via microstrip line). This indicates that additional                     receiving array in wireless power transmission over 1.54 k,” in 1976
gains in efficiency are likely from further optimization of                        IEEE M7T-S Int. Microwave Symposium Dig., 1976, pp. 139-141.
                                                                                      A. Alden and T. Ohno, “Single foreplane high power rectenna,”
the unit cell.                                                                     Electronics Letters, vol. 28, no. 11, pp. 1072-1073, May 21, 1992.
          Fig. 16shows  the                           for
                                  final configuration this                 [lo]         Peter Koert and James T. Cha, “Millimeter wave technology for
application. By increasing the load resistance, the required                       space power beaming,” IEEE Transactions on Microwave Theory
50 V output could be obtained for a low incident power                             and Techniques, vol. 40, no. 6, pp. 1251-1258, June 1992.
                                                                           [ 111        Takeo Ito, Yoshiyuki Fujino, and Masaharu Fujita,
density of 6.3 mW/cm2. This corresponds to only 3.4 W of                           “Fundamental experiment of a rectenna array for microwave power
transmit power at a distanceof 37.1 cm.                                            reception,” IEICE Trans. Commun, vol. E76-B, no 12, pp. 1508-
          Since the expected output voltage of 18 diodes in                        1513, December 1993.
                                                                           [ 121        James 0. McSpadden, Taewhan Yoo, and Kai Chang,
series could exceed the desired output voltage, a method of                        “Theoretical and experimental investigation of a rectenna element for
producingmorethanone           50 V outputwasdesired.By                            microwave power transmission,” IEEE Transactions on Microwave
attaching commercially available boost regulator circuits it                       Theory and Techniques, vol. 40, no. 12, pp. 2359-2366, December
was possible to obtain two 50 V outputs. The series output                         1992.
                                                                           [ 131        J.-F. Ziircher, “The SSFIP, a global concept for high
of 5 patches, or 10 circuits, was used to drive one regulator                      performance broadband planar antennas,” Elecrronics Letters, vol.
and the remainder to drive the additional regulator. Fig. 17                       24, no. 23, pp. 1433-1435, 1988.
shows typical output when the boost regulator circuits were                [ 141        Yoshiyuki Fujino, Takeo Ito, Masaharu Fujita, Nobuyuki Kaya,
driving large
       the impedance                of avoltmeter.Although                         Hiroshi Matsumoto, Kazuaki Kawabata, Hisashi Sawada and
                                                                                   Toshihiro Onodera, “A driving test of a small dc motor with a
                             has yet tested
interesting, this capability not been         for                                  rectenna array,” IEICE Trans. Commun, vol. E77-B, no 4,pp. 526-
suitability when driving actuators that will present a lower                       528, April 1994.
impedance.                                                                 [151         James 0.McSpadden and Kai Chang, “A dual polarized circular
                                                                                   patch rectifying antenna at2.45 GHz formicrowave power
                                                                                   conversion and detection,” in 1994 IEEEM U - S Int. Microwave
                        VI. CONCLUSIONS                                            Symposium Dig., 1994, pp. 1749 - 1752.
         A compact rectenna capable of producing a 50 V                    [ 161        T. Yoo, J. 0.McSpadden, and K. Chang, “35 GHZ rectenna
                                                                                   implemented with a patch and a microstrip dipole antenna,” in 1992
output suitable for driving mechanical activators has been                         IEEE M U - S Int. Microwave Symposium Dig., 1992, pp. 345-348.
demonstrated. advantage            of the       coupled
                                         aperture                          [ 171        D. M. Pozar, “Microstrip antenna aperture coupled to a
      configuration previous rectenna
rectenna            over     dipole                                                microstripline,” Electronic Letters, vol. 21, no. 2, pp. 49-50, January
designs: it is more amenable to dual polarization incidence                        17, 1985.
                                                                           [181         David M. Pozar, “A reciprocity method of analysis for printed
and more suitable to series combination outputs.                                   slot and slot-coupled microstrip antennas,” IEEE Transactions on
                                                                                   Antennas and Propagation, vol. AP-34, no. 12, pp. 1439-1446,
                                                                                   December 1986.
                  ACKNOWLEDGMENT                                           [19]         M. 1. Aksum, S. -L. Chuang, and Y. T. Lo, “On slot-coupled
The authors gratefully acknowledge     W. Wiesbeck and F.                          microstrip antennas and their application to cp operation - theory and
Rostan for providing the initial designof the dual polarized                       experiment,” IEEE Transactions on Antennas and Propagation, vol.
                                                                                   38, no. 8, pp. 1224-1230, August 1990.
aperture      microstrip      The      and
                        antenna. support                                   [20]         A. Ittipiboon, R. Oostlander, Y. M. M. Antar, and M. Cuhaci, “A
assistance of R. M. Dickinson and S. H. Zingales of JPL are                        modal expansion method of analysis and measurement on aperture-
similarly acknowledged, as is assistance from R. M. Perez                          coupled microstrip antenna,” IEEE Transactiqns on Antennas and
on the various measurements.                                                       Propagation, vol. 39, no. 11, pp. 1567-1573, November 1991.
                                                                           [21]         E. Heidrich, F. Rostan, and R. Zahn, “Dual polarised microstrip
                                                                                   array for spaceborne sar-application,” Proceedings of Joint 3rd
                        REFERENCES                                                 International Conference on Electromagnetics in Aerospace
                                                                                   Applications and 7th European Electromagnetic Structures
[ 11  Jet Propulsion Laboratory, California Institute of Technology, JPL           Conference, Politecnico di Torino, Italy, pp. 423-426, September 14-
    Task Plan No. 80-4734                                                           17.
[2]   W. C. Brown, “Design definition of a microwave power reception       [22]         F. Rostan, E. Heidrich, and W. Wiesbeck, “High-performance C-
    and conversion system for use on a high altitude powered platform,”            band microstrip patch subarray with dual polarization capabilities,”
       Progress in Electromagnetics Research Symposium, Noordwijk, The
       Netherlands, July 11-15 1994.
[23]       F. Rostan, E. Heidrich, W. Wiesbeck, “Design of aperture-
       coupled patch arrays with multiple dielectric layers,” 23rd European
       Microwave Conference, Madrid, Spain, September 6-9, 1993.
[24]       F. Rostan, G. Gottwald, and E. Heidrich, “Wideband aperture-
       coupled microstrip patch array for tv satellite reception,”Eighth
       International Conference on Antennas and Propagation, Heriot-Watt
       University, Edinburgh, UK, March 30 - April 2, 1993.
[25]       Hewlett Packard Product Note 8510-8A, “Network analysis
       applying the HP 8510 TRL    calibration for non-coaxial
[26]       J.J. Nahas, “Modeling and computer simulation of a microwave-
       to-DC energy conversion element,” IEEE Transacfions on
       Microwave Theory and Techniques, vol. MTT-23, no. 12, pp. 1030-
       1035, December 1975.
              L                 U
0    0   0   0    0    0   0
cv   7       7
    I   0
      """             - ""=-""""""                "'
                                                   a       0
                                                   A       L


I                                             W                                              II
I                                                                    c\!
I                                                                    0
I                                                                     II
I                                                               t:
I                                                               0
I                                             A
                                                            D "
;"""""""""_ I
                                                  F                        II
                                              0        I
; I
                           0"        "

                                                       I                            > o w
                                                       I                            mzu,
                                              U        I
                                              W        I
                                                       I                                h
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I
I                                                      I

    LL                                                     5
    'T"                                                    3
                     ....         P
LL   .............
i i
n    n
n    (d
a,   I!
a    0
a.   a,
n    L
a    w-
0    B
n    v)
         3                                                           r-
         cn                                                          c!
    b                                      a0                   al
    t:                                     t:                   t:
    0                                       0                   0
    a                                      a                    a
I   =jt,       .


                                            rr. I

    t:                                                          t:
    0                                                           0
    n                                                           a
k                                 E
1                       W
                    i   0
        I       i
        j       I
        I       1       0
            1   i
    "     .
         -"    .   ""   _.
                                     r 4   -
                                           0   0   t 5
                                                   i i
           I   -
               9   Y
""                         P
                           I n
0     0
v3    vl

To top