USF Libraries
USF Digital Collections

Microwave frequency doubler integrated with miniaturized planar antennas

MISSING IMAGE

Material Information

Title:
Microwave frequency doubler integrated with miniaturized planar antennas
Physical Description:
Book
Language:
English
Creator:
Presas, Suzette Marie
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Frequency doubling reflectenna
Diode doubler
Conversion efficiency
Compact microstrip antenna
Meandered antennas
Dissertations, Academic -- Electrical Engineering -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: In this thesis the development of a high efficiency harmonic re-radiator, consisting of a diode doubler and conjugate-matched receive and transmit antennas, is described. Diode-based frequency multipliers and rectifiers, coupled with antennas, are of interest for quasi-optical applications, for energy-scavenging and for sensing applications. The device studied operates by receiving an interrogating signal at a frequency of 1.3 GHz and re-radiating a signal at 2.6 GHz. The primary goal of this research was to develop a passive, miniature and effective frequency doubler integrated with planar antennas. The system is referred to as a frequency doubling reflectenna, (FDR). Prediction of accurate performance was achieved by employing precise modeling and measurement methods. The FDR can be utilized in data collection applications. The footprint of the FDR is occupied primarily by the receive and transmit antennas.Therefore, a significant portion of the research focused on the development of compact and efficient planar antennas, which would provide for a miniature FDR. A first-generation FDR design was designed, which utilized quarter-wavelength shorted microstrip patch antennas. The choice of antennas provided a small prototype with dimensions equal to 44 mm by 17 mm. In order to further reduce the size of the harmonic re-radiator, meandered planar antennas were investigated and optimized for efficient operation. A second-generation FDR design, which utilized meandered microstrip patch antennas, was produced and a size reduction of 75% was achieved. Both first- and second-generation harmonic re-radiator designs were designed for low input power operation and provided maximum measured conversion efficiencies of approximately 4.5% and 1.8%, with the input to the diode doubler at -14.5 and -17.5 dBm, respectively.Re-configurable microwave devices, which dynamically operate at different frequencies, are often desirable for sensing applications. Therefore, to conclude this research, a tunable FDR was realized using a semiconductor varactor that provided the dynamic capacitance required for the tunability.
Thesis:
Thesis (M.S.E.E.)--University of South Florida, 2008.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Suzette Marie Presas.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 109 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002001162
oclc - 319637461
usfldc doi - E14-SFE0002614
usfldc handle - e14.2614
System ID:
SFS0026931:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 002001162
003 fts
005 20090429153643.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 090429s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002614
035
(OCoLC)319637461
040
FHM
c FHM
049
FHMM
090
TK145 (Online)
1 100
Presas, Suzette Marie.
0 245
Microwave frequency doubler integrated with miniaturized planar antennas
h [electronic resource] /
by Suzette Marie Presas.
260
[Tampa, Fla] :
b University of South Florida,
2008.
500
Title from PDF of title page.
Document formatted into pages; contains 109 pages.
502
Thesis (M.S.E.E.)--University of South Florida, 2008.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
520
ABSTRACT: In this thesis the development of a high efficiency harmonic re-radiator, consisting of a diode doubler and conjugate-matched receive and transmit antennas, is described. Diode-based frequency multipliers and rectifiers, coupled with antennas, are of interest for quasi-optical applications, for energy-scavenging and for sensing applications. The device studied operates by receiving an interrogating signal at a frequency of 1.3 GHz and re-radiating a signal at 2.6 GHz. The primary goal of this research was to develop a passive, miniature and effective frequency doubler integrated with planar antennas. The system is referred to as a frequency doubling reflectenna, (FDR). Prediction of accurate performance was achieved by employing precise modeling and measurement methods. The FDR can be utilized in data collection applications. The footprint of the FDR is occupied primarily by the receive and transmit antennas.Therefore, a significant portion of the research focused on the development of compact and efficient planar antennas, which would provide for a miniature FDR. A first-generation FDR design was designed, which utilized quarter-wavelength shorted microstrip patch antennas. The choice of antennas provided a small prototype with dimensions equal to 44 mm by 17 mm. In order to further reduce the size of the harmonic re-radiator, meandered planar antennas were investigated and optimized for efficient operation. A second-generation FDR design, which utilized meandered microstrip patch antennas, was produced and a size reduction of 75% was achieved. Both first- and second-generation harmonic re-radiator designs were designed for low input power operation and provided maximum measured conversion efficiencies of approximately 4.5% and 1.8%, with the input to the diode doubler at -14.5 and -17.5 dBm, respectively.Re-configurable microwave devices, which dynamically operate at different frequencies, are often desirable for sensing applications. Therefore, to conclude this research, a tunable FDR was realized using a semiconductor varactor that provided the dynamic capacitance required for the tunability.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Thomas M. Weller, Ph.D.
653
Frequency doubling reflectenna
Diode doubler
Conversion efficiency
Compact microstrip antenna
Meandered antennas
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2614



PAGE 1

Microwave Frequency Doubler Integrated with Miniaturized Planar Antennas by Suzette Marie Presas A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering Department of Electrical Engineering College of Engineering University of South Florida Major Professor: Thomas M. Weller, Ph.D. Andrew M. Hoff, Ph.D. James T. Leffew, Ph.D. Date of Approval: May 22, 2008 Keywords: frequency doubling reflectenna diode doubler, conversion efficiency, compact microstrip antenna, meandered antennas, tunable Copyright 2008, Suzette Marie Presas

PAGE 2

To my father Pompeyo Aguilar, for showing me among many things the concept of To my mother Sonia Wiemer, for teaching me about persistence. To my Nina, for giving me her love and strength; you are the light that keeps me going.

PAGE 3

i TABLE OF CONTENTS LIST OF TABLES iii LIST OF FIGURES iv ABSTRACT viii PREFACE x CHAPTER 1 INTRODUCTION 1 1.1 Frequency Multiplication ............................................................................ 2 1.2 Frequency Multipliers and Detector s Integrated with Antennas ................ 4 1.3 Frequency Doubling Reflectenna ............................................................... 8 1.4 Overview and Contributions of the Research ............................................. 10 CHAPTER 2 FREQUENCY DOUBLING REFLECTENNA 1.3 GHz 2.6 GHz 11 2.1 Introduction ................................................................................................. 11 2.2 Circuit Overview ......................................................................................... 13 2.2.1 Schottky Diode Modeling ............................................................... 14 2.2.2 Circuit Modeling of the Frequency Doubling Reflectenna ............. 18 2.3 Quarter-Wavelength Shorted Patch Antennas ............................................ 23 2.3.1 Design ............................................................................................. 23 2.3.2 Simulated and Measured Radiation Patterns .................................. 27 2.4 Power Measurement Techniques ................................................................ 32 2.5 Results ...................................................................................................... 35 2.6 Conclusions ................................................................................................. 38 CHAPTER 3 COMPACT PLANAR ANTENNAS: MEANDERED ANTENNAS 40 3.1 Introduction ................................................................................................. 40 3.2 Review of Antenna Parameters .................................................................. 41

PAGE 4

ii 3.3 Limitations of Electrically Small Antennas ................................................ 46 3.4 Size Reduction Techniques for Planar Antennas ........................................ 49 3.5 Meandered Slot Antenna ............................................................................. 52 3.6 Meandered Shorted Patch Antenna ............................................................. 62 3.6.1 Antenna Design ............................................................................... 62 3.6.2 Substrate Thickness and Antenna Efficiency Considerations ........ 68 3.6.3 Radiation Patterns ........................................................................... 75 3.7 Conclusions ................................................................................................. 77 CHAPTER 4 A MINIATURIZED AND TUNABLE FREQUENCY DOUBLING REFLECTENNA 79 4.1 Introduction ................................................................................................. 79 4.2 Miniaturized 1.3 GHz 2.6 GHz FDR Design ........................................... 80 4.3 Miniaturized and Tunable FDR Design ...................................................... 84 4.4 Results ...................................................................................................... 92 4.5 Conclusions .................................................................................................. 96 CHAPTER 5 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK 98 5.1 Summary ...................................................................................................... 98 5.2 Recommendations .......................................................................................100 REFERENCES 103 APPENDICES 108 Appendix A: Copper Etching Process Flow.........................................................109

PAGE 5

iii LIST OF TABLES Table 2.1: P-N Junction Diode Model Parameters ....................................................... 15 Table 2.2: HSCH-9161 Beamlead Diode Parameters .................................................. 17 Table 3.1: Parameters for a Half-Wavelength CPW-Fed Slot Antenna Design (Dimensions in Millimeters) ....................................................................... 52 Table 3.2: Parameters for Meandered Slot Antenna Design (Refer to Figure 3.5) (Dimensions in Millimeters) ....................................................................... 54 Table 3.3: Parameters for the Meandered Sl ot Antenna Design (Refer to Figure 3.6) (Dimensions in Millimeters) ....................................................................... 55 Table 3.4: Parameters for Slot Antenna Designs (Refer to Figures 3.4 -3.6) .............. 60 Table 3.5: Parameters for Slot Antenna Designs (Refer to Figures 3.4 -3.6) .............. 61 Table 3.6: Parameters for 1.3 GHz a nd 2.6 GHz Meandered Shorted Patch Antenna Designs of Figure 3.13 (Dimensions in Millimeters) ................... 65 Table 3.7: Comparison of Simulated Pa rameters of Quarter-Wavelength Shorted Patch Antennas and Meande red Shorted Patch Antennas ............. 74 Table 3.8: Size Reduction between Qu arter-Wavelength Shorted Patch Antennas and Meandered Shorted Patch Antennas .................................... 74

PAGE 6

iv LIST OF FIGURES Figure 1.1: Operating Principle of a Frequency Multiplier Using Nonlinear Capacitance, [3] ...................................................................... 3 Figure 1.2: Quasi-Optical Frequency Diode Doubler Scheme, [9] ............................ 5 Figure 1.3: Rectenna Block Diagram, [13] ................................................................ 6 Figure 1.4: Frequency Doubling Refl ectenna Concept, [34] ..................................... 9 Figure 2.1: Layout of a Frequenc y Doubling Reflectenna with Conjugate-Matched Impedances .............................................................. 12 Figure 2.2: Circuit Model for the HSCH-9161 Zero-Bias Beamlead Detector Diode ......................................................................................... 14 Figure 2.3: Comparison of Measured and Modeled Current versus Bias Voltage of the Schottky Diode in the Forward Bias Region ........ 18 Figure 2.4: Comparison of Measured and Modeled Current versus Bias Voltage of the Schottky Diode in the Reverse Bias Region ............ 18 Figure 2.5: Equivalent Circuit Mode l of a Single FDR Antenna for Computer-Aided Analysis ....................................................................... 19 Figure 2.6: Equivalent Circuit Mode l of the Diode and the Input and Output Tuning Circuits of the FDR for Computer-Aided Analysis .................................................................................................... 20 Figure 2.7: Overall Equivalent Circ uit Model of Frequency Doubling Reflectenna for Compute r-Aided Analysis .............................................. 20 Figure 2.8: Simulated Conversion Gain and DC Current for the Diode Doubler ......................................................................................... 21 Figure 2.9: Antenna Test Board ................................................................................. 23 Figure 2.10: Comparison of Measured and Simulated S11 of the Preliminary 1.3 GHz Antenna Fabricated on an Arlon Substrate ................................ 24

PAGE 7

vFigure 2.11: Simulated Input Impedance: 1.3 GHz Antenna (left) and 2.6 GHz Antenna (right) .......................................................................... 25 Figure 2.12: Comparison of Measured and Simulated S11 of the 1.3 GHz Antenna ...................................................................................... 26 Figure 2.13: Comparison of Measured and Simulated S11 of the 2.6 GHz Antenna........................................................................................ 27 Figure 2.14: Coordinate System Used for Antenna Simulation ................................... 27 Figure 2.15: Axis Definition for Antenna Measurements ............................................ 29 Figure 2.16: Comparison of Measured and Simulated Co-Polarized E-Plane (left) and H-Plane (right), Pattern of the 1.3 GHz Antenna, (dB) .................... 30 Figure 2.17: Comparison of Measured and Simulated Co-Polarized E-Plane (left) and H-Plane (right), Pattern of the 2.6 GHz Antenna, (dB) .................... 31 Figure 2.18: Cross-Polarized Measurement of the E and H-Plane Radiation Pattern, 1.3 GHz Antenna (left) and 2.6 GHz Antenna (right), (dB) .................... 32 Figure 2.19: Fabricated Frequency Doubling Reflectenna ........................................... 33 Figure 2.20: Diagram of the Harmonic Re-radiator Test Measurement Setup............... 34 Figure 2.21: Hardware Test-Bench Used for FDR Measurements .............................. 35 Figure 2.22: Comparison of Measured a nd Expected Output Power and Diode Doubler Conversion Gain at a Sour ce Frequency of 1.3 GHz ................. 36 Figure 2.23: Measured Received Power for Different Source Frequencies.................... 37 Figure 2.24: Comparison of Expected Doubler Conversion Gain and Output Power at a Source Frequency of 1.3 GHz and Measured Performance at the Fabricated FDRs Optimum Source Frequency ............................. 37 Figure 3.1: Spherical Coordinate System ................................................................... 42 Figure 3.2: Antenna Within a Sphere of Radius r [42] ............................................. 47 Figure 3.3: Configuration of Different Planar Antennas: Microstrip Patch (left); Printed Dipole (middle); Slot Antenna (right), [2] .................................. 50 Figure 3.4: CPW-Fed Slot An tenna, (Design A) ........................................................ 52 Figure 3.5: CPW-Fed Meandered Slot Antenna (Design B) ...................................... 54

PAGE 8

viFigure 3.6: CPW-Fed Meandered Slot Antenna (Designs C, D, E) ........................... 55 Figure 3.7: Simulated S11, (Magnitude), of CPW-Fed Slot Antenna Designs A E ..................................................................... 56 Figure 3.8: Simulated S11, (Phase), of CPW-Fed Slot Antenna Designs, A E .................................................................... 57 Figure 3.9: Simulated Real Input Impedance for CPW-Fed Slot Antenna Designs, A E ................................................................... 58 Figure 3.10: Simulated Imaginary Input Impedance for CPW-Fed Slot Antenna Designs, A E ................................................................... 58 Figure 3.11: 1.3 GHz Simulated EPlane Patterns for CPW-Fed Slot Antenna Designs, A E ................................................................... 59 Figure 3.12: Geometry of the Meandered Shorted Patch Antenna, (1.3 GHz Design) ..................................................................................... 65 Figure 3.13: 1.3 GHz, (top), and 2.6 GHz, (bottom) Meandered Shorted Patch Antennas (at 2.6 GHz; via at x = 1.3 mm and y = 0.5 mm) ..................... 66 Figure 3.14: Comparison of Measured and Simulated S11 of the 1.3 GHz Antennas .......................................................................... 67 Figure 3.15: Comparison of Measured and Simulated S11 of the 2.6 GHz Antennas .......................................................................... 67 Figure 3.16: Plot of Simula ted Input Impedance vs. Frequency for Meandered Shorted Patch Antennas: 1.3 GHz (left) and 2.6 GHz (right) ................. 68 Figure 3.17: Plot of Simulated E-Plane Radiation Pattern for 3 Different Substrate Thicknesses for 1.3 GHz Meandered Shorted Patch Antenna ................. 72 Figure 3.18: Plot of Simulated E-Plane Radiation Pattern Gain vs. Frequency for 3 Different Substrate Thicknesses for 1.3 GHz Meandered Shorted Patch Antenna .................................................................................................... 73 Figure 3.19: Comparison of Measured a nd Simulated Co-Polarized E-Plane (left) and H-Plane (r ight), Pattern of the 1.3 GHz Antenna, (dB) ........... 76 Figure 3.20: Comparison of Measured a nd Simulated Co-Polarized E-Plane (left) and H-Plane (r ight), Pattern of the 2.6 GHz Antenna, (dB) ........... 76 Figure 3.21: Cross-Polarized Measurement of the E and H-Plane Radiation Pattern of the 1.3 GHz Antenna (left) and 2.6 GHz Antenna (right), (dB) .......... 77

PAGE 9

viiFigure 4.1: Overall Equivalent Circu it Model of Miniaturized Frequency Doubling Reflectenna for Com puter-Aided Analysis .............................. 82 Figure 4.2: Simulated Conversion Gain for First-Generation FDR Diode Doubler and Miniaturized FDR Diode Doubler ...................................... 83 Figure 4.3: Illustration of Miniaturized Frequency D oubling Reflectenna ................ 84 Figure 4.4: Circuit Model for the Metelics MSV34,067-0805 Varactor .................... 86 Figure 4.5: Layout of Tunable Frequency Doubling Reflectenna .............................. 88 Figure 4.6: Overall Equivalent Circuit Model of Tunable Frequency Doubling Reflectenna for Com puter-Aided Analysis .............................. 89 Figure 4.7: Simulated Conversion Gain for the Tunable FDR Diode Doubler: Maximum On Conversion Gain was Obtained when the Varactor was Biased at 17 V, (On-State) ....................................................................... 90 Figure 4.8: Simulated Conversion Gain for First-Generation FDR Diode Doubler, the Miniaturized FD R Diode Doubler and the Tunable FDR Diode Doubler when the Var actor was Biased at 17 V .................. 91 Figure 4.9: Comparison of Measured a nd Expected Doubler Power Received for the Miniaturized FDR at a Source Frequency of 1.3 GHz ................. 93 Figure 4.10: Comparison of Measured and Expected Doubler Conversion Gain for the Miniaturized FDR at a Source Frequency of 1.3 GHz ........ 93 Figure 4.11: Comparison of Measured Doubler Power Received for the Tunable FDR at a Source Frequency of 1.3 GHz (Solid Lines Represent Performance of the Diode when Var actor was Biased at 17 V) .............. 95 Figure 4.12: Comparison of Measured Doubler Conversion Gain for the Tunable FDR at a Source Frequency of 1.3 GHz (Solid Lines Represent Performance of the Diode when Varactor Biased at 17 V) ..................... 95

PAGE 10

viii Microwave Frequency Doubler Integrated with Miniaturized Planar Antennas Suzette Marie Presas ABSTRACT In this thesis the development of a high efficiency harmonic re-radiator, consisting of a diode doubler and conjugate-m atched receive and transmit antennas, is described. Diode-based frequency multipliers and rectifiers, coupled with antennas, are of interest for quasi-optic al applications, for energy-scavenging and for sensing applications. The device studied operat es by receiving an in terrogating signal at a frequency of 1.3 GHz and re-radiating a signa l at 2.6 GHz. The primary goal of this research was to develop a passive, miniatur e and effective frequency doubler integrated with planar antennas. The system is re ferred to as a frequency doubling reflectenna, (FDR). Prediction of accurate performance was achieved by employing precise modeling and measurement methods. The FDR can be utilized in data collection applications. The footprint of the FDR is occupied primarily by the receive and transmit antennas. Therefore, a significant portion of the research focused on the development of compact and efficient planar antennas, whic h would provide for a miniature FDR. A first-generation FDR design was designed, wh ich utilized quarter -wavelength shorted microstrip patch antennas. The choice of antennas provided a small prototype with

PAGE 11

ixdimensions equal to 44 mm by 17 mm. In order to further reduce the size of the harmonic re-radiator, meandered planar ante nnas were investigated and optimized for efficient operation. A second-generati on FDR design, which utilized meandered microstrip patch antennas, was produced a nd a size reduction of 75% was achieved. Both firstand second-generation harmonic re-radiator designs were designed for low input power operation and pr ovided maximum measured co nversion efficiencies of approximately 4.5% and 1.8%, with the input to the diode doubler at -14.5 and -17.5 dBm, respectively. Re-configurable microwave devices, which dynamically operate at different frequencies, are often desirable for sensing a pplications. Therefore, to conclude this research, a tunable FDR was realized using a semiconductor varactor that provided the dynamic capacitance required for the tunability.

PAGE 12

x PREFACE I am humbled by the opportunity of en countering great indi viduals that have believed in me throughout my life. My d eep appreciation goes to all of them. I will be eternally grateful to Professor Tom Weller for introducing me to the field of microwaves and for his guida nce and insight on this work and various other matters. I especially appreciate his enthusiasm of conveying a true understanding of the subject and his exceptional dedication to the field, both of which are treme ndous inspirations to me. I am extremely thankful to Dr. James Leffew for his enlight enment on Electrical Engineering topics and for pati ently editing the different porti ons of this thesis. I am grateful to Dr. Andrew Hoff for his sugges tions during the course of this work. Thank you, Henrry La Rosa for sharing your knowledge and time during all of our EE studies; you have shown me how to become a better engineer. Dr. Saravana P. Natarajan, you are a kind and generous charac ter, and I am grateful for all your help, especially for your assistance with fabrication and for your practical views of life. Bojana Zivanovic, I will always cherish our memories together, our travels, and the much-needed times when you have provided yo ur electromagnetic expertise, reason, and humor. I am indebted to Sergio Melais for providing guidance on antenna radiation pattern analysis and for always making my days at 412 just brighter. Bojana and Sergio, you are family. I am thankful to Diana Ar istizabal and Sam Baylis for furthering my understanding of RF techniques and early on serv ing as my microwave research mentors.

PAGE 13

xiThe encouragement given by Alberto Rodrguez to become an electromagnetic academic is much appreciated. My sincere regards go to Ebenezer Odu for your caring nature. I appreciate assistance with nanocrystalline diamond provided by Srinath Balachandran. For their company in ENB 412, I am gratef ul to: Quenton Bonds, Evelyn Benabe, Tony Price, Aswin Jayaraman, Dr. To m Ricard and Daniel Sosa Mart in. For conversations that enriched my life in various ways, my regards go to: James Mcknight, Scott Skidmore and Lance Nettles. I appreciate the fabrica tion assistance provid ed by Subbu Krishnan. Special thanks go to Norma Paz for her sweet encouragement at various times. For financial support, I am in gratitude to Raytheon Systems (Fullerton, CA), to the NSF (NIRT Project ECS-0404137), to the NSF IGERT Program (Grant DGE 0221681), and to the USF Graduate School, especi ally Mr. Rod Hale. Special thanks to Modelithics, Inc. for allowing me to use their lab facilities and measurement equipment. I would have never come this far wit hout my familys and friends love and support. I wish to express my deep appreciation to friends that have always believed in what I can accomplish: Gail, Alex, Mario, Jessica, and Diana. For her incredible kindness and strength, I would like thank Loly Valmaa. I am grateful to my mother for encouraging me to approach life with a graceful spirit, and to my father, my scholar, for teaching me to love all knowledge for the sa ke of knowledge with the passion that only you could have. I would like to thank my sist er Tatiana, to me the most unique person in the world; your doses of comedy and sensitiv ity remarkably enrich my life every day. Finally I would like to give eternal lo ving thanks to Axel; I will be forever appreciative of your noble and phi losophical spirit and your char isma. My infinite thanks go to you for revealing such a marvelous world of hopefulness to me.

PAGE 14

1 CHAPTER 1 INTRODUCTION Great advances have been made in micr owave integrated circ uit technology since the development of planar transmission lines and microwave so lid-state devices. Microstrip lines have become an integral pa rt of microwave integrated circuits due to their low-cost, small size, ease in manufact uring, ease of active and passive device integration and good repeatability and reproducibility associated with these transmission lines, [1]. In addition, since their first prac tical implementations in the 1970s, microstrip antennas have been one of the most commonly used radiators for integrated circuit applications. In the area of solid-state devi ces, research associated with semiconductor diodes has led to the development of com ponents such as frequency multipliers. Numerous studies have concen trated on building more efficient and compact integrated microwave circuits. As a result, microstrip antennas have been integrated with other microstrip circuits to produce small designs and to maximize the efficiency of solid-state devices, which may be employed in the circuitr y, [2]. Additionally, advances realized in computer-aided design and in manufacturing and processing technique s greatly assisted the development of novel integrated designs. The research described in this thesis focused on producing a simple, low-cost, low-power and miniature diode frequency multiplier integrated with planar antennas.

PAGE 15

2Integrating antennas with other components has been accomplished in quasi-optical applications and wireless power transmissi on. However, the main application for the devices described in this thesis is in wireless sensors nodes. This research concentrated on a passive harmonic re-radiator scheme. The reradiator receives an interr ogation signal at a fundamental frequency and re-radiates a signal back to a transceiver at the second ha rmonic. Harmonic re-radiation can alleviate the clutter radiation that occurs at the fre quency of the incident radiation in a radar transceiver system. Since sensor nodes may be located in remote areas and subjected to harsh conditions, a robust, compact, effici ent and reliable device was desired. Consequently, the design consists of a Scho ttky beamlead diode, which is used as the frequency multiplier, and compact microstrip antennas, which receive and transmit the signal. Schottky beamlead diodes have b een proven to perform well at microwave frequencies and microstrip antennas are of ten utilized in low-power transmitting and receiving applications, [2]. 1.1 Frequency Multiplication Frequency multipliers have been used pr imarily as signal generators to produce high-frequency local oscillator signals. Fr equency multiplication is possible due to the inherent nonlinearities present in certain electronic compone nts. When these nonlinear devices are excited with a sinus oidal waveform at one partic ular frequency, the resulting response is a distorted waveform composed of harmonics. The operating principle of harmonic generation for a device employing nonlin ear capacitance is presented in Figure 1.1.

PAGE 16

Figure 1.1: Operating Principl e of a Frequency Multiplier Using Nonlinear Capacitance, [3] A nonlinear charge-voltage characteristic can be expressed as a power series about an operating point, VB, which denotes the bias volta ge, [3]. The power series expression for the bias voltage is given by: 3 3 2 2 1)( VbVbVbbVVQo B. (1.1) The input voltage applied to the device is given by: )cos( twVVg d (1.2) Substitution of equation (1.2) into equation (1.1) yields: 3

PAGE 17

)3cos()2cos()cos( )(3 2 1twQtwQtwQQtQg g g o, (1.3) which is an expression that high lights the input signal harmonics. In passive frequency multiplier designs, a device with a nonlinear current-voltage (varistor) or capacitance-voltage (varactor) characteristic may be utilized. In fact, real devices present both nonlinearitie s. Devices utilized in a frequency conversion process must have strong nonlinearities and repeatab le electrical properties. As a result, semiconductor p-n junction diodes and metal-semic onductor junction diodes, which are usually called Schottky barrier or Schottky di odes, are often utilized Schottky barrier diodes are usually preferred since they are majority carrier devices and metalsemiconductor junctions can be fabricated in a more precise and repeatable manner, [3]. Multipliers based on nonlinear transmission lines have also been reported, [4-6]. 1.2 Frequency Multipliers and Det ectors Integrated with Antennas Diode-based frequency multipliers and r ectifiers, coupled with antennas, have been of interest in quasi-optical applications, in wireless power transmission and in wireless sensors applications. Numerous works have been presented on frequency multipliers integrated with antennas for qua si-optical receiver a pplications and power combining architectures at millimeter and sub-millimeter wave frequencies, [7-11]. One of these architectures is illustrated in Figure 1.2. 4

PAGE 18

Figure 1.2: Quasi-Op tical Frequency Diode Doubler Scheme, [9] Since the output power of solid-state devices decreases as the frequency of operation increases, quasi-optical techniques ar e used to combine the power of several solid-state devices. Moreover, diodes coupled with antennas have al so been utilized in wireless power transmission applications, [ 12-14]. For instance, a high-sensitivity CMOS rectifier, which can recharge a seconda ry battery for sensor network systems and be utilized in a low-power wireless transm ission system has been reported, [14]. Similarly, rectennas (rectifyi ng antennas) have been used to receive power where no physical connections are possible. Several types of rectennas have been presented where the main parameter of interest is the RF-to-dc conversion efficiency. A dual-fr equency rectenna operating at both 2.45 GHz and 5.8 GHz has been successf ully designed. Th e design utilized a printed dual-frequency dipole antenna. The antenna was integrated with low-pass and bandstop filters to block highe r order harmonics generated from a GaAs Schottky barrier diode, [12]. The diode is directly connected to the filters and the conversion efficiency maximized by matching the diodes input impeda nce to the filters im pedance. Chang et al., also developed a similar design where the rectenna circuitry consists primarily of a high-gain antenna and a Schottky diode. These components are integrated with a band5

PAGE 19

reject filter, a dc filter and a resistive load A basic block diagram of this device is presented in Figure 1.3, [13]. Figure 1.3: Rectenna Block Diagram, [13] In both of these rectenna designs, RF-to-dc conversion efficiencies of approximately 80% were achieved for received powers on the rectennas antenna of 15.5 20 dBm. In the context of wireless sensors, Ketterl et al. introduced a reflectenna design that consists of a receive antenna, a device, such as a switch, that modulates a received continuous wave and a transmit antenna that re directs the modulated signal back to the transmission source, [15]. This design wa s intended for a low-power application. Additionally, a miniature harmonic radar transc eiver design, which ut ilizes a short length of wire to form the antenna and a beamlead Schottky diode as the doubling element, has been reported and utilized for insect tracking, [16]. The authors of this study employed a harmonic radar receiver that was tuned to th e second harmonic of the transmitter. This configuration eliminates the conventional radar scatter. Riley et al., also presented transponders, which consisted of a similar de sign, for insect tracking purposes as well, [17]. Other types of wireless sensor circui ts include RF identification (RFID) transponders and surface acoustic wave (SAW) sensors. These sensors can be classified as active devices, which are powered by a ba ttery and as passive transponder devices, 6

PAGE 20

7which are not battery powered. One of the first applications of SAW sensors was for temperature sensing, [18]. A one-port re flective SAW device was connected to an antenna and interrogated by an RF signal, [18-21]. The SAW device usually consists of a piezoelectric substrate with metallic structures which are interdigital transducers on the polished surface. A SAW is stimulated at th e surface due to an el ectric signal at the interdigital transducer. Temperature, mech anical stress or bending can cause a delay change or frequency shift of the retransmitted signal sent by the SAW device to a transceiver. SAW sensors for humidity, pre ssure, position, accelerat ion, wear, magnetic field and electric current ha ve been developed, [18]. In a similar fashion, RFID tags also receive an interrogator signal from a unit designated as the reader. Thes e transponders are utilized in areas such as manufacturing, retailing, transport and securi ty, [22]. The RFID tags cons ist of an antenna and of an integrated circuit (IC) chip, whic h is capable of storing informa tion. Data transfer occurs between the reader and the tag through antenna s linked to both ends. Passive RFID tags, which are the most widely used, are powe red by restoring the electromagnetic energy emitted from the reader. The transponders communicate with the reader by modulating the RF energy received and by creating a b ackscatter signal. A study of modulated backscatter RFID transp onders can be explored further in [23]. A complete analysis for the estimation of the backscattered radiated fiel d of a UHF RFID tag is presented in [24]. Different types of antennas have been proposed for RFID tags such as dipole antennas, [25], planar inverted-F antennas, [26-27], and printed patch antennas, [28]. The detector diode most frequently used in the IC of the RFID tag is the Schottky barrier diode.

PAGE 21

8 Several authors have demonstrated that higher conversion efficiencies can be obtained by using active frequency multipliers in tegrated with antennas, [29-30]. Itoh et al. have extensively explored the concept of active integrated antennas with frequency multipliers, [31-33]. Although higher conversion efficiencies are obtainable, some active frequency multipliers involve complex circuitry, due to biasing considerations, which may lead to higher manufacturing costs. 1.3 Frequency Doubling Reflectenna The research presented in this thesis centers on the design of a diode frequency doubler integrated with antennas. The devi ce is referred to as a Frequency Doubling Reflectenna, (FDR). This device was create d primarily for sensing applications. The FDR receives a signal at a fundamental freque ncy and re-radiates the signal back to a transceiver at the second harmonic. It was of interest to obtain a design with minimum conversion loss for a low-power input (-30 dBm) as well as sufficie nt sensitivity for the data collection application. The energy of the interrogator signal was used effectively to power the solid-state device employed, which produced a simple passive design, [34]. Figure 1.4 demonstrates the FDR concept.

PAGE 22

Figure 1.4: Frequency Doubli ng Reflectenna Concept, [34] Efficiency and compactness were primarily achieved by conjugate matching the input impedances of the receive and transmit antennas to the diode te rminals. Moreover, reduction in size was attained by utilizing el ectrically small patch antennas. Proper simulation methods and modeling techniques proved to be essential for the accurate prediction of device performance. It was al so of importance to demonstrate the concept of a tunable FDR. In the tunable device, maximum re-radiation occurs at the first harmonic in an on-state, and no re-radiation occurs in an off-state. Tunability was provided by a semiconductor varactor, which is a variable reactance de vice. Features of the device such as its miniature size, frequenc y selectivity and reliable operation, make the FDR very useful in sensing applicatio ns. Although this thesis refers to a FDR operating at 1.3 GHz 2.6 GHz, the design can be considered a proof-of-concept and scaled to other operating frequencies. 9

PAGE 23

101.4 Overview and Contributions of the Research The primary goals of this research were Accurately predict the behavior of a frequency doubler integrated with receive and transmit antennas, Design the FDR in a manner that would maximize conversion efficiency, Produce a simple and compact integrated device. Chapter 2 presents the work performed in designing, fabricating, and testing a 1.3 GHz 2.6 GHz frequency doubling reflectenna. Chapter 2 also addresses topics such as the Schottky diode model used in the doubler, the circuit schematic modeling of the FDR, the quarter-wavelength shorted patch an tennas utilized in the first-generation design and the measurement test-bench employ ed. It is noteworthy to mention that a conversion efficiency of 1% was obtained for a low input power of -30 dBm. The following two chapters describe the progression toward a miniature and tunable FDR design. Chapter 3 provides a study of current electr ically small antenna technology with an emphasis on meandered antennas. In addition, Chapter 3 also presents a description of the design of the miniature meandered patch antennas utilized in a second-generation 1.3 GHz 2.6 GHz FDR design. The miniature meandered patch antennas represent approximately an 85% re duction in size from the patch antennas utilized in the first-generation design. In Chapter 4, a miniature and tunable FDR is presented. A summary of the fi ndings of this research as well as suggestions for future work are presented in Chapter 5.

PAGE 24

11 CHAPTER 2 FREQUENCY DOUBLING REFLECTENNA 1.3 GHz 2.6 GHZ 2.1 Introduction Frequency multipliers based on diodes offe r low-cost, high reliability and can be easily integrated with other semiconductor components by utilizing planar technology. Diode-based frequency multipliers and rectifiers, coupled with antennas, are of interest in quasi-optical applications, energy-scavenging, and sensing applications, [7-33]. In addition to obtaining a compact device, an integrated diode frequency multiplier/antenna module alleviates the requireme nt to separately design for the two components and then combine them under certain impedance constraints. In this chapter, a frequency doubling reflecte nna, (FDR), which receives a 1.3 GHz interrogation signal and re-radiates a 2. 6 GHz signal, is described. The topology of the FDR is presented in Figure 2.1.

PAGE 25

Diode ZSZLZS *ZL 1.3 GHz Antenna Tuning Topologies 2.6 GHz Antenna 44 mm17 mm Figure 2.1: Layout of a Freque ncy Doubling Reflectenna with Conjugate-Matched Impedances The FDR is intended for sensing applications. Therefore, to enhance the ability of the interrogating transceiver to detect a very low-power return signal, the FDR utilizes frequency discrimination. A GaAs Schottky diode was used as the doubling element. The GaAs Schottky diode is a proven reliab le device for generating power at microwave frequencies compared to other semiconductor p-n junction diodes. The conversion loss of a diode frequency multiplier is minimized when the source impedance is very close to the complex conjugate of the input impedan ce of the diode and lik ewise with the load impedance, [3]. Both the receive and tran smit antenna were designed to be conjugatematched to the diode multiplier in order to minimize the footprint and maximize the conversion efficiency. As illustrated in Figure 2.1, ZS and ZL are the input impedances of the receive and transmit antenna s, respectively. The input and output impedances of the diode circuit are presented as the complex conjugates of Zs and ZL. Integrating antennas with other components has been investigated in a similar fashion in quasi-optical receiver applications, [9]. Utilizing this method of component in tegration provided a doubler conversion efficiency of 1% at a -30 dBm input power level. A lthough DC bias could improve 12

PAGE 26

conversion efficiency, it was not applied in th e harmonic re-radiator in order to maintain a simple design. Quarter-wavelength, shorte d microstrip patch antennas were used to reduce the overall size of the FDR, [2]. In this design, the antennas provide a natural DC return for the doubler. In addition, they also inherently provide the proper harmonic impedance at 2.6 GHz at the input and at 1. 3 GHz at the output. Due to the conjugatematch approach, the resulting design was very narrowband and the desired sensitivity was achieved. Inclusion of the correct harm onic terminations in circuit simulations for the FDR yielded comparisons between measur ed and predicted conversion efficiency within tenths of a dB. 2.2 Circuit Overview A GaAs Schottky diode was utilized in the harmonic re-radiator to obtain frequency multiplication. The GaAs Schottky diode is a nonlinear device, which accepts the signal received by the input antenna of the FDR and produ ces the required harmonic. Tuning circuits were used to provide op timum source and load impedances to the Schottky diode at the input and output freque ncy. The output antenna of the FDR reradiates the second order harm onic of the received signal. In order to predict the performance of the harmonic re-radiator, simulations were performed using Agilents Advanced Design System (ADS) CAD software. The main FDR design parameter of interest wa s the doubler conversion efficiency which is given by in out nP P (2.1) 13

PAGE 27

Pin is the input power to the diode multiplier, at 1.3 GHz, and Pout is its output power, at 2.6 GHz. Semiconductor diodes are lossy passiv e devices. As a result, the conversion efficiency of a diode frequency multiplier is less than unity. 2.2.1 Schottky Diode Modeling A Schottky-barrier diode is a semic onductor device that exhibits nonlinear properties due to its metal-semiconductor contact. An electrostatic barrier arises due to unequal work functions between the me tal, which forms the anode and the semiconductor, which forms the cathode. Conduction is controlled by thermionic emission of majority carriers across the barrier. The elec trical characteristics of a Schottky diode are mainly determined by the metal and semiconductor contacting surfaces. High-frequency operation is optimized when the diode possesses low series resistance, low junction capacitance, high carrier mobility and high carrier saturation velocity, [3]. The nonlinear model for the HSCH-9161 beam lead diode used in the doubler is presented in Figure 2.2. Figure 2.2: Circuit Model for the HSCH-9161 Zero-Bias Beamlead Detector Diode 14

PAGE 28

15The model consists of parasitic elements and an anti-parallel diode pair. The diode designated D1 represents the characteristic under forward bias of the HSCH-9161 and D2 (in the forward direction) give s the V-I curve under reverse bias of the HSCH-9161. The anti-parallel diode pair is utilized due to the high leakag e of the diode under reverse bias. The p-n junction diode model parameters are presented in Table 2.1. Table 2.1: P-N Junction Diode Model Parameters Name Description Unit IS Saturation current A Rs Ohmic resistance N Emission coefficient Cj0 Zero-bias junction capacitance pF Vj Junction potential V M Grading coefficient XTI Saturation-current temperature exponent EG Energy gap eV BV Reverse break down voltage V IBV Current at reverse break down voltage A

PAGE 29

16 The ADS default p-n junction diode model parameters are presented in Table 2.2. Also shown in Table 2.2 are the SPICE mode l parameters that were obtained from the Agilent HSCH-9161 zero-bias beamlead detector diode data sheet. These parameters were used in order to convert the ADS basic p-n junction diode model. These SPICE model parameters were initially utilized in the computer-aided simulations. However, the ideality factor, ( N) and zero-bias junction capacitance, ( Cjo), parameters were optimized in the zero-bias region. Direct current I-V measurements, in the forward and reverse bias regions, were performed using a Keithley 4200 Semiconductor Characterization System. The characterization was accomplished by employing ground-signal-ground probing on a biased diode, which was mounted on a test fixture. After comparing measured and simulated I-V data, the ideality factor and ze ro-bias junction capacitance obtained were N = 1.35 and Cjo = 0.035 pF for D1 and N = 56 and Cjo = 0.035 pF for D2. An ideality factor that is not equal to 1 usually indicates tunneli ng of electrons through the electrostatic barrier between the metal and the semiconduc tor of the Schottky diode. The HSCH-9161 device model parameters presented include the optimized N and Cjo values. Throughout the research, these last set of parameters were utilized for the computer-aided simulations of circuits th at involved the HSCH-9161 beamlead detector diode.

PAGE 30

17Table 2.2: HSCH-9161 B eamlead Diode Parameters HSCH-9161 SPICE Model Parameters HSCH-9161 Device Model Parameters Name P-N Junction Diode Model Default D1 D2 D1 D2 Is 10-14 12 x 10-6 84 x 10-6 12 x 10-6 84 x 10-6 Rs 0.0 50 10 50 10 N 1.0 1.2 40 1.35 56 Cj0 0.0 0.030 0.030 0.035 0.035 Vj 1.0 0.26 0.26 0.26 0.26 M 0.5 0.5 0.5 0.5 0.5 XTI 3.0 2 2 2 2 EG 1.11 1.42 1.42 1.42 1.42 BV 10 10 10 10 IBV 0.001 10-12 10-12 10-12 1 Figures 2.3 and 2.4 illustrate good agreement, between the diode model and the measured responses, for the two diode samp les in both the forward and reverse bias region near zero-bias.

PAGE 31

0 0.2 0.4 0.6 0.8 1Voltage (V) 0.0001 0.001 0.01 0.1 1 10 100C u r r e n t ( m A ) Diode Model (Sim) Sample 1 (Meas) Sample 2 (Meas) Figure 2.3: Comparison of Measur ed and Modeled Current versus Bias Voltage of the Schottky Di ode in the Forward Bias Region -0.1 -0.08 -0.06 -0.04 -0.02Voltage (V) -0.05 -0.04 -0.03 -0.02 -0.01 0 0.01C u r r e n t ( m A ) Diode Model (Sim) Sample 1 (Meas) Sample 2 (Meas) Figure 2.4: Comparison of Measur ed and Modeled Current versus Bias Voltage of the Schottky Di ode in the Reverse Bias Region 2.2.2 Circuit Modeling of the Frequency Doubling Reflectenna The Schottky diode is a nonlinear de vice, which generates the harmonics necessary for service as a frequency multiplier Therefore, the effect of accounting for 18

PAGE 32

harmonic terminations in the computer-aided analysis was studied, [3]. Figure 2.5 presents the equivalent circuit model for an FDR antenna, which can represent either the receiving or transmitting antenna. Figure 2.5: Equivalent Circuit Model of a Single FDR Antenna for Computer-Aided Analysis In the design, produced during this research, the source and the load of the entire FDR circuit schematic are represented by the 1.3 GHz and 2.6 GHz antenna, respectively. The values of the impedances at the fundamental and harmonic frequencies were extracted from numerical electromagnetic simulations. Blocks that represented these antenna impedances at the frequencies of interest were added at the load and at the source of the diode doubler. To insure accurate simulation, data files were created and used as filter elements. These filter elements ensured that the proper source/load termination was presented to the doubler circu it at a given harmonic frequenc y. The equivalent circuit model for the diode and the i nput and output tuning circuits without the antennas, are presented in Figure 2.6. 19

PAGE 33

Figure 2.6: Equivalent Circuit Model of the Diode and the Input and Output Tuning Circuits of the FD R for Computer-Aided Analysis Figure 2.7 presents the overall equivalent circuit model for the FDR. 1.3 GHz antenna ZfoZ2foZ3foZ4fofilter elements 2.6 GHz antenna ZfoZ2foZ3foZ4fofilter elements diode model tuning circuits 1.3 GHz antenna 2.6 GHz antenna tuning circuits Equivalent Circuit Model of FDRFDR Figure 2.7: Overall Equivalent Ci rcuit Model of Frequency Doubling Reflectenna for Computer-Aided Analysis 20

PAGE 34

In order to determine the nu mber of harmonic impedances, ( n ), to be considered in the circuit simulations, an analysis was performed to discern differences in performance between the n and the n 1 order harmonics. Figure 2.8 presents comparisons of the conversion gain of the diode doubler for several combinations of fundamental and harmonic impedances. -50 -40 -30 -20 -10 0Voltage (V) -40 -35 -30 -25 -20 -15 -10C o n v e r s i o n Ga i n ( d B ) 0 0.2 0.4 0.6 0.8 1 1.2D C C u r r e n t ( m A) fundamental 2nd order 3rd order 4th order DC Current Figure 2.8: Simulated Conversion Gain and DC Current for the Diode Doubler The difference in predicted conversio n loss when considering only the 2nd order harmonic compared to including the 3rd order harmonic was ~5 dB at -30 dBm power input. However, the difference in th e conversion loss between the 3rd order and 4th order harmonics was almost null. Therefore, only harmonic impedances up to the 4th order harmonic were considered. The expected conv ersion loss to the FDR, at an input power level of -30 dBm, was approximately 20 dB This conversion loss was chosen as a benchmark for the remaining design considerations. 21

PAGE 35

22 Another important observation was th e predicted maximum conversion gain, which was -13.7 dB. The maximum conversi on gain was obtained at the onset of the diodes DC conduction at an input power of -15.8 dBm, [35]. This is also illustrated in Figure 2.8. This behavior agrees with the fact that the region of the diodes I-V curve with the most nonlinearity occurs at th e knee-voltage. This region of maximum nonlinearity serves to optimize the conversi on efficiency. At higher power levels, the conversion of RF-to-DC power re duces the multiplier efficiency. Ideally, no matching circuits are desi red in order to minimize the doubler conversion loss. However, the diode frequency doubler has increased sensitivity to variations in the source and load impeda nce, [35]. Therefore, additional network elements were added to the basic circuit topology for fine-tuning the conjugate-match between the doublers impedance and the antenna s impedance. A useful consequence of using quarter-wavelength shorted patch antennas, (Ref. Section 2.3), was that the input of the 1.3 GHz antenna appears as a short-ci rcuit at 2.6 GHz, which provided the desired 2nd harmonic termination. The open-circuited stub on the input side of the network was only used to fine-tune the conjugate-mat ch. At the output, the matching topology consisted of one series 4.3 nH inductor, one shunt 2.2 nH inductor and a shunt opencircuited stub. With these tuning elements, the input and output impedance of the doubler were (40 j278) and (44.7 + j369) respectively.

PAGE 36

2.3 Quarter-Wavelength Shorted Patch Antennas 2.3.1 Design The antennas were designed to be ap proximately quarter-wavelength shorted patch antennas. The ideal conjugate matching was to be performed without matching circuits in order to minimize losses. Thus, the radiation resistan ce of the antennas was initially set to a few ohms to match the im pedance of the doubler. To validate antenna simulations, preliminary test structures, cons isting of the two ante nnas and short and open calibration standards, were fabricated on an Arlon substrate of thickness 20 mils and a r of 4.5. A circuit prototyping milling machine was used and grounding vias were filled with conductive epoxy. Figure 2.9 illustrates the board utilized to test these antennas. 1.3 GHz Antenna 2.6 GHz Antenna Short Open Figure 2.9: Antenna Test Board A 3.5 mm coaxial calibration was performed, which set the reference plane at the edge of the board for measurements. Meas urements of the calibration standards were used to develop a model for the PCB edge-m ount connector and the feedline leading up to the reference plane, which was located ap proximately 1 mm from the antenna input. After the open-circuit standard was modeled it was compared with the measurements for 23

PAGE 37

the short-circuit standard. The comp arison indicated the need to add 1-2 resistors at the via locations in order to emulate the short-circuit response. The via resistance, which was also present in the shorted patch antenna circuits, effected the antennas return loss. Instead of being in the expected 0.5 dB range at the frequencies of interest, the return loss was of order of 10 dB. This ef fect is displayed in Figure 2.10. 1 1.5 2 2.5 3Frequency (GHz) -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0M a g n i t u d e o f S 1 1 ( d B ) Simulated Measured Figure 2.10: Comparison of Meas ured and Simulated S11 of the Preliminary 1.3 GHz Antenna Fabr icated on an Arlon Substrate The antenna efficiency depends on the radiation efficiency in accordance with Lr rRR R e (2.2) where Rr and RL are the radiation resistance and the loss resistance. The loss resistance, after transformation from the via location to the antenna input, increased by at least an order of magnitude with respec t to the radiation resistance and dramatically reduced the antenna efficiency. This shortcoming revealed an interesting characteristic of the quarter24

PAGE 38

wavelength antenna. The quarter-wavelength an tenna performance is highly sensitive to its shorting via resistance. A second harmonic re-radiator was designed, which accounted for this constraint. As mentioned previously, tuni ng circuits at the input and output of the diode frequency doubler resulted in a real im pedance on the order of 40 which provided an increase in the radiation resistance of the antennas. To conjugate-match to the diode doubler, the values of width and length of each of the antennas were varied. The dimensional variations were used to change the real a nd imaginary values of the input impedance. The ratio of width to length can be related to the ratio of Im{ZIN} to Re{ZIN}. An iterative optimization process was applied to th e ratio in order to obtain the desired input impedance for the antenna. Figure 2.11 presents the simulated input impedance investigations for the antennas. 1 1.5 2 2.5 3Frequency (GHz) 0 20 40 60 80 100R e a l Z i n ( ) -100 0 100 200 300 400I m a g i n a r y Z i n ( ) 34.31 305.03 RE {Zin} IM {Zin} 1 1.5 2 2.5 3Frequency (GHz) 0 20 40 60 80 100R e a l Z i n ( ) -400 -300 -200 -100 0 100I m a g i n a r y Z i n ( ) 40.34 -355.716 RE {Zin} IM {Zin} Figure 2.11: Simulated Input Impedance: 1.3 GHz Antenna (left) and 2.6 GHz Antenna (right) The resulting 1.3 GHz antenna measured 28 mm x 21.2 mm and the 2.6 GHz antenna measured 11.25 mm x 11 mm. 25

PAGE 39

Electromagnetic simulations revealed th at the overall loss resistance was reduced by shorting the antenna at each side with sets of three parallel vias. To further minimize this resistance, the vias were soldered to ground using copper wires instead of silver epoxy. The measured resistance in the vias was of the order of 0.1 The board substrate was also changed to 60 mil Taconic with an r of 6.15. Prior measurements of this material revealed it possessed a consistent r at the desired frequencies. Figures 2.12 and 2.13 present a comparison between the measured and simulated S11 data referenced to 50 for both antennas. The expected S11 da ta was obtained by assuming an infinite ground plane. The assumption was justified si nce the excitation ports available in the simulator yielded more accurate return loss values when compared to values obtained using excitation ports define d for a finite ground plane. 0.5 1 1.5 2 2.5 3Frequency (GHz) -0.6 -0.4 -0.2 0M a g n i t u d e o f S 1 1 ( d B ) -200 -100 0 100 200P h a s e ( d e g r e e s ) Figure 2.12: Comparison of Meas ured and Simulated S11 of the 1.3 GHz Antenna. Solid lines represent the simulated S11 26

PAGE 40

0.5 1 1.5 2 2.5 3Frequency (GHz) -1.6 -1.3 -1 -0.7 -0.4 -0.1M a g n i t u d e ( d B ) -200 -120 -40 40 120 200P h a s e ( d e g r e e s ) Figure 2.13: Comparison of Meas ured and Simulated S11 of the 2.6 GHz Antenna. Solid lines represent the simulated S11 2.3.2 Simulated and Measured Radiation Patterns A theoretical overview of antenna radiation patterns is presented in Section 3.2. The coordinate system utilized for electromagnetic simulations of the antenna is illustrated in Figure 2.14. Figure 2.14: Coordinate System Used for Antenna Simulation 27

PAGE 41

28 The radiation pattern of a pa tch antenna operating in the TM10 mode is in the broadside direction, [36]. This direction would correspond to the = 0 axis for the quarter-wavelength shorted patch antennas. The shorted patch antennas were excited by a microstrip feedline, which was parallel to th e direction at which the resonance occurs. Thus, they are mostly linearly polarized in the direction parallel to the microstrip feedline, which corresponds to = 90 In a typical patch antenna, there may be a resonance that occurs at a higher frequency along the width of the antenna, which is due to higher order modes. However, at the operating frequencies of 1.3 GHz a nd 2.6 GHz, higher order modes are not present for these quarter-wavelength shorted patch an tennas. Cross-polarization occurs since more current flows along the width of the quarter-wavelength antenna. The increased current flow does not occur in the half-wave antenna since the width to length ratio is lower. Nevertheless, the dominant polari zation is along the length of the antenna. The antenna feedline is parallel to the Eplane and perpendicular to the H-plane. The electromagnetic simulato r utilized the coordinates and to express the location of the observation point. was kept fixed at 90 for E-plane simulated patterns and was swept 360 was kept fixed at 0 for H-plane simulated patterns and was swept 360 Simulations were initially performed by considering the ground plane of the antenna to be infinite and radiation patterns revealed no back-side radiation. Practically, compact antennas such as the quarter-wavelength shorted patch antennas studied are designed to sit on a finite gr ound plane. Therefore, antenna simulations were performed using a finite ground plane. The resulting radi ation patterns displayed considerable backside radiation. When the size of the gr ound plane was greater than the patch antenna

PAGE 42

dimensions by approximately six times the substrate thickness, h, all around the edges of the patch, the results were similar to having an infinite ground plane, [36]. The size of the ground plane from the edge of the patch fo r both antennas is 7 mm which is less than 6h. Therefore, back lobes are expected in radiation patterns of these antennas. Radiation pattern measurements we re performed by placing the quarterwavelength shorted patch antenna inside an anechoic chamber. The test antenna acted as the receive antenna. A 1.3 GHz custom dua l-patch array was used as the transmit antenna and remained stationa ry. The receive antenna was rotated along a specified axis through 360 A commercial Yagi tube antenna was used as the receive antenna for the 2.6 GHz quarter-wavelength shorted patch. Ro tational axes are illust rated in Figure 2.15. Figure 2.15: Axis Definition for Antenna Measurements In a fashion similar to the simulations, measurements were obtained by keeping the coordinate constant while sweeping the coordinate. The = 0 angle was defined as the 29

PAGE 43

orientation where the receive and transmit ante nna faces were aligned with each other. A rotation was performed about the axis pe rpendicular to the feedline for E-plane measurements. In contrast, H-plane measur ements were obtained by rotating about the axis parallel to the feedline. Radiation patterns were normalized to the maximum copolarization value obtained. Figure 2.16 and Figure 2.17 present comparisons, at the particular frequency of interest, between the expected (finite ground plane) and m easured results of the copolarized Eand H-planes for the 1.3 GHz a nd 2.6 GHz antennas, respectively. The plots display good agreement between the simulated and measured radiation patterns. It is worth noting that the quarte r-wavelength antenna only ra diates from one edge. Therefore, the E-plane pattern becomes broade r since the array eff ect of two radiating edges for a half-wavelength antenna is absent, [2]. 0 45 90 135 180 225 270 315 -40 -30 -30 -20 -20 -10 -10 EPlane Co-Pol (Sim) EPlane Co-Pol (Meas) 0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 HPlane Co-Pol (Sim) HPlane Co-Pol (Meas) Figure 2.16: Comparison of Measured and Simulated Co-Polarized E-Plane (left) and H-Plane (right), Pattern of the 1.3 GHz Antenna, (dB) 30

PAGE 44

0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 EPlane Co-Pol (Sim) EPlane Co-Pol (Meas) 0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 -50 -50 HPlane Co-Pol (Sim) HPlane Co-Pol (Meas) Figure 2.17: Comparison of Measured and Simulated Co-Polarized E-Plane (left) and H-Plane (right), Pattern of the 2.6 GHz Antenna, (dB) Measured Eand H-plane cross-polarizati on patterns are presented in Figure 2.18. Measurements were normalized to the maxi mum co-polarization value obtained for each antenna. 31

PAGE 45

0 45 90 135 180 225 270 315 -10 -8 -8 -6 -6 -4 -4 -2 -2 EPlane Cross-Pol (Meas) HPlane Cross-Pol (Meas) 0 45 90 135 180 225 270 315 -10 -8 -8 -6 -6 -4 -4 -2 -2 2 2 EPlane Cross-Pol (Meas) HPlane Cross-Pol (Meas) Figure 2.18: Cross-Polarized Measurement of the E and H-Plane Radiation Pattern, 1.3 GHz Antenna (left) and 2. 6 GHz Antenna (right), (dB)) Considerable cross-polarization can be observed. Some possible causes for the crosspolarization levels are the lowered width to length ratio of the ante nna and the utilization of a microstrip feed, which can increase these cross-polarization levels. Crosspolarization levels for shorted patch antennas have been reported to be higher than for half-wave patches, [2], [36]. Therefore, these compact microstrip antennas should be employed in applications where cross-polarization can be tolerated. 2.4 Power Measurement Techniques On the transmitter side, a network analyzer was used to provide a continuous wave signal at 1.3 GHz and 0 dBm. This signa l was further attenuated to different test levels using a variable attenuator. The signal was then amplified using an amplifier with a power output of 22 dBm. Two isolators, which operated at 1.3 GHz, were used to 32

PAGE 46

ensure proper signal transmission at the proper frequency and to minimize any harmonic interference. A 20 dB coupler was used as a test point to observe the amplifiers drive level with the aid of a spectrum analyzer. Using a 3 dB power splitter, the 1.3 GHz signal was sent into a custom transmitting dua l-patch array. The FDR was placed on a tripod in the anechoic chamber where it r eceived the 1.3 GHz signal, doubled the frequency and transmitted it back a signal to a commercial 2.6 GHz Yagi tube antenna. A fabricated FDR device is displayed, fully assembled, in Figure 2.19. Figure 2.19: Fabricated Fr equency Doubling Reflectenna The received power transmitted by the harmonic re-radiator was measured using a spectrum analyzer. Path spreading losses we re calculated using the free-space path loss formula d PL 4 log2010dB, (2.3) where d is the distance from the transmitter and is the signal wavelength. Both the distance and wavelength parameters are in meters. The hardware test-bench utilized co mponents and cables, which introduce loss into the system. Therefore, proper charac terization of each element was performed to account for the losses being introduced. To validate the characterization for each of the 33

PAGE 47

components of the system and the 1.3 GHz transmitting dual-patch gain, a preliminary test was conducted using a custom single 1. 3 GHz patch antenna as the receiver. The power levels received by the patch antenna correlated with ex pected values for the power received. Once the validation was complete d and an accurate measure conducted of the power received by the actual FDR, the singl e patch antenna was replaced by the FDR for testing. Figures 2.20 presents a diagram of the test measurement setup utilized for the FDR tests. Figure 2.20: Diagram of the Harmonic Re-radiator Test Measurement Setup. The input power level to the FDR was -12.85 dBm Figure 2.21 presents a picture of the comp lete arrangement utilized for performing measurements on the FDR. 34

PAGE 48

Figure 2.21: Hardware Test-Ben ch Used for FDR Measurements The FDR was placed in an anechoic chambe r where it received the transmitted 1.3 GHz signal. The FDR doubled the received signal frequency and re-radiated it at 2.6 GHz. Output power measurements are conducted using a spectrum analyzer, which was connected to the 2.6 GHz receiving antenna. 2.5 Results The output power received at 2.6 GHz antenna and the multiplier conversion loss curves for three FDRs are presented in Fi gure 2.22. Conversion gain values were calculated by using the relationship CG = Pout(dB) Pin(dB) + Receiving system loss(dB). (2.4) 35

PAGE 49

-40 -35 -30 -25 -20 -15 -10Power Input to Diode Doubler (dBm) -125 -105 -85 -65 -45 -25 -5P o w e r R e c e i v e d a t 2 6 G H z ( d B m ) -125 -105 -85 -65 -45 -25 -5C o n v e r s i o n G a i n ( d B ) Expected Meas FDR1 Meas FDR2 Meas FDR3 Figure 2.22: Comparison of Measur ed and Expected Output Power and Diode Doubler Conversion Gain at a Source Frequency of 1.3 GHz Good correlation is demonstrated between the measured curves for FDRs 1 and 2. The received 2.6 GHz power for FDR 3 differed by approximately 7 dB from the other two re-radiators. Sin ce the measured values differed from those obtained from simulation, the frequency sensitivity of th e FDR was analyzed. Measurements were performed using source signals at frequencies close to 1.3 GHz with an input power to the FDR of -31.53 dBm. The results of the an alysis are presented in Figure 2.23. The frequency at which the maximum doubler conversion efficiency occurred varied somewhat from 1.3 GHz for each sample. 36

PAGE 50

1.27 1.275 1.28 1.285 1.29 1.295 1.3 1.305Frequency (GHz) -110 -105 -100 -95 -90 -85 -80P o w e r R e c e i v e d a t 2 6 G H z ( d B m ) -80.71 -81.57 -86.84 FDR1 FDR2 FDR3 Figure 2.23: Measured Received Power for Different Source Frequencies. Diode Doubler Input Power was -31.53 dBm Swept-power performance was measured at the source frequency that yielded the highest level of received power. This data was compared with the simulated data at a source frequency of 1.3 GHz. The comparison is presented in Figure 2.24. -40 -35 -30 -25 -20 -15 -10Power Input to Diode Doubler (dBm) -110 -90 -70 -50 -30 -10P o w e r R e c e i v e d a t O p t F r e q ( d B m ) -110 -90 -70 -50 -30 -10C o n v e r s i o n G a i n ( d B ) Expected Meas FDR1 Meas FDR2 Meas FDR3 Figure 2.24: Comparison of Expect ed Doubler Conversion Gain and Output Power at a Source Fre quency of 1.3 GHz and Measured Performance at the Fabricated FDRs Optimum Source Frequency 37

PAGE 51

38FDRs 1 and 2, demonstrated good agreement, a deviation of 0.7%, at a source frequency of 1.291 GHz. Peak power levels for FDR 3 occurred at 1.28 GHz, which was a deviation of 1.6%. The 1.3 GHz dual-patch ar ray, see Figure 2.19, which was used as the transmit antenna possessed a narrowband response and provided less gain at 1.28 GHz than at 1.3 GHz. Over an input power range of -40 to -10 dBm, the FDR measured and simulated conversion gain of the doubler compared very well for FDRs 1 and 2. The conversion gain was -19.87 dB when the inpu t power to the FDR was -31.53 dBm, and a maximum value of -13.7 dB for an input power of -15.53 dBm. These values are an improvement over what has been reported on a Schottky diode doubler/antenna transponder design, [17]. By considering the simulated value of gain of the input and output antennas, the measured conversion gain of the entire FDRs 1 and 2 at the optimum frequency of operation was -22.75 dB for an input power level at the FDR receiving antenna of -28.5 dBm. This FDR was designed for low-power input and zerobias operation. In other frequency multiplier s, where the input signal strength can be as high as 5 dBm, conversion loss values of close to10 dB are expected, [8]. 2.6 Conclusions A compact conjugate-matched freque ncy doubling reflectenna was designed, fabricated and tested. At an input frequency of 1.3 GHz a multiplier conversion efficiency of 1% was obtained at an i nput power of -30 dBm. A high degree of performance accuracy was obtained by the nonlin ear simulations performed in the design process. The harmonic re-radiators were pr oduced using careful, although not extremely precise, manufacturing methods with a small spread in performance. The conjugate-

PAGE 52

39matched design maximized conversion efficien cy and increased the sensitivity of the design. Since the FDRs performance is relatively narrowband its operation could be controlled by using integrated control structures such as tuning capacitors. This FDR is a promising device for sensing applications.

PAGE 53

40 CHAPTER 3 COMPACT PLANAR ANTENNAS: MEANDERED ANTENNAS 3.1 Introduction The reduction of the size of wireless sens or devices is often desired due to the various applications in which they may be employed. In many wireless devices the antenna occupies the vast majority of the overa ll device area. This is the case with the FDR, where the receiving and transmitting antennas dominate the footprint of the design. Despite the work that has been cond ucted in miniaturized antennas, ongoing investigations are still prev alent since a reduction in the size of the antenna typically degrades its performance. In an integrat ed device with antennas, degraded antenna performance impacts the overall efficiency of the system. Therefore, studies, which aim to miniaturize the antenna wit hout sacrificing parameters such as gain and efficiency, are important. Small antenna research often targ ets planar antennas due to their desirable light weight, low volume and conforma l characteristics, [2], [36], [37]. In this chapter a brief re view of important antenna pa rameters and a discussion of the fundamental limitations of electrically small antennas is presented. Some of the techniques, utilized by other au thors, to miniaturize planar antennas are also mentioned. In order to produce a miniature harmonic re-ra diator, two planar antenna designs were considered. A meandered slot antenna de sign and a meandered shorted patch antenna

PAGE 54

41design were investigated. A meandered sl ot antenna was designed for an operating frequency of 1.3 GHz. The meandered shorted patch antennas were designed for operating frequencies of 1.3 GHz and 2.6 GHz. The meandered shorted patch antenna provided an antenna size reduction of ~85% compared to the quarter-wavelength shorted patch antennas used in the first-generati on FDR design. The gain of the miniature meandered patch antennas was optimized by increasing the substrate thickness without exceeding a value, which would allow th e propagation of higher order modes. 3.2 Review of Antenna Parameters Antennas can either convert an RF signal from a transmitter to a propagating electromagnetic wave or conve rt a propagating wave into an RF signal in a receiver. Therefore, antennas are reciprocal devices a nd the properties define d below apply either to an antenna utilized as a tr ansmitting or a receiving device. An antennas radiation pattern is a plot of the transmitted or received signal strength versus position around the antenna. The operating frequency, the size and the gain are antenna parameters that are related to each other. An antenna needs to have minimum physical dimensions in order for eff ective radiation to occur. As the frequency of operation of an antenn a increases its size decreases. In addition, since the gain of an antenna is proportional to its cross-sectional area divided by the wave length squared, an electrically small antenna will usually have lower gain than a larger antenna, [38]. Wireless communications are possible due to the propagation of electromagnetic energy. An antenna converts a guided elec tromagnetic wave on a transmission line to a

PAGE 55

plane wave propagating in free-sp ace. The spherical coordina te system utilized in the analysis of antenna parameters is illustrated in Figure 3.1. Figure 3.1: Spherica l Coordinate System The antenna radiates a spherical wave. However, at large distances this wave becomes an approximation of a plane wave. The distance where this occurs is referred to as the far-field distance. The far-field distance is defined as 22 D Rffm, (3.1) where D is the maximum dimension of the antenna and is the wavelength. In the far-field zone of the antenna, the radiated electric field is given by r e F F rErjko )],(),([),,(^ ^ V/m (3.2) where is the electric field vector, and are unit vectors in th e spherical coordinate system, r is the radial distance from the origin, E^^ 42

PAGE 56

ko = 2 / (3.2a) wavelength, = c/f = (3.2b) f /1038 with frequency in Hz, and),( F and),( F are the pattern functions. The electric field propagates in the radial di rection. The electric field ma y be polarized in either the or directions but not in the radi al direction. Polarization of this type is characteristic of transverse electromagnetic, (TEM), waves. The magnetic fields associated with this TEM wave are given by 0 E H (3.3) and 0 E H (3.4) where 3770 (3.4a) The radiation pattern of an antenna denot es the magnitude of the far-zone field strength versus position around the antenna at a fixed distance. The field can be plotted and),( F.from the pattern functions ),( F The pattern functions depend on the 43

PAGE 57

44 If the p is plott polarization of the antenna. atterned versus the angle it depicts an elevation plane pattern and if plotted versus the angle it depicts an azimuthal plane pattern. T he radiation intensity provi des the variation in radiat ed power versus position around tion tion efficiency was previ ously defined, in Chapter 2, in terms of e radi r. ntenna. eref = the antenna. Directivity provides a meas ure of the directionality of an antenna pattern or the focusing ability of an antenna. Directivity is defined as the ratio of the maximum radiation intensity in the lobe with the maximum value to the average radia intensity over all space. Ther efore, directivity is a dimensionless ratio of power, which is usually expressed in dB. The antenna radia thation resistance and the loss resistance. This definition does not take into account surface wave loss, which will be discussed in Section 3.6. The radiation resistance Rr is associated with the radiation of an antenna. Loss occurs in an antenna mainly due to dissipative losses in the metals and the dielectric materials used to fabricate the radiato A reflection, (impedance), efficiency eref can also be described as a result of the reflections that arise due to the mismatch be tween the transmission lin e and the a Reflection efficiency is given by ) 1(2, (3.5) here = (Zin Z0)/(Zin + Z0). (3.5a) w

PAGE 58

45he antenna input impedance is Zin and the characteristic impedance of the line is Z0. T The overall efficiency is given by cdrefoeee (3.6) ain, related to an antenna, accounts for its losses and is defined as the product of 9] G directivity and efficiency. Antenna gain is usually expressed in dB and given by, [3 DeGcd (3.7) The bandwidth of an antenna is the range of frequencies on eith er side of a center frequen ed as the orientation of the radiated (3.8) hen E0x = 1 and E0y = 0 the field is linearly polarized. Whe 1 and and E0y = 1 cy where the antenna characteristics are within an acceptable value of those at the center frequency. Input impedance is said to be the impedance presented by an antenna at its terminals, namely the input terminals of the antenna. The polarization of an electromagnetic wave is defin electric field vector For the case of a plane wave propagating along the z axis, the electric field may expressed as zjk yxeyExEE0) (^ 0 ^ 0 W n E0x =

PAGE 59

the field is linearly polarized in the 45 direction in the or azimuthal plane. Thus, the wave will be linearly polarized if both E0x and E0y have the same phase. In contrast, if E0x and E0y have a 90 shift, the field is said to be circ ularly polarized. If the magnitudes of E0x and E0y are not equal or the phase difference is not exactly 90 the field is said to be elliptically polarized. The polarization of th e antenna will dictate the polarization of the radiated field, [38]. ^ 3.3 Limitations of Electrically Small Antennas Since 1947, several studies have been reported on the fundamental limitations of electrically small antennas, [40 43]. Wheeler commented on the reduction of the radiation resistance, efficiency and bandwidth of a small an tenna, [40]. In order to quantify the radiation of an antenna, Wheeler defined a radi ation power factor that was equal to the antenna resistance divided by th e antenna reactance. The radiation power factor for small antennas is less than one. With the purpose of further characterizing small antennas, Wheeler later defined the te rm radiansphere as a hypothetical sphere having a radius of /2 from the center of an antenna, [44]. This sphere can be viewed as the space occupied by the stored energy of a small antennas electric or magnetic field. As a result, a small antenna is one that occu pies a small fraction of one radiansphere in space, [45]. An idealized small spherical an tenna has a radiation pow er factor equal to the ratio of its volume to that of the radiansphe re. Therefore, if an antenna is limited by a maximum dimension but not by an occupied vo lume, the radiation power factor can be increased by utilizing the volume of a sphere with a diameter equal to this maximum dimension, [44]. 46

PAGE 60

Chu described the performance of an ante nna in terms of its power gain and its bandwidth. He defined the largest linear dimension of the antenna to be 2r, such that the complete antenna structure is enclosed inside a spherical surface of radius r. The quality factor Q is directly proportional to the mean electric energy stored beyond the input terminals of the antenna times 2 divided by the power dissi pated in radiation. Moreover, the bandwidth of the ante nna is indirectly proportional to Q. In addition, Chu found that if the Q of an antenna is required to be low, there is a definite limit to the gain of a practical antenna that is approximately equal to 4r/ [41]. The concept of an antenna enclosed in a sphere of radius r was also utilized by Hansen and is illustrated in Figure 3.2. Hansen defined k = 2 / and explained that higher modes may not be present for kr <1. For kr << 1, Q varies inversely as the cube of the radius of the sphere. The radiation Q increases rapidly as the size of the antenna decreases, [42]. Figure 3.2: Antenna Within a Sphere of Radius r, [42] Mclean derived an exact expression for the radiation Q associated with the TM01 mode, which is given by 47

PAGE 61

kr rk Q 1133. (3.9) If kr << 1 equation (3.9) reduces to 331 rk Q (3.10) A circularly polarized antenna given by a combination of TE01 and TM01 modes, will yield the lowest achievable radiation Q and will be approximately half that of a linearly polarized antenna. In addition, a narrower maximum achievable bandwidth is obtained with a larger radiation Q, [43]. Recently, additional studies have been conducted on the quality factor of electrically small antennas, [46 -48]. Best found that the quality factor of an electrically small wire antenna is mainly determined by the antennas height and effective volume in relation to the resonant wave length, [48]. Therefore, m odifying the geometry of an antenna within a fixed height and fixed cy lindrical diameter does not significantly improve the performance of a small self-resona nt antenna. A small antenna, which Best defined as one with a boundary limit at kr = 0.5, usually has a radiation resistance that is very low and an input reactance that is very high, [47]. E ssentially, the best compromise between bandwidth and efficiency will usuall y be achieved when most of the allotted volume in an antenna design is utilized in radiation, [49]. 48

PAGE 62

493.4 Size Reduction Techniques for Planar Antennas Miniaturizing techniques u tilized to reduce the size of different types of antennas consist primarily of loading the antenna with lumped elements, of u tilizing high-dielectric materials or conductors and use of the antenna casing to increase radiation. A small antenna will usually have a high input reactance, which can be compensated by lumpedelement loading. This method will either redu ce the efficiency of the antenna or increase its quality factor. In addition, an antenna may be miniatur ized by altering the dielectric or magnetic characteristics of the encasing material. However, modification of the encasing material can reduce the bandwidth of the antenna and produ ce higher dielectric losses. An antennas overall physical size ca n also be reduced by modifying its geometry and shape, [49]. The desire for smaller mobile communi cation devices has en abled some of the advances in the area of compact planar antennas. Effective methods utilized to design miniature antennas, which do not exhibit degrad ed gain or cross-polarization, have been reported. Several authors have presented co mprehensive reviews of compact microstrip antennas, [2], [36], [37]. Planar antennas such as microstrip antennas, printed monopole and dipole antennas and slot antennas have been studied extensively. These types of antennas, except for the printed monopole, ar e illustrated in Figure 3.3.

PAGE 63

Figure 3.3: Configuration of Different Planar Antennas: Microstrip Patch (left); Printed Dipole (middle); Slot Antenna (right), [2] A microstrip antenna may simply consist of a radiating patch of metallization on a grounded dielectric substrate. Microstrip antennas are char acteristically light weight, low-profile and low-cost. However, they may be limited by a narrow-bandwidth, lower gain and poor efficiency. Printed dipoles can be considered as narrow rectangular patches. Printed slot antennas are those, whic h consist of a slot in the ground plane of a grounded substrate. Printed slot antennas can be fed by a microstrip line or a coplanar waveguide, [2]. As with the shorted patch antennas described in Chapter 2, other edge-shorted rectangular patch antennas ha ve also been reported. Th e quarter-wavelength antennas described in the previous chap ter utilized shorting vias to reduce the antenna size. However, structures employing shorting walls or shorting plates instead of via holes have been developed. When a shorting pin is load ed at the tip of an equilateral-triangular microstrip antenna, its size can be reduced by as much as 94%, [50]. Size reduction is also ac hieved by introducing several slits in the non-radiating edges of a rectangular patch antenna. The sl its create an effect wh ere the patchs surface 50

PAGE 64

51currents are effectively meandered. Therefor e, the current path is lengthened, which reduces the antennas resonant frequency, [ 51]. Similarly, a bow -tie patch has been designed by cutting triangular no tches at both non-radiating si des of the patch. Surface current path lengthening can al so be achieved by utilizing th ree dimensional structures such as the U-shaped patch, the folded patc h or the double-folded patch, which is termed a planar inverted-f. The patchs surface cu rrent is bent along the antennas resonant dimension and no lateral currents are generated, thereby reducing cross-polarization levels. Rectangular patches that contain embe dded slots have been implemented to produce compact antennas. Slot geometries can range from cross slot s, rectangular slots and circular slots. These sl ots cause meandering of the pa tch surface currents, which is an effect that generates compact antenna designs. Compact microstrip antennas exhibiting dual-frequency operati on have also been realized. These designs achieve their effect by embedding a pair of slots parallel and close to the radiating edges of a meandered rectangular antenna or a bow-tie pa tch. A miniature dual-band folded patch antenna has been proposed, [52]. As mentioned in the previous section, the gain and effici ency characteristics of an antenna might degrade as the physical size of the structure decreases. Therefore, several techniques have been implemented in order to increase the gain of compact microstrip antennas. These include utilizing a high-permit tivity dielectric superstrate or integrating active circuitry into th e antenna structure. Both techniques have been found to increase the gain of the antenna, [53 56].

PAGE 65

3.5 Meandered Slot Antenna In order to reduce the overall footprin t of the FDR design, a meandered slot antenna was considered as an alternative de sign to the quarter-wavelength shorted patch antenna. The slot antenna geometry provi des several advantages over the quarterwavelength shorted patch antennas. Employi ng a coplanar waveguide, (CPW), fed slot antenna alleviates the shortcomings on the ra diation efficiency created by the shorted patch antennas via resistance. The geometry of a basic CPW-fed slot antenna employing a center feed is illustrated in Figure 3.4. Figure 3.4: CPW-Fed Sl ot Antenna, (Design A) Table 3.1 presents the parameters for antenna Design A, which is illustrated in Figure 3.4. Table 3.1: Parameters for a Half-Wavelength CPW-Fed Slot Antenna Design (Dimensions in Millimeters) Design L / W L1 / S / W1 A 64 / 2 1.5 / 1 / 0.5 52

PAGE 66

53Slot antennas can be less sensitive to to lerances due to manufacturing and can be polarized in a certain way us ing a combination of strip and slot conductors, [2]. Meandering or folding a slot antenna can further reduce its size. Use of a CPW line to feed the slot antenna is useful in the desi gn since mutual coupling between adjacent lines is minimized, [2]. Therefore, meandered slot antennas can be placed in close proximity, which reduces the physical size of the overall circuit. In addition, a center-fed slot antenna has a very high radiation resistance. Agilents Momentum was used to analyze antenna characteristics such as return loss, input impedance and gain. A traditi onal CPW-fed slot antenna such as the one presented in Figure 3.4 was designed to be a pproximately half-wavelength at the desired operating frequency of 1.3 GHz. The board su bstrate employed for the design was 60 mil Taconic, ( r = 6.15). Analysis of the properties of this traditional geometry, which was designated Design A, was conducted with the pu rpose of providing a ba seline for analysis of subsequent meandered slot geometries. Th e meandered slot geometry investigated in this research and designated Design B is illustrated in Figure 3.5.

PAGE 67

Figure 3.5: CPW-Fed Meandere d Slot Antenna (Design B) Table 3.2 presents the parameters for antenna Design B, which is illustrated in Figure 3.5. Table 3.2: Parameters for Meandered Sl ot Antenna Design (Refer to Figure 3.5) (Dimensions in Millimeters) Design L / W L1 / L2 / W1 / W2 / W3 / W4 / W5 / W6 B 22 / 8.5 1.5 / 4 / 0.5 / 3.5 / 3 / 1 / 1 / 3 When compared to the basic slot antenna design, the size reduction obtained in the total length of the antenna with the meandered sl ot antennas geometry was 65.6% for Design B. Additional meandered slot geometries investigated in this research and designated Design C, Design D and Design E are illustrated in Figure 3.6. 54

PAGE 68

Figure 3.6: CPW-Fed M eandered Slot Antenna (Designs C, D, E) Table 3.3 presents the parameters for antenna Designs C, D and E, which are illustrated in Figure 3.6. Table 3.3: Parameters for the Meandered Sl ot Antenna Design (R efer to Figure 3.6) (Dimensions in Millimeters) Design L / W L1 / W1 / W2 / W3 / W4 / W5 / W6 / W7 / W8 / W9 C 12 / 12 2 / 3 / 1 / 0.5 / 1 / 0.5 / 0.5 / 1 / 0.5 / 0.5 D 11.5 / 12 2 / 2.5 / 1 / 0.5 / 1 / 0.5 / 0.5 / 0.5 / 0.5 / 0.5 E 9 / 10.25 2 / 1 / 0.5 / 0.5 / 0.5 / 0.5 / 0.5 / 1 / 0.25 / 0.25 55

PAGE 69

When compared to the basic slot antenna design, the size reduction obtained in the total length of the antenna with the meandered slot antenna geometry was 81.2% for Design C. The physical dimensions of the meandered slot antenna were optimized utilizing the electromagnetic simulator. The dimensions obtained for the meandered slot antenna designs C E represent a re duction of ~ 50% when compared to the 1.3 GHz shorted patch antenna described in Chapter 2. Simulation results of the return loss; bot h magnitude and phase, of the five slot antenna designs considered were obtained. Figure 3.7 presents a comparison plot for the magnitudes and Figure 3.8 presents a comparison plot for the phases. 0 0.5 1 1.5 2 2.5 3Frequency (GHz) -2 -1.8 -1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0S 1 1 ( d B ) A B C D E Figure 3.7: Simulated S11, (Magnitude), of CPW-Fed Slot Antenna Designs A E 56

PAGE 70

0 0.5 1 1.5 2 2.5 3Frequency (GHz) -200 -150 -100 -50 0 50 100 150 200S 1 1 P h a s e ( De g r e e s ) A B C D E Figure 3.8: Simulated S 11, (Phase), of CPW-Fed Slot Antenna Designs, A E It is important to observe that when calcula ting the values for the magnitude of S11, the simulator considered the design to be matched to 50 The input impedance of the antenna designs was not 50 Therefore, the values displayed in Figure 3.7 are low for most designs. The input impedances for the five designs were simulated and are presented in Figures 3.9 and 3.10. A compar ison plot of the real part of the input impedance for the designs investigated is pres ented in Figure 3.9. The imaginary part of the input impedances is presented in the comparison plot of Figure 3.10. 57

PAGE 71

0 0.5 1 1.5 2 2.5 3Frequency (GHz) 0 5 10 15 20 25 30 35 40 45 50R e a l Z i n ( ) A B C D E Figure 3.9: Simulated Real Input Impedance for CPW-Fed Slot Antenna Designs, A E 0 0.5 1 1.5 2 2.5 3Frequency (GHz) -950 -750 -550 -350 -150 50 250 450 650 850I m a g Z i n ( ) A B C D E Figure 3.10: Simulated Imaginary Input Impedance for CPW-Fed Slot Antenna Designs, A E 58

PAGE 72

The simulated Eplane patterns of the five slot antenna designs are presented in Figure 3.11. It can be observed that the mea ndered slot designs pres ent radiation patterns similar to the ones of the traditional CPW-fed slot antenna. -180 -130 -80 -30 20 70 120 170Theta (deg) -50 -40 -30 -20 -10 0Ga i n ( d B ) A B C D E Figure 3.11: 1.3 GHz Simulate d EPlane Patterns for CPW-Fed Slot Antenna Designs, A E The values for the resonant frequencies of the different antenna geometries studied are presented in Table 3.4. 59

PAGE 73

60Table 3.4: Parameters for Slot Antenna Designs (Refer to Figures 3.4 -3.6) Design L / W (mm) fo (GHz) A 64 / 2 1.4 B 22 / 8.5 1.446 C 12 / 12 1.439 D 11.5 / 12 1.417 E 9 / 10.25 1.336 In order to produce an antenna design that would resonate at a frequency close to that of Design A, the total length of the m eandered slot for Designs B E was slightly larger than the total length of the slot for Design A. However, the increase in the total length of the meandered slot di d not significantly increase the overall size of the antenna geometry, (refer to Tables 3.1 3.3). A comparison of the simulated input impedance, gain and efficiency for the different slot antenna geometries inve stigated is presented in Table 3.5.

PAGE 74

61Table 3.5: Parameters for Slot Antenna Designs (Refer to Figures 3.4 -3.6) Design L / W (mm2) Re {Zin} / Im {Zin} ( ) Gain (dB) Efficiency (%) A 64 / 2 62.0 / 303.9 1.7 94.3 B 22 / 8.5 6.0 / 218.2 0.4 72.2 C 12 / 12 3.6 / 261.3 -3.5 30.0 D 11.5 / 12 5.2 / 310.5 -3.7 28.8 E 9 / 10.25 49.9 / 908.9 -9.0 8.5 The meandered slot antenna designs exhi bited low real input impedance values when compared to a basic CPW-fed slot antenna. Meandering of the slot did not result in a significant decrease of the antenna effi ciency for Design B. A decrease in the efficiency for Designs B E was observed. For the meandered slot geometry illustrated in Figure 3.6, decreasing parameters such as the width of the sl ot and the separation between meandered sections resulted in a decrease in the gain of the antenna. Integrating a CPW-fed antenna into th e FDR design would have resulted in having to re-design the layout for the ha rmonic re-radiator. CPW-to-microstrip transitions would have had to be studied and implemented. Additionally, implementing slot antennas into the design w ould have lead to a change in the radiation pattern from unidirectional to bidirectional. For ease in the design methodology, a shorted meandered patch antenna design was eventually chosen for a second-generation FDR design. This compact microstrip antenna design is described in the next section.

PAGE 75

3.6 Meandered Shorted Patch Antenna 3.6.1 Antenna Design In general, for a rectangular pa tch operating in the fundamental TM10 mode and designed for a thin dielectric, the length L should be rf c L2 (3.11) where c is the speed of light, f is the resonant frequency and r is the relative permittivity of the substrate, [37]. If the patch is operating in the fundamental TM0 mode, then the field varies every /2 along the length a nd no variation occurs along the width of the antenna. Lowering of the fundamental re sonant frequency has been achieved by meandering the excited patch surface current paths in the radiati ng patch, [37]. The meandering leads to an extended current path for a fixed length. Thus, a meandered shorted patch antenna was devised from the design of the quarter-wavelength shorted patch antennas used in the first-generation FDR design. The total length of the meandered shorted patch antenna was calculated using equation (3.11). In the same manner as with the quarter-wavelength antenna described in Chapter 2, shorting vias were used to decrease the size of the patch from a half-wave to a quarter-wavelength. The to tal length of the antenna corresponding to a quarterwavelength was meandered in different sections. A study was conducted on the correlation between the number of meandere d sections for a meandered patch antenna and its resonant frequency, [54]. Utiliz ing the findings of the Lancaster study and considering the difference betw een the antenna substrate utilized, an initial number of 62

PAGE 76

63five meandered sections was used. The num ber of meandered sections was optimized using the electromagnetic simulator to produ ce an antenna with the desired response. The design method used for these meande red shorted patch an tennas was similar to the one used for the quarter-wavelength s horted patch antennas of the first-generation FDR design. The meandered shorted patch antenn as were initially designed so that their lowest resonance occurred approximately when the total length of the meandered line was a quarter-wavelength. Th e real impedance of the firs t-generation quarter-wavelength shorted patch antennas was optimi zed to be on the order of 40 As discussed in the previous sections, as the elec trical length of an antenna de creases, so does its radiation resistance and efficiency. This effect was observed in the performance results obtained from initial antenna simulations. The input resistance of the meandered shorted patch antenna was much lower than the intended value of 40 As the real input impedance of the antenna increased, the antenna efficien cy decreased. Using equation (2.2) from Chapter 2, the reduction in antenna efficiency indicated that a higher value of real input impedance was due to an increment in loss resistance RL, which was undesirable. Since a compact microstrip antenna will yield low efficiency values, special attention was given to obtaining the maxi mum efficiency possible for the meandered shorted patch antennas. It has been reported that gain and efficiency for a shorted patch antenna is less than those for a half wave patch antenna, [2]. This reduction in efficiency is intensified due to the small dimensions of the meandered shorted patch antenna as discussed in Section 3.3. As reported by Best, an electrically small antenna will have a low radiation resistance and a high input reactance, [47]. Thus, efforts were made to optimize the

PAGE 77

64design in order to minimize the input reactance and maximize the input resistance without degrading the antenna efficiency. Several physical antenna pa rameters were optimized to obtain the desired antenna parameters with the aid of the ADS Momentum electromagnetic simulator. The optimized parameters included: the total length of the meander line of the antenna, the number and location of shorting vias, the number of meandered sections, the slot size between sections, the width of the meander line, the total length and width of the structure, the geometry, dimensions and location of the feedline. In an attempt to reduce the overall loss resistance, the antenna was shorted with a set of two parallel vias. It has been noted that, a single short at th e corner of the antenna creates the most compact design for a rectangular patch antenna. The vias on the meandered shorted patch antennas were loca ted in the corner of the non-ra diating edge of the antenna opposite to the feedline for the 1.3 GHz an tenna, [36]. Figure 3.12 illustrates the geometry for the 1.3 GHz meandered shorted patch antenna.

PAGE 78

Figure 3.12: Geometry of the Meandered Shorted Patch Antenna, (1.3 GHz Design) Table 3.6 presents the final parameters for the design of each of the antennas. Table 3.6: Parameters for 1.3 GHz and 2.6 GHz Meandered Shorted Patch Antenna Designs of Figure 3.13 (Dimensions in Millimeters) Design L / W L1 / L2 / L3 / L4 / L5 / L6 / L7 / W1 / W2 / W3 / W4 1.3 GHz Antenna 5.05 / 10.54 1.05 / 0.26 / 0.14 / 0.40 / 0.80 / 0.27 / 2.0 / 1.69 / 2.3 / 0.50 / 0.50 2.6 GHz Antenna 11.5 / 12 1.05 / 0.26 / 0.15 / 0.30 / 0.25 / 0.25 / 1.5 / 1.69 / 1.4 / 0.50 / 0.50 Figure 3.13 provides a layout for both the 1.3 GHz and the 2.6 GHz designs, which indicates the locations of the para meters provided in Table 3.6. 65

PAGE 79

W L L5L4W1 L5 L6 W3 L7 L8 L1 L2W2 L3 W4 L5 Figure 3.13: 1.3 GHz, (top), and 2.6 GHz, (bottom) Meandered Shorted Patch Antennas (at 2.6 GHz; via at x = 1.3 mm and y = 0.5 mm) 66

PAGE 80

Figure 3.14 presents a comp arison plot between the measured and simulated S11 data, referenced to 50 for the 1.3 GHz antenna. 1 1.5 2 2.5 3Frequency (GHz) -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0S 1 1 M a g n i t u d e ( d B ) -200 -150 -100 -50 0 50 100 150 200S 1 1 p h a s e ( d e g r e e s ) Mag (Meas) Mag (Sim) Phase (Meas) Phase (Sim) Figure 3.14: Comparison of Measured a nd Simulated S11 of the 1.3 GHz Antennas Figure 3.15 presents a compar ison plot between the measured and simulated S11 data, referenced to 50 for the 2.6 GHz antenna. 1 1.5 2 2.5 3 Frequency (GHz) -12 -10 -8 -6 -4 -2 0S 1 1 M a g n i t u d e ( d B ) -200 -150 -100 -50 0 50 100 150 200S 1 1 P h a s e ( d e g r e e s ) Mag (Meas) Mag (Sim) Phase (Meas) Phase (Sim) Figure 3.15: Comparison of Measured a nd Simulated S11 of the 2.6 GHz Antennas 67

PAGE 81

Section 2.3.1 describes the procedure utilized to obtain return loss measurements and the process used to obtain a model to properly perform simulated and measured response comparisons. The expected S11 data was obtained by assuming an infinite ground plane since the excitation ports avai lable in the simulator yielde d more accurate return loss values compared to using excitation por ts defined for a finite ground plane. Plots of the simulated input impedance as a function of frequency are presented for both the 1.3 GHz and 2.6 GHz, meandere d shorted patch ante nnas in Figure 3.16. 1 1.5 2 2.5 3Frequency (GHz) 0 50 100 150 200 250 300 350R e { Z i n } ( ) -350 -250 -150 -50 50 150 250 350I m a g { Z i n } ( ) Re {Zin} Imag {Zin} 1 1.5 2 2.5 3Frequency (GHz) 0 50 100 150 200 250 300 350R e { Z i n } ( ) -800 -600 -400 -200 0 200 400 600 800I m a g { Z i n } ( ) Re {Zin} Imag {Zin} Figure 3.16: Plot of Simula ted Input Impedance vs. Fre quency for Meandered Shorted Patch Antennas: 1.3 GHz (left) and 2.6 GHz (right) It is important to observe that in order to obtain the desired input impedance for the 2.6 GHz antenna a via was placed near the feed of the antenna. The resulting 1.3 GHz antenna had an input impedance of (12.7 + j335) and the 2.6 GHz antenna had an input impedance of (19.3 + j25.15) 3.6.2 Substrate Thickness and Antenna Efficiency Considerations Antenna electromagnetic simulations revealed that increasing the substrate thickness from 1.524 mm to 7.874 mm increased th e simulated efficiency and gain of the meandered shorted patch antenna. Therefore, an analysis on the effects of this significant 68

PAGE 82

increase in substrate thickness on the perf ormance parameters of the antenna was performed. Since microstrip antennas can be considered lossy cavities, they can be modeled using a cavity model. The principa l assumption for this model is that the substrate thickness h << 0. The model was applied to several patch sh apes including rectangul ar patches. The interior region of the patch is considered a cavity bounded by electric walls on the top and the bottom with a magnetic wall along the edge. The meandered patch antenna can be considered a rectangular patch with slits. Therefore, the cavity model relationships can also be applied to the meandered patch antenna, [2]. The following analysis is derived from utilizing cavity model relationships. Surface waves are excited on microstrip antennas if the substrate has an r > 1. The surface wave is launched into the substrate where it can be reflected back and diffracted by the edges. Surface waves are TM and TE modes of the substrate. The lowest TM mode, the TM0 mode, has no cutoff frequency. The value of the substrate thickness determines that only the TM0 surface wave propagates. In order to allow the TM0 surface wave mode to propagate and a void the excitation of other modes, the relationship 14 10 rh (3.12) 69

PAGE 83

must be satisfied. In equation (3.12), h is the thickness of the substrate, o is the waveguide wavelength in free-space and r is the relative permittivity of the substrate. For the meandered shorted patch antenna designs h = 7.874 mm and equation (3.12) was satisfied for both the 1.3 GHz and 2.6 GHz antennas. Thus, only the TM0 mode surface wave propagated through the substrate. A thicker substrate increases the radiat ed power and reduces conductor loss. However, it might also increase dielectric loss and surface wave loss, [2]. The power lost in the lossy dielectric of the substrate is denoted by, Pd. The power lost due to the finite metallization conductivity is denoted by Pc. The power radiated in the form of a space wave is denoted by, Pr. The dielectric loss is given by t dW P tan (3.13) where tan is the loss tangent of the dielectric, and WT is the energy stored at resonance. The conductor loss is given by 0fh W PT c (3.14) where is the conductivity of the conductor. Equation (3.14) indicates that Pc decreases with increasing substrate thickness h, [2]. The power radiated from the patch, Pr, is determined by integrating the radiation fiel d over the hemisphere above the patch using 70

PAGE 84

2 0 2 0 2 2 2 0sin) ( 2 1 ddrEE Pr. (3.15) In equation (3.15), E and E are complicated functions of and substrate parameters. The total quality factor of the antenna can be approximated by r Tr TP W Q. (3.16) In addition, r T hW (3.17) and 2hP r (3.18) Therefore, hQ r T / (3.19) From equations (3.16) and (3.19) an d given that the total power loss, PT, is inversely proportional to QT, it follows that the total power lo ss may be reduced by increasing the substrate thickness h. An analysis of the impact of substr ate thickness on the antenna radiation efficiency follows. It is impor tant to point out that authors often define antenna radiation efficiency as the ratio of pow er radiated to the sum of radiated power and surface wave 71

PAGE 85

power, [2], [57]. This definition negl ects power dissipated in the conductor and dielectric. The antenna radiation efficiency can be expressed as surdcr r i r rPPPP P P P e (3.20) where Psur is the power loss associated with the surface waves. If Pc + Pd 0, then surr r rPP P e (3.21) Expected E-plane patterns fo r three substrate thicknesses of the Taconic substrate, ( r = 6.15), for the 1.3 GHz antenna are presented in Figures 3.17 -100 -80 -60 -40 -20 0 20 40 60 80 100Theta ( ) -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0Ga i n ( d B ) 60 mil substrate 250 mil substrate 310 mil substrate Figure 3.17: Plot of Simulated E-Pl ane Radiation Pattern for 3 Different Substrate Thicknesses for 1.3 GHz Meandered Shorted Patch Antenna 72

PAGE 86

Expected E-Plane patterns for th ree substrate thicknesses of the T aconic substrate, ( r = 6.15), for the 2.6 GHz antenna are presented in Figures 3.18 -100 -80 -60 -40 -20 0 20 40 60 80 100 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 60 mil substrate 250 mil substrate 310 mil substrate Figure 3.18: Plot of Simulated E-Plane Radiation Pattern Gain vs. Frequency for 3 Different Substrate Thicknesses for 1.3 GHz Meandered Shorted Patch Antenna The data presented in Figures 3.17 and 3. 18 indicate that the gain of the antennas increased as the substrate thickness increased. The thickness of the board was not further augmented to avoid bulkiness of the FDR. Table 3.7 presents a comparison between the performance parameters of the quarter-wavelength shorted patch antennas em ployed in the first-generation FDR design and the meandered shorted patch antennas. 73

PAGE 87

74Table 3.7: Comparison of Simulate d Parameters of Quarter-Wavelength Shorted Patch Antennas and Mea ndered Shorted Patch Antennas Design L / W (mm) Gain (dB) Efficiency (%) 1.3 GHz /4 Shorted Patch 22.5 / 17 -2.3 28.4 1.3 GHz Meandered Shorted Patch 9.54 / 5.05 -7.7 12.6 2.6 GHz /4 Shorted Patch 12.55 / 11 2.2 71.65 2.6 GHz Meandered Shorted Patch 7.25 / 3.25 -4.9 24.7 Table 3.8 illustrates the detrimental effect of reducing the size of the meandered shorted patch antennas on the para meters of gain and efficiency. Table 3.8: Size Reduction between Quarter-Wavelength Shorted Patch Antennas and Meandered Shorted Patch Antennas Frequency Size Reduction (%) Gain Re duction (%) Efficiency Reduction (%) 1.3 GHz 87.4 70 55 2.6 GHz 82.9 55 65 The reduction in efficiency correlates very well with the discussion presented in Section 3.3. However, the increased substr ate thickness played an important role in maintaining the efficiency and gain paramete rs of the meandered shorted patch antennas within acceptable limits, which compare to the antennas of the first-generation FDR design.

PAGE 88

753.6.3 Radiation Patterns Radiation pattern simulations and measured results were obtained using the same methodology described in Section 2.3.2. Antenna simulations were performed using ADS Momentum. The radiation pattern of the meandered shorte d patch antennas was taken to be in the broadside direction (Refer to Figure 2.14, Chapter 2). The microstrip line feed was assumed to be parallel to th e direction that the resonance occurred. Therefore, the radiation patterns were linearly polarized in the direction parallel to the microstrip feedline, which corresponded to = 90 The antenna feedline was parallel to the E-Plane and perpendicular to the Hplane. As described in Section 2. 3.2 for E-Plane simulated patterns, was kept fixed at 90 and was swept 360 H-plane simulated patterns maintained fixed at 0 and was swept 360 Electromagnetic simulations of th e meandered shorted patch antennas were conducted utilizing a finite ground plane. The resulting radi ation patterns again displayed considerable back-side radiation. Th e radiation patterns were normalized to the maximum co-polarization value obtained. Figure 3.19 presents the co-polarized E and H-planes for the 1.3 GHz antenna at the particular frequency of in terest. Figure 3.20 presents th e co-polarized E and H-planes for the 2.6 GHz antenna at the particular fre quency of interest. Th e plots indicate good agreement between the simulated and measured results.

PAGE 89

0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 EPlane Co-Pol (Sim) EPlane Co-Pol (Meas) 0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 HPlane Co-Pol (Sim) HPlane Co-Pol (Meas) Figure 3.19: Comparison of Measured and Simulated Co-Polarized E-Plane (left) and H-Plane (right), Patte rn of the 1.3 GHz Antenna, (dB) 0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 EPlane Co-Pol (Sim) EPlane Co-Pol (Meas) 0 45 90 135 180 225 270 315 -50 -40 -40 -30 -30 -20 -20 -10 -10 HPlane Co-Pol (Sim) HPlane Co-Pol (Meas) Figure 3.20: Comparison of Measured and Simulated Co-Polarized E-Plane (left) and H-Plane (right), Patte rn of the 2.6 GHz Antenna, (dB) E and H-plane cross-polarization measur ed patterns are presented in Figure 3.21. 76

PAGE 90

0 45 90 135 180 225 270 315 -10 -8 -8 -6 -6 -4 -4 -2 -2 2 2 EPlane Cross-Pol (Meas) HPlane Cross-Pol (Meas) 0 45 90 135 180 225 270 315 -30 -25 -25 -20 -20 -15 -15 -10 -10 EPlane Cross-Pol (Meas) HPlane Cross-Pol (Meas) Figure 3.21: Cross-Polarized Measurement of the E and H-Plane Radiation Pattern of the 1.3 GHz Antenna (left) and 2.6 GHz Antenna (right), (dB) The measurements were normalized to th e maximum co-polarization value obtained for each antenna. Considerable cross-polar ization is demonstrated. Some possible causes for the cross-polarization levels are al so discussed in Section 2.3.2, of Chapter 2. A lowered width to length ratio of the antenna and the utilization of a microstrip feed can increase the polarization levels. Cross-polarization levels for shorted patch antennas have been reported to be higher than for half-wave patches, [2], [36]. Crosspolarization occurs since more current flows along the width of the quarterwavelength antenna compared to a half-wav elength antenna since the width to length ratio is lower. 3.7 Conclusions A review of antenna parameters, limitations of electrically small antennas and some of the techniques utilized in redu cing the size of planar antennas has been 77

PAGE 91

78presented. A study of different meandered sl ot antenna design geometries for operation at 1.3 GHz was performed. It was shown from the meandered slot antenna designs presented that the reduction in size did not result in a significant reduction of the antennas efficiency parameter. Effective meandering of the slot antenna geometry resulted in a minimal shift of the antenna s resonant frequency compared to the traditional CPW-fed slot antenna. Meandere d shorted patch antennas operating at 1.3 GHz and 2.6 GHz were designed, fabricated and measured. Good agreement occurred when comparing the expected and measured response for these electrically small antennas. The gain parameter for the mea ndered shorted patch antennas was improved by increasing the substrate thickness to 7.874 mm. The meandered shorted patch antennas studied yielded a reduc tion of approximately 85% in the overall size of the FDR design.

PAGE 92

79 CHAPTER 4 A MINIATURIZED AND TUNABLE FREQUENCY DOUBLING REFLECTENNA 4.1 Introduction The motivation for this research was a desire to design a compact frequency doubling reflectenna, which could be utilized primarily in sensing applications. The previous chapters of this thesis describe d the foundations for the realization of this miniature harmonic re-radiator. In Chapter 2, the methods used to design and to produce an efficient first-generation 1.3 GHz 2.6 GHz FDR were presented. It was observed that in order to obtain a mini aturized FDR, the size of the input and output antennas had to be reduced. A discussion on electrical ly small antennas and the methodology utilized to achieve compact antennas, which were im plemented in a second-generation FDR were outlined in Chapter 3. This chapter presents a 1.3 GHz 2.6 GHz miniature FDR design, which utilizes the meandered shorted patch antennas discussed in Chapter 3. Since the FDR operates at low input power levels, this research focu sed on producing an efficient device, which would exhibit low conversion lo ss values. Careful attention was focused, once more, on utilizing adequate performa nce prediction tools and measurement techniques. The miniature FDR obtained presented a size re duction of 74% and a maximum measured

PAGE 93

80conversion efficiency of 1.8%. The desi gn of a compact tunable FDR will also be covered in this chapter. It was discovered that a re-configurable harmonic re-radiator can easily be obtained by employing variable impedance devices in its design. 4.2 Miniaturized 1.3 GHz 2.6 GHz FDR Design The design of the miniaturized freque ncy doubling reflectenna followed the same principles as presented for the first-generati on design, which was discussed in Chapter 2. The FDR consisted of the receive and the tr ansmit meandered shorted patch antennas, the diode doubler and matching circuits. As in th e first-generation harmonic re-radiator, the HSCH-9161 GaAs Schottky diode wa s used as the doubling element. Maximum power transfer was achieved wh en the doublers circuit impedance was matched to the antennas input impedance. The implementation of the meandered shorted patch antennas in the miniaturized FDR design resulted in a decrease of 8.8 dB of gain when compared to the quarter-wavelength shorted patch antennas used in the firstgeneration harmonic re-radiator. Therefore, maximum power transfer was especially important. The meandered shorted patch antennas pr esented lower real input impedance and higher input reactance when compared to the quarter-wavelength shorted patch antennas used in the first-generation FDR. Thes e parameter characteristics are natural for electrically small antennas. The an tennas presented input impedances of (12.7 + j335) at 1.3 GHz and (19.3 + j25.15) at 2.6 GHz.

PAGE 94

81 The main objective of the FDR design was to achieve maximum conversion efficiency. Therefore, impedance matchi ng circuits were designed to maximize the power transfer between the r eceive and transmit antennas and the input and output of the diode doubler. The input and output matching ci rcuits consisted of distributed elements and surface mount components. Implementation of both types of elements ensured that low loss was achieved while compactness of the whole design was maintained. The input matching circuit consisted of an open-circuited stub, a series 0.2 pF capacitor and a 100 nH shunt induc tor. It was observed that the inclusion of a shunt inductor at input of the network improved the ove rall conversion efficiency of the design. At the output, the matching circuit consisted primarily of series distributed lines and a series 3.3 nH inductor. The overall circuit model for the miniaturized FDR design, which includes the impedance matching netw orks, is presented Figure 4.1.

PAGE 95

Figure 4.1 requency Doubling Reflectenna for Computer-Aided Analysis Agilents Advanced Design System, (ADS), CAD software was utilized in order to predict the performance of this second-generation FDR. Conversion efficiency is given, as presented in Chapter2 by equation (2 .1). Equation (2.1) is repeated here for convenience as equation (2.1): : Overall Equivalent Circ uit Model of Miniaturized F in out nP P (2.1) 82

PAGE 96

Chapter 2 presented a discussion on the criteria used to extract and determine the number of harmonics considered in the circuit simula tions. Utilizing the results from that study, harmonic impedances up to the 4th order were utilized to pr edict the conversion loss of the diode doubler. It is im portant to note that this st udy was not done for the devices presented in this Chapter. Figure 4.2 presents the conversion gain of the diode doubler for the miniaturized harmonic re-radiator. -50 -40 -30 -20 -10 0Pin to FDR at 1.3 GHz (dBm) -40 -35 -30 -25 -20 -15 -10C o n v e r s i o n G a i n ( d B ) 1st Generation FDR Miniature FDR Figure 4.2: Simulated Conversion Ga in for First-Generation FDR Diode Doubler and Miniaturized FDR Diode Doubler The maximum predicted conversion gain was -1 6.7 dB at an input power to the diode doubler of -22.7 dBm. This maximum convers ion gain was 3 dB lower than the firstgeneration FDR design. The decrease in conve rsion gain can be attributed to the loss, which occurs when impedance matching the high input reactance of the antennas to the impedance at the input and output of the diode doubler. 83

PAGE 97

Figure 4.3 presents the design of a miniaturized FDR device. Figure 4.3: Illustration of Miniaturized Frequenc y Doubling Reflectenna Substrate thickness was used to increase the gain of the meandered shorted patch antennas employed in the second-generati on multiplier design. Therefore, the miniaturized FDR device was realized on a 310 mil thick Taconic substrate board with an r of 6.15. The harmonic re-radiator measur ed 22 mm x 8.8 mm, which represents a size reduction of 75% in terms of length and widt h when compared to the first-generation FDR device. 4.3 Miniaturized and Tunable FDR Design A final objective of the research presente d in this thesis was to determine the feasibility of utilizing the FDR device for th e RF transmission of modulated data. In order to demonstrate such a capability, a tuna ble harmonic re-radiator was designed. As 84

PAGE 98

85a sensor node, the FDR device would be encoding information by modulating the reradiation. This information could then be decoded by a central transceiver system. Collection of data transmitted by the FDR sensor node was performed at the 2nd harmonic. The 2nd harmonic was used in order to alle viate the return that occurs from unwanted objects at the fundamental. The tunable FDR was developed to operate eith er in an on-state or an off-state. In the on-state, the maximum achievable power transfer existed between the antennas and the diode doubler. Thus, the re-radiation at 2fo was the maximum achievable. In an offstate, no re-radiation occurs at the 2nd harmonic. In the on-state, maximum power transf er between the antennas and the diode doubler occurred when they were impedance ma tched. Therefore, maximum re-radiation occurred at the 2nd harmonic. In contrast, if the i nput impedance of the antennas is not well-matched to the source and load impedance of the diode, reduced re-radiation will occur. The FDR device would then be in an off-state. During circuit simulations it was observed that the performance of the FDR device was most sensitive to effects at th e input to the diode doubler. The impedance match between the receive antenna and the sour ce terminal of the diode doubler can be modified by employing a variable reactance device, which modifies the impedance at the input of the network. Therefore, the variable reactance device could be operated in two states. One state would provide the reactance to ensure impedance match between the receive antenna and the input to the diode A second state would impede maximum power transfer between the elements. This e ffect resulted in the harmonic re-radiators desired onand off-states.

PAGE 99

In this research and as a proof-of-con cept, a semiconductor varactor was chosen as the variable react ance device. The varactor employed was the Metelics MSV34,0670805. A model of the Metelics varactor wa s developed by using S-parameter data and RF impedance data. Figure 4.4 diagrams the model of the varactor employed in computer-aided simulations. Figure 4.4: Circuit Model for the Metelics MSV34,067-0805 Varactor In the model, C represents the series capacitance of the varactor, which can be described as the junction capacitance of a reversed biased junction as a function of the reverse bias voltage applied to the varactor. The junction capacitance, C, is given by m j r joV V C VC )1( )( (4.1) In equation (4.1), three constants specify the junction capacitance. Cjo is the zero-bias junction capacitance, Vj is the junction potential and m is a unitless grading coefficient, which is a value that depends on the nature of the semiconductor junction of the varactor. Vr denotes the reverse bias potentia l applied to the varactor. The L parameter represents the series inductance and the R parameter represents the effective series resistance. The L and R parameters are the result of parasitic effects of the varactor component at RF 86

PAGE 100

frequencies. The constants of the junc tion capacitance equation for the Metelics MSV34,067-0805 model, were found to be Cjo = 2.05 pF, Vj = 0.58 V m = 0.37. The inductance value was found to be 1.74 nH. The effective series resistance is given by R = (4.2) 5.0freqba where a and b are constants. For this particular varactor a = 1.8, b = 1.3 x 10-5. The varactor was to be implemented at the input matching network. For ease in design, the varactor was implemented at the input matching network of the miniaturized FDR design. A 1.5 pF capacitor was placed in series with the varactor to obtain appropriate impedance for the shunt confi guration that was comparable to the one obtained with the miniaturized harmonic re-radiator. Bias was supplied to the varactor through a 10 K resistor. The input matching circuit consisted of a shunt 1. 5 pF capacitor, a 9.5 nH seri es inductor and a 8.2 nH shunt inductor. As in the miniaturized FDR de sign, at the output matching network, the matching circuit consisted primarily of seri es distributed lines and a series 3.3 nH inductor. Figure 4.5 presents the layout fo r the miniaturized FDR design, which includes the impedance matching networks and the varactor implementation. 87

PAGE 101

Figure 4.5: Layout of Tunable Frequency Doubling Reflectenna Figure 4.6 presents the overall equivale nt circuit model for the tunable FDR. 88

PAGE 102

Figure 4 quency Doubling Reflectenna for Computer-Aided Analysis The varactor was operated at two differe nt voltages. At a bias of 17 V, the varactor provided an impedance such that the conversion gain obtained from the diode doubler was optimum. When the varactor was biased with 0 V, the reactance it provided produced a significant change in the performance of the devi ce. A significant difference in the conversion gain obtained from the di ode doubler translated in to a change in the frequency of operation of the device. Figure 4.7 illustrates this difference in conversion gain for the two different bias voltages. .6: Overall Equivalent Ci rcuit Model of Tunable Fre 89

PAGE 103

90 -50 -40 -30 -20 -10 0Pin to FDR at 1.3 GHz (dBm) -100 -90 -80 -70 -60 -50 -40 -30 -20 -10C o n v e r s i o n G a i n ( d B ) Conv Gain 0 Volts Conv Gain 17 VoltsF On Conversion Gain was Obtained when the Varactor was Biased at 17 V, (On-State) Both the miniaturized and tuneable devi ces were fabricated and assembled in the same manner. The Taconic board with a substrate thickness of 310 mil and r = 6.15, which was required for this design, was not readily available from the manufacturer. Therefore, a board with a th ickness of 250 mil and one w ith a thickness 60 mil were bonded together using non-conductive epoxy. The copper on the substrate boards was patterned by using the standard copper et ching techniques, which are described in Appendix A. The surface mount component s were assembled on the board by using standard solder reflow techniques. Silv er epoxy bonding was utilized to assemble the GaAs Schottky diode. The vias were soldered to ground using copper wires. igure 4.7: Simulated Conversion Gain fo r the Tunable FDR Diode Doubler: Maximum

PAGE 104

Figure 4.8 presents the conversion ga in of the diode doubler for the three harmonic re-radiator designs. -50 -40 -30 -20 -10 0Power input to Diode Doubler (dBm) -45 -40 -35 -30 -25 -20 -15 -10C o n v e r s i o n G a i n ( d B ) 1st Generation FDR Miniature FDR Tunable FDR Figure 4.8: Simulated Conversion Ga in for First-Generation FDR Diode Doubler, the Miniaturized FDR Diode Doubler and the Tunable FDR Diode Doubler when the Va ractor was Biased at 17 V The maximum predicted conversion gain for the tunable design was -19.6 dB at an input power to the diode doubler of -19.4 dBm. This maximum conversion gain was 6 dB lower than the first-generation FDR design. As described for the miniature design, the decrease in conversion gain can be attr ibuted to the loss, which occurs when impedance matching the high input reactance of the antennas to the impedance at the input and output of the diode doubler. Also, the addition of a semiconductor varactor to the device increases loss due to a reduction of the quality factor of the network. 91

PAGE 105

924.4 Results Measurements on the miniature and tuna ble FDR devices, whic h were fabricated, were performed with the test measurement se tup described in Section 2.4. The FDR was placed in an anechoic chamber, where it rece ived the transmitted 1.3 GHz signal, doubled its frequency and reradiated it at 2.6 GHz. For an illustration of the hardware test-bench utilized for performing measurements refer to Section 2.4. Output power measurements were conducted using a spectrum analyzer which was connected to the 2.6 GHz receiving antenna. There were two major differences to note from the first-generation measurement setup. A narrow bandpass f ilter was connected between the spectrum analyzer and the 2.6 GHz receiving antenna in order to filter out any unwanted harmonics. A high power amplifier with a power output of 40 dBm to provide input power levels to the diode doubler that were greater than -23.5 dBm was employed. The 22 dBm power output amplifier, described in Section 2.4, was used for input power levels to the multiplier, which were lower than 23.5 dBm. Path spreading losses were calculated using the free-space path loss formula. The output power received at the 2. 6 GHz antenna for two fabricated FDR devices was measured using the setup descri bed above. The multiplier conversion gain values were calculated as well by using the relationship CG = Pout(dB) Pin(dB) + Receiving system loss(dB). (4.3) The output power received at the 2.6 GHz antenna is presented in Figure 4.9.

PAGE 106

-40 -35 -30 -25 -20Power Input to the Diode Doubler (dBm) -120 -115 -110 -105 -100 -95 -90 -85 -80 -75 -70P o we r R e c e i v e d a t 2 6 G Hz ( d B m ) Expected Meas FDR1 Meas FDR2 Figure 4.9: Comparison of Measured and Expected Doubler Power Received for the Miniaturized FDR at a Source Frequency of 1.3 GHz The multiplier conversion loss curves for two miniature FDRs are presented in Figure 4.10. -40 -35 -30 -25 -20Power Input to the Diode Doubler (dBm) -40 -35 -30 -25 -20 -15C o n v e r s i o n Ga i n ( d B ) Expected Meas FDR1 Meas FDR2 Figure 4.10: Comparison of Measured and Expected Doubler Conversion Gain for the Miniaturized FDR at a Source Frequency of 1.3 GHz 93

PAGE 107

94 Good correlation can be observed between the measured curves for the miniature FDRs 1 and 2 and their expected perform ance. The measured conversion gain for miniature FDR 2 was -25.6 dB when the inpu t power to the diode doubler was -31.5 dBm and a maximum value of -17.4 dB was obtained for an input power of -17.5 dBm. This translates into a maximum measured conversion efficiency of 1.8%. By considering the simulated value of gain of the input and output antennas, the measured conversion gain of the entire miniature FDR 2 was -33.5 dB for an input power level at the FDR receiving antenna of -23.5 dBm. Measured values differed from simulated values as the input power level to the diode doubler decreased. The variation can be attr ibuted to the fact that only up to fourth order harmonics were consider ed when performing circuit simulations. A study such as the one performed for the first-generation ha rmonic re-radiators, which observed the effect on the conversion loss values obtained when considering higher order harmonics was not conducted for these miniature devices. In addition, the frequency sensitivity of the miniature FDRs was not analyzed. Th e frequency at which the maximum doubler conversion efficiency occurred was not measured The slight shift in performance might be due to fabrication tolerances. The 1.3 GHz dual-patch array, which was used as the transmit antenna, possessed a narrowband response and provided less gain at frequencies other than at 1.3 GHz. If the multipliers perform optimally at a frequency not equal to 1.3 GHz, the gain used for calculations of the expected power received at 2.6 GHz might be an overestimation. The measured output power received at the 2.6 GHz antenna for the tunable FDR is presented in Figure 4.11.

PAGE 108

-36 -31 -26 -21 -16 -11Power Input to the Diode Doubler (dBm) -120 -110 -100 -90 -80 -70 -60P o w e r R e c e i v e d a t 2 6 G H z ( d B m ) -67.2 -70.6 -77 -86.1 -94.1 -106.4 -73.1 -79.5 -87 -98.2 -101.9 -113.5 Meas FDR1 17V Meas FDR2 17V Meas FDR1 0V Meas FDR2 0V Figure 4.11: Comparison of Measured Doubler Power Received for the Tunable FDR at a Source Frequency of 1.3 GHz (Solid Lines Represent Performance of the Diode when Varactor wa s Biased at 17 V) The multiplier conversion loss curves for the miniature tunable FDR are presented in Figure 4.12. -36 -31 -26 -21 -16 -11Power Input to the Diode Doubler (dBm) -45 -40 -35 -30 -25 -20 -15C o n v e r s i o n G a i n ( d B ) -20.5 -18.9 -20.3 -23.4 -27.4 -31.7 -26.4 -27.8 -30.3 -35.5 -35.2 -38.8 Meas FDR1 17V Meas FDR2 17V Meas FDR1 0V Meas FDR2 0V Figure 4.12: Comparison of Measured Doubler Conversion Gain for the Tunable FDR at a Source Frequency of 1.3 GHz (Solid Lines Represent Performance of the Diode when Varactor Biased at 17 V) 95

PAGE 109

96 The measured results obtained from the tunable harmonic re-radiators fabricated illustrate the re-configurability of the device. In Figures 4.11 and 4.12, measured values are labeled for the tunable FDR 2 performa nce curves. As can be observed, the difference in bias voltage provided to the varactor caused a difference in conversion gain values in the range of 6 10 dB. This conversion gain variation demonstrates the tunability of the device. The conversion gain for the tunable FDR 2 was -27.4 dB when the input power to the diode doubler was -27.5 dBm. A maximu m value of -18.9 dB was obtained for an input power to the diode doubler of -12.2 dBm. These values translate to a maximum measured conversion efficiency of 1.3%. The conversion gain of the entire tunable harmonic re-radiator 2 was -35.4 dB for an input power level to the FDR receiving antenna of -19.5 dBm. This value was obtai ned by considering the simulated values for gain of the input and output antennas of the FDR device. These values correspond to optimum FDR operation which occurs when the varactor is bi ased at 17 V. 4.5 Conclusions A miniaturized frequency doubling reflectenna was designed, fabricated and tested. At an input frequency of 1.3 GHz a multiplier conversion efficiency of 1.3% was obtained at an input power of -30 dBm. Meandered shorted patch antennas were successfully implemented in the design of th e harmonic re-radiator. The harmonic reradiators were produced using careful manufacturing methods and yielded a small spread in performance. It was demonstrated that the operation of th e FDR device could be controlled by using a varactor element, whic h provided variable capacitance at the input

PAGE 110

97of the network. Due to the miniature and tunable designs achieved, the FDR exhibits great potential as a sensor node.

PAGE 111

98 CHAPTER 5 SUMMARY AND RECOMMENDATIO NS FOR FUTURE WORK 5.1 Summary This thesis presented the research associated with the design of efficient and compact 1.3 GHz 2.6 GHz frequency doubling reflectenna devices. The work focused on creating a simple and quasi-planar passive design, which would not require extremely precise manufacturing methods to produce. It was demonstrated that a miniature harmonic re-radiator, which operates efficien tly at low input power levels, could be produced. Significant attention was give n to optimization of the c onversion efficiency of the FDR. In order to optimize the conversion efficiency, the input impedances, at the frequencies of interest, of the receive and transmit antennas were approximately conjugate-matched to the input and output impedances of the diode doubler. Compact and low loss matching circuits were designed to maximize the power transfer between the antennas and the diode. The conversion gains of the harmonic re-radiators, described in this thesis, displayed an improvement over t hose reported for similar devices such as a Schottky diode doubler/antenna transponder desi gn. A first-generation design yielded a conversion efficiency of 1% at a power input to the diode doubler of -30 dBm. Literature reports of comparable frequency multipliers coupled with antennas were operated at

PAGE 112

99higher input power levels, whic h provided for higher conversion efficiencies in those devices. It was demonstrated that measurements obtained from both the first-generation and miniaturized FDR devices co rrelated well to the simulated values. In order to ensure accurate circuit simulations, the diode mode l parameters provided by the manufacturer were optimized. In addition, it was demonstr ated that utilizing an adequate number of harmonic impedances, which were derived fr om the input and output antennas in the circuit simulations, provided for a more accura te prediction of the performance of the FDR. The antennas utilized in the design of the harmonic re-radiators were shorted patch antennas. Implementation of quarte r-wavelength shorted pa tch antennas produced a compact first-generation FDR device. Thes e antenna designs were optimized for length and width to obtain desired i nput impedances at desired fre quencies. This research focused on reducing the size of the first-gene ration harmonic re-radiat or. A miniaturized FDR was developed as a s econd-generation device. Development of the secondgeneration device involved an in depth study of meandered shorted patch antennas. The meandered antenna provided the principle mi niaturization mechaniz ation for the secondgeneration device. In comparison with th e quarter-wavelength shorted patch antennas utilized in the first-generation design, th e electrically small meandered shorted patch antennas yielded a significant re duction in expected efficiency levels. It has been wellreported in the literature that antenna effici ency decreases as the volume that the antenna occupies decreases. It was dem onstrated that by increasing the thickness of the substrate, used to implement the meandere d shorted patch antennas, the expected values of antenna

PAGE 113

100efficiency and gain also incr eased. A study of the vias, wh ich were used to short the patch antennas, revealed that via resistan ce must be minimized in order to obtain expected antenna efficiency and performance le vels. It was observed that high levels of cross-polarization were exhibite d by the shorted patch antenn as used for the harmonic reradiators. Therefore, shorted patch antennas can be utilized only if cons iderable levels of cross-polarization can be tolerated. The first-generation FDR design meas ured 44 mm x 17 mm. The size of the miniaturized FDR device was reduced by 75%. The second-generation design resulted in a device that measured 22 mm x 8.8 mm. Due to the high input reactance values obtained from the meandered shorted patch antennas, which were utilized in the miniaturized design, losses associated with the matching networks of the miniaturized FDR device were higher. Therefore, the conve rsion efficiency, at an input power to the diode of -30 dBm, was 3 dB lower than the first-generation design. A final accomplishment of this research was a demonstration of the tunability of the FDR device. It was demonstrated that the tunability of the FDR could be easily obtained by utilizing a variable reactance device at the input of the network. Therefore, the use of the frequency multiplier as a re -configurable device holds much potential. 5.2 Recommendations Efficient passive multipliers, which can al leviate the clutter radiation that occurs at the frequency of the incide nt radiation in a ra dar transceiver system, are useful. The harmonic re-radiator described in this thesis is a promising device in sensing applications. The following are a few suggestions aimed at improving and exte nding this research.

PAGE 114

101 In this research, a power input to the diode of -30 dBm was chosen as a benchmark value for design purposes. The impe dance values of the diode, at the source and load, were dictated by this input power level. However, time-domain methods and harmonic balance simulations could be employed to provide a more in-depth analysis of the optimum power levels. Such an analysis would provide values for source and load impedances, which could maximize the conversi on efficiency of the diode doubler. Then the input and output antennas of the network could be designed to be the conjugate-match of the optimum diode source and load impedances. More consideration can be given to th e compact antennas used for the harmonic re-radiator design. Even though the planar dimensions of the meandered shorted patch antenna designs decreased in size, substrate thickness was increased to optimize for efficiency and gain. In addition, the shorte d patch antennas used in this design exhibited high levels of cross-polarizati on. Different types of electri cally small antennas should be investigated to alleviate these shortcomings. As an alternate choice, compact slot an tennas could be utilize d if the FDR design was to be implemented using a CPW layout conf iguration. Even if a CPW layout is not employed for the circuitry of the re-radiator these antennas could still be used in conjunction with microstrip elements if a low loss and compact CPW-to-microstrip transition design was being considered. Slot antennas are less sensit ive to manufacturing tolerances. Slot antennas can also be polar ized in a desired manner using a combination of strip and slot conductors. In addition, a center-fed slot antenna can provide a high radiation resistance. The 1.3 GHz meandered slot antennas st udied and described in this

PAGE 115

102thesis presented expected efficiency valu es, which were comparable to the firstgeneration quarter-wavelengt h shorted patch antennas. A study of alternate multipli cation elements could also be undertaken. Different diodes could be tested in orde r to select one that provided the best performance. Active devices such as FETs can also be employed to increase the conversi on efficiency of the harmonic re-radiator if appropriate biasing schemes are investigated. The research presented in this thesis can lead to the development of an FDR device, which can serve as a sensor for RF transmission of modulated data. The possibility of utilizing the FDR in a modulation scheme was not explored in this research. However, the design simplicity associated with obtaining a tunable FDR was demonstrated. As an extension of this rese arch, analyzing alternate low loss integrated tuning structures, to replace the semiconductor varactor utilized in the present design, should be investigated. Additional research associated with implementation of the FDR as a viable sensing device should be undertaken. The harm onic re-radiator could be tested in a multi-tone environment. Such testing would assess performance trends and levels when the device is subjected to several signals at different frequencies. The data presented in this research were all broadside measurements. Measurements at different angles of incide nce should be performed. A more detailed study of how the polarization leve ls affect the re-radiated si gnal from the FDR needs to be performed.

PAGE 116

103 REFERENCES [1] K. Chang, Microwave Solid -State Circuits and Applications, John Wiley & Sons: New York, p.51, 1994 [2] R. Garg, P. Bhartia, I. Bahl and A. Ittipiboon, Microstrip Antenna Design Handbook, Artech House: Norwood, MA, pp.1-657, 2001 [3] M. T. Faber, J. Chramiec and M. E. Adamski, Microwave and Millimeter-Wave Diode Frequency Multipliers, Artech House: Norwood, MA, pp.1-87, 1995 [4] N. Paravastu, Novel Frequency Multi plier Architectures for Millimeter Wave Applications. Ph.D. diss, University of Virginia, Charlottesville, VA, Jan. 2005 [5] E. Carman, K. Giboney, M. Case, M. Kamegawa, R. Yu, K. Abe and M.J.W. Rodwell, -39 GHz Distributed Harm onic Generation on a Soliton Nonlinear Transmission Line, IEEE Microwave and Guided Wave Letters, vol. 1, no. 2, pp.28-31, Feb. 1991 [6] E. Carman, M. Case, M. Kamegawa, R. Yu, K. Giboney and M.J.W. Rodwell, V-band and W-band Broadband, Monolithic Distributed Frequency Multipliers, IEEE Intl. Microwave Symp. Digest, vol. 2, pp.819-822, June 1992 [7] S. Helbing, M. Cryan, F. Alimenti, P. Mezzanotte, L. Roselli and R. Sorrentino, Design and Verification of a Novel Cro ssed Dipole Structure for Quasi-Optical Frequency Doublers, IEEE Microwave Gu ided Wave Letters, vol. 10, pp.105107, Mar. 2000 [8] M. Kim, V. M. Lubecke, S. C. Martin R. P. Smith and P. H. Siegel, A Planar Parabola-Feed Frequency Multiplier, IEEE Microwave and Wireless Comp. Letters, vol. 7, pp.60-62, Mar. 1997 [9] S. B. Yeap, C. G. Parini, J. A. Dupuy and M. R. Rayner, FDTD Simulation and Measurement of a 90 GHz Quasi-Optical Annular Slot Recei ver, Proceedings, IEEE Microwaves, Antennas and Propagation, vol. 152, pp.117-123, Apr. 2005 [10] S. Hollung, J. Stake, L. Dillner a nd E. Kollberg, A 141 GHz Integrated QuasiOptical Slot Antenna Tripler, Proceedings, IEEE Microwaves, Antennas and Propagation Society, In ternational Symposium, vol. 4, pp.2394-2397, Aug. 1999

PAGE 117

104[11] N. Paravastu and R.M. Weikle II, A 40-80 GHz Quasi-Optical Balanced Doubler using Nested Ring-Slot Antennas, IEEE Antennas and Propagation Society, International Symposium, vol.1, pp.270-273, 2002 [12] Y.H. Suh and K. Cha ng, A High-Efficiency Dual -Frequency Rectenna for 2.45 and 5.8 GHz Wireless Power Transmissi on, IEEE Transactions on Microwave Theory and Techniques, vol.50, no.7, pp.1784-1789, July 2002 [13] B.Strassner and K. Chang, .8 GHz Circularly Polarized Rectifying Antenna for Wireless Microwave Power Transmission, IEEE Transactions on Microwave Theory and Techniques, vol.50, no.8, pp. 1870-1876, Aug. 2002 [14] T. Umeda, H. Yoshida, S. Sekine, Y. Fujita, T. Suzuki and S. Otaka, A 950MHz Rectifier Circuit for Sensor Networks with 10m-distance, IEEE International Solid-State Circuits Conference Di gest, vol. 1, pp.256-597, Feb. 2005 [15] T. P. Ketterl, M icroand Nano-Scale Switche s and Tuning Elements for Microwave Applications Ph.D. diss, Univ ersity of South Florida, Tampa, FL, Mar. 2006 [16] B.G. Colpitts and G. Boiteau, Har monic Radar Transceive r Design: Miniature Tags for Insect Tracking, IEEE Transactions on Antennas and Propagation, vol.52, no.11, pp.2825-2832, Nov. 2004 [17] J. R. Riley and A. D. Smith, Des ign Considerations fo r a Harmonic Radar to Investigate the Flight of Insects at Lo w Altitude, Computers and Electronics in Agriculture, Elsevier, vol. 35, pp.151-169, Aug. 2002 [18] A. Polh, A Review of Wire less SAW Sensors, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control, vol.47, no.2, pp.317-332, Mar. 2000 [19] A. Pohl and F. Seifert, New App lications of Wirelessly Interrogable Passive SAW Sensors, IEEE Transactions on Microwave Theory and Techniques, vol.46, no.12, pp.2208-2212, Dec. 1998 [20] L.M. Reindl and I.M. Shrena, W ireless Measurement of Temperature using Surface Acoustic Wave Sensors, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control, vol.51, no.11, pp.1457-1463, Nov. 2004 [21] A.Pohl and F. Seifert, Wirele ssly Interrogable SAW-Sensors for Vehicular Applications, Proceedings, IEEE Instrumentation and Measurement Technology Conference, Quality Measurements: The Indispensable Bridge between Theory and Reality, vol.2, June 1996, pp.1465-1468

PAGE 118

105[22] I.D. Robertson and I. Jaialy, RF Id Tagging Explained, Communications Engineer, vol.1, no.1, pp.20-23, Feb. 2003 [23] J. D. Griffin and G. D. Durgin, G ains for RF Tags using Multiple Antennas, IEEE Transactions on Antennas and Propa gation, vol.56, no.2, pp.563-570, Feb. 2008 [24] F. Fuschini, C. Piersanti, F. Paolazzi and G. Falciasecca, Analytical Approach to the Backscattering from UHF RFID Transponders, IEEE Antennas and Wireless Propagation Letters, vol.7, pp.33-35, 2008 [25] Z. Fang, R. Jin and J. Geng, A symmetric Dipole Antenna Suitable for Active RFID Tags, Electronics Letters vol.44, no.2, pp.71-72, Jan. 2008 [26] W. Choi, H.W. Son, J. Yeo, J. J ae-young, J.H. Bae and C.G. Young, PlatformInsensitive Antenna for RFID Sensor Tag in the UHF Band, Proceedings, IEEE Antennas and Propagation Society, Inte rnational Symposium, pp.2277-2280, June 2007 [27] M. Hirvonen, P. Pursula, K. Jaa kkola and K. Laukkanen, Planar Inverted-F Antenna for Radio Frequency Identificatio n, Electronics Letters, vol.40, no.14, pp.848-850, July 2004 [28] H.-E. Nilsson, J. Siden, T. Olsson, P. Jonsson and A. Koptioug, Evaluation of a Printed Patch Antenna for Robust Microwave RFID Tags, IEEE Transactions, Microwaves, Antennas & Propagation, vol.1, no.3, pp.776-781, June 2007 [29] M.J. DeVincentis, S. Ulker and R.M. Weikle II, A Balanced HEMT Doubler for Quasi-Optical Applications, IEEE Micr owave and Guided Wave Letters, vol.9, no.6, pp.239-241, June 1999 [30] D. Singh, P. Gardner and P.S. Hall, Integrated Push-Push Frequency Doubling Active Microstrip Transponder, Electron ics Letters, vol.33, no.6, pp.505-506, 13 Mar. 1997 [31] J. Birkeland and T. Itoh, A Microstrip Based Active Antenna Doppler Transceiver Module, 19th European Microwave Conference, pp.172-175, Oct. 1989 [32] C.W. Pobanz and T. Itoh, A Micr owave Non-Contact Identification Transponder using Sub-Harmonic Interrogation, IEEE Transactions on Microwave Theory and Techniques, vol.43, no.7, pp.1673-1679, July 1995 [33] Y. Chung and T. Itoh, A New Architecture for an AlGaN/GaN HEMT Frequency Doubler using an Active In tegrated Antenna Design Approach, Conference Proceedings Asia-Pacifi c Conference Proceedings, Nov. 2002

PAGE 119

106[34] T. Weller, Activity Tag Interim Re port, Univ. of South Florida, Tampa, FL, Raytheon Systems, Fullerton, CA, Interim Report, Feb. 2006 [35] J. Grajal, V. Krozer, E. Gonzlez, F. Maldonado and J. Gismero, Modeling and Design Aspects of Millimeter-Wave a nd Submillimeter-Wave Schottky Diode Varactor Frequency Multipl iers, IEEE Transactions on Microwave Theory and Techniques, vol. 48, pp. 700-711, Apr 2000 [36] G. Kumar and K.P. Ray, Broadband Microstrip Antennas, Artech House, Norwood, MA, pp. 1-248, 2003 [37] K.L. Wong, Compact and Broadband Microstrip Antennas, New York: John Wiley & Sons, 2002, pp.1-325 [38] D.M. Pozar, Microwave and RF Desi gn of Wireless Systems, New York: John Wiley & Sons, 2001, pp.111-145 [39] C.A. Balanis, Antenna Theory, New York: John Wiley & Sons, 2nd ed., 1997, pp.28-570 [40] H.A. Wheeler, Fundamental Limitations of Small Antennas, Proceedings of the IRE, vol. 35, no. 12, Dec. 1947, pp.1479-1484 [41] L.J. Chu, Physical Limitation of Omni-Directional Antennas, Journal of Applied Physics, vol. 19, pp.1163-1175, Dec. 1948 [42] R.C. Hansen, "Fundamental limitations in antennas," Proceedings of the IEEE, vol.69, no.2, pp.170-182, Feb. 1981 [43] J. S. McLean, "A re-examination of the fundamental limits on the radiation Q of electrically small antennas," IEEE Tran sactions on Antennas and Propagation, vol.44, no.5, pp.672-676, May 1996 [44] H.A. Wheeler, "The Radiansphere ar ound a Small Antenna," Proceedings of the IRE vol.47, no.8, pp.1325-1331, Aug. 1959 [45] H. Wheeler, "Small antennas," IEEE Transactions on Antennas and Propagation, vol.23, no.4, pp.462-469, Jul 1975 [46] J.C.-E. Sten, A. Hujanen, P.K. Koivisto "Quality factor of an electrically small antenna radiating close to a conducting plane," IEEE Transactions on Antennas and Propagation, vol.49, no.5, pp.829-837, May 2001 [47] A.D. Yaghjian, S.R. Best, "Imped ance, bandwidth, and Q of antennas," IEEE Transactions on Antennas and Propagation, vol.53, no.4, pp.1298-1324, April 2005

PAGE 120

107[48] S.R. Best, "A discussion on the quality factor of impedance matched electrically small wire antennas," IEEE Transactions on Antennas and Propagation, vol.53, no.1, pp.502-508, Jan. 2005 [49] A.K. Skrivervik, J.-F. Zurcher, O. Staub, J.R. Mosig, "PCS antenna design: the challenge of miniaturization," I EEE Antennas and Propagation Magazine, vol.43, no.4, pp.12-27, Aug 2001 [50] K. L. Wong and S.C. Pan, Compact triangular microstrip antenna, Electron Letters, vol 33, pp. 433-434, Mar. 1997 [51] D. Supriyo, R. Mittra, Compact microstrip patch antenna, Microwave and Optical Technology Letters, vol. 13, no. 1, pp.12-14, 1996 [52] K. L. Lau, K.C. Kong, K. M. Luk, "A Miniature Folded Shorted Patch Antenna for Dual-Band Operation," IEEE Trans actions on Antennas and Propagation, vol.55, no.8, pp.2391-2398, Aug. 2007 [53] T. Ozdernir, P. Frantzis, K.F. Sabet, L.P.B. Katehi, K. Sarabandi; J.F. Harvey, "Compact wireless antennas using a supe rstrate dielectric lens," IEEE Antennas and Propagation Society In ternational Symposium, vol.3, no., pp.1678-1681, 2000 [54] H.Y. Wang, M.J. Lancaster, "A perture-coupled thin-film superconducting meander antennas," IEEE Transactions on Antennas and Propagation, vol.47, no.5, pp.829-836, May 1999 [55] J. Lin; T. Itoh, "Active integrated antennas," IEEE Transactions on Microwave Theory and Techniques, vol.42, no.12, pp.2186-2194, Dec 1994 [56] V. A. Thomas; K.-M. Ling, M.E. Jones, B. Toland, J. Lin, T. Itoh, "FDTD analysis of an active antenna," I EEE Microwave and Guided Wave Letters, vol.4, no.9, pp.296-298, Sep 1994 [57] D.R. Jackson, N.G. Alexopoulos, "S imple approximate formulas for input resistance, bandwidth, and efficiency of a resonant rectangular patch," IEEE Transactions on Antennas and Propagation, vol.39, no.3, pp.407-410, Mar 1991

PAGE 121

108 APPENDICES

PAGE 122

109Appendix A: Copper Etching Process Flow The photoresist utilized in the process flow below is SC-1827 positive photoresist. a. Clean the laminate board with ace tone/methanol and nitrogen dry, b. Dispense HMDS onto the board and spin at 3500 rpm for 30 seconds to dry completely, c. Spin SC 1827 at 500 rpm for 15 seconds followed by 2000 rpm for 30 seconds, d. Soft bake in the oven at 105 C for 20 minutes, (lay the board flat), e. Expose using broadband Quintel mask aligner for 55 seconds, f. Immersion develop with slight agit ation using MF 319 for ~75 seconds, g. Clean with deionized, (DI), water and nitrogen dry, h. Microscope inspect the image definition, i. If protecting laminate backsi de, repeat steps b and c, j. Hard bake in the oven at 110 C for 20 minutes, (lay the board flat), k. Ferric Chloride etch with slight agitation for ~8 minutes at 55 C, l. Clean with DI water and nitrogen dry, m. Microscope inspect the image definition, n. Remove photoresist with acetone/methanol.