|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
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 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2011 flu ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004982
Ultra-low power electronics for autonomous micro-sensor applications
h [electronic resource] /
by Rebeka Davidova.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 107 pages.
(M.S.E.E.)--University of South Florida, 2011.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: This thesis presented the research, design and fabrication associated with a unique application of rectenna technology combined with lock-in amplification. An extremely low-power harmonic transponder is conjoined with an interrogator base-station, and utilizing coherent demodulation the Remote Lock-In Amplifier (RLIA) concept is realized. Utilizing harmonic re-radiation with very low-power input, the 1st generation transponder detects a transmitted interrogation signal and responds by retransmitting the second harmonic of the signal. The 1st generation transponder performs this task while using no additional power besides that which accompanies the wireless signal. Demonstration of the first complete configuration provided proof of concept for the RLIA and feasibility of processing relevant information under "zero" power operating conditions with a remote transponder. Design and fabrication of a new transponder where the existing zero-bias transponder was modified to include a DC bias to the diode-based frequency doubler is presented. Applied bias voltage directly changed the impedance match between the receiving 1.3 GHz antenna and the diode causing a change in conversion loss. Testing demonstrated that a change in conversion loss induces an amplitude modulation on the retransmission of the signal from the transponder. A test of bias sweep at the optimal operating frequency was performed on the 2nd generation transponder and it was seen that a change of ~ 0.1 V in either a positive or negative bias configuration induced an approximate 15 dB change in transponder output power. A diode-integrated radar detector is designed to sense microwaves occurring at a certain frequency within its local environment and transform the microwave energy to a DC voltage proportional the strength of the signal impinging on its receiving antenna. The output of the radar detector could then be redirected to the bias input of the 2nd generation transponder, where this DC voltage input would cause a change in conversion loss and modulate the retransmitted interrogation signal from the transponder to the base station. When the base station receives the modulated interrogation signal the information sensed by the radar detector is extracted. Simulations and testing results of the fabricated radar detector demonstrate capability of sensing a signal of approximately -53.3 dBm, and accordingly producing a rectified DC voltage output of 0.05 mV. A comparison is made between these findings and the transponder measurements to demonstrate feasibility of pairing the radar detector and the 2nd generation transponder together at the remote sensor node to perform modulation of interrogation signals.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Weller, Thomas .
Frequency Doubling Reflectenna
x Electrical Engineering
t USF Electronic Theses and Dissertations.
Ultra Low Power Electronics for Autonomous Micro Sensor Applications by Rebeka Davidova 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. Gokhan Mumcu, Ph.D. Jing Wang, Ph.D. Date of Approval: March 2 1 201 1 Keywords: lock in amplifier, sensor network, frequency doubling reflecten na, diode doubler, conversion efficiency Copyright 2011, Rebeka Davidova
DEDICATION This thesis is dedicated to the most incredible woman I know. She is a woman whose commitment to her children takes precedence above all else in her life. She has a spirit of selfless devotion and of endless compassion. Her presence is soft and her eyes sh ow kindness. This is for my mother, Michele.
ACKNOWLEDGEMENTS I wish to acknowledge Dr. Thomas Weller for his guidance, his generous time, fortunate to have the opport unity to become part of the USF WAMI group and the experience s I have gained are invaluable. I would also like to thank my committee members, Dr. Gokhan Mumcu and Dr. Jing Wang. I would like to acknowledge all those who have he lped me along this journey : Bojana, my sister, for your endless support, honest advice, and selfless friendship ; my dear Chami my first real friend in Tampa I will never forget our afternoon walks, your amazing Sri Lankan cooking and our time in Turino, Italy ; Yohannes, for your countless motivationa l speeches and loyal friendship I will neve r forget Chapter Two; SBS, for teaching me to sit by the river of thoughts your positive words and creativity made me smile ; QB, for your inspiring words and upl ifting spirit ; David (Il Maestro) for your lovely serenades in the middle of the day and for always making me laugh regardless of my mood ; my dear Ebenezer, one of the most caring people I know I will never forget the chocolate bar ; all those in my 412 family who never he sitated to offer a helping hand : Ibrahim Nassar, Tony Price, Scott Skidmore, Evelyn Banabe, Michael Grady, and Maria Cordoba. Special thanks are due to my professors who have guided and formed my understanding of RF, a subject that has become close to my heart.
i TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ...... iii LIST OF FIGURES ................................ ................................ ................................ ...... iv ABSTRACT ................................ ................................ ................................ .................. vi CHAPTER 1 INTRODUCTION ................................ ................................ ................ 1 1.1 Lock In Amplifiers and Phase Sensitive Detection ................................ 2 1.2 Harmonic Re Radiator: 1 st and 2 nd Generation ................................ ........ 5 1.3 Diode Integrated Radar Detector ................................ ............................. 7 1.4 Overview and Con tributions of the Research ................................ .......... 9 CHAPTER 2 BACKGROUND OF RECTENNAS SIGNIFICANCE IN CURRENT WORK AND RELATED RESEARCH ....................... 11 2.1 Introduction ................................ ................................ .............................. 11 2.2 Diode Appli cations and Schottky Diode s ................................ ................ 12 2.3 Schottky Diode Characteristics and Application Specific Parameters ................................ ................................ ................................ 13 2.4 Wireless Power Transfer as an Applicat ion of Diode Rectification ........ 20 2.4.1 Development of Microwave Power Transmission (MPT) and Co ................................ ............ 21 2.4.2 MPT in So lar Power Satellites (SPS) and R elated E xperiments ................................ ................................ .............. 23 2.4.3 MPT for Wireless Power Distribution Systems in Buildings ................................ ................................ ................... 27 2.5 Rectennas and Remote Wireless Sensing Networks ................................ 29 2.5.1 Rectennas and Re trodirective Array Systems ........................... 29 2.5.2 Rectennas for Monitoring and Sensing of Infrastructure ................................ ................................ ............. 30 2.5.3 Energ y Harvesting and Recycling for U se in Sensor Networks ................................ ................................ ................... 31 2.6 Wireless Passive Surface Acoustic Wave Sensors ................................ .. 32 2.7 Rectennas U sing Harmonic Re R adiation for Sensing Applications ................................ ................................ ............................. 34 2.8 Conclusions ................................ ................................ .............................. 37
ii CHAPTER 3 REMOTE INTERROGATION SYSTEMS WITH HARMONIC RE RADIATION ................................ ........................... 38 3.1 Introduction ................................ ................................ .............................. 38 3.2 Introduc tion to Lock In Amplifiers ................................ ......................... 39 3.2.1 Phase Sensitive Detection ................................ ......................... 40 3.3 Detector and Lock In Amplifier Characterization ................................ ... 44 3.4 Background Work on Sensor Node Harmonic Repeaters 1 st Generation Transponder ................................ ................................ ...... 48 3.5 1 st Gene ration Tran sponder within RLIA System ................................ ... 50 3.6 Design of 2 nd Generation Transponder ................................ .................... 56 3.7 Conclusions ................................ ................................ .............................. 68 CHAPTER 4 APPLICATION OF RECTENNA TECHNOLOGY: DIODE INTEGRATED RADAR DETECTOR ................................ ... 69 4.1 Introduction ................................ ................................ .............................. 69 4.2 Design of Radar Detector ................................ ................................ ........ 70 4.3 Simulations of Radar Detector ................................ ................................ 72 4.4 Measurement Results of Radar Detector ................................ ................. 75 4.5 Conclusions ................................ ................................ .............................. 79 CHAPTER 5 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK ................................ ................................ .................. 80 5.1 Summary ................................ ................................ ................................ .. 80 5.2 Recom mendations for Future Work ................................ ......................... 83 REFERENCES ................................ ................................ ................................ ............. 85 APPENDICES ................................ ................................ ................................ ............. 90 Appendix A: Additional Tables of Measurements, L ist s and Specifications of E quipment ................................ ..................... 91 Appendix B : 1 st Generation Transponder Measurement Notes ...................... 92
iii LIST OF TABLES Table 2.1: Frequencies and relative a mplitudes of a diode detector within the square law region of output ................................ ................................ ...... 18 Table 3.1: (a) Link budgets for RLIA system at 3.16 mW and (b) 0.25 mW transmit powers respectively ................................ ................................ ........... 54 Table 3.2 : Comparison of P out of transponder at each tested bias over frequency sweep ................................ ................................ ............................. 63 Table 3.3 : Comparison of conversion loss for each transponder at each tested bias over frequency sweep ................................ ................................ .. 64 Table I: Sensitivity measurements of Narda detector (Model 4503 01) .................... 91 Table II: Measured and simulated sensitivity values of radar detector ....................... 93 Table III: List of parts used in testing configuration ................................ ..................... 94
iv LIST OF FIGURES Figure 1. 1 : Bl ock diagram of Lock In Amplifier ................................ ........................... 3 Figure 1. 2: Block d iagram of RLIA c oncept ................................ ................................ .. 4 Figure 1. 3 : Concept of Frequency Doubling Reflectenna,  ................................ ....... 5 Figure 1. 4 : Concept of biased transponder (2 nd generation transponder) ....................... 7 Figure 1. 5 : Concept of diode integrated radar detector ................................ .................. 8 Figure 2.1: Equivalent AC circuit model for a Schottky diode ................................ .... 14 Figure 2.2: Spectral output of detected modulated signal ................................ ............ 17 Figure 2.3: Square law region of a diode detector ................................ ....................... 19 Figure 2.4: Layout of Frequency Doubling Reflectenna t ransponder  .................... 35 Figure 2.5: Received power vs. frequency for 3 FDR devices ................................ ..... 36 Figure 2.6: Conversion g ain vs. frequency for 3 FDR devices ................................ .... 37 Figure 3.1: Phase sensitive detection performed by a lock in amplifier ...................... 41 Figure 3.2: Block diagram of detector c haracterization U sing Lock In Amplifier ................................ ................................ ................................ .... 45 Figure 3.3: Measured voltage sensitivity of Narda d etector (Model 4503 01) using Lock In Amplifier ................................ ................................ ............ 47 Figure 3.4: 1 st g eneration transponder layout: Z S and Z L are the input and output impedances of the diode. ................................ ................................ ........... 49 Figure 3.5: 1 st g eneration transponder expected and measured conversion gain vs. input power at 1.3 GHz ................................ ................................ ........ 49 Figure 3.6: 1 st g eneration transponder conversion gain vs. input power ...................... 50
v Figure 3.7: Block diagram of full Remote Lock In Amplifier interro gation system ...... 52 Figure 3.8: (a) Measured performance of the syst em at an interrogator to sensor distance of 1.5 m using a spectrum analyzer; (b) lock in amplifier .......... 56 Figure 3.9: Layout of 2 nd generation transponder ................................ ........................ 57 Figure 3.10: 2 nd generation transponder testing configuration ................................ ....... 58 Figure 3.11 : P out of transponder at 0 V bias over frequency sweep ............................... 59 Figure 3.12 : Conversion l oss of transponder at 0 V bias over frequency sweep ........... 59 Figure 3.13 : P out of transponder at +/ 0.05 V bias over frequency sweep ..................... 61 Figure 3.14 : P out of transponder at +/ 0.15 V bias over frequency sweep ..................... 61 Figure 3.15 : Comparison of P out of transponder at each tested bias over frequency sweep ................................ ................................ ........................... 63 Figure 3.16 : Comparison of conversion loss for each transponder at each tested bias over frequency sweep ................................ ................................ 64 Figure 3.17 : Bias vs. P out of transponder at 1275 MHz T x frequency ............................ 65 Figure 3.18 : Bias vs. P out of transponder at 1300 MHz T x frequency ............................ 67 Figure 4.1: Layout of diode integrated radar detector ................................ .................. 71 Figure 4.2 : Schematic for the conjugate matched radar signal detector ...................... 73 Figure 4.3 : Simulated conversion efficiency for the radar signal detector ................... 74 Figure 4.4 : Radar signal detector output voltage and current versus input power (left) and I V curve for the Schottky diode used in the multiplier of the harmonic transceiver (right) ................................ ............ 74 Figure 4.5 : Block diagram of testing configuration of the radar detector .................... 76 Figure 4.6: Measured vs. si mulated results of the radar detector at P Tx = 1300 MHz .... 77 Figure I: Sensitivity curve of Narda Schottky barrier detector (4503 01) ............... 94
vi ABSTRACT This thesis presented the research, design and fabrication associated with a unique application of rectenna technology combined with lock in amplification. An extremely low power harmonic transponder is conjoined with an interrogator base station, and utilizing coherent demodulation the Remote Lock In Amplifier (RLIA) concept is realized. Utilizing harmonic re radiation with very low power input, the 1 st generation transponder detects a transmitted interrogation signal and responds by retransmitting the second harmonic of the signal. The 1 st generation transponder performs this task while using no additional power besides that which accompanies the wireless signal. Demonstration of the first complete configuration provided proof of concept for the RLIA and feasibility of processing relevant i conditions with a remote transponder. Design and fabrication of a new transponder where the existing zero bias transponder was modified to include a DC bias to the diode based frequency doubler is presented Applied bias voltage directly changed the impedance match between the receiving 1.3 GHz antenna and the diode causing a change in conversion loss. T esting demonstrated that a change in conversion loss induces an amplitude modulation on the retransmission of the s ignal from the transponder. A test of bias sweep at the optimal operating frequency was performed on the 2 nd generation transponder and it was seen that
vii a change of ~ 0.1 V in either a positive or negative bias configuration induced an approximate 15 dB ch ange in transponder output power. A diode integrated radar detector is designed to sense microwaves occurring at a certain frequency within its local environment and transform the microwave energy to a DC voltage proportional the strength of the signal im pinging on its receiving antenna. The output of the radar detector could then be redirected to the bias input of the 2 nd generation transponder, where this DC voltage input would cause a change in conversion loss and modulate the retransmitted interrogatio n signal from the transponder to the base station. When the base station receives the modulated interrogation signal the information sensed by the radar detector is extracted. Simulations and testing results of the fabricated radar detector demonstrate cap ability of sensing a signal of approximately 53.3 dBm, and accordingly producing a rectified DC voltage output of 0.05 mV. A comparison is made between these findings and the transponder measurements to demonstrate feasibility of pairing the radar detecto r and the 2 nd generation transponder together at the remote sensor node to perform modulation of interrogation signals.
1 CHAPTER 1 INTRODUCTION There is no argument that radio has made a profound impact on humanity over the past hundred years. Some of the primary applications of this technology have been within the realm of mobile and telecommunications, navigation, and links for computer controlled systems. Advances continue to be made, p articularly within microwave technology, involving new applications. Microwave power transmission was the desired application behind the conception of the ], and since that time rectenna technology has continued to develop. N ew arenas have been discovered for this technology and methods of use have expanded, including remote sensing [2 ] monitoring of infrastructure [3 ] and energy harvesting and recycling [4 5 ]. The research described in this thesis focuses on a unique appl ication of sensing involving rectenna technology combined with phase sensitive detection implemented by a lock in amplifier within an interrogating transceiver. A harmonic re radiator is employed within the system as a remote sensor and signal interrogatio n is performed to communicate relevant information from its location. The goal of this research is to implement phase sensitive detection to enhance the ability of the interrogating transceiver to detect a very low power return signal from the remote senso r.
2 1.1 Lock In Amplifiers and Phase Sensitive Detection Lock in amplifiers are devices used to detect extremely small AC signals reaching as low as 100 dB m or more. It is not uncommon for lock in amplifiers to operate with sensitivities in the nano volt range, allowing for monitoring and extracting of information below the system noise floor. Lock in amplifiers have long been used in laboratories for accurate small signal detection and comp lex impedance measurement [6 8 ]. The technique employed by l ock in amplifiers to perform such tasks is referred to reference frequency. Noise signals at all other frequencies are rejected by the lock in amplifier and thus do no t contribute to the measurement. Phase sensitive detection is performed by the lock in amplifier by utilizing a phased locked loop that synchronizes to the external reference channel input (Figure 1.1) creating an internal reference signal. This internal r eference signal is then multiplied by the signal channel input. The product of the two signals is amplified and passed through a low pass filter, removing the AC components. The resulting signal will be zero unless the signal frequency is identical to the reference frequency, in which case a DC voltage proportional to the signal channel input will be seen at the lock tput (DC Out in Figure 1.1) [9 ].
3 Figure 1.1 : Block d iagram of Lock In Amplifier With the goal of implementing phase sensitive detection with a lock in amplifier in a remote fashion, the proposed scheme in Figure 1.2 was created. The scheme utilizes a lock in amplifier within an interrogating base station (interrogator node) and a low to zero power transponder as part o f the remote sensor node. The configuration is named a Remote Lock In Amplifier (RLIA) and it is unique in that the high power consuming electronics within the RLIA system are separated from the low power remote components, while the system as a whole perf orms phase sensitive detection to sense and gather remote intelligence.
4 Figure 1.2: Block diagram of RLIA concept.
5 1.2 Harmonic Re Radiator: 1 st and 2 nd Generation Employed within the RLIA system is a harmonic re radiat or, referred to as a Frequency D oubling Rectenna (FDR). The FDR was developed for l ow power sensing applications [2 ]. An illustration of the FDR c oncept is depicted in Figure 1.3 Using the principles of frequency multiplication, this device receives an interrogation signal at 1.3 GHz and retransmits the 2 nd harmonic at 2.6 GHz. The FDR is initially used in the first stage of demonstrating the RLIA concept by assuming the role of the transponder or remote sensor node. The printed circuit board design consists of two quarter wav elength patch antennas (1.3 GHz and 2.6 GHz) and a diode based doubler. The FDR was designed for minimum conversion loss for a low power input application, yet possesses sufficient sensitivity for efficient data collection within a remote interrogation sch eme. Figure 1.3 : Concept o f Frequency Doubling Rectenna [2 ] (1 st g eneration t ransponder)
6 The goal in the first phase of this research involved verifying the capability to effectively communicate with the 1 st g eneration low power remote transponder within the RLIA system. The Int errogator (RLIA) from Figure 1.2 was constructed using standard testing equipment and testing was performed at a distance between the interrogating node and remote se nsor node that would provide an input power giving the best possible conversion efficiency from the transponder (Chapter Three, Figure 3.7, Appendix A ). S weep s of tra nsmitted power and frequency were performed to characterize the 1 st g eneration transponder and determine the minimum detectable signal level arriving to the interrogator node from the remote sensor node. The entire RLIA system was characterized through measured link budgets, which were then verified with the calculated link budgets. The second phase of implementing the RLIA system involved a redesign o f the 1 st g eneration transponder to include a DC bias connection to the diod e doubler, as seen in Figure 1.4 (2 nd g eneration transponder). An applied bias voltage to the diode directly changes the impedance match between the receiving 1.3 GHz antenna and the doubler causing a change in conversion loss. This change in conversion loss will induce an amplitude modulating effect on the retransmission of the signal from the transponder by reducing the output power of the processed signal. Ideally, a small chang e in bias voltage will cause a decrease or increase in the ou t put power of the transponder to essentially turn according to the diagram seen in Figure 3.10 of Chapter Three (Appendix A) and it was verified that the new design in fact could process relevant information back to the interrogator node through this modulation process. Such tests provided the initial proof of concept for utilizing the
7 RLIA system with a frequency discrimination scheme where a remote network of several devices could be interrogated simultaneously and each individual device communicates detected information back to a base station relevant to its environment. Figure 1.4 : Concept of biased t ransponder (2 nd g eneration t r ansponder). 1.3 Diode Integrated Radar Detector In the final stage of this research a diode integrated radar detector is designed, fabricated and tested. Providing the functionality of the sensor element and reactance modulator as seen in Figure 1.2 the radar detector verifies feasibility of using a separate device concurrently with the 2 nd g eneration transponder to realize the full RLIA system. As seen in Figure 1.5, the radar detector is designed by modifying the 1 st generation transponder and replacing the transmitting 2.6 GHz patch antenna with a pad for a DC voltage connection. Additionally, a shunt capacitor is introduced to short the RF components of the signal, and a shunt resistor is included to maximize voltage sensitivity.
8 Being essentially a re ctenna, the radar detector senses microwaves occurring at a certain frequency within its local environment and transforms the microwave energy to a DC voltage proportional the strength of the signal impinging on the receiving antenna. The output of the rad ar detector could then be redirected to the bias input of the 2 nd g eneration transponder, where this DC voltage input would cause a change in conversion loss and modulate the retransmitted interrogation signal to the base station. When the base station rec eives the modulated interrogation signal using the lock in amplifier the information sensed by the radar detector will be extracted. However, when the radar detector is not sensing the presence of microwave energy at the frequency of interest it will simpl y have a zero DC voltage output, which will not affect the conversion efficiency of the transponder. Figure 1.5 : Concept of diode integrated radar d etector
9 1.4 Overview and Contributions of the Research The primary goals of this research were: Develop using standard testing equipment and the 1 st g eneration transponder the RLIA system, and demonstrate the capability to effectively communicate with the low power sensor. Design, fabricate and test the 2 nd g eneration transponder and verify modulation capabi lities using the same RLIA system. Design, fabricate and test a diode integrated radar detector to determine feasibility of employing such a device with the 2 nd g eneration transponder to effectively communicate information sensed with in the environment of the remote sensor node (Figure 1.2) Chapter Two presents background research related to rectennas and various applications and well as an introduction to the 1 st generation transponder (harmonic re radiator) used within the RLIA system. Chapter Three focu ses on characterization of the various components of the RLIA system and integration of the transponder within the system. The second part of Chapter Three presents the design and fabrication of the 2 nd generation transponder. Test results are presented to verify the feasibility of using the transponder in the RLIA system to implement a frequency discrimination communication scheme. Chapter Four presents the design and fabrication of a diode integrated radar detector to investigate feasibility of introducin g a separate sensing element and utilizing it concurrently with the 2 nd generation transponder. The ultimate goal in this research is to have the two devices comprise a remote sensor node and provide the initial proof of
10 concept where information occurring at a specific frequency is detected and this is communicated back to the interrogator node in the form of a modulated interrogation signal.
11 CHAPTER 2 BACKGROUND OF RECTENNAS, SIGNIFICANCE IN CURRENT WORK AND RELATED RESEARCH 2.1 Introduction Rectennas have assumed a pivotal role in the growing field of low power sensing and monitoring, where small passive devices are needed that fulfill cost, size, weight and power issues. Additionally, rectenna arrays show significant promise for the continued development of wireless power transfer, and the exploration of this concept within new arenas has demonstrated a potentially bold impact on future technologies. Within this chapter, a discussion of rectenna technologies is presented and some of the many applications are explored. Beginning with a brief look at the history of wireless microwave power transfer, this chapter investigates related applications to Solar Power Satellites (SPS), aircraft powering, and wireless power distribution syst ems in buildings. Rectenna array technologies for various applications are also discussed, including the different antenna polarizations employed, choices of diodes, and operational parameters. Applications within wireless sensor networks are then present ed, focusing on retrodirective array systems and their use within sensor networks, monitoring and sensing of infrastructure, and energy harvesting and recycling. Lastly, a Frequency Doubling Rectenna (FDR) used as a harmonic re radiator is described as ano ther application of rectenna technology for sensing purposes. It is the subject of further discussion
12 throughout this thesis and is redesigned and employed within a transponder based interrogation network scheme. 2.2 Diode Applications and Schottky Diode s Because a diode is a non linear circuit element, it will generate harmonics of a given sinusoidal input signal. According to Balanis, this non linearity may be exercised to provide useful applications of signal detection, demodulation, switching, frequency mu ltiplication, and oscillation [10 ]. Frequency conversion is a common application of the Schottky b arrier diode, and as noted in [11 ] encompasses signal detection (demodulation of an amplitude modulated signal), mixing (frequency shifting), and rectification (conversion of an AC to DC signal). As most wireless sensor devices perform at least one of the aforementioned tasks, a Schottky diode is a commonly used element in the design of such devices. For high frequency applications, a Schottky diode is preferable to the classic pn junction diode. As opposed to the pn junction diode, the Schottky diode consists of a semiconductor metal junction that maintains a much lower junction capacitance and forward voltage drop, providing better conversion effic iency.
13 2.3 Schottky Diode Characteristics and Application Specific Parameters As a common application of diodes, rectification is used for signal strength indication. As previously discussed, and also cited in [4 ], within low RF frequencies (kHz to lo wer MHz), both pn diodes and transistors are used as rectifiers. However, at microwave frequencies (1 GHz and higher), either GaAs or Si Schottky diodes are necessary, due to the required transit times. Within a rectifier application such devices convert a fraction of a RF signal to DC power. The result of this conversion is seen as an output of the diode in the form of a DC voltage proportional to the input signal. Let the diode voltage be equal to V=V 0 +v where the DC bias voltage is V 0 and the small AC signal voltage is v The diode current can be expressed in expanded Taylor series form as the sum of the DC bias current, I 0 and the AC current, i, as : (2.1) w here the bias current is I 0 =I(V 0 ) and G d is the dynamic conductance of the diode, where and R j is the junction resistance. In practice, the AC characteristics of a diode involve reactive effects due to the structure and packaging of the diode. Figure 2.1 shows an equivalent AC circuit model for a Schottky diode.
14 Figure 2.1: Equivalent AC circuit model for a Schottky diode. If the diode voltage is comprised of a sinusoidal small signal RF voltage and a DC bias voltage, it will be equal to: (2.2) Therefore, equation (2.1) can be rewritten as: (2.3) (2.4) w here I 0 is the bias current and is the DC rectified current. Additionally, and due to inherent non linearity of the diode element, the output of the diode will produce
15 0 0 and higher order. Often times, for the purposes of signal detection, harmonics produced by circui t elements will be removed by a simple low pass filter. However, in some applications of remote sensing, the production of these signal harmoni cs are put to use, as seen in [2 ]. Two important parameters in the application of signal detection for a diode ar e current sensitivity, and voltage sensitivity, Current sensitivity for a diode is defined as the change in DC output current due to a given RF power input from the received signal. Following from (2.1), using the first term in the sum, the RF power input to the diode is Additionally, (2.4) shows t hat the change in DC current will be Therefore, the current sensitivity can be defined as: (2.5) Voltage sensitivity, can be defined in terms of the voltage drop across the junction resistance when the diode is open circuited as: (2.6) Vol tage sensitivity is quantified in units of and a typical diode detector will range from 400 to 1500 as seen in [11 ].
16 An area of concern associated with the Schottky diode especially within the applications of detection and wireless power transfer is the loss occurring inside the element itself, which inevitably contributes to circuit loss and overall device efficiency. Such losses are associated with the voltage drop of the Schottky barrier and the series resistan ce of the diode. As noted in [12 ], a significant reduction in diode series resistance can be achieved during the design process of the substrate, active epitaxial region, and diode contacts. For some detecting applications, a diode is employed to demodulate an amplitude modu lated RF signal. In this case, the diode voltage can be expressed as: (2.7) where is the RF carrier frequency and is the modulated frequency. The output spectrum of the detected amplitude modulated signal can be se en in Figure 2.2, where the linear output terms of the diode voltage have frequencies of and
17 Figure 2.2: Spectral output of detected modulated signal Table 2.1 shows the frequencies and relative amplitudes of the terms of the diode output that are proportional to the square of the diode voltage. Square law behavior from the output of a diode detector is limited and can only be obtained within a certain region of input powers.
18 Table 2.1: Frequencies and relative amplitudes of a diode detecto r within the square law region of output. Outside of this region where input signal power levels are either too large or too small the square law region is not applicable. If the input signal power level is too large, the diode output will become saturated, approaching a linear and then constant i versus P characteristic. In the opposite regard, if the input signal power level is too small it will be lost within the noise floor, becoming undetectable. Figure 2.3 shows a typical voltage sensitivity curve for a diode. Within the square law region, In Frequency Relative Amplitude 0
19 applications where power levels are interpreted by the output voltage provided by the diode, the square law reg ion of operation is essential [11 ]. Some examples of diode application s of this type can be seen in , [14 ], and [4 ]. For the purposes of optimized diode sensitivity or to generate modulation characteristics, a bias voltage may be applied to the diode. At a certain operating voltage, a diode may either produce its maximum output or it may reduce its output by a sufficient Figure 2.3: Square law region of a diode detector As seen in [15 ], wherein a diode is utilized to generate the second harmonic of a received interrogation signal, maximum radi ation is achieved in the on state when a varactor diode is operating under reverse bias. Signal transmission is then reduced by approximately 10 dB when the applied bias changes to 0 V.
20 2.4 Wireless Power Transfer as an Application of Diode Rectification At the end of the nineteenth century Nicola Tesla conceived the idea of transmitting el ectrical power without wires [16 ]. He worked for years on a concept that would employ the Earth and Ionsphere as a giant capacitor, storing energy that would be distri buted by means of wireless power transmission. Unfortunately, due to a fateful after learning the project was designed for wireless power transmission and not for the purpose of telegraphy the work was abandoned. It is unknown if the project could have achieved the potential success Tesla believed it would. Interestingly, for the majority of the twentieth century, radio had been used primarily as a means of transmitting information and intelligence, seemingly being disregarded for the purposes of power transmission perhaps due to the lack of ne cessary microwave technology [17 ]. After the invention and initial development of microwave technologies, usage was predominantly limited to communications and radar. In the most elementary fashion, wireless power transmission can be defined as a three stage process. The first step involves conversion of DC electrical power to RF power. Secondly, the RF power is transmitted through space to specific location. The third step th en involves capturing the RF power through a reception device and converting it back to DC po wer through a receiver system [1 ]. Wireless power transmission has been employed in a vast array of applications including helicopter and airplane powe ring [17 ][18 ], solar powered satellite to Earth tran smission of microwave energy [19 ], nonlinear plasma wave excitation in the ionosphere b y high power microwave beams [17 ], and microwave power transmission for wireless power distribution systems in buildi ngs [20 ].
21 2. 4.1 Development of Microwave Power Transmission (MPT) and Conception of By the early 1960s, work done by R. H. George at Purdue showed that dense arrays of closely spaced diodes within an expanded waveguide could achieve as much as 40 W of DC power output from mic rowaves in the range 2 3 GHz [21 ]. Prior to this point, the semiconductor diode had been ignored as a microwave power rectifier due to its individual inherently low power handling capability. Consequently, this work demonstrated how power handling capability could be combined to produce reasonable amounts of DC power. However, one issue still to be resolved at that time was the problem relating to low RF power collection efficiency. As a solution to the limiting potential of the firs t diode array power rectifiers, the conce [1 ]. The proposed solution to the low efficiency of the prior work was to remove the individual full wave rectifiers from the waveguide, incorporate them with half wave dipoles, and int roduce a reflecting plane behind them [22 ]. In an effort to drive a proposed microwave powered helicopter experiment that later developed into a contract between Raytheon and the Air Force the development of the rectenna continued [ 1 ]. The outcome was the first rectenna array, conceived at Raytheon in 1963 and built and tested at Purdue. It was composed of 28 half wave dipoles, and each was terminated in a bridge rectifier comprised of four IN82G point c ontact, semiconductor diodes [23 ]. The development of a microwave beam powered helicopter at Raytheon accelerated the evolution of rectenna technology and was a pivotal experiment in the further study of free space microwave power transfer. The microwave powered helicopter was essentially a hovering platform with an electric motor, a propeller, and a 4
22 ft 2 rectenna array. The rectenna array was assembled from 4480 semiconductor IN82G diodes and had a maximum power output of 270 Watts. The microwave power source employed was a 5 kW magnetron that operated at 2 450 MHz. Combined capture and rectification efficiency of the rectenna array was significant, reaching levels of 55 percent, with rectification efficiencies of the diodes themselves being measured at 80 percent. Additionally, the helicopter flew continuous ly for an impressive 10 hours at an altitude of 50 feet. The helicopter device demonstrated an unmanned propulsion serve as a useful tool in the area of co mmunication and surveillance [18 ]. Following the success of the helicopter project, Brown (et al) published findings from continued research in power reception efficiency in free space with rectifier devices as seen in [12 ]. Rectennas were designed using GaAs Schot tky barrier diodes combined with rectifier circuit technology and input and output filtering, reaching 82 86 percent RF to DC efficiency the highest recorded RF to DC conversion effic iency according to Matsumoto [17 ]. A single diode half wave rectifier con figuration was the focus of the continued work, as it was believed to embody an agreeable compromise between efficiency, power handling capability, cost, and printed circuit adaptation. Additionally, GaAs diodes were specifically used because of their abil ity to individually supply 5 10 W of DC power output when used in a rectenna element. Further development of the rectenna was conducted with a focus on the design of low power density operation. As seen in [2 4], a 2.45 GHz rectenna was designed to absorb s mall amounts (as low as 50 mW) of microwave power. This rectenna had the capacity to absorb incident power density levels that were 10,000 times lower than
23 contemporary rectennas and convert the power into DC at levels that were still useful. To achieve th is, the standard rectenna design in which there was an individual rectifier in each microwave collecting element was abandoned. A broadside array was adopted as the collecting aperture and several advantages were realized, a few of which were: (1) the abil ity to filter out harmonics by introducing a filter element, (2) a reduction in the number of required diodes, (3) maximum efficiency realized from the diodes, and (4) the ability to employ the array as eit her a transmitter or receiver [2 4]. Continuing fo exp eriment, Matsumoto published [17 ], where a project named MILAX is described a fuel free microwave driven airplane equipped with lightweight microstrip rectennas. In order to control a target movin g both vertically and horizontally with a microwave power beam, a phased array transmitting antenna was designed. 2.4.2 MPT in So lar Power Satellites (SPS) and Related E xperiments Over the years, since the beginning of rectenna technology, research and ad vances have been made within the subject of rectennas used in free space microwave power transfer. One area of particular interest is solar power satellites (SPS). Solar power satellites, which are geostationary orbital stations, are characterized by havin g immense power capacity (5 GW). Such satellites generate electrical power through solar cells and transmit microwaves from the station to a site on Earth composed of rectenna arrays. SPS, a concept introduced by Dr. Peter Glaser of the Arthur D Little Company in 1968 [25 ], was a development that was predicted by Brown to have a profound effect on the future direction of wireless power transfer technology. Many key issues stemming from
24 the technology of SPS becam e apparent to Glaser, and in [26 ] he addre sses the economic and environmental considerations directly related to the microwave beaming utilized in MPT for SPS. In collaboration, Brown and Glaser organized the first technical session on SPS at the International Symposium of the International Microw ave Power Institute at the Hague Goldstone facility, the first long distance wireless power transfer experiment using SPS technology was demonstrated. Power was transferred by a microwave beam over a distance of 1 mile, with a DC output of 30 kW, utilizing a 288 ft 2 rectenna array [1 ]. It was a significant milestone in the progress of SPS technology, as the distance and amount of power transfer documented were almost two orders of magnitude greater than had been accomplished in prior experiments. In 1976 77 important electrical and mechanical improvements were made to rectenna technology through work supported by LeRC. One of the most significant changes was the move from a three plane system to a two plane system; this new format would eventually lead to thin film etched circuit format used in aerospace applications [ 1], [2 7]. Brown further focused his research and work of high powered rectennas toward SPS applications later pu blishing [28 ]. Through the years, since the conception of SPS, research has continued in this technology. In 1995, Shinohara and Matsumoto developed a new rectenna array using circular polarization. The rectenna utilized circular microstrip antennas (CMAs) and a a high receiving power application; rectenna input powers from 100 to 10,000 mW were utilized. Within a sub array configuration, the total number of rectenna e lements on the
25 array was 2,304; this gave a peak RF to DC efficiency of 64 percent wit h an input RF power of 2 4 W [19 Space W ireless ] where a target for MPT is acquired and maintaine d. The experiment involved a planar phased array antenna housed on the space shuttle bay that beams to the target rectenna on the free flying Wake Shield Facility (WSF) with a pilot guide signal beam. The WSF follows behind the shuttle, taking advantage of the vacuum than the most advanced laboratory vacuum chambers on Earth). This naturally occurring vacuum environment was shown to be useful for epitaxial thin film m aterial s processing in space [30 ]. The linearly polarized phased array antenna would transmit 1 kW of power a distance of 100 meters, allowing the rectenna on the WSF to prod uce a DC output power of 65W [29 ]. As part of ongoing research related to SPS and MPT, M atsumoto investigated possible nonlinear plasma wave excitation in the ionosphere caused by high power microwave beams through two rocket experiments: the Microwave Ionosphere Nonlinear Interaction Experiment (MINIX) and the International Space Year Microw ave Energy Tran smission in Space (ISY METS) [17 ]. These two experiments provided significant results on non linear wave interaction in ionospheric plasma caused by microwave power beams, as well as proving that MPT in space is possible. Losses through plas ma wave excitation were shown to be less than 1 percent of total transmitted power [ 31 ]. As of 2010, developments in microwave power transmission related to SPS have continued to evolve, particularly in Japan at the Research Institute for Sustainable Huma nosphere, Kyoto. Current research for MPT is focused on developing a phased
26 array power transmitter to control beam direction that is highly efficient and low cost. Cost and efficiency are paramount for the sake of establishing MPT commercial applications, and according to [ 32 ], are the reasons more commercial applications of MPT do not currently exist. A transmitting phased array for SPS and MPT is currently being developed at Kyoto University as a part of a government program. The phased array is composed of 256 elements, operating at 5.8 GHz, having an output power of 1.5 kW and a total DC to RF conversion efficiency of greater than 40 percent. Additionally, a receiver phased array is being developed with an efficiency of greater than 50 percent and an ou tput power of 0.1 mW for each individual rectenna element. The purposes of advancement are not limited to SPS, they are in part due to the anticipation of future commercial MPT applications, such as ubiquitous power sources for wireless cell phone charging MPT for wireless power distribution systems in buildings [ 20 ], and wireless charging of electric vehicles [ 32 ]. However, there are still obstacles to overcome before MPT can be considered for commercial applications. One important issue is frequency regu lation in an already crowded spectrum, especially within the ISM band. Another concern, according to [ 20 ] is the shortage of effective and economical MPT applications. Most research in the realm of MPT has been single point to single point transmission; ho wever there are few recognized advantages to MPT over wired power transmission in this regard.
27 2.4.3 MPT for Wireless Power Distribution Systems in Buildings With respect to a single point to multiple point configuration, especially with low power devic es, there is much promise for MPT. Devices such as ICs, sensors and RF IDs that operate within the microwatt range are reasonably applicable in MPT, where a single point source could power many devices at one time. Buildings, being closed area environments where high power MPT systems can operate within current frequency regulations, provide many opportunities for wireless power transfer. Shinohara, et al, have proposed an indoor wireless power distribution system (WPDS), whereby power is distributed wirel essly using existing building components for example, deck plates. Within this system, such components are utilized as microwave transmission waveguides; the microwaves propagate within the medi um and there is no radiation [20 ]. The use of existing buildin g components reduces the initial cost of construction for automated buildings. Additionally, rectennas are placed as DC outlets on the floor and can be moved practically anywhere since microwaves exist everywhere within the floor. The proposed system could frequently used, such as computers, fax and copy machines, refrigerators, etc. It has been estimated with WPDS that one DC outlet needs < 50W and within this proposed work a single ro om has been provided with microwave power >3kW, at an operating frequency of 2.45 GHz. For the purposes of this project, a highly efficient rectenna is designed using Schottky barrier diodes. To achieve a rectification of 100W, a 64 way power divider confi guration is used with 256 total diodes. The rectenna has provided 55 percent RF to DC conversion efficiency with an input of 100W, at a physical size of 125 mm x 110 mm, x 95mm [ 20 ]. Several companies within the United States have recognized the
28 potential of MPT and have taken an active approac h to development in this area , , [3 5]. MPT applications of this kind display a very promising future in many areas of automation, and could evolve to become a source of immense energy conservation through in creased usage efficiency. Many applications of wireless power transfer with rectennas have utilized a variety of antenna polarization techniques and power density operations. Linear polarization was used for l ow power density rectennas in [2 4], which prov ed to be useful for transponders in a sensor or communications system in which the interrogation signal also supplies power to the transponder. Additionally, lin ear polarization was seen in [12 ] for the purpose of in space wireless energy transmission. Dual pol arized antennas were seen in  and [26 ], useful for low power signal detection and operation of mechanical actuators. Circularly polarized receiving antennas were employed for the purposes of , , and [37 ]. These polarization techniques w ere shown to be effective for microwave power transmission within SPS and distributed monitoring sensor systems within infrastructure, with efficiencies as high as 78 percent. Low power density operation w as the focus of , [14 ], and [4 ], and it was pro ven to be successful within the realm of wireless powering of low power indoor sensor networks and remote sensors, as well as recycling of ambient microwave energy.
29 2.5 Rectennas and Remote Wireless Sensing Networks As previously investigated, diode integrated rectennas have been shown to be useful within many applications of wireless power. However, these devices are instrumental within yet a broader scope of applications one important application being sensing, monitoring, and communication within r emote wireless networks. The spectrum of monitoring and sensing is extensive, including: temperature and humidity sensing, light and laser detection, chemical and biological sensing, sound and vibration detection, and radio/IED/radar detection. Rectennas p lay a key role within remote sensor networks not only by providing power to the system; they are an essential element to communicating collected information within a sensor network back to a data collection device. 2.5.1 Rectennas and Retrodirective Array Systems Research on rectenna arrays has continued within this important field, and numerous approaches have been taken to find fitting solutions to current obstacles and to strea mlining existing systems. In [38 ], an adap tive power controllable 5.8 GHz Retrodirective Array System is proposed, which is useful for wireless sensor servers that behave as access points between wireless sensors and remote data collectors. A Retrodirective Array System is described as being able to respond to an incoming signal beam such as monitoring, data collection, in space a pplications, etc. Within the array system is an integrated rectenna and analog switch that controls a battery source. It is able to
30 conserve battery power in idle mode and wake when operation is necessary, thus prolonging its lifetime. An incoming RF signa l wakes the system by accumulating power at the input that is converted to DC voltage, thereafter activating a switch and battery. The arriving signal is split between the rectenna and receiver, where the majority of the power is directed to the rectenna. In the absence of an interrogation signal, the switch remains op en and the system is off. In [39 ] a 60 GHz Retrodirective Array System is developed for multimedia sensor server applications. This system streamlines battery efficiency of remote sensors thro ugh an incorporated interrogation scheme that uses required usage times. When the received power is above 0.013 mW/cm 2 the system powers on and operates as a Retrodirec tive Array. In addition to streamlining power efficiency, the proposed system uses a microwave frequency of 60 GHz to be compatible with higher data rate requirements, which are often needed in certain remote monitoring applications. 2.5.2 Rectennas for Monitoring and Sensing of Infrastructure The life span and current conditions of the reinforcement of infrastructures such as buildings, roads, and bridges and is of major concern for public safety in modern times. Developing sensor network technologies c an be significantly beneficial in helping assess the continuously changing conditions of infrastructure. For example, miniaturized embedded sensors can monitor steel reinforcement tendons inside the concrete covering of a bridge, and can be deployed during the construction process, as discussed in [37 ]. It has been shown that detecting cracks or deformations in infrastructures can be achieved
31 through the use of embedded active acoustic wave transducers operating within the ISM frequency band [3 ]. However, t here is a need for an intermediate technology to process and transmit the collected information from these sensors to an appropriate destination, as well as power the embedded devices. A mobile test unit using an on board r ectenna has been proposed in [37 ] installed supervisory control sensors that will be in the vicinity of such embedded sensors, and these control sensors will perform the function of data transmission and powering. Circularly polarized antennas within the rectenna devices are again recommended as a means of obtaining constant DC voltage regardl ess of rectenna rotation. In [37 ], a rectenna featuring a circularly polarized microstrip patch antenna is proposed, which c an provide wireless battery powering at 5.5 GHz and data delivery within the 5.15 5.35 GHz WLAN band. 2.5.3 Energy Harvesting and Recycling for Use in Sensor Networks It is very probable that the future will see a ubiquitous presence of remote sensors, beyond the aforementioned infrastructure applications. Areas such as intelligent office spaces, medical monitoring, surveillance systems, military, construction and manufacturing, and even agriculture will very likely become enveloped with the technology i n the coming decades. Along with this expected presence comes the need to efficiently harness and deliver the energy required by low powered sensor networks. There is much research in the field of energy harvesting and energy recycling utilizing rectenna t echnology for these purposes.
32 RFID is a popular technology for sensor networks and should continue to be in the future. For the purposes of powering batteryless RFID sensors, a low profile planar rectenna has been proposed which scavenges ambient RF power from the surroundings and delivers to senso rs within a wireless network [40 ]. The described rectenna technology is integrated with the RFID sensors themselves and operates at 2.4 GHz. A two stage Dickson zero bias Schottky rectifier circuit with a mi niaturized antenna provides a RF to DC conversion efficiency of approximately 54 percent at an input power of 10 dBm. Recycling of ambient microwave energy with broadband rectenna arrays has been described in [4 ], utilizing low power density microwave radi ation. The purposes of this work are also focused on the wireless powering of industrial sensors within a network structure. A 64 element dual circularly polarized spiral rectenna array was designed and characterized over a frequency range of 2 18 GHz. Mea surement of the rectenna was made with broadside linear polarized radiation at 3 GHz with an incident power density ranging from tens of nW/cm 2 to 0.1 mW/cm 2 Rectification efficiency was seen to reach the 20 percent range for an incident power density of 0.1 mW/cm 2 with arbitrary polarizat ion [4 ]. 2.6 Wireless Passive Surface Acoustic Wave Sensors Another form of passive technology shown to be useful in the realm of sensing is surface acoustic wave (SAW) technology. By exploiting the piezoelectric effec t of the material on which they are fabricated on (quartz, lithium niobate, lanthanum gallium silicate, etc), SAW sensors convert acoustic waves to electronic signals. The conversion is typically achieved through an inter digita l transducer within the devi ce. Much research
33 within the subject of SAWs has been conducted within the past twenty years [41 ]. Sensing applications related to many areas such as optical, thermal, biological, chemical, and pressure have seen developments. In particular, passive wirele ss SAW sensors have shown to be very useful for in space applications involving structural health monitoring (SHM) of space vehicles as seen in [42 ]. In an area where constant close range monitoring of the physical structure of space vehicles is crucial, passive SAW sensors offer many advantages. Wireless passive SAW sensors offer a solution to many cost, mass, volume, and power constraints within SHM instrumentation on a space vehicle, and often eliminate the need for heavy wiring and battery requirements associated with wired SHM systems. SAW sensors also operate more robustly within a wider range of extreme temperatures often experienced in an in space environment. In addition to in s pace vehicular applications, [43 ] explores the use of SAW sensors in rough environments related to motor vehicles. In such environments, SAW sensors have proven to be useful in providing remote readouts to computers located within the vehicle in severe environments that other types of se nsors cannot withstand, for example, within heat, dirt, mechanical vibration, or electromagnetic interference. Interrogable SAW sensors used within a network have al so been explored. As seen in [44 ], a technique to allow multiple sensors to operate simulta neously was developed and demonstrated using a network analyzer and sine pulse interrogation. NASA is continuing to sponsor the research and development of wireless passive SAW sensors due to their inherent environmental robustness and ability to operate well in multi sensor networks.
34 2.7 Rectennas U sing Harmonic Re R adiation for Sensing Applications Within many remote wireless network applications, monitoring and sensing is made possible by integrating diodes into low power wireless devices that operate by modifying and re radiating a received interrogation signal. In [1 ] a high efficiency diode doubl er also named a frequency doubling reflectenna (FDR) is implemented and is shown to be an efficient model for a remote sensor. Within this described work, a modulated form of the second harmonic of a received interrogation signal serves as a return communi cation signal from the remote sensor within a network. In field application, the return communication signal from a remote sensor may contain described in [2 ] is intended for sensing applications and utilizes frequency discrimination to enhance the ability of an interrogating transceiver base station to detect a signal of very low amplitude. Through the employment of a diode, harmonic forms of the incoming interrogation signal at 1.3 GHz are produced and a single harmonic at 2.6 GHz is retransmitted. The topology of the FDR is presented in Figure 2.4.
35 Figure 2.4: Layout of Frequency Doublin g Reflectenna t ransponder [2 ] A GaAs Schottky diode is chosen for the doubling element, as this type of diode has proven to perform with high conversion efficiencies at microwave frequencies. The transponder also consists of two quarter wavelength, shorted microstrip patch antennas, an d matching stubs to provide a conjugate match to the diode. The matching elements associated with the antennas provide a source impedance that is very close to the complex conjugate of the input impedance to the diode, thus minimizing conversion loss and m aximizing device efficiency. It is reported in [2 ] that doubler conversion efficiency of 1% was achieved at 30 dBm input power level. As previously discussed, a bias can improve conversion efficiency and provide modulation characteristics, however for the described work, in the interest of maintaining a simple design, the bias has been ne glected. Figure 2.5 shows relative measured power output levels from the tag according the respective frequencies of operation. This will be further discussed and comparisons will be made in Chapter 3, with a description of how the FDR was incorporated int o the work of this thesis. Figure 2.5 shows the varying output power (at 2f 0 ) versus input power
36 (at f 0 in this work, which is closest to 1.291 GHz operating frequency. Figure 2.6 shows changing conversion gain and DC current over a varied input power to the FDR of 0 to 50 dBm. The proposed harmonic re radiating rectenna device could be useful for quasi optical applications, energy scavenging, and monitoring and sensing [2 ] Figure 2.5: Received power vs. frequency for 3 FDR devices
37 Figure 2.6: Conversion g ain vs. frequency for 3 FDR devices 2.8 Conclusions In summary, a rectenna was described as a diode and antenna integrated device, useful in a broad spectrum of low and high power applications. A background of diode theory was presented, as the Schottky diode is a key component to a rectenna device. Rectenna and rectenna array application technologies have been explored, from wireless microwave power transfer to remo te wireless sensor networks. Much of the ongoing research within rectenna technologies exhibits promise for continued improvements in communications, safety, and energy conservation.
38 CHAPTER 3 REMOTE INTERROGATION SYSTEMS WITH HARMONI C RE RADIATION 3.1 Introduction Although the concept of microwave detection has existed since WWII, when radar was the first major application of microwave technology, significant advances still continu e to be made within the field [11 ]. Such advances are especially not iceable within the realm of integrated devices for sensor communication networks. The issues of cost, size, and efficiency continue to be addressed through the developments of higher efficiency m iniaturized devices, such as [45 ]. For particular sensing net works involving remote communication devices, enhancements in size and efficiency of devices at times is not sufficient for detectability of low strength signals. To assist with the task of measuring low strength signals, lock in amplifiers have been emplo yed. The goal of the work described within this chapter was to incorporate the harmonic re radiator from [2 ] which possessed significant size and operating efficiencies into a unique oach utilizes lock in amplification in a remote connection scheme that separates the heavy electronics from the sensor. The interrogation equipment that require more power for operation are located at a mobile base station platform, allowing the sensor to operate
39 3.2 Introduction to Lock In Amplifiers One notable contributor to the evolution of signal detection is the lock in amplifier. Lock in amplifiers are devices used to detect extremely small AC signals reaching as low as 100 dBm or more. It is not uncommon for lock in amplifiers to operate with s ensitivities well into the nano volt range, where signals may be detected in the presence of an excessive amount of noise, which would otherwise obscure a signal. These devices have proven to be useful in many different measurement applications including: complex impedance measurement for nano wire gas se nsing [6 ], accurate measurement of modulat ed scattered electric fields [7 ], and measurement of open path scattered light from fine particles such as diesel particulate emissions [8 ]. The technique employed by lock in amplifiers to perform such tasks is referred to reference frequency. Noise signals at all other frequencies are rejected by the lock in amplifier and thus d o not contribute to the measurement. According to Fourier's theorem, representation of a signal in the frequency domain is possible as a sum of many sine waves occurring at different amplit udes, frequencies and phases [46 ]. Time domain representation of si gnals, as in the case of normal oscilloscopes, often does not bear enough information about the various frequencies that make up a signal un less it is a clean sine wave [9 ]. However, a lock in amplifier avoids this by way of multiplying each component of t he input signal with a pure sine wave at the reference frequency contemporaneously. In general, unless two frequencies are identical, the average of the product of the two sine waves will be zero. The multiplication done by the lock in amplifier produces a DC signal proportional to the component of the received signal that
40 has a frequency exactly locked to the reference frequency. Following the multiplication that occurs, a low pass filter removes the components at all other frequencies other than the refer ence frequency. 3.2.1 Phase Sensitive Detection The process of Phase Sensitive Detection utilized by lock in amplifiers essentially deciphers an experimental signal from noise. The measurements performed by lock in amplifiers require the input of a refer ence frequency. The experiment being performed is stimulated at this fixed reference frequency via either a function generator or oscillator, and the lock in amplifier detects the output response of the experiment, which will contain the desired signal inf ormatio n at the reference frequency [9 ]. All other frequencies are unnoticeable to the lock in amplifier. As shown in Figure 3.1 the reference signal, provided by the function generator, is a square wave. This signal is directed into the lock in amplifier so as to provide the reference according to the
41 Figure 3.1: Phase sensitive detection performed by a lock in amplifier Lock in amplifiers employ a Phase Locked Loop (PLL) fixed on the external reference, which in turn generates an internal reference signal. Figure 3.1 shows the external reference square wave, the experimental signal, and the internal reference signal of the loc k in amplifier. According to [9 ], the internal reference signal is defined as: (3.1) The experimental signal is amplified and multiplied by the reference signal through a phase sensitive detector (PSD) or multiplier. The output of the PSD provides a signal that is equal to the product of the tw o sine waves, as shown below [9 ]:
42 (3.2) thus, The output of the PSD consists of two AC signals, one existing at the sum frequenc y of and the other at the difference frequency of When the output of the PSD is subsequently passed through a low pass filter, AC signals will be removed. If the difference freq uency will be a DC signal proportional to the signal am plitude and will be equal to [9 ]: (3.4) With traditional analog lock in amplifiers, the reference and experimental signals are analog voltage signals. The two are multiplied in an analog multiplier and passed through at least one stage of RC filters. The lock in amplifier could not successfully detect an experimental signal at the reference frequency without the use of narrow band detection. Since the input to the lock in amplifier consists of signal plus noise, it is important for it to have a low pass filter with a very narrow bandwidth. The PSD and
43 low pass filter are limited to detecting signals very close to the reference frequency. This causes noise signals that occur at frequencies far from the reference to be attenuated and is a key concept in the operat ion of the lock in amplifier [9 ]. If noise signals occurring at frequencies very close to the reference frequency exist the output of the PSD will observe very low frequency AC signals. The strength of these leaked noise signals depends on the roll off and bandwidth of the low pass filter. The bandwidth of the low pass filter ultimately determines the bandwidth of detecti on for the lock in amplifier; a wider bandwidth will allow some of these noise signals to pass and a smaller bandwidth will attenuate the noise signals close to the reference. The experimental signal (the only signal occurring at the exact reference freque ncy) will be unaffected by the low pass filter, as it will be a true DC signal, and this is the sign al of interest for measuring [9 ]. Another necessary component of the lock in amplifier is the PLL. For proper measurement of the experimental signal, it is required that the reference and the experimental frequencies be equal and have equal phases unchanging with time. A change in the two phases will cause to change, and subsequently will not be a DC signal. Therefore, t he lock in amplifier utilizes a PLL to lock the internal reference oscillator to the external reference, providing a consistent reference sine wave at with a fixed phase shift of This feature ensures that a change in the external referen ce frequency will not disturb the measurements [9 ].
44 3.3 Detector and Lock In Amplifier Characterization In Chapter Two a discussion of the characteristics of the Schottky barrier diode was presented. Another important characteristic of a diode detector, particularly when an application requires measurement of extremely low power signals, is Tangential Sen sitivity (TSS). According to [47 ], the TSS of a diode detector is the lowest signal power level for which the detector will have a specified si gnal to noise ratio at its output. Measurements of TSS along with voltage sensitivity are useful in radiome ter applications. As seen in [48 ], TSS measurements can be performed to estimate noise equivalent power (NEP) of the detector itself. In this TSS mea surement procedure, input power levels are set such that the value where the noise peaks without an RF signal equals the lowest noise with an RF signal. Then using an oscilloscope and video amplifier the actual value of TSS can be measured. This type of pr ocedure could enhance voltage sensitivity of a measurement system where the absolute highest sensitivity values are critical. In characterization phase for the coaxial detector and lock in amplifier, various testing equipment was incorporated into a bench top configuration and measurements were performed. This was a necessary step in the progression of the RLIA concept since future tests would require a reference for measurement verification purposes. The block diagram in Figure 3.2 shows the equipment used in the configuration: Standard Research Systems Model SR530 lock in amplifier, Narda Zero Bias Crystal detector Model 4503 01, variable attenuator, standard signal generator, and standard function generator. The documented measurements constituted the eff ective sensitivity of the detector and verified proper function of the lock in amplifier. A voltage sensitivity curve was
45 constructed from the results, and the Narda detector data sheet was used as a reference to verify the sens itivity measurements (Append ix A ). Figure 3.2: Block diagram of detector c haracterization U sing Lock In Amplifier The role of the signal generator in the bench top configuration is to provide the fundamental signal source (1.3 GHz) to the detector through a variable attenuator, which is then fed to the lock in amplifier. The function generator provides the reference signal for the lock in amplifier (1 kHz square wave, 100 mV) as well as the chopping frequency (1 kHz sine wave, 100 mV), which is the imparted modulation onto the so urce signal. Without both the reference signal and the modulating chopping frequency, the lock in amplifier is unable to perform its function. Between the signal generator and the detector is the variable attenuator, emulating channel loss. The presence o f the variable attenuator in the configuration provides a way to consistently reduce the signal power level, imitating the effects of free space path loss, while keeping the source at a maximum signal output level. In order to have a reliable test environ ment, the performance of the standard equipment was verified. Before constructing the detector/lock in characterization set up,
46 initial measurements were performed on the signal generator using a power meter to verify the output power and obtain calibrated signal values. The signal generator was programmed to give a continuous wave output of 1.3 GHz at an amplitude of 0 dBm, then the output was measured by the power meter with the variable attenuator attached. Calibration measurements were performed in order to establish a 0 dBm reference point including losses from cables and the variable attenuator. This reference point was achieved with a 2.0 dBm output from the signal generator. The coaxial detector was then introduced into the circuit at the output of the variable attenuator, in series with a multimeter (which was later replaced by the lock in amplifier). With t his configuration it was possible to compare the dc voltage output of the detector to the known calibrated power input to the detector. Attenuation was decreased in increments of 1dBm and the output voltages (mV) from the detector were recorded as seen in Figure 3.3 (Table I, Appendix A).
47 Figure 3. 3: Measured voltage sensitivity of Narda d etector (Model 4503 01) using Lock In Amplifier 0.05 0.5 5 50 500 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 5 10 15 20 Vdc Output of Detector (mV) Calibrated Input Power (dBm)
48 The signal source was set to its maximum power output of 14.5 dBm (a calibrated power output of 11.58 dBm accountin g for losses) and the variable attenuator was increased in 1 dB steps. The measured signal voltage at the output of the detector vs. input power is shown in Figure 3.3. Stable measurements were acquired from 11.58 dBm to ~ 50 dBm, at which point fluctuati ons in voltage output were seen, indicating the TSS threshold for the detector. 3.4 Background Work on Sensor Node Harmonic Repeaters 1 st G eneration Transponder As previously introduced in Chapter Two, a high efficiency frequency doubling transponder was developed for low power sensing applications [2 ] Using the principles of frequency multiplication, this device receives an interrogation signal at 1.3 GHz and re transmits the harmonic at 2.6 GHz. This transponder is used to verify the RLIA concept and is considered the 1 st generation device. The printed circuit board design consists of two quarter wavelength patch antennas (1.3 GHz and 2.6 GHz) and a diode based d oubler. Each antenna is operated near its respective resonant frequency, where the impedance changes rapidly as a function of frequency. By this and conjugate matched conditions on both sides of the multiplier (Figure 3.4), a bandwidth of less than 0.5%, i s achieved. In order to provide an intrinsic match to the diode doubler impedance under very small signal conditions, the transponder antennas are designed such that the the transmit patch antenna is ~40 conversion gain vs. input power for the 1 st generation transponder at 1.3 GHz is shown in
49 Figure 3.5. When properly functioning, the 1 st generation transponder will have a conver sion gain of approximately 20 dB at an input power level of 30 dBm, and approximately 15 dB at an input power level of 20 dBm (Figure 3.6). Peak multiplier efficiency of ~2% at 30 dBm input power was seen as in the figure, and this was achieved over a frequency sweep of 1.27 GHz to 1.305 GHz. Efficiency is also seen to remain relatively flat over a 30 40 dB input power range [2 ] Figure 3.4: 1 st generation transponder layout: Z S and Z L are the input and output impedances of the diode. Z S and Z L are the source antenna input impedance and load antenna input impedance, respectively [ 2 ] Figure 3.5: 1 st generation transponder expected and measured conversion gain vs. input power at 1.3 GHz.
50 Figure 3.6: 1 st generation transponder conversion gain vs. input power As cited in Chapter Two, other diode integrated rectenna designs have shown greater RF to DC conversion efficiencies, however, this is in the case of much higher transmit powers (wireless power transfer for example). 3.5 1 st G eneration Transponder within RLIA System The following section presents the configuration of the 1 st generation transponder within the RLIA system, measurements and analysis. The first major test involving the full RLIA system focused on demonstrating t he capability to effectively communicate with the low power remote transponder and to verify system link budgets. This name can be used due to the fact that the device itself requires no additional power to operate, other than that carried within the radio signal itself. This is an important feature for many remote sensing schemes where long life with minimal maintenance is a concern. The 1 st
51 generation transponder demonstrated proof of concept by processing relevant information conditions. The full RLIA test set up was configured so that when an interrogation signal is transmitted at a particular frequency, a return signal is received from the transponder at double the frequency with the sensor signal modulated onto the carrier. The full test configuration was performed with a distance of 1.5 meters and is shown in Figure 3.7. The transmitter block is comprised of a 1.3 GHz signal source, a function generator that modulates the 1.3 GHz signal, and a transmit antenna with a gain o f 5.05 dBi. The function generator provides a chopping frequency (sine wave form at a frequency of 1 kHz with amplitude of 100mV pp ) that is imparted onto the transmitted signal. The signal arriving to the receiver block may be very weak or may be accompani ed by the presence of an overwhelming amount of noise; therefore the chopping frequency will assist the lock in amplifier in locating the signal components which contain the desired information. The receiver block consists of the necessary components to pr operly filter and coherently demodulate the retransmitted signal from the transponder. It is comprised of a receive antenna with a gain of 7.08 dBi, a variable attenuator (to incrementally reduce signal strength at a constant rate), filter and amplificatio n stages, a diode detector and a lock in amplifier (Figure 3.7).
52 Figure 3.7: Block diagram of full Remote Lock In Amplifier interrogation system
53 As aforementioned, coherent demodulation with a chopping and reference frequency is an intrinsic feature that is key to the function of the lock in amplifier. However, when a detector is introduced and used concomitantly with the lock in, it is important that the signal reaching the input of the detector be free of undesired interference signals. Sinc e the diode detector outputs a particular DC voltage that is directly proportional to its input power, ideally the input signal to the detector should be solely comprised of the desired information signal. If interference signals (for example ambient signa ls from WLAN, cellular, microwave ovens, etc.) at significant power levels are allowed to pass through the system and impinge on the detector, inaccurate voltage readings at the output of the detector will result. In light of this, sufficient filtering is crucial including a possible down conversion stage to produce a signal that is free of interference. The RLIA system is equipped with filters on both transmit and receive blocks and a down conversion stage yielding an intermediate frequency of ~60 MHz. Th e down conversion stage facilitates removal of strong interference signals within the 1.8 to 2.8 GHz band. End to end interrogator receiver gain excluding the antenna is 75 dB. Link budgets based upon the configuration in Figure 3.7 are listed in Tables 3. 1 and 3.2. The difference between the two budgets is the interrogator transmitted power, which is 5.00 dBm and 6.02 dBm respectively. In the budget in Table 3.1 (a) the transmitted power at 1.3 GHz is 3.16 mW, which yields an input power of 29.19 dBm to the transponder. At this input power level the harmonic transceiver conversion efficiency is approximately 20 dB. The budget in Table 3.1(b) consists of a transmit power, input power to the transponder, and harmonic conversion efficiency of 0.25 mW, 40.2 1 dBm, and 30 dB
54 respectively. In the first part of testing the RLIA configuration, a spectrum analyzer was employed; therefore a system bandwidth of 10 kHz is reported in both budgets that correspond to spectrum analyzer settings. Measurements were cond ucted in an open lab environment and a comparison between calculated and measured output power is given in the bottom two lines of each budget table. As seen in the tables, for transmit powers of 3.16 mW and 0.25 mW the difference in measured and calculate d output powers is 1.12 dB and 4.27 dB respectively. These differences could be attributed to additional multipath occurring during experimentation or small interference signals leaking into the system. Table 3.1 (a) Link budgets for RLIA system at 3.16 mW and (b) 0.25 mW transmit powers respectively (a) (b)
55 The second test portion of the RLIA system involved using both a spectrum analyzer and a lock in amplifier to detect the received signal measured over a swep t transmit power range. For these results the distance from the interrogator to the sensor node was again 1.5 m. Figure 3.8(a) shows the results using a spectrum analyzer, where a minimum detectable power of ~ 55 dBm is demonstrated. To verify the output system noise floor, first the interrogator noise floor was calculated according to kTB, where k is 23 J/ K), T is temperature in degrees kelvin ( k = 0.599 ), and B is the bandwidth of the system, which is taken to be th e bandwidth of the spectrum analyzer ( B = 10 MHz). I t is then possible to take the interrogator noise floor value from Tables 3.1 (a) and (b) ( 130.83 dBm) and add the receiver block gain of 75 dB, arriving at a value of 55.83 dBm. Figure 3.8(b) shows the measured results of testing with a lock in amplifier in place of the coaxial diode detector and signal analyzer combination. In this testing configuration, minimum detectable signal levels were seen approximately 5 dB below the configuration in Figure 3.8(a). It can be deduced that the detectable signal using the lock in amplifier is at least 5 dB into the system noise floor, since conversion efficiency of the harmonic transceiver decreases at least linearly with input power, as seen in Figure 3.6.
56 (a) (b) Figure 3.8: (a) Measured performance of the system at an interrogator to sensor distance of 1.5 m using a spectrum analyzer; (b) lock in amplifier 3.6 Design of 2 nd G eneration Transponder As the RLIA scheme is further developed, a new version of the transponder is introduced. The existing zero bias transponder is modified to include a DC bias line with a series inductor and shunt capacitor to block any RF leakage. An applied bias voltage to the diode will directly change the impedance match between the receiving 1.3 GHz antenna and the doubler causing a change in conversion loss. This change in conversion loss in duces an amplitude modulating effect on the retransmission of the signal from the transponder. The 2 nd generation transponder as seen in Figure 3.9 was fabricated on Taconic, 60 mil (Er=6.15) copper substrate and utilizes the shown Coilcraft inductors (040 2) and Johanson capacitors (0402).
57 Figure 3.9: Layout of 2 nd generation transponder Referring to Figure 3.6, as input power to the transponder drops below 30 dBm, conversion efficiency subsequently drops and the same occurs when input power to the transponder reaches above 20 dBm. Simply increasing the power level of the transmit signal does not convert to a stronger retransmitted signal from the transponder. Conclusively, the maximum power the transponder is able to retransmit back to the receiver is approximately 40 dBm ( 20 dBm 20 dB [conversion loss]). With the best possible conversion efficiency in mind, testing of the 2 nd generation transponder was performed with an incident power to the transponder within the range of 20 dBm to 30 dBm. In determining the appropriate power level required from the signal generator to result in an incident power level to the transponder within this range, the proper distance between the 1.3 GHz antenna and the transponder needed to be calculated. To achieve a transponder input power within the range of 20 to 30 dBm it was necessary to fix the source at its maximum level of 14.5 dBm. A link budget of the testing configuration of the 2 nd generation transponder can be seen in Figure 3.10 (Table I I, Appendix A ). The
58 free space path loss computation method was used to determine that a transmitted signal of ~12 dBm at a distance of 4 feet would yield an incident input power to the transponder at 1.3 GHz within the optimal range. The receive antenna gain of the tr ansponder itself is 0 dB, therefore the receive power of the transponder is 24.3 dBm, as seen in Figure 3.10. Figure 3.10: 2 nd generation transponder testing configuration In characterizing the transponder, the first test was to identify the optimal transmit frequency, which would translate to the highest output power and conversion efficiency from the device. The source amplitude was kept at a constant 12.2 dBm, and frequen cy was swept from 1265 to 1315 MHz. The transponder bias was set to a fixed value of 0 V. Measurement results are shown in Figures 3.11 and 3.12, where it can be seen that optimal transmit frequencies and conversion efficiencies occur at 1275 MHz and 1300 MHz.
59 Figure 3.11: P out of transponder at 0 V bias over frequency sweep Figure 3.12: Conversion l oss of transponder at 0 V bias over frequency sweep
60 Referring to Figures 3.6 and 3.12 and comparing the 1 st and 2 nd generation transponders, it can be observed that both devices exhibit similar conversion loss behavior at input powers around 24 dBm. The next test in the characterization of the 2 nd generation transponder involved fixing the bias and input power level a nd sweeping the frequency to find the peak retransmitted signal power level, as was done in the previous test; however, in this test the bias voltage was fixed at +/ 0.05 V and then at +/ 0.15 V. The results of this test determined whether a change in bi as voltage would cause a shift in frequency. This information is critical in the case of employing a frequency diversity scheme for interrogating sensors in the field. It is imperative to know the exact spectral spacing required for two consecutive sensors to operate without interfering with each other. If a change in bias voltage did in fact cause a shift in frequency, this would cause a single transponder to experience a shifting peak; and without knowing the amount of frequency shift for a single sensor, proper spectral spacing may not be obtained. The consequence of this being that when interrogation is in progress, it would be impossible to determine if a retransmitted signal is being produced by one single transponder, or if another transponder close i n frequency is actually producing the signal. As in the former test, the transmit frequency was swept from 1265 to 1315 MHz in steps of 5 MHz. Input power was fixed at the maximum 12.2 dBm to achieve a 24.3 dBm input level and, as seen in Figures 3.13 and 3.14, P out of the transponder is provided according to the respective frequencies.
61 Figure 3.13: P out of transponder at +/ 0.05 V bias over frequency sweep Figure 3.14: P out of transponder at +/ 0.15 V bias over frequency sweep
62 Figures 3.13 and 3.14 reveal the results of changing the applied bias from +/ 0.05 mV to +/ 150 mV. With a zero and +/ 50 mV bias the peak frequency remains at 1275 MHz, however when the applied bias is changed to +/ 150 mV a shift in the peak frequency to 1300 MHz is observed. It can be concluded from the behavior of the transponder that if a frequency discrimination scheme were to be implemented using this and other like devices, frequencies of operation may be desired (in this case, 1275 MHz) where bot h maximum and minimum output powers can be realized with a changing bias. The reason for this is that for modulation to occur for one particular transponder, observed tha t a zero bias and +/ 50 mV biases at 1275 MHz allowed signal transmission at a much higher output power and a +/ 150 mV bias at the same frequency caused the signal power to decrease by 15 dB. A comparison of the performance characteristics of the 2 nd g eneration transponder in a frequency sweep from 1265 MHz to 1315 MHz is shown in Figure 3.15 ( output power) and Figure 3.16 ( conversion efficiency); tabulated values are listed in Tables 3.2 and 3.3 The 0 V and +/ 50 mV bias settings exhibit relatively c onstant trends in conversion efficiency behavior, each providing maximum conversion efficiency at 1275 MHz. The +/ 150 mV bias settings, however, yield output powers from the transponder that are approximately 15 dB below that of the 0 V and +/ 50 mV bia s settings. These combined results again verify feasibility of using the 2 nd generation transponder for modulation purposes. A change in power output of 15 dB caused from only a 0.1 V
63 Figure 3.15: Comparison of P out of transponder at each t ested bias over frequency sweep Table 3.2 : Comparison of P out of transponder at each tested bias over frequency sweep
64 Figure 3.16: Comparison of conversion loss for each transponder at each tested bias over frequency sweep Table 3.3 : Comparison of conversion loss for each transponder at each tested bias over frequency sweep
65 In the previous tests the transponder bias voltage was held constant while the transmitted frequency was swept. Measurements involving a sweep of the bias voltage needed to be made to determine the optimal biasing condition for operation of the transponder, as well as to verify with previous tests the conditions in bias voltage that would allow for modulation. Tests were performed at the previously det ermined optimal transmit frequencies of 1275 MHz and 1300 MHz with the same input power level to the transponder of 24.3 dBm. Bias voltage was swept from 0.3 V to 0.3 V and power output levels from the receiving antenna were recorded as shown in Figures 3.17 and 3.18. Figure 3.17: Bias vs. P out of transponder at 1275 MHz T x frequency 63.0 61.0 59.0 57.0 55.0 53.0 51.0 49.0 47.0 45.0 43.0 41.0 0.40 0.20 0.00 0.20 0.40 Pout of Transponder (dBm) Bias (V)
66 Analysis of the data in Figure 3.17 reveals the specific behavior of the biased diode transponder. The peak occurring at a bias voltage of approximately 0.01 V provides optimal performance of the transponder and the best conversion efficiency. Beginning with a bias of 0.3 V, output power from the transponder climbs steadily until it reaches 0.024 V, after which it fluctuates between 41 to 43 dBm, and then p roceeds to sharply decline after 0.051 V. In a negative bias configuration, a change from 0.024 V to 0.13 V 20.7 dB (output power drops from 63.1 nW to 0.54 nW), which is sufficient to produce a digita l logic zero in retransmission. Likewise, moving in the opposite direction, a change in bias from 0.009 V to 0.151 V 16.1 dB (output power drops from 74.1 nW to 1.82 nW), equally as sufficient to pro duce a digital logic zero.
67 Figure 3.18: Bias vs. P out of transponder at 1300 MHz T x frequency Analysis of the data in Figure 3.18 reveals the specific behavior of the biased diode transponder operating at 1300 MHz. As seen from the figure and correlating to Figure 3.15, the 2nd generation transponder operates less optimally at 1300 MHz. The peak output power occurring at 1300 MHz transmit frequency is 45.8 dBm, which is 4.5 dB less than the peak output power at a t ransmit frequency of 1275 MHz. Additionally, it is observed that a comparable change in bias that caused the results at 1275 MHz (Figure 3.17) has much less of an effect on output power of the transponder at 1300 MHz (Figure 50.5 50.0 49.5 49.0 48.5 48.0 47.5 47.0 46.5 46.0 45.5 0.40 0.30 0.20 0.10 0.00 0.10 0.20 0.30 0.40 Pout of Transponder (dBm) Bias (V)
68 3.18). In a negative bias confi guration, a change from 0.072 V to only produces a change in conversion loss of 4.1 dB (compared to 15.2 dB in Figure 3.17), which may not be sufficient to produce a clear digital logic zero in retransmission. Likewise, moving in th e opposite direction, a change in bias from 0.0 V to 0.098 V causes a change in conversion efficiency of only 3.5 dB. This data verifies the results shown in Figure 3.15 where it can be seen that modulation is less likely possible at 1300 MHz due to much smaller fluctuations of amplitude from a change in bias. These characteristics are important when individual transponders are designed and fabricated for use in a sensing network employing frequency discrimination that is comprised of many devices. For the purposes of interrogation, each sensor would need to be identified by its frequency of operation. 3.7 Conclusions The Remote Lock In Amplifier concept was developed and tested. The capability to effectively communicate with the 1 st generation low powe r remote transponder within the RLIA system was demonstrated. A 2 nd generation biased transponder was built to implement modulation of the retransmitted signal in an effort to employ frequency discrimination within a sensor network. Measurements of both tr ansponders were compared and it was seen that application of the RLIA sensor system in a field environments is possible.
69 CHAPTER 4 APPLICATION OF RECTE NNA TECHNOLOGY: DIOD E INTEGRATED RADAR DETECTOR 4.1 Introduction The following chapter consists of the design fabrication, and testing of a diode integrated rectenna radar detector. The functionality of the radar detectors is to p rov ide sensing capabilities for recognizing the presence of phenomena at a frequency of interest located at the remote s ensor node, and correspondingly induce modulation to the signal being sent from the 2 nd generation transponder to the interrogator node. Through this configuration, the complete RLIA system is realized for use in a frequency discrimination interrogation sc heme. Being essentially a rectenna, the radar detector senses microwaves occurring at a certain frequency within its local environment and transforms the microwave energy to a DC voltage proportional the strength of the signal impinging on the receiving a ntenna. The output of the radar detector can then be redirected to the bias input of the 2 nd g eneration transponder, where this DC voltage input would ca use a change in conversion loss, thus modulating the retransmitted interrogation signal to the base sta tion. When the base station receives the modulated interrogation signal using the lock in amplifier the information sensed by the radar detector will be extracted. However, when the radar detector is not sensing the presence of microwave energy at the freq uency of interest it will simply have a zero DC voltage output, which will not affect the conversion
70 efficiency of the transponder. In a final realization of the RLIA system, the detector and transponder would both be incorporated as part of the remote sen sor node. 4.2 Design of Radar Detector As seen in Figure 1.4, the radar detector is designed by modifying the 1 st generation transponder and replacing the transmitting 2.6 GHz patch antenna with a pad for a DC voltage connection. Also included in the des ign layout is a quarter wavelength resistor for the purpose of maximizing voltage sensitivity and an 11pF RF blocking capacitor The diode integrated radar detector is f abricated on Taconic, 60 mil (Er=6.15) copper substrate incorporating surface mount components. A patch for soldering a DC wire connection is included after the DC blocking capacitor and shunt resistor.
71 Figure 4.1 : Layout of diode integrated radar detector Ideally, when the radar detector is employed in the RLIA system with the transponder as part of an interrogation scheme using frequency discrimination, it will sense phenomena occurring at a unique frequency from the interrogation signal be ing sent to the transponder. However, for the purposes of this research, a frequency of operation for the radar detector of 1300 MHz was chosen. This choice was made due to ease of modifying the existing design of the transponder into the radar detector an d taking advantage of the previously existing narrow band conjugate matching technique used with the 1 st and 2 nd generation transponders. By taking this approach, several advantages were gained: the need for additional filtering for radar signal frequency selection could be minimized or eliminated as well as the need for discrete matching components, and using 1300 MHz for the frequency of operation for the radar detector allowed verification of the input power to the receive antenna of the transponder to b e done. Additionally, due to the compactness of the radar detector, it may be possible to integrate multiple devices
72 onto a single remote sensor node. This would provide the capacity to cover multiple frequencies at once or to harvest a higher amount of mi crowave energy at a single frequency. In the final realization of the RLIA system, where the remote sensor will be located in a field environment, the detector would be designed to operate at a different frequency, for example a frequency required for sens ing the presence of two way radio use. In Chapter Three it was observed that at the optimal operating frequency a change in bias of only 0.1 V could cause output power from the transponder to change by 15 dB, and it was concluded that this kind of change i n conversion efficiency is sufficient to 4.3 Simulation s of Radar Detector Simulations of the diode integrated radar detector were performed in Agilent ADS and the results of presented in this section. Figure 4.2 shows the schematic for the radar detector including the equivalent circuit models for the 1.3 GHz patch antenna (dete ctor antenna), the Schottky diode based doubler (power detector), and the wire connection (output load). Referring to Figure 4.2, the detector antenna block is comprised of four impedance blocks (Z1P5, Z1P8, and Z1p9). Each impedance block provides the co mplex input impedance of the antenna at the fundamental frequency and 2 nd through 4 th harmonics. Immediately following each impedance blocks, in series, is an ideal band pass filter that is open circuited out of band The band pass filters guarantee that o nly one impedance block is active at a given harmonic. A full wave analysis of the antenna was performed with conjugate matching to the power detector at the radar signal frequency, which allowed values to be determined for simulation purposes. The detecto r
73 antenna is followed by a shunt stub for conjugate matching to the power detector, followed by the power detector. The power detector is comprised of a Schottky diode 04 02) model that provides maximum voltage sensitivity to the detector, and the capacitor is an 11pF (Modelithics Johanson 0402) model that behaves as an RF short. On the output side of the power detector is a high age connection. Figure 4.2: Schematic for the conjugate matched radar signal detector
74 As seen in Figure 4.3, the simulated results of the radar detector design predicted an RF to DC conversion efficiency varying from ~0.02% to ~32% for an input power range of 55 dBm to 10 dBm. Additionally, simulation results verify that the radar detector design would be capable of inducing modulation onto the return signal sent from the 2 nd generation transponder to the interrogating node. Figure 4.3: Simulated conversion efficiency of the radar signal detector Figure 4.4: Radar signal detector output voltage and current versus input power (left) and I V curve for the Schottky diode used in the multiplier of the harmonic transceiver (right)
75 Figure 4.4 ( left ) shows the V out and I out curves for the radar detector, and it is seen that with an input signal power of approximately is produced. The I V curve in Figure 4.4 ( right ) indicates that these voltage and current conditions are sufficient to forward bias the diode doubler on the transponder. It can be concluded from the results of the simulations that if the presented radar detector was employed in a field environment for remote sensing, any signal greater than 20 dBm could be detected. The supplied bias from the detector due to this sensed signal would accordingly provide the required change in bias to the transponder that is needed to communicate back to the interrogatin g base station relevant information in the form of a modulated return signal. 4.4 Measurement Results of Radar Detector To perform measurements of the radar detector the same standard testing equipment was used as was previously for testing the transp onder ( Tabl e III Appendix A). A distance of 4 feet was used to achieve an input power to the receiving antenna of the radar detector of 24.3 dBm. The input power to the receiving antenna is equal to the input power of the detector itself due to the fact that the receiving antenna gain is 0 dB. A block dia gram of the testing configuration for the measurements of the radar detector can be seen in Figure 4.5. The transmitting antenna sends a 1.3 GHz signal modulated with a 1 kHz sine wave, which is again the chopping frequency that the lock in amplifier will use for detection of the very small DC voltages at the output of the radar detector. A function generator is used to provide the reference frequency to the lock in amplifier that matches the chopping frequency imparted onto the transmitted signal.
76 Figu re 4.5: Block diagram of testing configuration of the radar detector The first set of measurements of the radar detector involved a transmit power sweep beginning at the maximum calibrated source output of 12.2 dBm (input power of 23.4 dBm to the radar d etector), and decreasing in 1 dB steps until the minimum detectable output from the radar detector is reached The measured and simulat ed results are shown in Figure 4.6 (Table II, Appendix A), where it is seen that very good agreement was reached between these sets of data. Beginning with a 24.3 dBm power input to the detector, output voltage is approximately 20.5 mV (19.4 mV in simulation). As transmit power is decreased in 1 dB steps, output voltage is seen to decrease approximately according t o which is in accordance to the square law region descri bed in section 2.3 of Chapter Two This square law behavior exists until input power reaches approximately 45 dBm, after which point the voltage output from the radar detector is less stable. The reason for this instability is due to TSS of the detector, as discussed in section 3.3 of Chap ter Three. However, in comparison to the DC voltage output of the Narda detector (Figure 3.3), the radar detector is seen to detect lower input powers within the square law region. The Narda detector is seen to operate within the square law region only unt il approximately 35 dBm, which is 10 dB above the edge of the square law region for the radar detector.
77 Figure 4.6: Measured vs. simulated results of the radar detector at P Tx = 1300 MHz.
78 A comparison of the output voltage provided by the radar detector to the output voltage provided by the Narda detector (Figures 3.3, 4.6, and Appendix A) at an input power of approximately 45 dBm show that the radar detector has better performance charact eristics, which are helpful within the proposed application. At an input power of 45.4 dBm, the Narda detector provides 0.069 mV of DC voltage at its output. The radar detector, however, provides 0.25 mV of DC voltage at its output with an input power of 45.3 dBm, which is approximately 360 percent of what the Narda detector can provide with this input power. This is a useful quality for the purposes of providing sufficient DC voltage biases to the transponder when smaller detected signals are present. To consider the feasibility of combining the radar detector and transponder together at the remote sensor node it is necessary to observe, as seen in Figure 3.17, that an approximate 40 mV change in bias was sufficient to provide a change in P out of the retr ansmitted signal equal to 7.8 dB. To determine the required input power to the radar detector to produce an output DC voltage of 40 mV, it is possible to calculate the reponsivity of the radar detector, which is defined as a nd extrapolate from this value The resistivity of the detector was found to be 5518.2 therefore it was found that an input power to the radar detector of 20.4 dBm would be sufficient to produce an output DC voltage of 40 mV, which could ultimat ely be redirected to the bias of the transponder and induce amplitude modulation.
79 4.5 Conclusions Design simulations in Agilent ADS and testing results of the radar detector were included to demonstrate minimum signal detection capabilities at 1300 MHz. It was seen that the fabricated radar detector was capable of sensi ng a signal of approximately 53 dBm, and accordingly producing a rectified DC voltage output of 0.05 mV. By calculating the responsivity of the radar detector, it was also discovered that an input power of 20.4 dBm to the device would be sufficient in creating enough DC voltage to Referring to the measurements presented in Chapter Three of the 2 nd generation transponder in a bias sweep configuration (Figure 3.17), it is seen that an approximate 40 mV change in bias was sufficient to cause a 7 .8 dB drop in conversion efficiency.
80 CHAPTER 5 SUMMARY AND RECOMMENDATIONS FOR FUTURE WORK 5.1 Summary This thesis presented the research, design and fabrication associated with a unique application of rectenna technology combined with lock in amplification. An extremely low power harmonic transponder is conjoined with an interrogator base station, and utilizing coherent demodulation the Remote Lock In Amplifier (RLIA) concept is realized. The lock in amplifier performs phase sensitive detection using a phase locked frequency. Th e lock in amplifier provides benefit to the interrogation system through its ability to detect signals below the noise floor by singling out signal components at specific reference frequencies. Without phase sensitive detection, extremely weak signals or s ignals arriving in the presence of an excessive amount of noise would otherwise be obscured. Development of the RLIA began with a bench top configuration in cluding standard test equipment to perform characterization of a Narda 4503 01 coaxial Schottky barr ier diode detector (Figure 3.2) The purpose of this testing was to verify both the functionality of the lock in amplifier as well as sensitivity of the diode detector. Sensitivity was measured and results were found to correlate sheet. Data related to detector sensitivity served as a reference for future measurements.
81 Using the sensitivity graph (Figure 3.3) and measured output voltage from the detector during experimentation with the full RLIA system, detector input power levels could be deduced, thus validating link budgets. After characterizing the coaxial detector and verifying functionality of all testing equipment, a complete configuration of the RLIA system was constructed including the 1 st g eneration remote transponder (Figure 3.7) Utilizing harmonic re radiation with very low power input, the 1 st g eneration transponder (Figure 3.4) detects a transmitted interrogation signal and responds by retransmitting the second harmonic of the signal. The 1 st g eneration transponde r performs this task while using no additional power besides that which accompanies the wireless signal. Demonstration of the first complete configuration provided proof of concept for the RLIA and feasibility of processing A spectrum analyzer and lock in amplifier were used in th e measurements and the data was compared (Figure 3.8) Results showed that the configuration including the lock in amplifier provided detection of the return signal occurring below the noise floor (~5 dB below what was possible using a spectrum analyzer), thus verifying prior assertions of the RLIA. Following the initial demonstration of the RLIA system d esign and fabrication of a new version of the transponder was introduced. The existing zero bias transponder was modified to include a modulating DC bias to the diode based frequency doubler Applied bias voltage directly changed the impedance match between the receiving 1.3 GHz antenna and the d iode by changing the diode impedance causing a change in conversion loss. It was shown through testing that a change in conversion loss induces an
82 amplitude modulating effect on the retransmission of the signal from the transponder (Figure 3.15) A test of bias sweep at the optimal operating frequency was performed on the 2 nd g eneration transponder and it was seen that a change of ~ 0.1 V in either a positive or negative bias configuration induced an approximate 15 dB change in transponder output power (Fig ure 3.16) This performance characteristic can be advantageously used in implementing a frequency discrimination interrogation scheme for remote sensor networks. Chapter Four of this thesis consisted of the research associated with and design of a radar d etector. The radar detector was comprised of a quarter wavelength patch antenna operating at 1.3 GHz, a Schottky diode, and maximize voltage sensitivity and an 11 pF RF blocking capacitor (Figure 4.1). The radar det ector was designed to sense microwaves occurring at a certain frequency within its local environment and transform the microwave energy to a DC voltage proportional the strength of the signal impinging on its receiving antenna. The output of the radar dete ctor could then be redirected to the bias input of the 2 nd generation transponder, where this DC voltage input would cause a change in conversion loss and modulate the retransmitted interrogation signal from the transponder to the base station. When the base station receives the modulated interrogation signal the information sensed by the radar detector is extracted. When the radar detector is not sensing the presence of microwave energy at the frequency of interest, the conversion efficiency of the transponder will not be affected and the interrogating base stati on will not receive any information. Design simulations in Agilent ADS and testing results of the radar detector were included to demonstrate minimum signal detection capabilities at 1300 MHz. It was seen
83 that the fabricated radar detector was capable of sensing a signal of approximately 53.3 dBm Additionally it was found that an input power of 20.4 dBm to the radar detector would accordingly produce a rec tified DC voltage output of 40 mV. Referring to the measurements presented in Chapter Three of the 2 nd generation transponder in a bias sweep configuration (Figure 3.17), it was seen that an approximate 40 mV change in bias was sufficient to cause a 7 .8 dB dr op in conversion efficiency. The s e two sets of data verified the feasibility of pairing the rad ar detector and the 2 nd generation transponder together at the remote sensor node to perform modulation of interrogation signals. 5.2 Recommendations for Future Work The future work of the Remote Lock In Amplifier concept should involve advances in both t he remote transponder (sensor node) and interrogating bas e station (interrogator node). Future research on the sensor node will be directed to ward specific operational charac teristics and optimized functionalities, demonstrating increased capability within a network. Efforts will be directed at increased power efficiency of the remote sensors to provide extremely long lifetimes with enhanced sensing and potential on/off functi onality. Continued development on the in terrogator node will be aimed to ward increasing compactness and evolving the bench top interrogator/receiver that was demonstrated in this work to a compact hand held or portable device that communicates with sensor nodes. Some expected advantages of the next generation interrogation node ov er the bench top version will be (1) the ability to simultaneously interrogate multiple sensor
84 nodes, and (2) the ability to communicate with other interrogator nodes to optimize network quality of service. Overall f uture research will be focused on a prim ary goal of optimally combining multiple interrogators with multiple low power or zero power sensors
85 REFERENCES  Brown, W.C.; "The History of Power Transmission by Radio Waves," Microwave Theory and Techniques, IEEE Transactions on vol.32, no.9, pp. 1230 1242, Sep 1984  Presas, S.M.; Weller, T.M.; Silverman, S.; Rakijas, M.; "High efficiency diode doubler with conjugate matched antennas," Microwave Conference, 2007. European vol., no., pp .250 253, 9 12 Oct. 2007  Bernhard, J.T.; Hietpas, K.; George, E.; Kuchima, D.; Reis, H.; "An interdisciplinary effort to develop a wireless embedded sensor system to monitor and assess corrosion in the tendons of prestressed concrete girders," Wire less Communication Technology, 2003. IEEE Topical Conference on vol., no., pp. 241 243, 15 17 Oct. 2003  Hagerty, J.A.; Helmbrecht, F.B.; McCalpin, W.H.; Zane, R.; Popovic, Z.B.; "Recycling ambient microwave energy with broad band rectenna arrays ," Microwave Theory and Techniques, IEEE Transactions on vol.52, no.3, pp. 1014 1024, March 2004  Le, T.; Mayaram, K.; Fiez, T.; "Efficient Far Field Radio Frequency Energy Harvesting for Passively Powered Sensor Networks," Solid State Circuits, IEEE Journal of vol.43, no.5, pp.1287 1302, May 2008  Jiawei Xu; Guy Meynants; Merken, P.; "Low power lock in amplifier for complex impedance measurement," Advances in sensors and Interfaces, 2009. IWASI 2009. 3rd International Workshop on vol., no., pp.110 114, 25 26 June 2009  Jung Hwan Choi; Jung Ick Moon; Seong Ook Park; "Measurement of the modulated scattering microwave fields using dual phase lock in amplifier," Antennas and Wireless Propagation Letters, IEEE vol.3, no., pp. 340 3 43, 2004 doi: 10.1109/LAWP.2004.839629  Davies, R.; Meuli, G.; "Development of a digital lock in amplifier for open path light scattering measurement," Industrial Electronics & Applications (ISIEA), 2010 IEEE Symposium on vol., no., pp.50 55, 3 5 Oct. 201045 
86  C. A. Balanis, Antenna Theory: Analysis and Design, 2 nd edition, Wiley, N.Y., 1997  D. M. Pozar, Microwave Engineering, 3 rd edition, Wiley, N.Y., 2005  Brown, W. C.; Kim, C. K.; "Recent Progress in Power Reception Efficiency in a Free Space Microwave Power Transmission System," Microwave Symposium Digest, 1974 S MTT International vol.74, no.1, pp. 332 333, Jun 1974  Walsh, C.; Rondineau S.; Jankovic, M.; Zhao, G.; Popovic, Z.; "A conformal 10 GHz rectenna for wireless powering of piezoelectric sensor electronics," Microwave Symposium Digest, 2005 IEEE MTT S International vol., no., pp. 4 pp., 12 17 June 2005  Selvakumaran, R.; Liu, W.; Boon Hee Soong; Luo Ming; Sum, Y.L.; "Design of low power Rectenna for wireless power transfer," TENCON 2009 2009 IEEE Region 10 Conference vol., no., pp.1 5, 23 26 Jan. 2009  Aguilar, S.M.; Weller, T.M.; "Tunable harmonic re radiat or for sensing applications," Microwave Symposium Digest, 2009. MTT '09. IEEE MTT S International vol., no., pp.1565 1568, 7 12 June 2009  Electrical World and Engineer, March 5, 190 4.  Matsumoto, H.; "Research on solar power satellites and microwave power transmission in Japan," Microwave Magazine, IEEE vol.3, no.4, pp. 36 45, Dec 2002  Brown, W.C.; "Experiments Involving a Microwave Beam to Power and Position a He licopter," Aerospace and Electronic Systems, IEEE Transactions on vol.AES 5, no.5, pp.692 702, Sept. 1969  Shinohara, N.; Matsumoto, H.; "Experimental study of large rectenna array for microwave energy transmission," Microwave Theory and Technique s, IEEE Transacti ons on vol.46, no.3, pp.261 268, Mar 1998  Shinohara, N.; Miyata, Y.; Mitani, T.; Niwa, N.; Takagi, K.; Hamamoto, K.; Ujigawa, S.; Jing Ping Ao; Ohno, Y.; "New application of microwave power transmission for wireless power distri bution system in buildings," Microwave Conference, 2008. APMC 2008. Asia Pacific vol., no., pp.1 4, 16 20 Dec. 2008  Microwave Power Engineering, vol. 1, New York: Academic Press, 1968, pp. 275 294.
87  Brown, William C., George, Roscoe H. et al., U.S. Patent 3434678, March 25, 1969.  Power Engineering, vol. II, E. C. Okress, Ed. New York: Aca demic, 1968, pp. 273 275.  Brown, W.C.; "An experimental low power density rectenna," Microwave Symposium Digest, 1991., IEEE MTT S International vol., no., pp.197 200 vol.1, 10 14 Jul 1991  Glaser, P.E.; "The potential of satellite solar power," Proceedings of the IEEE vol.65, no.8, pp. 1162 1176, Aug. 1977  Glaser, P.E.; "An overview of the solar power satellite option," Microwave Theory and Techniques, IEEE Transactions on vol.40, no.6, pp.1230 1238, Jun 1992  Brown, Terminal of a Free Contractor Rep. PT 4964, NASA CR 135194, Aug. 1977.  Brown, W.C.; Eves, E.E.; "Beamed microwave power transmission and its application to space ," Microwave Theory and Techniques, IEEE Transactions on vol.40, no.6, pp.1239 1250, Jun 1992  McSpadden, J.O.; Little, F.E.; Duke, M.B.; Ignatiev A.; "An in space wireless energy transmission experiment," Energy Conversion Engineering Conference, 1996. IECEC 96. Proceedings of the 31st Intersociety vol.1, no., pp.468 473 vol.1, 11 16 Aug 1996  STS 80 Columbia ORFEUS SPAS 2/Wake Shield Fa cility 3, KSC Release No. 124 96, October 1996, NASA; http://wwwpao.ksc.nasa.gov/kscpao/release/1996/124 96.htm  ed nonlinear Radio Sci. Bull. vol. 273, pp. 11 35, June 1995.  Shinohara, N.; "Development of high efficient phased array for microwave power transmission of Space Solar Power Satellite/Station," Antennas and Propagation Society I nternational Symposium (APSURSI), 2010 IEEE vol., no., pp.1 4, 11 17 July 2010  Powercast Corporation; web address: http://www.powercastco.com/ ; Feb 17, 2011.  WiTricity; web address: http://www.witricity.com/index.html ; Feb 17, 2001.
88  Wireless Power Consortium; web address: http://www.wirelesspowerconsortium.com/ ; Feb 17, 2001.  Strassne r, B.; Kai Chang; "A circularly polarized rectifying antenna array for wireless microwave power transmission with over 78% efficiency," Microwave Symposium Digest, 2002 IEEE MTT S International vol.3, no., pp.1535 1538, 2002  Ali, M.; Yang, G.; Dougal, R.; "A new circularly polarized rectenna for wireless power transmission and data communication," Antennas and Wireless Propagation Letters, IEEE vol.4, no., pp. 205 208, 2005  Lim, S.; Leong, K.M.K.H.; Itoh, T.; "Adaptive power contro llable retrodirective array system for wireless sensor server applications," Microwave Theory and Techniques, IEEE Transactions on vol.53, no.12, pp. 3735 3743, Dec. 2005  Lim, S.; Itoh, T.; "A 60 GHz retrodirective array system with efficient p ower management for wireless multimedia sensor server applications," Microwaves, Antennas & Propagation, IET vol.2, no.6, pp.615 625, Sept. 2008  Olgun, U.; Chi Chih Chen; Volakis, J.L.; "Low profile planar rectenna for batteryless RFID sensors," Antennas and Propagation Society International Symposium (APSURSI), 2010 IEEE vol., no., pp.1 4, 11 17 July 2010  Benes, E.; Groschl, M.; Seifert, F.; Pohl, A.; "Comparison between BAW and SAW sensor principles," Ultrasonics, Ferroelectrics and Frequency Control, IEEE Transactions on vol.45, no.5, pp.1314 1330, Sep 1998  Wilson, W.C.; Malocha, D.C.; Kozlovski, N.; Gallagher, D.R.; Fisher, B.; Pavlina, J.; Saldanha, N.; Puccio, D.; Atkinson, G.M.; "Orthogonal Frequency Coded SAW Sensors for Aerospace SHM Applications," Sensors Journal, IEEE vol.9, no.11, pp.1546 1556, Nov. 2009  Pohl, A.; Seifert, F.; "Wirelessly interrogable SAW sensors for vehicular applications," Instrumentation and Measurement Technology Conference, 1996. IM TC 96. Conference Proceedings. 'Quality Measurements: The Indispensable Bridge between Theory and Reality'., IEEE vol.2, no., pp.1465 1468 vol.2, 1996  Kozlovski, N.Y.; Malocha, D.C.; "SAW passive wireless multi sensor system," Ultrasonics Sympos ium (IUS), 2009 IEEE International vol., no., pp.1541 1544, 20 23 Sept. 2009 [45 ] Aguilar, S.M.; Weller, T.M.; "Tunable harmonic re radiator for sensing applications," Microwave Symposium Digest, 2009. MTT '09. IEEE MTT S International vol., no., pp.1565 1568, 7 12 June 2009
89 [46 ] A. D. Poularikas; S. Seely;, Signals and Systems, PWS Publishers, Boston, MA, 1985 [47 ] Hewlett [48 ] Linli Xie; Yonghong Zhang; Yong Fan; Conghai Xu; Yuanbo Jiao; "A W band detector with high tangential signal sensitivity and voltage sensitivity," Microwave and Millimeter Wave Technology (ICMMT), 2010 International Conference on vol., no., pp.582 531, 8 11 May 2010
91 Appendix A Additional Tables of Measurements, Lists and Specifications of Equipment Table I : Sensitivity measurements of Narda detector (Model 4503 01) Calibrated Power (dBm) Vdc (mV) 11.58 586.7 10.58 521.5 9.58 460.5 8.58 406.7 7.58 356.4 6.58 314.3 5.58 274.3 4.58 239.3 3.58 207.3 2.58 180.5 1.58 154.7 0.58 133.3 0.42 113.6 1.42 96.0 2.42 81.5 3.42 68.9 4.42 57.8 5.42 48.0 6.42 39.7 7.42 33.1 8.42 27.3 9.42 23.2 10.42 18.8 11.42 15.3 12.42 12.4 13.42 10.06 14.42 8.07 15.42 6.49 16.42 5.17 17.42 4.16 18.42 3.32 19.42 2.73 20.42 2.16 21.42 1.74 22.42 1.39 23.42 1.12 24.42 0.90 25.42 0.73 26.42 0.58 27.42 0.48 28.42 0.39
92 Appendix A (Continued) Table I : (Continued) Calibrated Power (dBm) Vdc (mV) 29.42 0.33 30.42 0.27 31.42 0.23 32.42 0.20 33.42 0.17 34.42 0.15 35.42 0.13 36.42 0.12 37.42 0.11 38.42 0.097 39.42 0.092 40.42 0.087 41.42 0.082 42.42 0.079 43.42 0.080 44.42 0.070 45.42 0.069 46.42 0.068 47.42 0.069 48.42 0.070 49.42 0.069 50.42 0.061 51.42 0.061 52.42 0.059 53.42 0.061 54.42 0.060 55.42 0.059 56.42 0.058 57.42 0.059 58.42 0.059 59.42 0.062 60.42 0.061 61.42 0.060 62.42 0.059 63.42 0.059
93 Appendix A (Continued) Table II: Measured and simulated sensitivity values of radar detector T x Power (dBm) P in Radar Detector (dBm) V DC meas. (mV) V DC simul. (mV) 12.2 24.3 20.5 19.4 11.2 25.3 16.4 16.2 10.2 26.3 14.8 13.5 9.2 27.3 12.8 11.2 8.2 28.3 10.6 9.2 7.2 29.3 7.6 7.6 6.2 30.3 6.5 6.2 5.2 31.3 5.4 5 4.2 32.3 4.7 4.1 3.2 33.3 3.4 3.3 2.2 34.3 2.8 2.7 1.2 35.3 2.2 2.1 0.2 36.3 1.7 1.7 0.8 37.3 1.6 1.4 1.8 38.3 1.3 1.1 2.8 39.3 0.9 0.88 3.8 40.3 0.68 0.7 4.8 41.3 0.58 0.56 5.8 42.3 0.37 0.44 6.8 43.3 0.36 0.35 7.8 44.3 0.27 0.28 8.8 45.3 0.25 0.24 9.8 46.3 0.24 0.18 10.8 47.3 0.18 0.14 11.8 48.3 0.12 0.11 12.8 49.3 0.08 0.09 13.8 50.3 0.05 0.07 14.8 51.3 0.04 0.06 15.8 52.3 0.06 0.05 16.8 53.3 0.05 0.04
94 Appendix A (Continued) Table III: List of parts used in testing configuration Manufacturer: Model and Specs: Equipment Type: Hewlett Packard 8594E (9k 2.9 GHz) Spectrum Analyzer Stanford Research Systems SR530 Lock In Amplifier Agilent 33120A (15 MHz) Function Generator Agilent ESG D4000A (250 kHz 4 GHz) Signal Generator Narda 4503 01 Schottky Barrier Detector Narda 614A (2.6 3.95 GHz, Gain=5.048 dBi) Waveguide Sciperio (custom made = 1.3 GHz, Gain=7.075 dBi) Waveguide Figure I: Sensitivity curve of Narda Schottky barrier detector (4503 01)
95 Appendix B 1 st Generation Transponder Measurement Notes The following summarizes the challenges confronted during testing of the 1 st generation transponder within the RLIA set up and techniques of overcoming them. B.1 Transmitter Block It was known that a properly functioning 1 st generation transponder had a peak conversion efficiency of around 20 dB, occurring at input power levels within the range of approximately 30 dBm to 20 dBm. Also known, was that as the input power to the transponder falls below 30 dBm the conversion efficiency also starts to decrease. The same effect occurs when the input power to the transponder reaches above 20 dBm. Simply sending more transmit power from the interrogator was not the solution to maximizing the retransmitted 2.6 GHz signal power from the transponder. Testing had to b e done to insure that the transmitted 1.3 GHz signal was between 20 dBm and 30 dBm and that it was also a clean transmit signal, free of harmonics. The first part of testing included the front end receiver block as seen in Figure 3.7 (not include the dow n conversion stage). To accomplish a clean transmit signal, first a low pass filter (VLF 1800+) was added between the signal source and the transmitting antenna. However signal harmonics were still seen on the spectrum analyzer at the receiving end of the system. Investigation needed to be done to see if the signal harmonics were originating from the signal source. It was estimated that the total path loss from transmitting antenna to receive antenna including the conversion loss of the transponder was clos e to 100 dB. Therefore, any signal harmonics originating from the signal source needed to be suppressed by at least this value. It was found that as 3 5
96 Appendix B (Continued) successive filters were added in series, the transmitted signal was not sufficiently clean. However introducing 7 VLF 1800+ filters in series provided a transmit signal that was sufficiently free of source generated harmonics to achieve a clean received signal at 2. 6 GHz. This however, was not in itself sufficient for a clean receive signal from the transponder. In addition work had to be done to the receiver side of the interrogator. B.2 Receiver Block In addition to introducing filters at the transmit end of the R LIA system, additional filtering became required to eliminate interference signals from entering the system. Not only were 2.6 GHz coaxial band pass filters (VBFZ 2527+) required for filtering interference signals, as seen in Figure 3.7, they were also req uired for eliminating harmonics caused by the amplifiers within the receiver block, as well as harmonics created from non linear behavior of the spectrum analyzer. In addition to the coaxial band pass filters, a coupled line band pass filter (percent bandw idth = 1%) was designed to narrow the noise bandwidth of the system. The coupled line band pass filter was not by itself successful at narrowing the noise bandwidth. During testing, it was seen that the filter was behaving as antenna and was itself introdu cing interference signals into the system. To overcome this, a metal enclosure was utilized to house the filter and block additional interference from being injected into the system at the receiver block end. After this measure was taken, the receive signa l seen on the spectrum analyzer was greatly improved.
97 Appendix B (Continued) A down conversion stage with an IF of 61 MHz was implemented to eliminate unwanted signals that were appearing around 2.6 Ghz. Eliminating these interference signals was no t crucial in terms of viewing the signal on the spectrum analyzer, because the desired received signal can be visually identified from the interference signals. However, proper filtering of interference is necessary for the purposes of reading voltage leve ls on the lock in amplifier. The total power being input to the detector will provide a voltage reading to the lock in amplifier, and if this total power input is not solely comprised of the desired signal, the lock in amplifier will give an inaccurate reading. It was found that the combination of the transmit signal filtering, added receiver block filtering and additional down conversion was sufficient to eliminate unwanted signals at the desired frequencies.