Integration of Radio Frequency Harvesting with Low Power Sensors

Dissertation

Presented in Partial Fulﬁllment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Brock J. DeLong, B.S., M.S.

Graduate Program in Electrical and Computer Engineering

The Ohio State University

2018

Dissertation Committee:

John L. Volakis, Advisor Asimina Kiourti, Co-Advisor Liang Guo c Copyright by

Brock J. DeLong

2018 Abstract

This dissertation gives guidelines for state-of-the-art power harvesters and for optimizing its components, e.g., rectifier, matching network, and antenna, in various applications. A single diode rectifier using a quarter-wave matching circuit with a measured efficiency of 73.7% is also presented. Several experimental demonstrations are included for powering a number of sensors and devices, such as a clock, computer mouse, calculator, thermometer, medical insulin pump, and super capacitor with power management circuitry.

To increase the amount of RF harvested power, an array of rectifying antennas

(rectennas) is presented and used in experiments up to 60 meters. Wireless power transfer demonstrations at near ﬁeld distances are also presented. For the latter, we show a strong tolerance to misalignment while delivering high levels of power (1.2 mW over 42 cm). As an application, a medical pump is successfully powered over this distance. Further, bandwidth widening techniques are presented along with rectiﬁer optimizations.

To reduce the overall dimensions of the rectenna, miniaturization techniques are discussed. This leads to a rectenna size of 1.5 x 2.5 cm2, making it ideal for medical or on-body applications. This rectenna was used to successfully activate a body-worn thermometer across 65 cm. In the case of implantable devices, a dielectric matching

ii layer was found useful and validated using pig skin. A related SAR analysis ensured the safety of the proposed RF powering harvesting techniques.

iii Dedicated to my lovely wife, Megan

iv Acknowledgments

I would like to thank Megan, my wonderful wife, who has supported me through the process of graduate school. She has supported and loved me well through this journey, and I could not have done it without her.

I would like to thank my parents, Kevin and Laurie DeLong, who encouraged me to pursue and stick with graduate school. They have encouraged me tremendously through this process.

I would like to thank my colleagues, Cedric Lee, Ushe Chipengo, Satheesh Bojja

Venkatakrishnan, Shubhendu Bhardwaj, Jingni Zhong, Md Asiful Islam (Asif), and

Roland Tallos. Getting to work alongside these skilled people was incredible. I will certainly miss our weekly lunch meetings. I would like to thank my junior colleagues

Jack Blauert, Vigyanshu Mishra, Saad Alharbhi, Keren Zhu, and Ramandeep Vilhkhu for making my last year at the ESL memorable.

I would like to thank Dr. Volakis and Dr. Kiourti, my advisors and mentors, who found me, pushed me to be better, provided support and help, and who I am forever indebted to for their persistent guidance. What they do on a daily basis is nothing short of amazing.

Finally, I would like to thank God, who gave me a mind to think, and reason, and explore, and ﬁnd good work, and ultimately to understand my need of a savior, Jesus

Christ.

v Vita

September 4, 1991 ...... Born - Van Wert, OH

2014 ...... B.S. Electrical and Computer Eng., The Ohio State University, USA summa cum laude 2017 ...... M.S. Electrical and Computer Eng., The Ohio State University, USA

Publications

Research Publications

B. J. DeLong, A. Kiourti, J. L. Volakis, “Cutting The Cord: A Button-Sized Rectenna for Wireless Patient Monitoring Using Radiated Near-Field Signals at 2.4 GHz,” 2018 IEEE International Symposium on Antennas and Propagation (APSURSI), Boston, MA, 2018

B. J. DeLong, C. W. L. Lee, A. Kiourti, S. B. Venkatakrishnan and J. L. Volakis, “Wireless energy harvester from 700-900 MHz,” 2018 United States National Com- mittee of URSI National Radio Science Meeting (USNC-URSI NRSM), Boulder, CO, 2018, pp. 1-2.

B. J. DeLong and A. Kiourti and J. L. Volakis, “A Radiating Near-Field Patch Rectenna for Wireless Power Transfer to Medical Implants at 2.4 GHz,” IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology,vol. 2, no. 1, pp. 64-69, March 2018.

R. Vilkhu, B. DeLong, A. Kiourti, P. Das Ghatak, S. Mathew-Steiner and C. K. Sen, “Power harvesting for wearable electronics using fabric electrochemistry,” 2017 IEEE International Symposium on Antennas and Propagation & USNC/URSI National Ra- dio Science Meeting, San Diego, CA, 2017, pp. 213-214.

vi B. DeLong, A. Kiourti and J. L. Volakis, “A 2.4-GHz wireless sensor network using single diode rectennas,” 2016 IEEE International Symposium on Antennas and Propagation (APSURSI), Fajardo, 2016, pp. 403-404.

B. DeLong, Q. Yuan and J. L. Volakis, “Long range, safe power transmission using iteratively-tuned rectiﬁcation,” 2015 International Symposium on Antennas and Propagation (ISAP), Hobart, TAS, 2015, pp. 1-2.

B. DeLong, C. C. Chen and J. L. Volakis, “Wireless energy harvesting for medical applications,” 2015 IEEE International Symposium on Antennas and Propagation & USNC/URSI National Radio Science Meeting, Vancouver, BC, 2015, pp. 1213-1213.

Fields of Study

Major Field: Electrical and Computer Engineering

Studies in: RF Power Harvesting Prof. A. Kiourti, Prof. J. L. Volakis Microwave Circuit Design Prof. A. Kiourti, Prof. J. L. Volakis Antenna Design and Miniaturization Prof. A. Kiourti, Prof. J. L. Volakis

vii Table of Contents

Page

Abstract ...... ii

Dedication ...... iv

Acknowledgments ...... v

Vita ...... vi

List of Tables ...... xi

List of Figures ...... xii

1. Introduction ...... 1

1.1 History of Wireless Power Transfer ...... 1 1.2 Overview of Applications for Wireless Power Transfer ...... 5 1.2.1 Far-Field Power Transfer ...... 5 1.2.2 Near-Field Power Transfer ...... 7 1.2.3 Mid-Field Transfer for Medical Applications ...... 8 1.3 Organization of this Thesis ...... 9

2. RF Harvester Design Guidelines ...... 10

2.1 Antenna Design ...... 11 2.2 Rectification Theory and Design ...... 11 2.2.1 Rectification Theory ...... 11 2.2.2 Practical Rectifier Modeling ...... 14 2.2.3 Matching Network Design ...... 17 2.2.4 RF Pass and DC Block Design ...... 20 2.2.5 Rectification Elements ...... 21 2.2.6 RF Block and DC Pass Design ...... 22

viii 2.3 Load Design ...... 22 2.4 Experimental Demonstration of WPT for Wireless Charging . . . . 23

3. Max Power Delivery for Sensor Applications using Arrays and Wideband Rectennas ...... 29

3.1 Voltage Multiplication Using Multi-Stage Rectenna ...... 30 3.2 Demonstrations at 2.4 GHz ...... 32 3.2.1 Long Range Measurements ...... 32 3.2.2 Powering Up Devices ...... 35 3.3 Wireless Energy Harvester from 700-900 MHz ...... 38 3.3.1 Wideband Antenna Design ...... 38 3.3.2 Wideband Rectiﬁer Design ...... 42 3.4 System Integration and Measurements ...... 43

4. Minimal Loss Rectiﬁer Using Single Diode ...... 46

4.1 Quarter-Wave Rectifier Operation ...... 47 4.1.1 Quarter-Wave Rectifier Components ...... 47 4.1.2 Fabrication and Measurement of Quarter-Wave Rectifier . . 49 4.2 Comparisons with Other Rectifiers ...... 54 4.3 Integration with Quarter-Wave Antenna ...... 56 4.3.1 Demonstrations and Measurements ...... 60

5. Near-Field RF Harvesting for Medical Devices ...... 63

5.1 Introduction to Near-Field Harvesting ...... 63 5.2 Near-Field vs. Far-Field Harvesting ...... 64 5.3 Rectenna Design ...... 66 5.4 Radiating Near-Field Antenna Measurements ...... 70 5.5 Results and Demonstrations ...... 72 5.6 Conclusion ...... 74

6. Applications and Considerations for Medical Sensors ...... 76

6.1 Miniaturizing Power Harvesting for Medical Applications ...... 76 6.2 Miniature Rectenna Design ...... 77 6.3 Transmitting Power through the Skin ...... 82 6.4 SAR Guidelines ...... 85 6.5 Conclusion ...... 88

ix 7. Conclusion ...... 89

7.1 Summary of Work ...... 89 7.2 Outlook as a Technology ...... 90 7.3 Future Work ...... 93

Bibliography ...... 97

Glossary ...... 106

x List of Tables

Table Page

1.1 Various commercial sensors ...... 6

2.1 SPICE parameters for the Skyworks SMS7630 and Broadcom HSMS2860 Diode ...... 15

2.2 Various rectiﬁers seen in literature ...... 16

3.1 Examples of various transmitting scenarios for ﬁxed point-to-point links according to FCC rules and regulations ...... 32

3.2 Recharge times of a super capacitor over various distances ...... 37

3.3 Dimensions of wideband rectenna ...... 40

3.4 Comparison of this work with other wideband harvesters in literature 45

4.1 Comparison with similar rectiﬁers in literature ...... 57

5.1 Relevant coil misalignment studies in literature when separated by 15 cm...... 71

5.2 Comparison of the proposed versus reported midﬁeld WPT systems . 74

6.1 Relevant coil misalignment studies in literature compared to this work *Separation at 5 cm, and including system losses ...... 80

xi List of Figures

Figure Page

1.1 Nikola Tesla's tower at Wardenclyﬀe [6]. Copyright c 1911 . . . . . 2

1.2 W.C. Brown's wirelessly powered helicopter. [9] Copyright c 1958 . . 4

1.3 Ambient power harvesting system by Olgun et al. [16] Copyright c 2012 6

1.4 Early works in wireless near-ﬁeld phone charging. [28] Copyright c 2004 8

2.1 Building blocks common of WPT systems ...... 10

2.2 Classically modeled, simulated, and measured IV curve for SMS7630 diode ...... 12

2.3 Input and output voltages in the frequency and time domain . . . . . 13

2.4 Building blocks for a rectiﬁcation circuit at RF ...... 15

2.5 ADS Schematic for Rectiﬁer Design using Source Pull ...... 17

2.6 Rectiﬁcation eﬃciency for various input impedances ...... 18

2.7 Rectification efficiency vs. input VSWR for quarter wave rectifier while frequency, power, and load are held constant ...... 19

2.8 Common approaches for rectiﬁer matching networks ...... 19

2.9 BQ25570 power management circuitry from Texas Instruments [49][50] 23

2.10 Block diagram of all components for a WPT system ...... 23

xii 2.11 Entire rectenna on one Printed Circuit Board (PCB). All units are in mm...... 24

2.12 Measured vs. Simulated Rectiﬁer Eﬃciency for Voltage Quadrupler . 25

2.13 Measured power around the back of a laptop while it was constantly uploading ...... 26

2.14 Recharge of super capacitor using rectenna and power management circuit ...... 27

2.15 Discharge of super capacitor while using computer mouse ...... 28

3.1 Illustration of 3x3 rectenna array at 2.4 GHz ...... 31

3.2 Rectenna experiments in a room ...... 33

3.3 Voltage over distance in a room setting ...... 34

3.4 Power over distance in a room setting ...... 35

3.5 Rectenna experiments in a building hallway ...... 36

3.6 Wireless rectenna demonstration ...... 39

3.7 Wideband rectenna in simulation and fabrication ...... 41

3.8 Simulated and measured VSWR for the tapered oﬀset dipole antenna 42

3.9 Circuit schematic of wideband rectiﬁer ...... 43

3.10 Measured wideband rectiﬁer eﬃciency ...... 43

3.11 Rectiﬁed power (dBm) versus input power density ...... 44

4.1 Basic quarter-wave rectiﬁer schematic ...... 47

4.2 Quarter-Wave rectiﬁer, a) ADS schematic, and b) simulated source- pull eﬃciency ...... 50

4.3 Fabricated quarter-wave rectiﬁer, all dimensions are in mm ...... 51

xiii 4.4 Transmission lines designed for de-Embedding the SMA-to-Microstrip transition...... 52

4.5 Interpolated insertion loss for various transmission lines ...... 53

4.6 Measured conversion eﬃciency of single diode quarter-wave rectiﬁer . 53

4.7 Comparison of the quarter-wave rectiﬁer with other common rectiﬁers when optimized for 8 dBm ...... 55

4.8 Comparison of the quarter-wave rectiﬁer with other common rectiﬁers when optimized for -10 dBm ...... 57

4.9 Illustration and fabrication of the quarter-wave rectenna (units are mm) 58

4.10 Field distribution when the diode is a) OFF (RF open) and b) ON (RF short) ...... 59

4.11 Simulated quarter-wave rectiﬁer eﬃciency ...... 59

4.12 Simulated gain of the shorted quarter-wave microstrip antenna when Open and Short (Diode OFF and ON, respectively)) ...... 60

4.13 Quarter-wave rectenna demonstration ...... 61

4.14 Rectiﬁed output power from the quarter-wave rectenna ...... 62

5.1 Voltage quadrupling rectiﬁer topology ...... 66

5.2 Simulated, measured, and theoretical output voltage from voltage quadrupler, and simulated output from a voltage doubler ...... 69

5.3 Lateral misalignment tests ...... 70

5.4 Radiating near-ﬁeld rectenna setup ...... 72

5.5 Measured voltage and current across a 1.8 kΩ load ...... 73

5.6 Actuated medical pump ...... 74

xiv 6.1 Back (left) and front (right) of rectenna circuit with dimensions (in mm): a=25, b=15, c=12, d=2.25, e=11.25, f=1.25, g=3.3, h=2.1, i=3.8, j=3.8, k=3.8, l=0.66 ...... 78

6.2 Miniature rectenna S11 ...... 79

6.3 Miniature rectenna eﬃciency ...... 80

6.4 Misalignment Analysis for Miniature Rectenna ...... 81

6.5 Demonstration of WPT rectenna button turning on a temperature sensor on a body phantom...... 81

6.6 a) Dielectric, and b) conductivity properties of bulk muscle ...... 83

6.7 Illustrations and simulations of a quarter-wave dielectric matching layer 84

6.8 Pig skin, a) without a matching layer, and b) with a matching layer (b) 85

6.9 Measured power coupling between internal/external antenna with and without a matching layer ...... 86

6.10 SAR demonstrations while radiating when, a) antenna closely located to surface of skin, and b) antenna with dielectric matching layer . . . 87

6.11 Lateral misalignment loss in pig skin ...... 88

7.1 Smart desks concepts ...... 91

7.2 Power requirements for a commercial calculator ...... 92

7.3 Power requirements for a thermometer ...... 93

7.4 Operational calculator using wireless power ...... 94

xv Chapter 1: Introduction

1.1 History of Wireless Power Transfer

The concept of Wireless Power Transfer (WPT) has been around since the days of

Nikola Tesla [1]. The ﬁrst mentioning of power transmission through the air was by

Tesla's patents at the end of the 19th century and early 20th century [2]–[4]. Believing strongly in the invention of such a system, he states, “the economic transmission of power without wires is of all-surpassing importance to man,” even going on to say, “it will bring peace and harmony on Earth” [5]. Tesla's Wardenclyﬀe power station was

a large scale attempt to deliver wireless power to homes and industries, and is shown

in Fig. 1.1. His theory was not built on the theories of electromagnetic radiation;

instead, Tesla states, “the transmission of electrical energy [in his device] is one of true

conduction and is not to be confounded with the phenomena of electrical radiation”.

In his mind, the Earth's atmosphere would be used as a conductor that would be

excited, and whose resonating power would be accessible at any point around the

world [4]–[6]. Sadly though, the dreams of Tesla did not become a reality in his

lifetime. But his vision of widely distributed wireless power continues to this day,

and these aspirations will be the focus of this work.

1 Figure 1.1: Nikola Tesla's tower at Wardenclyﬀe [6]. Copyright c 1911

The appeal of wireless power is immediately tangible in everyday life; from cell phones, to laptops, to electric vehicles, the ability to wirelessly power devices would be a revolutionary leap forward in technology. But how is such a phenomenon possible?

The answer lies in wireless energy sources. Traditional energy sources from coal, wind, gas, etc. rely on wired connections to deliver power. However, there are some energy sources not bound by wires. These include: bio-fuels, solar cells, piezoelectricity, ultrasonic transducers, inductive coupling, and radiated power [7]. However, of all these choices, the most ubiquitous option (or that which has the possibility to cover the most area and the most remote places) is wireless radiated power.

2 The ﬁrst modern demonstrations of long range wireless power transmission were performed by William C. Brown of Raytheon in 1959 [8][9]. Brown carried out several microwave-powered ﬂight experiments for an unmanned helicopter (see Fig. 1.2), including one to an altitude of 50,000 feet. He was responsible for the introduction of the concept for the rectifying antenna, or 'rectenna'. These antennas contained diodes at the terminals, converting the incident Radio Frequency (RF) signal to usable

DC power. In addition to wirelessly powering a helicopter, W.C. Brown performed an experiment with NASA in which an astounding 34 kW of power was transmitted wirelessly across 15 km with an eﬃciency of over 80% [10][11].

These investigations were furthered by Peter Glaser who developed the concepts of a solar power satellite (SPS) in 1968. Among other applications, Glaser proposed the concept of an energy harvesting satellite, in which solar energy would be converted to microwaves on the satellite, and subsequently beamed down to Earth [12]. However, due to budgetary constraints this project was eventually abandoned. As [8] points out,“had the SPS program actually gone forward, it would have been the greatest mega-engineering feat of all time, dwarﬁng the entire Apollo program many times over”.

From the late 1970's to the early 1990's, a rapid growth of interest in wireless power transmission in the microwave regime emerged. Eﬃciencies for rectennas increased from a few percent to over 85% at high powers at 2.45 GHz, and power requirements shrunk considerably, viz. 50 mW [13]. Improvements came through the invention and use of low voltage drop Schottky diodes, and the development of enhanced matching circuits to reduce harmonic content created by the diode. Addi- tionally, higher frequency components were created as well, such as a 35 GHz rectenna

3 Figure 1.2: W.C. Brown's wirelessly powered helicopter. [9] Copyright c 1958

from Chang et. al [14]. Radio Frequency Identiﬁcation (RFID) has also been a major example of wireless powering in recent decades.

From the early 2000's and onward, focus on WPT has been in two directions, namely: 1) far-field radiation, and 2) near-field coupling. These techniques and applications for Wireless Power Transfer (WPT) will be the topic of this work, as well as their applications for low power sensors. Additionally, a new method for wireless powering in the 'radiating near-field' will be discussed.

4 1.2 Overview of Applications for Wireless Power Transfer

1.2.1 Far-Field Power Transfer

Far ﬁeld radiation applications have explored the use of radio waves, and especially ambient signals such as WiFi, AM/FM radio, cell phone signals, and television for use in wireless applications [15]. In these studies, a low power wireless device such as a sensor or detector could be used indeﬁnitely for wireless monitoring with no wires or batteries. An example of one such temperature sensor being power via ambient energy harvesting by Olgun et al. is shown in Fig. 1.3 [16].

The primary focus has been on introducing higher efficiency rectification circuits at low powers, and utilizing the most efficient methods of wireless power transfer.

Typically, ambient power ranges for WiFi or television exist in the -20 to -40 dBm range [17], or 10 µW to 100 nW. While these power levels may seem low, eﬀorts have been made to capture this low power and brieﬂy turn on low power devices. Modern day companies have been pursuing the ideas of WPT for phone recharging and low power sensor operation as well [18][19].

Some devices and their turn-on power are given in [20] and [21]. By way of example, a few commercially available low power sensors are presented in Table 1.1 as well. As can be seen, these low power sensors operate in the µW range. However, many applications in sensing use a a duty-cycle (periodic on-oﬀ time), which lowers the average consumed power even further.

In addition to ambient power, power harvesting can occur when there is intentional illumination. Key to harvesting these ambient RF signals is the introduction of high eﬃciency rectiﬁcation circuits that turn-on even though incoming signals are of very low power, viz. on the order of µW. In either case (ambient or intentionally applied

5 Figure 1.3: Ambient power harvesting system by Olgun et al. [16] Copyright c 2012

Table 1.1: Various commercial sensors

Ultra-Low Power Sensors Ref. Device Power Requirement [22] Temperature Sensor 3 µW [23] Motion Sensor 18 µW [24] Temperature Sensor 30 µW [25] Pressure Sensor 75 µW

power), the far-field losses necessitate very efficient circuitry in order to operate. As such, many advances have come to the circuitry that makes power harvesting possible, e.g. the rectifier. The rectifier is responsible for converting the RF signal from the antenna into Direct Current (DC) power, which operates the sensor. A detailed review of modern rectifiers in literature is given in [26].

While dreams of free wireless energy everywhere once existed, realistic limitations of propagation and human safety have set boundaries on radiating power sources,

6 which limits the widespread availability of wireless power. However, prompted by the wide availability of WiFi, there has been a resurgence of work in WPT systems.

These concepts feed into the so-called Internet of Things (IoT), in which everything becomes integrated: cars, computers, phones, appliances, etc. This mesh of sensors would serve to optimize and quantize the world in which we live. A wide variety of sensing applications exist, such as speed, rotation, temperature, humidity, light, chemical, medical, etc. As such, far-ﬁeld WPT will continue to be a studied ﬁeld for many years to come.

1.2.2 Near-Field Power Transfer

A very diﬀerent approach to WPT occurs for wireless power transfer in the near-

field. Near-field WPT depends mostly on magnetic coupling of coils, and not radiated power. These coils typically resonate in range of 0.3-30 MHz [21], and occasionally into the 100s of MHz. Mathematical derivations of source-loop to load-loop coupling have advanced significantly in recent years, in addition to the equivalent circuitry that describes the resonators, e.g. [27]. Wireless inductive charging for devices such as phones and even cars have attracted much attention as of late, even becoming popular in the commercial sector. An image of early applications in near-field power transfer is shown in Fig. 1.4 [28]. This field has seen rapid advancements due to enhanced planar coupling techniques [28]–[30]. Due to the close proximity of the coils, immense levels of energy can be transferred, even into the kilowatt range [31].

Near-field capacitive coupling is another form of wireless power transfer that gen- erates an electric field between two capacitive plates. This method of wireless power transfer is growing due to its applications in the medical field [32].

7 Figure 1.4: Early works in wireless near-ﬁeld phone charging. [28] Copyright c 2004

The most prominent form of near-ﬁeld power transfer has been magnetically coupled coils. From the 1960's onward, concepts in wirelessly coupled, implantable devices have been extensively explored. Wireless medical devices as well as the massive miniaturization of sensors have allowed for the potential for diagnosis and patient care without repetitive surgeries [21]. This ﬁeld will only continue to grow as on-body and in-body medical devices and monitoring systems develop.

1.2.3 Mid-Field Transfer for Medical Applications

In recent years, the far-field and near-field demonstrations have become blurred together as loosely coupled coils have began to be view as quasi-radiating struc- tures [33]. This new area has been especially interesting in the medical field. The

8 primary beneﬁt of the so-called midﬁeld transfer is that the rigid alignment toler-

ance of near-ﬁeld devices are loosened when the power is radiated. Poon et al. has pioneered this work with coils [34], [35].

1.3 Organization of this Thesis

In this work, the primary focus will be on the design and application of WPT devices. Chapter 2 will outline the guidelines for design and implementation of a modern rectenna. Chapter 3 will discuss power enhancing and wideband methods to operate low power sensors. Chapter 4 introduces and investigates an efficient single diode rectifier and a novel antenna structure as well. Chapter 5 introduces the concept of WPT in the radiating near-field, and presents a medical application. Chapter 6 discusses wireless power applications and considerations for medical designs. Finally,

Chapter 7 summarizes the work and concludes with future steps in the journey of wireless power.

9 Chapter 2: RF Harvester Design Guidelines

This chapter will outline the steps to build a rectifying antenna (rectenna) from beginning to end. The application demonstrated here will be to recharge a supercapacitor wirelessly, which will then operate a computer mouse. The basic building blocks of a free space WPT system are shown in Fig. 2.1. It consists of: 1) transmitting antenna, supplying power wirelessly through an electromagnetic signal, 2) receiving antenna, capturing the electromagnetic signal, 3) rectiﬁer, converting the RF signal on the antenna into DC, and 4) load, which can be a battery, sensor, device, etc.

These building blocks will be discussed in the following sections. Section 2.1 will cover the guidelines for antenna design, Section 2.2 will discuss the theory and implementation of the rectiﬁer, Section 2.3 will break down the principles of load design, and ﬁnally Section 2.4 will demonstrate the practical experiment.

Figure 2.1: Building blocks common of WPT systems

10 2.1 Antenna Design

The receiving antenna is responsible for capturing electromagnetic energy wirelessly and converting it to an RF signal. The antenna can come in any variety of shapes and sizes. Three parameters that must be chosen before the design begins are: 1) resonant frequency, 2) desired gain pattern, and 3) input impedance of the antenna. The resonant frequency of the antenna will determine the overall size of the antenna, as well as the speciﬁc frequency bands that the user wishes to rectify.

The gain pattern determines how directional the receiving antenna is, e.g. focused in one direction, or omnidirectional. Finally, the impedance of the antenna must be a well-known value so that it can be matched to the rectiﬁer.

2.2 Rectiﬁcation Theory and Design

Before discussing the practical application of the rectifier, it is essential to understand rectification theory. Rectification is the process by which the RF signal coming into the antenna is converted to DC. It is important to have DC because most low power devices (e.g., sensors) operate using DC power. We will first discuss the theory behind rectification in order to gain an intuitive understanding, and then we will move to practical implementation.

2.2.1 Rectiﬁcation Theory

A rectiﬁcation circuit is a non-linear circuit; that is, the current and voltage (IV) relationship of the diode is not linear. This point is illustrated by Fig. 2.2, where the input voltage is increased linearly and the diode is seen to “turn on” after a suﬃcient input voltage is achieved. This non-linearity can be exploited for several

11 Figure 2.2: Classically modeled, simulated, and measured IV curve for SMS7630 diode

useful functions in RF systems, such as mixers, detectors, oscillators, and rectiﬁers,

to name a few [36]. A perfect diode acts simply as a switch (which is also non-linear)

that is closed when the Alternating Current (AC) signal is in one polarity, and it is

open when the AC signal is in the other polarity. The classic non-linear model [36]

of the current on a diode can be written as

αV I V = Is e − 1 (2.1)

−6 −15 where Is is the saturation current (typically between 10 -10 A), α is equal to q/nKT , where q is the charge of an electron (1.602 ∗ 10−19 C), n is the ideality factor

(between 1 and 2), k is Boltzmann's constant (1.38 ∗ 10−23 m2kgs−2K−1), T is the

temperature (290 K for room temperature), and V is the voltage. Thus, a typical

value of α is approximately 1/(25 mV ) [36]. An ideal IV plot for a generic diode

using Eq. (2.1) as well as simulated and measured IV curve for a SMS7630 diode is

shown in Fig. 2.2.

Since the exponential term in Eq. 2.1 is quite large, it is often simpliﬁed as

12 (a) Diode schematic (b) Time domain view (c) Frequency domain view

Figure 2.3: Input and output voltages in the frequency and time domain

αV I V = Is e , (2.2) which shows the current as a non-linear function the voltage. The non-linearity of the diode is what is responsible for creating DC as well as the harmonic components.

This is due to the fact that the exponential in (2.2) can be written as a Taylor series expansion [37] in the form

∞ X xn x2 x3 ex = = 1 + x + + + ··· (2.3) n! 2 3! n=0 and thus, Eq. 2.2 becomes

αV 2 αV 3 IV = I 1 + αV + + + ··· , (2.4) s 2 3! which shows that a DC component (unity) is created, as well as the original signal and higher harmonics. By simply viewing the diode response in the frequency domain as in Fig. 2.3c, we observe this phenomenon. Fig. 2.3 shows the time domain and spectral response of a time-varying signal applied to a diode. All rectiﬁer simulations took place in Keysight Advanced Design System (ADS).

13 From this vantage point we can define the RF-to-DC rectification efficiency as

P η = DC,out (2.5) PRF,in

in which PRF,in is the amount of power in the fundamental frequency, and PDC,out is the power in the output DC signal. This will be the key figure of merit when comparing various rectification circuits. Using this definition for efficiency, we are now able to compare various efficiencies for rectification circuits.

In reality however, the diode contains a myriad of parameters that affect its operation under various circumstances, making it difficult to model using a closed form equation. Indeed, a diode's input impedance changes whenever its input power, frequency, and/or load conditions change. To account for these effects, we turn to the

Keysight ADS. The simulation includes other more subtle effects of the diode, including parameters that affect rectification efficiency such as parasitic resistance and the reverse breakdown voltage of a diode. These parameters are typically given in

Simulation Program with Integration Circuit Emphasis (SPICE) format. The diode

SPICE parameters for the Skyworks SMS7630 [38] and Broadcom HSMS2860 diode

[39] are given in Table 2.1. In [40] we can ﬁnd the description of the physical meaning of these parameters.

Now that we have discussed the theoretical background of rectiﬁcation, it is appropriate to now discuss the practical implementation. The rectiﬁer block shown in

Fig. 2.1 will now be broken down further.

2.2.2 Practical Rectiﬁer Modeling

Various rectification circuits are shown in Table 2.2. To build an efficient rectifier, four elements are needed, as depicted in Fig. 2.4. Namely, there is: 1) an input

14 Table 2.1: SPICE parameters for the Skyworks SMS7630 and Broadcom HSMS2860 Diode

Diode SPICE Parameters Parameters Units SMS7630 HSMS2860 Is A 5E-6 5E-8 Rs Ω 20 6.0 N - 1.05 1.08 TT sec 1E-11 - CJ0 pF 0.14 0.18 M - 0.4 0.5 EG eV 0.69 0.69 XTI - 2 2 FC - 0.5 - BV V 2 7.0 IBV A 1E-4 1E-5 VJ V 0.34 0.65

Figure 2.4: Building blocks for a rectiﬁcation circuit at RF

matching network, 2) an RF Pass/DC Block element, 3) a rectifying element(s), and

4) an RF Block/DC Pass element. To begin, we will discuss the input matching network.

15 Table 2.2: Various rectiﬁers seen in literature

Various Rectifier Topologies Rectifier Type Rectifier Topology Notes

Bridge Full-wave rectiﬁer

Current Doubler Self-explanatory

Quadruples the Greinacher voltage

Half Wave Max 50% eﬃcient Creates standing wave between matching network Quarter Wave and quarter-wave line, above 50% eﬃciency possible

Simple single diode Series rectiﬁer

Simple single diode Shunt rectiﬁer

Increases voltage for Villard each added stage. Requires balun.

Voltage Doubler Self-explanatory

16 2.2.3 Matching Network Design

To match a rectifier well, its reflection coefficient, |S11|, must be minimized for maximum rectification. This is important because the rectifier's efficiency depends on how much power it receives. Rectifier circuits can be matched either with lumped elements or tunable microstrip stubs.

To minimize the time spent on matching network design, a simulated source-pull technique is used in ADS. A source-pull applies the same amount of power at each input impedance. An ADS schematic of this concept is shown in Fig. 2.5. In this

figure, we see a source pull, rectifier, and load. The source pull applies a fundamental tone at a particular power. The power is probed as it flows to the rectifier. The rectifier shown in this case is voltage quadrupling Dickson charge pump using SMS7630 diodes. Finally, the load is fixed as a 1.8 kΩ resistor. The corresponding RF-to-DC conversion efficiency is measured at each input impedance and is shown in Fig. 2.6.

Figure 2.5: ADS Schematic for Rectiﬁer Design using Source Pull

17 Figure 2.6: Rectiﬁcation eﬃciency for various input impedances

Using the input impedance given by Fig. 2.6, one could design an optimally matched rectifier. Such a matching network will transform the impedance of the antenna to match that of the rectifier. As mentioned, matching the rectifier is key since the incident RF power heavily affects the RF-to-DC conversion efficiency. This concept is illustrated by Fig. 2.7. In this figure, a simulated rectifier is shown while other factors are held constant except for the input impedance; the conversion efficiency swiftly drops as the Voltage Standing Wave Ratio (VSWR) is increased.

Assuming we know the ideal input impedance, Zin, from the source pull technique described in Section 2.2.2, we can design a matching network using one of two processes: 1) lumped element tuning, or 2) microstrip stub tuning. For lumped element

18 Figure 2.7: Rectification efficiency vs. input VSWR for quarter wave rectifier while frequency, power, and load are held constant

(a) Lumped elements (b) Open stub

Figure 2.8: Common approaches for rectiﬁer matching networks

tuning, L-, T-, and π-matching networks are typical. For microstrip stub tuning, open stub matching is typical. These examples are illustrated in Fig. 2.8. The matching process in reality can be very sensitive for both processes.

For the lumped Capacitor, Inductor, Resistor (CLR) components in the GHz range, like the one shown in [41] and [42], the primary challenge is due to its parasitics, which makes the model non-ideal. In fact, at a certain frequency known as the Self-

Resonant Frequency (SRF), the parasitic eﬀects overtake the nominal value of the

19 component. The result is a capacitor looking like an inductive impedance, or an inductor looking like a capacitive impedance. Unless the models account for these parasitic eﬀects, such as the ones in [43], additional post-fabrication tuning of inductor and capacitor values will most likely be required. Additional losses may arise as well due to parasitic resistance in the model, as well as soldering eﬀects. However, lumped element matching can occupy a substantially smaller footprint as compared to microstrip stub tuning.

For microstrip stub tuning, matching is performed by open or shorted stub lengths in parallel to the rectifier. The length of the stub and its positioning can be used as tuning variables. This method does not experience parasitic effects like the lumped elements. However, it will occupy a larger footprint and can face additional de-tuning due to fringing fields, and these effects must usually be modeled with a full-wave solver, such as ADS Momentum. Meandering techniques may be used for miniaturization. However, there may be losses due to signal radiation in the stubs. It is common practice to use copper tape for post-fabrication tuning.

2.2.4 RF Pass and DC Block Design

Directly after the matching network, there is usually a large input capacitor. The purpose of this capacitor is to pass the RF signal to the rectification elements, while blocking the rectified DC from flowing backward. If the DC signal is allowed to flow backward, there is a chance it can be shorted to ground by the matching network.

As an aside, it is also important to protection the Vector Network Analyzer (VNA) and signal generator machines from DC signals, as they typically cannot handle large

20 amounts of DC power at the port. To avoid extra losses, these capacitors should have

a low Equivalent Series Resistance (ESR), as explained in [44].

2.2.5 Rectiﬁcation Elements

A diode's SPICE parameters are used to deﬁne its IV characteristics. In simulation, a Harmonic Balance tool is then used in order to account for the harmonic frequencies that are generated by the diode's non-linear IV curve. This tool is useful

for calculating the RF-to-DC conversion eﬃciency. Additionally, the Large Signal

S-Parameter (LSSP) tool is useful for matching the rectiﬁer across a bandwidth.

Rectiﬁcation elements seen in literature, almost without exception, have been

either Schottky diodes or Complementary Metal Oxide Semiconductor (CMOS) [45]

transistors. In the low power (-20 to 10 dBm) microwave range, Schottky diodes are

standard. These diodes’ non-linear response is capable of producing a DC component

as well as various harmonics, as described in Section 2.2.1. The diode that is chosen

depends on the application and the power requirements for the rectiﬁer. As the study

in [45] shows, the SMS7630 is an ideal diode for the low power range. In our tests,

we choose the SMS7630 diode since it rectiﬁes well at 0 dBm (1 mW). Additionally,

the SMS7630 diode is commercially available and inexpensive compared to a CMOS

alternative.

Typically, a diode's SPICE parameters are used to model it in simulation. How- ever, there are other methods that exist. These methods include behavioral, circuit- based, and X-parameter1 models. Behavioral models are discussed in great detail

in [46]. Circuit-based models are described in [47]. Finally, X-parameters measure

the amplitude and phase information of each harmonic or spectral component, and

1Registered trademark of Agilent Technologies

21 more information for this technique can be found in [48]. In the future, these modeling techniques may become standard practice for diode modeling.

2.2.6 RF Block and DC Pass Design

After the rectification elements, it is necessary to have a filter and/or matching network to block any remaining RF signal from going forward while passing DC to the load. In the traditional scenario, a shunt capacitor is used to simply short all remaining RF components to ground. However, more advanced designs can be created using an output matching network, which reflects the RF signal back into the rectification elements, utilizing wasted harmonics.

2.3 Load Design

The load receives the DC power passed from the RF Block/DC Pass element(s).

This is usually a battery, sensor, etc., and is modeled as a resistor. The load affects the overall rectification circuit’s efficiency and input impedance. Thus, it is important to choose a load that rectifies efficiently. For the load, we choose a 1.8 kΩ resistor to represent the power management chip. In this case, we will use the power management circuit shown in Fig. 2.9.

The purpose of the power management chip is to boost the voltage from the rec- tiﬁer to a suﬃcient level so that it can recharge a battery or capacitor. The BQ25570 power management chip requires only 0.3 V to activate its circuitry from a cold-start scenario [49]. After it has been activated, it can continue to boost the voltage and run the circuitry down to 0.1 V. It also monitors for overvoltage/undervoltage cases, and helps to mitigate these scenarios.

22 Figure 2.9: BQ25570 power management circuitry from Texas Instruments [49][50]

2.4 Experimental Demonstration of WPT for Wireless Charg- ing

At this point, we have covered all the necessary components for Wireless Power

Transfer. Combining the block diagram elements shown in Fig. 2.1 and Fig. 2.4, we show all the components for WPT in one block diagram in Fig. 2.10. Now, it will be demonstrated how these pieces can come together to build a practical WPT system.

Figure 2.10: Block diagram of all components for a WPT system

Based on the block diagram in Fig. 2.10, we present all components of a rectenna on a single Printed Circuit Board (PCB) in Fig. 2.11. The elements in this ﬁgure are described as follows:

23 Figure 2.11: Entire rectenna on one Printed Circuit Board (PCB). All units are in mm.

(a) Antenna - A simple patch element, in this case, optimized for resonance at 2.4

GHz.

(b) Matching Network - Two meandered open stubs (to save space) that can be

tuned to account for any shifts in frequency.

matching network to pass the RF signal to the load, while blocking DC from

ﬂowing backward.

(d) Rectiﬁcation Element - Four SMS7630 Schottky diodes in a voltage quadrupler

arrangement.

24 (e) RF Block/DC Pass Element - Two 100 nF shunt capacitors to block the RF

signal from propagating forward, while passing DC to the load.

(f) Load - Two lead wires (+/-) that can be fed to the power management circuit.

It should be mentioned that the rectifier and antenna should first be fabricated and measured separately before they are combined on one PCB. For example, the rectifier was first fabricated as a stand-alone unit and tested to determine its efficiency. The rectifier was then fed by a signal generator at various power levels to measure the

RF-to-DC conversion efficiency. The load was fixed at 1.8 kΩ. With minor tuning between the simulated and fabricated models, the conversion efficiency is shown in

Fig. 2.12. A coaxial-to-microstrip feed provided the input power for PRF,in, while the

2 voltage across the output resistor provided PDC,out, or V /R. As seen, we achieved a maximum simulated eﬃciency of 55%, and a measured eﬃciency of 47.7% at 11 dBm.

At 0 dBm, we see a 40% conversion eﬃciency.

Figure 2.12: Measured vs. Simulated Rectiﬁer Eﬃciency for Voltage Quadrupler

25 As mentioned at this beginning of this chapter, we would like to try to recharge a super capacitor using WPT. The power source in this case was laptop’s WiFi signal. During this test, the rectenna was placed very closely to the laptop’s radiating antenna. As such, this antenna would actually be receiving energy in the near-ﬁeld.

Fig. 2.13 shows the received energy levels around the back of the laptop in dBm. The most power is measured at the upper left corner, with a maximum measured power of approximately 0 dBm. The laptop was continuously uploading during these tests.

Figure 2.13: Measured power around the back of a laptop while it was constantly uploading

The output DC leads were then fed to the BQ25570 power management chip, which was then fed to a 4 Farad super capacitor. While the laptop was uploading, the voltage across the super capacitor was measured. The results are shown in Fig. 2.14.

As indicated by this graph, it took approximately 10 hours to completely recharge the

26 super capacitor using the laptop’s signals. Additionally, the majority of the charging took place in the ﬁrst three and a half hours, rising from zero to two volts.

Figure 2.14: Recharge of super capacitor using rectenna and power management circuit

Once the super capacitor was charged, it was then connected directly to a computer mouse. The computer mouse was a bluetooth-connected, wireless mouse. It was then used normally throughout the day and, as Fig. 2.15 indicates, the mouse lasted a little over four hours. Thus, we have used wireless power to operate a computer mouse.

27 Figure 2.15: Discharge of super capacitor while using computer mouse

28 Chapter 3: Max Power Delivery for Sensor Applications using Arrays and Wideband Rectennas

Wireless Power Transfer (WPT) can enable remote charging and operation of small electronic devices like sensors. Herein, we discuss intentionally applied energy at 2.4 GHz and from 700-900 MHz as a source for power harvesting. Indeed, the concepts of utilizing an array for wireless power harvesting has been explored, as in [51] where an 8x8 spiral array was utilized to harvest ambient signals from 2-18

GHz. Another wideband array was shown in [52] where energy harvesting in the

London Underground network showed eﬃcient harvesting (40%) while covering the

Digital Television (DTV), Global System for Mobile (GSM), and 3G bands between

0.3-3 GHz. However, these applications consider the cases when transmitted power is ambient, and basic information such as frequency and polarization is unknown. In these cases, the power is practically very low. In this chapter, we apply a known signal with a known power level, frequency, and polarization. We utilize arrays and wideband rectennas to perform novel experiments in which we characterize power and voltage delivery for actual sensors and devices over long distances (up to 60 meters).

The work shown in Ch. 2 demonstrates the operation of a single rectenna at 2.4

GHz. Here, we will demonstrate how to further enhance and optimize the amount of power we can receive using rectenna arrays and wideband receivers. Section 3.1

29 will discuss the arrangement and operating principles of the array. Section 3.2 will demonstrate various operating cases in diﬀerent environments, such as in a room, or in a hallway. Also in this section, we demonstrate various low power devices being operated by this array. Finally, in section 3.3 we will demonstrate wideband rectenna principles, which will demonstrate eﬃcient power delivery in the 700-900 MHz bands.

3.1 Voltage Multiplication Using Multi-Stage Rectenna

For long range wireless power transmission, we implement an array of rectennas. Fig. 3.1 shows an illustration of this concept. As the illustration indicates, the rectenna elements are approximately spaced at half wavelength separations. In this case, a half wavelength at 2.4 GHz corresponds to 6.25 cm in free space.

Similar to a battery, each negative terminal of the rectenna can be connected via wire to the positive terminal of the next rectenna. The result is simple voltage addition as each rectenna stage was added. Alternatively, if all rectennas were connected in parallel, a boost in the current would be observed. It was decided that it would be better for each antenna element to have its own rectiﬁer due to the added complexity of ensuring proper phasing if all patch elements were combined. Additionally, extra losses due to the interconnects of the patches would degrade performance.

These multi-stage rectennas each provided voltage quadrupling. Thus, all nine elements provided a total of 36 times voltage multiplication. We emphasize the voltage in this case because the power management chip requires a minimum voltage of 100mV to operate after cold start [49]. In each experimental case, there was a signal generator connected to a single transmitting antenna. In order to comply with Federal Communications Commission (FCC) standards as of early 2018 [53], the

30 Figure 3.1: Illustration of 3x3 rectenna array at 2.4 GHz

maximum Equivalent Isotropically Radiated Power (EIRP) for a ﬁxed point-to-point link is limited according to FCC Title 47 Part 15.247(b) and (c). This can be stated as the following equation

EIRPmax = Pt[dBm] + Gt[dB] − Pm[dB] (3.1)

where Pt is limited to 30 dBm (1 Watt), Gt is the transmitting antenna gain, and Pm is 1 dB for each 3 dB of antenna gain over 6 dB [54]. A few scenarios for maximum

EIRP are illustrated in Table 3.1.

31 Table 3.1: Examples of various transmitting scenarios for ﬁxed point-to-point links according to FCC rules and regulations

Various EIRP examples for ﬁxed point-to-point links Allowable? Transmitter Antenna Gain EIRP [dBm] Power [dBm] [dB] Yes 18 18 36 Yes 30 6 36 No 32 4 36 Yes 29 9 38 No 30 8 38

3.2 Demonstrations at 2.4 GHz

A rectenna array was fabricated and is shown in Fig. 3.2a. Using this array, two types of experiments were carried out: 1) maximum voltage and power receivable over long ranges (not in the near-ﬁeld), and 2) real-time device actuation and recharging over these distances. These experiments took place in either a room or hallway setting, as would be typical in real life. The transmitter is shown in Fig. 3.6b. In all scenarios, the transmitted power was kept within FCC limits. These tests were carried out in order to show the feasibility of WPT for low power sensors in these environments.

3.2.1 Long Range Measurements

The ﬁrst measurements captured the real-time voltage as the physical separation between the transmitter and rectenna array receiver was varied. In this case, the load was inﬁnitely large, and thus this can be considered the open-circuited DC voltage produced by the rectenna array. Fig. 3.3 shows the open-circuited voltage that the rectenna array is able to produce. In these experiments, the array was kept linearly polarized.

32 (a) Fabricated 3x3 rectenna array

(b) Signal generator with 2 dBi half-wave dipole antenna

Figure 3.2: Rectenna experiments in a room

33 Figure 3.3: Voltage over distance in a room setting

The transmitter was set to 17 dBm, and the gain of the transmitting rubber duck antenna was 2 dBi. Thus, the transmitting EIRP was 19 dBm, well below the FCC limits. Fig. 3.3 shows that the voltage stays suﬃciently high over a distance of 5 meters. We note that the voltage always stays at or above 0.3 V.

The second experiment that took place was power versus distance, again in a room setting. In this case, the transmitter EIRP was kept at 19 dBm, but a load of 4.7 kΩ was placed across the terminals of the rectenna array receiver. Fig. 3.4 shows the power received in a room setting. Since the transmit power was kept low (sub-100 mW), the received power over the room distance (5 meters) dropped to approximately

-46 dBm, or 25 nW. However, if the EIRP was increased, the overall level of received power would correspondingly increase, and one could expect powers in the microwatt range.

The third, and ﬁnal experiment measuring the rectenna array was for voltage levels in a building hallway. The transmitter power was decreased to 11 dBm and the transmitting antenna was changed to a patch array with a gain of 19 dBi; thus, the

34 Figure 3.4: Power over distance in a room setting

EIRP was now 30 dBm. Fig. 3.5b shows the transmitter and receiver rectenna array in the hallway setting. As the receiver distance was varied, again the open-circuited voltage was measured. This is shown in Fig. 3.5c.

As the plot in Fig. 3.5b demonstrates, the voltage stays above 0.1 V across nearly the entire hallway. Interestingly, the received DC voltage tends to oscillate as it moves down the hallway. This is most likely due to the constructive and destructive nature of the 2.4 GHz signal as it propagates.

3.2.2 Powering Up Devices

The second campaign of experiments involved operating actual devices remotely using the intentionally applied RF signal. As applications for the Internet of Things

(IoT) emerge, sensors and devices grow to become truly wireless using the techniques described in this section. This especially includes devices that are located in hard to reach locations, such as on a ceiling, or behind a wall. Additionally, wireless sensor operation oﬀers a means of delivering power to hazardous environments, such

35 (a) Signal generator with a (b) Transmitter and re- 19 dBi patch array ceiver rectenna array

Figure 3.5: Rectenna experiments in a building hallway

36 as inside a furnace, or in a radioactive environment. Removing a battery or wired solution provides a critical ﬂexibility and increases up-time for the sensor.

To begin, the power management chip was connected to the rectenna array. This chip was set to boost the voltage on the incoming receiver to 1.5 V, the equivalent of a standard AA battery. The output of the power management chip was sent to a super capacitor with a capacitance of 0.1 mF. The test scenario took place in a room setting again. However, this time the transmit power was reduced to 11 dBm and the transmitter gain was 19 dBi, making the EIRP 30 dBm.

In each test case, the capacitor was completely discharged. The time it took to complete a charge up to 1.5 V was measured at each distance. Table 3.3 shows the times to charge up the super capacitor. As can be seen, the amount of time to charge up to a 1.5 V increases non-linearly as the distance increases. This is due to drop in efficiency from: 1) path loss, 2) rectification loss, and 3) power management circuit efficiency.

Table 3.2: Recharge times of a super capacitor over various distances

Recharge times for a 0.1 mF super capacitor Distance [m] Charge Time [mm:ss] Volts 0.5 00:10 1.5 1.0 00:13 1.5 1.5 00:27 1.5 2.0 01:08 1.5 2.5 08:54 1.5 3 22:20 1.5

The ﬁnal scenario for operating a device incorporated an actual clock that was powered wirelessly over a distance of 0.9 meters. For this test, the transmitter was

37 located on a table top surface, and the transmit power was set to 16 dBm (40 mW) with a rubber duck antenna gain of 2 dBi. Normally this clock required a single AAA battery, but in this case the battery was removed and replaced by the feeds of the rectenna array. The wirelessly powered clock began operating immediately when the signal generator was operated. This demonstration is shown in Fig. 3.6.

3.3 Wireless Energy Harvester from 700-900 MHz

Signals in the 700-900 MHz range (also known as the Ultra High Frequency (UHF) band) are various, and have been heavily used for mobile communication, broadcast, and navigation applications [55]. As such, these signals are nearly ubiquitous in application for daily life. Given their availability and their ability to penetrate walls, trees, buildings, etc. [56], these frequencies are attractive for more continuous wireless energy harvesting. The ability to penetrate walls, for example, could be used for wireless recharging of various sensors around home, such as smoke alarms, IoT sensors, and small medical devices. Further, with more frequencies to receive from, we increase the amount of power our WPT device can accept.

Typical wireless power harvesting devices focused on the 2.4 GHz range, given the wide availability of WiFi [57]. However, many of the applications for 2.4 GHz have been limited to close range, or even near-ﬁeld power transmission [58]. With this in mind, herein we propose and demonstrate a UHF rectenna that covers the 700-900

MHz bands.

3.3.1 Wideband Antenna Design

To achieve a fractional bandwidth of 25% (700-900 MHz), we propose a planar tapered oﬀset dipole antenna, shown in Fig. 3.7a. The associated dimensions of the

38 (a) Rectenna array with transmitter on desk

(b) Wireless clock powered wirelessly using rectenna array

Figure 3.6: Wireless rectenna demonstration

39 antenna and rectiﬁer circuit are given in Table 3.3. It is noted that a slight taper near the edges of the dipole leads to improved wideband performance. An important feature of the oﬀset dipole is the large metallic ground plane, shown in Fig. 3.7a.

This ground plane was designed to be electrically large and serve as a suﬃciently large ground for the rectiﬁer circuit.

The oﬀset tapered dipole was simulated and optimized in Ansys HFSS. The substrate was chosen to be Rogers TMM 10, with a relative permittivity of r=9.2, a dielectric loss tangent of tan δ=0.0022, and a thickness of 60 mil with 1 oz. cladding.

Table 3.3: Dimensions of wideband rectenna

Wideband Rectenna Widths and Lengths Dimension Length (mm) Dimension Length (mm) w 81.3 l8 2.2 d1 10.7 l9 3.9 d2 148.4 l10 4.1 d3 119.8 w1 1.4 l1 40.1 w2 2.4 l2 22.4 w3 0.3 l3 6.9 w4 2.1 l4 6.9 w5 2.2 l5 2.9 w6 0.3 l6 4.1 w7 0.9 l7 9.7 - -

The antenna was fabricated without the rectiﬁer circuit, and a UFL connector was placed at the feed point of the rectiﬁer. Vias were drilled to connect the back-facing dipole arm to the UFL connector. The simulated and measured VSWR is shown in

Fig. 3.8. VSWR less than 2 is shown between 700 MHz and 900 MHz. Also, the

40 (a) Front and back face of wideband rectenna

(b) Rectiﬁer with dimensions

Figure 3.7: Wideband rectenna in simulation and fabrication

41 simulated broadside gain of the dipole showed a gain of above 2.2 dBi from 700-900

MHz.

Figure 3.8: Simulated and measured VSWR for the tapered oﬀset dipole antenna

3.3.2 Wideband Rectiﬁer Design

The broadband rectiﬁer shown in Fig. 3.9 was designed and fabricated to oper-

ate across the bands of interest. The two diodes used were Skyworks SMS7630. A

source-pull simulation using Keysight ADS showed that an input impedance matching

network can be employed at our desired frequencies and power levels. Harmonic Bal-

ance and Large Signal S-Parameter (LSSP) engines were also used for characterizing

the diodes’ non-linear performance.

The fabricated rectiﬁer performance is shown in Fig. 3.10. Similar to the antenna,

the rectiﬁer and matching network were designed on a Rogers TMM 10 substrate.

The ﬁgure shows the approximate power delivered to the rectiﬁer when the incident

power density was 1 µW/cm2. Also, Fig. 3.10 shows that a peak conversion eﬃciency of 64.3% is achieved at 700 MHz with an input power of 10 dBm.

42 Figure 3.9: Circuit schematic of wideband rectiﬁer

Figure 3.10: Measured wideband rectiﬁer eﬃciency

3.4 System Integration and Measurements

Once the offset tapered dipole and wideband rectifier were independently validated, the two were combined to form the RF harverster. For testing, the 50 Ω antenna port was fed into the 50 Ω port of the rectifier and fabricated onto a single rectenna shown in Fig. 3.7a.

43 The ﬁnal rectenna was then characterized in an anechoic chamber. To do so, a signal generator was used to feed into a power ampliﬁer, with the output signal feeding into the UHF horn antenna. The transmitting horn was placed a distance of

210 cm away in order to meet the far-ﬁeld condition, viz. beyond 2D2/λ, which was a suﬃcient distance so that the power density onto the rectenna was 1-10 µW/cm2.

The transmitted power density was calculated at this distance using Eq. (3.2).

P × G × G in a t (3.2) 4π × r2

In (3.2), Pin is the input power from a signal generator, Ga is the gain of the ampliﬁer, Gt is the gain of the transmit horn, and r is the distance the signal has radiated. The rectiﬁed power versus power density for the fabricated rectenna is given in Fig. 3.11, showing a 10dB variation across the band.

Figure 3.11: Rectiﬁed power (dBm) versus input power density

44 A table of comparisons showing similar wideband rectenna designs is shown in

Table 3.4. In this table, we see that the designed rectenna shows a smaller footprint

while eﬃciently covering the 700-900 MHz band with a small load (500 Ω).

Table 3.4: Comparison of this work with other wideband harvesters in literature

Wideband harvesters in literature Ref. Bandwidth Input Max Eﬃ- Footprint Antenna Load Power ciency Gain Range This 700-900 -10 to 5 62 120.65 2 500 Work [59] 700-900 -10 to 0 65 182 3.6 2500 [60] 850-1940 2.7 to 7.2 60.4 169 2 500 [61] 470-860 -10 to 10 60 - - 12000 [61] 840-975 -17 to 0 58 - - 10000

In summary, a novel wideband rectenna was designed using a tapered oﬀset dipole

with greater than 25% fractional bandwidth and a wideband rectiﬁer circuit. The

RF-to-DC conversion eﬃciency of the rectiﬁer had a maximum value of 64.3%. The

rectenna operates eﬃciently even with incident power densities down to 1 µW/cm2.

This wideband device can be used for sensor integration and in-home IoT applications.

45 Chapter 4: Minimal Loss Rectiﬁer Using Single Diode

This chapter proposes and demonstrates a quarter-wave single-shunt diode recti-

fier for RF energy harvesting with optimized efficiency. The proposed rectifier has

73.7% RF-to-DC conversion efficiency at 8 dBm, and rectifies efficiently down to lower power levels as well. Similar to the rectifier shown in [26], this rectifier consists of a single diode and a quarter-wave transmission line. However, we herein show that this rectifier, with the proper matching network, provides the highest simulated efficiency of any of the aforementioned rectifiers. The matching network is designed using source-pull techniques as described in Section 2.2.3. Such a rectifier can be scaled to various power levels and frequencies of interest as well, and can be re-optimized for other frequencies as needed.

While the single diode is capable of rectifying only one half of the AC signal, the λ/4 transmission line serves to reflect the positive wave cycle and recycle it back into the diode. This is how over 50% rectification efficiency is possible with a single diode. We note that this rectifying element can be easily integrated with an antenna to create a rectenna for wireless energy systems.

In Section 4.1, the operating principles of the quarter-wave rectiﬁer are discussed as well as a detailed fabrication and testing procedure. A comparison with other

46 rectiﬁer circuits is shown in Section 4.2. In Section 4.3, the quarter-wave rectiﬁer is integrated with an antenna and tested.

4.1 Quarter-Wave Rectiﬁer Operation

In this section, we discuss the rectiﬁcation process at low power for our proposed rectiﬁer for WPT applications. As wireless signals are both limited in power by the

FCC and face strong path loss, a focus has been on rectennas that eﬃciently rectify low RF power. For low power signals, we consider signals in the -20 dBm to 10 dBm range. Table 2.2 shows some comparisons of rectiﬁers that operate well in this range.

The quarter-wave rectiﬁer discussed in this section is simple and easy to understand, yet provides a very high eﬃciency using just a single diode.

4.1.1 Quarter-Wave Rectiﬁer Components

Figure 4.1: Basic quarter-wave rectiﬁer schematic

The proposed quarter-wave rectiﬁer consists of six main components. They are listed as follows.

47 (a) Antenna - Assumed to be present at this point, receiving a constant RF tone.

(b) Matching Network - Transforms the impedance of the antenna to the impedance

of the rectiﬁer. It is implemented with microstrip line matching stubs in this

case. It also serves to block any reﬂected signals or harmonics from propagating

backward towards the antenna.

simultaneously blocks the rectiﬁed DC signal from ﬂowing backwards.

(d) Rectiﬁcation Element - Placed in shunt to rectify the AC signal. Several diode

choices exist as the rectifying element, but here we ﬁnd that the HSMS2860

diode performs best in the 0-10 dBm range.

(e) RF Block/DC Pass Element - A quarter-wave microstrip line that transforms

the AC-short of the shunt capacitor (in parallel with the resistor) to an AC-

open. The impedance looking into the quarter-wave microstrip is open, meaning

the AC signal is reﬂected. As long as the matching network does not allow

the signal to propagate backward, the reﬂected signal will be remixed into the

diode. Thus, positive and negative cycles are rectiﬁed due to this component.

Additionally, harmonics also encounter a large mismatch while looking into the

quarter-wave transmission line.

(f) Load - Can be a sensor or battery, but in this case is modeled with a 2 kΩ

resistor.

As with any low power rectifier, several factors are capable of drastically changing the input impedance, and thus efficiency. The main influencing factors for input

48 impedance include: 1) rectiﬁer topology, 2) input power, 3) frequency, and 4) load.

As such, for this design, we define the rectifier topology as the quarter-wave rectifier,

the input power to be 8 dBm, the frequency to be set at one tone, and the load to be

2 kΩ. The optimized input impedance is determined using the source-pull technique

as deﬁned in Section 2.2.3. The ADS schematic is shown in Fig. 4.2a. As the input

impedance was being swept, a constant input power of 8 dBm was applied. We see

that in Fig. 4.2b, one input impedance creates an eﬃciency of over 80%.

As the Smith Chart in Fig. 4.2b shows, the impedance that creates the highest

eﬃciency is quite large, making it an additional challenge for matching. As such, great

care should be taken in fabricating this rectiﬁer due to its diﬃculty in matching.

4.1.2 Fabrication and Measurement of Quarter-Wave Recti- ﬁer

The quarter-wave rectiﬁer was then developed using microstrip lines. The sub-

strate used was Rogers 5880, with a dielectric of r=2.2, substrate thickness of 60

mils, 1 oz. cladding, and tan δ=0.0009. Additionally, a 68 pF inductor was added to the load in order to block the RF eﬀects caused by the voltage probes. The fabricated rectiﬁer board is shown in Fig. 4.3.

Since the diode can be diﬃcult to model at RF, some manual tuning was required for the matching network. This was accomplished by adding a copper tape open stub on the microstrip length before the series capacitor, or (b) in Fig. 4.3. Other than this, the rectiﬁer was measured without any changes.

To characterize the true rectification efficiency, the SMA-to-microstrip port needs to be de-embedded as well as the cable that feeds the rectifier. In order to de-embed the SMA port, various lengths of transmission lines were developed. These lengths

49 (a) ADS schematic for quarter-wave rectiﬁer using source pull

(b) Rectification efficiency for various input impedances on the quarter-wave rectifier

Figure 4.2: Quarter-Wave rectiﬁer, a) ADS schematic, and b) simulated source-pull eﬃciency

50 Figure 4.3: Fabricated quarter-wave rectiﬁer, all dimensions are in mm

are shown in Fig. 4.4. Each board utilized the same SMA-to-microstrip port as the fabricated quarter-wave rectiﬁer, as well as the same substrate material. As in [62], the loss of each transmission line length was measured as the transmission coeﬃcient,

|S21|, on a calibrated VNA. Once the losses were measured, a first order polynomial was fit to the slope of the measured losses. This graph is shown in Fig. 4.5. When the electrical length is interpolated at 0 degrees, we assume the insertion loss to be due only to the ports. In this case, the interpolated insertion loss is -0.087 dB at 0 degrees. Thus, if we assume the port losses to be symmetric, we divide this loss by two to find that the coaxial port reduces efficiency of the system by 0.98%.

51 Figure 4.4: Transmission lines designed for de-Embedding the SMA-to-Microstrip transition.

Next, we consider the loss of the cable. Doing a simple |S21| measurement of the

cable shows an insertion loss of -0.38 dB, or 8.37% reduction in eﬃciency. Now that

we know the insertion loss of the SMA-to-microstrip port and the cable, the total

eﬃciency can be summarized by Eq. 4.1.

2 V /Rload η = out,DC (4.1) Pin,RF × (1 − Lconnector) × (1 − Lcable)

In (4.1), the conversion eﬃciency is measured where Vout,DC is the output voltage across the load, Lconnector is the loss from the SMA-to-microstrip connector, Lcable is the loss due the cable, Rload is the load in Ohms, and Pin,RF is the peak amplitude of the input RF signal. The measured eﬃciency versus the optimally-matched simulation is shown in Fig. 4.6.

52 Figure 4.5: Interpolated insertion loss for various transmission lines

Figure 4.6: Measured conversion eﬃciency of single diode quarter-wave rectiﬁer

As shown in Fig. 4.6, the simulated and measured quarter-wave rectifier agree to within a few percent. The maximum measured efficiency is 73.7%, with over 60% efficiency at 0 dBm, and over 30% at -10 dBm. We note that this methodology of modeling and simulating can be performed for any rectifier at various frequency and power levels. This particular measurement and simulation utilized 1.8 GHz as the input tone. Now that we have analyzed the performance of this rectifier, we will compare its performance with other rectifiers in this power range.

53 4.2 Comparisons with Other Rectiﬁers

In this section we present a comparison of various low power rectification circuits, as well as the quarter-wave rectifier presented in Section 4.1. In theory, an infinite number of rectification circuits may exist, as any number of diodes and elements can be placed in any configuration to make a rectifier. However, power harvesting applications typically utilize circuits that rectify efficiently at low powers, implying smaller circuits with less components [63]. A visual representation of some these rectification circuits as well as our proposed quarter-wave rectifier circuit is illustrated in Table 2.2.

To create a fair comparison between rectifier circuit topologies, we must utilize the same input frequency, input power, and load. As such, we define the input frequency to be 2.4 GHz, the input power to be 8 dBm, and the load to be 2 kΩ for each test case. Additionally, the same diode model, capacitor, and inductor (when present) values were used, which were the HSMS 2860 diode, 0.1 µF capacitor, and 100 nH inductor, respectively. As discussed in Section 2.2.3, the ideal input impedance that gives the highest conversion efficiency can be determined via source-pull simulation.

Optimizing for the highest RF-to-DC conversion eﬃciency, Fig. 4.7 shows several common rectiﬁer topologies tuned for 8 dBm.

A brief description of each rectiﬁer is explained as follows.

(a) Bridge - Broadly used rectiﬁer for general AC-to-DC conversion. It merely

requires four diodes and a smoothing capacitor at the output. However, at each

diode intersection, a voltage drop occurs. As such, this rectiﬁer is ineﬃcient at

ultralow power levels.

54 Figure 4.7: Comparison of the quarter-wave rectiﬁer with other common rectiﬁers when optimized for 8 dBm

(b) I Doubler - Serves to duplicate the incoming current at the output terminals.

well at higher power levels.

(d) Half Wave - Consists of only one diode and one smoothing capacitor. Maximum

eﬃciency possible is 50%.

(e) Quarter-Wave - As described in Section 4.1.1, reﬂects all signals past the rec-

tiﬁer due to the quarter-wave transmission line. As a result, well over 50% is

possible using a single diode.

(f) Single Series - Duality of the single shunt rectiﬁer [63]. The single series diode

rectiﬁer creates an RF pass/DC block response with its shunt inductor. The

single diode performs all rectiﬁcation, and has a DC pass/RF block capacitor

at the output. Resistive losses are seen through the inductor.

55 (g) Single Shunt - Exists with all the essential elements required for rectiﬁcation

including: 1) RF pass/DC block capacitor, 2) rectiﬁcation element, and 3) RF

block/DC pass inductor. Again, resistive losses are seen through the inductor.

(h) V Doubler - Operates by exciting the upper (series) diode in the positive cycle,

and lower (parallel) diode in the negative cycle. It is a widely used rectiﬁer

circuit for power harvesting applications due to its high output voltage and

ease of matching.

The quarter-wave rectifier is similar to the voltage doubling rectifier, but it was found that by replacing the series diode from the voltage doubler with the quarter- wave transmission line, the voltage is no longer doubled, but the efficiency of the rectifier increases. As Fig. 4.7 shows, the quarter-wave rectifier provides the highest efficiency (greater than 85% in simulation) at 8 dBm when compared to all other rectification circuits. Another comparison is showed when the same rectifiers are tuned for the -10 dBm power range in Fig. 4.8. In this case, the quarter-wave rectifier is again the most efficient. This plot does not include the Greinacher rectifier since it was not efficient at these lower power ranges. This work is compared to other similar works in this band in Table 4.1.

4.3 Integration with Quarter-Wave Antenna

Since the quarter-wave rectiﬁer necessitates the presence of a microstrip line, this feature lends itself to integration of a microstrip patch antenna. Since traditional microstrip patch antennas are a half wavelength in size, we will utilize a shorting pin in order to realize a quarter-wave microstrip patch. Indeed, the quarter-wavelength

56 Figure 4.8: Comparison of the quarter-wave rectiﬁer with other common rectiﬁers when optimized for -10 dBm

Table 4.1: Comparison with similar rectiﬁers in literature

Comparison between this rectifier and other rectifiers in literature Ref. Input Power [dBm] Conversion Efficiency [64] 0 57% [65] 10 66.8% [66] 0 70.4% [67] 8 72.8% This work 8 73.7%

transmission line and the quarter-wave microstrip patch are one in the same in this design. The quarter-wave rectenna herein described includes all elements of the quarter- wave rectiﬁer as described in Section 4.1, except for the input capacitor. Interestingly, the antenna utilized in this section will used for both RF and DC operations.

Fig. 4.9 shows the quarter-wave shorted patch antenna, or quarter-wave rectenna, as an illustration as well as the fabricated model. The λ/4 microstrip acts as the antenna as well as a transmission line. Typical half wavelength patches operate on

57 (a) Illustration (b) Fabrication

Figure 4.9: Illustration and fabrication of the quarter-wave rectenna (units are mm)

the principle that the voltages on opposite ends of the patch are maximum, but with opposite polarity, and zero in the middle. Since the ﬁelds near the center of a patch antenna are low, a grounding pin can be placed here. A discussion on the shorting pin is found in [68] and [69]. A reduction in gain is expected, but the antenna still resonates and can be tuned for the frequency of interest.

The capacitor on the top end of the rectiﬁer in Fig. 4.9a acts as a short pin for

RF frequencies, since it is immediately grounded. The diode in this ﬁgure can be idealized as an open half of the time, and a short the other half. As such, we can now model the ﬁelds on the quarter-wave microstrip using HFSS. Fig. 4.10 shows the

ﬁeld distribution when it is aligned with a transmitter. The transmitter is located in the far-ﬁeld.

This figure shows that the voltage distribution on the quarter-wave patch antenna is greatest near the diode. Thus, the rectifier circuit is able to perform efficiently in this setup. The quarter-wave rectenna was fabricated on Rogers 5880 material, with a dielectric constant of r=2.2, loss tangent of tanδ=0.0009, thickness of 60 mil, and copper cladding of 1 oz. The width of the patch was chosen to provide a 50 Ohm

58 (a) Diode OFF (b) Diode ON

Figure 4.10: Field distribution when the diode is a) OFF (RF open) and b) ON (RF short)

impedance as a transmission line. The fabricated model is shown in Fig. 4.9b. The diode used was a Skyworks SMS7630, and the capacitor was 0.1 µF.

The simple circuit setup in Fig. 4.9 was simulated in ADS. For this simulation, the source was placed in parallel with the diode, since this is assumed to be the location of the greatest concentration of the fields. The simulated rectification efficiency of the quarter-wave rectenna is shown in Fig. 4.11. As can be seen from Fig. 4.11, the rectification efficiency peaks at approximately 30% at -7 dBm.

Figure 4.11: Simulated quarter-wave rectiﬁer eﬃciency

59 As opposed to traditional patches, this antenna's width is much narrower and its length is cut in half due to the shorting pin. As such, the simulated gain of this patch is lower than traditional patch antennas. The gain of the shorted quarter-wave microstrip is shown in Fig. 4.12. The maximum gain is approximately 0.5 dBi at 2.4

GHz.

Figure 4.12: Simulated gain of the shorted quarter-wave microstrip antenna when Open and Short (Diode OFF and ON, respectively))

4.3.1 Demonstrations and Measurements

In order to test the quarter-wave rectenna, a secondary patch array was used to transmit power. In this case, a signal generator was used as an RF power source. The patch array had a gain of 9 dBi. The receiver rectenna was placed in the radiating near-ﬁeld, which can be computed as 0.62pD3/λ, where D is the largest dimension, and λ is the wavelength. In this case, D was 17.5 cm and λ was computed at 2.4 GHz.

The signal generator was swept across various frequencies to evaluate the quarter-wave rectenna. An image of the experimental setup is shown in 4.13.

60 Figure 4.13: Quarter-wave rectenna demonstration

The transmitting patch array was set to 16.5 dBm, allowing for an overall EIRP of

25.5 dBm. The frequency was then swept, and the DC output power was measured.

A load of 1.8 kΩ was placed across the terminals.

The power rectiﬁed by the antenna is shown in Fig. 4.14. As can be seen from

Fig. 4.14, the patch antenna resonates properly at the desired frequency. The maximum power is delivered around the 2.4 GHz range, which corresponds to the resonant frequency of this rectenna.

Given the compact nature of this rectenna, being roughly a quarter-wavelength in size, it is envisioned that several of these rectenna can be easily mass-produced into an array. Additionally, since this rectenna only necessitates one diode and one capacitor, this provides a signiﬁcant reduction in complexity over previous designs.

61 Figure 4.14: Rectiﬁed output power from the quarter-wave rectenna

Further works include improvement to the resonating patch structure, the addition of a matching network, and integration into an array.

62 Chapter 5: Near-Field RF Harvesting for Medical Devices

5.1 Introduction to Near-Field Harvesting

Traditionally, most wearable or implantable medical devices (e.g., sensors, pacemakers, etc.) necessitate battery replacement [7]. However, this can be impractical or infeasible for hospital patients or the elderly. An example would be an insulin pump that would be embedded underneath the skin; the replacement of such a battery would be both an encumbrance to the patient, as well as a medical hazard due to the repetitive surgeries it would require. As an alternative, a wireless RF system may be used to safely and wirelessly transmit power to such a medical device [7]. Ad- ditionally, such an RF system would serve as a sanitary means of medical powering without the need for invasive operations.

Advances in antenna miniaturization, circuit design, and biocompatible materials are bringing forward new opportunities for unobtrusive diagnosis and patient care using wireless implantable devices [21]. It was reported that there are several million individuals using implanted medical devices like pacemakers [70]. With most of these devices, the typical method of powering them is to use Lithium batteries, although some use other methods of power generation, including piezoelectric, electrostatic,

63 ultrasonic transducers, and optical charging [7]. In this chapter, we propose a dif- ferent method for powering implantable sensors. This method is less sensitive to misalignment as it relies on using Wireless Power Transfer (WPT) in the radiated near-field to deliver power to a device on or in the body. Indeed, midfield wireless power transmission has been thoroughly investigated by Poon et al. as a viable means of WPT for medical applications [33]; however, instead of utilizing spirals and coils as the means of power reception, this work utilizes patch antennas in the radiating near-field.

Wireless powering of devices has been extensively researched since the early 1990s.

Chang et al. developed eﬃcient rectenna designs at low powers, achieving 82% for a

50-mW rectenna design [14][71]. Popovic and Hagerty researched the recycling of ambient RF signals, as well as developed modern rectenna matching techniques [51][72].

Olgun et al. demonstrated the wireless powering of a sensor using only ambient WiFi energy [15][17]. Costanzo et al. has been notable for her work in near-ﬁeld inductive links as well [73][74].

5.2 Near-Field vs. Far-Field Harvesting

Two approaches have been, generally, considered for WPT: 1) near-field coupling, and 2) far-field radiation. Near-field coupling operates on the principle of nearby magnetically coupled coils that resonate at low frequencies (0.3-30 MHz) [7]. Indeed, since the 1960s, the concept of wirelessly powered medical implanted has been explored [35]. But although inductive links have been capable of delivering high levels of power [75], misalignment and sensitivity to coil adjustments have been a major challenge [7]. Coil separations of 1-2 diameters tend to maximize the coil quality

64 factor [76] and lateral misalignment can cause severe degradations, limiting the eﬀec- tiveness when the coils are misaligned.

Far field radiation is appropriate when considering WPT from ambient signals, such as WiFi and television [15]. For far field applications, it is important to have high efficiency rectification circuits that turn-on even though incoming signals are of very low power, e.g. on the order of µW. Ambient power has been measured in the range from -20 to -40 dBm [17], depending on the distance from the RF source.

Several studies have been conducted on rectiﬁer eﬃciencies at the 2.4 GHz range [26].

However, most of these systems employ a single diode or voltage doubler as the rectifier; this work utilizes the voltage quadrupler presented in Section 2.4 that is tuned to have a maximum efficiency of 47.7% at 11 dBm and 40% at 0 dBm, which we demonstrate to deliver a sufficient voltage to operate an insulin pump wirelessly over 42 cm. Additionally, most midfield designs thus far have considered only spiral or coil antennas for transmission [34], [77], [78]; herein we present a patch antenna which provides a high tolerance to displacement while occupying a small footprint.

In this chapter, we propose a new method to wirelessly power sensor devices from near-ﬁeld radiating patch antennas. The proposed radiating method overcomes alignment issues between the exterior transmitter and receiver. This is achieved by relying on radiation, rather than direct coupling between the device and the RF source. To test the operation in the radiating near-ﬁeld region, a patch rectenna is developed and tested at 2.4 GHz.

This chapter will ﬁrst discuss rectenna design in Section 5.3. We will then perform a misalignment analysis in Section 5.4. Results and demonstrations using a medical

65 insulin pump are presented in Section 5.5. Finally, we summarize the ﬁndings of this chapter in Section 5.6.

5.3 Rectenna Design

Figure 5.1: Voltage quadrupling rectiﬁer topology

Previous works demonstrated wireless power transfer at WiFi frequencies for over- the-air wireless power transmission across large distances [79]. Herein, we present all components of a rectenna on a single PCB for operation at WiFi frequencies and in the radiating near-ﬁeld. The basic components of the proposed rectenna shown in Section 2.2.5 remain the same. The circuit schematic of this voltage quadrupling rectiﬁer topology is shown in Fig. 5.1. In summary, the elements of the overall rectenna include:

66 (a) Transmitting Element - This is typically placed outside the body and close to

the skin. It is generally a transmitting coil, but in our case we use a patch array

as the external transmitter in order to establish a proper radiating near-ﬁeld.

(b) Receiving Antenna - This is responsible for receiving the electromagnetic energy

wirelessly.

the rectiﬁer.

(d) RF Pass/DC Block Element - This passes the RF signal from the antenna to

the rectiﬁer, while blocking DC from ﬂowing backward.

(e) Rectiﬁcation Elements - These are ideally diodes, whose non-linear response is

capable of producing a DC component as well as various harmonics that are

ﬁltered.

(f) RF Block/DC Pass element - This is responsible for passing the DC signal to

the load while blocking the RF signal (modeled as a series inductor or shunt

capacitor).

(g) Output Load - This can represent a battery, sensor, medical device, etc., and is

usually modeled as a resistor.

The rectenna patch element was designed to operate at 2.4 GHz. For this proof- of-concept demonstration, the patch antenna occupied a footprint of 27.5 x 19.75 mm2 neglecting the surrounding substrate, and was fed by a 50 Ω source. Though outside the scope of this work, further miniaturization of the patch can be performed using high-permittivity substrates and meandering techniques [80]. For example, a

67 similar patch on a substrate with dielectric constant r=12.2 would shrink the overall size roughly 25%. When implanting directly beneath the skin, the high dielectrics of the body will provide a means of shrinking the patch down signiﬁcantly as well [68].

Additionally, the active circuitry may be folded onto the back of the patch, as in [81], to reduce the overall size by another 50%.

As shown in Fig. 5.1, the equivalent circuit of the rectiﬁer is shown as a two-stage voltage doubling rectiﬁer. An RF splitter was further added at the terminal of the patch element, to divide the RF signal between a right and a left matching branch.

The purpose of the two branches was for additional fine-tuning made possible by the open transmission line stubs. Using these tunable components, any frequency offset (due to the narrow bandwidth of the patch antenna) can be corrected. We note that this splitter is necessary for voltage quadrupling. Another such circuit is a Villard quadrupler, as in [82]; however, the Villard rectifier necessitates a balun, adding unwanted extra bulk.

Fig. 5.2 shows the simulated, measured, and theoretical output voltage from a voltage quadrupling rectifying circuit, as well as a simulated voltage doubler. The theoretical output voltage is simply four times the peak voltage of the input signal assuming 50 Ω input power. The simulated and measured traces for the quadrupler simply show the DC voltage across the output resistor. The simulated voltage doubler shows similar performance at lower powers; as can be seen, it fails to rectify eﬃciently at higher powers.

We remark that the insulin pump in this case requires a minimum of 1 V in order to oxidize. Indeed, the voltage quadrupler was chosen in order to provide the highest possible voltage to the implantable insulin pump. The insulin pump polymer, upon

68 Figure 5.2: Simulated, measured, and theoretical output voltage from voltage quadrupler, and simulated output from a voltage doubler

oxidation, expels cations causing the actuation arm to shrink. The higher the input power, the faster the actuation arm turns on. As such, the turn-on power of the implantable pump according to Fig. 5.2 is approximately 0 dBm.

The voltage quadrupler was fabricated on a Rogers TMM10 substrate having a dielectric constant r=9.2, and a loss tangent tan δ=0.0022. We note that the high dielectric permittivity was chosen to help realize a small rectenna size. The thickness of the substrate is 60 mil, and 0.1 µF was used for every capacitor. Further, the rectiﬁer diodes were Skyworks SMS7630 [38]. These diodes oﬀer excellent performance

69 (a) Illustration of the transmitting and (b) Lateral misalignment loss as a func- receiving antenna in the lateral misalign- tion of oﬀset distance when the vertical ment test separation is ﬁxed to 42 cm

Figure 5.3: Lateral misalignment tests

for the expected power range [45]. The load in this case is a 1.8 kΩ resistor, which

was found experimentally to be the approximate resistance of the insulin pump.

The rectiﬁer was then fabricated as a stand-alone unit and tested. This rectiﬁer

was essentially the same rectiﬁer shown in Chapter 2, Fig. 2.5. As shown, this rectiﬁer

achieved a maximum simulated eﬃciency of 55%, and a measured eﬃciency of 47.7%

at 11 dBm. Further, an eﬃciency of 40% was measured at 0 dBm.

5.4 Radiating Near-Field Antenna Measurements

The transmitting patch array measured approximately 34x33 cm2 while the receiving patch was the same dimension as the one shown in Fig. 5.1. The separation

70 distance was ﬁxed at 42 cm as shown in Fig. 5.3a. This distance was chosen because it places the receiving rectenna in the radiating near-ﬁeld, which occurs between

r D3 2D2 0.62 < x < (5.1) λ λ where D is the longest dimension of the antenna, and λ is the wavelength at 2.4

GHz [83]. In this case, D is assumed to be the distance from the ﬁrst patch to the last patch, as described in [84]; that is, the cut along which the patches are excited.

When the receiving antenna is a distance outside this range, then it will be in the far-ﬁeld.

The received power from the transmitting array was then measured as the receiving patch was swept in the x and y directions. Fig. 5.3b shows the loss due to the separation (42 cm) and lateral misalignments in the radiating near-ﬁeld. When misaligned by 15 cm, an additional 7.5 dB of path loss is added. A table with comparison to coil studies (of comparable size) found in literature is included in Table 5.1. We remark that similar or lower misalignment loss is seen for a smaller receiver size, making the radiated near-ﬁeld approach very suitable for on-body medical applications.

Table 5.1: Relevant coil misalignment studies in literature when separated by 15 cm

Coil Loss and Misalignment Studies in Literature Ref. Rx An- Separation Tx Size Rx Size Lateral tenna (cm) (cm2) (cm2) Loss (dB) [85] Coil 10 1017.9 7.1 11.3 [86] Spiral 18 254.5 63.6 4.2 This Work Patch 42 1122 5.4 7.5

71 5.5 Results and Demonstrations

To assess the performance of the given rectenna structure, a transmitter and re-

ceiver were connected to a controllable power supply, as shown in Fig. 5.4. A possible

clinical application is the case where the patient stands near the external transmitter

to activate the implanted pump. We note that the external transmitter is a linearly

polarized planar antenna with a nominal gain of 19 dBi, and the transmitting power

was set to 16.49 dBm. Thus, the entire EIRP was 35.49 dBm. This EIRP level is

just below the FCC limit for a point-to-point link of 36 dBm [87].

Figure 5.4: Radiating near-ﬁeld rectenna setup

The receiver rectenna was placed at a distance of 42 cm (see Fig. 5.4) and optimally angled and positioned to the transmitter, reducing the path loss and multipath eﬀects.

72 A multimeter was then connected to the output of the rectiﬁcation circuit across the

1.8 kΩ resistor placed in series and parallel set-up. The measurement is shown in

Fig. 5.5, and the goal was to extract the current and voltage across at the output of the rectiﬁer. We found that a constant current of 0.8 mA and voltage of 1.5 V was delivered using the wireless transmitter and rectenna receiver. As such, the amount of power delivered was 1.2 mW over this distance. A comparison of this work compared to other works in the midﬁeld that deliver power to a medical device is shown in

Table 5.2.

Figure 5.5: Measured voltage and current across a 1.8 kΩ load

The speciﬁc implantable pump was intended to squeeze a small pouch of insulin when located inside the body. The Polypyrrole (PPy) actuation arm is shown in

Fig. 5.6, and is submerged in Phosphate Buﬀered Saline (PBS) solution, which provides cations (mostly Na+) for the actuation. PBS is found naturally in the body,

73 Table 5.2: Comparison of the proposed versus reported midﬁeld WPT systems

Midﬁeld WPT Systems in Literature Ref. Rx An- Rx. Ant. Tx Dis- Power De- Source tenna Size (cm) tance livered Power [78] Coil 0.03 5 125 µW 0.5 W [77] Spiral 6.35 6.5 cm 4 mW 1 W [34] Loop 0.9 5-6 cm 173 µW 0.5 W [33] Coil 0.12 5 cm 200 µW 0.5 W This Work Patch 5.4 42 cm 1.2 mW 44.5 mW

and is a conductive ﬂuid in the blood stream. The left side of this ﬁgure shows the arm with no RF illumination, and on the right the actuated PPy arm is bending after the successful reception of the near-zone RF power illumination.

Figure 5.6: Actuated medical pump

5.6 Conclusion

A radiating near-field RF harvester (rectenna and source/transmitter configuration) was designed and built, showing up to 47.7% efficiency at 2.4 GHz. Using a

74 transmitting antenna/source located in the radiated near-ﬁeld zone of the rectenna, a total of 1.2 mW was delivered across 42 cm. A misalignment analysis was performed showing a maximum of 7.5 dB when oﬀset by 15 cm.

This type of wireless power delivery oﬀers a robust alternative to the position- sensitive coils, traditionally used for medical wireless power transfer. In the future, this rectenna design can be miniaturized even further by placing the rectiﬁer behind the rectenna on a multilayer substrate and implanting under the skin. The miniaturized rectenna circuit could then be embedded onto the insulin pump and integrated with all components in a single System-in-Package (SiP) medical device.

75 Chapter 6: Applications and Considerations for Medical Sensors

6.1 Miniaturizing Power Harvesting for Medical Applications

Patient monitoring in the hospital setting (heart rate, blood pressure, O2 levels, etc.) is a common, yet vital practice to ensure continuous patient care. However, these sensors necessitate either a cord or battery for continuous use, creating an in- creasingly complex array of cords that tether the patient to the monitoring apparatus.

Additionally, patient mobility becomes more complex as the person must travel with wires attached to speciﬁc devices. These wires reduce the quality of life for patients by limiting mobility, creating trip or choke hazards, causing discomfort or disconnections due to wire pulling, and slowing transfer between rooms.

To address this above issue, an electromagnetic solution is herein presented using a WPT device. The power is transformed from an external transmitter placed nearby the patient (i.e., bedside) and transferred to an on-body sensor, free from wires. No- tably, the majority of literature on medical WPT is for implantable sensors [7][21], such as pacemakers, drug pumps, or cochlear implants. These devices typically employ inductive coil receivers for WPT, as in [33]. However, coil misalignments as well as separations by more than one or two coil diameters can cause signiﬁcant loss of

76 WPT eﬃciency [86]. As an alternative, we propose a button-sized on-body WPT rectenna that receives power from a nearby transmitter, eliminating the need for excess wires traditionally used for patient monitoring.

In this chapter, Section 6.2 discusses miniaturization of the previous rectenna designs. Section 6.3 examines transmission of power through a skin layer. In Sec- tion 6.4, we discuss the SAR limitations of an implanted medical rectenna. Finally, we conclude with a summary in Section 6.5.

6.2 Miniature Rectenna Design

The proposed rectenna is a three layer device that includes, 1) a patch antenna, responsible for receiving the incoming RF power, 2) a shared ground plane on the other side of the substrate, and 3) a rectiﬁer that converts the RF signal to DC for powering the on-body sensor. The patch and rectiﬁer were fed together with wire bonding.

Antenna Design

The designed antenna was a small patch with slots, resembling an E-shaped patch.

It occupied a total footprint (including substrate) of 1.5 x 2.5 cm2. It was fabricated on Rogers TMM 10 substrate with a thickness of 60-mil and 1 oz. cladding. The ground of the patch antenna was a shared ground with the rectiﬁer. Additionally, the output of this patch antenna was wirebonded to the rectiﬁer on the bottom layer.

Rectiﬁer Design

A voltage doubling rectiﬁer was designed in Keysight ADS, and was chosen for its small size and ease of fabrication. The rectiﬁer used the SMS7630 diodes from

Skyworks. An input series- and output shunt-capacitor of 100 nF were used. Also,

77 an RF choke inductor of 68 nH was used at the output. Similar to the antenna, the rectiﬁer was fabricated on a 60-mil thick Rogers TMM 10 substrate with a footprint of 1.5 x 2.5 cm2.

Figure 6.1: Back (left) and front (right) of rectenna circuit with dimensions (in mm): a=25, b=15, c=12, d=2.25, e=11.25, f=1.25, g=3.3, h=2.1, i=3.8, j=3.8, k=3.8, l=0.66

The antenna and rectifier were fabricated (see Fig. 6.1) and measured. The antenna reflection coefficient, |S11| is shown in Fig. 6.2. The RF-to-DC conversion efficiency is shown in Fig. 6.3. The simulated gain was 0 dBi with 3-dB beamwidth of greater than 100 degrees. Also, the simulated SAR showed a maximum of 0.0175

78 W/kg averaged over 1 g of human tissue, and the maximum Eﬀective Isotropic Radi- ated Power of the transmitter was 35.5 dBm, satisfying IEEE and FCC regulations, respectively.

Figure 6.2: Miniature rectenna S11

For WPT misalignment analysis, the rectifier’s antenna was placed in the radiating near-field zone of the transmitter. This near-field distance was computed using

Eq. 5.1. Based on this formula, the minimum distance is d=65 cm, and d is the distance between the transmitter and receiver. To evaluate the WPT eﬃciencies, a 20 cm diameter area was swept, as shown in Fig. 6.4, and the path loss was measured.

As indicated in Table 6.1, this method is able to operate over much longer distances with a smaller receiver, while having minimal eﬀects due to misalignment.

This eﬀort culminated with a demonstration of powering a temperature sensor on a human phantom (see Fig. 6.5). In this case, the transmitter sent a 16.5 dBm signal (45 mW), nearly four times less than peak cell phone usage. This power was

79 Figure 6.3: Miniature rectenna eﬃciency

suﬃcient to wirelessly operate the rectenna and successfully activate the sensor. The sensor itself drew at least 20-45 µW.

Table 6.1: Relevant coil misalignment studies in literature compared to this work *Separation at 5 cm, and including system losses

Coil Loss and Misalignment Studies in Literature Ref. Separation Rx Size Tx Size Path Loss Lateral (cm) (cm2) (cm2) Loss [86] 18 63.6 254.5 -23.1 2.66 [85] 10 7.1 1017.9 -8.7* dB -3.7 dB This Work 65 3.75 1332.3 -22.5 dB -2.64 dB

80 Figure 6.4: Misalignment Analysis for Miniature Rectenna

Figure 6.5: Demonstration of WPT rectenna button turning on a temperature sensor on a body phantom.

81 6.3 Transmitting Power through the Skin

For implanted or subcutaneous sensors, there are enormous losses and challenges

with communication and transmission into the body. This is primarily due to the

high dielectrics of the skin and tissue, as well as the conductivity. In order to address

this issue, the losses associated with in-vivo wireless power transmission must be

addressed. Herein, we present a concept to enhance power delivery into the body using

a dielectric matching layer, or buﬀer, which helps transfer the free space radiating

impedance of the antenna into the impedance of the body.

Modeling of the skin is a particularly challenging pursuit mostly due to the various

environments in which the body can be involved. However, instead of modeling

local and minute portions of the body, the various layers of the body (skin, fat, and

muscle) can be modeled macroscopically, as shown in Thakor et al. [32]. The original

measurements and characterizations of human body measurements were conducted

by Gabriel et al. from 10 Hz to 20 GHz [88]. Similarly, some models have used simply a homogeneous material to simulate the implanted media, as in [89] where a

2/3 human muscle block is used. The dielectric and conductivity values of various human tissues are shown in [88], [90] and are presented for bulk muscle between 100

MHz and 5 GHz in Fig. 6.6.

As can be seen by 6.6a, the dielectric of bulk muscle is quite large compared to free space. It is from this vantage point that we present the concepts for a dielectric matching layer. To transmit as much energy as possible into the body, it is important to minimize reﬂections as the signal propagates. Similar to transmission line theory, reﬂections between two boundaries may be minimized by using a quarter wavelength

82 (a) Muscle dielectric (b) Muscle conductivity

Figure 6.6: a) Dielectric, and b) conductivity properties of bulk muscle

section. For propagating waves, we use a dielectric layer. Given two bulk dielectric

slabs with r,1 and r,2, the reﬂection coeﬃcient at the boundary can be described as

η2 − η1 Γ12 = (6.1) η2 + η1 √ where the wave impedance ηn can be deﬁned as 377/ r,n.

Assuming low reﬂection loss as stated in [91], the reﬂection from two boundaries can be described as

−j2βd Γ12 + Γ23 × e (6.2)

where Γ12 is the reflection between the first and second boundary, Γ23 is the reflection

between the second and third boundary, β is the propagation constant given by 2π/λ,

and d is the size of the matching layer. When d is set to λ/4 in thickness, and Eq. 6.2

is set to zero, the required dielectric value can be calculated. This concept is shown

in Fig. 6.7b and the improvement is shown in Fig. 6.7c.

83 (a) Illustration of quarter- (b) Quarter-wave matching wave matching in HFSS (c) Transmission loss

Figure 6.7: Illustrations and simulations of a quarter-wave dielectric matching layer

In this simulation, there is one external dipole and one internal dipole. As Fig. 6.7

shows, a dielectric matching layer does indeed add approximately 2 dB of improved

antenna coupling at 2.4 GHz.

Now when we consider the actual human body layer, the analysis introduces some

errors. These are primarly due to: 1) the large contrast between dielectric layers, and

2) the lossy nature of human tissue. However, we can consider the tissue layers as

a bulk medium with one eﬀective permittivity (tissue) to aid as a starting point for

developing a matching solution.

To model bulk human tissue in real life, we will use pig skin. Pig skin has an

approximate dielectric of r=25, and a loss tangent of tanδ=0.322. The thickness of the layer is roughly 5 mm. Using the pig skin model and the miniature patch designed in Section 6.2, we proceed to model the path loss when one of the antennas is embedded underneath pig skin.

The dielectric matching layer can be found by setting Eq. 6.2 to zero, which yields

√ 1tissue = 2 (6.3)

84 (a) No matching layer (b) Matching layer

Figure 6.8: Pig skin, a) without a matching layer, and b) with a matching layer (b)

where 1 = 1 (air) and tissue = 25 (pig skin). As such, the dielectric matching layer is

calculated to be 5. A very close option for this would be the Rogers 4003C material,

which has a dielectric of r=3.55, as well as a loss tangent of tanδ=0.0027. Utilizing

this material, matching sheets were made. The measurement setup with and without

the matching layer is shown in Fig. 6.8. √ The physical thickness of the matching sheet was calculated by λ0/(4 ∗ 3.55), which calculates to approximately 16 mm. The measured results are shown in Fig. 6.9.

As can be seen, an improvement of between 0.5 and 2 dB is seen across the entire band.

6.4 SAR Guidelines

Now that we have investigated the minimization of losses using wireless power transmission through the human body, we now must investigate the safety of this device to operate inside the human body.

85 Figure 6.9: Measured power coupling between internal/external antenna with and without a matching layer

This analysis begins with the deﬁnition of the Speciﬁc Absorption Rate (SAR),

which is the accepted measure the rate of energy deposited per unit mass of tissue [68].

The IEEE C95.1-2005 standard limits the SAR in 1 g of tissue to be less than 1.6

W/kg. This standard is to protect against adverse health eﬀects when humans are

exposed to electromagnetic ﬁelds between 3 kHz and 300 GHz. The SAR is deﬁned

σ|E|2 SAR = (6.4) ρ

where σ is deﬁned as the conductivity of the tissue (S/m), ρ is the mass density of the tissue (kg/m3), and E is the root mean square (rms) electric ﬁeld strength in the tissue (V/m).

Using Ansys HFSS, the ﬁelds inside human tissue can be simulated. Fig. 6.10 shows the averaged SAR in human tissue under two diﬀerent scenarios: 1) closely located, and 2) separated with matching layer. In each case, the amount of input

86 (a) On-Surface (b) Matching

Figure 6.10: SAR demonstrations while radiating when, a) antenna closely located to surface of skin, and b) antenna with dielectric matching layer

power was decreased until a SAR value of 1.6 W/kg was achieved. In the ﬁrst ﬁgure

(Fig. 6.10a) we see that the SAR is a maximum of 1.6 W/kg when the input power is 95 mW. This low input power is due to the intense electric ﬁeld created between the antennas at the surface of the skin. In the second ﬁgure (Fix. 6.10b) we see the matching layer. In this case, the SAR is 1.6 W/kg when in the input power is 297 mW. This is quite a large amount of power, which allows for a high transmission level into the body.

We conclude this study with a misalignment analysis. As discussed in Section 6.2, misalignment in coils causes signiﬁcant degradation in performance when separated by more than one or two coil diameters. We perform a study on lateral misalignment using the pig skin. As Fig. 6.11 shows, when the external antenna is displaced by one inch in both directions, the maximum additional loss is 6 dB.

87 Figure 6.11: Lateral misalignment loss in pig skin

6.5 Conclusion

In summary, we have introduced two concepts in this chapter: 1) the miniature power harvester, and 2) the dielectric matching layer. The miniature power harvester is shown to have a strong tolerance to misalignments, and it is able to work over large distances. This power harvester is ideal for medical applications. The dielectric matching layer is proven to add up to 2 dB of extra coupled power between the internal and external antenna. Such a device with its matching layer could be used for a myriad of medical sensors.

88 Chapter 7: Conclusion

7.1 Summary of Work

This work followed in the dreams and works of Nikola Tesla in the modern pursuit of wireless power for small devices and sensors. A summary of guidelines is given in

Chapter 2 for the young engineer making a power harvester. It summarizes various utilities of the rectiﬁers, matching network design, and demonstrates a rectenna recharging a super capacitor. Chapter 3 further expands on the work in Chapter

2 by creating an array of rectennas in order to receive power over longer distances.

In this work, power is transmitted in a room and a hallway setting over 60 meters.

An experiment is also developed in which a super capacitor is recharged over several meters, and a clock is turned on wirelessly. This chapter concludes with a method to widen the bandwidth of the rectenna by using a tapered oﬀset dipole antenna and a wideband input/output matching network.

In Chapter 4, an investigation into single diode operation is performed. A matching network and prototype is developed, with de-embedded SMA connectors. A maximum RF-to-DC conversion efficiency of 73.7% is measured with 8 dBm input power. We go on to compare this rectifier with other rectifiers, finding that indeed

89 the quarter-wave rectiﬁer is the most eﬃcient. This chapter concludes with the con-

cept of a quarter-wave rectenna. In this design, the antenna and the quarter-wave

microstrip line are one in the same. A demonstration of this technology is shown at

the end of the chapter.

Chapter 5 details principles for a near-ﬁeld RF harvester, in which the rectenna

is placed within the radiating near-ﬁeld of the transmitter. In this work, the same

rectenna as in Chapter 2 is utilized. We ﬁnd that using RF harvesters in the radiating

near-ﬁeld show strong tolerance to misalignment. A demonstration is shown in which

the rectenna wirelessly activates an insulin pump over a distance of 42 cm.

The goal of Chapter 6 was to miniaturize the rectenna found in Chapter 5 speciﬁ-

cally for medical applications. The overall dimensions of the rectenna is shrunk down

to 1.5 x 2.5 cm2 and designed to operate at 2.4 GHz. Misalignment analysis is per-

formed across a 20 cm x 20 cm area showning very strong tolerance to misalignment.

This rectenna was used to successfully activate a body-worn thermometer across 65

cm. At the latter end of this study, various aspects are presented in regards to wire-

less power transfer to implanted devices. The concept of a dielectric matching layer

for rectennas is presented and validated using pig skin. Finally, SAR requirements

are explained and simulated with and without a matching layer.

7.2 Outlook as a Technology

In the author's opinion, the outlook of wireless power transfer is immensely bright.

Given the trend in recent decades for smaller devices, lower power levels, and less wires, the next logical step seems to be WPT. The author envisions incremental steps

90 of technology, similar to how the internet was ﬁrst a luxury, but soon became a

common commodity in almost every venue.

Similar to the inductive phone charging stations that exist today, the author

envisions the next step of wireless power will come in close ranges. For example,

imagine a smart desk in which every device on the surface of the desk could be recharged wirelessly. Such a scenario is pictured in Fig. 7.1.

Figure 7.1: Smart desks concepts

Even in this simple ﬁgure, the calculator, watch, key fob, phone, mouse, and

keyboard could be recharged — that's six items! Such a technology would be ideal

for the short-ranged radiating near-ﬁeld WPT proposed in this work. Additionally,

91 the low power levels adhered to FCC standards, providing a good groundwork for safe WPT.

We continue by looking at some power consumption levels for various devices.

Fig. 7.2 shows the power requirements to operate a calculator in real-time, while

Fig. 7.3 shows the power to operate a thermometer with LCD display. As can be seen, the requirements of both of these devices are in the µW range, which is well within the limits of current WPT technology. The red circles in these plots show the points where the power was sampled.

Figure 7.2: Power requirements for a commercial calculator

This work is furthered by a demonstration in which the calculator tested in Fig. 7.2 is powered using two the rectenna receivers from Chapter 6. Using minimal amounts of transmitting power (45 mW), the calculator is able to be turned on over a few feet.

Enthusiasm has also grown in the commercial sector over the idea of 'cutting the cord' on phone charging. This industry may be the ﬁrst to invest and be inﬂuenced in this particular area.

92 Figure 7.3: Power requirements for a thermometer

7.3 Future Work

Future work for wireless power transfer will focus on several aspects of the design mentioned herein. The rectifier will continue to be a circuit that must be studied, understood, and designed. Indeed, matching and optimizing a rectifier circuit is still a technically challenging and very manual process. In the future, it is envisioned that all aspects of the rectifier, from the non-linear diode to the transmission lines, will be able to be placed in a simulator such as ADS and Momentum and optimized in a single click. Additionally, the operational power range of the rectifier is expected to go down further in the future.

93 Figure 7.4: Operational calculator using wireless power

Antenna design will always be an important aspect of WPT, as it is the receiving mechanism for this wireless power. Indeed, miniaturization and seamless integration of rectennas into everyday life will be an on-going work. Flexible and/or comformal antennas that could ﬁt into clothing or on an irregular surface would be important ergonomically.

Another area with limitless potential is the realm of higher frequencies. If a rectenna could be designed to operate at 600 THz (which is no simple task!), then light itself could be rectiﬁed and used to power electronics. Such a revolutionary technology could change the industry of solar panels and power generation in general as we know it.

94 Speaking of wireless power, Tesla writes [5], “All that was great in the past was ridiculed, condemned, combated, suppressed only to emerge all the more powerfully, all the more triumphantly from the struggle.” Towards that end, this work has taken steps to understand, utilize, miniaturize, and apply WPT technology to daily life via sensor and low power device integration. As this technology continues to develop, shrink, become more eﬃcient, and be used in more interesting ways, it will certainly be exciting to watch as it grows.

95 The author would like to thank Hewlett-Packard (HP) and the Juvenile Diabetes

Research Foundation (JDRF) for sponsoring various aspects of this research. An additional thanks goes out to Dr. Gregory Wainwright and Dr. Chi-Chih Chen for their work in developing the rectenna circuitry at the beginning of this study.

96 Bibliography

[1] M. Cheney, Tesla. Simon & Schuster, 2001. [2] N. Tesla, System of transmission of electrical energy, 1897. [3] ——, Apparatus for transmission of electrical energy, 1897. [4] ——, Apparatus for transmitting electrical energy, 1907. [5] ——, “The transmission of electrical energy without wires as a means for fur- thering peace,” Electrical World and Engineer, pp. 21–24, Jan. 1905. [6] A. B. Reeve, “Tesla and his wireless age,” Popular Electricity, vol. 4, no. 2, p. 97, Jun. 1911. [7] A. Amar, A. Kouki, and H. Cao, “Power approaches for implantable medical devices,” vol. 15, pp. 28 889–28 914, Nov. 2015. [8] A. Fisher, “Secret of perpetual ﬂight? beam-power plane,” Popular Science, vol. 232, no. 1, pp. 62–65, Jan. 1988. [9] W. Brown, J. Mims, and N. Heenan, “An experimental microwave-powered helicopter,” in 1958 IRE International Convention Record, vol. 13, Mar. 1965, pp. 225–235. doi: 10.1109/IRECON.1965.1147518. [10] W. Brown, “Experiments in the transportation of energy by microwave beam,” in 1958 IRE International Convention Record, vol. 12, Mar. 1964, pp. 8–17. doi: 10.1109/IRECON.1964.1147324. [11] R. Mahan. (2008). 1975 nasa jpl goldstone demonstration of wireless power transmission, [Online]. Available: https://c-sbsp.org/2008/03/10/1975- nasa-jpl-goldstone-demonstration-of-wireless-power-transmission/. [12] P. E. Glaser, “An overview of the solar power satellite option,” IEEE Transac- tions on Microwave Theory and Techniques, vol. 40, no. 6, pp. 1230–1238, Jun. 1992, issn: 0018-9480. doi: 10.1109/22.141356. [13] W. C. Brown, “Electronic and mechanical improvement of the receiving terminal of a free-space microwave power transmission system,” NASA Technical Reports, Aug. 1977.

97 [14] J. O. McSpadden, T. Yoo, and K. Chang, “Theoretical and experimental investigation of a rectenna element for microwave power transmission,” IEEE Transactions on Microwave Theory and Techniques, vol. 40, no. 12, pp. 2359– 2366, Dec. 1992, issn: 0018-9480. doi: 10.1109/22.179902. [15] U. Olgun, C. C. Chen, and J. L. Volakis, “Investigation of rectenna array con- figurations for enhanced rf power harvesting,” IEEE Antennas and Wireless Propagation Letters, vol. 10, pp. 262–265, 2011, issn: 1536-1225. doi: 10.1109/ LAWP.2011.2136371. [16] ——, “Efficient ambient wifi energy harvesting technology and its applications,” in Proceedings of the 2012 IEEE International Symposium on Antennas and Propagation, Jul. 2012, pp. 1–2. doi: 10.1109/APS.2012.6349384. [17] U. Olgun, “Efficient microwave energy harvesting technology and its applications,” PhD thesis, The Ohio State University, 2012. [18] M. Butcher. (2015). Nikola labs launches iphone 6 case which harvests electricity from the air, [Online]. Available: https://techcrunch.com/2015/05/04/ nikola-labs-launches-iphone-6-case-which-harvests-electricity- from-the-air/. [19] J. Rossignol. (2017). Energous says its truly wireless charging technology will ship later this year - in next iphone? [Online]. Available: https : / / www . macrumors.com/2017/01/04/energous-wireless-charging-2017. [20] A. P. Chandrakasan, N. Verma, and D. C. Daly, “Ultralow-power electronics for biomedical applications,” Annual Review of Biomedical Engineering, vol. 10, no. 1, pp. 247–274, 2008. doi: 10.1146/annurev.bioeng.10.061807.160547. [21] X. Wei and J. Liu, “Power sources and electrical recharging strategies for implantable medical devices,” vol. 2, pp. 1–13, Mar. 2008. [22] Unknown. (2011). Low-power, digital temperature sensor with smaart wire in- terface, [Online]. Available: http://www.ti.com/lit/ds/symlink/tmp104. pdf. [23] Z. Albus. (2015). Ultra-low-power motion detection using the msp430f2013, [Online]. Available: http://www.ti.com/lit/an/slaa283b/slaa283b.pdf. [24] 2018. [Online]. Available: https://www.nordicsemi.com/eng/Products/ Bluetooth-low-energy/nRF51822. [25] M. Clifford, “Proceedings sensor 2009,” AMA Service, 2009, pp. 187–192. [26] C. R. Valenta and G. D. Durgin, “Harvesting wireless power: Survey of energy- harvester conversion efficiency in far-field, wireless power transfer systems,” IEEE Microwave Magazine, vol. 15, no. 4, pp. 108–120, Jun. 2014, issn: 1527- 3342. doi: 10.1109/MMM.2014.2309499.

98 [27] H. Hu and S. V. Georgakopoulos, “Multiband and broadband wireless power transfer systems using the conformal strongly coupled magnetic resonance method,” IEEE Transactions on Industrial Electronics, vol. 64, no. 5, pp. 3595–3607, May 2017, issn: 0278-0046. doi: 10.1109/TIE.2016.2569459. [28] B. Choi, J. Nho, H. Cha, T. Ahn, and S. Choi, “Design and implementation of low-profile contactless battery charger using planar printed circuit board wind- ings as energy transfer device,” IEEE Transactions on Industrial Electronics, vol. 51, no. 1, pp. 140–147, Feb. 2004, issn: 0278-0046. doi: 10.1109/TIE. 2003.822039. [29] Y. Jang and M. M. Jovanovic, “A contactless electrical energy transmission system for portable-telephone battery chargers,” IEEE Transactions on Indus- trial Electronics, vol. 50, no. 3, pp. 520–527, Jun. 2003, issn: 0278-0046. doi: 10.1109/TIE.2003.812472. [30] X. Liu and S. Y. Hui, “Simulation study and experimental verification of a uni- versal contactless battery charging platform with localized charging features,” IEEE Transactions on Power Electronics, vol. 22, no. 6, pp. 2202–2210, Nov. 2007, issn: 0885-8993. doi: 10.1109/TPEL.2007.909301. [31] Z. Li, C. Zhu, J. Jiang, K. Song, and G. Wei, “A 3-kw wireless power transfer system for sightseeing car supercapacitor charge,” IEEE Transactions on Power Electronics, vol. 32, no. 5, pp. 3301–3316, May 2017, issn: 0885-8993. doi: 10.1109/TPEL.2016.2584701. [32] R. Jegadeesan, K. Agarwal, Y. X. Guo, S. C. Yen, and N. V. Thakor, “Wire- less power delivery to flexible subcutaneous implants using capacitive coupling,” IEEE Transactions on Microwave Theory and Techniques, vol. 65, no. 1, pp. 280–292, Jan. 2017, issn: 0018-9480. doi: 10.1109/TMTT.2016.2615623. [33] A. Ma and A. S. Y. Poon, “Midfield wireless power transfer for bioelectronics,” IEEE Circuits and Systems Magazine, vol. 15, no. 2, pp. 54–60, Secondquarter 2015, issn: 1531-636X. doi: 10.1109/MCAS.2015.2418999. [34] A. J. Yeh, J. S. Ho, Y. Tanabe, E. Neofytou, R. E. Beygui, and A. S. Y. Poon, “Wirelessly powering miniature implants for optogenetic stimulation,” Applied Physics Letters, vol. 103, no. 16, p. 163 701, 2013. doi: 10.1063/1.4825272. eprint: https://doi.org/10.1063/1.4825272. [Online]. Available: https: //doi.org/10.1063/1.4825272. [35] J. Ho and A. Poon, “Energy transfer for implantable electronics in the electromagnetic midfield (invited paper),” Progress In Electromagnetics Research, vol. 148, pp. 151–158, 2014. doi: http://10.2528/pier14070603. [36] D. M. Pozar, Microwave Engineering, Fourth. John Wiley & Sons, Inc, 2012. [37] M. Girardi. (Unknown). Commonly used taylor series, [Online]. Available: http: //people.math.sc.edu/girardi/m142/handouts/10sTaylorPolySeries. pdf.

99 [38] Skyworksinc.com. (2017). Skyworks : Products : Sms7630 series, [Online]. Avail- able: http://www.skyworksinc.com/Product/511/SMS7630_Series. [39] broadcom.com. (2018). Hsms-2860, [Online]. Available: https://www.broadcom. com/products/wireless/diodes/schottky/hsms-2860. [40] Unknown. (). Spice modeling of a diode from datasheet, [Online]. Available: http://www.youspice.com/spice-modeling-of-a-diode-from-datasheet/. [41] L. Teschler. (2015). When inductors self resonate, [Online]. Available: https: //www.analogictips.com/when-inductors-self-resonate/. [42] Unknown. (1999). Srf and prf for rf capacitors, [Online]. Available: https : //www.johansontechnology.com/srf-prf-for-rf-capacitors. [43] ——, (2017). Modelithics, [Online]. Available: https://www.modelithics. com/. [44] ——, (2013). What are impedance/esr frequency characteristics in capacitors? [Online]. Available: https://www.murata.com/en-us/products/emiconfun/ capacitor/2013/02/14/en-20130214-p1. [45] S. Hemour, Y. Zhao, C. H. P. Lorenz, D. Houssameddine, Y. Gui, C. M. Hu, and K. Wu, “Towards low-power high-eﬃciency rf and microwave energy harvesting,” IEEE Transactions on Microwave Theory and Techniques, vol. 62, no. 4, pp. 965–976, Apr. 2014, issn: 0018-9480. doi: 10.1109/TMTT.2014.2305134. [46] J. Wood, Behavioral Modeling and Linearization of Rf Power Ampliﬁers. Artech House, 2014. [47] P. Roblin, Nonlinear RF circuits and nonlinear vector network analyzers. Cam- bridge University Press, 2011. [48] Unknown. (). Agilent nvna and x-parameter simulation in ads, [Online]. Avail- able: https://www.keysight.com/upload/cmc_upload/All/1_NVNA_and_X- parameters_in_ADS.pdf. [49] (2015). Bq25570 nano power boost charger and buck converter for energy harvester powered applications, [Online]. Available: http://www.ti.com/lit/ds/ symlink/bq25570.pdf. [50] (2014). User’s guide for bq25570 battery charger evaluation module for energy harvesting, [Online]. Available: http : / / www . ti . com / lit / ug / sluuaa7a / sluuaa7a.pdf. [51] J. A. Hagerty, F. B. Helmbrecht, W. H. McCalpin, R. Zane, and Z. B. Popovic, “Recycling ambient microwave energy with broad-band rectenna arrays,” IEEE Transactions on Microwave Theory and Techniques, vol. 52, no. 3, pp. 1014– 1024, Mar. 2004, issn: 0018-9480. doi: 10.1109/TMTT.2004.823585.

100 [52] M. Pinuela, P. D. Mitcheson, and S. Lucyszyn, “Ambient rf energy harvesting in urban and semi-urban environments,” IEEE Transactions on Microwave Theory and Techniques, vol. 61, no. 7, pp. 2715–2726, Jul. 2013, issn: 0018-9480. doi: 10.1109/TMTT.2013.2262687. [53] (2018). Ecfr - code of federal regulations, [Online]. Available: https://www. ecfr.gov/cgi-bin/text-idx?SID=ba501a744f7008618a550f9c4d2553f1& mc=true&node=pt47.1.15&rgn=div5#se47.1.15_1247. [54] (2015). Scc-ares-races.org, [Online]. Available: https://www.scc-ares-races. org/mesh/doc/WiFi_Part_15_Power_Limits_v150424.pdf. [55] (2011). National telecommunications and information administration. (n.d.). united states frequency allocation chart., [Online]. Available: https://www. ntia.doc.gov/page/2011/united-states-frequency-allocation-chart. [56] A. Justice. (2010). 900 mhz: The wireless workhorse, [Online]. Available: http: //gridconnect.com/blog/general/900-mhz-the-wireless-workhorse/. [57] U. Olgun, C. c. Chen, and J. L. Volakis, “Design of an efficient ambient wifi energy harvesting system,” IET Microwaves, Antennas Propagation, vol. 6, no. 11, pp. 1200–1206, Aug. 2012, issn: 1751-8725. doi: 10.1049/iet-map.2012.0129. [58] B. DeLong, C. C. Chen, and J. L. Volakis, “Wireless energy harvesting for medical applications,” in 2015 IEEE International Symposium on Antennas and Propagation USNC/URSI National Radio Science Meeting, Jul. 2015, pp. 1213– 1213. doi: 10.1109/APS.2015.7304995. [59] V. Palazzi, M. D. Prete, and M. Fantuzzi, “Scavenging for energy: A rectenna design for wireless energy harvesting in uhf mobile telephony bands,” IEEE Microwave Magazine, vol. 18, no. 1, pp. 91–99, Jan. 2017, issn: 1527-3342. doi: 10.1109/MMM.2016.2616189. [60] M. Arrawatia, M. S. Baghini, and G. Kumar, “Broadband bent triangular omnidirectional antenna for rf energy harvesting,” IEEE Antennas and Wireless Propagation Letters, vol. 15, pp. 36–39, 2016, issn: 1536-1225. doi: 10.1109/ LAWP.2015.2427232. [61] J. Kimionis, A. Collado, M. M. Tentzeris, and A. Georgiadis, “Octave and decade printed uwb rectifiers based on nonuniform transmission lines for energy harvesting,” IEEE Transactions on Microwave Theory and Techniques, vol. 65, no. 11, pp. 4326–4334, Nov. 2017, issn: 0018-9480. doi: 10.1109/TMTT.2017. 2697851. [62] Z. Wang, “Electronic textile antennas and radio frequency circuits for body- worn applications,” PhD thesis, The Ohio State University, 2014. [63] T. Ohira, “Power efficiency and optimum load formulas on rf rectifiers featuring flow-angle equations,” IEICE Electronics Express, vol. 10, no. 11, pp. 20 130 230– 20 130 230, 2013. doi: 10.1587/elex.10.20130230.

101 [64] S. E. F. Mbombolo and C. W. Park, “An improved detector topology for a rectenna,” in 2011 IEEE MTT-S International Microwave Workshop Series on Innovative Wireless Power Transmission: Technologies, Systems, and Applica- tions, May 2011, pp. 23–26. doi: 10.1109/IMWS.2011.5877083. [65] D. Wang and R. Negra, “Design of a dual-band rectifier for wireless power transmission,” in 2013 IEEE Wireless Power Transfer (WPT), May 2013, pp. 127– 130. doi: 10.1109/WPT.2013.6556899. [66] B. R. Franciscatto, V. Freitas, J. M. Duchamp, C. Defay, and T. P. Vuong, “High-efficiency rectifier circuit at 2.45 ghz for low-input-power rf energy harvesting,” in 2013 European Microwave Conference, Oct. 2013, pp. 507–510. doi: 10.23919/EuMC.2013.6686703. [67] M. Roberg, T. Reveyrand, I. Ramos, E. A. Falkenstein, and Z. Popovic, “High- efficiency harmonically terminated diode and transistor rectifiers,” IEEE Trans- actions on Microwave Theory and Techniques, vol. 60, no. 12, pp. 4043–4052, Dec. 2012, issn: 0018-9480. doi: 10.1109/TMTT.2012.2222919. [68] A. Kiourti and K. S. Nikita, “A review of implantable patch antennas for biomedical telemetry: Challenges and solutions [wireless corner],” IEEE An- tennas and Propagation Magazine, vol. 54, no. 3, pp. 210–228, Jun. 2012, issn: 1045-9243. doi: 10.1109/MAP.2012.6293992. [69] P. J. Bevelacqua. (). Pifa - the planar inverted-f antenna, [Online]. Available: http://www.antenna-theory.com/antennas/patches/pifa.php. [70] M. A. Wood and K. A. Ellenbogen, “Cardiac pacemakers from the patient’s perspective,” Circulation, vol. 105, no. 18, pp. 2136–2138, 2002, issn: 0009- 7322. doi: 10.1161/01.CIR.0000016183.07898.90. eprint: http://circ. ahajournals . org / content / 105 / 18 / 2136 . full . pdf. [Online]. Available: http://circ.ahajournals.org/content/105/18/2136. [71] J. O. McSpadden, L. Fan, and K. Chang, “Design and experiments of a high- conversion-efficiency 5.8-ghz rectenna,” IEEE Transactions on Microwave The- ory and Techniques, vol. 46, no. 12, pp. 2053–2060, Dec. 1998, issn: 0018-9480. doi: 10.1109/22.739282. [72] J. A. Hagerty, “Nonlinear circuits and antennas for microwave energy conver- sions,” PhD thesis, University of Colorado, 2003. [73] C. Florian, F. Mastri, R. P. Paganelli, D. Masotti, and A. Costanzo, “Theoret- ical and numerical design of a wireless power transmission link with gan-based transmitter and adaptive receiver,” IEEE Transactions on Microwave Theory and Techniques, vol. 62, no. 4, pp. 931–946, Apr. 2014, issn: 0018-9480. doi: 10.1109/TMTT.2014.2303949.

102 [74] R. Trevisan and A. Costanzo, “A uhf near-field link for passive sensing in industrial wireless power transfer systems,” IEEE Transactions on Microwave Theory and Techniques, vol. 64, no. 5, pp. 1634–1643, May 2016, issn: 0018-9480. doi: 10.1109/TMTT.2016.2544317. [75] B. Lenaerts and R. Puers, “An inductive power link for a wireless endoscope,” Biosensors and Bioelectronics, vol. 22, no. 7, pp. 1390–1395, 2007, issn: 0956- 5663. doi: https : / / doi . org / 10 . 1016 / j . bios . 2006 . 06 . 015. [On- line]. Available: http://www.sciencedirect.com/science/article/pii/ S0956566306002892. [76] K. N. Bocan and E. Sejdic, “Adaptive transcutaneous power transfer to implantable devices: A state of the art review,” Sensors, vol. 16, no. 3, 2016, issn: 1424-8220. doi: 10.3390/s16030393. [Online]. Available: http://www.mdpi. com/1424-8220/16/3/393. [77] R. Das and H. Yoo, “A multiband antenna associating wireless monitoring and nonleaky wireless power transfer system for biomedical implants,” IEEE Transactions on Microwave Theory and Techniques, vol. 65, no. 7, pp. 2485– 2495, Jul. 2017, issn: 0018-9480. doi: 10.1109/TMTT.2017.2647945. [78] A. Abid, J. M. O’Brien, T. Bensel, C. Cleveland, L. Booth, B. R. Smith, R. Langer, and G. Traverso, “Wireless power transfer to millimeter-sized gastroin- testinal electronics validated in a swine model,” Scientific Reports, vol. 7, 2017. doi: 10.1038/srep46745. [79] B. DeLong, Q. Yuan, and J. L. Volakis, “Long range, safe power transmission using iteratively-tuned rectification,” in 2015 International Symposium on An- tennas and Propagation (ISAP), Nov. 2015, pp. 1–2. [80] J. L. Volakis, C.-C. Chen, and K. Fujimoto, Small antennas. McGraw-Hill, 2010. [81] A. Kiourti, C. W. L. Lee, J. Chae, and J. L. Volakis, “A wireless fully passive neural recording device for unobtrusive neuropotential monitoring,” IEEE Transactions on Biomedical Engineering, vol. 63, no. 1, pp. 131–137, Jan. 2016, issn: 0018-9294. doi: 10.1109/TBME.2015.2458583. [82] S. Salsabila and A. Munir, “1.8ghz printed bow-tie dipole rectenna with voltage quadrupler for rf energy harvesting,” in TENCON 2017 - 2017 IEEE Region 10 Conference, Nov. 2017, pp. 2739–2742. doi: 10.1109/TENCON.2017.8228327. [83] C. A. Balanis, Antenna Theory - Analysis and Design (3rd Edition), 3rd ed. John Wiley & Sons, 2005. [84] M. A. Salhi, A. Kazemipour, G. Gentille, M. Spirito, T. Kleine-Ostmann, and T. Schrader, “Near- and far-field characterization of planar mm-wave antenna arrays with waveguide-to-microstrip transition,” Journal of Infrared, Millimeter, and Terahertz Waves, vol. 37, no. 9, pp. 857–873, 2016. doi: 10.1007/s10762- 016-0273-x.

103 [85] M. Q. Nguyen, Z. Hughes, P. Woods, Y. S. Seo, S. Rao, and J. C. Chiao, “Field distribution models of spiral coil for misalignment analysis in wireless power transfer systems,” IEEE Transactions on Microwave Theory and Techniques, vol. 62, no. 4, pp. 920–930, Apr. 2014, issn: 0018-9480. doi: 10.1109/TMTT. 2014.2302738. [86] K. Fotopoulou and B. W. Flynn, “Wireless power transfer in loosely coupled links: Coil misalignment model,” IEEE Transactions on Magnetics, vol. 47, no. 2, pp. 416–430, Feb. 2011, issn: 0018-9464. doi: 10.1109/TMAG.2010. 2093534. [87] Unknown. (). Fcc rules and regulations, [Online]. Available: https://www. air802.com/fcc-rules-and-regulations.html. [88] S. Gabriel, R. W. Lau, and C. Gabriel, “The dielectric properties of biological tissues: Ii. measurements in the frequency range 10 hz to 20 ghz,” Physics in Medicine & Biology, vol. 41, no. 11, p. 2251, 1996. [Online]. Available: http: //stacks.iop.org/0031-9155/41/i=11/a=002. [89] P. Soontornpipit, C. M. Furse, and Y. C. Chung, “Design of implantable microstrip antenna for communication with medical implants,” IEEE Transactions on Microwave Theory and Techniques, vol. 52, no. 8, pp. 1944–1951, Aug. 2004, issn: 0018-9480. doi: 10.1109/TMTT.2004.831976. [90] R. D.Andreuccetti and C.Petrucci. (1997). An internet resource for the calcu- lation of the dielectric properties of body tissues in the frequency range 10 hz - 100 ghz, [Online]. Available: Based%20on%20data%20published%20by%20C. Gabriel%20et%20al.%20in%201996.%20[Online].%20Available:%20http: //niremf.ifac.cnr.it/tissprop/. [91] C. A. Balanis, Advanced Engineering Electromagnetics, 2nd ed. J. Wiley & Sons, 2012, p. 73.

104 Acronyms

AC Alternating Current. 12, 45, 47

ADS Advanced Design System. 13, 14, 17, 20, 41, 76

CLR Capacitor, Inductor, Resistor. 19

CMOS Complementary Metal Oxide Semiconductor. 21

DC Direct Current. 6, 11, 13–15, 17, 20–22, 26, 31, 34, 47, 56, 66, 68

DTV Digital Television. 28

EIRP Equivalent Isotropically Radiated Power. 30, 33, 34, 36, 60, 71

ESR Equivalent Series Resistance. 20

FCC Federal Communications Commission. 29, 46, 71, 77

GSM Global System for Mobile. 28

HFSS High Frequency Structure Simulator. 39, 57, 83, 85

IEEE Institute of Electrical and Electronics Engineers. 77, 85

IoT Internet of Things. 7, 34, 37, 44

105 IV Current-Voltage. 12, 21

LSSP Large Signal S-Parameter. 21, 41

PBS Phosphate Buﬀered Saline. 73

PCB Printed Circuit Board. 23, 25, 64

PPy Polypyrrole. 72, 73

RF Radio Frequency. 3, 11, 12, 14, 15, 17, 20–22, 45, 46, 56, 62, 64, 66, 73

RFID Radio Frequency Identiﬁcation. 4

SAR Speciﬁc Absorption Rate. 76, 85, 86

SiP System-in-Package. 74

SPICE Simulation Program with Integration Circuit Emphasis. xi, 14, 15, 21

SRF Self-Resonant Frequency. 19

UHF Ultra High Frequency. 37

VNA Vector Network Analyzer. 20, 50

VSWR Voltage Standing Wave Ratio. 18, 41

WPT Wireless Power Transfer. 1, 4, 5, 7, 9, 10, 23, 25, 28, 37, 46, 63, 64, 75, 76

106