Design and analysis of rectifying circuits and antennas for wireless power transmission and ambient RF energy harvesting

57 Figure 3.9 a Simulated and measured efficiencies and output DC voltages of the rectenna with a 2800-Ω resistor as the function of the input power level at 2.45 GHz.. 60 Figure 3.12 Me

DESIGN AND ANALYSIS OF ANTENNAS AND RECTIFYING CIRCUITS FOR WIRELESS POWER TRANSMISSION AND AMBIENT RF ENERGY

Acknowledgements

First of all, I would like to express my sincere gratitude to my supervisor, Prof Guo Yongxin, for his invaluable guidance and constructive support throughout my doctoral study Without his professional guidance and inspiration, this thesis would not be possible

I am deeply grateful to Dr Zhong Zheng, for his detailed and constructive comments and help on my works I also would like to thank Mr He Miao, who has been a collaborator of some works, and a co-author of some of my papers It was great pleasure to work with them

I would like to thank my friends in Microwave Research Laboratory, who have been very kind and supportive in my research life, especially Mr Agarwal Kush, Dr Bao Xiaoyue, Dr Bi Xiaojun, Dr Chu Hui, Dr Duan Zhu, Miss Lei Wen, Mr Liu Changrong, Mr Long Yunsheng, Miss Ren Rui, Dr Wang Lei, and Miss Xu Lijie

My graduate life at NUS would not have been fun and interesting without them Many thanks go to all the staff of Microwave Research Laboratory and ECE Department, especially Mdm Guo Lin, Mdm Lee Siew Choo, and Mr Sing Cheng Hiong for their kind assistances in all the technical and administrative support

My deepest appreciation goes to my family My parents gave much love, and their continuous encouragements are my great source of power, which enables me to overcome the frustrations in writing this thesis This thesis is dedicated to them Finally, I would like to thank all people who have helped and inspired me through

my doctoral study

Table of Contents

Acknowledgements i

Table of Contents ii

Summary v

List of Tables vii

List of Figures viii

List of Acronyms xii

Chapter 1 Introduction 1

1.1 Background and Motivation 1

1.2 Literature Review 5

1.2.1 Rectennas for Low-Input-Power Applications 5

1.2.2 Rectifiers With Wide Operating Input Power Ranges 9

1.2.3 60-GHz Rectennas 11

1.3 Thesis Outline 13

1.4 Original Contributions 17

1.5 Publication List 19

1.5.1 Journal Papers 19

1.5.2 Conference Papers 20

1.5.3 Patents 21

Chapter 2 Technology-Independent Table-Based Non-Linear Diode Model for Rectenna Design 22

2.1 Introduction of Nonlinear Circuits 22

2.1.1 Nonlinear Circuits 22

2.1.2 Methods of Nonlinear Circuits Analysis 24

2.1.3 Quasi-Static Assumption 26

2.2 Equivalent Circuit of the Schottky Diode 27

2.2.1 Basic Nonlinear Components in the Equivalent Circuit 27

2.2.2 Equivalent Circuit Model of a Schottky Barrier Diode 31

2.3 Non-Quasi-Static Table-Based Diode Model 35

2.3.1 Model Algorithm 35

2.3.2 Through-Reflect-Line (TRL) Calibration and Diode Measurement 41

2.3.3 Comparison Between Simulated and Measured Results 44

2.4 Summary 45

Chapter 3 Design of a High-Efficiency 2.45-GHz Rectenna for

Low-Input-Power Energy Harvesting 47

3.1 Introduction 47

3.2 Rectenna Design 48

3.2.1 Antenna Design 49

3.2.2 Rectifier Design 55

3.3 Rectenna Measurement 56

3.4 Summary 62

Chapter 4 A Dual-Band Rectenna Using Broad-Band Quasi-Yagi Antenna Array for Ambient RF Power Harvesting 63

4.1 Introduction 63

4.2 Quasi-Yagi Antenna Array 68

4.3 Rectifier Design and Measurement 72

4.4 Rectenna Measurement in the Ambience 78

4.5 Summary 80

Chapter 5 Expansion of Rectifier’s Operating Input Power Range for Wireless Power Transmission Applications 82

5.1 Introduction 83

5.2 First Expansion Method 85

5.2.1 Operation Mechanism 85

5.2.2 Experimental Results 93

5.3 Second Expansion Method 95

5.3.1 Operation Mechanism 95

5.3.2 Experimental Results 99

5.4 Performance Comparison and Discussion 100

5.5 Summary 102

Chapter 6 Study on 60-GHz Antennas and Rectifiers for Millimeter Wave Power Transmission 103

6.1 Introduction 104

6.2 60-GHz LTCC Linearly Polarized U-Slot Patch Antenna Array 107

6.2.1 Single Element 107

6.2.2 Antenna Array 111

6.2.3 Scattering Parameters 113

6.2.4 Radiation Patterns 115

6.3 60-GHz LTCC Circularly Polarized U-Slot Patch Antenna Array 117

6.3.1 Single Element 117

6.3.2 Antenna Array 119

6.3.3 Scattering Parameters 121

6.3.4 Radiation Patterns 125

6.3.5 Performance Comparison 131

6.4 60-GHz Rectifier Design 132

6.5 Summary 137

Chapter 7 Conclusion and Suggestions for Future Works 138

7.1 Conclusion 138

7.2 Suggestions for Future Works 141

Bibliography 143

Secondly, a high-efficiency 2.45-GHz rectenna which can harvest low input RF power effectively is developed In the design process, the antenna is co-designed with the rectifier and their matching performance is optimized at low input power points Measurement results have fully demonstrated that it can be used for WPT applications with low input power levels

Thirdly, a new rectenna is proposed It employs a broadband 1×4 quasi-Yagi antenna array and a dual-band rectifier for harvesting the ambient RF power at both global system for mobile communications (GSM) 1800 and universal mobile telecommunications system (UMTS) 2100 frequency bands The prototypes are developed and experimental results confirm the concept

Next, two innovative methods for extending the operating input power range of a traditional rectifier are studied In the first method, a field-effect transistor (FET) switch is utilized, so that the configuration of the rectifier can automatically adapt to the input power level In the second means, the breakdown voltage of the rectifier’s diode is enhanced while its built-in voltage is preserved by using a metal-oxide-semiconductor field-effect transistor (MOSFET) Compared to traditional rectifiers, both proposed rectifiers exhibit greatly enlarged operating input power ranges

Lastly, the possibility of rectennas operating at 60-GHz band is explored For the antenna part, a U-slot patch antenna is introduced at the 60-GHz band Based on that,

a linearly-polarized (LP) antenna array and a circularly-polarized (CP) of 4×4 elements are developed at 60-GHz band using the low temperature co-fired ceramic (LTCC) technology Good performances of both antenna arrays have been demonstrated through measurements on the probe station The rectifier is designed to fabricate on the printed circuit board (PCB) of Rogers-5880 Its capability of working around 60 GHz is verified by experimental results With proper wire-bonding technology, the antennas and the rectifier are promising to be integrated as 60-GHz rectennas

List of Tables

Table 1.1 Overview of Rectennas and Rectifiers in the Prior Literature 8Table 4.1 Measured Ambient RF Power Densities of Different Public Telecommunication Bands 65Table 5.1 Performance Comparison 101Table 6.1 Key Data of Several 60-GHz Circularly Polarized Antenna Arrays 131

List of Figures

Figure 1.1 Scenarios of wireless charging (a) Proximity charging (b) Distant

charging (c) Ambient RF energy harvesting 11

Figure 2.1 I-V characteristic of a nonlinear resistor 29

Figure 2.2 Equivalent circuit of a Schottky diode 32

Figure 2.3 Disagreement between simulation and measurement results caused by inaccurate diode model [57] 34

Figure 2.4 Model comparison (a) Equivalent circuit model (b) Table-based model 36

Figure 2.5 Circuit topology of the table-based model for simulation in ADS 38

Figure 2.6 Illustration of SDDs in ADS The left one has one port, while the right one has fourteen ports 39

Figure 2.7 Illustration of the parameters in the model and DAC 40

Figure 2.8 Simulation of the TRL standards in ADS 42

Figure 2.9 Insertion phase differences between the delay line and through versus frequency (a) The first delay line (b) The second delay line 42

Figure 2.10 Photographs of the fabricated standards for TRL calibration (a) Through (b) The first delay line (c) The second delay line (d) Open-ended line 43

Figure 2.11 Photograph of the diode on PCB for measurement 43

Figure 2.12 Measured and simulated DC I-V curves 45

Figure 2.13 Measured and simulated S11 under different voltage biases 45

Figure 3.1 Circuit configuration of the proposed rectenna 49

Figure 3.2 Configuration of the antennas (a) Top view of the proposed one (b) Side view of the proposed one (c) A two-element microstrip patch antenna array for reference 50

Figure 3.3 Input impedance of the antenna versus frequency for various stub length d 51

Figure 3.4 (a) Photograph of the fabricated antenna with transition and matching circuit (b) Configuration of the transition and matching circuit part 53

Figure 3.5 Simulated and measured |S11| of the antenna with transition and matching circuit 54

Figure 3.6 Simulated and measured radiation patterns at 2.45 GHz for the antenna in (a) E-plane and (b) H-plane 54

Figure 3.7 Current distribution on the antenna at 2.45 GHz 54

Figure 3.8 Rectenna measurement system 57

Figure 3.9 (a) Simulated and measured efficiencies and output DC voltages of the rectenna with a 2800-Ω resistor as the function of the input power level at 2.45 GHz (b) Photograph of the fabricated rectenna 58

Figure 3.10 (a) Simulated power level of each harmonic term at the point between the antenna and the rectifier circuit (b) Simulated power level of each harmonic term on the resistor 59

Figure 3.11 Influence of various insertion loss of the band-pass filter on the efficiency (The load resistance is 2800 Ω.) 60

Figure 3.12 Measured efficiency values of the rectenna as functions of, (a) input power and (b) power density with different resistors at 2.45 GHz 61

Figure 4.1 Illustration of the ambient RF Power measurement 65

Figure 4.2 Layout of the quasi-Yagi subarray (a) Top view (b) Side view 67

Figure 4.3 Simulated and measured |S11| and gain of the 1×2 subarray 68

Figure 4.4 Photographs of the 1×4 quasi-Yagi array (a) Top side (b) Back side 69

Figure 4.5 Simulated and measured |S11| and gain of the 1×4 quasi-Yagi array 70 Figure 4.6 Simulated and measured E-plane and H-plane patterns at (a) 1.85 and (b) 2.15 GHz for the 1×4 quasi-Yagi array 70

Figure 4.7 Topology of the proposed rectifier 72

Figure 4.8 Photograph of the fabricated rectifier 73

Figure 4.9 Measured return losses of the rectifier at different input power levels 73

Figure 4.10 Measured efficiencies of the rectifier against frequency at different input power levels 75

Figure 4.11 Measured efficiencies of the rectifier by using single and dual tones 76

Figure 4.12 Measured efficiency and output voltage of the rectifier against power density with dual-tone (1.84 and 2.14 GHz) input 76

Figure 4.13 Photographs of (left) the experimental setup for the rectenna measurement in the ambience and (right) the indication of the voltmeter. 77

Figure 4.14 Block diagram of RF energy harvesting system 80

Figure 5.1 Rectifier circuit topologies (a) Single shunt-mounted diode (b) Five shunt-mounted diodes (c) Five shunt-mounted diodes with one FET as a switch (d) Five shunt-mounted diodes with two FETs as a switch (N = 2) 85

Figure 5.2 Simulated efficiencies versus input power for the four different

rectifier topologies shown in Figure 5.1 87

Figure 5.3 Simulated efficiencies versus input power for the topology (N = 2) in Figure 5.1(d) with different load resistances 88

Figure 5.4 (a) Transfer characteristics of FETs with different threshold voltages (Vds = 1 V) (b) Efficiencies of the rectifiers use FETs with different threshold voltages (for topology of Figure 5.1(c)) 91

Figure 5.5 Circuit configuration of the proposed rectifier 92

Figure 5.6 (a) Photograph of the fabricated rectifier (b) Simulated and measured efficiencies and output voltages of the rectifier versus input power level 93

Figure 5.7 Rectifier circuit topologies (a) Single series-mounted diode (HSMS-2852 or HSMS-2860) (b) N series-mounted diodes (HSMS-(HSMS-2852) (c) Single diode (HSMS-2852 ) with a MOSFET (BF998) 96

Figure 5.8 Simulated efficiencies of the rectifiers in Figure 5.7 97

Figure 5.9 Circuit schematics of the proposed rectifier 99

Figure 5.10 Simulated and measured efficiencies and output voltages of the proposed rectifier against input power level 100

Figure 6.1 Geometry of the single element (a) Top view (b) Side view 108

Figure 6.2 Simulated |S11| for the single elements with and without U-slot 109

Figure 6.3 Geometry of the 4 × 4 U-slot patch antenna array 110

Fig 6.4 Geometry of GCPW to stripline transition (a) Top view (b) Side view (c) Zoom out view of the M1, M2 and M3 layers 110

Figure 6.5 Photographs of (a) fabricated antenna array and (b) S-parameter measurement on a probe station 111

Figure 6.6 Simulated and measured |S11| of the antenna array 112

Figure 6.7 Antenna test setup (a) Pattern measurement (b) Polarization study 113

Figure 6.8 Simulated and measured gain of the antenna array 113

Figure 6.9 Simulated and measured radiation patterns on (a) E-plane and (b) H-plane at 57 GHz 114

Figure 6.10 Simulated and measured radiation patterns on (a) E-plane and (b) H-plane at 60 GHz 114

Figure 6.11 Simulated and measured radiation patterns on (a) E-plane and (b) H-plane at 64 GHz 115

Figure 6.12 Geometry of the single element (a) Top view (b) Side view 118

Figure 6.13 Simulated |S11| and axial ratio of the single element 119

Figure 6.14 Geometry of the 4×4 U-slot patch antenna array 120

Figure 6.15 Geometry of GCPW to stripline transition (a) Top view (b) Side view (c) Zoom out view of the M1 and M2 layers 121

Figure 6.16 Photographs of the fabricated antenna array 121

Figure 6.17 Simulated (εr = 5.9) and measured |S11| of the antenna array 123

Figure 6.18 Simulated |S11| for εr = 5.9 and εr = 5.7, and measured |S11| 123

Figure 6.19 Illustration of the input impedance for the (a) simulation, and (b) measurement 125

Figure 6.23 Radiation patterns at 55 GHz Horn at 0o position: E-field of horn is in the y-direction; and horn at 90o position: E-field of horn is in the x-direction (a) xz-plane, horn at 0o (b) yz-plane, horn at 0o (c) xz-plane, horn at 90o (d) yz-plane, horn at 90o 128

Figure 6.24 Radiation patterns at 60 GHz (a) xz-plane, horn at 0o (b) yz-plane, horn at 0o (c) xz-plane, horn at 90o (d) yz-plane, horn at 90o 129

Figure 6.25 Radiation patterns at 65 GHz (a) xz-plane, horn at 0o (b) yz-plane, horn at 0o (c) xz-plane, horn at 90o (d) yz-plane, horn at 90o 129

Figure 6.27 Topology of the diode model in ADS 132

Figure 6.28 Circuit configuration of the proposed rectifier in ADS 133

Figure 6.29 Topology of the proposed rectifier 134

Figure 6.30 Simulated efficiency of the rectifier versus input power level 134

Figure 6.31 Simulated output DC voltage of the rectifier versus input power level 135

Figure 6.32 Photograph of the fabricated rectifier 135

Figure 6.33 Photograph of the experimental setup for the rectifier measurement 135

List of Acronyms

WPT Wireless power transmission

GSM Global system for mobile communications UMTS Universal mobile telecommunications system SPS Solar power satellites

PCE Power conversion efficiency

RFID Radio-frequency identification

UHF Ultra high frequency

JPL Jet Propulsion Laboratory

LP Linearly polarized

CP Circularly polarized

ISM Industrial, scientific, and medical

FET Field-effect transistor

MOSFET Metal-oxide-semiconductor field-effect transistor LTCC Low temperature co-fired ceramic

AR Axial ratio

GCPW Grounded coplanar waveguide

SDD Symbolically-defined device

DAC Data access component

VNA Vector network analyzer

PCB Printed circuit board

TRL Through-reflect-line

Chapter 1 Introduction

1.1 Background and Motivation

Wireless power transmission (WPT) is the transmission of electrical energy from a

power source to an electrical device without power cords It is especially useful in

cases where interconnecting wires are inconvenient, hazardous, or impossible

Basically, WPT can be divided into three types [1]: inductive, capacitive or resonant

reactive near-field coupling; far-field directive power beaming; and far-field

nondirective power transfer

In the first type, the resonance inductive coupling or electrodynamic induction is

the near-field wireless transmission between the two coils (primary coil and

secondary coil) that are tuned to resonate at the same frequency [2] Normally, the

coils are much smaller than the corresponding free space wavelength at the frequency

of power transfer Also, compared with the wavelength λ, the distance between the

two coils is small, which is usually smaller than 2D2/λ, where D is the diameter of the

coil This type of WPT benefits from several advantages: its non-radiative energy

transfer is safe for people and animals; its wastage of power is less; it does not

interfere with radio waves; and it provides relatively high efficiency through highly

resonant strong coupling However, the main drawback is that it can only operate in

the close proximity to the transmitter This type of WPT is mainly used for wireless

charging applications, which remove the need for multiple power sockets

The second type of WPT is to use a narrow-beam antenna or antenna array to

transmit the power in a well-defined direction toward the receiving antenna [3], [4]

Brown is a pioneer in this field who made important contributions to this emerging

technology He demonstrated a helicopter that received all the power needed for flight

from a microwave beam for up to 10 h [3] The high efficiency of this type of WPT

comes from the large-aperture antennas or antenna arrays so that most of the power

can be delivered to the rectifier for rectification This type of WPT is especially

useful for the space solar power satellites (SPS) systems [5] The SPS is like a high

power station located on the geostationary orbit The station uses large-scale solar

cells in space to generate electrical power The power is then transmitted in the form

of microwaves from the SPS to the rectenna site on the ground With SPS, large-scale,

environmentally clean power can be supplied

The third type of WPT is the far-field nondirective power transfer For such a WPT

system, it has two key features [1]: 1) the incident power densities are low; 2) the

position and orientation of the transmitting and the receiving antennas may vary and

are not particularly specified This research work focuses on this type of WPT

For the first feature of the third WPT type, since the incident power density is low,

a rectenna should have high RF-to-DC power conversion efficiency (PCE) at low

input power levels Also, the frequency bandwidth of the transmitted signal could be

narrowband, multiband or broadband, so that the rectenna should have the capability

for narrowband, multiband, or broadband operation accordingly To fulfill the needs,

the first of this research work intends to investigate on these different rectennas with

high RF-to-DC PCEs for low-input-power applications

On the other hand, for the second feature of the third WPT type, since the distance

between the transmitting and the receiving antennas could change, the input power

level of the rectification circuit would also vary Thus, rectification circuits that can

maintain high RF-to-DC PCE over variable power levels are required To meet this

requirement, the second aim of this research work is to design a new rectification

circuit which is able to perform high RF-to-DC PCE over a wide input power range

In addition, as the applications of radio-frequency identification (RFID) are

widespread, there is a tendency to move the operating frequency of RFID to the

extremely high frequency, e.g., 60 GHz [6–8] High operating frequency can benefit

the RFID from several aspects First, the data rate of the communications could be

significantly increased Also, at this frequency, a small directive antenna can be

mounted on a reader device to choose a transponder by pointing to it, which is not

possible for current ultra high frequency (UHF) RFID systems Among the passive

tag operation, the antenna and the rectifier are the two key components Since few

research papers report the rectennas at 60 GHz, the third objective of this research

work attempts to design an antenna and a rectifier at 60 GHz with our available

fabrication technology

For the system of the third WPT type, the RF-to-DC PCE is a critical factor There

is only one definition for the RF-to-DC PCE of the rectifier part, written as:

21

where VDC is the measured output DC voltage on the resistor, RL is the resistance

value, and Pin is the input power For the efficiency of the rectenna, there are three

definitions according to different measurement methods The first definition is made

by assuming the effective area of the rectenna as the geometric area AG,

2 1

1( 0 , 90 )

calculated RF-to-DC PCE is the conservative lower bound In the second definition,

the RF power delivered to the diode is obtained using the Friis formula,

2

2

21

4

DC L

where Pt, Gt, and r are the transmitted power, gain and distance between the

transmitter and receiver, respectively Gr is found from measurement or simulation of

an equivalent antenna without the rectifier Thus, this definition does not consider the

non-linear loading of the antenna by the feed, coupling between the rectifier and

antenna, mismatch and ohmic losses This definition involves many parameters and

small errors in r, Gr, Pt, Gr have a large effect The third definition, which is adopted

in this thesis, uses measured power density S at the rectenna plane and estimates the

antenna effective area Aeff through gain measurement or simulation,

2 3

As in (1.4), it does not take in to account the mismatch loss between the antenna and

rectifier Thus it overestimates the input power of the rectifier and the calculated

RF-to-DC PCE will be slightly lower than the true value

The following section will review prior research on rectennas for low-input-power

applications, rectifiers with wide operating input power ranges, and rectennas

operating at 60 GHz, respectively

1.2 Literature Review

1.2.1 Rectennas for Low-Input-Power Applications

The rectenna and its name were invented by W C Brown in the 1960s [9] Since

then, considerable research has been conducted on rectennas for WPT applications In

the 1970s, P E Glaser proposed the idea of SPS [10], which greatly promoted the

rectenna technology although it was not realized in the visible form Some

improvements on rectennas were made in this decade In 1974, R M Dickinson

constructed a power transmission system in laboratory with a high DC-to-DC

efficiency of 54% [11], which is a total efficiency including the source efficiency,

transmission efficiency, and the rectenna efficiency One year later, Jet Propulsion

Laboratory (JPL) and Raytheon demonstrated a large power transmission system in

the Mojave Desert with a one-mile transmission distance [12] In the same year, W C

Brown designed a huge rectenna array containing around 5000 elements at 2.38 GHz

[13] The RF-to-DC PCE of each rectenna element is as high as 82% In the 1980s,

rectenna’s ratio of the output power to the weight became an essential factor, since

more interests turned to improvements of rectennas for the aircraft applications In

1982, W C Brown and J F Triner produced a novel rectenna using thin-film printed

circuits [14] The ratio of the output power to the weight was further enhanced by W

C Brown using Kapton film in 1987 [15] At that time, although the rectenna

technology had been greatly improved, the input power levels or input power

densities for most of the works are high Normally, the input power level is more than

20 dBm and the power density is higher than 10 mW/cm2

From 1990, more and more works focused on rectennas with lower input power

levels or lower power densities In 1992, T W Yoo derived a closed-form equation

for RF-to-DC PCE to analyze the diode for the rectenna [16] Base on the theoretical

analysis, a rectenna was design and it performed RF-to-DC PCE of 40% with about

15-dBm input power at 10 GHz The rectenna was improved by the same research

group in 1998 [17], it showed RF-to-DC PCE of 80% at 5.8 GHz when the input

power is 15 dBm Till that time, a linearly-polarized (LP) antenna was mostly used

for the rectenna design After 2000, while the operating input power level kept

reducing, various kinds of antennas were applied for rectennas, e.g.,

circularly-polarized (CP) antennas [18]–[30], dual-frequency antennas [20], [31], [32],

harmonic-suppressing antennas [33]–[35], electrically small antennas [36], fractal

antennas [37], and so on

Normally, it is hard to compare the rectennas directly, since they work at various

power levels, and more critically, different efficiency definitions may be used in

different papers Table 1.1 lists some published rectennas which can operate at

relatively low input power levels The RF-to-DC PCEs listed in Table 1.1 are as

reported in the papers It can be seen from Table 1.1 that several rectennas can

perform satisfactory RF-to-DC PCEs at low input power densities, like rectennas in

[30], [38], [45], [47] However, as measured in [48], the average ambient RF power

densities of the three dominating public telecommunication bands (900,

GSM-1800, and UMTS-2100) are on the order of a few 0.01 µW/cm Hence, these

rectennas are actually incapable of harvesting the ambient RF energy effectively

Table 1.1 Overview of Rectennas and Rectifiers in the Prior Literature

Reference Freq

RF-to-DC PCE (%)

[23] 2-18 64 CP spiral elements 62 μW/cm2

20 [30] 2.45 Shorted annular ring-slot 10 μW/cm2

50

80 [36] 1.5754 Electrically small planar 4.51 dBm 80.4

57

1.2.2 Rectifiers With Wide Operating Input Power Ranges

Recently, there is a growing interest in WPT technology [3], which is quite useful

for wireless charging applications Typically, there are three scenarios for the

applications of wireless charging [49], [50] As shown in Figure 1.1, in the first

scenario, the wireless energy is transferred from a compact emitter in close proximity

to the sensor area Since the distance is short, the RF power level is usually high The

second situation is to use a high-gain transmitting antenna to send the RF energy from

a relatively long distance In this case, due to the path loss, the RF power level is

lower The third strategy utilizes the RF power present in the ambience and the

typical RF power level is very low [48] Therefore, it would be better for a rectenna

to achieve a high RF-to-DC PCE over a wide range of input power levels so that it

can operate well in the three scenarios discussed above

It is well known that among the WPT systems, a rectifier is crucial to enhancing

the transmission RF-to-DC PCE Various kinds of rectifiers [17], [18], [24], [37],

have been designed for WPT applications In well-matched rectifiers, the diode is the

most critical component to achieve high RF-to-DC PCEs since it is the main source

of loss [17] Proper diode selection for a rectifier design depends on the input power

levels Diodes with low threshold voltage are preferred for low-input-power

applications because a large portion of input power is consumed to overcome the

threshold voltage when the input power is low On the other hand, diodes with high

breakdown voltage are desirable for high-input-power applications since the

breakdown voltage limits the diode’s power handling capability Therefore, in order

to design a rectifier with wide operating input power range, diodes with low threshold

voltage and high breakdown voltage should be selected However, the threshold

voltage is directly related to the breakdown voltage through intrinsic properties of the

diode’s material and structure [5] For example, increasing the breakdown voltage

raises the threshold voltage Thus, due to the intrinsic nonlinearity of the diode, each

of those traditional rectifiers can only perform satisfactory RF-to-DC PCE in a

narrow input power range [49], [50] Once the input power level goes out of the

operating range, its PCE degrades very sharply This turns out to be a critical problem

to limit the wireless charging applications in which the input power level could

change significantly since high efficiency over a wide input power range is desirable

to greatly reduce the charging time In the prior literature, few works are tackled on

this problem In [49], [50], an attempt was made to overcome the limitation The

designed reconfigurable rectifier is able to collect RF power over a wide input power

range However, complicated detector and switch circuits are involved in the design

These circuits occupy a large size and are different to be designed and fabricated at

different operating frequencies Therefore, a new rectifier topology is required to

achieve a wide operating input power range as well as to simplify the design process

and relieve the requirement for advanced fabrication technology

Since the invention of a rectenna, considerable research has been performed on

rectennas for WPT applications Most of the development of RF power transmission

technology initially concentrated on the frequency of 2.45 GHz which is located in an

industrial, scientific, and medical (ISM) frequency band Since the attenuation by

atmosphere and scattering by rain are negligible at a frequency which is less than 3

GHz [17], the frequency of 2.45 GHz is considered to be a proper frequency for

power transmission between space-to-ground, ground-to-space, and ground-to-ground

[16] On the other hand, the operating frequency could be high for the space-to-space

applications so that compact antenna and rectenna could be used to increase the

distance of the power transmission However, only a few rectennas at high

frequencies are reported A 35-GHz rectenna [16] was designed with PCE of 39%

The rectenna used a microstrip dipole antenna and a Ka-band mixer diode In [51], a

rectenna element, a 1×2 rectenna array and a 2×2 rectenna array at 35 GHz were built

with PCE of 35% at the power density of 30 mW/cm2 In [52], a dual-band rectenna

which can operate at 35 GHz was developed in the CMOS 0.13-μm process

One successful application of WPT technology is the RFID system Nowadays,

there are some attempts to apply the RFID at the extremely high frequency, e.g., 60

GHz [6–8] In [6], the commonly used UHF RFID was compared with the RFID at 60

GHz It showed that although the remote powering is limited to a shorter distance,

very wide, even gigabit data bandwidth can be realized at 60 GHz A 60-GHz

harvesting RFID tag [7] is fabricated in Intel 90-nm CMOS technology The

power-harvesting got rid of the need of any extra battery while the use of high frequency

enabled the integration of an antenna and a rectifier circuit In [8], a reader module

with high data rate at 60 GHz for short-range communications was proposed For the

operation of the passive tag operation, the antenna and rectifier are the critical

components However, there is no report on 60-GHz rectennas This work put some

attempt on design of the antenna and rectifier for 60-GHz rectennas

1.3 Thesis Outline

This thesis consists of seven chapters Chapter 2 presents a novel non-quasi-static

table-based diode model for WPT applications This model is

technology-independent and is suitable for different diodes Without complicated curve fitting

and de-embedding process, this model is directly generated from S-parameters and

DC measurements A new way to deal with DC characteristic is proposed Excellent

fitness is achieved at both AC and DC fields This large-signal model can be used in

Agilent ADS software Simulated and measured results of a commercial diode prove

the accuracy of the model

Chapter 3 introduces a high-efficiency 2.45-GHz rectenna which can harvest low

input RF power effectively A new antenna with a simple structure and high gain of

8.6 dBi is proposed for the rectenna The antenna is designed to directly match the

rectifying circuit at 2.45 GHz and mismatch it at the second and third harmonics so

that the use of band-pass filter between the antenna and rectifying circuit can be

eliminated The rectenna shows a maximum conversion efficiency of 83% with a load

resistance of 1400 Ω Furthermore, the overall conversion efficiency can remain 50%

for the low, -17.2 dBm (corresponding power density 0.22 µW/cm2) input power

level

Based on the rectenna discussed in Chapter 3, a dual-band rectenna which can

harvest ambient RF power of GSM-1800 and UMTS-2100 bands efficiently is

proposed in Chapter 4 The novel rectenna is based on a broadband 1×4 quasi-Yagi

antenna array with bandwidth from 1.8 to 2.2 GHz, and high gains of 10.9 and 13.3

dBi at 1.85 and 2.15 GHz, respectively Also, a dual-band rectifier which can

sufficiently enhance the RF-to-DC PCE at ambient RF power level is designed for the

rectenna Measurement results show that a PCE of 40% and an output DC voltage of

224 mV have been achieved over 5-kΩ resistor when the dual-tone input power

density is 455 µW/m2 Additionally, output DC voltage varies between 300 to 400

mV can be obtained by collecting the relatively low ambient RF power

Chapter 5 focuses on the methods to expand the operating input power range of a

traditional rectifier Two methods are presented to realize the objective For the first

method, by utilizing a depletion-mode field-effect transistor (FET) switch, the

configuration of the rectifier can automatically adapt to the input power level

Compared with traditional rectifiers, it can provide a consistent high RF-to-DC PCE

over a significantly extended operating input power range Measured results show

that the PCE of this proposed adaptive rectifier keeps above 50% in the input power

range spanning from -14 up to 21 dBm Additionally, maximum PCE of more than 75%

is achieved in the input power range from 5 to 15 dBm For the second method, the

proposed rectifier utilizes a depletion-mode metal-oxide-semiconductor field-effect

transistor (MOSFET) to enhance the diode’s breakdown voltage while its built-in

voltage is preserved As a result, in comparison with traditional rectifiers, this

proposed rectifier can achieve a high RF-to-DC PCE in a much wider input power

range Measurement results show that a conversion efficiency of more than 50% can

be obtained over the input power range from -13.5 to 16.7 dBm, proving that this

rectifier is suitable for the wireless power transmission applications with varying

input power level

In Chapter 6, an attempt at a 60-GHz rectenna is described The two parts of the

rectenna, an antenna and a rectifier, are developed separately For the antenna part,

both LP and CP antennas are investigated The LP antenna designed is a 60-GHz

wideband U-slot patch antenna array of 4×4 elements on low temperature cofired

ceramic (LTCC) A U-slot patch antenna is used as the array element to enhance the

impedance bandwidth and a stripline feeding scheme with quarter wave T-junctions is

applied to feed each radiator Meanwhile, a grounded coplanar waveguide (GCPW) to

stripline transition with low insertion loss is designed for the antenna measurement on

the probe station The fabricated antenna array has a dimension of 15.5×17.5×0.9

mm3 The simulated and measured impedance bandwidths and radiation patterns are

investigated and compared Measured results show that the proposed antenna array

has a wide impedance bandwidth from 56.3 GHz to 64.3 GHz for |S11| < -10 dB

Additionally, it exhibits a peak gain of 17 dBi and a beam-shaped pattern with

reasonable 3-dB beam width of 20° On the other hand, the CP antenna designed is a

60-GHz wideband CP slot patch antenna array of 4×4 elements on LTCC A CP

U-slot patch antenna is used as the array element to enhance the impedance bandwidth

and a stripline sequential rotation feeding scheme is applied to achieve wide axial

ratio (AR) bandwidth Meanwhile, a GCPW to stripline transition is designed for

probe station measurement The fabricated antenna array has a dimension of

14×16×1.1 mm3 The simulated and measured impedance bandwidths, AR

bandwidths, and radiation patterns are investigated and compared Measured results

show that the proposed antenna array has a wide impedance bandwidth from 50.5

GHz to 67 GHz for |S11| < -10 dB, and a wide AR bandwidth from 54 GHz to 65.5

GHz for AR < 3 dB In addition, it exhibits a peak gain of 16 dBi and a beam-shaped

pattern with 3-dB beam width of 20o Moreover, its AR keeps below 3 dB within the

3-dB beam width For the rectifier part, a series-mounted diode rectifier is designed

Measurement proves that it can work around 60 GHz

Finally, the conclusion and suggestions for future work will be given in Chapter 7

1.4 Original Contributions

In this thesis, the following original contributions have been made:

1 A novel non-quasi-static table-based diode model for WPT applications is

proposed This model is technology-independent and is suitable for different

diodes It can be directly generated from S-parameters and DC measurements

without complicated curve fitting and de-embedding process The accuracy of

the model is proved through the simulated and measured results of a

commercial diode

2 A high-efficiency 2.45-GHz rectenna which can harvest low input RF power

effectively is developed The rectenna’s efficiency for low input power is

improved from two aspects Firstly, a high-gain yet compact antenna is

designed for the rectenna The use of a high-gain antenna is able to receive

more power for rectification to a great extent when the incident power is low

Secondly, the antenna is co-designed with the rectifier and their matching

performance is optimized at low input power points Thus, the novel rectenna

is sufficiently capable of recycling low RF power

3 A new rectenna with broad-band 1×4 quasi-Yagi antenna array and a

dual-band rectifier is designed to harvest the ambient RF power of GSM-1800 and

UMTS-2100 bands It has exhibited an output DC voltage varies between 300

mV to 400 mV, when measured in the ambience As many sensors monitor

physical quantities not frequently, their duty cycle of operation and average

power requirement are low Hence, with a suitable energy management circuit,

many sensor systems can be operational by harvesting ambient RF power using

this proposed rectenna

4 A novel reconfigurable adaptive rectifier is developed for WPT applications

The configuration of the rectifier can automatically adapt to the input power

level by utilizing an FET switch Compared with traditional rectifiers, it can

provide a consistently high RF-to-DC PCE over a significantly extended

operating input power range The RF-to-DC PCE of this proposed adaptive

rectifier can keep above 50% in the input power range spanning from -14 dBm

up to 21 dBm

5 A novel rectifier with extended operating input power ranges is proposed By

utilizing depletion-mode MOSFET, the breakdown voltage of the rectifier’s

diode is enhanced while its built-in voltage is preserved In comparison with

traditional rectifiers, this proposed rectifier can achieve a high RF-to-DC PCE

in a much wider input power range Measurement results show that a

conversion efficiency of more than 50% can be obtained over the input power

range from -13.5 dBm to 16.7 dBm, proving that this rectifier is suitable for the

wireless power transmission applications with varying input power level

6 For the first time, a U-slot patch antenna is introduced to the 60-GHz band

Based on the U-slot patch antenna, two antenna arrays (one is LP and the other

one is CP) of 4×4 elements are developed at 60 GHz using the LTCC

technology Both antenna arrays achieve wide operating frequency band The

CP antenna array performs good AR characteristic The designs take the

advantages of both traditional U-slot patch antenna and LTCC technology The

planar structure and the stripline feeding scheme provide a good solution for

integration applications

1.5 Publication List

The following publications are generated in the course of the research

1.5.1 Journal Papers

1 H C Sun, Z Zhong, and Y.-X Guo, “An Adaptive Reconfigurable Rectifier for

Wireless Power Transmission,” IEEE Microwave and Wireless Components

Letters, vol 23, no 9, pp 492–494, Sep 2013

2 H C Sun, Z Zhong, and Y.-X Guo, “Design of Rectifier With Extended

Operating Input Power Range,” IET Electronics Letters, vol 49, no 18, pp

1175–1176, Aug 2013

3 H C Sun, Y.-X Guo, M He, and Z Zhong, “A Dual-Band Rectenna Using

Broad-Band Yagi Antenna Array for Ambient RF Power Harvesting,” IEEE

Antennas and Wireless Propagation Letters, vol 12, pp 918–921, 2013

4 H C Sun, Y.-X Guo, and Z L Wang, “60-GHz Circularly Polarized U-Slot

Patch Antenna Array on LTCC,” IEEE Transactions on Antennas and

Propagation, vol 61, no 1, pp 430–435, Jan 2013

5 H C Sun, Y.-X Guo, M He, and Z Zhong, “Design of a High-Efficiency

2.45-GHz Rectenna for Low-Input-Power Energy Harvesting,” IEEE Antennas and

Wireless Propagation Letters, vol 11, pp 929–932, 2012

1.5.2 Conference Papers

1 Z Zhong, H C Sun, Y.-X Guo, “Design of Multi-channel Rectifier with High

PCE for Ambient RF Energy Harvesting,” 2013 International Symposium on

Antennas and Propagation (ISAP), Nanjing, China, Oct 23–25, 2013

2 H C Sun, Y.-X Guo, and Z Zhong, “A Novel Rectifier for Wireless Power

Transmission With a Wide Input Power Range,” 2013 IEEE International

Workshop on Electromagnetics (iWEM), Hong Kong, China, Aug 1–3, 2013

3 H C Sun, and Y.-X Guo, “Wideband 60-GHz LTCC Circularly Polarized Patch

Antenna With a U-Slot,” 2012 IEEE MTT-S International Microwave Workshop

Series on Millimeter Wave Wireless Technology and Applications (IMWS),

Nanjing, China, Sep 18–20, 2012

4 H C Sun, Y.-X Guo, and Z Zhong, “A High-Sensitivity 2.45 GHz Rectenna for

Low Input Power Energy Harvesting,” 2012 IEEE Antennas and Propagation

Society International Symposium (APSURSI), Chicago, USA, Jul 8–14, 2012

5 M He, H C Sun, Z Zhong, Y.-X Guo, and M Y Xia,

“Technology-Independent Table-Based Diode Model for Rectenna Design in RF Energy

Harvesting,” 2012 IEEE Antennas and Propagation Society International

Symposium (APSURSI), Chicago, USA, Jul 8–14, 2012

1.5.3 Patents

1 Y.-X Guo, Z Zhong, and H C Sun, “A Creative Non-Breakdown Rectifier

Topology for Wireless Power Transmission,” US Provisional Patent Filed, No.:

61/856,791

2 Y.-X Guo, Z Zhong, and H C Sun, “Adaptive Reconfigurable Architecture for

Wireless Power Transmission,” US Provisional Patent Filed, No.: 61/769,837

Chapter 2 Technology-Independent Table-Based Non-Linear Diode Model for Rectenna Design

The RF-to-DC PCE of the rectenna is one of the key factors among the whole

WPT and RF energy harvesting systems For the passive components (transmission

line, inductor, and capacitor etc.) of the rectenna, there are many accurate circuit and

electromagnetic models However, for the diode, its model is not easy to be

constructed due to its nonlinearity Inaccurate diode model will cause the

disagreement between the simulated and measured results of the rectenna Therefore,

an accurate diode model which can be easily built through measurement is very

important In this chapter, some classic methods for modeling of nonlinear devices,

their equivalent circuit models and drawbacks are presented Additionally, a new

non-static table-based diode model is proposed and verified by measurements

2.1 Introduction of Nonlinear Circuits

2.1.1 Nonlinear Circuits

All the circuits are nonlinear [53], while linear circuits are those of weak nonlinear

characteristics so that they can be analyzed using linear models The small-signal

amplifier is an example of linear circuits On the other hand, a mixer is an example of

nonlinear circuits, which fully utilizes the nonlinear characteristic It is well known

that solid-state devices are nonlinear Even for passive devices, such as resistors,

capacitors and so on, could be nonlinear under some special circumstances For

instance, the RF transition adapter could have insertion loss caused by

inter-modulation effect under the condition of high input power levels

According to the extent of nonlinearity, nonlinear circuits can be classified into

two categories: one is the strongly nonlinear circuits and the other one is the weakly

nonlinear circuits A weakly nonlinear circuit can be modeled by the Taylor series

expansions of its voltage over current (I/V), electric charge over voltage (Q/V) or

magnetic flux over current (Φ/I) functions at a bias point The condition of this

definition is that all the functions and their derivatives are continuous, and the DC

operating point does not change with time Almost all the transistors and passive

devices satisfy this nonlinear condition in normal operating range However, the

Schottky-barrier diode and the transistor under high power levels do not meet this

requirement, because their I/V-characteristics are exponential functions Weakly

nonlinear circuits can be analyzed by power series and Volterra series, while strongly

nonlinear circuits need to be investigated using harmonic balance and time-domain

methods

2.1.2 Methods of Nonlinear Circuits Analysis

For the simulation of nonlinear circuits, the following methods are mainly used:

load pull, large-signal S-parameters, time-domain analysis, and frequency-domain

analysis

The load pull method can be used to characterize the large-signal nonlinear circuits

For instance, the procedure of characterizing a power amplifier is: first, drawing the

curves of the load impedance versus the gain and output power in the Smith chart;

and then finding the optimal load impedance to balance the gain and output power

The main drawback of this method is that it only considers the influence of the load

impedance on the fundamental frequency of the circuit, while the effect on the

harmonics is omitted The modern load pull technology can include the influence of

the harmonics However, this measurement is limited by expensive equipment and

complicated measurement process

The large-signal S-parameter is based on the small-signal S-parameter The input

signal at the port is of high power level Since essentially the S-parameter is utilized

to characterize a linear system and the large-signal S-parameter expresses nonlinear

characteristic using linear methods, many problems may come out For example,

when a large-signal S22 of an FET is defined, a source should be applied at the output

port and then the reflected power should be measured For the transistor with strong

nonlinearity, when the excitation at the input port varies a lot, the reflected power at

output port will also change accordingly However, large-signal S-parameter assumes

the reflected power is stable Hence, some errors will be generated This disadvantage

could be overcome by X-parameter The X-parameter was proposed by Root et al in

the first decade of twenty-first century It can be used to model nonlinear black boxes,

nonlinear circuits and nonlinear systems With the capability of quick simulation and

good agreement with the measurement results, it is a novel tool of nonlinear analysis

The X-parameter measurement equipment produced by Agilent will play an

important role in the design of the nonlinear circuits

Time-domain analysis method is widely used to analyze the low-frequency analog

and digital circuits In the physical principle, it solves the differential equations of the

nonlinear circuit in the time domain The differential equations are nonlinear, but they

can be solved using numerical methods Time-domain analysis is not suitable for

high-frequency circuits and the circuits which need to be expressed in frequency

domain

Frequency-domain methods are broadly employed to analyze the nonlinear circuits

There are mainly two methods, harmonic balance and Volterra series analysis

Harmonic balance can be applied in analysis of single-frequency strongly nonlinear

circuits with high power excitation, such as power amplifiers, mixers and those

circuits using diodes and transistors On the other hand, Volterra series analysis can

be used for weakly nonlinear circuit with multiple small-signal excitations The

small-signal amplifier used at the receiving end is an example The diode model

proposed in this chapter can be utilized in harmonic balance simulation Also, the

rectifying circuits for WPT are analyzed by harmonic balance simulation

2.1.3 Quasi-Static Assumption

In the quasi-static assumption, all the nonlinear components will change with the

varying control voltage instantly without delay in time Some components have

memory property, like capacitor and inductor Their voltages depend on the preceding

voltage or current However, the capacitance, inductance and resistance of the

nonlinear components are independent on the previous voltage or current, and they

only depend on the existing control voltage The majority of the nonlinear models are

on the basis of this assumption

The quasi-static assumption is critical to the analysis of the linear and nonlinear

circuits Firstly, solid-static devices can be expressed by equivalent circuit using some

discrete components and then simulated Hence, they can be analyzed by method of

linear circuit analysis Secondly, the time delay in the silicon and gallium

semiconductors is only on the order of a few picoseconds, which can be neglected