Title High Power Microwave Wireless Power Transmission ...

Title High Power Microwave Wireless Power Transmission Systemwith Phase-Controlled Magnetrons( Dissertation_全文 )

Author(s) Yang, Bo

Citation 京都大学

Issue Date 2020-11-24

URL https://doi.org/10.14989/doctor.k22843

Right

“Experimental Study on a 5.8 GHz Power-Variable Phase-Controlled Magnetron”, IEICE Trans. Electron, Vol.E100-C,No.10, doi: 10.1587/transele.E100.C.901. “Evaluation of theModulation Performance of Injection-Locked Continuous-Wave Magnetrons”. IEEE Trans. ED, vol.66, no.1, doi:10.1109/TED.2018.2877204. Bo Yang, Tomohiko Mitani, andNaoki Shinohara, “A 5.8 GHz Phased Array System UsingPower-Variable Phase-Controlled Magnetrons for WirelessPower Transfer”.IEEE Trans. MTT,doi:10.1109/TMTT.2020.3007187. “Modeling andExperiments of an Injection-locked Magnetron with VariousLoad Reflection Levels”, IEEE Trans. ED, vol. 67, no. 9,doi:10.1109/TED.2020.3009901. “A High-efficiencyMicrowave Power Combining System based on Frequency-tuning Injection-Locked Magnetrons”, IEEE Trans. ED, inprint, 2020, doi: 10.1109/TED.2020.3013510

Type Thesis or Dissertation

Textversion ETD

Kyoto University

High Power Microwave Wireless Power

Transmission System with

Phase-Controlled Magnetrons

Bo YANG

October 2020

High Power Microwave Wireless PowerTransmission System with


Bo YANG

October 2020

i

High Power Microwave Wireless PowerTransmission System with


Bo YANG

Supervisor: Prof. Naoki Shinohara

Department of Electrical EngineeringKyoto University

This dissertation is submitted for the degree ofDoctor of Philosophy

October 2020

Acknowledgment

The present thesis summarizes the research while I am studying at the Department ofElectrical Engineering, Graduate School of Engineering, Kyoto University. I would like toacknowledge all the people supporting my doctoral study.

I would like to express my sincere and special appreciation to my supervisor, ProfessorNaoki Shinohara for giving me an opportunity to pursue the promising study. His valuablesuggestions and advice always indicated the guideline of my study. I sincerely thank ProfessorYoshiharu Omura for reviewing the thesis and suggesting the more clear explanations. I alsogreatly appreciate Associate Professor Yasuhito Gotoh for reviewing the thesis and advisingthe expression methods of the research results.

I am deeply grateful to Associate Professor Tomohiko Mitani for his insightful commentsand helpful suggestions through my whole research. I would like to show my gratitudefor his helpful suggestions. I would like to express my acknowledgments to ProfessorHirotsugu Kojima and Associate Professor Yusuke Ebihara, and Dr. Yoshikatsu Ueda fortheir encouragement and helpful suggestions at the monthly space group seminar.

I would like to thank Dr. Ce Wang, Mr. Xiaojie Chen, Mr. Jie Chu for overcomingdifficulties and completing the experiments together. Thanks to Katsumi Kawai for modifyingthe grammar of this thesis. I also would like to thank Mr. Seishirou Kojima, Mr. TakeshiNogi, Mr. Takashi Hirakawa, Dr. Yikai Hsieh and, all other researchers and students ofspace group of Research Institute for Sustainable Humanosphere (RISH) for their helpfulcomments and support. I thank to relevant departments supporting research and life duringthe COVID-19 pandemic.

I gratefully appreciate Kyoto University Special Admission Support Program for DoctoralStudent (KSPD) that providing one year scholarship.

I gratefully appreciate the financial support from Japan Society for the Promotion ofScience (JSPS) that could make me devote most of my time to this research. All theexperiments in the present study were conducted through a collaborative research program:Microwave Energy Transmission Laboratory (METLAB), at RISH, Kyoto University

Finally, thanks to Xi Yang and Niuniu for the greatest support.

Abstract

In this study, we are conducting research on high-power Wireless Power Transmission(WPT)systems that are deeply involved in microwave WPT technology. The constructed WPTsystems with injection-locking magnetrons as the DC–RF converter is aim to practical use.The magnetron are widely used in microwave oven. Magnetron has practical advantagessuch as high efficiency, large power, low cost, and light weight, comparing to semiconductorsDC–RF converters. On the other hand, it is difficult to realize phase control of the magnetron,and it has a drawback that it is noisy, so it was mainly used for microwave heating microwaveovens that do not consider the influence of noise. However, it is necessary to control the phaseof the magnetron for the WPT system which involved in controlling the power transmissiondirection.

To solve this problem, phase-controlled magnetrons methods realized from microwaveoven magnetrons have been developed. On the contrary, it is difficult to control the mag-netrons output power through the developed phase-controlled methods. In a bid to solvethis problem, this thesis developed a 5.8 GHz power-variable phase-controlled magnetron(PVPCM) which controls the phase of magnetron output by a phase shifter and controls thepower by the anode current of the magnetron. This method is different from the previous2.45 GHz phase-controlled magnetron which utilizes an injection method and a phase lockedloop by the anode current, since the frequency of 5.8 GHz magnetron hardly changes withthe anode current. Our experiments show that the developed 5.8 GHz PVPCM had a variableoutput power with 1% power stability from 160 W to 329 W, the phase accuracy was nearly±1◦, and the response time was less than 100 `s. Stable output power, high phase-controlledaccuracy, and fast response speed microwave sources based on the PVPCMs are suitable forphased array system for microwave WPT.

A phased array system with four developed PVPCMs was bulid by applying the injection-locking method and phase-locked-loop method. To reduce the cost and ensure the durabilityof the phased array, a waveguide slot array antenna was designed and used for the outputantenna of power-variable phase-controlled magnetrons. The slot antenna has an expectedangle deflection of 22.5◦, a gain of 24.9 dBi, and the half bandwidth of the main lobe was

10◦. The experiments demonstrated the properties of microwave beam forming and WPTbased on the magnetron phased array system. In horizontal directions, a beam scanningrange of ±3◦ was obtained by adjusting the output phase of the magnetrons. Furthermore, thereceived DC power reaches 142 W at a distance of 5 m when the output microwave power ofthe magnetron phased array is 1304 W.

This thesis also proved that 2.45 GHz and 5.8 GHz band continuous-wave magnetrons canbe used to perform amplitude, phase, and frequency modulations by applying an injection-locking method. The magnetron behaved like an amplifier, and its output could follow theinjection signal. In addition, we have achieved the transmission of amplitude shift keying dataat 200 kb/s as well as phase-shift keying and frequency-shift keying at 10 Mb/s. Moreover,we quantitatively discussed several demodulation performances of the injection-lockingmagnetrons. Finally, the transmission of audio and video information was demodulated usingthe injection-locking magnetrons.

Two low cost power supplies were improved. The improved low-cost power suppliesapplied to phase-controlled magnetron system, modulation magnetron system, magnetronload fluctuation experiment and magnetron output combining system.

Finally, the thesis introduced the wireless information and power transmission systemsand highly efficient WPT systems. The wireless TV transmission distance was about 3.5m, and we succeeded in developing a wireless TV system. The constructed 5.8 GHzmagnetron wireless power transmission system receives the received power of 109 W withthe transmission distance of 5.6 m, and the total efficiency of the system reaches up to 9.73%.We discussed for constructing a high-efficiency microwave wireless power transmissionsystem with higher efficiency.

Table of contents

List of figures ix

List of tables xv

1 Introduction 11.1 Wireless Power Transmission technologies . . . . . . . . . . . . . . . . . . 1

1.1.1 History of Wireless Power Transmission . . . . . . . . . . . . . . . 11.1.2 Applications of Wireless Power Transmission . . . . . . . . . . . . 61.1.3 Theory of Wireless Power Transmission . . . . . . . . . . . . . . . 9

1.2 High Power Microwave Wireless Power Transmission Systems . . . . . . . 121.3 High Power Microwave Transmitter . . . . . . . . . . . . . . . . . . . . . 15

1.3.1 GaN HEMT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151.3.2 Klystrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161.3.3 TWT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171.3.4 Gyrotrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171.3.5 Magnetrons & CFA . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.4 Outline of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2 Power-Variable Phase-Controlled Magnetron 212.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.2 Injection-Locking Magnetron . . . . . . . . . . . . . . . . . . . . . . . . 222.3 Design of a PVPCM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2.3.1 Characteristics of 5.8 GHz Magnetrons . . . . . . . . . . . . . . . 262.3.2 Configuration of PLL . . . . . . . . . . . . . . . . . . . . . . . . . 282.3.3 Determination of Transfer Functions . . . . . . . . . . . . . . . . . 29

2.4 Measurement Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 312.4.1 Outline of Experiments . . . . . . . . . . . . . . . . . . . . . . . . 312.4.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . 32

2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3 Magnetron Phased Array 393.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393.2 Principle of Power-Variable Phase-Controlled Magnetrons and Magnetron

Phased Array System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403.3 Design of Slot Array Antennas . . . . . . . . . . . . . . . . . . . . . . . . 433.4 Demonstration Experiments of Magnetron Phased Array . . . . . . . . . . 49

3.4.1 Beam Forming Experiments . . . . . . . . . . . . . . . . . . . . . 513.4.2 Wireless Power Transfer Experiments . . . . . . . . . . . . . . . . 55

3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

4 Modulation Performance with Injection-Locking Magnetrons 654.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654.2 Magnetron Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 664.3 Modulation Performance Evaluation Experiments . . . . . . . . . . . . . . 69

4.3.1 Injection-Locking Magnetron for ASK . . . . . . . . . . . . . . . 694.3.2 Injection-Locking Magnetron for PSK . . . . . . . . . . . . . . . 724.3.3 Injection-Locking Magnetron for FSK . . . . . . . . . . . . . . . 75

4.4 Discussion of the Modulation Performance . . . . . . . . . . . . . . . . . 754.5 Demonstration Experiments . . . . . . . . . . . . . . . . . . . . . . . . . 80

4.5.1 Demonstration of a Phase-Controlled Magnetron for Transmittingan Audio Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

4.5.2 Demonstration of an Injection-Locking Magnetron for Transmittinga Video Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

4.5.3 Demonstration of an Injection-Locking Magnetron for transferring data 834.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

5 Power Supplies of the Magnetron 895.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895.2 Full Wave Doubler Rectifier . . . . . . . . . . . . . . . . . . . . . . . . . 89

5.2.1 Injection-Locking Magnetron with Full Wave Doubler Rectifier . . 945.2.2 Modulation Magnetron with Full Wave Doubler Rectifier . . . . . . 965.2.3 Phase-controlled Magnetron with Full Wave Doubler Rectifier . . . 99

5.3 Switched-Mode Power Supply . . . . . . . . . . . . . . . . . . . . . . . . 1035.3.1 Magnetron Various Load Reflection Experiments . . . . . . . . . . 1055.3.2 Magnetrons Output Power Combining Experiments . . . . . . . . . 109

5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

6 Microwave Wireless Power Transmission Systems 1156.1 Information Transmission System with Beam Forming . . . . . . . . . . . 1176.2 Wirelessly-Powered TV . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

6.2.1 Experimental Arrangement . . . . . . . . . . . . . . . . . . . . . 1196.2.2 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . 121

6.3 Charging in the Microwave Oven . . . . . . . . . . . . . . . . . . . . . . . 1226.4 Efficiency Measurement of the WPT System . . . . . . . . . . . . . . . . . 125

6.4.1 Magnetron WPT System with Horn Antenna . . . . . . . . . . . . 1256.4.2 Magnetron WPT System with Patch Antenna . . . . . . . . . . . . 126

6.5 Discussion of a High Efficiency WPT System . . . . . . . . . . . . . . . . 1306.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

7 Conclusion 133

References 135

Publication List 145

List of figures

1.1 History of wireless power transmission(2-1). . . . . . . . . . . . . . . . . . 41.2 History of wireless power transmission(2-2). . . . . . . . . . . . . . . . . . 51.3 Applications of Wireless Power Transmission . . . . . . . . . . . . . . . . 81.4 Classification of wireless power transfer. . . . . . . . . . . . . . . . . . . . 91.5 WPT Research achieved efficiency with distance. . . . . . . . . . . . . . . 111.6 RF vacuum devices (black line) [52] in 1999 and solid-state devices (color

area) [53] in 2014 state of technology for single device performance. . . . . 161.7 Principle diagram of Magnetron. . . . . . . . . . . . . . . . . . . . . . . . 19

2.1 Schematic of the injection locked magnetron. . . . . . . . . . . . . . . . . 232.2 Injection locked spectrum and free-running spectrum of a 5.8 GHz Mag-

netron in max hold mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.3 Relationship of injection power and lock frequency range on 5.8 GHz mag-

netrons with difference serial numbers. . . . . . . . . . . . . . . . . . . . . 252.4 Phase difference between the reference signal and the injection locked mag-

netron output. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252.5 Relationship of anode current and oscillation frequency on 5.8 GHz mag-

netrons with difference serial numbers. . . . . . . . . . . . . . . . . . . . 262.6 Relationship of anode current and output power on 5.8 GHz magnetrons with

difference serial numbers. . . . . . . . . . . . . . . . . . . . . . . . . . . 272.7 Schematic diagram of a 5.8 GHz PVPCM. . . . . . . . . . . . . . . . . . 282.8 Block diagram of a phase control loop. . . . . . . . . . . . . . . . . . . . . 292.9 Diagram of a low pass filter and a phase control circuit for a 5.8GHz PVPCM. 302.10 Bode diagrams of open-loop transfer function for a PLL. . . . . . . . . . . 312.11 Experimental schematic diagram of a 5.8 GHz PVPCM. . . . . . . . . . . 322.12 Photo of a 5.8 GHz PVPCM system (1: 5.8 GHz magnetron, 2: directional

coupler, 3: dummy load, 4: current meter, 5: circulator, 6: spectrum analyzer.) 33

2.13 Magnetron phase difference when the PLL circuit starts working measuredthrough oscilloscope. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

2.14 Magnetron phase difference and output power when the PLL circuit startsworking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

2.15 Phase stability when the PLL circuit was working. . . . . . . . . . . . . . . 352.16 Waveform of the PCM output and injection signal measured through oscillo-

scope. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362.17 Magnetron phase difference when the output power was changed. . . . . . 362.18 Response time of the PLL circuit when the phase was changed. . . . . . . . 37

3.1 Diagram of a power-variable phase-controlled magnetron. . . . . . . . . . 413.2 Measured upper and lower frequency boundaries of magnetrons to the injec-

tion power. The magnetron frequency could be locked to the injection signalfrequency within the boundary. red line: S/N007, blue line: S/N012, greenline: S/N010, black line: S/N 005. The green zone indicates the lockingrange within which all the four magnetrons can be locked. . . . . . . . . . 43

3.3 Configurations of the developed slot array antenna. . . . . . . . . . . . . . 453.4 Simulation results of the slot array antenna. . . . . . . . . . . . . . . . . . 463.5 Photographs of the developed slot array antenna. . . . . . . . . . . . . . . 463.6 |𝑆11 | of the slot array antenna. . . . . . . . . . . . . . . . . . . . . . . . . 483.7 Beam patterns of the slot array antenna on the H-plane at 5.8 GHz. . . . . . 483.8 Block diagram of the 5.8GHz phased array. . . . . . . . . . . . . . . . . . 493.9 Photograph of the magnetron phased array system. . . . . . . . . . . . . . 503.10 Simulation results of the combined beam patterns of the 2 × 2 slot antenna

array on the E plane at 5.8 GHz. . . . . . . . . . . . . . . . . . . . . . . . 523.11 Photograph of the LED lamp rectenna elements. . . . . . . . . . . . . . . . 533.12 Beam forming experiments of the magnetron phased array system. . . . . . 543.13 Display of the LED lamp array and simulation beam pattern of states

No.1 No.5 (Black frame: LED lamp array). . . . . . . . . . . . . . . . . . 553.14 5.8 GHz rectenna array system. . . . . . . . . . . . . . . . . . . . . . . . . 563.15 Full view of 5.8 GHz WPT experimental system. . . . . . . . . . . . . . . 573.16 Power density distribution of the case No. 5. . . . . . . . . . . . . . . . . . 583.17 Transmission efficiency and RF-RF-DC efficiency with distances. (Black

line shows the maximum possible efficiency) . . . . . . . . . . . . . . . . 603.18 Transmission efficiency and RF-RF-DC efficiency with distances. (Schematic

diagram of the deviated beam pattern.) . . . . . . . . . . . . . . . . . . . . 61

4.1 Experimental system of the injection-locked magnetron. . . . . . . . . . . 674.2 Characteristics of the lockedfrequency range of the 2.45 GHz and 5.8 GHz

band magnetrons versus the injection power. . . . . . . . . . . . . . . . . . 684.3 Phase noise of the 5.8 GHz Magnetron with the different injection power. . 684.4 Change in the output power of the 5.8 GHz band magnetron with the anode

current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704.5 Characteristic of a 2.45GHz band magnetron output power with magnetic

coil current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714.6 Photograph of a 2.45 GHz band magnetron with coil. . . . . . . . . . . . . 724.9 Result of the data signal and demodulation signal using a 5.8-GHz band

injection-locked magnetron ASK system. . . . . . . . . . . . . . . . . . . 724.7 Change in the output power of the magnetrons with the filament power. . . 734.8 Change in the output power of the magnetrons with the injection power levels. 734.10 Results of the PSK data signal and the demodulating signal. . . . . . . . . . 744.11 Results of the FSK data signal and the demodulating signal. . . . . . . . . . 754.12 Spectral curve of the injection-locked magnetron FSK system with different

data signals (Data signal on the upper trace: a square wave, lower trace: PN9code). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

4.13 Changes in the error vector magnitude of the injection-locked magnetronsystem for PSK with the injection power levels. . . . . . . . . . . . . . . . 79

4.14 Changes in the error vector magnitude of the injection-locked magnetronsystem for PSK with the transmission rates. . . . . . . . . . . . . . . . . . 80

4.15 Schematic of the PM injection-locked magnetron system. . . . . . . . . . . 814.16 Waveform of the PM injection-locked magnetron. . . . . . . . . . . . . . . 824.17 Schematic of the FM injection-locked magnetron system. . . . . . . . . . . 834.18 Photo of the FM injection-locked magnetron system (1: high voltage power

supply, 2: DVD player, 3: FM modulator, 4: amplifier, 5: horn antenna, 6:circulator, 7: magnetron, 8: TV). . . . . . . . . . . . . . . . . . . . . . . . 83

4.19 A wireless power transfer system for electric trolley. . . . . . . . . . . . . . 844.20 Schematic of data transfer system by FSK modulation. . . . . . . . . . . . 854.21 Transmitted data vs. received data. . . . . . . . . . . . . . . . . . . . . . . 864.22 Photos of data transfer system (1: Modulator, 2: High-Voltage Power Supply

and Amplifier, 3: Magnetron, 4: Circulator, 5: Transmitting antenna, 6:Electric trolley, 6.1: Demodulator, 6.2: Driver circuit). . . . . . . . . . . . 87

5.1 Power supply of the magnetron(a.microwave oven power supply, b.half wavevoltage doubler with smoothing capacity, c.full wave voltage doubler) . . . 91

5.2 Simulation results of the power supply voltage(a.microwave oven powersupply, b.half wave voltage doubler with smoothing capacity, c.full wavevoltage doubler, dot line: magnetron oscillation voltage) . . . . . . . . . . 92

5.3 Photo of the microwave power supply. . . . . . . . . . . . . . . . . . . . . 945.4 Photo of the full wave doubler rectifier. . . . . . . . . . . . . . . . . . . . 945.5 Spectral of the magnetron. (dot line:microwave oven, blue line: full wave

voltage doubler without injection signal, black line:full wave voltage doublerwith injection signal) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

5.6 The anode voltage and anode current of the magnetron worked by full wavevoltage doubler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

5.7 FSK system block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . 975.8 FSK modulation/demodulation results. . . . . . . . . . . . . . . . . . . . . 975.9 Injection power vs. lock frequency range. . . . . . . . . . . . . . . . . . . 985.10 Characteristics of the magnetron (upper trace: oscillation frequency vs.

output power, lower trance: anode voltage vs. output power based on the fullwave voltage doubler). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

5.11 Block diagram of a phase-controlled magnetron with full wave double rectifier.1005.12 Magnetron output phase. (upper: worked without PLL circuit, lower: worked

with PLL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1015.13 Voltage waveform of the injection signal and magnetron output without PLL.

(green line: injection signal, yellow line: magnetron) . . . . . . . . . . . . 1025.14 Voltage waveform of the injection signal and magnetron output with PLL at

1% voltage ripple. (green line: injection signal, yellow line: magnetron) . . 1025.15 Photo of the full wave doubler rectifier with 1% voltage ripple. . . . . . . . 1035.16 Circuit of the low cost power supply. . . . . . . . . . . . . . . . . . . . . . 1045.17 Photo of the improved power supply. . . . . . . . . . . . . . . . . . . . . . 1055.18 Block diagram of the experimental system. . . . . . . . . . . . . . . . . . 1075.19 Photograph of the experimental system.Components and devices: (1) mag-

netron; (2) coupler; (3) E–H tuner; (4) circulator; (5) dummy load; (6)power supply; (7) fan; (8) power sensor; (9) high-voltage probe; (10) currentprobe; (11) signal generator; (12) power amplifier; (13) power meter; (14)oscilloscope; and (15) signal analyzer. . . . . . . . . . . . . . . . . . . . . 108

5.20 Experimental spectra of the free-running magnetron, injection signal, andinjection-locking magnetron (both resolution bandwidth (RBW) and videobandwidth (VBW) are 5 kHz). . . . . . . . . . . . . . . . . . . . . . . . . 109

5.21 Block diagram of the experimental system. . . . . . . . . . . . . . . . . . 111

5.22 Photograph (b) of the experimental system. Components and devices: (1)magnetron; (2) waveguide coupler; (3) coaxial coupler; (4) power sensor; (5)magic T; (6) horn antenna; (7) wind-cooled dummy load; (8) waveguide tocoax adapter; (9) 3.75-cm straight waveguide; (10) fan; (11) signal analyzer;(12) power amplifier; (13) VNA; (14) signal generator; (15) dc power supply;(16) power amplifier; (17) oscilloscope; (18) current-probe amplifier; (19)current probe; (20) high-voltage probe; (21) power supply; and (22) ceramicabsorber wall. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

6.1 Photo of the OAM antenna. . . . . . . . . . . . . . . . . . . . . . . . . . . 1166.2 Simulation beam pattern of the OAM antenna. . . . . . . . . . . . . . . . . 1166.3 Measurement of beam phase in near field of the OAM antenna. . . . . . . . 1176.4 Experimental system of the information and power transfer system with

magnetron phased array. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1186.5 Experimental system of the injection-locked magnetron. . . . . . . . . . . 1196.6 Photo of a 2.45GHz rectifier circuit board. . . . . . . . . . . . . . . . . . . 1206.7 Rectifier efficiency of the rectifier circuit with different load (input 50mW). 1206.8 Rectifier efficiency of the rectifier circuit with different input power (load:60Ω).1216.9 Photo of the wirelessly-powered TV. . . . . . . . . . . . . . . . . . . . . . 1216.10 Block diagram of the quick charging system. . . . . . . . . . . . . . . . . 1236.11 Photo of the receiver. (a. a rectifier circuit, b. rear view of rectenna array, c.

front view of rectenna array. d. receiver) . . . . . . . . . . . . . . . . . . . 1246.12 Rectifier circuit efficiency with different load @ input power 6 W. (Orange

line: rectifier efficiency, blue line: reflection efficiency) . . . . . . . . . . . 1246.13 Rectifier circuit efficiency with different input power @60 Ω load. (Orange

line: rectifier efficiency, blue line: reflection efficiency) . . . . . . . . . . . 1256.14 Photo of Magnetron WPT experiment with horn antenna. . . . . . . . . . . 1266.15 Photo of Magnetron WPT experiment with patch antenna. . . . . . . . . . 127

List of tables

1.1 MICROWAVE WIRELESS POWER TRANSMISSION SYSTEMS . . . . 141.2 HIGH POWER MICROWAVE DEVICE [58] . . . . . . . . . . . . . . . . 19

3.1 DESIGN PARAMETERS OF THE SLOT ARRAY ANTENNA . . . . . . 473.2 PHASE DIFFERENCE AMONG PCMS AND BEAM DIRECTION . . . . 543.3 EXPERIMENTAL RESULTS OF WIRELESS POWER TRANSFER . . . 573.4 EXPERIMENTAL RESULTS OF DIFFERENT DISTANCES . . . . . . . 573.5 COMPARISON OF OUR MAGNETRON POWER TRANSMITTER WITH

THOSE IN OTHER RESEARCHES . . . . . . . . . . . . . . . . . . . . . 62

4.1 PARAMETERS AND CONDITIONS OF THE INJECTION-LOCKEDMAGNETRON EXPERIMENTS . . . . . . . . . . . . . . . . . . . . . . . 69

4.2 MODULATION· PERFORMANCE RESULTS OF THE INJECTION-LOCKEDMAGNETRON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

4.3 PARAMETER OF THE FSK MODULATION EXPERIMENTS . . . . . . 86

5.1 PARAMETERS OF THE INJECTION-LOCKING MAGNETRON . . . . 100

6.1 PARAMETERS OF THE WIRELESSLY-POWERED TV. . . . . . . . . . 1226.2 EXPERIMENT RESULTS OF MAGNETRON WPT SYSTEM WITH HORN

ANTENNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1286.3 EXPERIMENT RESULTS OF MAGNETRON WPT SYSTEM WITH PATCH

ANTENNA IN DIFFERENT FREQUENCY . . . . . . . . . . . . . . . . . 1286.4 EXPERIMENT RESULTS OF MAGNETRON WPT SYSTEM WITH PATCH

ANTENNA IN DIFFERENT DISTANCE . . . . . . . . . . . . . . . . . . 1296.5 COMPARE OF MAGNETRON WPT SYSTEM WITH DIFFERENT AN-

TENNAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Chapter 1

Introduction

1.1 Wireless Power Transmission technologies

1.1.1 History of Wireless Power Transmission

Wireless power transmission( WPT) can be traced back to Nikola Tesla’s research as earlyas 100 years ago. Tesla raised one idea of "World Wireless System" which constructeda resonant WPT system that can receive free energy anywhere on the earth[1]. He buildWardenclyffe Tower (also known as the Tesla Tower) for WPT that transmitted 150 kHz300 kW power[2]. M. Hutin and M. Le-Blanc proposed a patent of “Transformer Systemfor Electric Railways” in 1894[3]. They proposed a transformer with an electromagneticinduction method for powering the electrical railways. In Japan, Yagi and Uda inventedthe famous Yagi-Uda antenna[4][5]. They used the electromagnetic wave of about 68 MHz(wavelength:400 cm) to conduct WPT experiments with distinct distances. A vacuum tubewas used as a power receiving rectifier, and output power of about 200 mW was obtained atan output of 2-3 W at a distance of 1.5 m. The transmission efficiency was about 10% andless than 1% at 5 m[6]. 40 years later, W. C. Brown (Raytheon’s Spencer Laboratory) greatlypromoted the development of WPT via microwave almost a half century ago. In May 1963,Brown conducted a 400 W continuous wave generated by a magnetron while the motor atthe receiving 100 W DC power transmission[7]. This should be the first successful WPTexperiment in the world. In 1964, He used microwave to drive a helicopter and continuedto fly for 10 hours at an altitude of 18 meters. The area of the rectifier antenna array is 4square feet. 4480 pieces 1N82G diodes are used to output a maximum DC power of 270W in this device. The experiment used a magnetron operating at 2.45 GHz 5 kW power.The diameter of the elliptical reflector of the transmitter was 3 meters[7],[8]. In 1975, theDC-RF-RF-DC efficiency of a WPT system achieved 54±1% at a distance of 1.7 meters

2 Introduction

(@2.45GHz, DC receive power: 495W)[7]. So far, this is still the world record for thehighest efficiency of microwave WPT. The RF-DC rectification efficiency by diodes wasachieved to 90% (@2.45GHz, RF input power: 8W) [8]. W.C.Brown also developed theamplitron with DC-RF efficiency of more than 70% (@3GHz, RF output power: 400kW,1964) [7]. Amplitron is used in the Apollo program to transmit astronaut television images onthe moon[9]. In 1969, Dr. Glaser proposed the idea of space solar power station (SSPS)[10].The SSPS idea is a huge solar panel system in outer space that transmits energy to the groundvia WPT. Even more in a WPT demonstration 450kW microwave power from a CW klystronwas transferred over than 1 mile, and the DC energy harvesting of 30kW (at 2.388GHz, 1975)[11]. A 4.5 m wing span model (eighth scale) of Stationary High Altitude Relay Platform(SHARP) took its maiden flight in1987 at Communications Research Center [12]. Theseresults of these experiments are still at the high levels of WPT in recent.

After the 1980s, the microwave WPT research began in Japan, and many research experi-ments were carried out with Kyoto University as the representative. In 1983, H.Matsumoto(Kyoto University) led the first microwave WPT rocket experiment in the World, the Mi-crowave Ionosphere Nonlinear Interaction Experiment(MINIX).[13] MINIX is a parent-childrocket experiment that was performed to theoretically predict and demonstrate the nonlineareffect of the radio waves of SSPS microwaves on the ionosphere. A joint study by the KyotoUniversity research group was MILAX (MIcrowave Lifted Airplane eXperiment), using a2.45GHz phased array system on a car to charge the flying airplane. [14]The airplane flewon altitude of 10 m and length of 350 m by the microwave power supply (@ 2.411 GHz, RFoutput power: 1.2 kW, DC receiving power 88 W, 1992). The next year, the transmitter ofthe MILAX was used to in the second microwave WPT rocket experiment of Experimentof Microwave Energy Transmission in Space(ISY-METS)[15]. ISY-METS demonstratedthat activity phased array in space and experiments of nonlinear plasma interaction. In 1995,N.Kaya 𝑒𝑡.𝑎𝑙 carried out the ETHER (Energy Transmission to a High Altitude long en-durance airship ExpeRiment) project test that drives an unmanned airship through microwaveWPT[16]. In the experiment, 10 kW power(2.45 GHz) was transmitted from a parabolicantenna placed on the ground, and a DC output was taken out from a rectenna attached to thebottom of the airship, to drive a propulsion unit of the airship. The airship had a total lengthof 16 m and a maximum diameter of 6.6 m. A 2.7 m × 3.4 mm rectenna with a total of 1200elements was mounted on the bottom of this, producing maximum output of 5.9 kW. In 1996,Kyoto University, Kobe University, and Kansai Electric Power Co., Inc. jointly conducted a2.45 GHz, 5 kW Point-to-Point microwave transmission using a magnetron for 42 m from aparabolic antenna with a diameter of 3 m. The maximum of 742W DC power was receivedwith a 2304-element rectenna array of 3.2 m x 3.6 m arranged at half wavelength intervals. A

1.1. WIRELESS POWER TRANSMISSION TECHNOLOGIES 3

high power rectenna was developed for incident power reaches 2.5W per element [17]. In2009, two phase-controlled magnetrons mounted on an airship were array antennas with a2-element honeycomb radial line slot antenna. The power was supplied by two 28V30AHlithium-ion batteries. It is a retro directive system with a 5.8GHz pilot signal from the ground.An airship was moored at an altitude of 33 m, we were able to confirm power transmissionwhen an electronic buzzer connected to a 4-element rectenna on the ground rang[25]. InJapan, the 5.8 GHz band WPT was demonstrated to be a phased array transmitting 1.8 kWmicrowave. The control accuracy of the microwave beam was achieved at 0.15 degrees(distance 55 m, DC receive 340 W, 2015) [26].

During the same period, products such as electric toothbrushes, shavers, and cordless tele-phone that could be charged in wireless appeared on the market. Since the IC cards(Felica)developed by SONY with 13.56MHz inductive WPT[18]. IC card WPT product becamecommonly in life, so far WPT really came into people’s lives. In 1995, A program was pro-posed for Individual and Public Urban Transport (Transport Urbain Libre Individuel et Public– TULIP) of the electric car rental subscription service with wireless charging in France[24].In 2007, a research group of MIT carried out on coupled magnetic resonance wireless powertransfer (WPT)[19][20], which has been well known by more people. Since 2002, the ap-plication of wireless charging technology in electric vehicles(EV) has gradually formed atrend. Daimler Mercedes Benz, Toyota, Nissan leaf, Mitsubishi iMiEV and other manufac-turers have released or demonstrated models with wireless charging [27][28][29]. EuropeIPT Technology and China ZTE launched wireless charging products for EV bus[30][31].UK government has tested the "dynamic wireless power transfer" technologies on countrymotorways and major A roads where EVs could travel long distances without needing to stopto charge the car’s battery[32].

Recently, two WPT standards have been proposed[21]. One is that the Wireless PowerConsortium WPC has put forward a𝑄𝑖 standard which operates at 100–205 kHz[22]. Anotherstandard is the establishment of the AirFuel Alliance, which combines the Power MattersAlliance and Alliance for Wireless Power working at 277–357 kHz[23]. Untill now, wirelesscharging products are also getting closer to life. The 𝑄𝑖 standard smart phones can bewirelessly charged on the table at STARSTRUCK shops. In May 2020, Japan Land WirelessCommunication Committee reviewed the proposal for microwave WPT in three bands of920MHz, 2.45GHz, and 5.8GHz. Japan is expected to become the first country to pass themicrowave band WPT legislation.

After more than 100 years of development, wireless charging technology has beenimproved and applied to daily products. The distance of wireless charging covers several cm

4 Introduction

His

tory

of th

e W

PT

En

d o

f 1

9th

Ce

ntu

ry1

96

41

92

6

19

68

19

92

19

83

© R

ISH

, K

yo

to U

niv

.

Nic

ola

Te

sla

”Wo

rld

Sy

ste

m”

Re

son

an

ce W

PT

ev

ery

wh

ere

on

th

e E

art

h

US P

atent

527,8

57

Fir

st P

ate

nt

of

Wir

ele

ss C

ha

rge

r

(3k

Hz-

AC

) fo

r E

V in

18

94

On

th

e f

ea

sib

ilit

y o

f p

ow

er

tra

nsm

issi

on

by

ele

ctri

c w

av

es

Yag

i a

nd

Ud

a

MP

T t

o F

lyin

g D

ron

e

W.

C.

Bro

wn

Sp

ace

so

lar

po

we

r st

ati

on

P. E

. G

lase

r

19

75

W.

C.

Bro

wn

JPL,

US

Kyo

to U

niv

.

MIL

AX

Kyo

to U

niv

. &

…

Fig.

1.1

His

tory

ofw

irel

ess

pow

ertr

ansm

issi

on(2

-1).


His

tory

of th

e W

PT

19

93

© R

ISH

, K

yo

to U

niv

.

19

95

Fe

lica

ISY

-ME

TS

13

.56

MH

z In

du

ctiv

e W

PT

Tra

nsp

ort

Urb

ain

Lib

re I

nd

ivid

ue

l e

t P

ub

lic

19

97

MIT

Ind

uct

ive

Re

son

an

t C

ou

pli

ng

20

07

Wir

ele

ss P

ow

er

Co

nso

rtiu

m

(WP

C)

20

09

Air

Fu

el

20

15

Wir

ele

ss C

ha

rgin

g f

or

EV

-Bu

s in

Ja

pa

n

20

11

Ele

ctri

c m

oto

rwa

ys

in U

K

JAX

A S

SP

S

20

20

Ind

uct

ive

Wir

ele

ss C

ha

rge

r fo

r P

ub

lic

Ele

ctri

c B

us

in E

uro

pe

Fig.

1.2

His

tory

ofw

irel

ess

pow

ertr

ansm

issi

on(2

-2).

6 Introduction

to several kilometers, and the charging power ranges from several mW to several hundredkW. For some important events in the history of wireless charging as shown in Fig. 1.1 andFig. 1.2.

1.1.2 Applications of Wireless Power Transmission

WPT technology, especially the application of wireless charging, can greatly facilitate dailylife. The wireless charging technology does not need to require connection joints with a lowwear rate. As shown in Fig. 1.3, the current application of WPT mainly includes wirelesscharging electric toothbrush, cordless phone, IC card, mobile phone charging, EV charging,drone charging, and island power transmission and SSPS.

Low frequency electromagnetic induction technology is used in electric toothbrushes,shavers, and cordless telephone, and their principles are similar to that of transformers.Specifically, the charging base for wireless charging is the primary coil of the transformer,and the receiving coils on the toothbrush and the telephone are the secondary coils of thetransformer. When the alternating current is connected to the primary coil, the secondary coilproduces an induced voltage. After rectification, the battery can be charged. The efficiencycan be as high as about 80% to 95%, without high frequency radiation, has passed the safetycertification of various countries, and the price is relatively low.

IC card, The RF reader sends a set of fixed frequency (13.56 MHz) electromagneticwaves to the IC card. There is a LC series circuit in the card, whose resonance frequency isthe same as the frequency emitted by the card reader. Therefore, so that under the excitationof electromagnetic waves, the LC resonance circuit resonates, which makes the capacitorgenerate electric charge; at the other end of the capacitor, a unidirectional electronic pumpis connected to transfer the electric charge in the capacitor to another capacitor for storage.When the accumulated charge reaches a certain voltage, the capacitor can be used as a powersupply to provide working voltage for other circuits, and transmit the data in the card orreceive the data from the card reader.

Mobile phone wireless charging technology is becoming mature. There is strong electro-magnetic field energy in wireless charging area. Adding some coils here produces a voltage.The induced voltage can be converted into the energy received by the mobile phone battery.However, mostly mobile phone wireless charging only carry out one-to-one short-distancecharging (𝑄𝑖 standard, within 1 cm, within 5 W[22]). Some research have developed onMultiple Input Multiple Output (MIMO) WPT[33][34]. In October 2017, the iPhone 8 wasreleased to support wireless charging Qi standard functions. Wireless charging technologyhas attracted more people’s attention again. We can look forward to wireless charging to beas convenient as connecting mobile phones to Wi-Fi in the future.


Internationally, automobile manufacturers such as Audi, BMW, Mercedes-Benz, Volvo,Toyota, Nissan leaf, Mitsubishi and other makers including IPT Technology, ZTE, DAI-HEN, etc. have all started to study the wireless charging technology of electric vehicles[27][28][29][30][31][35]. There are two main forms of wireless charging for cars, one isfixed-point charging when the car is parked and the other is charging while the car is drivingon the charging roads in UK[32]. Charging while driving technology can greatly improvethe driving distance of electric vehicles, making it not inferior to or even more than gasolinevehicles. On the other hand, the construction of charging roads requires a large amount offunds for road reconstruction. Fixed-point charging technology is developed by most manu-facturers. However, due to the lack of a unified industry standard, it is currently impossibleto charge products across manufacturers, which is the bottleneck of current development.

The longer WPT application is island power transmission. In France, there is a planto transmit 10kw of electricity from the summit to a mountain village on Reunion Island.When WPT 2001 was held on this island, it launched 800 W 2.45 GHz microwave as ademonstration, and obtained 65 W output 40 m away[36]. There is a research similar to theisland power transmission application, where wind power plants at sea use microwaves todeliver electricity to shore[37].

Since 1968, Peter E. Glaser proposed the SPSS concept[10], SSPS became the longestapplication of WPT. The United States, Japan, EU, and China have proposed relatedSSPS programs. The more representative programs are the DOE/NASA satellite powersystem[38], SPS2000[39], Tethered-SPS[40], JAXA model[41], SSPS USEF model[42],SPS-ALPHA[43], Multi-Rotary Joints SPS[44]. These programs proposed that solar panelsin the space environment use sunlight to generate electrical energy, and then supply it to theground through microwaves. In space, the amount of sunlight is not affected by the weather,and consequently it is stable throughout the day, the night, and the year. It is said that theamount of power generation as a comprehensive value will be about ten times that on theground. SSPS usually chooses geostationary orbit (at an altitude of 36,000 km). It faces theground receiving equipment in the power supply area for 24 hours, and at this altitude, thepower generation satellite is located outside the earth’s shadow. From space to ground, thereis a so-called "radio wave window" (frequency band from 1 GHz to 10 GHz, in which theradio wave attenuation is relatively small), and in the microwave system, the frequency of2.45 GHz to 5.8 GHz can be selected in the ISM band .The SSPS also faces another problem,how to send tens of thousands of tons of SSPS to the space orbit at low cost.

8 Introduction

WP

T s

yste

m a

pplic

ations

mm

cmm

10

m3

60

00

km

10

km

Sp

ace

so

lar

po

we

r st

ati

on

Qi

UK

Oss

ia dro

ne

Ind

uct

ive

Co

up

lin

gR

eso

na

nce

Co

up

lin

gR

ad

io w

ave

Fig.

1.3

App

licat

ions

ofW

irel

ess

Pow

erTr

ansm

issi

on


1.1.3 Theory of Wireless Power Transmission

According to the principle, WPT technology can be divided into three categories, electromag-netic waves radiation, field coupling, and vibration waves radiation as shown in Fig. 1.4. Theelectromagnetic wave radiation mainly includes laser[48], microwave and other frequenciesband, the coupling has inductive coupling and capacitive coupling based on the principle ofelectromagnetic induction, and the resonance coupling represented by MIT research[19], andthe vibration wave mainly transmits energy through ultrasonic waves[46].

WPT

Laser

UltrasoundVibration wave

Electromagnetic wave

MicrowaveMicrowave

Coupling Inductive Coupling

Resonance Coupling (Magnetic)

Capacitive Coupling

Other Frequencies

Fig. 1.4 Classification of wireless power transfer.

The electromagnetic waves radiation WPT system consists of three parts. The first partis the transmitter that is the electromagnetic wave generator, which realizes the conversionof DC–electromagnetic waves; the second part is the transmitting antenna or radiator, theelectromagnetic wave propagates through free space, and reaches the receiver of the thirdpart, the rectifying antenna or solar panel, which transfers the electromagnetic wave energyand convert to DC power. This method requires the system with a high DC–RF conversionefficiency and transmission efficiency. The transmission of electromagnetic energy in freespace follows the Friis formula as shown as Eq. 1.1 [48].

10 Introduction

𝑃𝑟 =

(_

4𝜋𝐷

)2𝑃𝑡𝐺 𝑡𝐺𝑟 (1.1)

Here, 𝑃𝑡 , 𝑃𝑟 , 𝐺 𝑡 and 𝐺𝑟 are the transmitter power, receiver power, antenna gain of thetransmitter and receiver, respectively. 𝐷 is the transmission distance. _ is the wavelength. Eq.1.1 shows that a high transmission efficiency requires a high-gain receiving antenna and atransmitting antenna. The electromagnetic waves radiation WPT is suitable for long-distancepower transmission and is not limited by the transmission medium. The representative studyis that W. C. Brown developed WPT system achieved 54±1% DC-RF-RF-DC efficiency at1.7 meters distance (@2.45 GHz, DC receive power: 495 W, Receiver antenna diameter:nearly 1.5 m)[7].

The inductive coupling and capacitive coupling WPT system based on the principle oftransformer utilizes the method of electromagnetic induction. It separates the transformer’stightly coupled magnetic circuit of the system, and the primary and secondary windingsare wound on different structures to realize the energy transfer between the power supplyand the load unit without physical connection. The primary side and the secondary siderealize power transmission through electromagnetic induction. The primary side of thetransformer is input by the AC power rectified and filtered into DC power, and then invertedby a high-frequency inverter to provide high-frequency AC power. An AC voltage is inducedon the secondary side. The AC power is rectified into DC power. The DC–RF conversionefficiency can worked in high efficiency. The higher the transmission efficiency requires thehigh frequency and amplitude, the short primary and secondary distances. The theory ofinductive coupling and capacitive coupling WPT system limited the transmission distance.The inductive coupling technology is the most mature and widely used, and there are manycommercial products. For example, Daihen’s products for EV charging system achieved92% AC-DC-85 kHz AC-85 kHz AC-DC efficiency at transmission distance of 0.25 m (DCreceive power: 10 kW, Coil diameter: 0.75 m)[35].

The resonance coupling represented a 2 m transmission distance with 40% 9.9 MHz RF-9.9 MHz RF efficiency (RF receive power: 60 W, Coil diameter: 0.5 m) by MIT research[19].The principle is the same as the resonance principle of sound. The primary coil and thesecondary coil arranged in a magnetic field can supply power from one to another. Theresonance coupling is a special mode of electromagnetic induction coupling. The differencebetween them is the quality factor of the circuit. Resonant devices (inductors and capacitors)are used to make the transmitter and receiver reach a resonance frequency, thereby generating


magnetic resonance and energy transfer. The advantage of resonance coupling is longertransmission distance than inductive coupling and capacitive coupling.

At present, there is still relatively little research on the use of ultrasonic WPT. Thetransmission of sound waves (vibration waves) radiation is similar to electromagnetic wavesradiation. The system replaces the antenna of the electromagnetic wave radiation systemwith an acoustic wave transducer. Ignoring the absorption loss in the transmission medium,the transmission efficiency also follows the Friis formula. In addition to the transmissionloss, the power loss in the medium cannot be ignored. Compared with electromagneticwaves radiation, the advantage of using sound waves WPT technology that its transmissionspeed is much lower than the transmission speed of electromagnetic waves. Therefore, inthe case of the same wavelength, the frequency used by the sound wave WPT will be lower,and according a smaller energy loss. Sound waves WPT is very suitable for mediums withrelatively large attenuation of electromagnetic waves, such as in water, or environmentswithout electromagnetic interference, but not in vacuum. Among the sound waves, thecharacteristics of ultrasonic waves are particularly obvious. When the ultrasonic wavespropagate in the medium, the ultrasonic waves have a relatively high frequency. So they willdrive the particles in the medium to have a higher vibration frequency, which will producegreater power than ordinary sound waves. In addition, ultrasound also has good directivity,which can focus the sound waves well at the receiving point.

0%

20%

40%

60%

80%

100%

0.1 1 10 100 1000 10000

Efficiency

Distance(m)

[34]

[18]

[6]

[7]

[11][14]

[16] [25]

Coupling WPT

Microwave WPT

Fig. 1.5 WPT Research achieved efficiency with distance.

12 Introduction

The above three technologies differ greatly in performance indicators. They differ inoutput power, transmission distance, efficiency, frequency and other aspects. Due to the fewresearch cases of laser and ultrasound, Fig. 1.5 only summarizes the research of couplingWPT and microwave WPT that the achieved efficiency levels with distance. The researchof microwave energy transmission type is the earliest, the transmission power as well asthe efficiency are low, but the transmission distance is relatively long. If the antennas forman array, the transmission distance and transmission power can reach a considerable scale.It has a unique advantage in transmission distance. Microwave beam forming technologycan be used to transfer power to the moving target. Therefore, its mainly research areas arein SPS, drones charging, and islands power transmission. Coupling charging technologyis widely used in large-scale transportation, mechanical equipment and portable electronicequipment power transmission and charging. The transmission efficiency drops sharplywith the increase of distance. At present, it is mainly the power transmission at close range.Resonance coupling technology basically meets the requirements of portable electronicdevices and household appliances, and can be used as a good solution in commercialapplications. Ultrasonic energy transmission research is in its infancy, and it is expected thatthis technology will be used for power transmission in special occasions such as underwater.The purpose of this study is to build a high-power long-distance power transmission system,the microwave WPT technology is most suitable.

1.2 High Power Microwave Wireless Power TransmissionSystems

Since the research of W.C.Brown, many verification experiments have been carried out in thefield of microwave WPT. Table 1.1 lists a relatively complete system with relatively high-power WPT experimental parameters, and compares the relevant experimental parameters.

In the Table 1.1, the highest microwave transmission efficiency is recorded by W.C.Brownsystem in 1975[7]. The DC–RF conversion efficiency of the magnetron used at the transmitterwas 69%, and the output was connected to a horn antenna with a length of more than 2meters. There was a receiving rectenna array 1.6 meters directly in front of the horn antenna,and the collection efficiency reached 95%. Rectenna arrays are juxtaposed according to thesame energy density distribution to improve the synthesis efficiency. The overall efficiency ofRF–DC reaches 82%. The efficiency of each item had reached the limit that can be reached atthat time. This is a milestone study in the history of WPT. Nearly 25 years later, another highefficiency microwave WPT was developed by Dr. Kwan-Ho Kim of Korea Electro technology

1.2. HIGH POWER MICROWAVE WIRELESS POWER TRANSMISSION SYSTEMS13

Research Institute (KERI) group from 1997 to 2000. The measured total DC–RF–RF–DCefficiency was 44% and the calculated total efficiency 50.6% with 50 m transmission distance.The calculated DC–RF conversion efficiency, RF–RF collection efficiency, and RF–DCrectification efficiency were 84.3%, 77.4% and 79.2% [47], respectively.

In 1975, the microwave WPT demonstration conducted at the JPL gold stone factorycreated another record[11]. A 450 kW transmitter at 2,388 MHz from a parabolic antennawith a diameter of 26 m was used for power transmission. The transmission distance was 1.54km, and a 7.3 m × 3.5 m rectenna array was used for power reception, and it was receivedmore than 30 kW DC power. So far, it is the WPT experiment with the highest transmitpower and received power and the longest transmission distance.

SHARP [12], MILAX[14], Semi-Autonomous BEam Rider (SABER)[49], and ETHER[16] systems are powered by the ground to airborne targets (airplanes, airship). SABER flewa model helicopter with a microwave-driven rotor with a diameter of 1.15 m. A 2.45 GHzmagnetron boosted to 1 kW was connected to a slot waveguide antenna and connected to arectenna array hanging under the helicopter. This rectenna array was used in an experiment.It supplied 180 W to a helicopter motor with a transmission distance of 3 m. MILAX isthe earliest use of semiconductor FET(GaAs) as the microwave amplifier in the list. Theconstructed phased array combined with CCD camera to track and transmit power to theairplane in flight. Although auxiliary cars are needed, this should be the first time to use aphased array with beam forming WPT experiment.

After the MILAX experiment, Point-to-Point MPT[17], Reunion Island[36] all usedmagnetrons for the experiment. Among them, Point-to-Point MPT also carried out up to 5months of endurance experiments and rainy power transmission experiments, which providedvaluable data for the practicality of WPT technology.

20 years later of the MILAX experiment, with the development of semiconductor technol-ogy, the semiconductor HEMT (GaN) is used in H. Ma 𝑒𝑡.𝑎𝑙 experiment [50] and the groundtest practical application demonstration of JAXA 𝑒𝑡.𝑎𝑙 [26]. The H. Ma 𝑒𝑡.𝑎𝑙 experimentconstructed the parabolic antenna with a diameter of 2.4 m as the transmitter antenna and2.4 m × 2.4 m patch antenna array as the receiving antenna. The beam collection efficiencyis 49% at a distance of 11 m, and the overall DC-RF-RF-DC efficiency is 16.9%. Here, theDC-RF conversion efficiency of the GaN transmitter is 59.3% when the output is 87.68 W.Although related research shows that the GaN amplifier delivered a maximum power-addedefficiency of 79% at 5.65 GHz @33.3 dBm [51]. But in tens of W power level microwaveamplifiers, it worked in high efficiency level. The phased arrays in JAXA’s demonstrationdemonstrated the high-precision beam control technology, which can direct high-power

14 Introduction

Tabl

e1.

1M

ICR

OW

AVE

WIR

EL

ESS

POW

ER

TR

AN

SMIS

SIO

NSY

STE

MS

Res

earc

hTr

ansm

itter

devi

ceTr

ansm

itter

RF

Pow

er(W

)R

ecei

ving

DC

Pow

er(W

)D

ista

nce

(m)

Effi

cien

cyR

F-R

F-D

CFr

eque

ncy

(GH

z)E

xper

imen

tTi

me

W.C

.Bro

wn

[7]

Mag

netr

on(6

9%)

635

495

1.7

78%

2.45

1975

R.M

.Dic

kins

on[1

1]K

lyst

ron

4500

0030

000

1540

6.67

%2.

388

1975

SHA

RP[

12]

Mag

netr

on10

0015

015

01.

50%

2.45

1987

MIL

AX

[14]

GaA

sFE

T12

5088

107.

04%

2.41

119

92Po

int-

to-P

oint

MPT

[17]

Mag

netr

on50

0074

242

14.8

4%2.

4519

95SA

BE

R[4

9]M

agne

tron

1000

180

318

%2.

4519

95E

TH

ER

[16]

Mag

netr

ons

1000

030

0035

30%

2.45

1995

K.H

Kim

Mag

netr

on(8

4.3%

)19

0010

2050

53.6

8%2.

4520

00R

euni

onIs

land

[36]

Mag

netr

on80

065

408.

12%

2.45

2001

JAX

Aet

.al[

26]

GaN

1800

340

5530

.60%

5.8

2015

H.M

aet

.al[

50]

GaN

(59.

3%)

>50

-11

28.6

4%2.

4520

15W

irel

ess

TV

(CH

6.2)

Mag

netr

on32

950

3.5

15.2

0%2.

4520

17M

agne

tron

Phas

edar

ray(

CH

3.4)

Mag

netr

ons(

60.7

%)

1204

142

511

.79%

5.8

2019

Mag

netr

on(C

H6.

4)M

agne

tron

(60.

46%

)72

410

95.

615

.12%

5.8

2020

1.3. HIGH POWER MICROWAVE TRANSMITTER 15

microwave beams to the receiving antenna with high precision and high efficiency. Provide avery feasible solution for the realization of the future SSPS system.

In the Table 1.1, there are the systems of this thesis (Chapter 3.4 and Chapter 6.4) areused 5.8 GHz band magnetrons. The Wireless TV experiment (Chapter 6.2) is a uniquesystem which transfer the power and information. Details of these experiments are describedlater.

1.3 High Power Microwave Transmitter

V. L. Granatstein 𝑒𝑡.𝑎𝑙 [52] and S. Oliver[53] have summarized the power-frequency perfor-mance of semiconductor devices and vacuum devices. The combined graph is shown in Fig.1.5. Fig. 1.5 shows that semiconductor devices are used in low-frequency and low-powerfields, and vacuum tube devices are used in high-frequency and high-power fields. Withthe development of semiconductor technology, semiconductor devices continue to achievehigher power and higher frequency. However, one single device is basically a vacuum tubeproduct in the field above 1 kW and the field above 100 GHz. In this section, it is introducedthat several devices suitable for high-power systems in microwave WPT. It mainly composedof Gallium Nitride High Electron Mobility Transistor (GaN HEMT), Klystron, TravelingWave Tube(TWT), Gyrotron, Magnetron and crossed-field amplifier (CFA).

1.3.1 GaN HEMT

Gallium nitride (GaN) is a semiconductor material with a wide direct band gap and canbe used in high-power, high-speed optoelectronic components. For example, GaN can beused in violet laser diodes. The earliest used in light-emitting diodes (LED). Y. AkasakiYu, H. Amano and S. Nakamura won the 2014 Nobel Prize in Physics for inventing blueLEDs[54]. GaN has the characteristics of high power, high voltage, high frequency andlow loss, which have exceeded the limits of Si power devices widely used today. The effectof making microwave power devices by GaN (in power density) is often far superior to allexisting semiconductor materials. It is highly expected to be used as next-generation powerdevices.

H. Ma et.al experiment [50] and JAXA’s demonstration [26] using GaN HEMT to buildthe microwave amplifier in the transmitter. GaN HEMT has the advantages of semiconductordevices and very low noise. JAXA’s demonstration demonstrated the high-precision beamcontrol performance. For microwave WPT transmitters, GaN HEMT works as a switchingdevice for amplifiers. In the class F amplifier circuit with a theoretical efficiency of 100%,

16 Introduction

Solid-State Devices

Vacuum

Devices

Magnetron

Max:1GWMax:150MW

Fig. 1.6 RF vacuum devices (black line) [52] in 1999 and solid-state devices (color area) [53]in 2014 state of technology for single device performance.

the highest 79% PAE of the amplifier is currently achieved [51]. However, to achieve high-power microwave output requires multi-stage amplification, which inevitably leads to lowerefficiency. A single amplifier circuit to construct a single array unit to complete the phasedarray should be the most efficient way. However, when the power of a single array element issmall, the size of the array becomes very huge. It requires GaN HMET to conduct furtherresearch on high efficiency and high power. GaN HEMT is processed using a semiconductorNano/Micro processing technology. From the cost calculated per unit power, both GaNHEMT individuals and GaN amplifiers are very expensive. This is a low noise, low efficiencyand expensive solution for WPT.

1.3.2 Klystrons

Klystron is a type of microwave tube that uses periodic modulation of electron injectionvelocity to achieve oscillation or amplification. Klystron also named velocity modulatedtube since its theory. It first modulates the velocity of the electron beam in the input cavity,converts it to density modulation after drifting. Then the clustered electron blocks exchangeenergy with the microwave field of the output cavity gap, and the electrons give the kineticenergy to the microwave field to complete oscillation or amplification. There are mainly


two types of rectilinear type and reflective type klystron. The rectilinear type klystron has astructure in which a plurality of cavity resonators are connected in series. Microwaves aregenerated by repeating velocity modulation and density modulation between the plurality ofresonators. The power of a single tube can reach tens of kW in continuous wave, and thepulse wave can reach reach 150 MW level[55]. The DC-RF conversion efficiency from isaround 60–70%. The surplus kinetic energy of the electron beam is converted into heat bythe collector. Some papers show the expected efficiency range can reach 90% [56][57]. It ismainly used as an amplifier in the field of TV broadcasting and accelerator in the early UHFband. The reflective klystron is used as an oscillator and has low power (usually less than1W). A cavity resonator is used alone, and one resonator is used as an input and an output ofa high frequency electric field at the same time to cause electrons to retrograde at a reflectingelectrode to generate a microwave by resonance.

1.3.3 TWT

TWT utilizes the interaction that occurs between the electric field on the axis and the electronwhen the velocity of the electron flow emitted from the electron gun in vacuum is almostequal to the velocity of the radio wave on the delay circuit traveling in the same direction.Then, the microwave is amplified. The electron flow is finally collected by the collector.There are also compound devices that combine klystron and traveling wave tube technology,named twystrons[55]. Compared with klystron, TWT has unmatched bandwidth which iswidely used in satellite communications. The MPM (Microwave Power Module) is a singlemodule that integrates the high-frequency amplifier and power supply system. It is a new typeof microwave amplification device that combines a super mini TWT, which plays the role ofmicrowave amplification. TWT has a wide bandwidth, high gain, high power, high frequencycharacteristics. Thus, MPM is expected as a potential device in the field of communications.

1.3.4 Gyrotrons

Gyrotron is a large electron tube characterized by oscillating high-power microwaves with ashort wavelength using the motion of electrons that rotate at high speed along a magneticfield as an energy source. While entering the high frequency oscillator (cavity resonator), therotational power of the electron beam becomes high frequency. The rotation frequency 𝜔𝑐is shown in the Eq. 1.2. 𝑒, 𝑚 and 𝐵 are electron charge, mass, and magnetic flux density,respectively.

𝜔𝑐 =

( 𝑒𝑚

)𝐵 (1.2)

18 Introduction

Gyrotron as a large output oscillator is possible. The oscillation frequency band is 20 to250 GHz, and the output is usually 1 to 2 MW. A research had obtained an output of pulsewidth of 70ns and peak power of about 1 GW under an electron beam of 3.3 MV and 80kA[57]. Gyrotron mainly used for plasma heating in nuclear fusion, but also used for ionengines etc.

1.3.5 Magnetrons & CFA

The magnetron is the oldest microwave tube in use today and can be widely used as anunparalleled high-power oscillator. It is especially applied to microwave ovens and radar.Even before the widespread use of microwave ovens, high-frequency dielectric heating hasbeen used in industrial fields such as drying and bonding wood. Similar to other vacuumtubes, semiconductor devices have been tried, but their high output, high efficiency, and lowcost are still beyond the reach of semiconductor devices.

A magnetron is a special bipolar vacuum tube that applies a magnetic field from a magnet.Lorentz force acts on the electrons moving in the magnetic field, and the orbits of theelectrons are bent. As shown in Fig. 1.7, when the electrode structure of the bipolar vacuumtube is devised and a magnetic field is applied from the outside, the electrons emitted fromthe central cathode do not reach the outer anode, and circulate while rotating around thecathode. If the velocity 𝑣 of the electrons and the phase velocity of the electric field in theresonant cavity on the anode side are both equal to 𝐸/𝐵 as shown in Eq. 1.3, the electronswill be swarmed by the interaction of the electric fields. The electrons approach the cavityand the electric field gains energy and increases. Unlike other microwave vacuum tubes,the magnetron’s collector and anode are integrated, which is the cause of the magnetron’shigh efficiency. This vibration is resonated in the cavity provided on the anode side, and theenergy is taken out as a radio wave from the antenna.

𝑣 =𝐸

𝐵(1.3)

The disadvantage of magnetron is that the unstable oscillation frequency and the shortlife which is mainly decide by filament. W.C. Brown improved magnetron principles toinvent a new vacuum tube amplifier which he called an amplitron [60]. The electric andmagnetic fields in amplitron are perpendicular to each other (cross field) and hence the nameis cross field amplifier. Its working principle is basically the same as the magnetron, andmany characteristics are also very similar. The difference is that the CFA has an input port,which is an amplifier, and the magnetron is an oscillator. Due to the structural change of CFA,the unstable frequency is avoid. If the short-life filament is transformed into cold cathode,


vacuum

Fig. 1.7 Principle diagram of Magnetron.

Table 1.2 HIGH POWER MICROWAVE DEVICE [58]

Devices Power level Frequency BandwidthEfficiency(Typical)

Lifetime Cost:W

GaN ∼100W Hz∼10GHz Normal 50% 800kh HighKlystrons 100W∼100MW MHz∼100GHz Normal 60% 30kh NormalTWT W∼100kW GHz∼100GHz Widely 50% 200kh HighGyrotron kW∼GW 10GHz∼THz Narrow 40% 50kh NormalCFA 10W∼10MW GHz∼10GHz Narrow 80% 40kh NormalMagnetron 10W∼MW 100MHz∼10GHz Narrow 70% 2kh Low

the life of CFA can reach more than 40,000 hours [61]. This should be the most suitabledevice for WPT, but it is difficult to buy related products.

Table 1.2 summarizes the parameters of these 6 kinds of microwave devices, which cancompare their advantages and disadvantages. The purpose of this research is to constructa high-power, long-distance, high-efficiency WPT system. There is no requirement forbandwidth in WPT research. Considering the convenience of purchase, magnetron is the bestchoice. The service life of 2000 hours is very lower than other devices, but it is far enoughfor research.

20 Introduction

1.4 Outline of the Thesis

This thesis consists of 7 chapters. The first chapter describes the the history, applications,principles of WPT. Then, the high power wireless power transmission system and high powermicrowave device are introduced . Chapter 2 is a study of phase-controlled magnetron to solvethe magnetron noise problem. The original phase control method using the anode current ofthe 2.45 GHz magnetron is that the 5.8 GHz magnetron cannot be used. We analyzed that thephase difference between the injected signal and the magnetron does not depend on the I- 𝑓characteristic, designed a phase-locked loop circuit, and developed a magnetron phase controlmethod with high versatility. In Chapter 3, four units phase-controlled magnetrons whichdeveloped in Chapter 2 were used to build the magnetron phased array. The magnetronswere arranged in a 2×2 form. By constructing a microwave WPT system, the output phaseand power of the magnetron phased array can be adjusted, and the direction control (beamforming) function was verified. Chapter 4 discovers the modulation performance of themagnetron which is expected to realize high-speed communication. In the conventionaltechnology, the magnetron was researched for communication with low transmission speed,but as a result of this research, the transmission speed was improved and a communicationspeed of 10 Mb/s could be realized. In addition, demonstration experiments such as voice,video, and motor operation control signal were conducted through magnetron communication.Chapter 5 describes the low cost power supply for magnetron. We introduce a half-wavevoltage doubler rectifier circuit by improving the power supply of the microwave oven and aswitching power supply. With low cost power supply, it becomes possible to realize phaseshift control magnetron, high-speed communication, power combining of magnetron, etc.Chapter 6 introduces the completed WPT systems by this research. It describes the magnetronphased array that can communicate, wireless TV, magnetron WPT system. Then, how tobuild a high efficiency WPT system was discussed. Finally, in Chapter 7 described theconclusion of this thesis and future issues.

Chapter 2

Power-Variable Phase-ControlledMagnetron

2.1 Introduction

Magnetrons have been widely used as microwave heating applications typified as microwaveovens. The advantages of magnetrons include high efficiency, low cost, and high outputpower. However, the magnetrons as transmitters have a wide bandwidth, difficult phasecontrol, poorly stable output frequency, and high phase noise. If they are used for wirelesspower transmission (WPT) and other applications, the magnetrons require a method to obtainrigorously stable output power, frequency and phase.

An ordinary approach of a phase-controlled magnetron (PCM) with stable output fre-quency and phase is utilizing an injection locking method to make the oscillation frequencyof magnetron stable, and a phase locked loop (PLL) method to control the magnetron phase[62]-[64]. The 2.45 GHz PCM has been developed and been used for WPT and a com-munication transmitter. Our research group has developed the PCM as the transmissionapparatus of a WPT system for several years. In 2001, we had developed a kW-class highpower phased array system called SPORTS (Space POwer Radio Transmission System)[65].Further, we developed a 2.45 GHz phase and amplitude controlled magnetron (PACM) forWPT system[66]-[68]. Also, another group developed a multi-magnetron microwave sourceto build a near-field 2.45 GHz microwave power transmission system[69]. Also, the PCMcan be used for a communication transmitter, whose transmission data rate at 2 Mb/s hasbeen achieved[70].

In previous studies, the PCM was achieved by controlling the anode current to lockthe phase of magnetron[62][63][66][67]. This method requires that the anode current and

22 CHAPTER 2. POWER-VARIABLE PHASE-CONTROLLED MAGNETRON

the oscillation frequency of the magnetron have a linear relationship. However, a 5.8 GHzmagnetron has different frequency-current characteristics from a 2.45 GHz magnetron.Furthermore, the PACM requires adding a coil to the magnetron to control the magneticfield[66]. This affects the convenience of PCAM and limit the use of PACM to otherapplications. The previous 2.4 GHz PCM studies were based on turn-off of the filamentcurrent[63], [66]. It makes the magnetron spectrum quiet and pure[66]. However, turn-off ofthe filament current of 5.8 GHz magnetron will reduce the output power and the signal-noiseratio. Moreover, there were few studies established a phase locked method for a 5.8 GHzmagnetron.

This work aims to build a 5.8 GHz power-variable phase-controlled magnetron (PVPCM)system with stable output power, frequency, and phase. In this study, we combine theinjection locking method and the PLL method by controlling a phase shifter to lock the phaseof the 5.8 GHz magnetron. The anode current of the magnetron is used to control the outputpower.

2.2 Injection-Locking Magnetron

As described in Sect. 1, the ordinary approach of the PCM was utilizing the injection lockingmethod and the PLL method[62]-[64]. The injection locking method is injecting a referencesignal to the magnetron to lock the oscillation frequency. The reference signal frequencyis set close to the self-oscillation frequency of the magnetron. The lock frequency rangeis expressed as Equations (2.1) and (2.2) [71], where Δ 𝑓 is the lock frequency range, 𝑓 isthe reference signal frequency, 𝑃𝑖 is the reference signal power, 𝑃𝑜 is the magnetron outputpower, and 𝑄𝑒 is the external Q-factor of the magnetron. Then, the phase difference \between the reference signal and the magnetron can be expressed by using the frequencydifference Δ 𝑓 ′ between the self-oscillation frequency and the reference signal frequency, asfollows,

Δ 𝑓

𝑓=

2𝑄𝑒

√𝑃𝑖

𝑃𝑜(2.1)

\ = sin−1 Δ 𝑓′𝑄𝑒

𝑓

√𝑃𝑖𝑃𝑜

(2.2)

These equations show that in the lock frequency range, the frequency and phase of magnetronoutput are locked with the reference signal. Therefore, controlling the reference signalparameters achieves the frequency and phase synchronization of the magnetron output.

2.2. INJECTION-LOCKING MAGNETRON 23

In our experiments, a reference signal was injected into the magnetron via a circulator asshown in Fig. 2.1. The reference signal frequency was 5.773 GHz which was the same as theself-oscillation frequency of a 5.8 GHz magnetron (Panasonic, M5802) in this experimentand the reference signal was amplified to 5 W which is high enough to lock the magnetron.The 5.8 GHz CW magnetron was supplied by a DC high voltage power supply with cathodevoltage -4.84 kV and anode current 150 mA. The filament current was 7.4 A by AC 3.35 V.The magnetron output was connected to a 500 W dummy load.

Reference signal

loadAmplifier

Circulator

5.8GHz

Magnetron

High-Voltage

Power Supply

Network analyzer

Fig. 2.1 Schematic of the injection locked magnetron.

Figure 2.2 shows the spectrum of an injection locking magnetron and a free-runningmagnetron in the max-hold mode. The output spectrum of magnetron was measured by aspectrum analyzer. The frequency of free-running magnetron was shifted in almost 5 MHzbandwidth and the frequency of injection locked magnetron was locked. We measured thelock frequency range with the injection power from 2 W to 10 W, the results are shown inFig. 2.3. Here, the measured results confirm that the lock frequency range is proportional tothe square root of the injection power as described in Eq. (2.1). When the injection powerwas 10 W (𝑃𝑖/𝑃𝑜:-15.2dB), the frequency was locked within almost 5 MHz band. Thephase difference between the reference signal and the injection locked magnetron output wasmeasured by the network analyzer as shown in Fig. 2.1. The phase reference was determinedat the cable end (calibration plane) of each network analyzer port. Figure 2.4 shows thephase difference when the magnetron was working inside the locked frequency range. Thephase shifted in almost ±15◦ with 60 Hz and a higher frequency noise. The 60 Hz noise


comes from the filament current and the higher frequency noise comes from the power supply.These results show that the injection locking method can lock the magnetron frequency withthe reference signal but the noise can not be avoided. Therefore it is necessary to control themagnetron phase by PLL method.

-70

-60

-50

-40

-30

-20

-10

0

5.74 5.76 5.78 5.80 5.82 5.84

Inte

nsi

ty [

dB

c]

Frequency [GHz]

Injection locked Free-running

Fig. 2.2 Injection locked spectrum and free-running spectrum of a 5.8 GHz Magnetron inmax hold mode.

2.2. INJECTION-LOCKING MAGNETRON 25

0

1

2

3

4

5

6

2 3 4 5 6 7 8 9 10

Lo

ck f

req

uen

cy r

ang

e Δ

f [M

Hz]

Injection power[W]

S/N012 S/N010 S/N007 S/N005

Fig. 2.3 Relationship of injection power and lock frequency range on 5.8 GHz magnetronswith difference serial numbers.

0

20

40

60

80

100

120

0 0.02 0.04 0.06 0.08 0.1

Mag

net

ron p

has

e dif

fere

nce

[°]

Time[s]

Injection locked magnetron

Fig. 2.4 Phase difference between the reference signal and the injection locked magnetronoutput.


2.3 Design of a PVPCM

2.3.1 Characteristics of 5.8 GHz Magnetrons

In order to design a PVPCM, we measured characteristics of the anode current and outputpower with the oscillation frequency of 5.8 GHz magnetrons. The measurement results inrated conditions are shown in Figs. 2.5 and 2.6.

5.760

5.765

5.770

5.775

5.780

5.785

5.790

5.795

5.800

5.805

5.810

0 50 100 150

Mag

net

ron

fre

qu

ency

[G

Hz]

Anode current [mA]

S/N012 S/N010 S/N009

S/N007 S/N005

Fig. 2.5 Relationship of anode current and oscillation frequency on 5.8 GHz magnetronswith difference serial numbers.

Our research group used to develop a 2.45 GHz PCM with the PLL method whichconsisted of a high voltage power source and a magnetron as a voltage-controlled oscillator(VCO), and the output signal feedback to control the anode current[67]. This PLL methodrequires that the anode current and the oscillation frequency of the magnetron have a linearrelationship like a VCO[70], [72]. However, from Fig. 2.5, these 5.8 GHz magnetrons do notshow any relationships between them. It is difficult to utilize the PLL method by controllingthe anode current to control the phase of the 5.8 GHz magnetron.

Equation (2.2) shows that controlling the reference signal phase can control the phaseof the magnetron output. Hence, we set an analog phase shifter to control the phase of thereference signal in the 5.8 GHz PCM. The magnetron output is fed back to control the phaseshifter, then a feedback loop is constituted for controlling the phase. This PLL method does

2.3. DESIGN OF A PVPCM 27

not require the characteristics of the anode current and the magnetron oscillation frequency.Then, the anode current is possible to control the output power of the magnetron as shownin Fig. 2.6. All the 5.8 GHz magnetrons had similar power-current characteristics. Thatwill make the power control system of magnetron much simpler. Based on this analysis,we control the phase of magnetron output by a phase shifter and control the power by theanode current of magnetron to design a PVPCM system as shown in Fig. 2.7 as a basicconfiguration.

0

50

100

150

200

250

300

350

400

0 50 100 150

Mag

net

ron o

utp

ut

[W]

Anode current [mA]

S/N012 S/N010 S/N009

S/N007 S/N005

Fig. 2.6 Relationship of anode current and output power on 5.8 GHz magnetrons withdifference serial numbers.


Reference signal

Phase shifter loadAmplifier

Circulator

5.8GHz

Magnetron

Directional coupler

Double Balanced Mixer

LO RF

IF

LPF

Control circuit

Magnetron power supply system

High-Voltage

Power Supply

Power control signal

Filament current supply

Fig. 2.7 Schematic diagram of a 5.8 GHz PVPCM.

2.3.2 Configuration of PLL

A block diagram of the PLL part is shown in Fig. 2.8. Here, \𝑜 (𝑡) is the phase of themagnetron, \𝑖 (𝑡) is a reference signal, and 𝐹𝑀 (𝑠) is the phase detection sensitivity of thephase comparator. The transfer functions of the low pass filter, phase control circuit and thephase shifter are 𝐹𝐿 (𝑠), 𝐹𝐶 (𝑠), 𝐹𝑃 (𝑠), respectively. The open loop transfer function of thephase control loop 𝐿 (𝑠) is expressed in Eq. (2.3).

𝐿 (𝑠) = 𝐹𝑀 (𝑠)𝐹𝐿 (𝑠)𝐹𝐶 (𝑠)𝐹𝑃 (𝑠) (2.3)

The output phase signal \𝑜 (𝑡) of the magnetron from the directional coupler is comparedwith the phase of reference signal \𝑖 (𝑡) by the double balanced mixer (DBM). The phasedifference signal is input to the control circuit to control the phase shifter. The phase shifterchanges the phase of the reference signal which is injected into the magnetron. By usingthe PLL, the phase difference between reference signal and output of magnetron graduallyconverges to 0◦.

2.3. DESIGN OF A PVPCM 29

LO

RF

IFLPF Control circuit Phase shifter

θi(t) θo(t)

FL(s) FC(s) FP(s)

FM(s)

Fig. 2.8 Block diagram of a phase control loop.

2.3.3 Determination of Transfer Functions

In this section, we define the transfer function of the block diagram, and design a PLL circuit.As the phase comparator, the DBM outputs the multiplication of two input signals. If the

harmonics are ignored, the transfer function 𝐹𝑀 (𝑠) of the mixer can be represented by thefollowing Eq. (2.4).

𝐹𝑀 (𝑠) = 𝐾𝑑 (2.4)

From measurements, 𝐾𝑑 value is 2.17×10−3 V/◦.The phase shifter is RVPT0408GBC RF-LAMBDA. It is an analog voltage control

phase shifter in 4-8 GHz band. From the datasheet of the phase shifter, the shift phase isproportional to the control voltage. From measurements, the proportional 𝐾𝑝 was 37.5◦/V,and the time constant 𝑇 was 2.20×10−7 s. This delay time contains the transfer time fromphase shifter to the directional coupler. We consider changing the phase has a first orderdelay with the control voltage changed, and the transfer function of the phase shifter can bewritten in Eq. (2.5).

𝐹𝑃 (𝑠) =𝐾𝑝

(1+ 𝑠𝑇) (2.5)

In this PVPCM system, a low pass filter is used to remove the high harmonics of theDBM output. In order to get a stable system, an incomplete integrated circuit was designedas a control circuit to control the phase shifter. The gain of the control circuit was related tothe steady state error. The larger gain makes a smaller steady state error. The capacitancein the circuit is related to the response time, hence all the capacitances value should be asmall value. The resistance and capacitance values are related to the system stability. Theseparameters of the low pass filter and the control circuit were adjusted to get stable gain andphase margin. Simultaneously, these parameters should make the steady state error as smallas possible, and the response speed as fast as possible. The low pass filter and the controlcircuit were finally designed as shown in Fig. 2.9. The transfer functions of the low pass filter


Vm

Vc

100KR4

820kR3

100pC1

3

21

NJM4558

3

21

NJM4558

47R1

330

R2

100p

C2

Vm

Vc

Fig. 2.9 Diagram of a low pass filter and a phase control circuit for a 5.8GHz PVPCM.

and the control circuit can be written as Eqs. (2.6) and (2.7), respectively. Here, the controlcircuit also can use the indefinite integration circuit which is without 𝑅3 in Fig. 2.9. Eq. (2.8)shows the transfer functions of the indefinite integration circuit that can work stable easily.

𝐹𝐿 (𝑠) =1

(1+ 𝑠𝑅1𝐶1)(2.6)

𝐹𝐶 (𝑠) = − 𝑅3/𝑅21+ 𝑠𝑅3𝐶2

(2.7)

𝐹𝐶 (𝑠) = − 1(1+ 𝑠𝑅2𝐶2)

(2.8)

The open loop transfer function of PLL 𝐿 (𝑠) is expressed as Eq. (2.9). The calculationformula of steady state error is written as Eq. (2.10). When the phase difference of \𝑜 and\𝑖 is set as the largest value 90◦ by a step response, the steady state error Y becomes 0.46◦.Bode diagrams of the open loop transfer function are shown in Fig. 2.10. From the bodediagrams, the gain margin 𝐺𝑚 was 38.6 dB and the phase margin 𝑃𝑚 was 63.7◦. This PLLcan be worked in a stable state.

𝐿 (𝑠) = −𝐾𝑝𝐾𝑑𝑅3

𝑅2(1+ 𝑠𝑇) (1+ 𝑠𝑅1𝐶1) (1+ 𝑠𝑅3𝐶2)(2.9)

Y = lim𝑡→∞

𝑒(𝑡) = lim𝑠→0

11+ 𝐿 (𝑠)

\𝑜 − \𝑖𝑠

𝑠 (2.10)

2.4. MEASUREMENT EXPERIMENTS 31

-300

-240

-180

-120

-60

0

1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 1.E+10

Ph

ase

[deg

]

Frequency [Hz]

100 102 104 106 108 1010

P.M.: 63.7deg

f: 353kHz

-250

-200

-150

-100

-50

0

50

100

1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 1.E+10

Gai

n [

dB

]

Frequency [Hz]

100 102 104 106 108 1010

G.M.: 38.6dB

f: 4.85MHz

Fig. 2.10 Bode diagrams of open-loop transfer function for a PLL.

2.4 Measurement Experiments

2.4.1 Outline of Experiments

Figure 2.11 shows an experimental schematic diagram of a 5.8 GHz PVPCM. A photo ofthe 5.8 GHz PVPCM is shown in Fig. 2.12. The 5.773 GHz reference signal was generatedfrom a signal generator, then went through a phase shifter and a 40 dB power amplifier. Thenthe reference signal was injected to the magnetron via a circulator. The magnetron outputpower was consumed by the dummy load. Also, a part of the reference signal was input tothe 𝐿𝑂 port of the DBM. The phase difference signal 𝑉𝑚 was measured by an oscilloscope,to observe the phase-locked state. When the phase is locked, the phase difference signal 𝑉𝑚value will be 0. The phase difference between the magnetron output from 𝐷𝑖𝑣𝑖𝑑𝑒𝑟1 and thereference signal from 𝐷𝑖𝑣𝑖𝑑𝑒𝑟2 was measured in a network analyzer.


A PC set a power control signal to the analog output of 16-bit DA converter (NI 9263).The analog output set a signal to a high voltage power supply to control the anode voltageand anode current. We set the anode voltage at -4.45 kV and adjusted the anode current tocontrol the magnetron output power. The average output power and the oscillation spectrumof the magnetron were measured by a power meter and a spectrum analyzer, respectively.The power signal was fed back to the PC by a GPIB communication and via a program.

Digital Oscilloscope

(Tektronix MSO2024)

Magnetron

(Panasonic

M5801J)

Circulator

High-Voltage

Power Supply

(Glassman

PS/LT005R360-20)

PC

Analog output

(NI 9263)

Directional coupler

Dummy load

(SPC).

PL

L c

on

trol c

ircu

itPower Meter

(Agilent 4419B)Spectrum Analyzer

(Agilent N9020A)

Network Analyzer

(Agilent E8364C)

Signal Generator

(Agilent N5183A)

40dB Amplifier

(R&K A252HP-R)

Phase shifter2

(RF-LAMBDA

RVPT0408GBC)

Double balance mixer

(R&K MX370)

Divider1

Divider2

Isolator

Isolator30dB

AC110V

Phase shifter1

(RF-LAMBDA

RVPT0408GBC)

S1

VmVc

LO

RF

IF

Transformer

GPIB-USB

(NI)

From

Divider2

To Network

Analyzer

Fig. 2.11 Experimental schematic diagram of a 5.8 GHz PVPCM.

2.4.2 Experimental Results

We set 150 mA of the anode current to make the magnetron worked in the rated condition.The injection power was 5 W. The output power was 329 W within 1% stability. The filamentcurrent of the magnetron was 7.4 A by AC 3.35 V. Before the PLL circuit worked, weobserved the phase difference between the reference signal and the magnetron output withthe 60 Hz fluctuation as shown in Fig. 2.4. When turning on the PLL circuit, the magnetronwas in a phase-locked state as shown in Fig. 2.13. The measurement results of the phase

2.4. MEASUREMENT EXPERIMENTS 33

Fig. 2.12 Photo of a 5.8 GHz PVPCM system (1: 5.8 GHz magnetron, 2: directional coupler,3: dummy load, 4: current meter, 5: circulator, 6: spectrum analyzer.)

difference and the output power are shown in Fig. 2.14. The 60 Hz noise was disappearedand the output power was in a stable state. Here, the phase difference was locked in anarrow range. Figure 2.15 shows the phase stability when the PLL circuit was working.The phase-locked stability was nearly ±1◦ and the phase difference was fixed to 0◦. Thewaveform of the PCM output was shown in Fig. 2.16.

Next, we set the injection power at 10 W and adjusted the magnetron output power.Setting the output power by 20 W step from 329 W, the phase of the magnetron is shownin Fig. 2.17. When the power was changed from 160 W to 329 W, the phase difference ofmagnetron was kept in a stable state.

Also, we measured the response time of PVPCM. In this system, the switch S1 changesthe phase of the phase shifter1. We measured the phase difference from the IF port of theDBM by oscilloscope to observe the response time. After the phase was shifted, the phasedifference returned to 0◦ within 69 `s as shown in Fig. 2.18.

When the magnetron output power was lower than 160W, the PVPCM went out thephase locked state. Also the anode current fluctuation can change the magnetron oscillationfrequency as shown in Fig. 2.5. When the anode current fluctuation or other environmentchanges make the frequency difference Δ 𝑓 ′ outside the lock frequency range as shown inFig. 2.3, the PVPCM system will be unlocked. The recovery process of the PVPCM systemis improving the inject power level to make the frequency difference Δ 𝑓 ′ inside the lockfrequency range. Then, resetting the power of control circuit, the PVPCM can be returned tothe locking state.


-100

-80

-60

-40

-20

0

20

0 20 40 60 80 100

Ph

ase

[°]

Time[ms]

No PLLPLL worked

Fig. 2.13 Magnetron phase difference when the PLL circuit starts working measured throughoscilloscope.

2.5 Summary

A 5.8 GHz PVPCM system was designed and demonstrated. With 5 W (𝑃𝑖/𝑃𝑜:-18.2dB)injection power, the phase-locked stability of the PVPCM was lower than ±1◦. The responsetime was less than 100 `s. The PVPCM output could be varied from 160 W to 329 W in aphase and power stability state with 10 W (𝑃𝑖/𝑃𝑜:-15.2dB) injection power.

The developed PVPCM system can be applied to other magnetrons which are not ap-plicable to the anode current PLL method, because it does not require a linear relationshipbetween the anode current and the oscillation frequency.

In the next, we will make several PVPCM components for a phased array system of WPT.Furthermore, the PVPCM is available for communication applications.

2.5. SUMMARY 35

140

160

180

200

220

240

260

280

300

320

340

-20

0

20

40

60

80

100

120

0 20 40 60 80 100

Ou

tpu

t p

ow

er[W

]

Mag

net

ron

ph

ase

dif

fere

nce

[°]

Time[s]

Phase difference Output power

Fig. 2.14 Magnetron phase difference and output power when the PLL circuit starts working.

-3

-2

-1

0

1

2

3

0 0.1 0.2 0.3 0.4 0.5

Mag

net

ron

ph

ase

dif

fere

nce

[°]

Time[s]

Phase locked mangetron

Fig. 2.15 Phase stability when the PLL circuit was working.


Fig. 2.16 Waveform of the PCM output and injection signal measured through oscilloscope.

0

50

100

150

200

250

300

350

-10

0

10

20

30

40

50

0 20 40 60 80 100

Outp

ut

pow

er[W

]

Mag

net

ron p

has

e dif

fere

nce

[°]

Time[s]

Phase difference Output power

Fig. 2.17 Magnetron phase difference when the output power was changed.

2.5. SUMMARY 37

PVPCM experiment results

21

-20

0

20

40

60

80

100

120

Phas

e diff

eren

ce[°

]

Time [μs]0 50 100

Reference signal

Phase shifter2

load

Control circuit

Divider Amplifier Circulator

Power signal 5.8GHzMagnetron

High-Voltage Power Supply

Comparer

Directional coupler

Double Balanced MixerLO RF

IF

Phase shifter1

RF-DC

Low pass filter

Response time less than 50 us.

Phase changing of the Phase shifter 1

Fig. 2.18 Response time of the PLL circuit when the phase was changed.

Chapter 3

Magnetron Phased Array

3.1 Introduction

Nikola Tesla proposed wireless power transfer (WPT) in far-field for more than one hundredyears [73]. Far-field WPT is more flexible to the transmission distance such as space solarpower stations [74] where the wires and near-field coupling cannot reach. Brown et al.promoted the development of WPT almost half a century ago. They developed the amplitron,which is one of the microwave amplification vacuum tubes. The DC-RF efficiency of theamplitron was more than 70% (@3 GHz, RF output power: 400 kW, 1964) [75]. The RF-DCrectification efficiency by diodes was achieved at 90% (@2.45 GHz, RF input power: 8 W)[76], and the DC-RF-RF-DC efficiency of a WPT system achieved 54 ± 1% at a distance of 2m (@2.45 GHz, DC receive power: 495 W, 1975) [77]. A long-distance WPT experiment wasdemonstrated over one mile at a microwave power of 450 kW generated from a continuouswave klystron. The rectified DC power reached 30 kW and it is the highest rectified powerlevel of the microwave WPT so far (@ 2.388 GHz, 1975) [78]. These experimental resultsare still high levels in WPT history.

Other research topics on WPT use phased array technologies. In 1992, MIcrowave LiftedAirplane eXperiment (MILAX) was conducted by using a 1.2 kW solid-state 2.45 GHzphased array system on a car to power a flying airplane. The airplane was driven by a stableDC rectified power of 88 W and flew at an altitude of 10 m and a flight distance of 350 m[79]. In 1993, the solid-state phased array [79] was also conducted in the ISY–METS rocketexperiment. A dipole antenna at different distances in space measured the received power[80]. Recently, a 1.8 kW solid-state 5.8 GHz phased array WPT was developed. A rectifiedDC power of 340 W with a distance of 55 m was obtained when a large majority of rectennaelements in the received panel were optimally activated by precise beamforming [81].

40 CHAPTER 3. MAGNETRON PHASED ARRAY

In the above WPT studies with phased array systems, the high-cost, low-efficiency, andlow power of the semiconductor phased array might be the biggest obstacle to apply forthe high-power WPT application relative to microwave vacuum electronic tubes. In thisstudy, we used the magnetrons, which are one of the microwave vacuum tubes, to builda phased array system. Magnetrons are widely used in microwave ovens. In contrast tothe semiconductor amplifiers, magnetrons have higher energy density and lower cost, butunstable frequency and phase. In 2001, a kW-class magnetron phased array system calledthe Space POwer Radio Transmission System (SPORTS) was developed [82]. The outputpower of the SPORTS, however, could not be controlled because the magnetron phase wascontrolled by its anode current. In 2015, Liu et al. developed a 2 × 2 2.45 GHz magnetronarray for near-field WPT. They obtained a rectified DC power of 70 W with a transmissiondistance of 5.5 m [83]. In 2019, Chen et al. used the master-slave injection-locking techniqueto develop a 2 × 2 magnetron array with an output power of over 3500 W at 2.45 GHz[84]. Nevertheless, the beamforming in these works is unavailable. In our previous study,we developed a different phase-controlled magnetron (PCM) method that could control themagnetron phase and power simultaneously [85]. Considering the development of a largewireless power transfer system, we propose and demonstrate a 2 × 2 phased array systembased on 5.8 GHz power-variable PCMs.

3.2 Principle of Power-Variable Phase-Controlled Mag-netrons and Magnetron Phased Array System

Fig. 3.1 shows a PCM system diagram. It uses the injection-locking method to lock themagnetron frequency and phase-locked-loop (PLL) method to control the magnetron phase.The anode current of the magnetron controls the magnetron output power.

3.2. PRINCIPLE OF POWER-VARIABLE PHASE-CONTROLLED MAGNETRONS AND MAGNETRON PHASED ARRAY SYSTEM41IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

II. PRINCIPLE OF POWER-VARIABLE PHASE-CONTROLLED

MAGNETRONS AND MAGNETRON PHASED ARRAY SYSTEM

Fig. 1 shows a PCM system diagram. It uses the injection-

locking method to lock the magnetron frequency and phase-locked-loop (PLL) method to control the magnetron phase. The anode current of the magnetron controls the magnetron output power.

In the injection-locking method for a magnetron, a reference signal with low power is injected into the magnetron through a circulator, and the output frequency of the magnetron can be locked with the injection signal [13]. The injection signal frequency is set close to that of the self-oscillation frequency of the magnetron. From the Alder equation [14], the locked

frequency range Δf was described as: Δf = 2𝑓 (𝑃𝑖/𝑃𝑜)/𝑄𝑒, where f is the injection signal frequency, 𝑃𝑖 is the injection signal power, 𝑃𝑜 is the magnetron output power, and 𝑄𝑒 is the external Q-factor of the magnetron. When the injection-locking condition is satisfied, the magnetron output frequency becomes stable by the injection signal. In this study, the injection signal from the signal generator (Agilent N5183A) is injected into the magnetron with an amplifier to achieve frequency synchronization.

In the PCM, the magnetron output phase can be locked with the injection signal via a control circuit of the PLL. As shown in Fig. 1, the phase difference between the output signal and the reference signal is fed back to the PLL control circuit The magnetron output phase will then be locked automatically. The PLL control circuit converts the phase difference signal to the driven voltage of the phase shifter β and shows the phase-lock state by an LED lamp. The phase shifter γ can adjust the phase difference between the magnetron output and the injection signal. The previous PCM experiments showed the PLL response time is less than 50 μs, phase control accuracy less than ±1°, and the output power varies from 160 W to 450 W [13]. When the output power was lower than 160 W, the self-oscillation frequency of the magnetron would deviate from the locked frequency range. The frequency and phase-locking of the magnetron failed. Enlarging the injection signal power or closing the injection signal frequency can solve this problem.

A magnetron phased array system requires all the magnetrons worked at the same frequency. The manufacturing error, however, causes various self-oscillation frequencies in different magnetrons. According to the Alder equation, the locked frequency range Δf depends on the power ratio. To keep all the magnetrons working at the same frequency, we selected the magnetrons with similar self-oscillation frequencies. Then, a lower injection power can lock all the magnetrons. The anode current also affects the self-oscillation frequency of the magnetron. When controlling the output power of the magnetron, it should be worked in the locked frequency range Δf of the magnetron to prevent the PCM from the unlocked state. Fig. 2 shows the upper and lower

frequency boundaries, within which the magnetron frequency could be locked to the injection signal frequency. The frequency boundaries of four magnetrons were measured to the injection power, and the red, blue, green and orange colors correspond to the serial numbers of the measured magnetrons. For instance, the area indicated by the black arrow is the locked frequency range Δf of magnetron (S/N005) at an injection power of 7 W. All the magnetrons can be locked in the green zone where the magnetrons can be used as phased array transmitters.

III. DESIGN OF SLOT ARRAY ANTENNAS

In this study, we used four PCMs to build a 2 × 2

Magnetron(Panasonic

M5802)

Circulator

High-Voltage Power Supply

(Glassman PS/LT005R360-20)

PC

Analog output(NI 9263)

Directional coupler

Output Load

PLL control circuit

Signal Generator(Agilent N5183A)

40dB Amplifier(R&K A252HP-R)

Phase shifter β(RF-LAMBDA

RVPT0408GBC)

Double balance mixer(R&K MX370)

Divider

Isolator

Isolator

30dB

Phase shifter α(RF-LAMBDA

RVPT0408GBC)

LO

RFIF

Heat circuit

Phase-Controlled Magnetron(PCM)

Phase shifterγ

Driver circuit

Fig.1 Diagram of a power-variable phase-controlled magnetron. system.

Fig.2 Measured upper and lower frequency boundaries of magnetrons to the injection power. The magnetron frequency could be locked to the injection signal frequency within the boundary. red line: S/N007, blue line: S/N012, green line: S/N010, orange line: S/N 005. The green zone indicates the locking range within which all the four magnetrons can be locked.

5.782

5.784

5.786

5.788

5.790

5.792

0 1 2 3 4 5 6 7 8 9 10

Freq

uenc

y [G

Hz]

Injection Power [W]

△f

Fig. 3.1 Diagram of a power-variable phase-controlled magnetron.

In the injection-locking method for a magnetron, a reference signal with low power isinjected into the magnetron through a circulator, and the output frequency of the magnetroncan be locked with the injection signal [85]. The injection signal frequency is set close to thatof the self-oscillation frequency of the magnetron. From the Alder equation [86], the lockedfrequency range Δ 𝑓 was described as: Δ 𝑓 = 2 𝑓

√((𝑃𝑖/𝑃𝑜))/𝑄𝑒 , where 𝑓 is the injection

signal frequency, 𝑃𝑖 is the injection signal power, 𝑃𝑜 is the magnetron output power, and 𝑄𝑒

is the external Q-factor of the magnetron. When the injection-locking condition is satisfied,the magnetron output frequency becomes stable by the injection signal. In this study, theinjection signal from the signal generator (Agilent N5183A) is injected into the magnetronwith an amplifier to achieve frequency synchronization.

In the PCM, the magnetron output phase can be locked with the injection signal via acontrol circuit of the PLL. As shown in Fig. 3.1, the phase difference between the outputsignal and the reference signal is fed back to the PLL control circuit The magnetron outputphase will then be locked automatically. The PLL control circuit converts the phase differencesignal to the driven voltage of the phase shifter 𝛽 and shows the phase-lock state by an LEDlamp. The phase shifter 𝛾 can adjust the phase difference between the magnetron outputand the injection signal. The previous PCM experiments showed the PLL response time


is less than 50 `𝑠, phase control accuracy less than ±1◦, and the output power varies from160 W to 450 W [85]. When the output power was lower than 160 W, the self-oscillationfrequency of the magnetron would deviate from the locked frequency range. The frequencyand phase-locking of the magnetron failed. Enlarging the injection signal power or closingthe injection signal frequency can solve this problem.

A magnetron phased array system requires all the magnetrons worked at the samefrequency. The manufacturing error, however, causes various self-oscillation frequenciesin different magnetrons. According to the Alder equation, the locked frequency range Δ 𝑓

depends on the power ratio. To keep all the magnetrons working at the same frequency, weselected the magnetrons with similar self-oscillation frequencies. Then, a lower injectionpower can lock all the magnetrons. The anode current also affects the self-oscillationfrequency of the magnetron. When controlling the output power of the magnetron, it shouldbe worked in the locked frequency range Δ 𝑓 of the magnetron to prevent the PCM fromthe unlocked state. Fig. 3.2 shows the upper and lower frequency boundaries, within whichthe magnetron frequency could be locked to the injection signal frequency. The frequencyboundaries of four magnetrons were measured to the injection power, and the red, blue, greenand orange colors correspond to the serial numbers of the measured magnetrons. For instance,the area indicated by the black arrow is the locked frequency range Δ 𝑓 of magnetron (SN005)at an injection power of 7 W. All the magnetrons can be locked in the green zone where themagnetrons can be used as phased array transmitters.

3.3. DESIGN OF SLOT ARRAY ANTENNAS 43

-50

-40

-30

-20

-10

0

10

5.2 5.4 5.6 5.8 6 6.2 6.4

Am

pli

tud

e [d

B]

Frequency[GHz]

No.1

No.2

No.3

No.4

Simulation

-60

-50

-40

-30

-20

-10

0

10

-90 -67.5 -45 -22.5 0 22.5 45 67.5 90

Am

pli

tud

e[d

Bc]

Angle[°]

Simulation

No.1

No.2

No.3

No.4

5.782

5.784

5.786

5.788

5.790

5.792

0 1 2 3 4 5 6 7 8 9 10

Fre

qu

ency

[G

Hz]

Injection Power [W]

-50

-40

-30

-20

-10

0

10

-25 -20 -15 -10 -5 0 5 10 15 20 25

Am

pli

tud

e [d

Bc]

Angle [°]

0°60°120°-120°-60°

Fig. 3.2 Measured upper and lower frequency boundaries of magnetrons to the injectionpower. The magnetron frequency could be locked to the injection signal frequency within theboundary. red line: S/N007, blue line: S/N012, green line: S/N010, black line: S/N 005. Thegreen zone indicates the locking range within which all the four magnetrons can be locked.

3.3 Design of Slot Array Antennas

In this study, we used four PCMs to build a 2 × 2 magnetron phased array system. Thissystem required the antenna having a high gain and being able to work at a high-power level.A horn antenna, a parabolic antenna, and a slot antenna are satisfied under this condition.The horn antenna and parabolic antenna, however, require a larger volume. Consideringreliability and processing cost, we chose planar slot antennas to connect to the PCM output.

We designed a back forward-wave slot array antenna using the CST studio software.Fig. 3.3 shows the dimensional parameters of the slot array antenna.The feed waveguideworks like a divider and provides the same amplitude and phase to each element. The slotantennas were constructed as a horizontal array and fed by a slot waveguide. The slot topand slot space were set as half of the wavelength in the waveguide. The wavelength in

the waveguide can be calculated by√((1/_)2 − (1/_𝑐𝑢𝑡𝑜 𝑓 𝑓 )2)). Here, _𝑐𝑢𝑡𝑜 𝑓 𝑓 is twice the

waveguide width. The waveguide standard is WR-159 (waveguide width 40.3 mm, height20.15 mm). The wavelength in the waveguide at 5.8 GHz is 67.36 mm. To achieve a widebandwidth, the radiation waveguide (slot antenna) was designed in the travelling wave mode.The slot space of the slot antenna was a little shorter than half-wavelength in the waveguideand there is a phase difference between each slot. Thus, the main beam of the slot antenna


has an angle deviation from the center. The slot antennas were constructed as a horizontalarray and fed by the feed waveguide. Based on the overall dimensions of the magnetronphased array, the number of slots in a radiating waveguide was calculated as 14, and thenumber of slots in the feed waveguide was calculated as 8.

Through the simulation, the radiation capacity of the slot basically depends on the slotlength, but it is not sensitive with the slot offset. The radiation phase of a slot depends onboth slot length and slot offset. The output amplitude and phase of each slot of the feedwaveguide are supposed to be equal. Then the value of slot length and slot offset can bedetermined. The optimization of the CST studio fine-tuned the design parameters of the slotarray antenna, as listed in Table 3.1. The slot width was set to 2.6 mm. Fig. 3.4 shows thesimulation model and a 3-D beam pattern. On the H plane, the main beam of the slot antennahas a 22.5◦ angle deviation from the center. The developed slot array antenna has 112 slotsand a volume of 494 mm × 290 mm × 40 mm.

Since the slot array antenna designed by CST Studio is manufactured by sheet metalprocessing, it is almost impossible to directly manufacture it with a 3D model. To solvethis problem, we used one aluminum plate for the slot surface of the eight parallel radiatingwaveguides and bent both sides to increase the strength. 8 groove type aluminum platesare arranged side by side and fixed to the slot surface with screws. Temporarily tighten thefeed waveguide and screws on both sides (bent surface) of the aluminum plate on the slotsurface. The input port of the slot array antenna is connected to the network analyzer via thewaveguide-coaxial cable converter. Finely adjust the positions of the feeding and radiatingwaveguides while looking at the |𝑆11 | parameter. When the |𝑆11 | parameter reaches theminimum value at the center frequency of the magnetron, fully tighten the screws. Affix allgaps with aluminum tape to prevent microwave leakage. As described above, the assemblyof the slot array antenna is completed. Fig. 3.5 shows photographs of the slot array antenna.


Slot offset

Slot length

Slot width

Slot top

Slot space

Feed Waveguide

Radiation Waveguide

No.1

No.14

No.8

Fig. 3.3 Configurations of the developed slot array antenna.


Fig. 3.4 Simulation results of the slot array antenna.

Front Rear

Fig. 3.5 Photographs of the developed slot array antenna.


Table 3.1 DESIGN PARAMETERS OF THE SLOT ARRAY ANTENNA

Feed waveguide(mm) Radiation waveguide(mm)No. length offset No. length offset No. length offset1 24.3 4.2 1 23.2 3.9 9 24.8 8.22 23.3 4.8 2 23.6 3.6 10 24.7 7.93 24.8 5.9 3 23.6 3.7 11 24.6 7.64 22.8 6.7 4 23.8 4.5 12 24.6 7.45 26.5 7.2 5 24.0 5.2 13 24.8 8.66 26.4 6.5 6 24.3 6.0 14 25.0 10.47 25.9 4.7 7 24.5 7.08 25.3 4.4 8 24.7 8.0

In this study, we manufactured and assembled four slot array antennas and measuredtheir |𝑆11 | parameter using a network analyzer (Agilent E8364C). Fig. 3.6 shows the simu-lation and measurement results of the four antennas. The simulation was –21dB while themeasurement results of |𝑆11 | on 5.8 GHz were –28 dB, –23 dB, –23 dB, and –18 dB. Due tothe manufacturing processing error and the assembly error, the measurement results of |𝑆11 |slightly deviate from the simulation results. The beam pattern of the slot array antenna wasmeasured in an anechoic chamber. Fig. 3.7 shows the measurement and simulation resultsof the beam pattern at 5.8 GHz. As shown in Fig. 3.7, the measurement results concurredwith the simulation. The designed gain of the slot array antenna was 24.9 dBi (side lobe:–15.6 dB; main beam angle: 22.5◦; half-value width: 10◦), while the measured gain of theNo.1 antenna was 23.9 dBi (side lobe: –14.6 dB; main beam angle: 22.4◦; half-value width:11.1◦). The measured gains of the No. 2, No. 3, and No. 4 antennas were 23.1 dBi, 22.8 dBi,and 23.4 dBi, respectively. The volume of the manufactured 2 × 2 array antenna is 988 mmlength × 580 mm width × 40 mm depth. The measured gain of the 2 × 2 array constructed bythe developed phased array antennas was 29.2 dBi.


-50

-40

-30

-20

-10

0

10

5.2 5.4 5.6 5.8 6 6.2 6.4

|S11|[d

B]

Frequency[GHz]

No.1

No.2

No.3

No.4

Simulation

-60

-50

-40

-30

-20

-10

0

10

-90 -67.5 -45 -22.5 0 22.5 45 67.5 90

Am

pli

tude

[dB

c]

Angle[°]

Simulation

No.1

No.2

No.3

No.4

5.782

5.784

5.786

5.788

5.790

5.792

0 1 2 3 4 5 6 7 8 9 10

Fre

qu

ency

[G

Hz]

Injection Power [W]

-50

-40

-30

-20

-10

0

10

-25 -20 -15 -10 -5 0 5 10 15 20 25

Am

pli

tud

e [d

Bc]

Angle [°]

0°60°120°-120°-60°

Fig. 3.6 |𝑆11 | of the slot array antenna.

-50

-40

-30

-20

-10

0

10

5.2 5.4 5.6 5.8 6 6.2 6.4

Am

pli

tud

e [d

B]

Frequency[GHz]

No.1

No.2

No.3

No.4

Simulation

-60

-50

-40

-30

-20

-10

0

10

-90 -67.5 -45 -22.5 0 22.5 45 67.5 90

Am

pli

tud

e[d

Bc]

Angle[°]

Simulation

No.1

No.2

No.3

No.4

5.782

5.784

5.786

5.788

5.790

5.792

0 1 2 3 4 5 6 7 8 9 10

Fre

qu

ency

[G

Hz]

Injection Power [W]

-50

-40

-30

-20

-10

0

10

-25 -20 -15 -10 -5 0 5 10 15 20 25

Am

pli

tud

e [d

Bc]

Angle [°]

0°60°120°-120°-60°

Fig. 3.7 Beam patterns of the slot array antenna on the H-plane at 5.8 GHz.

3.4. DEMONSTRATION EXPERIMENTS OF MAGNETRON PHASED ARRAY 49

3.4 Demonstration Experiments of Magnetron Phased Ar-ray

We developed a 2 × 2 high-power phased array by using four PCMs and four slot arrayantennas. Each PCM was connected to a slot array antenna. Fig. 3.8 shows the blockdiagram of the magnetron phased array system. The signal generator outputs the injectionpower via a divider and phase shifters to the PCMs. The magnetrons were operated at thesame frequency when they worked in the injection-locking status. The phase shifters 𝛿1-4 were used to synchronize the input phases of the injection signals to the PCMs 1-4. ALabVIEW programme remotely controlled the magnetron output power and filament power.The LabVIEW programme uploads the beamforming coefficients into the phase shifters 𝛼,shown in PCMs through the voltage output unit (NI 9263, 16-bit Voltage Output Module).Fig. 3.9 shows the photograph of the magnetron phased array system.

PC

LabVIEW

Signal Generator

(Agilent N5183A)

Isolator

Phase shifter δ1 Phase shifter δ2 Phase shifter δ3 Phase shifter δ4

PCM1

Power supply 1

Phase shifter α1

PCM2

Power supply 2

Phase shifter α2

PCM3

Power supply 3

Phase shifter α3

PCM4

Power supply 4

Phase shifter α4

Slot antenna 1 Slot antenna 2 Slot antenna 3 Slot antenna 4

Analog output

(NI 9263)

Divider

Fig. 3.8 Block diagram of the 5.8GHz phased array.

50 CHAPTER 3. MAGNETRON PHASED ARRAYIEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

magnetrons are locked. Afterward, the PLL control circuits are

turned on to stabilise and manually adjust the magnetron

output phases. When all the magnetrons worked in the phase-

locking state, a vector voltmeter (Keysight N9928A) was

connected to directional couplers to detect and upload the data

of the phase difference and amplitude ratio. Finally, the phase

shifters and anode currents were automatically adjusted to

achieve the same output phase and output power by the

LabVIEW programme. To verify the performance of the

magnetron phased array system, we conducted beamforming

and wireless power transfer experiments.

A. Beam Forming Experiments

The phase shifter α was controlled to change the magnetron

output phase. The beam directivity of the array antenna is

related to the magnetron’s phases. As shown in (1), the

magnetron output phase difference can calculate the main lobe

deflection angle �. Here, ∆� is the phase difference between

the two adjacent antennas, d is the distance between the two

adjacent antennas, λ is the wavelength. In this system, the

horizontal and vertical distances between the element centers

are 290 mm and 470 mm, respectively. When the incoming

phase differences of horizontal and vertical adjacent elements

are respectively equalled to 107° and 116°, the main lobe

angle can be controlled to ±3° in horizontal and ± 2° vertical

directions. By uploading the above phase strategy to the

control system, the beam scanning can cover an area of ±3° in

horizontal and 22.5° ± 2° in vertical. Fig. 10 shows the

simulation results of the combined beam patterns of the 2 × 2

slot antenna array at 5.8 GHz.

sin � ��∆�

2� 1�

In the experiments, we used a 1.2 × 1.2 metre LED lamp

rectenna array consisting of 600 rectenna elements to display

the validity of beam scanning. The rectenna element receives

the microwave and rectifies the DC power to light the

corresponding LED lamp. The luminance of the lamps can

also symbolize the intensity of the received microwave. Fig.

11 shows the photograph of the LED lamp rectenna elements.

We placed the LED lamp array about 8.5 m in front of the

magnetron phased array and 4.5 m in height, as shown in Fig.

12. The LED lamp array was 9.2 m away from the center of

the slot array antennas. The elevation angle of the vertical

direction from the phased array antennas to the center of the

LED lamp array became 22.5°, which is exactly the radiating

angle of the array antennas.

After finishing setting the magnetron phased array, we

adjusted all the anodes current of the PCMs to 100 mA. At this

time, each PCM output was kept in the same phase. State 1 in

the upper section of Fig. 12 shows the appearance of the LED

lamp rectenna array under this situation. Adjusting the phase

shifters α which are analog phase shifters, could control the

beam in high resolution to scan an area both in vertical and

horizontal directions and the LED lamp array could display

Fig. 9. Photograph of the magnetron phased array system.

1300 mm

2060 mm

1420 mm

PCM3 PCM4

PCM1 PCM2

Fig. 10. Simulation results of the combined beam patterns of the 2 × 2 slot antenna array on the E plane at 5.8GHz.

-50

-40

-30

-20

-10

0

10

-25 -20 -15 -10 -5 0 5 10 15 20 25

Am

pli

tud

e [d

Bc]

Angle [°]

0°60°120°-120°-60°

Fig. 11. Photograph of the LED lamp rectenna elements.

Fig. 3.9 Photograph of the magnetron phased array system.

The operation procedure of the magnetron phased array is as follows. First, all themagnetrons are turned on and kept for close to 30 s, then the LabVIEW programme turns offthe filament power to obtain relative stable outputs. Next, the injection frequency and powerof each magnetron are respectively set as 5.788 GHz and 6 W, then all the magnetrons arelocked. Afterward, the PLL control circuits are turned on to stabilise and manually adjustthe magnetron output phases. When all the magnetrons worked in the phase-locking state,a vector voltmeter (Keysight N9928A) was connected to directional couplers to detect and


upload the data of the phase difference and amplitude ratio. Finally, the phase shifters andanode currents were automatically adjusted to achieve the same output phase and outputpower by the LabVIEW programme. To verify the performance of the magnetron phasedarray system, we conducted beamforming and wireless power transfer experiments.

3.4.1 Beam Forming Experiments

The phase shifter 𝛼 was controlled to change the magnetron output phase. The beamdirectivity of the array antenna is related to the magnetron’s phases. As shown in equation(3.1), the magnetron output phase difference can calculate the main lobe deflection angle\. Here, Δ𝜑 is the phase difference between the two adjacent antennas, 𝑑 is the distancebetween the two adjacent antennas, _ is the wavelength. In this system, the horizontal andvertical distances between the element centers are 290 mm and 470 mm, respectively. Whenthe incoming phase differences of horizontal and vertical adjacent elements are respectivelyequalled to 107◦ and 116◦, the main lobe angle can be controlled to ±3◦ in horizontal and ±2◦ vertical directions. By uploading the above phase strategy to the control system, the beamscanning can cover an area of ±3◦ in horizontal and 22.5◦ ± 2◦ in vertical. Fig. 3.10 showsthe simulation results of the combined beam patterns of the 2 × 2 slot antenna array at 5.8GHz.

sin\ =_Δ𝜑

2𝜋𝑑(3.1)


-50

-40

-30

-20

-10

0

10

5.2 5.4 5.6 5.8 6 6.2 6.4

Am

pli

tud

e [d

B]

Frequency[GHz]

No.1

No.2

No.3

No.4

Simulation

-60

-50

-40

-30

-20

-10

0

10

-90 -67.5 -45 -22.5 0 22.5 45 67.5 90

Am

pli

tude

[dB

c]

Angle[°]

Simulation

No.1

No.2

No.3

No.4

5.782

5.784

5.786

5.788

5.790

5.792

0 1 2 3 4 5 6 7 8 9 10

Fre

qu

ency

[G

Hz]

Injection Power [W]

-50

-40

-30

-20

-10

0

10

-25 -20 -15 -10 -5 0 5 10 15 20 25

Am

pli

tud

e [d

Bc]

Angle [°]

0°60°120°-120°-60°

Fig. 3.10 Simulation results of the combined beam patterns of the 2 × 2 slot antenna array onthe E plane at 5.8 GHz.

In the experiments, we used a 1.2 × 1.2 metre LED lamp rectenna array consisting of 600rectenna elements to display the validity of beam scanning. The rectenna element receives themicrowave and rectifies the DC power to light the corresponding LED lamp. The luminanceof the lamps can also symbolize the intensity of the received microwave. Fig. 3.11 shows thephotograph of the LED lamp rectenna elements. We placed the LED lamp array about 8.5 min front of the magnetron phased array and 4.5 m in height, as shown in Fig. 3.12. The LEDlamp array was 9.2 m away from the center of the slot array antennas. The elevation angle ofthe vertical direction from the phased array antennas to the center of the LED lamp arraybecame 22.5◦, which is exactly the radiating angle of the array antennas.


Front Rear

Fig. 3.11 Photograph of the LED lamp rectenna elements.

After finishing setting the magnetron phased array, we adjusted all the anodes currentof the PCMs to 100 mA. At this time, each PCM output was kept in the same phase. State1 in the upper section of Fig. 3.12 shows the appearance of the LED lamp rectenna arrayunder this situation. Adjusting the phase shifters 𝛼 which are analog phase shifters, couldcontrol the beam in high resolution to scan an area both in vertical and horizontal directionsand the LED lamp array could display the scanning area, i.e. up, down, left, and right, asdepicted as states 2, 3, 4, and 5 in Figs. 3.12, 3.13. Fig. 3.13 also shows the simulation beampattern of the states 1 to 5. The phase difference among PCMs and beam direction are listedin Table 3.2. Using the LabVIEW programme to realise a timing cycle switching of states 2to 5, it can simulate the application of multiple charging points and those targets could becharged in turn. The system worked well for more than 10 minutes and the beam scanningoperation continuously kept at a precise level. During the state-switching procedure, thecenter of the main lobe was deflected about one metre, which agreed with the theoreticalprediction. As a comparative experiment, the invalid beamforming state when turning offthe injection signal is shown as state 6 in Figs. 12, 13. In the experiments, we verified themicrowave beamforming performance of the magnetron phased array with the slot arrayantennas. The PCMs could work stably for a long time, also demonstrating the phase andamplitude-controlled capabilities of the PCMs.


Table 3.2 PHASE DIFFERENCE AMONG PCMS AND BEAM DIRECTION

Phase difference Beam directionState PCM1 PCM2 PCM3 PCM4 Coordinate1 0◦ 0◦ 0◦ 0◦ 0◦, 0◦

2 0◦ 120◦ 60◦ 180◦ –3◦, 2◦

3 60◦ 180◦ 0◦ 120◦ –3◦, –2◦

4 180◦ 60◦ 120◦ 0◦ 3◦, –2◦

5 120◦ 0◦ 180◦ 60◦ 3◦, 2◦

6 Different frequency –, –

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

the scanning area, i.e. up, down, left, and right, as depicted as

states 2, 3, 4, and 5 in Figs. 12, 13. Fig. 13 also shows the

simulation beam pattern of the states 1~5. The phase

difference among PCMs and beam direction are listed in Table

II. Using the LabVIEW programme to realise a timing cycle

switching of states 2 to 5, it can simulate the application of

multiple charging points and those targets could be charged in

turn. The system worked well for more than 10 minutes and

the beam scanning operation continuously kept at a precise

level.

During the state-switching procedure, the center of the main

lobe was deflected about one metre, which agreed with the

theoretical prediction. As a comparative experiment, the

invalid beamforming state when turning off the injection

signal is shown as state 6 in Figs. 12, 13. In the experiments,

we verified the microwave beamforming performance of the

magnetron phased array with the slot array antennas. The

PCMs could work stably for a long time, also demonstrating

the phase and amplitude-controlled capabilities of the PCMs.

TABLE II

PHASE DIFFERENCE AMONG PCMS AND BEAM DIRECTION

B. Wireless Power Transfer Experiments

Fig. 14 shows a 5.8 GHz rectenna array system (IHI Aero

Space) that was used as a microwave power receiver in the

WPT experiments [9]. This rectenna array system consists of

36 sub-arrays, each sub-array has 64 rectenna elements that

can tolerate a maximum microwave power intensity of 378

W/m2. These sub-arrays form a symmetrical octagonal

structure. In the center of the octagonal structure, there is a

null space for the retro directive module of the microwave

power transmission, which was not applied in the experiments.

The rectenna circuit element can achieve a maximum

microwave rectification efficiency of 63% at 200 mW to 800

mW input power. All the rectenna elements of a sub-array

combine the DC power and output to the load control unit,

keeping the load resistance working at 24 Ω. The rectified

output DC power of each sub-array was monitored in real-time.

The 36 load control units are connected to the synthesis unit

and charge the battery unit. There are 24 sensors installed in a

cross-shape to detect the microwave power density. Through

these sensors, we can observe the microwave whether it

exceeds the rated power density and spot the position of the

main lobe. A monitor PC communicates with the control units

the through optical fiber, reads the data of all sensors, and

controls the connection relays of each part. Combing the DC

output of the 2304 rectenna elements, the best rectification

efficiency can reach 50%. Furthermore, the rectenna array

Phase difference Beam direction

State PCM1 PCM2 PCM3 PCM4 Coordinate

1 0° 0° 0° 0° 0°，0°

2 0° 120° 60° 180° −3°，2°

3 60° 180° 0° 120° −3°，−2°

4 180° 60° 120° 0° 3°，−2°

5 120° 0° 180° 60° 3°，2°

6 Different frequency -，-

Fig. 12. Beam forming experiments of the magnetron phased array system.

8.5 m

4.5 m 9.2 m

22.5°

1 2 3 4 5 6

Fig. 14. 5.8 GHz rectenna array system.

2.7 m

2.7 m

3.5 m

Fig. 13. Display of the LED lamp array and simulation beam pattern of

states 1~5. (Black frame: LED lamp array)

1

2

3 4

5

Fig. 3.12 Beam forming experiments of the magnetron phased array system.

3.4. DEMONSTRATION EXPERIMENTS OF MAGNETRON PHASED ARRAY 55IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES










level.










TABLE II






























1 0° 0° 0° 0° 0°，0°

2 0° 120° 60° 180° −3°，2°

3 60° 180° 0° 120° −3°，−2°

4 180° 60° 120° 0° 3°，−2°

5 120° 0° 180° 60° 3°，2°



8.5 m

4.5 m 9.2 m

22.5°

1 2 3 4 5 6


2.7 m

2.7 m

3.5 m



1

2

3 4

5

Fig. 3.13 Display of the LED lamp array and simulation beam pattern of states No.1 No.5(Black frame: LED lamp array).

3.4.2 Wireless Power Transfer Experiments

Fig. 3.14 shows a 5.8 GHz rectenna array system (IHI Aero Space) that was used as amicrowave power receiver in the WPT experiments [81]. This rectenna array system consistsof 36 sub-arrays, each sub-array has 64 rectenna elements that can tolerate a maximummicrowave power intensity of 378𝑊/𝑚2. These sub-arrays form a symmetrical octagonalstructure. In the center of the octagonal structure, there is a null space for the retro directivemodule of the microwave power transmission, which was not applied in the experiments.The rectenna circuit element can achieve a maximum microwave rectification efficiency of63% at 200 mW to 800 mW input power. All the rectenna elements of a sub-array combinethe DC power and output to the load control unit, keeping the load resistance working at24 Ω. The rectified output DC power of each sub-array was monitored in real-time. The 36load control units are connected to the synthesis unit and charge the battery unit. There are24 sensors installed in a cross-shape to detect the microwave power density. Through thesesensors, we can observe the microwave whether it exceeds the rated power density and spotthe position of the main lobe. A monitor PC communicates with the control units the throughoptical fiber, reads the data of all sensors, and controls the connection relays of each part.


Combing the DC output of the 2304 rectenna elements, the best rectification efficiency canreach 50%. Furthermore, the rectenna arraysystem can adjust the pitch angle using a wheel.

As the power transmitter, we measured the efficiency of the magnetron phased arraysystem. The measured maximum output power of the magnetron phased array was 1680W when the anode current and voltage of each magnetron output is 187 mA and 3680V,respectively. The DC-RF efficiency of the magnetron phase array is 61.0%.

The magnetron phased array system and the rectenna array system were placed in theanechoic chamber, as shown in Fig. 3.15. The pitch angle of the rectenna array was setat –22.5◦. The horizontal distance between the center of the magnetron phased array andrectenna array was 5 m. The anode current of each PCM was set as 150 mA. We measured thetransmitted microwave power and rectified DC power, as shown in Table 3.3. Two operatingmagnetrons, i.e. PCM1 and PCM2 measured the result cases No.1 to No.4. They representthe circumstances of the filament ON state, filament OFF state, injection-locking state, andphase-locked state, respectively.











level.










TABLE II






























1 0° 0° 0° 0° 0°，0°

2 0° 120° 60° 180° −3°，2°

3 60° 180° 0° 120° −3°，−2°

4 180° 60° 120° 0° 3°，−2°

5 120° 0° 180° 60° 3°，2°



8.5 m

4.5 m 9.2 m

22.5°

1 2 3 4 5 6


2.7 m

2.7 m

3.5 m



1

2

3 4

5

Fig. 3.14 5.8 GHz rectenna array system.


Table 3.3 EXPERIMENTAL RESULTS OF WIRELESS POWER TRANSFER

No PCM1(W) PCM2(W) PCM3(W) PCM4(W) DC Power(W) [𝑅𝐹−𝑅𝐹−𝐷𝐶1 365 339 0 0 60 8.52%2 340 346 0 0 65 9.48%3 349 339 0 0 68 9.69%4 345 333 0 0 75 10.85%5 0 0 316 319 76 11.97%6 339 336 319 310 142 10.89%7 305 306 0 0 43 7.04%8 0 306 0 330 40 6.29%9 0 0 295 330 47 7.52%10 305 0 295 0 43 7.17%

Table 3.4 EXPERIMENTAL RESULTS OF DIFFERENT DISTANCES

Distance 2.7 m 3.0 m 3.5 m 3.8 m 4.3 m 4.7 m 5.2 mDC Power(W) 67 65 70 72 70 63 54[𝑅𝐹−𝑅𝐹−𝐷𝐶 7.77% 7.54% 8.12% 8.35% 8.12% 7.31% 6.26%


system can adjust the pitch angle using a wheel.

As the power transmitter, we measured the efficiency of the

magnetron phased array system. The measured maximum

output power of the magnetron phased array was 1680 W

when the anode current and voltage of each magnetron output

is 187 mA and 3680V, respectively. The DC-RF efficiency of

the magnetron phase array is 61.0%.

The magnetron phased array system and the rectenna array

system were placed in the anechoic chamber, as shown in Fig.

15. The pitch angle of the rectenna array was set at −22.5°.

The horizontal distance between the center of the magnetron

phased array and rectenna array was 5 m. The anode current of

each PCM was set as 150 mA. We measured the transmitted

microwave power and rectified DC power, as shown in Table

III. Two operating magnetrons, i.e. PCM1 and PCM2

measured the result cases No.1~No.4. They represent the

circumstances of the filament ON state, filament OFF state,

injection-locking state, and phase-locked state, respectively.

TABLE III

EXPERIMENTAL RESULTS OF WIRELESS POWER TRANSFER

No PCM1 PCM2 PCM3 PCM4 DC

Power η RF-RF-

DC

1 365 339 0 0 60 8.52%

2 340 346 0 0 65 9.48%

3 349 339 0 0 68 9.69% 4 345 333 0 0 75 10.85%

5 0 0 316 319 76 11.97%

6 339 336 319 310 142 10.89%

7 305 306 0 0 43 7.04% 8 0 306 0 330 40 6.29%

9 0 0 295 330 47 7.52%

10 305 0 295 0 43 7.17%

TABLE IV

EXPERIMENTAL RESULTS OF DIFFERENT DISTANCES

Distance 2.7 m 3.0 m 3.5 m 3.8 m 4.3 m 4.7 m 5.2 m

DC Power 67 65 70 72 70 63 54

η RF-RF-DC 7.77% 7.54% 8.12% 8.35% 8.12% 7.31% 6.26%

We found that the efficiency ηRF-RF-DC (transmission

efficiency × microwave rectifier efficiency) increased when

the magnetron became stable. It suggests that the controllable

beam scanning will contribute to a higher transmission

efficiency. Case No.5 shows that the highest transfer

efficiency was observed when PCM3 and PCM4 were on. Fig.

16 shows the power density distribution of case No.5. The

distributed power densities of rectenna array were detected by

24 power sensors. An infrared image of the rectenna array

system in Fig. 16 was captured by a thermographic camera,

which can also reveal the received power distribution. The

highest received power is found in case No.6, when the output

power of the magnetron phased array was 1304 W. When the

pitch angle of the rectenna array is −12.5°, case No.7~No.10

illustrate the conditions that the adjacent two PCMs in

horizontal or vertical directions worked. The values of their

efficiency are near the same level. These values, however, are

lower than in case No.5 since they were measured in a

different transmit angle. Here, the result simulated the effect

of different incident angles on the transmission efficiency in

the WPT systems.

We then changed the transmission distance between the

center of the magnetron phased array and the rectenna array.

The anode current of each PCM was set as 150 mA. The

magnetron phased array output 862 W microwave and the

radiated beam remains constant. Table IV shows the measured

data when the distance varies from 2.7 m to 5 m. When the

distance is too close, the received power is focused on a few

sub-arrays where the rectified power is limited even if the

received microwave is continuously increasing. Besides, the

imperfection of the central area of the rectenna array will also

influence the transfer efficiency, otherwise, some parts of the

received microwave energy will distribute outside the rectenna

array panel when the transmission distance is relatively long.

We further analyzed the relationship between transmission

efficiency and transmission distance in the experiments. The

transmission distance of the experiment was less than 5.5m,

which is within the near-field region. Therefore, the

transmission efficiency can be expressed as the following (2)

[15],

�� 1 � ��

� �� /��

� � � ��/4�

Monitor PC

Measuring instrument

Optical fiber

Fig. 15. Full view of 5.8 GHz WPT experimental system.

Fig. 16. Power density distribution of the case No.5.

101.5 74.7 91.7 75.3

116.4 118.3 153.5 127.2

66.3

107.0 26.1 0.0

0.0

24.2

29.3 115.6

0.0 0.0

0.0

0.0 0.0

30.5

45.0

23.6

Power density

(W/m2)

Fig. 3.15 Full view of 5.8 GHz WPT experimental system.

We found that the efficiency [𝑅𝐹−𝑅𝐹−𝐷𝐶 (transmission efficiency × microwave rectifierefficiency) increased when the magnetron became stable. It suggests that the controllable


beam scanning will contribute to a higher transmission efficiency. Case No.5 shows that thehighest transfer efficiency was observed when PCM3 and PCM4 were on. Fig. 3.16 showsthe power density distribution of case No.5. The distributed power densities of rectennaarray were detected by 24 power sensors. An infrared image of the rectenna array systemin Fig. 3.16 was captured by a thermographic camera, which can also reveal the receivedpower distribution. The highest received power is found in case No.6, when the output powerof the magnetron phased array was 1304 W. When the pitch angle of the rectenna array is–12.5◦, case No.7 No.10 illustrate the conditions that the adjacent two PCMs in horizontalor vertical directions worked. The values of their efficiency are near the same level. Thesevalues, however, are lower than in case No.5 since they were measured in a different transmitangle. Here, the result simulated the effect of different incident angles on the transmissionefficiency in the WPT systems.

Feed waveguide(mm) Radiation waveguide(mm)No. length offset No. length offset No. length offset1 24.3 4.2 1 23.2 3.9 9 24.8 8.22 23.3 4.8 2 23.6 3.6 10 24.7 7.93 24.8 5.9 3 23.6 3.7 11 24.6 7.64 22.8 6.7 4 23.8 4.5 12 24.6 7.45 26.5 7.2 5 24.0 5.2 13 24.8 8.66 26.4 6.5 6 24.3 6.0 14 25.0 10.47 25.9 4.7 7 24.5 7.08 25.3 4.4 8 24.7 8.0

No PCM1 PCM2 PCM3 PCM4 DC Power η RF-DC1 365 339 0 0 60 8.52%2 340 346 0 0 65 9.48%3 349 339 0 0 68 9.69%4 345 333 0 0 75 10.85%5 0 0 316 319 76 11.97%6 339 336 319 310 142 10.89%7 305 306 0 0 43 7.04%8 0 306 0 330 40 6.29%9 0 0 295 330 47 7.52%

10 305 0 295 0 43 7.17%

101.574.7

91.775.3

116.4118.3153.5

127.266.3

107.0 26.10.0

0.0

24.2

29.3 115.6

0.00.0

0.0

0.00.0

30.5

45.0

23.6

Fig. 3.16 Power density distribution of the case No. 5.

We then changed the transmission distance between the center of the magnetron phasedarray and the rectenna array. The anode current of each PCM was set as 150 mA. Themagnetron phased array output 862 W microwave and the radiated beam remains constant.Table 3.4 shows the measured data when the distance varies from 2.7 m to 5 m. When thedistance is too close, the received power is focused on a few sub-arrays where the rectifiedpower is limited even if the received microwave is continuously increasing. Besides, theimperfection of the central area of the rectenna array will also influence the transfer efficiency,


otherwise, some parts of the received microwave energy will distribute outside the rectennaarray panel when the transmission distance is relatively long.

We further analyzed the relationship between transmission efficiency and transmissiondistance in the experiments. The transmission distance of the experiment was less than 5.5 m,which is within the near-field region. Therefore, the transmission efficiency can be expressedas the following equation (3.2) [87],

[0 = 1− 𝑒−𝜏2

𝜏 =𝐷𝑡 ×𝐷𝑟_𝐷

(3.2)

𝐷𝑡 and 𝐷𝑟 are the antenna diameters of the transmitter and receiver, respectively. Here,𝐷𝑡 is the diagonal of the slot array antenna, which is nearly 1.14 m. 𝐷𝑟 is edge-to-edgedistance, which is nearly 2.70 m. _ is the wavelength. 𝐷 is the transmission distance. Thetransmission efficiency [0 of this ideal model with distance is shown in the blue line inFig. 3.17. The power density distribution received by the rectenna array system is a powerdensity distribution formed by the slot array antenna at a transmission distance 𝐷. Thesimulation data of beam pattern were used to analyze the power distribution. The powerdensity distribution of the main lobe is approximately expressed as [88][89]

60 CHAPTER 3. MAGNETRON PHASED ARRAYIEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES

�� /4 (2)

� and �� are the antenna gain of the transmitter and receiver,

respectively. � is the effective area of the transmitting

antenna area, which is nearly 0.18 m2. �� is the effective area

of the receiving antenna area, which is nearly 2.4 m2. D is the

transmission distance. The transmission efficiency �� of this

ideal model with distance is shown in the blue line in Fig. 17.

The power density distribution received by the rectenna array

system is a power density distribution formed by the slot array

antenna at a transmission distance . The simulation data of

beam pattern were used to analyze the power distribution. The

power density distribution of the main lobe is approximately

expressed as [16] [17]

�� 2�

�� 2��/��

� � tan � �3�

P, �, �, r0 are the transmission power, the radius of the 1/e2

value contour of the beam at transmission distance D, the

beam angle of 1/e2 value which is calculated to 4.98°, the

distance from the receiving antenna center, respectively. In the

distance-varied experiments, the center of microwave beam

was deviated from the null space of the rectenna array as

shown as Fig. 18. The power density distribution at the

rectenna array ��, � needs a coordinate transformation by the

following (4).

�� !� " �� 2!� cos

! � sin ' � ∆ℎ . (4)

where ! is the deviation distance from the center of the

rectenna array, ' is the main beam angle in vertical direction

which is 22.5°, ∆ℎ is the altitude difference of between the

antenna centers of Tx and Rx system which is 1.3 m. The

received power Pr of the rectenna array can be expressed as (5)

Pr=* * ��, �+,-

�

�.

�/�/ . (5)

/� is the equivalent diameter of the rectenna array system

which is 2.7 m. When setting l to 0, the receiving power is the

ideal model power �� . The loss of transmission efficiency

from the deviation of the rectenna array null center is written

as

01 � �� /�� . (6)

The D-01 curve is shown in the green line in Fig. 17. The

power loss of the center null space of the rectenna array can be

also expressed

Pn=* * ��, �2,-

�

�.

�/�/ . (7)

Here, 3� is the equivalent diameter of the null space , which is

0.49 m. The loss of transmission efficiency from the center

null space is expressed in the following equation and shown in

the orange line in Fig. 17.

04 � �4/�� . (8)

The black line in Fig. 17 shows the transmission efficiency of

the WPT experiments which is expressed as

� � �� 5 �1 � 01 � 04� � �� 4�/� . (9)

The antennas of rectenna array were circularly polarized.

However, the developed transmitting slot array antennas were

linearly polarized. Then, almost a half power would be wasted

at the receiving antennas. Additionally, containing the rectifier

loss (minimum 50%), the RF-RF-DC efficiency of the

experimental results is the brown line as shown in Fig. 17. The

black line in Fig. 17 also shows the maximum possible RF-RF-

DC efficiency by the secondary horizontal axis. In addition, the non-uniform received microwave intensity

fluctuates the rectifier efficiency [18] [19]. It proves that the

higher rectified efficiency needs as many rectennas working

optimally as possible. The phase adjustment, therefore, is

necessary for achieving a uniform or a specific microwave

energy distribution at the received array.

Table V shows a detailed comparison of the magnetron

microwave power transmitter and transfer parameters with

other works. Compared with the magnetron WPT transmitter

developed by Zhang et al. [11] and Chen et al. [12], it is the

first time that a beamforming transmitter is developed based

on power-variable and PCMs. Furthermore, the received DC

power is much higher than [11] reported, even if the operating

frequency is higher.

Fig. 17. Transmission efficiency and RF-RF-DC efficiency with

distances. (Black line shows the maximum possible efficiency)

0%

5%

10%

15%

20%

25%

0%

20%

40%

60%

80%

100%

2.5 3.5 4.5 5.5 6.5

RF

-RF

-DC

eff

icie

ncy

Tra

nsm

issi

on E

iffi

cien

cy

Distance [m]

Deviation loss

Experiment result

of η RF-RF-DC

Ideal Model

Loss of center space

Fig. 18. Schematic diagram of the deviated beam pattern.

φ

! ��

�

∆ℎ

Fig. 3.17 Transmission efficiency and RF-RF-DC efficiency with distances. (Black lineshows the maximum possible efficiency)

𝐼 (𝑟0) =2𝑃𝜋𝜔2 exp

(−2𝑟

2𝜔2

)𝜔 = 𝐷 tan𝜓 (3.3)

𝑃, 𝜔, 𝜓, 𝑟0 are the transmission power, the radius of the 1/𝑒2 value contour of the beamat transmission distance 𝐷, the beam angle of 1/𝑒2 value which is calculated to 4.98◦, thedistance from the receiving antenna center, respectively. In the distance-varied experiments,the center of microwave beam was deviated from the null space of the rectenna array asshown as Fig. 3.18. The power density distribution at the rectenna array 𝐼 (𝑟, \) needs acoordinate transformation by the following equation (3.4).

𝑟20 = 𝑙

2 + 𝑟2 −2𝑙𝑟 cos\

𝑙 = 𝐷 sin𝜑−Δℎ. (3.4)

where 𝑙 is the deviation distance from the center of the rectenna array, 𝜑 is the main beamangle in vertical direction which is 22.5◦, Δℎ is the altitude difference of between the antenna



�� /4 (2)

� and �� are the antenna gain of the transmitter and receiver,

respectively. � is the effective area of the transmitting

antenna area, which is nearly 0.18 m2. �� is the effective area

of the receiving antenna area, which is nearly 2.4 m2. D is the

transmission distance. The transmission efficiency �� of this

ideal model with distance is shown in the blue line in Fig. 17.

The power density distribution received by the rectenna array

system is a power density distribution formed by the slot array

antenna at a transmission distance . The simulation data of

beam pattern were used to analyze the power distribution. The

power density distribution of the main lobe is approximately

expressed as [16] [17]

�� 2�

�� 2��/��

� � tan � �3�

P, �, �, r0 are the transmission power, the radius of the 1/e2

value contour of the beam at transmission distance D, the

beam angle of 1/e2 value which is calculated to 4.98°, the

distance from the receiving antenna center, respectively. In the

distance-varied experiments, the center of microwave beam

was deviated from the null space of the rectenna array as

shown as Fig. 18. The power density distribution at the

rectenna array ��, � needs a coordinate transformation by the

following (4).

�� !� " �� 2!� cos

! � sin ' � ∆ℎ . (4)

where ! is the deviation distance from the center of the

rectenna array, ' is the main beam angle in vertical direction

which is 22.5°, ∆ℎ is the altitude difference of between the

antenna centers of Tx and Rx system which is 1.3 m. The

received power Pr of the rectenna array can be expressed as (5)

Pr=* * ��, �+,-

�

�.

�/�/ . (5)

/� is the equivalent diameter of the rectenna array system

which is 2.7 m. When setting l to 0, the receiving power is the

ideal model power �� . The loss of transmission efficiency

from the deviation of the rectenna array null center is written

as

01 � �� /�� . (6)

The D-01 curve is shown in the green line in Fig. 17. The

power loss of the center null space of the rectenna array can be

also expressed

Pn=* * ��, �2,-

�

�.

�/�/ . (7)

Here, 3� is the equivalent diameter of the null space , which is

0.49 m. The loss of transmission efficiency from the center

null space is expressed in the following equation and shown in

the orange line in Fig. 17.

04 � �4/�� . (8)

The black line in Fig. 17 shows the transmission efficiency of

the WPT experiments which is expressed as

� � �� 5 �1 � 01 � 04� � �� 4�/� . (9)

The antennas of rectenna array were circularly polarized.

However, the developed transmitting slot array antennas were

linearly polarized. Then, almost a half power would be wasted

at the receiving antennas. Additionally, containing the rectifier

loss (minimum 50%), the RF-RF-DC efficiency of the

experimental results is the brown line as shown in Fig. 17. The

black line in Fig. 17 also shows the maximum possible RF-RF-

DC efficiency by the secondary horizontal axis. In addition, the non-uniform received microwave intensity

fluctuates the rectifier efficiency [18] [19]. It proves that the

higher rectified efficiency needs as many rectennas working

optimally as possible. The phase adjustment, therefore, is

necessary for achieving a uniform or a specific microwave

energy distribution at the received array.

Table V shows a detailed comparison of the magnetron

microwave power transmitter and transfer parameters with

other works. Compared with the magnetron WPT transmitter

developed by Zhang et al. [11] and Chen et al. [12], it is the

first time that a beamforming transmitter is developed based

on power-variable and PCMs. Furthermore, the received DC

power is much higher than [11] reported, even if the operating

frequency is higher.

Fig. 17. Transmission efficiency and RF-RF-DC efficiency with

distances. (Black line shows the maximum possible efficiency)

0%

5%

10%

15%

20%

25%

0%

20%

40%

60%

80%

100%

2.5 3.5 4.5 5.5 6.5

RF

-RF

-DC

eff

icie

ncy

Tra

nsm

issi

on E

iffi

cien

cy

Distance [m]

Deviation loss

Experiment result

of η RF-RF-DC

Ideal Model

Loss of center space

Fig. 18. Schematic diagram of the deviated beam pattern.

φ

! ��

�

∆ℎ

Fig. 3.18 Transmission efficiency and RF-RF-DC efficiency with distances. (Schematicdiagram of the deviated beam pattern.)

centers of Tx and Rx system which is 1.3 m. The received power 𝑃𝑟 of the rectenna arraycan be expressed as equation (3.5)

𝑃𝑟 =

∫ 2𝜋

0

∫ 𝐷𝑟

20

𝐼 (𝑟, \) 𝑑𝑟𝑑\. (3.5)

𝐷𝑟 is the equivalent diameter of the rectenna array system which is 2.7 m. When setting𝑙 to 0, the receiving power is the ideal model power 𝑃0. The loss of transmission efficiencyfrom the deviation of the rectenna array null center is written as

𝐿𝑑 =(𝑃0 −𝑃𝑟)

𝑃0. (3.6)

The 𝐷 − 𝐿𝑑 curve is shown in the green line in Fig. 3.17. The power loss of the center nullspace of the rectenna array can be also expressed

𝑃𝑟 =

∫ 2𝜋

0

∫ 𝑛𝑟

20

𝐼 (𝑟, \) 𝑑𝑟𝑑\. (3.7)

Here, 𝑛𝑟 is the equivalent diameter of the null space , which is 0.49 m. The loss of transmissionefficiency from the center null space is expressed in the following equation and shown in theorange line in Fig. 3.17.

𝐿𝑛 =𝑃𝑛

𝑃0(3.8)


Table 3.5 COMPARISON OF OUR MAGNETRON POWER TRANSMITTER WITHTHOSE IN OTHER RESEARCHES

Antenna𝑓0(GHz) 𝑃𝑡 (W)

BeamForming

Distance(m)

DCPower (W)

[11] 2×2 horn array 2.45 N/A N/A 5.5 67.3[12] 2×2 horn array 2.45 3552 N/A 48.8 N/AThis work 2×2 slot array 5.8 1304 ± 3° 5 142

The black line in Fig. 3.17 shows the transmission efficiency of the WPT experiments whichis expressed as

[ = [0 × (1− 𝐿𝑑 − 𝐿𝑛)

=(𝑃𝑟 −𝑃𝑛)

𝑃(3.9)

The antennas of rectenna array were circularly polarized. However, the developed transmit-ting slot array antennas were linearly polarized. Then, almost a half power would be wastedat the receiving antennas. Additionally, containing the rectifier loss (minimum 50%), theRF-RF-DC efficiency of the experimental results is the brown line as shown in Fig. 3.17.The black line in Fig. 3.17 also shows the maximum possible RF-RF-DC efficiency by thesecondary horizontal axis.

In addition, the non-uniform received microwave intensity fluctuates the rectifier effi-ciency [90][91]. It proves that the higher rectified efficiency needs as many rectennas workingoptimally as possible. The phase adjustment, therefore, is necessary for achieving a uniformor a specific microwave energy distribution at the received array.

Table 3.5 shows a detailed comparison of the magnetron microwave power transmitterand transfer parameters with other works. Compared with the magnetron WPT transmitterdeveloped by Zhang et al. [83] and Chen et al. [84], it is the first time that a beamformingtransmitter is developed based on power-variable and PCMs. Furthermore, the received DCpower is much higher than [83] reported, even if the operating frequency is higher.

3.5 Summary

We demonstrated a phased array system using four PCMs. By the designed slot antenna, themicrowave beam of the magnetron phased array was controlled within ±3◦ in the horizontalaxis. It is verified that the output phase and power of the magnetron phased array are

3.5. SUMMARY 63

adjustable. The measured maximum output of the magnetron phased array was 1870 W with61.0% DC-RF efficiency. In the WPT experiments, the magnetron phased array output 1340W and received 142 W DC power at a transmission distance of 5 m. We investigated theeffect on power transmission efficiency under the condition of multiple output combinationsand different distances. The magnetron phased array provides a scheme of low-cost, high-efficiency WPT. Furthermore, the microwave power transmission technology should beapplied to other WPT systems.

Chapter 4

Modulation Performance withInjection-Locking Magnetrons

4.1 Introduction

Continues wave (CW) magnetrons are extensively used in heating applications, e.g., mi-crowave oven. The advantages of magnetrons include low-cost, high-efficiency, and high-level output. However, when magnetrons are used as transmitters, they exhibit severalshortcomings such as an unstable output frequency and high levels of phase noise [92].Several studies have investigated magnetron noise. For example, a study conducted by ourresearch group examined the magnetron noise and developed a phase-controlled magnetron(PCM) [93]. Further, we developed a kilowatt (kW)-class high-power phased array systemfor wireless power transfer (WPT) that can be referred to as space power radio transmissionsystem (SPORTS) and a power-variable phase-controlled magnetron (PVPCM) for the WPTsystems [94]–Ref45. Some other researchers established a phase-locking 15-kW magnetronfor coherent power combining [97]. Tahir 𝑒𝑡 𝑎𝑙. have achieved frequency and phase modu-lations using an injection-locked 2.45 GHz magnetron as a transmitter for communicationin which the transmission of phase-shift keying (PSK) data was achieved at 2 Mbps [98].In their study, a fast pin diode switch was used to drive two RF sources in the modulationsystem, which discontinued the phase of the frequency-shift keying (FSK) signal that causedunnecessary spectrum radiation.

In addition, several studies have explored 2.45 GHz injection-locked magnetrons usingthe phase-locking method by controlling the anode current at low noise levels [92]–[98]. Inthis method, the output power of the magnetron is observed to change with the anode current,which requires a linear relation between the anode current and the oscillation frequency

66CHAPTER 4. MODULATION PERFORMANCE WITH INJECTION-LOCKING MAGNETRONS

of the magnetron. However, a 5.8 GHz magnetron exhibits different frequency-currentcharacteristics compared to a 2.45 GHz magnetron. Therefore, our group has addressedthis problem in a previous study by developing a 5.8 GHz PCM by controlling the phaseof the injection signal to lock the magnetron phase [99]. Our study was necessary to verifythe modulation system based on the performances of different magnetrons. Another studydeveloped a 1 kW class microwave band solid-state amplifier with an 8-way combiner toreplace the vacuum-tube devices [100]. Similarly, N. Hasegawa 𝑒𝑡 𝑎𝑙. have developed a170 W high efficiency amplifier [101]. Recently, extensive research has been conducted onhigh-power amplifiers in the microwave band [102]–[104]. In these studies, the amplifierswere combined by N-way combiners to improve the output, which may increase the cost andcomplicate the manufacturing process.

In this study, a high versatility injection-locked magnetron system, which can be appliedto the 2.45 GHz as well as the 5.8 GHz magnetrons and act as a communication transmitterfor amplitude-shift keying (ASK), PSK and FSK, was developed. Additionally, a modulatingsignal was injected into the magnetron to amplify it. This method was also used to verifythe transmission of the ASK data at 200 kbps as well as the PSK and FSK data at 10 Mbps.Moreover, the audio and video data were modulated on a magnetron output and were furtherdemodulated to evaluate the modulation performance of the injection-locked magnetron.

4.2 Magnetron Characteristics

The injection locking method of magnetrons is injecting a signal into the magnetron to lockthe oscillation frequency. The value of the injection signal frequency is set to be close tothat of the self-oscillation frequency of the magnetron. The locked frequency range, Δ 𝑓 ,can be expressed by Alder equation [105] as follows: Δ 𝑓 = 2 𝑓

√(𝑃𝑖/𝑃𝑜)/𝑄𝑒, where 𝑓 is the

injection signal frequency, 𝑃𝑖 is the injection signal power, 𝑃𝑜 is the magnetron output powerand 𝑄𝑒 is the external Q-factor of the magnetron.

In the locked frequency range, the frequency and phase of the magnetron output arelocked with the injection signal. Therefore, controlling the parameter of the injection signalresults in frequency and phase synchronization of the magnetron output. Additionally, theinjection of a modulating signal allows the magnetron output to follow the injection signaland to amplify the modulating signal.

This study intended to develop a highly versatile modulation system, which can beapplied to the 2.45 GHz and 5.8 GHz band magnetrons. However, the study mainly evaluatedthe modulation performance of the 5.8 GHz band magnetron because it has a differentperformance and exhibits a higher noise level than the 2.45 GHz band magnetron [99].

4.2. MAGNETRON CHARACTERISTICS 67

Fig. 4.1 illustrates an experimental system of the injection-locked magnetron modulation.The system comprises a signal generator (signal generator 1, Agilent N5183A), whichinjects a signal into the magnetron via an amplifier and a circulator. The injection signalfrequency exhibits the same value as compared to the value of the self-oscillation frequencyof the magnetron. During the experiments, the injection signal was sufficiently amplifiedto lock the magnetron frequency, and the magnetron output was measured via a directionalcoupler, whose output was connected to a dummy load. The experimental conditions of themagnetrons are presented in Table 4.1 . The locked frequency range, Δ 𝑓 , was measuredwith a spectrum analyzer (Agilent N9010A) using a directional coupler signal (results areillustrated in Fig. 4.2). According to the Alder equation, a large Δ 𝑓 value can be obtainedfor both the 2.45 GHz and 5.8 GHz band magnetrons by increasing the injection power.The phase noise of the 5.8 GHz magnetron with the different injection power are shown inFig. 4.3.

IF

Data Signal Signal generator1

Dummy load

Amplifier

Circulator

MagnetronHigh-Voltage

Power Supply

Demodulation Signal

RFLO

Signal generator2

LPF

Spectrum analyzer

FM demodulator

Directional

coupler

Oscilloscope

Power meter

Fig. 4.1 Experimental system of the injection-locked magnetron.


0

10

20

30

40

50

5.74 5.75 5.76 5.77 5.78

0

10

20

30

40

50

2.425 2.435 2.445 2.455 2.465

Inje

ctio

n p

ow

er [

W]

Frequency [GHz]

Δf Δf

Fig. 4.2 Characteristics of the lockedfrequency range of the 2.45 GHz and 5.8 GHz bandmagnetrons versus the injection power.

-160

-140

-120

-100

-80

-60

-40

-20

0

100 1000 10000 100000 1000000 10000000 100000000

Ph

ase

No

ise[

dB

c/H

z]

Frequency[Hz]

SG SG+AMP 注入なし 2W3W 4W 5W 8W10W 15W 20W 25W30W 33W 35W 40W

Free running

Fig. 4.3 Phase noise of the 5.8 GHz Magnetron with the different injection power.

4.3. MODULATION PERFORMANCE EVALUATION EXPERIMENTS 69

Table 4.1 PARAMETERS AND CONDITIONS OF THE INJECTION-LOCKED MAG-NETRON EXPERIMENTS

Band 5.8 GHz 2.45 GHzMaker Panasonic PanasonicType M5801J 2M236M42Anode Current 250 mA 140 mAAnode Voltage 4.48 kV (DC) 3.65 kV (DC)HVPS mode CC CCFilament Voltage 3.35 V (AC) OFFFilament Current 7.4 A OFFOutput Frequency 5.774 GHz 2.445 GHzOutput Power 614 W 309 W

4.3 Modulation Performance Evaluation Experiments

4.3.1 Injection-Locking Magnetron for ASK

The output power of a magnetron can be changed using four methods that involve controllingone of the following parameters: 1) anode current [99], 2) magnetic field [96], [106], [107],3) filament power and 4) injection power. Therefore, adjusting one of these parameters canmodulate an ASK signal.

In the first method, a high-voltage power supply (Glassman PS/LT005R360-20) wasused to control the anode current of the magnetron. We measured a 5.8 GHz magnetroncharacteristic with anode current is shown in Fig. 4.4. Fig. 4.4 shows the frequency shiftingspan is less than 5 MHz from 100 mA to 250 mA anode current. Base on Adler equationwhich was described in Sect.2, the Δ 𝑓 large than 5 MHz, the output frequency of magnetroncan be locked. From Fig. 4.4 the output power was relatively with anode current, thusmodulate the anode current can modulate the amplitude of magnetron output. It was observedthat the rise time of the high-voltage power supply was approximately 3 ms. However, apulsed-driven power supply can improve the rise time to 0.1 ms [108]. In this case, the risetime of the power supply limits the data rate to be achieved at a fast speed.

In the method 2), a coil outside the magnetron vacuum tube changed the magnetic fieldof the magnetron. Therefore, the coil current flow the magnetron output power and thecharacteristic of them was shown in Fig. 4.5. The photograph of the 2.45 GHz magnetronwith a coil is shown in Fig. 4.6. The magnetron output power was shift 20W (-12.3dB,5.88%) by the magnetic coil current. This power shift level was similar as the injection powerway.


315

320

325

330

335

340

345

0 0.2 0.4 0.6 0.8 1 1.2

Outp

ut

Pow

er [

W]

Magnetic coil current [A]

450

460

470

480

490

500

510

520

530

540

295

300

305

310

315

320

325

0 10 20 30 40

Outp

ut

pow

er [

W]

Outp

ut

pow

er [

W]

Filament Power[W]

5.8G

2.4G

5.75

5.755

5.76

5.765

5.77

5.775

5.78

0

100

200

300

400

500

600

700

50 100 150 200 250

Fre

quen

cy[G

Hz]

Outp

ut

Pow

er[W

]

Anode current[mA]

280

290

300

310

320

330

340

500

510

520

530

540

550

0 10 20 30 40 50 60

Outp

ut

Pow

er [

W]

Outp

ut

Pow

er [

W]

Injection power [W]

5.8G

2.45G

Fig. 4.4 Change in the output power of the 5.8 GHz band magnetron with the anode current.

In the method 3), changing the filament power of the magnetron, the magnetron outputpower also was been changed. Here, we measured a different characteristic between 2.45GHz band and 5.8 GHz band magnetron as shown in Fig. 4.7. The 2.45 GHz band and 5.8GHz band magnetron output power were shifted 25W (-11dB, 7.85%) and 60W (-9.46dB,11.3%), respectively. The filament power setting of the 2.45 GHz magnetron is differentfrom the 5.8 GHz because turn-off of the filament current makes the 2.45 GHz magnetronspectrum quiet and pure [8]. However, turn-off of the filament current of 5.8 GHz magnetronwill reduce the output power and the signal-noise ratio. The reason for this different of themis that after the filament particles of the 5.8 GHz magnetron are increased, the reactive currentis increased, and in the constant current (CC) mode of the high-voltage power supply, theoutput power becomes smaller and the efficiency is reduced when the current is constant.

In methods 2) and 3), controlling the parameters requires a drive power that is larger than10 W; however, a digital data signal exhibits a low power level.

However, method 4) comprises plenty of ASK products, which exhibit a fast responsetime and a low date signal power level. Fig. 4.8 depicts the characteristic change in the outputpower of the injection-locked magnetrons with the injection power. The maximum shiftsin the output power of the 2.45 GHz band and the 5.8 GHz band magnetrons were 40 W(–9.16 dB, 12.1%) and 30 W (–12.5 dB, 5.56%), respectively. Therefore, an injection-lockedmagnetron ASK system was developed by adjusting the injection power level.

A 5.8-GHz injection-locked magnetron modulation system was developed and assessed.As illustrated in Fig. 4.1 , the carrier signal was set to be a 5.774-GHz sine signal that wasgenerated from signal generator 1. A 100 kHz (200 bps, 1 Hz = 2 symbol rate) square wavewas provided as input to signal generator 1 as the external data signal. Signal generator 1


315

320

325

330

335

340

345

0 0.2 0.4 0.6 0.8 1 1.2

Outp

ut

Pow

er [

W]


450

460

470

480

490

500

510

520

530

540

295

300

305

310

315

320

325

0 10 20 30 40

Outp

ut

pow

er [

W]

Outp

ut

pow

er [

W]

Filament Power[W]

5.8G

2.4G

5.75

5.755

5.76

5.765

5.77

5.775

5.78

0

100

200

300

400

500

600

700

50 100 150 200 250

Fre

quen

cy[G

Hz]

Outp

ut

Pow

er[W

]

Anode current[mA]

280

290

300

310

320

330

340

500

510

520

530

540

550

0 10 20 30 40 50 60

Outp

ut

Pow

er [

W]

Outp

ut

Pow

er [

W]

Injection power [W]

5.8G

2.45G

Fig. 4.5 Characteristic of a 2.45GHz band magnetron output power with magnetic coilcurrent.

modulated the data signal on the carrier signal through ASK, and the amplitude modulationdepth was set to be 70%. Further, the modulated signal went through a 50-dB power amplifier(R&K A252HP-R). The power of the injection signal that was injected into the 5.8-GHzmagnetron via a circulator shifted from approximately 6.6 to 22 W. The magnetron outputpower was measured by a signal analyzer (Agilent N9010A VSA89601), and it was observedto shift by approximately 6%. Fig. 4.9 depicts the results of the demodulated signal. Thisinjection-locked magnetron output a lower amplitude depth than the injection signal, whichis different as compared to an amplifier because the magnetron output exhibits a fluctuatinggain with an increase in the injection power. The amplitude modulation depth of the injectionpower affects the magnetron output power shift level. The low part amplitude of the injectionpower should maintain the functionality of the magnetron in a frequency-locking state. Afast data rate requires a considerably extensive bandwidth; however, the bandwidth should benarrower than the locked frequency range, Δ 𝑓 , as depicted in Fig. 4.2 . In this experiment(𝑃𝑜/𝑃𝑖 = 13.5 dB), it was observed that the amplitude of magnetron output was shifted in the60 Hz noise delivered by the AC power filament. The filament of the magnetron was turnedoff after the magnetron began working to stop the 60 Hz noise. It should be noted that turningoff the filament current in case of a 5.8 GHz magnetron reduces the output power and thesignal-noise ratio, while it improves the signal-noise ratio in case of a 2.45 GHz magnetron.


Fig. 4.6 Photograph of a 2.45 GHz band magnetron with coil.

-22

-21

-20

-19

-7

-5

-3

-1

1

3

0 20 40 60 80 100

Dem

odula

tion s

ignal

[dB

m]

Dat

a si

gnal

[V

]

Time [μs]

Fig. 4.9 Result of the data signal and demodulation signal using a 5.8-GHz band injection-locked magnetron ASK system.

4.3.2 Injection-Locking Magnetron for PSK

Based on the injection method described in section 4.2, the magnetron output can followthe injection signal. This experiment was conducted by following a similar procedure tothat followed the case of the ASK experiment. In Fig. 4.1 , signal generator 1 generated


315

320

325

330

335

340

345

0 0.2 0.4 0.6 0.8 1 1.2

Outp

ut

Pow

er [

W]


450

460

470

480

490

500

510

520

530

540

295

300

305

310

315

320

325

0 10 20 30 40

Outp

ut

pow

er [

W]

Outp

ut

pow

er [

W]

Filament Power[W]

5.8G

2.4G

5.75

5.755

5.76

5.765

5.77

5.775

5.78

0

100

200

300

400

500

600

700

50 100 150 200 250

Fre

quen

cy[G

Hｚ

]

Outp

ut

Pow

er[W

]

Anode current[mA]

480

485

490

495

500

505

510

515

280

290

300

310

320

330

340

0 10 20 30 40 50

Outp

ut

Pow

er [

W]

Outp

ut

Pow

er [

W]

Injection power [W]

2.45G

5.8G

5.8GHz

2.45GHz

5.8GHz

2.45GHz

Fig. 4.7 Change in the output power of the magnetrons with the filament power.

315

320

325

330

335

340

345

0 0.2 0.4 0.6 0.8 1 1.2

Outp

ut

Pow

er [

W]


450

460

470

480

490

500

510

520

530

540

295

300

305

310

315

320

325

0 10 20 30 40

Outp

ut

pow

er [

W]

Outp

ut

pow

er [

W]

Filament Power[W]

5.8G

2.4G

5.75

5.755

5.76

5.765

5.77

5.775

5.78

0

100

200

300

400

500

600

700

50 100 150 200 250

Fre

quen

cy[G

Hｚ

]

Outp

ut

Pow

er[W

]

Anode current[mA]

480

485

490

495

500

505

510

515

280

290

300

310

320

330

340

0 10 20 30 40 50

Outp

ut

Pow

er [

W]

Outp

ut

Pow

er [

W]

Injection power [W]

2.45G

5.8G

5.8GHz

2.45GHz

5.8GHz

2.45GHz

Fig. 4.8 Change in the output power of the magnetrons with the injection power levels.

a 5.774 GHz sine wave as the carrier signal. A 5 MHz square wave was input into signalgenerator 1 as the data signal and was modulated on the carrier signal, which shifted 180◦

phase. Further, the modulating signal passed through a power amplifier (R&K A252HP-R)to increase the power to 14 W after which it was injected into the magnetron via a circulator.The magnetron output locked with the injection power to deliver a modulated signal to thedummy load. Next, signal generator 2 (Vaunix LSG-602) provided a 5.774 GHz signal asthe output (i.e., the same value as that of the carrier frequency). It was compared to themagnetron output signal by a double balanced mixer (R&K MX370), where the IF portof the mixer provided the demodulated signal through a low pass filter as the output. Anoscilloscope (Tektronix MSO2024) was used to measure the demodulated signal (Fig. 4.10


). The remaining experimental parameters are listed in Table 4.1 . The magnetron outputsignal was demodulated, and the data signal was achieved at 10 Mbps.

The rise time, 𝑡𝑟𝑖𝑠𝑒, of this injection-locked magnetron modulation system would limitthe maximum data transmission frequency, 𝑓𝑝, as according to the following relation:

𝑓𝑝 = 𝛼/𝑡𝑟𝑖𝑠𝑒 (4.1)

where 𝛼 is a constant (usually set to less than 0.35). The results of the experiment showed thatthe rise time, 𝑡𝑟𝑖𝑠𝑒, of this PSK system, which contained the phase shifter, the amplifier andthe magnetron, was 59 ns, the maximum data transmission frequency, 𝑓𝑝, was approximately5.9 MHz, 𝑃𝑜/𝑃𝑖 was 16.42 dB and the maximum transmission speed was 11.8 Mbps.

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

-40

-30

-20

-10

0

10

20

0 0.5 1 1.5 2D

emo

du

lati

on

sig

nal

[V

]

Mo

du

lati

on s

ignal

[V

]

Time[μs]

Fig. 4.10 Results of the PSK data signal and the demodulating signal.

The response time of the magnetron was expressed in another study as (𝑄𝑒/2𝜋 𝑓 )√𝑃𝑜/𝑃𝑖),

which was derived from the Alder equation [98], [100]. This expression depicts that a highinjection power causes the response time of the magnetron to decrease, which can increasethe transmission rate.

Further, a 60 Hz phase noise was observed at the magnetron output phase, which can becontrolled by the PCM system, as demonstrated in our previous study [99]. Furthermore, thesystem was operated under a condition that the magnetron should only work in the frequencylocked state. Therefore, if the magnetron is not in the locking state, the PSK system shouldadjust the injection power frequency so that it becomes close to the magnetron self-oscillationfrequency.

4.4. DISCUSSION OF THE MODULATION PERFORMANCE 75

4.3.3 Injection-Locking Magnetron for FSK

It is also possible to modulate a magnetron by FSK in a manner similar to the PSK modulation.FSK modulation offers several advantages such as strong noise immunity and low cost.Therefore, a similar experiment with identical experimental conditions (Table 4.1) wasconducted for FSK modulation. After the first step (input data) was performed, the carrierfrequencies were set to be 5.774 and 5.776 GHz (deviation frequency: 2 MHz). Moreover,the FSK signal was amplified to 14 W and injected via a circulator into the magnetron. Themagnetron was locked with the FSK signal, and the FSK signal was amplified and deliveredto the dummy load. Further, a frequency demodulator (Pakite PAT-630) demodulated themagnetron output signal. Finally, the oscilloscope measured the demodulated signal (Fig. 4.11). This injection-locked magnetron modulation system can amplify the FSK signal to 10 Mbps.The bandwidth of the injection-locked magnetron system for FSK should be smaller than thelocked frequency range, Δ 𝑓 . Additionally, a small value of modulation index (quotient of thedeviation frequency to the data rate) results in the production of less sidebands. Furthermore,multiple FSK modulations can improve the bandwidth efficiency. Therefore, over a constantlylocked frequency range or a constant bandwidth, a small value of the modulation index andmultiple FSK modulations can result in a fast transmission rate.

-2

-1

0

1

2

3

4

5

-40

-30

-20

-10

0

10

20

0 0.5 1 1.5 2

Dem

od

ula

tio

n s

ign

al [

V]

Mo

du

lati

on s

ignal

[V

]

Time[μs]

Fig. 4.11 Results of the FSK data signal and the demodulating signal.

4.4 Discussion of the Modulation Performance

The injection-locked magnetron acted as an amplifier for the modulation system. However,there are some differences between the injection-locked magnetron and the amplifier. Thegain of an amplifier constantly increases with the input power while the magnetron exhibits


a fluctuating gain, as depicted in Fig. 4.8 . Thus, the magnetron exhibits a lower gain witha considerably wide locked range. The injection power affects the frequency and phase ofthe magnetron when it works in an injection-locked state. Therefore, the magnetron outputphase and frequency could follow the injection signal.

Both periodic and aperiodic signals were detected because the data signals exhibiteda considerable difference between their magnetron modulation performances. Fig. 4.12depicts the spectral curve of the injection-locked magnetron for the FSK modulation systemoutput versus the periodic and aperiodic data signals. As illustrated, periodic signals do notaccurately reflect the locked frequency range or the data error level. In the experiment, apseudo noise (PN) sequence (PN 9: 29 −1 = 511 bit) was modulated on the carrier signalby a digital signal generator (Agilent E4432B), while a signal analyzer (Agilent N9010AVSA89601) was used to analyse the error vector magnitude (EVM) of the modulating signal.EVM compares the measured locations with the ideal locations. Therefore, it helps toevaluate the quality of the magnetron digital modulation. The remaining relevant operatingparameters are presented in Table 4.2.


-100

-80

-60

-40

-20

0

2437.64 2442.64 2447.64 2452.64 2457.64

Am

pli

tud

e[d

Bm

]

Frequency [MHz]

-80

-60

-40

-20

0

2437.64 2442.64 2447.64 2452.64 2457.64

Am

pli

tud

e[d

Bm

]

Frequency [MHz]

Fig. 4.12 Spectral curve of the injection-locked magnetron FSK system with different datasignals (Data signal on the upper trace: a square wave, lower trace: PN9 code).


Tabl

e4.

2M

OD

UL

AT

ION

·PE

RFO

RM

AN

CE

RE

SULT

SO

FT

HE

INJE

CT

ION

-LO

CK

ED

MA

GN

ET

RO

N

Inje

ctio

nsi

gnal

(Sig

nalg

ener

ator

&A

mpl

ifier

)M

agne

tron

outp

ut

Type

Dat

ara

teE

VM

Mag

Err

Phas

eE

rr(d

eg)

Freq

Err

(Hz)

EV

MM

agE

rrPh

ase

Err

(deg

)Fr

eqE

rr(H

z)A

SK20

0kb

ps1.

74%

1.48

%0.

5404

3489

.52.

29%

0.49

%1.

3074

3708

.7B

PSK

10M

bps

4.24

%4.

02%

0.77

6378

.137

8.85

%8.

33%

1.70

4833

97.1

QPS

K10

Mbp

s5.

29%

4.10

%1.

9337

3.60

689.

21%

7.17

%3.

2776

–18.

842

8PSK

5M

bps

5.62

%4.

06%

2.25

4241

.388

11.3

7%6.

36%

5.45

7778

.593

MSK

10M

bps

3.03

%0.

91%

3.65

253.

3269

6.56

%3.

02%

2.88

5830

.079

Type

Dat

ara

teFS

KE

rrM

agE

rrC

arrO

ffse

t(k

Hz)

Dev

iatio

n(H

z)FS

KE

rrM

agE

rrC

arrO

ffse

t(k

Hz)

Dev

iatio

n(H

z)2

FSK

10M

bps

2.61

%0.

83%

1.73

1397

5.96

6.38

%8.

12%

0.17

3092

8.88

4FS

K10

Mbp

s2.

22%

0.75

%–2

.411

697

2.16

5.13

%0.

64%

2.91

3892

9.53

8FS

K5

Mbp

s0.

78%

0.78

%0.

1751

799

3.53

1.45

%0.

62%

0.26

3698

1.90

16FS

K10

Mbp

s2.

09%

0.65

%0.

6610

197

6.22

4.20

%0.

61%

10.6

6210

25.4


Fig. 4.13 depicts that the 2.45 GHz band and 5.8 GHz band magnetrons exhibit similarEVM results at PSK modulation of 2 Mbps. A signal modulation can be achieved whenthe injection power is larger than 5 W. Therefore, the 5 W injection power level was set asthe condition for the magnetron to work under the injection-locked state at a transmissionbandwidth at 2 Mbps. The injection power is further increased under these conditions;however, the EVM value remains unchanged. Further, the transmission rate was changedfrom 0.5 to 2 Mbps, while the injection power was fixed at 5 W. It was observed that the2.45 GHz band and 5.8-GHz band magnetrons exhibited similar performances (Fig. 4.14). Consequently, it can be concluded that, under injection-locked conditions, adjusting thetransmission rate does not affect the transmission quality. Therefore, in a case where theinjection power is high enough to constantly lock the magnetron, the speed of modulationwill be limited by the response time of the injection-locked magnetron.

0%

10%

20%

30%

40%

50%

60%

0 2 4 6 8 10

EV

M

Injection Power [W]

2.455.8

0%

10%

20%

30%

40%

50%

0.5 0.75 1 1.25 1.5 1.75 2

EV

M

Transmission rate [Mbps]

2.45

5.8

Fig. 4.13 Changes in the error vector magnitude of the injection-locked magnetron systemfor PSK with the injection power levels.

80CHAPTER 4. MODULATION PERFORMANCE WITH INJECTION-LOCKING MAGNETRONS0%

10%

20%

30%

40%

50%

60%

0 2 4 6 8 10

EV

M

Injection Power [W]

2.455.8

0%

10%

20%

30%

40%

50%

0.5 0.75 1 1.25 1.5 1.75 2

EV

M

Transmission rate [Mbps]

2.45

5.8

Fig. 4.14 Changes in the error vector magnitude of the injection-locked magnetron systemfor PSK with the transmission rates.

Furthermore, while conducting the ASK, multi-value PSK and multi-value FSK modula-tions, a PN9 data signal was used to evaluate the quality of the modulating signal. In addition,a 2.45 GHz band magnetron was used to perform the measurements. To determine themodulation limit of the magnetron, the maximum gain under the injection-locked conditionswas set. The magnetron output power was 309 W (𝑃𝑜/𝑃𝑖: 13.43 dB) at an injection powerof 14 W. Further, the PN9 data signal of the modulating signal having a speed of 10 Mbpswas tested, as depicted in Table 4.2 (it was 5 Mbps for both 8 PSK and 8 FSK). It was alsoobserved that the magnetron increased the error rate in all the modulation methods. However,under identical conditions, the communication quality of the FSK modulation was observedto be better than that of the PSK modulation, which can be attributed to the higher anti-noiseability of the FSK modulation. In the vector signal analyzer (VSA 89601) software, the filterwas set as root raised cosine and the Alpha/BT value was 0.35. The EVM and FSK error wasmeasured when it changed less than ±1%; values of them lower than 10% can be consideredas valid data transmissions.

4.5 Demonstration Experiments

4.5.1 Demonstration of a Phase-Controlled Magnetron for Transmit-ting an Audio Signal

A PCM PM system (Fig. 4.15 ) was used to demonstrate an audio signal transfer. An audioof Hello was input into phase shifter 1 that demodulated it. The magnetron output phase was

4.5. DEMONSTRATION EXPERIMENTS 81

further compared with the phase of the modulating signal in the mixer. The phase differencesignal was input into the phase-locked loop (PLL) circuit to control phase shifter 2, whichchanged the phase of the signal injected into the magnetron. Using the PLL circuit, thephase difference was gradually converged to zero, and the magnetron output followed themodulating signal. The receiver antenna was compared with the signal generator whichworked at the same frequency as that of the transmitter, and the mixer outputted the phasedifference. Fig. 4.16 depicts a part of an audio signal from among which high-frequencynoise that is out of the human hearing range (i.e., 20 Hz to 20 k Hz) is observed to be presentin the demodulated signal. The rise time, 𝑡𝑟𝑖𝑠𝑒, of this PM system was observed to be 1.78`s. In addition, it was observed that this system can avoid the 60 Hz phase noise from thefilament or the environmental impact and can amplify the multi-value PSK signal. Therefore,the analogue PM system acted as a radio broadcast.

IF

Audio signal

Transmitter

Signal generator Phase shifter2

Antenna

Amplifier

Circu

lator

5.8GHz

Magnetron

High-Voltage

Power Supply

PLL control circuit

Double Balanced MixerLO RF

IFPhase shifter1

Antenna

Audio signal

Receiver

RFLO

Signal generator

LPF

Fig. 4.15 Schematic of the PM injection-locked magnetron system.


-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

-0.8

-0.3

0.2

0.7

1.2

0 5 10 15 20

Dem

od

ula

tio

n s

ign

al[V

]

Mo

du

lati

on s

ignal

[V]

Time[ms]

Fig. 4.16 Waveform of the PM injection-locked magnetron.

4.5.2 Demonstration of an Injection-Locking Magnetron for Transmit-ting a Video Signal

Similarly, the transmission and demodulation of a video signal were demonstrated by ananalogue frequency modulation (FM). In this case, an injection-locked magnetron FM systemwas designed (Figs. 4.17 and 4.18 ). The cores of the FM modulator and demodulator (PakitePAT630) are IC RTC6705 and RTC6709, respectively. The FM modulator generated a 5.8GHz band FM signal modulated with a video signal and two modulated audio sub-carriersat 6 and 6.5 MHz, respectively. A DVD player outputted the signal to the FM modulator,and the modulating signal was amplified to 10 W and injected into the magnetron. In thereceiver, the demodulator demodulated the video and audio signals and delivered them to theTV. Thus, the DVD signal was played on the TV. It should be noted that the long distance inthis system was simulated using a microwave absorber. Therefore, it can be concluded thatthe analogue FM system acted as a TV broadcast.


FM modulator

Antenna

Amplifier

Circulator

5.8 GHz

Magnetron

High-Voltage

Power Supply

DVD Player

Transmitter

FM demodulatorTV

Receiver

Antenna

Fig. 4.17 Schematic of the FM injection-locked magnetron system.

Fig. 4.18 Photo of the FM injection-locked magnetron system (1: high voltage power supply,2: DVD player, 3: FM modulator, 4: amplifier, 5: horn antenna, 6: circulator, 7: magnetron,8: TV).

4.5.3 Demonstration of an Injection-Locking Magnetron for transfer-ring data

With develop of industrial production, the factory automation has been continuously improved.Electric trolleys or transport robots that carry parts are widely used in the automated factory[109]. However, the electric trolleys need people to charge the battery. It’s proposed to usea wireless power transfer system to charge the power to the trolley. At the same time, the


system using the same microwave to control the trolley. Since the electric trolley needs alarge amount of power for charging, we suggest a magnetron to as the microwave transmitteras shown in Fig. 4.19.

Receiver

Transmitter

Fig. 4.19 A wireless power transfer system for electric trolley.

Regarding frequency and phase modulations by the injection-locked magnetron as atransmitter for communication, the transmission of phase-shift keying (PSK) data at 10 Mbpshas been achieved. In this experiments, we utilized a 5.8 GHz injection-locked magnetron fora wireless power and data transfer system whose antenna size can be smaller than 2.45 GHz.The magnetron outputs can follow the injection signal in the frequency locking range [105] .A data signal was modulated by frequency shift keying (FSK) on the injection signal, thenthis modulated signal was injected to the magnetron. The magnetron output the high-powerlevel microwaves carrying the data information, at the receiver, then we can rectify themicrowave power and demodulate this data signal. In this section, we present how to utilizethe injection-locked magnetron to transmit data.

Fig. 4.20 shows a block diagram of the injection-locked magnetron for the FSK modu-lation system. A LabVIEW program was designed and output the motor control signal tothe modulator (Pakite PAT-630 transmitter) through the RS-232 TxD port. The modulatormodulated the control signal on the microwave by FSK. This FSK modulated signal wasamplified to 10 W and via a circulator, injected to the magnetron (Panasonic M5802). Here,the modulation frequency was nearly at the magnetron oscillation frequency. Then, the


magnetron followed the modulated signal and amplified it. Then the high-power modu-lated microwaves were transmitted through the antenna. At the receiver, the transmittedmicrowaves were received and demodulated by a frequency demodulator (Pakite PAT-630receiver). Then a driver circuit was designed that transferred the demodulated data to controlthe motors.

PC

RS232

Signal generator

An

tenn

a

Amplifier

Circulator

5.8GHz

Magnetron

High-Voltage

Power Supply

An

tenn

a

Demodulator

TxD RxD

Fig. 4.20 Schematic of data transfer system by FSK modulation.

The parameters of the FSK modulation experiments are shown in Table 4.3. Fig. 4.21shows the transmitted the control data and demodulated data. The demodulation datacontrolled the motors successfully. Fig. 4.22 shows the photos of data transfer system.Through the driver circuit, the motors were controlled. Here, the magnetron that must beworked in the locking state can transmit the data. According to Alder’s equation, the lockingrange Δ 𝑓was related to the injection power level [105] . In this experiment, the locking rangeΔ 𝑓 was 5 MHz when the injection power was 10 W. The FSK bandwidth should be narrowerthan the locking range of the magnetron, otherwise the magnetron will lose its locking status.

Increasing the injection power, the transmission data rate can be faster. We also confirmedthis data transfer system provided much higher rate such as 115200 bps and worked well.We set the demodulated data as the RS-232 RxD data, and fed back to the PC to check thedata error. The transmitted data file was larger than 1 MB, and the received data had no biterror. We demonstrated a 5.8 GHz injection-locked magnetron could transmit the FSK data.


Table 4.3 PARAMETER OF THE FSK MODULATION EXPERIMENTS

Anode Current 250 mAAnode Voltage -4.48 kV(DC)Filament Current 7.4 AFilament Voltage 3.35 V(AC)Injected Power 10 WOutput Frequency 5.774 GHz-5.776 GHzOutput Power 655.3 WModulation FSK@9600 bps

0.8

1.3

1.8

2.3

2.8

-2.5

-1.5

-0.5

0.5

1.5

2.5

0 0.5 1 1.5 2R

S-2

32

Rx

D s

ign

al [

V]

RS

-23

2 T

xD

sig

nal

[V

]

Time [ms]

Fig. 4.21 Transmitted data vs. received data.

4.6 Summary

In this study, several modulation methods were assessed using injection-locked magnetrons,which behaved like a narrow bandwidth amplifier. The 5.8 GHz injection-locked magnetron(𝑃𝑜/𝑃𝑖: 13.5 dB) transmitted the modulating signal at a rate of 200 kbps in the ASKexperiment. The data transmission in the PSK and FSK(𝑃𝑜/𝑃𝑖: 16.42 dB) experimentsachieved a high speed of 10 Mbps. The 2.45-GHz injection-locked magnetron (𝑃𝑜/𝑃𝑖: 13.43dB) was assessed using ASK, multi-value PSK and multi-value FSK performances. TheEVM in these cases were observed to be less than 10% at a data rate of 10 Mbps.

When the magnetron worked in a frequency-locked state, the injection power level wasobserved to have no effect on the quality of the modulating signal. Increasing the injection

4.6. SUMMARY 87

1 3

4

5

2

6

Transmitter

Receiver6.1

6.2

Fig. 4.22 Photos of data transfer system (1: Modulator, 2: High-Voltage Power Supply andAmplifier, 3: Magnetron, 4: Circulator, 5: Transmitting antenna, 6: Electric trolley, 6.1:Demodulator, 6.2: Driver circuit).

power level was observed to produce an extensive frequency range, a high response time anda fast data transmission rate.

Additionally, as a demonstration of system viability, an audio signal was successfullytransmitted by the PCM PM system and demodulated; further, a video signal was alsotransmitted by the injection-locked magnetron FM system and demodulated.

We also demonstrated a 5.8 GHz injection-locked magnetron could transmit the motorcontrol data by FSK modulation. Through the transmitted data, we successfully controlledthe electric trolley. It’s expected for a wireless power and data transfer system for the electrictrolley.

Chapter 5

Power Supplies of the Magnetron

5.1 Introduction

Power supply is a key component of the magnetron system. The performance of the powersupply is related with the stability of the magnetron output. Our research group has developedthe phase-controlled magnetron as the transmission apparatus of a wireless power transfersystem [110]. In the other hand, another group has developed a phase-locking 15 kWmagnetron for coherent power combining [111]. Regarding frequency and phase modulationsby an injection-locked magnetron as a transmitter for communication, the transmission ofphase-shift-keying data at 2 Mbps has been achieved by Tahir et al. [112]. However, all thesepower supplies of the low noise magnetron systems are stabilized power supplies [110]-[112]which costs hundreds times than magnetron. It’s limited the magnetron to many practicalapplications, such as communication and wireless power transfer.

It was clarified that the noise factors of the magnetron are the stability of the power supplycircuit and the temperature of the electron emission filament inside the magnetron[113].The conventional phase-controlled magnetron use stabilized DC power sources to reducemagnetron noise by turning off filaments[114].

This chapter introduces two low cost power supplies which could be improved on voltageripple rate for magnetron systems.

5.2 Full Wave Doubler Rectifier

We utilized a full wave voltage doubler, which was improved from a power source ofmicrowave ovens and whose cost is almost nearly a magnetron.

90 CHAPTER 5. POWER SUPPLIES OF THE MAGNETRON

The magnetron used in this study is a Panasonic 2M236M42 with an oscillating voltageof 3.60 kV. In this chapter, we study a power supply circuit that can continuously oscillatethis magnetron. According to the measurement results of the study, the power supply circuitused in the microwave oven is a half-wave voltage doubler rectifier circuit (Fig. 5.1.(a)),and the output of the magnetron oscillates intermittently at a 60 Hz cycle. The oscillationfrequency changes greatly due to this intermittent oscillation, and the cause is that theoscillation frequency of the magnetron varies with the anode current. Since the magnetronhas a discontinuous output and the anode current of the magnetron changes drastically, theoscillation frequency also changes greatly. This is one of the reasons why the frequency ofthe microwave oven is noisy. In addition, if the filament is turned off after the magnetronoscillates in the half–wave voltage doubler rectifier for a microwave oven, the filament’selectron emission cannot be maintained due to the discontinuous output, which preventsoscillation.

5.2. FULL WAVE DOUBLER RECTIFIER 91

(a)

(b)

(c)

D

C

T

T

D2

D1

C1

D2

D1

T

L

C1

C2

C2

MGT

MGT

MGT

Fig. 5.1 Power supply of the magnetron(a.microwave oven power supply, b.half wave voltagedoubler with smoothing capacity, c.full wave voltage doubler)


Since the half-wave voltage doubler rectifier circuit is an economical circuit for themagnetron system shown in Fig. 5.1.(a), it is widely used in microwave ovens. Here, wethrough LTSpice performed a simulation with the parameters of the microwave oven used.The output voltage waveform is shown in Fig. 5.2.(a). Since the output voltage is unsmooth,the voltage waveform changes at a frequency of 60 Hz, and when the power supply outputvoltage falls below the magnetron oscillation voltage, the magnetron cannot oscillate. Evenif we adjusted the parameters of the power supply circuit in Fig. 5.1.(a), there was a timewhen the magnetron could not oscillate.

-6

-3

0

0 0.02 0.04 0.06 0.08 0.1

Time [s]

-6

-3

0

Vo

ltag

e [k

V]

-6

-3

0

(a)

(b)

(c)

Fig. 5.2 Simulation results of the power supply voltage(a.microwave oven power supply,b.half wave voltage doubler with smoothing capacity, c.full wave voltage doubler, dot line:magnetron oscillation voltage)

The output voltage must be maintained above the oscillation voltage so that the magnetroncan be output continuously. The circuit with the smoothing circuit added is shown in Fig.5.1.(b). As a result of the simulation obtained with LTSpice, the output voltage waveformis shown in Fig. 5.2.(b). If the capacitance of smoothing capacitor C2 in Fig. 5.1.(b) is


increased, the output voltage can be maintained above the oscillation voltage, and the requiredcapacity of the capacitor C2 can be calculated from Eq. (5.1).

𝐶 =Δ𝑄

Δ𝑉=𝐼Δ𝑡

Δ𝑉(5.1)

Δ𝑄 is the amount of electricity stored in capacitor C2, Δ𝑉 is the ripple voltage of thepower circuit output, 𝐼 is the current of the power circuit output, and Δ𝑡 is the discharge timeof capacitor C2. The power supply circuit is supplied from an AC voltage of 60 Hz, and thedischarge time is a half cycle, which is about 8 ms. If the ripple voltage ratio is set to 5%, thecapacitance of capacitor C2 needs to be about 4.5 `F when the magnetron current is 100 mA.

The smoothing capacitor C2 in Fig. 5.1.(b) is a high-voltage capacitor with a withstandvoltage of several thousand volts, The larger its capacity, the higher its cost and the larger itsvolume. To reduce the capacitance of the capacitor C2, it is necessary to reduce either thecurrent 𝐼 or the discharge time Δ𝑡 from Eq. (5.1) to obtain the same ripple rate. The formeris related to the output power of the magnetron and cannot be adjusted. The latter dependson the frequency of the power supply circuit, and by using the inverter power supply circuit,the discharge time Δ𝑡 can be largely saved, but the cost will be increased.

From the above discussion, we adopted the full-wave voltage doubler rectifier circuit anddesigned the circuit in Fig. 5.1.(c). The inductance L was set to 0 H, and the parameters ofother components were the same as in Fig. 5.1.(b). Fig. 5.2.(c) shows the output voltagewaveform of the LTSpice simulation. Comparing to Fig. 5.2.(b), the output voltage ripplewas smaller with the same capacitor capacity. Since this is full-wave rectification, the outputfrequency becomes doubles and the capacitor discharge time is shortened. In addition, aninductance is connected in series with the output to smooth the output current.

The power supply circuit of the consumer microwave oven shown in Fig. 5.3 wasimproved to a full-wave voltage doubler rectifier circuit shown in Fig. 5.4. Fig. 5.1.(c)shows the AC voltage (100 V, 60 Hz) supplies to the transformer, the D1, D2 (Hio CL01-12),C1 and C2 constitute the full wave voltage doubler rectifier circuit and L makes the outputsmoothly. Here, the C1 and C2 are 2.06 `F, the L is 15 H. then we measured the output ofthe magnetron. The continuous output of magnetron was realized by using the improvedpower supply circuit. That is to say, the output voltage of all cycles is above the oscillationvoltage. Even if the filament is turned off, the residual heat will continue to oscillate.


Fig. 5.3 Photo of the microwave power supply.

Fig. 5.4 Photo of the full wave doubler rectifier.

5.2.1 Injection-Locking Magnetron with Full Wave Doubler Rectifier

The magnetron with full wave doubler rectifier can be locked with an injection power. Afterthe oscillation, the filament is switched off and the oscillation frequency of the magnetronbecomes stable. As shown in Fig. 5.5, the output spectrum before and after the improve-ment of the power supply circuit of the microwave oven shows that the frequency noise is


considerably reduced even without the injection power. The black line in Fig. 5.5 shows thespectral of the injection-locking magnetron satisfies the condition of the injection power is5 W. All spectral were measured with the maximum value retained by spectrum analyzer(Agilent N9010A). As shown in Fig. 5.6, the full wave voltage doubler output from –3730V ~–3570 V which fulfills the magnetron oscillation voltage. The ripple rate of magnetronanode voltage was 4.16%, and the output power at that time fluctuates in a 60 Hz cycle.However, the long-term average output power is stable.

-80

-60

-40

-20

0

20

2.3 2.35 2.4 2.45 2.5 2.55 2.6

Amplitude[dBm]

Frequency[GHz]

Span:300MHz

-70

-60

-50

-40

-30

-20

-10

0

2.4 2.425 2.45 2.475 2.5

Amplitude[dBm]

Frequency[GHz]

Span：100MHz

Fig. 5.5 Spectral of the magnetron. (dot line:microwave oven, blue line: full wave voltagedoubler without injection signal, black line:full wave voltage doubler with injection signal)


-4

-3.6

-3.2

-2.8

-2.4

-2

0

50

100

150

200

250

300

350

400

450

0 0.02 0.04 0.06 0.08 0.1

An

od

e V

olt

age

[kV

]

Ou

tpu

tP

ow

er [

W]

Time[s]

Fig. 5.6 The anode voltage and anode current of the magnetron worked by full wave voltagedoubler

5.2.2 Modulation Magnetron with Full Wave Doubler Rectifier

We developed a frequency-shift keying (FSK) transmitting system by the improved injectionlocked magnetron full wave doubler rectifier.

Fig. 5.7 shows a block diagram of the FSK system by using an injection-locked magnetron.2 MHz (=2 Mbps) pulse signal was created as FSK data by a signal generator (AgilentN5183A). The modulation frequency was set at 2.448 GHz and 2.45 GHz, respectively. ThisFSK modulated signal was amplified to 10 W and via a circulator, injected to the magnetron(Panasonic M236-M42). The oven magnetron was locked with the FSK modulated signaland amplified it. Then the FSK modulated microwaves were transmitted through the antenna.The transmitted microwaves were received and demodulated by a frequency demodulator(Pakite PAT-260) in the receiver. Then, 2 Mbps FSK data were obtained and the demodulatedresult is shown in Fig. 5.8.


Figure 2. FSK system block diagram

Figure 3. FSK modulation/demodulation result

level was changed, and the voltage should be kept above

the magnetron oscillation voltage all the time. However,

in this FSK system the magnetron worked at a ripple rate

of 4.16% and the output power was changed from 300 W

to 400 W. The oscillation frequency shifting fo was 3MHz

(2.4495-2.4524 GHz), and their relationships can be

defined as follows:

fmax <( flock - fo)/a

( If flock<fo, the magnetron goes off the injection locking )

Here, a is the modulation index. To achieve a faster

transmission rate, the injection power should be improved,

which could also improve the locking frequency range

flock, or the ripple rate should be reduced, which could

reduce the oscillation frequency shifting fo as well.

Conclusions We demonstrated FSK data transmission by an injection-

locked magnetron with a full wave voltage doubler. At 10

W injection power and 4.16% power source ripple, 2

Mbps FSK data transmission was achieved. Increasing the

injection power and reducing the voltage ripple, the

transmission data rate can be faster.

Figure 4. Injection power vs. lock frequency range

Figure 5. Characteristics of the magnetron (upper trace:

oscillation frequency vs. output power, lower trance: anode voltage vs. output power based on the full wave voltage doubler)

Acknowledgments This work was supported by the collaborative research

program: Microwave Energy Transmission Laboratory

(METLAB), Research Institute for Sustainable

Humanosphere, Kyoto University.

References 1. R. Adler, “A Study of Locking Phenomena in

Oscillators,” Proceedings of I.R.E and Waves and

Electrons, vol. 34, pp. 351-357, 1946

2. N. Shinohara, H. Matsumoto, K. Hashimoto, “Phase-

controlled magnetron development for SPORTS:

Space power radio transmission system,” Radio

Science Bulletin, vol. 2004, No. 310, pp. 29-35, 2004

3. C. Liu, H. Huang, Z. Liu, F. Huo, and K. Huang,

“Experimental Study on Microwave Power

Combining Based on Injection-Locked 15-kW S-

Band Continuous-Wave Magnetrons,” IEEE

Transactions on Plasma Science, vol.44, No. 8,

pp.1291–1297, August 2016.

4. I. Tahir, A. Dexter and R. Carter, “Frequency and

Phase Modulation Performance of an Injection-

Locked CW Magnetron,” IEEE Transactions on

Electron Devices, vol. 53, No. 7, pp. 1721-1729,

2006

Signal Generator

Antenna

AmplifierC

irculato

r

2.45GHz

Magnetron

Full wave

voltage doubler

Information

Transmitter

AntennaInformation

Receiver

Demodulator

2.45GHz

-1

1

3

5

7

-6

-4

-2

0

2

0 0.5 1 1.5 2

Dem

od

ula

tion

sig

nal

[V

]

Mo

du

lati

on

sig

nal

[V

]

Time[μs]

0

1

2

3

4

5

2 3 4 5 6 7 8 9 10

Lo

ckin

gfr

equ

ency

ran

ge

[MH

z]

Injetion Power [W]

3.4

3.5

3.6

3.7

3.8

3.9

4

2.442

2.444

2.446

2.448

2.45

2.452

2.454

2.456

0 200 400 600

|An

od

e v

olt

age

| [

kV

]

Mag

net

ron o

scil

lati

on

freq

uen

cy [

GH

z]Output of the Magnetron [W]

Fig. 5.7 FSK system block diagram.









defined as follows:























Electrons, vol. 34, pp. 351-357, 1946










pp.1291–1297, August 2016.





2006

Signal Generator

Antenna

Amplifier

Circu

lator

2.45GHz

Magnetron

Full wave

voltage doubler

Information

Transmitter

AntennaInformation

Receiver

Demodulator

2.45GHz

-1

1

3

5

7

-6

-4

-2

0

2

0 0.5 1 1.5 2

Dem

od

ula

tion

sig

nal

[V

]

Mo

du

lati

on

sig

nal

[V

]

Time[μs]

0

1

2

3

4

5

2 3 4 5 6 7 8 9 10

Lo

ckin

gfr

equ

ency

ran

ge

[MH

z]

Injetion Power [W]

3.4

3.5

3.6

3.7

3.8

3.9

4

2.442

2.444

2.446

2.448

2.45

2.452

2.454

2.456

0 200 400 600

|An

od

e v

olt

age

| [

kV

]

Mag

net

ron o

scil

lati

on

freq

uen

cy [

GH

z]

Output of the Magnetron [W]

Fig. 5.8 FSK modulation/demodulation results.

The fastest transmission data rate 𝑓𝑚𝑎𝑥 was limited by the magnetron oscillation frequencyrange. The magnetron FSK system only worked in the frequency-locking state. Increasing theinjection power could improve the lock frequency range 𝑓𝑙𝑜𝑐𝑘 , as shown in Fig. 5.9. Here, the


power source ripple would affect the magnetron oscillation frequency. Fig. 5.10 shows thatthe oscillation frequency was changed by the magnetron output power, and the anode voltagecould change the output power. It is difficult to define the relationship between the ripple rateand the oscillation frequency range because the power level was changed, and the voltageshould be kept above the magnetron oscillation voltage all the time. However, in this FSKsystem the magnetron worked at a ripple rate of 4.16% and the output power was changedfrom 300 W to 400 W. The oscillation frequency shifting 𝑓𝑜 was 3 MHz (2.4495-2.4524GHz), and their relationships can be defined as follows:

𝑓𝑚𝑎𝑥 < ( 𝑓𝑙𝑜𝑐𝑘 − 𝑓𝑜)/𝑎 (5.2)

( If 𝑓𝑙𝑜𝑐𝑘 < 𝑓𝑜, the magnetron goes off the injection locking state.) Here, 𝑎 is themodulation index. Improving the injection power can achieve a faster transmission rate. Thelager injection power improves the locking frequency range 𝑓𝑙𝑜𝑐𝑘 as well as the lower ripplerate which could reduce the oscillation frequency shifting 𝑓𝑜 .









defined as follows:























Electrons, vol. 34, pp. 351-357, 1946










pp.1291–1297, August 2016.





2006

Signal Generator

Antenna

Amplifier

Circu

lator

2.45GHz

Magnetron

Full wave

voltage doubler

Information

Transmitter

AntennaInformation

Receiver

Demodulator

2.45GHz

-1

1

3

5

7

-6

-4

-2

0

2

0 0.5 1 1.5 2

Dem

od

ula

tio

nsi

gn

al [

V]

Mo

du

lati

on

sig

nal

[V

]

Time[μs]

0

1

2

3

4

5

2 3 4 5 6 7 8 9 10

Lo

ckin

gfr

equ

ency

ran

ge

[MH

z]

Injetion Power [W]

3.4

3.5

3.6

3.7

3.8

3.9

4

2.442

2.444

2.446

2.448

2.45

2.452

2.454

2.456

0 200 400 600

|An

od

e v

olt

age

| [

kV

]

Mag

net

ron o

scil

lati

on

freq

uen

cy [

GH

z]


Fig. 5.9 Injection power vs. lock frequency range.










defined as follows:























Electrons, vol. 34, pp. 351-357, 1946










pp.1291–1297, August 2016.





2006

Signal Generator

Antenna

Amplifier

Circu

lator

2.45GHz

Magnetron

Full wave

voltage doubler

Information

Transmitter

AntennaInformation

Receiver

Demodulator

2.45GHz

-1

1

3

5

7

-6

-4

-2

0

2

0 0.5 1 1.5 2

Dem

od

ula

tion

sig

nal

[V

]

Mo

du

lati

on

sig

nal

[V

]

Time[μs]

0

1

2

3

4

5

2 3 4 5 6 7 8 9 10

Lo

ckin

gfr

equ

ency

ran

ge

[MH

z]

Injetion Power [W]

3.4

3.5

3.6

3.7

3.8

3.9

4

2.442

2.444

2.446

2.448

2.45

2.452

2.454

2.456

0 200 400 600

|An

od

e v

olt

age

| [

kV

]

Mag

net

ron o

scil

lati

on

freq

uen

cy [

GH

z]


Fig. 5.10 Characteristics of the magnetron (upper trace: oscillation frequency vs. outputpower, lower trance: anode voltage vs. output power based on the full wave voltage doubler).

We demonstrated FSK data transmission by an injection-locked magnetron with a fullwave voltage doubler. At 10 W injection power and 4.16% power source ripple, 2 Mbps FSKdata transmission was achieved. The transmission data rate can be faster by increasing theinjection power or reducing the voltage ripple,

5.2.3 Phase-controlled Magnetron with Full Wave Doubler Rectifier

A phase-controlled magnetron system was constructed using the full-wave voltage doublerrectifier circuit in the previous section as the power supply circuit in Fig. 5.11. Table 1shows the detailed experimental parameters. As shown in Fig. 5.11, the signal generator(Agilent N5183A) outputs a 2.45 GHz sine wave, amplifies it to 5 W through the amplifier(R&K A250HP-R) and the phase shifter, passes it through a circulator, and injects it into themagnetron. A phase control experiment of a magnetron was carried out using the low-passfilter (LPF) and an integrator circuit as the PLL control circuit which is described in Chapter2.


��

��

��

��

��

��

�� !��

��

��

" #

��

��$��%��%��

#��&�'��

��

#��% 2.45GHz

Fig. 5.11 Block diagram of a phase-controlled magnetron with full wave double rectifier.

Table 5.1 PARAMETERS OF THE INJECTION-LOCKING MAGNETRON

𝑀𝑎𝑘𝑒𝑟 Panasonic𝑇𝑦𝑝𝑒 2M236-M42𝐴𝑛𝑜𝑑𝑒𝑐𝑢𝑟𝑟𝑒𝑛𝑡 140 mA𝐴𝑛𝑜𝑑𝑒𝑣𝑜𝑙𝑡𝑎𝑔𝑒 −3.68±4.16% kV(DC)𝐼𝑛 𝑗𝑒𝑐𝑡𝑖𝑜𝑛𝑝𝑜𝑤𝑒𝑟 5 W𝐹𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 2.45 GHz𝑂𝑢𝑡𝑝𝑢𝑡𝑝𝑜𝑤𝑒𝑟 309 W𝐹𝑖𝑙𝑎𝑚𝑒𝑛𝑡𝑝𝑜𝑤𝑒𝑟 OFF

The phase difference between the injection signal and the output of the magnetron can beobserved from the IF port of the mixer (R&K MX370). Fig. 5.12 shows the phase differencesignal before and after the operation of the PLL control circuit. During the operation ofthe PLL control circuit, the phase of the injection signal and the output of the magnetronare synchronized. Here, the phase difference of the phase-locked is not 0◦, but is lockedat any phase. Therefore, even if the phases are synchronized, the value of the phase variesdepending on the observation point. Fig. 5.13 and Fig. 5.14 show the voltage waveform theoutput waveform of the magnetron and the injection signal before and after the operation ofthe PLL control circuit through a high-speed oscilloscope (Agilent 86100A). According toFig. 5.14, the output phase of the magnetron is stable, but the amplitude varies. If the poweris changed in a short time (100 ms or less), this fluctuation can be reduced by lowering theripple rate of the power supply circuit. Also, we improved the C1 and C2 to 8 `F, the ripple


voltage ratio can be achieved to less than 1%. The phase control accuracy can be improvedto ±1%. Fig 5.15 shows the lower ripple full wave doubler rectifier.

-20

0

20

40

60

80

100

0 0.02 0.04 0.06 0.08 0.1

Ph

ase

dif

fere

nce

[°]

Times[s]

Fig. 5.12 Magnetron output phase. (upper: worked without PLL circuit, lower: worked withPLL)


Fig. 5.13 Voltage waveform of the injection signal and magnetron output without PLL. (greenline: injection signal, yellow line: magnetron)

Fig. 5.14 Voltage waveform of the injection signal and magnetron output with PLL at 1%voltage ripple. (green line: injection signal, yellow line: magnetron)

5.3. SWITCHED-MODE POWER SUPPLY 103

Fig. 5.15 Photo of the full wave doubler rectifier with 1% voltage ripple.

5.3 Switched-Mode Power Supply

There is a low cost switched-mode DC power supply (WepeX 1000B-TX, Megmeet) whichis worked in high voltage ripple. The circuit diagram of the power supply is shown in Fig.5.16. The switching frequency is 20 kHz. According equation (5.2), 25 pF capacity canachieved the 1%. However, there is a also 50Hz voltage ripple in this power. The C1 and C2


in Fig. 5.16 were paralleled the larger capacity 5000 mF and 4`F. The anode voltage ripplewas measured achieved at 1%. The photo of the improved power supply is shown in Fig.5.17.

The improved low cost power supply was successfully worked at the following magnetronvarious load reflection experiment and magnetrons output power combining experiment.

1

2

A B C

1

2

A B C

AC200V

Q2

Q1C4

C6

D2

D1 L

C1

C2

MGT

T

C3

GND

Fig. 5.16 Circuit of the low cost power supply.


Fig. 5.17 Photo of the improved power supply.

5.3.1 Magnetron Various Load Reflection Experiments

The magnetron frequency and phase are synchronized with the injection signal within acertain locking bandwidth. The magnetron performance must be estimated before usagein practical applications. We investigate the performance of an injection-locking 5.8 GHzcontinuous-wave magnetron with various load reflection levels. Experiments are performedwhile the load reflection is varied using an E–H tuner between a magnetron and a circulator. A


narrower locking bandwidth is observed under constant injection power with increasing loadreflection. The proper-mismatched system suppresses its sideband energy, thereby reducingphase noise. The experimental features qualitatively validate the theoretical analyses results.The investigation results also provide guidance for advanced applications in communicationand high-energy physics based on injection-locking magnetrons.

We developed a corresponding experimental system to explore the effects of the anodevoltage ripples on the injection-locked magnetron’s performance and verify our theoreticalanalysis. Fig. 5.21 and 5.22 shows the block diagram and photo of the system. Themagnetron (model: M5802-KRSC1) was manufactured by Panasonic Microwave Co. (Japan)with a 5.8 GHz continuous wave output. The filament current can be turned off after 5min of operation. A relatively sharp free-running spectrum was achieved. An oscilloscope(TDS-3054, Tektronix) was used to measure the anode voltage (high-voltage probe: P6015A,Tektronix) and current (AC/DC current probe: 1146A, Hewlett Packard). A reference signalwas generated using a signal generator (N5172B, Keysight) and amplified using a poweramplifier (CA5800BW50-4040R, R&K). The circulators provided a transmission path forthe injection of the amplified reference signal, thereby protecting the solid-state amplifier.Couplers were used to sample the signals and measure the power and spectrum using powermeters (A1914A, Agilent) and a signal analyzer (N9010A, Agilent), respectively. The outputmicrowave power was absorbed by a dummy load. Various load reflection levels wereintroduced through an E–H tuner (EMH-6H, Nihon Koshuha) placed between the magnetronand the circulator. The E–H tuner integrated the removable short-end slug at both E andH planes. Simultaneously, the load characteristic variation was measured using a vectornetwork analyzer (N9928A, Keysight). The effects of different load reflection levels on themagnetron’s injection locking performance were then investigated. The operating anodevoltage Va and the current Ia were 4.01 kV and 188 mA, respectively. The magnetronproduced a CW power of Pout = 370 W. Fig. 5.20 shows a comparison of the spectralvariations of the circumstances of free running, injection, and injection locking. The injectionlocking spectral shows the magnetron with the low cost noise was lower than –60 dB.


S.G.

Coupler #2

Magnetron

Dummy

Load #1

Dummy

Load #2

Coupler

#1

Power

Amplifier

Signal

Analyzer

E-H Tuner

Power

Meter

VNA

Coupler #3

Fig. 5.18 Block diagram of the experimental system.


(1)

(2) (2) (3)(4)

(4)

(2)(5)

(7) (7)

(6)(9)(10)

(2)

(8)

(8)

(5)

(15)

(14)

(11)

(12)

(13)

Fig. 5. Block diagram (a) and photograph (b) of the

experimental system. Components and devices: (1) magnetron;

(2) coupler; (3) E–H tuner; (4) circulator; (5) dummy load; (6)

power supply; (7) fan; (8) power sensor; (9) high-voltage probe;

(10) current probe; (11) signal generator; (12) power amplifier;

(13) power meter; (14) oscilloscope; and (15) signal analyzer.

Fig. 5.19 Photograph of the experimental system.Components and devices: (1) magnetron;(2) coupler; (3) E–H tuner; (4) circulator; (5) dummy load; (6) power supply; (7) fan; (8)power sensor; (9) high-voltage probe; (10) current probe; (11) signal generator; (12) poweramplifier; (13) power meter; (14) oscilloscope; and (15) signal analyzer.


5.786 5.788 5.790 5.792 5.794

-100

-80

-60

-40

-20

0

Inte

nsity (

dB

m)

Frequency (GHz)

Injection

Free-runnig

Injection-Locking

Fig. 7. Experimental spectra of the free-running magnetron,

injection signal, and injection-locking magnetron (both

resolution bandwidth (RBW) and video bandwidth (VBW) are 5

kHz).

Fig. 5.20 Experimental spectra of the free-running magnetron, injection signal, and injection-locking magnetron (both resolution bandwidth (RBW) and video bandwidth (VBW) are 5kHz).

5.3.2 Magnetrons Output Power Combining Experiments

The power capacity of a single oscillator or amplifier is notably limited and is insufficientto meet the increasing demand for microwave applications. Therefore, the use of power-combining techniques to produce a multifold microwave output has attracted tremendousresearch attention in recent years. To increase the energy-utilization rate of a microwave-power source, a dual-way 1-kW S-band magnetron microwave-power-combining systemwithout circulators or phase shifters connected to the magnetrons was proposed and validated.A magic-T waveguide was used to achieve power combination and to provide pathways forthe reference signal. This system utilizes the power-dividing characteristics of the magic Tto lock the magnetrons. Frequency tuning is applied to adjust the phase difference betweenthe two magnetrons’ signals to achieve a high combination efficiency. The behaviors ofthe frequency and the phases of the system were numerically calculated and analyzed. Themagnetron frequencies track the varied reference frequency; the variation of the phasedifference is restricted to 90◦ when the reference frequency is tuned within the overlappedlocking bandwidth between the two magnetrons. Experimental results indicate that the


microwave-power-combining efficiency of the proposed system can reach 94.5%, withattenuation of microwave energy caused only by the waveguides and the magic T.

We developed a corresponding experimental system based on WR430 waveguides ina chamber to explore the performance of the proposed power-combining system and toverify our theoretical analysis. Fig. 5.21 and Fig. 5.22 shows a block diagram and photoof the system. The magnetrons (model: 2M167B-M32) were manufactured by PanasonicMicrowave Co. (Japan) with a 2.45-GHz CW output. A commercial S-band magic T wasused (5G052, SPC Electronics Co.) to achieve the functions of both power combining andpower division. The characteristics of the magic T were measured using a vector-networkanalyzer (N9928A, Keysight). Moreover, a straight waveguide with a length of 3.75 cmwas connected to Port 2 of the magic T, as shown by the close-up view of Fig. 5.22. Anoscilloscope (TDS-3054, Tektronix) was used to visualize the anode voltage (high-voltageprobe: P6015A, Tektronix) and current (AC/DC current probe: TPC A300 module). Areference signal was produced using a signal generator (E4421B, Agilent) and amplifiedusing a power amplifier (CA2450BW100-4547M-C, R&K). The circulators were connectedto port 4 of the magic T to protect the solid-state amplifier and were therefore not connectedto the magnetrons. Couplers were used to sample the signals and to measure the power andspectrum using power meters (A1914A, Agilent and E4419B, Agilent) and a signal analyzer(N9010A, Agilent), respectively. The horn antennas were assumed to radiate the mismatchedpower and combined . Moreover, the two horn antennas were mutually orthogonal to preventmutual coupling. The radiated power was absorbed by the high-power ceramic-absorberwall. Simultaneously, the phase difference of the magnetrons output was measured using avector-network analyzer (N9928A, Keysight).

Two-way 1 kW S band magnetron power-combining system without any lossy andexpensive circulators or phase shifters connected to the magnetrons was built and measuredfor the first time. A magic-T waveguide was used for power combination of the twomagnetrons while simultaneously providing paths for the reference signals. The system’sbehaviors were numerically extrapolated for the first time. A stable combined-output powerand pure spectra during phase control were achieved via frequency tuning of the referencesignal. The proposed magnetron coherent-power-combining system exhibited a maximummicrowave-combining efficiency of 94.5% and a maximum phase-control scope of 73°. Thepower attenuation of each magnetron was minimized in the absence of connections betweencirculators and phase shifters.


Magnetron

#2

Magnetron

#1

Signal

Generator

Dummy

Load

Magic-TCoupler

#4

Coupler

#3

Coupler

#1

Coupler

#2

Power

Amplifiler

VNA

Power

Meter #1

Signal

Analyzer

Power

Meter #2

λg/4Horn

Antenna #1

Horn

Antenna

#2

Ceramic

Absorber

Circulator

#1

Circulator

#2

0 50 100 150 200 250 300

900

1050

1200

1350

Pow

er

(W)

Time (sec)

Without reference signal

Injection-locking condition

Fig. 5.21 Block diagram of the experimental system.


Fig. 4. Block diagram (a) and photograph (b) of the

experimental system. Components and devices: (1) magnetron;

(2) waveguide coupler; (3) coaxial coupler; (4) power sensor;

(5) magic T; (6) horn antenna; (7) wind-cooled dummy load; (8)

waveguide to coax adapter; (9) 3.75-cm straight waveguide;

(10) fan; (11) signal analyzer; (12) power amplifier; (13) VNA;

(14) signal generator; (15) dc power supply; (16) power

amplifier; (17) oscilloscope; (18) current-probe amplifier; (19)

current probe; (20) high-voltage probe; (21) power supply; and

(22) ceramic absorber wall.

Fig. 5.22 Photograph (b) of the experimental system. Components and devices: (1) mag-netron; (2) waveguide coupler; (3) coaxial coupler; (4) power sensor; (5) magic T; (6) hornantenna; (7) wind-cooled dummy load; (8) waveguide to coax adapter; (9) 3.75-cm straightwaveguide; (10) fan; (11) signal analyzer; (12) power amplifier; (13) VNA; (14) signalgenerator; (15) dc power supply; (16) power amplifier; (17) oscilloscope; (18) current-probeamplifier; (19) current probe; (20) high-voltage probe; (21) power supply; and (22) ceramicabsorber wall.

5.4 Summary

This chapter introduced two type improving methods of low-cost power supplies, the coreidea is to reduce the ripple voltage of the power supplies, thereby improving the stability of

5.4. SUMMARY 113

the magnetron output. The ripple of the full-wave rectifier power supply improved in themicrowave oven power supply reached 4.16%. Based on the injection-locking magnetron,the transmission rate of 2 Mbps was realized. The relationship between the ripple rate, theinjection power and communication bandwidth are also analyzed. After further increasingthe capacitance of the capacitor, the ripple is reduced to less than 1%, and the phase controlaccuracy of the phase-controlled magnetron is within ±1%. In the switched-mode powersupply, its ripple has also reached below 1%. In the magnetron load variation experiment, thefrequency spectrum of the injection-locking magnetron is stable, and the noise is suppressedto below -60 dB. In addition, the combined efficiency of the two magnetrons using this powersupplies reached 94.5%.

Chapter 6

Microwave Wireless Power TransmissionSystems

Chapter 3 described the performance of the magnetron phased array to verify the beamforming and WPT experiment. Chapter 4 described the communication verification of themodulation magnetron. This chapter introduces systems that transmit power and information,then shows how to build efficiently WPT systems.

There are several isolation ways to complete the information and power transmissionsystem. The isolation factors were increased as time, frequency, polarization and mode.The simplest system is the different transmission frequency for information and powertransmission. A very forward-looking method is to use the mode of the orbital angularmomentum (OAM) wave, which are considered to be the next generation of communicationtechnology revolution due to their high isolation in different modes. The mode defined as thephase difference times of the 360◦ in one period. The direction of orbital angular defined themode that has positive and negative values.

The phase of the OAM wave speed is vortex-like and the power distribution in the centralarea is very low as donut-like. An OAM antenna was designed by 32 patch antenna array(in mode=1) as shown in Fig. 6.1 The patch antennas were generated by 1-32 divider inthe same phase and each antenna were rotated in 11.25◦ to the beside antennas. The phasedifference 𝜑 each antenna is also 11.25◦ whose mode 𝑀 is 1. The beam pattern is shownin Fig. 6.2 and its phase diagram is shown in Fig. 6.3. The main lobe ring angle \ and theantenna radius 𝑅 can be describe as (6.1).

sin\ ≈ _𝑀

2𝜋𝑅. (6.1)

116 CHAPTER 6. MICROWAVE WIRELESS POWER TRANSMISSION SYSTEMS

Here, _ is the wavelength of the transmission frequency. The lower mode has a smallermain lobe angle. However, it’s difficult to improve the OAM antenna gain and the receivingantenna also has donut distribution beam. It will be worked in a low transmission efficiencyand the future work is design a concentration beam of the OAM antenna.

Finally, there is anther method to construct the information and power transmissionsystem whose transmission wave in the same microwave. Here, two demonstration systemswas introduced as the phased array in beam forming in Section.1 and wireless TV in Section.2.

R

N=32

Fig. 6.1 Photo of the OAM antenna.

Fig. 6.2 Simulation beam pattern of the OAM antenna.

6.1. INFORMATION TRANSMISSION SYSTEM WITH BEAM FORMING 117

Fig. 6.3 Measurement of beam phase in near field of the OAM antenna.

6.1 Information Transmission System with Beam Forming

The core of the magnetron phased array is the phased controlled magnetron (PCM). Chapter4.5.1 introduced the use of a PCM to transmit an audio signal. However, the response timeof the PCM requires 50 `𝑠. It determines that the communication speed will not exceed200 kbps. If the communication rate is higher, the phase-controlled magnetron will lose thephase-locked state. We modulated the magnetron phased array to transfer a video signalwithout PLL circuit.

We built a communication system with the magnetron phased array system which devel-oped in Chapter 3. We made some adjustments to the magnetron phased array system andits system diagram is shown in Fig. 6.4. The injected signal was replaced with a frequencymodulated (FM) signal that carries a camera video signal. The FM signal is divider to injectto four magnetrons. The phase shifter 𝛽 is used here for beam forming, which was used forthe PLL circuit in Chapter 3 . At the receiver, we set up two rectifying antennas to confirmthe effect of beam deflection. A frequency demodulator PAT-260 (Pakite) is used to receivethe FM signal and output it to a TV.

In the case where the four magnetrons are working in the injection locking state, theTV in the receiver successfully restores the video signal of the camera. Then, the phaseshifter 𝛽 is controlled to form the microwave beam. At the same time, and the TV signalcan still display a camera signal. The FM signal were divider to multiple microwaves andoutput the array antenna. This experiment verified that the multiple FM signal also can bedemodulation. The FM signal can still form phase interference through the array antenna,and can effectively control the microwave beam direction.


PC

LabVIEW

PAT630

FM Modulator

Phase shifter δ1 Phase shifter δ2 Phase shifter δ3 Phase shifter δ4

PCM1

Power supply 1

Phase shifter β1

PCM2

Power supply 2

Phase shifter β2

PCM3

Power supply 3

Phase shifter β3

PCM4

Power supply 4

Phase shifter β4

Slot antenna 1 Slot antenna 2 Slot antenna 3 Slot antenna 4

Analog output

(NI 9263)

Divider

Camera

PAT630

FM Demodulator

Rectenna 1 Rectenna 2 TV

Beam1 Beam2

Fig. 6.4 Experimental system of the information and power transfer system with magnetronphased array.

6.2 Wirelessly-Powered TV

A wireless power transfer system that transmitted microwave drove the power and videosignal of a TV which worked as a wirelessly-powered TV. At the transmitter, we modulateda video signal on a 2.45 GHz sine wave via frequency modulation, a 2.45 GHz injection-locked magnetron could amplify the frequency modulation signal. Utilizing injection lockingmethod, we injected the modulated signal to a 2.45 GHz magnetron and the magnetronamplified this modulation signal. At the receiver, we rectified the microwave energy tothe power source and demodulated this microwave to the video signal of the TV. Thewireless power transfer distance was 3.5 meters. At the aid of the transmitted microwave, wesuccessfully rectified 48 W DC power and demodulated the video signal.

6.2. WIRELESSLY-POWERED TV 119

6.2.1 Experimental Arrangement

We built a 2.45 GHz microwave power transfer TV system which was shown in Fig. 6.5.At the transmitter, the FM modulator (Pakite PAT240) which can modulate the video signaland audio signal on the 2.45 GHz band microwave (Bandwidth: 2 MHz). The modulatedsignal was amplified to 10 W then was injected to a 2.45 GHz magnetron. The magnetronwas locked with the injection signal and output the modulation signal via a circulator to ahorn antenna. Here, the full wave voltage doubler was improved from the power source of amicrowave oven. It is able to keep the magnetron oscillate in continuous wave [113]. Theripple rate of magnetron anode voltage is 4.16% and the oscillation frequency of magnetronwas shifted less 3 MHz bandwidth. By the 10 W injection power, the magnetron worked inan injection-locked state.

XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE

Injection-Locked CW Magnetron for a

wirelessly-powered TV

Bo Yang Tomohiko Mitani Naoki Shinohara

Research Institute for Sustainable Humanosphere

Kyoto University

Uji, Japan

{yang_bo, shino, mitani}@rish.kyoto-u.ac.jp

Abstract— It is shown a wireless power transfer system that

transmitted microwave drove the power and video signal of a

TV which worked as a wirelessly-powered TV. At the

transmitter, we modulated a video signal on a 2.45 GHz sine

wave via frequency modulation, a 2.45 GHz injection-locked

magnetron could amplify the frequency modulation signal.

Utilizing injection locking method, we injected the modulated

signal to a 2.45 GHz magnetron and the magnetron amplified

this modulation signal. At the receiver, we rectified the

microwave energy to the power source and demodulated this

microwave to the video signal of the TV. The wireless power

transfer distance was 3.5 meters. At the aid of the transmitted

microwave, we successfully rectified 48 W DC power and

demodulated the video signal.

Keywords—Magnetron, injection-locked, wireless power

transfer, frequency modulation, rectification

I. INTRODUCTION

Magnetrons as low cost and high efficiency microwave

sources, are widely applied in heating areas such as

microwave ovens. According to Alder's equation [1], our

research group has reduced the noise of the magnetron and

developed the phase-controlled magnetron as the transmitter

of a wireless power transfer system [2]. Using an injection-

locked magnetron, the transmission of phase-shift-keying

(PSK) data at 2 Mb/s has been achieved via Tahir et al. [3].

In the recent study, a full-wave voltage doubler circuit which

improved the power source of a microwave oven, was used

as the power source of the injection-locked magnetron. A 4

Mbps transmission of the frequency-shift-keying (FSK) data

was achieved [4]. Moreover, we evaluated several

modulation performance of the 2.45 GHz and 5.8 GHz band

injection-locked magnetron which achieved at 10 Mbps

(Po/Pi:13.43 dB) [5]. These research results show that the

magnetron noise can be limited in a low level, which is good

enough for wireless power transfer and communication.

In this study, we utilized this injection-locked magnetron

as the transmitter of the wirelessly-powered TV system. At

the receiver, through the received antenna and rectifier the

microwave was rectified to the DC power and was

demodulated to the video signal, as the power source and

signal of the TV.

II. EXPERIMENTAL ARRANGEMENT

We build a 2.45 GHz microwave power transfer TV system which was shown in Fig. 1. At the transmitter, the FM modulator (Pakite PAT240) which can modulate the video signal and audio signal on the 2.45 GHz band microwave (Bandwidth: 2 MHz). The modulated signal was amplified to 10 W then was injected to a 2.45 GHz

magnetron. The magnetron was locked with the injection signal and output the modulation signal via a circulator to a horn antenna. Here, the full wave voltage doubler was improved from the power source of a microwave oven. It is able to keep the magnetron oscillate in continuous wave [4]. The ripple rate of magnetron anode voltage is 4.16% and the oscillation frequency of magnetron was shifted less 3 MHz bandwidth. By the 10 W injection power, the magnetron worked in an injection-locked state.

DVD Player

FM Modulator Amplifier

Circulator

Full wave

voltage doublerMagnetron

Antenna

Transmitter

TV

FM demodulator Antenna

Rectifier

Receiver

DC/DC Converter

Fig. 1 Wirelessly-powered TV system diagram.

Fig. 2 Photo of a 2.45GHz rectifier circuit board

At the receiver, we set 54 patch antennas, each one connected to a rectifier circuit. The photo of a rectifier circuit[6] is shown in Fig. 2. The rectifier circuit efficiency with different load and input power are shown in Fig. 3 and Fig. 4, respectively. Through the experimental data, the maximum rectifier efficiency was 51.7% when it worked at 60 Ω load and 1.8 W input power. A rectifier circuit was

Fig. 6.5 Experimental system of the injection-locked magnetron.

At the receiver, we set 54 patch antennas, each one connected to a rectifier circuit. Thephoto of a rectifier circuit [114] is shown in Fig. 6.6. The rectifier circuit efficiency withdifferent load and input power are shown in Fig. 6.7 and Fig. 6.8, respectively. Through theexperimental data, the maximum rectifier efficiency was 51.7% when it worked at 60 Ω load


and 1.8 W input power. A rectifier circuit was connected to the demodulator (Pakite PAT240).The other 53 rectifier circuits outputs are connected to a DC/DC converter in parallel. TheDC/DC converter limited the voltage lower than 16 V which is the TV work voltage. Thevideo signal and audio signal of the demodulator was connected to the TV. The distancebetween the transmitter and the receiver was 3.5 m. The other parameters are shown in Table.6.1.

XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE

Injection-Locked CW Magnetron for a

wirelessly-powered TV

Bo Yang Tomohiko Mitani Naoki Shinohara

Research Institute for Sustainable Humanosphere

Kyoto University

Uji, Japan

{yang_bo, shino, mitani}@rish.kyoto-u.ac.jp

Abstract— It is shown a wireless power transfer system that

transmitted microwave drove the power and video signal of a

TV which worked as a wirelessly-powered TV. At the

transmitter, we modulated a video signal on a 2.45 GHz sine

wave via frequency modulation, a 2.45 GHz injection-locked

magnetron could amplify the frequency modulation signal.

Utilizing injection locking method, we injected the modulated

signal to a 2.45 GHz magnetron and the magnetron amplified

this modulation signal. At the receiver, we rectified the

microwave energy to the power source and demodulated this

microwave to the video signal of the TV. The wireless power

transfer distance was 3.5 meters. At the aid of the transmitted

microwave, we successfully rectified 48 W DC power and

demodulated the video signal.

Keywords—Magnetron, injection-locked, wireless power

transfer, frequency modulation, rectification

I. INTRODUCTION

Magnetrons as low cost and high efficiency microwave

sources, are widely applied in heating areas such as

microwave ovens. According to Alder's equation [1], our

research group has reduced the noise of the magnetron and

developed the phase-controlled magnetron as the transmitter

of a wireless power transfer system [2]. Using an injection-

locked magnetron, the transmission of phase-shift-keying

(PSK) data at 2 Mb/s has been achieved via Tahir et al. [3].

In the recent study, a full-wave voltage doubler circuit which

improved the power source of a microwave oven, was used

as the power source of the injection-locked magnetron. A 4

Mbps transmission of the frequency-shift-keying (FSK) data

was achieved [4]. Moreover, we evaluated several

modulation performance of the 2.45 GHz and 5.8 GHz band

injection-locked magnetron which achieved at 10 Mbps

(Po/Pi:13.43 dB) [5]. These research results show that the

magnetron noise can be limited in a low level, which is good

enough for wireless power transfer and communication.

In this study, we utilized this injection-locked magnetron

as the transmitter of the wirelessly-powered TV system. At

the receiver, through the received antenna and rectifier the

microwave was rectified to the DC power and was

demodulated to the video signal, as the power source and

signal of the TV.

II. EXPERIMENTAL ARRANGEMENT

We build a 2.45 GHz microwave power transfer TV system which was shown in Fig. 1. At the transmitter, the FM modulator (Pakite PAT240) which can modulate the video signal and audio signal on the 2.45 GHz band microwave (Bandwidth: 2 MHz). The modulated signal was amplified to 10 W then was injected to a 2.45 GHz

magnetron. The magnetron was locked with the injection signal and output the modulation signal via a circulator to a horn antenna. Here, the full wave voltage doubler was improved from the power source of a microwave oven. It is able to keep the magnetron oscillate in continuous wave [4]. The ripple rate of magnetron anode voltage is 4.16% and the oscillation frequency of magnetron was shifted less 3 MHz bandwidth. By the 10 W injection power, the magnetron worked in an injection-locked state.

DVD Player

FM Modulator Amplifier

Circulator

Full wave

voltage doublerMagnetron

Antenna

Transmitter

TV

FM demodulator Antenna

Rectifier

Receiver

DC/DC Converter

Fig. 1 Wirelessly-powered TV system diagram.

Fig. 2 Photo of a 2.45GHz rectifier circuit board

At the receiver, we set 54 patch antennas, each one connected to a rectifier circuit. The photo of a rectifier circuit[6] is shown in Fig. 2. The rectifier circuit efficiency with different load and input power are shown in Fig. 3 and Fig. 4, respectively. Through the experimental data, the maximum rectifier efficiency was 51.7% when it worked at 60 Ω load and 1.8 W input power. A rectifier circuit was

Fig. 6.6 Photo of a 2.45GHz rectifier circuit board.

connected to the demodulator (Pakite PAT240). The other 53 rectifier circuits outputs are connected to a DC/DC converter in parallel. The DC/DC converter limited the voltage lower than 16 V which is the TV work voltage. The video signal and audio signal of the demodulator was connected to the TV. The distance between the transmitter and the receiver was 3.5 m. The other parameters are shown in Table. I.

When the transmitter was activated, the TV was turned on by microwave power and the DVD data was properly displayed on the wirelessly-powered TV. 48W DC power was supplied to the TV and the demodulator. The photo of a wireless power transfer system of TV as shown in Fig. 5. As a wireless power transfer system, it requires a low power loss or a high transfer efficiency, which contains microwave conversion efficiency, antenna transfer efficiency and rectifier efficiency. In this system, the rectifier circuit didn’t work at the optimal load which can be improved. We also can improve the transfer efficiency via improving the effective antenna aperture.

0

10

20

30

40

50

0

2

4

6

8

10

10 100 1000

Eff

icie

ncy

[%]

Ou

tpu

t[m

W]

Load[Ω]

Output powerEfficiency

Fig. 3 Rectifier efficiency of the rectifier circuit with different

load(input 50mW).

0

10

20

30

40

50

60

0

200

400

600

800

1000

0 500 1000 1500 2000

Eff

icie

ncy

[%]

Ou

tpu

t[m

W]

Input[mW]

Out poweerEfficiency

Fig. 4 Rectifier efficiency of the rectifier circuit with different input

power(load:60Ω)

Fig. 5 Photo of the wirelessly-powered TV.

TABLE I. PARAMETERS OF THE WIRELESSLY-POWERED TV

III. CONCLUSIONS

We developed a wireless power and video transfer

system for the wirelessly-powered TV in the microwave

band. It demonstrated that a 2.45 GHz injection-locked

magnetron could transmit the microwave power and the

video signal by frequency modulation. Through the

transmitted microwave, we successfully rectified 48W DC

power for the TV and demodulated the video signal.

ACKNOWLEDGMENT

This work was supported by the collaborative research




REFERENCES

[1] R. Adler, “A Study of Locking Phenomena in Oscillators,” Proceedings of I.R.E and Waves and Electrons, vol. 34, pp. 351-357, 1946

[2] N. Shinohara, H. Matsumoto, K. Hashimoto, “Phasecontrolled magnetron development for SPORTS: Space power radio transmission system,” Radio Science Bulletin, vol. 2004, No. 310, pp. 29-35, 2004

[3] I. Tahir, A. Dexter and R. Carter, “Frequency and Phase Modulation Performance of an Injection-Locked CW Magnetron,” IEEE Transactions on Electron Devices, vol. 53, No. 7, pp. 1721-1729, 2006

[4] B. Yang, T. Mitani, N. Shinohara, “Experimental study on frequency modulation of an injection-locked magnetron based on full wave voltage doubler,” IVEC2018, Monterey, CA, USA, 2018

[5] B. Yang, T. Mitani, N. Shinohara, “Evaluation of the Modulation Performance of Injection-Locked Continuous-Wave Magnetrons,” IEEE Transactions on Electron Devices, Vol. 66, pp. 709–715, Jan. 2019

[6] J.Kojima, N.Shinohara, T.Mitani, T.Hashimoto, N.Kishi, H.Toyomura and A.Okazaki, “Development of highly efficient microwave wireless charging system for electric vehicle,” Technical Report of IEICE Information and Communication Engineers SPS2007-16 (2008-03)

Magnetron 2M236-M42(Panasonic)

Anode Current 140 mA

Anode Voltage -3.68 kV(DC)

Filament Current 7.4 A

Filament Voltage 3.35 V(AC 60Hz)

Injected Power 10 W

Output Frequency 2.448 GHz-2.450 GHz

Output Power 329 W(RF)

Rectified Power 48 W(DC)

Modulation Frequency ,odulation

Transmitter Antenna Gain 16 dBi (SPC)

TV LL-M1550A (Sharp)

Circulator

High voltage

power source

Magnetron

Horn antenna

Receiver antenna

TV

← 3.5m →

Power + Data

Fig. 6.7 Rectifier efficiency of the rectifier circuit with different load (input 50mW).

6.2. WIRELESSLY-POWERED TV 121



0

10

20

30

40

50

0

2

4

6

8

10

10 100 1000

Eff

icie

ncy

[%]

Ou

tpu

t[m

W]

Load[Ω]



load(input 50mW).

0

10

20

30

40

50

60

0

200

400

600

800

1000

0 500 1000 1500 2000

Eff

icie

ncy

[%]

Ou

tpu

t[m

W]

Input[mW]



power(load:60Ω)



III. CONCLUSIONS








ACKNOWLEDGMENT





REFERENCES












Injected Power 10 W







Circulator

High voltage

power source

Magnetron

Horn antenna

Receiver antenna

TV

← 3.5m →

Power + Data

Fig. 6.8 Rectifier efficiency of the rectifier circuit with different input power (load:60Ω).

6.2.2 Experimental Results

When the transmitter was activated, the TV was turned on by microwave power and the DVDdata was properly displayed on the wirelessly-powered TV. 48W DC power was supplied tothe TV and the demodulator. The photo of a wireless power transfer system of TV as shownin Fig. 6.9. As a wireless power transfer system, it requires a low power loss or a high transferefficiency, which contains microwave conversion efficiency, antenna transfer efficiency andrectifier efficiency. In this system, the rectifier circuit didn’t work at the optimal load whichcan be improved. We also can improve the transfer efficiency via improving the effectiveantenna aperture.



0

10

20

30

40

50

0

2

4

6

8

10

10 100 1000

Eff

icie

ncy

[%]

Ou

tpu

t[m

W]

Load[Ω]



load(input 50mW).

0

10

20

30

40

50

60

0

200

400

600

800

1000

0 500 1000 1500 2000

Eff

icie

ncy

[%]

Ou

tpu

t[m

W]

Input[mW]



power(load:60Ω)



III. CONCLUSIONS








ACKNOWLEDGMENT





REFERENCES












Injected Power 10 W







Circulator

High voltage

power source

Magnetron

Horn antenna

Receiver antenna

TV

← 3.5m →

Power + Data

Fig. 6.9 Photo of the wirelessly-powered TV.


Table 6.1 PARAMETERS OF THE WIRELESSLY-POWERED TV.

Magnetron 2M236-M42(Panasonic)Anode Current 140 mAAnode Voltage -3.68 kV(DC)Filament Current 7.4 AFilament Voltage 3.35 V(AC 60Hz)Injected Power 10 WOutput Frequency 2.448 GHz-2.450 GHzOutput Power 329 W(RF)

Rectified Power 48 W(DC)Modulation Frequency ,odulationTransmitter Antenna Gain 16 dBi (SPC)TV LL-M1550A (Sharp)

We developed a wireless power and video transfer system for the wirelessly-powered TVin the microwave band. It demonstrated that a 2.45 GHz injection-locked magnetron couldtransmit the microwave power and the video signal by frequency modulation. Through thetransmitted microwave, we successfully rectified 48W DC power for the TV and demodulatedthe video signal.

6.3 Charging in the Microwave Oven

Due to current battery capacity limitations, electronic products require frequent charging.The magnetron in the microwave oven output almost 700 W microwave which is high enoughfor quickly charging. In this regard, we provide a quick charging solution, using a microwaveoven to build a wireless power transmission system, and charging to super capacitors. Theblock diagram of the quick charging system is shown in Fig. 6.10.

To build the high power receiver. We designed a single shut circuit of the rectifier whichconsisted 2 diode as shown in Fig. 6.11. a. We measured the load and power characteristicsof the rectifier. The rectifier circuit efficiency with different load and different input powerare shown in Fig. 6.12 and Fig. 6.13, respectively. The best efficiency is 78% when therectifier worked under the condition of 9 W input power and 60 Ω load. There are 7 patchantennas connect to the 7 rectifier circuits by the via as shown in Fig. 6.11. The outputs ofthe rectifiers were connected in parallel to a DC-DC converter which converter the voltageto 13 V. The DC-DC converter output the power to charge a super capacity unit. The supercapacity unit was consisted of 5 pieces 2.7 V/3000 F capacitors in series work as a mobilebattery. The super capacity unit output 5V via another DC-DC converter. The receiver of

6.3. CHARGING IN THE MICROWAVE OVEN 123

0

20

40

60

80

100

0 2000 4000 6000 8000 10000

Eff

icie

ncy

[%

]

Input power[mW]

0

20

40

60

80

100

0 20 40 60 80 100

Eff

icie

ncy

[%

]

Load[Ω]

Microwave oven

Rectenna

DC-DC converter

Super capacity

DC-DC converter

13V

5V

a b

c d

Fig. 6.10 Block diagram of the quick charging system.

the wireless charging system was built by these above rectenna, super capacity and DC-DCconverter.

Since the rectenna array only withstand a maximum of 63 W of power. In the microwaveoven, the receiver was tested with water for attenuation and measured 40 W DC power. Torealize a practical charging system, the technical difficulty lies in designing a high-powerrectifier circuit and how to solve the heat dissipation problem. The next step is designing ahigher power level rectenna, and charging the super capacity without water attenuate in themicrowave oven.


0

20

40

60

80

100

0 2000 4000 6000 8000 10000

Eff

icie

ncy

[%

]

Input power[mW]

0

20

40

60

80

100

0 20 40 60 80 100E

ffic

iency

[%

]Load[Ω]

Microwave oven

Rectenna

DC-DC converter

Super capacity

DC-DC converter

13V

5V

a b

c d

Fig. 6.11 Photo of the receiver. (a. a rectifier circuit, b. rear view of rectenna array, c. frontview of rectenna array. d. receiver)

0

20

40

60

80

100

0 2000 4000 6000 8000 10000

Eff

icie

ncy

[%

]

Input power[mW]

0

20

40

60

80

100

0 20 40 60 80 100

Eff

icie

ncy

[%

]

Load[Ω]

Microwave oven

Rectenna

DC-DC converter

Super capacity

DC-DC converter

13V

5V

a b

c d

Fig. 6.12 Rectifier circuit efficiency with different load @ input power 6 W. (Orange line:rectifier efficiency, blue line: reflection efficiency)

6.4. EFFICIENCY MEASUREMENT OF THE WPT SYSTEM 125

0

20

40

60

80

100

0 2000 4000 6000 8000 10000

Eff

icie

ncy

[%

]

Input power[mW]

0

20

40

60

80

100

0 20 40 60 80 100

Eff

icie

ncy

[%

]

Load[Ω]

Microwave oven

Rectenna

DC-DC converter

Super capacity

DC-DC converter

13V

5V

a b

c d

Fig. 6.13 Rectifier circuit efficiency with different input power @60 Ω load. (Orange line:rectifier efficiency, blue line: reflection efficiency)

6.4 Efficiency Measurement of the WPT System

In order to verify how to get a high-efficiency WPT system, we used several antennas tomeasure the transmission efficiency. This section measured magnetron WPT system withdifferent antennas.

6.4.1 Magnetron WPT System with Horn Antenna

A simple magnetron WPT system was built as shown in Fig. 6.14. The transmitter is amagnetron output to the horn antenna, and there is a rectified antenna array (IHI Aerospace)as the receiver in front of the horn antenna. We adjusted the distance between the transmitterand the receiver to measure the efficiency of the WPT system. At the same time, the effectof the filament power of the magnetron ON/OFF on efficiency were observed. Table 6.2shows the experimental results. From the experimental results, the highest total efficiencyis achieved at the distance is longer than 2.5 meters, The relationship between efficiencyand distance is analyzed in detail in Chapter 3.4.2. The ON/OFF of the filament will affectthe efficiency of the magnetron, but it will not invisibly receive efficiency and rectificationefficiency.


Fig. 6.14 Photo of Magnetron WPT experiment with horn antenna.

6.4.2 Magnetron WPT System with Patch Antenna

We performed the same experiment by a patch antenna array instead of a horn antenna, asshown in Fig. 6.15. The actively antennas distribution is 5 pieces in Row 1 and Row 5, 7pieces in Row 2 and Row 4, 8 pieces in Row 3. A 1-32 waveguide divider was connectedto the patch antenna via coaxial cable. Here, the transmission |𝑆21 | from the waveguide toeach patch antennas is -18.1 dB, while the transmission |𝑆21 | without loss is -15 dB, so thetransmission loss of the divider and coaxial cable is 3 dB. In this experiment, we adjustedthe frequency of the magnetron to measure the effect of on the efficiency. Table 6.4 showsthe experimental results with different frequencies. The data shows that the frequency has agreater impact on the efficiency of the magnetron, and basically has no effect on the collectionefficiency. We also measured the efficiency with different transmission distances as shown inTable 6.2. The overall efficiency reached the highest 9.73% at 5.6 meters.


Fig. 6.15 Photo of Magnetron WPT experiment with patch antenna.

128 CHAPTER 6. MICROWAVE WIRELESS POWER TRANSMISSION SYSTEMSTa

ble

6.2

EX

PER

IME

NT

RE

SULT

SO

FM

AG

NE

TR

ON

WPT

SYST

EM

WIT

HH

OR

NA

NT

EN

NA

dist

ance

(m)

Fila

men

tpo

wer

Ano

decu

rren

t(m

A)

Ano

decu

rren

t(-

V)

Tran

smis

sion

pow

er(W

)

Rec

eive

rpo

wer

(W)

DC

-RF

effic

ienc

yR

F-R

F-D

Cef

ficie

ncy

Tota

lef

ficie

ncy

4.75

ON

250

4480

702

6162

.68%

8.69

%5.

45%

4.75

OFF

250

4480

740

6366

.07%

8.51

%5.

63%

3.85

ON

250

4480

695

6362

.05%

9.06

%5.

63%

3.85

OFF

250

4480

731

6665

.27%

9.03

%5.

89%

2.5

ON

250

4480

701

8562

.59%

12.1

3%7.

59%

1.78

ON

250

4480

690

7561

.61%

10.8

7%6.

70%

Tabl

e6.

3EX

PER

IMEN

TR

ESU

LTS

OF

MA

GN

ETR

ON

WPT

SYST

EMW

ITH

PATC

HA

NTE

NN

AIN

DIF

FER

ENT

FREQ

UEN

CY

Freq

uenc

y(G

Hz)

dist

ance

(m)

Fila

men

tpo

wer

Ano

decu

rren

t(m

A)

Ano

decu

rren

t(-

V)

Tran

smis

sion

pow

er(W

)

Rec

eive

rpo

wer

(W)

DC

-RF

effic

ienc

yR

F-R

F-D

Cef

ficie

ncy

Tota

lef

ficie

ncy

5.78

92.

55O

FF14

939

0033

537

57.6

5%11

.04%

6.37

%5.

822.

55O

FF15

037

2024

729

44.2

7%11

.74%

5.20

%5.

791

2.55

OFF

150

3850

320

3655

.41%

11.2

5%6.

23%

5.76

72.

55O

FF25

044

2066

074

59.7

3%11

.21%

6.70

%


Tabl

e6.

4E

XPE

RIM

EN

TR

ESU

LTS

OF

MA

GN

ET

RO

NW

PTSY

STE

MW

ITH

PAT

CH

AN

TE

NN

AIN

DIF

FER

EN

TD

ISTA

NC

E

dist

ance

(m)

Fila

men

tpo

wer

Ano

decu

rren

t(m

A)

Ano

decu

rren

t(-

V)

Tran

smis

sion

pow

er(W

)

Rec

eive

rpo

wer

(W)

DC

-RF

effic

ienc

yR

F-R

F-D

Cef

ficie

ncy

Tota

lef

ficie

ncy

2.55

OFF

250

4420

660

7459

.73%

11.2

1%6.

70%

3.45

OFF

250

4420

647

8058

.55%

12.3

6%7.

24%

4.4

OFF

250

4420

650

8658

.82%

13.2

3%7.

78%

5.45

OFF

250

4420

646

9258

.46%

14.2

4%8.

33%

5.45

OFF

267

4420

690

9858

.47%

14.2

0%8.

30%

6.18

OFF

250

4500

716

9763

.64%

13.5

5%8.

62%

6.8

OFF

250

4480

712

8663

.57%

12.0

8%7.

68%

5.6

OFF

250

4500

724

109

64.3

6%15

.11%

9.73

%

Tabl

e6.

5C

OM

PAR

EO

FM

AG

NE

TR

ON

WPT

SYST

EM

WIT

HD

IFFE

RE

NT

AN

TE

NN

AS

Mag

netr

onQ

uant

ityM

AX

outp

utpo

wer

(W)

DC

–RF

effic

ienc

yTr

ansm

issi

onpo

wer

(W)

Rec

eive

rpo

wer

(W)

dist

ance

(m)

RF-

RF-

DC

effic

ienc

yto

tal

effic

ienc

ysl

otar

ray

ante

nna

416

3760

.70%

1204

142

511

.79%

6.88

%ho

rnan

tenn

a1

740

66.0

7%70

185

2.5

12.1

3%7.

59%

patc

han

tenn

a1

724

60.4

6%72

410

95.

615

.12%

9.73

%


6.5 Discussion of a High Efficiency WPT System

For WPT, a very critical parameter is the total efficiency of the system. Table 6.5 summarizesthe data of the WPT experiment of the 5.8 GHz magnetron in this paper. Among them, thephased array magnetron are four magntrons which is different from other magnetrons system.The highest total efficiency in this paper is to use patch antenna to reach 9.73% at 5.6 m. It isnecessary to further improve the total efficiency.

For DC to RF inverters, the magnetron should be the most efficient of all devices. Themarketed 900 MHz magnetron achieves 90% efficiency, but there is no 2.45 GHz or 5.8GHz magnetron reaches this level. If combined with the problem of the life time of themagnetron, developing high-efficiency CFA or another improved magnetron should be aresearch direction.

The collection efficiency of wave beam depends on factors such as antenna gain andpolarity matching. The collection efficiency of W.C Brown reaches 95%, which is acceptableto WPT. The receiving array antenna used in this experiment did not exactly match thetransmitting antenna. For a system with high collection efficiency, the transmitting antennaand the receiving antenna must be well matched.

The receiver efficiency depends on microwave rectification efficiency and combiningefficiency. The efficiency of the 2.45 GHz rectifier circuit can reach more than 90%, butthis efficiency will change greatly under different loads and different power level. A betterreceiving system, according to the energy density distribution of the beam, set up a rectifiercircuit to achieve the best input power state. When combing with the rectifier circuit,the optimal load state should be considered. DC-DC output connection is the simplestsolution, but this will have about 10% loss. In the high-power rectifier device, Vanke’s CWCexperiment data is 2.45 GHz, 10 kW 83% rectification efficiency. However, there is norelevant information to show that the team continues related research, and it is also veryregrettable that CWC has not been applied to the relevant WPT system.

6.6 Summary

In this chapter, several situations of non-interference of signal-to-energy simultaneous trans-mission are described, and the design and test results of the OAM antenna are introducedin detail. Then, two information and power transmission systems using magnetrons wereintroduced, and both realized video transmission. In addition, the realization and technicaldifficulties of charging in a microwave oven. Then, a horn antenna and a micro-strip antennaarray were used for WPT experiments comparing with the experimental results of the slot

6.6. SUMMARY 131

antenna The total efficiency were measured with different frequency, different distance,injection locking and other different parameters. Finally, the high-efficiency WPT system isanalyzed.

Chapter 7

Conclusion

In this thesis, the history of WPT, microwave WPT, high power microwave transmitter wereintroduced in the Chapter 1. Chapter 2 devised a new Power-variable Phase-Controlledmethod of the Magnetron (PVPCM). The experimental results show that the developed 5.8GHz phase-controlled magnetron had a variable output power range from 160 to 329 W.Furthermore, the phase accuracy was nearly ±1◦, and the response time was less than 100 `s.

Chapter 3 introduced a magnetron phased array system, which was developed for theWPT. Four PVPCMs were proposed as a phased array system. A waveguide slot array antennawas designed and used for the output of the PVPCMs. We demonstrated the properties ofmicrowave beam forming and WPT based on the aforementioned magnetron phased arraysystem. In the horizontal and vertical directions both, a beam scanning range of ±3° wasobtained by adjusting the output phase of the magnetrons. Moreover, using a 5.8 GHzrectenna array system (IHI Aerospace) as the receiver, a microwave WPT system was built. Itis verified that the output phase and power of the magnetron phased array are adjustable. Themeasured maximum output of the magnetron phased array was 1870 W with 61.0% DC–RFefficiency. The received DC power reaches 142 W at a distance of 5 m when the outputmicrowave power of the magnetron phased array is 1304 W. It is the first time a beamformingtransmitter is developed based on power-variable magnetrons.

Development of the modulation magnetron was described in Chapter 4. In the past,magnetrons can communicate at a low transmission rate, which has been studied in theUnited States. While investigating the phase-controlled magnetrons, we discovered themodulation of the magnetron output and its enabled for communication. The transmissionrate was improved, and it achieved the transmission of phase-shift keying and frequency-shiftkeying at 10 Mbps. Finally, the transmission of the audio and video information and relaycontrolling data were demodulated.

134 CHAPTER 7. CONCLUSION

The ripple rate of full wave voltage doubler and switched-mode circuit of the lowcost power supply were improved to below 1% in Chapter 5. These power supplies weresuccessful used in phase-controlled magnetron, frequency modulation system, magnetronload fluctuation experiment, magnetron output combining system. It proposed a improvedmethod of the low cost power supplies for magnetron system.

Chapter 6 summarized the information and power transfer systems and efficiency mea-surement experiment. The magnetron phased array which was injected with frequencymodulation signal, also can form the microwave beam. It’s a new way for information andpower transfer system. Another information and power transfer system of a wireless TVsystem was invented at a low cost. The video signal and the power consumption of the TV aretransmitted with the same microwave from the magnetron. The receiving antenna is placed3.5 m away and the microwave is rectified to supply to the TV. Also, there is a demodulatorthat demodulated microwave into a video signal. It’s the first time for developing the wirelessTV that tested a 300 W class microwave wireless power and information transmission system.

The highest DC–RF–RF–DC efficiency of the 5.6 m distance WPT experiments in thisresearch was achieved at 9.73% . The paper suggested further on improving the efficiency ofthe microwave WPT system.

References

[1] N. Tesla, "Transmission of energy without wires," Sci Am (Suppl), vol. 57, pp. 237-260,1904.

[2] H. Winfield Secor,"Tesla’s Million Dollar Folly,"Export American Industries, pp. 52-53,March 1, 1916

[3] M. Hutin and M. Leblanc, "Transformer System for Electric Railways," US PatentUS527857A, 1894

[4] H. Yagi and S. Uda, "Projector of the Sharpest Beam of Electric Waves," Proc. ImperialAcademy of Japan, pp. 49-52. February 1926. doi:10.2183/pjab1912.2.49

[5] S. Uda, "On the WireleSS Beam Of Short Electric Waves," vol.46, pp.273-282,No.452,1926 doi:10.11526/ieejjournal1888.46.273

[6] Yagi, H., and S. Uda, "On the feasibility of power transmission by electric waves," Proc.Third pan-pacific congress held in Tokyo, Vol.2, pp.1306-1313, 1926

[7] W. C. Brown, The History of Power Transmission by Radio Waves," IEEE Transactionson Microwave Theory and Techniques, vol.32, Issue: 9, pp.1230-1242.Sep 1984, DOI:10.1109/TMTT.1984.1132833

[8] W. C. Brown, "The History of the Development of the Rectenna", Johnson Space CenterSolar Power Satellite Microwave Transmission and Reception in NASA, pp.271-280,1980

[9] Andrey D. Grigoriev, Vyacheslav A. Ivanov, Sergey I. Molokovsky, "Microwave Elec-tronics," Springer, pp247, 2018

[10] P. E. Glaser, "Power from the sun: Its future," Science, vol. 162, no.3856, pp. 857–861,Nov. 1968, DOI: 10.1126/science.162.3856.857

[11] Dickinson R. M, "Evaluation of a microwave high-power reception-conversion array forwireless power transmission," NASA STI/Recon Technical Report NASA-CR-145625,1975

[12] SHARP(Stationary High Altitude Relay Platform) [Online]. Available:https://www.friendsofcrc.ca/Projects/SHARP/sharp.html [Accessed: 25- June-2020]

136 REFERENCES

[13] H.Matsumoto, MINIX Plan [Online]. Available:http://www.isas.jaxa.jp/j/isasnews/backnumber/1982/ISASnews020.pdf [Accessed:25- June- 2020]

[14] N.Kaya, H. Matsumoto,"METS Rocket Experiment and MILAX Airplane Demonstra-tion," The Journal of Space Technology and Science 8.2 (1992)

[15] N.Kaya, H. Kojima, H. Matsumoto, M. Hinada and R. Akiba, "ISY-METS RocketExperiment for Microwave Energy Transmission, Acta Astronautica," 34, 43-46, 1994.doi:10.1016/0094-5765(94)90241-0

[16] Y. Fujino, M. Fujita, N. Kaya, S. Kunimi, M. Ishii, N. Ogihara, N. Kusaka and S. Ida,"A dual polarization microwave power transmission system for microwave propelledairship experiment," in proceedings of The International Symposium on antennas andPropagation JAPAN, 1996, pp. 393-396.

[17] N.Shimokura, N.Kaya, N.Shinohara, H.Matsumoto. "Point-to-Point Microwave PowerTransmission Experiment." Transactions of the Institute of Electrical Engineers ofJapan. B, 116(6), 648-653.(1996).

[18] J.Boyd, "Here comes the wallet phone [wireless credit card]." IEEE Spectrum 42.11(2005): 12-14.

[19] A. Kurs, A. Karalis, R. Moffatt, J. D. Joannopoulos, P. Fisher, and M. Solja£i¢, "Wire-less power transfer via strongly coupled magnetic resonances,"Science, vol. 317, no.5834, pp. 83-86, Jul. 2007.

[20] A. Karalis, J. D. Joannopoulos, and M. Solja£i¢, "Effcient wireless non-radiativemid-range energy transfer,"Ann. Phys., vol. 323, no. 1, pp. 34-48, Jan. 2008.

[21] Kim, Han Joon, Hirayama, Hiroshi , Kim, Sanghoek, Han, Ki, Zhang, Rui, Choi,Ji-Woong. (2017). Review of Near-Field Wireless Power and Communication forBiomedical Applications. IEEE Access. PP. 1-1. 10.1109/ACCESS.2017.2757267.

[22] Wireless Power Consortium Website. [Online]. Available:http://www.wirelesspowerconsortium.com[Accessed: 25- June- 2020]

[23] AirFuel Alliance Website. [Online]. Available: http://www.airfuel.org[Accessed: 25-June- 2020]

[24] G. Calabrese ,The Greening of the Automotive Industry,Business & Economics,pp58,2016

[25] K.Hashioto, H.Yamakawa, N.Shinohara, T.Mitani, F.Takahashi, H.Yonekura, T.HiranoT.Fujiwara, K.Nagano, and S.Kawasaki, Microwave Power Transmission Experimentfrom an Airship,Technical Report of IEICE,SPS2009-03, April.

[26] JAXA, "Ground demonstration testing of microwave wireless power transmission," [On-line]. Available: http://www.kenkai.jaxa.jp/eng/research/ssps/150301.html [Accessed:25- Feb- 2020]

REFERENCES 137

[27] [Online]. Available: https://www.daimler.com/innovation/case/electric/eq-ecosystem.html [Accessed: 25- June- 2020]

[28] [Online]. Available: https://www.nissan-global.com/JP/TECHNOLOGY/OVERVIEW/wcs.html[Accessed: 25- June- 2020]

[29] [Online]. Available: https://monoist.atmarkit.co.jp/mn/articles/1109/27/news077.html[Accessed: 25- June- 2020]

[30] [Online]. Available: https://ipt-technology.com/e-mobility/ [Accessed: 25- June- 2020]

[31] [Online]. Available: https://www.zte.com.cn/china/about/news/427348.html [Accessed:25- June- 2020]

[32] [Online]. Available: https://www.gov.uk/government/news/off-road-trials-for-electric-highways-technology [Accessed: 25- June- 2020]

[33] T. Aoki, Q. Yuan, D. Quang-Thang, M. Okada and H. Hsu, "Maximum TransferEfficiency of MIMO-WPT System," 2018 IEEE Wireless Power Transfer Conference(WPTC), Montreal, QC, Canada, 2018, pp. 1-3, doi: 10.1109/WPT.2018.8639417

[34] S. Kim and B. Lee, "Control of magnetic field distribution by excitation phasesof transmitters in MIMO WPT system," 2016 URSI Asia-Pacific Radio ScienceConference (URSI AP-RASC), Seoul, 2016, pp. 861-863, doi: 10.1109/URSIAP-RASC.2016.7601280..

[35] [Online]. Available: https://www.daihen.co.jp/products/wireless/ [Accessed: 25- June-2020]

[36] A. Celeste, P. Jeanty and G. Pignolet, "Case study in reunion island," Acta Astronaut.,vol. 54, no. 4, pp. 253-258 2004.

[37] T. Matsumuro, Y. Ishikawa, T. Mitani, N. Shinohara, M. Yanagase and M. Matsunaga,"Study of a single-frequency retrodirective system with a beam pilot signal using dual-mode dielectric resonator antenna elements," Wireless Power Transfer, vol. 4, no. 2, pp.132 2017.

[38] A.D. Kotin, "Satellite power system (SPS) resource requirements (critical materials,energy and land)," 1978.

[39] M. Omiya and K. Itoh, "Development of a Functional System Model of the Solar PowerSatellite, SPS2000," in proceedings of The International Symposium on Antennas andPropagation JAPAN, 1996, pp. 417-420.

[40] S. Sasaki, K. Tanaka, K. Higuchi, N. Okuizumi, S. Kawasaki, N. Shinohara, K. Sendaand K. Ishimura, "A new concept of solar power satellite: Tethered-SPS," Acta Astro-naut., vol. 60, no. 3, pp. 153-165 2007.

[41] M. Mori, H. Kagawa and Y. Saito, "Summary of studies on space solar power systemsof japan aerospace exploration agency (JAXA)," Acta Astronaut., vol. 59, no. 1-5, pp.132-138 2006.

138 REFERENCES

[42] T. Saito, Y. Kobayashi and H. Kanai, "Concept Study of Space Solar Power Systems inUSEF," in 55th International Astronautical Congress of the International AstronauticalFederation, the International Academy of Astronautics, and the International Instituteof Space Law, pp. R. 1.02, 2004.

[43] J. Mankins, N. Kaya and M. Vasile, "SPS-ALPHA: the first practical solar power satel-lite via arbitrarily large phased array (a 2011-2012 NIAC project)," in 10th InternationalEnergy Conversion Engineering Conference, 2012, pp. 3978.

[44] X. Hou, M. Li, L. Niu, L. Zhou, Y. Chen, Z. Cheng and H. Jia, "Multi-rotary jointsSPS," Online Journal of Space Communication, vol. 2016, no. 18 2016.

[45] K. Jin and W. Zhou, "Wireless laser power transmission: a review of recent progress,"IEEE Transactions on Power Electronics, vol. 34, pp. 3842-3859, 2018.

[46] Zhaohui, LI Zhijian ZHUANG Ganlin WU and L.B.Z. Xiulai, "Modeling of ultra-sonic wireless electrical energy transfer system," Journal of South China University ofTechnology (Natural Science), vol. 46, no. 3, pp. 72 2018.

[47] N.Shinohara, “Microwave and Millimeter Wave Power Transmission in the Far Field,”IEEE Journal of Microwaves, print in 2020

[48] H. T. Friis, “A note on a simple transmission formula,” Proc. IRE 34, 254–256 (1946).doi:10.1109/JRPROC.1946.234568

[49] J.W. Hawkins, The saber microwave-powered helicopter project and related WPTresearch at the University of Alaska Fairbanks AIP conference proceedings, 1092-1097,1998. doi:10.1063/1.54725.

[50] H. Ma, X. Li, L. Sun, H. Xu and L. Yang, "Design of high-efficiency microwave wirelesspower transmission system," Microwave Opt Technol Lett, vol. 58, pp. 1704-1707,2016.

[51] M. Kamiyama, R. Ishikawa and K. Honjo, "5.65 GHz High-Efficiency GaN HEMTPower Amplifier With Harmonics Treatment up to Fourth Order," in IEEE Microwaveand Wireless Components Letters, vol. 22, no. 6, pp. 315-317, June 2012, doi:10.1109/LMWC.2012.2197385.

[52] V. L. Granatstein, P. K. Parker, and C. M. Armstrong, "Scanning the TechnologyVacuum Electronics at the Dawn of the Twenty-First Century," Proc. IEEE, vol. 87, pp.702–716, May 1999

[53] S. Oliver, "Optimize a Power Scheme for these Transient Times", Electronic Design,Sep 30, 2014.

[54] [Online]. Available: https://www.nobelprize.org/prizes/physics/2014/summary/ [Ac-cessed: 25- June- 2020]

[55] D. Sprehn, R. M. Phillips, and G. Caryotakis, “The design and performance of 150- MWS-band klystrons,” SLAC-PUB-6677, Stanford Linear Accelerator Center, September1994.

REFERENCES 139

[56] Baikov, A.Y., Grushina, O.A. & Strikhanov, M.N. Simulation of conditions for themaximal efficiency of decimeter-wave klystrons. Tech. Phys. 59, 421–427 (2014).https://doi.org/10.1134/S1063784214030037

[57] A. Y. Bajkov and D. M. Petrov, "Problems of creation of powerful and super-powerklystrons with efficiency up to 90%," Proceedings International University Conference’Electronics and Radiophysics of Ultra-High Frequencies’ (Cat. No.99EX356), SaintPetersburg, Russia, 1999, pp. 5-8, doi: 10.1109/UHF.1999.787864.

[58] A. Gilmour, Klystrons, traveling wave tubes, magnetrons, crossed-field amplifiers, andgyrotrons, Artech House, 2011.

[59] J. Benford, J.A. Swegle and E. Schamiloglu, High power microwaves, CRC press,2015.

[60] W. Brown, J. Skowron, G. MacMaster and J. Buckley, "The super power CW amplitron,"in 1963 International Electron Devices Meeting, 1963, pp. 52-52.

[61] Crossed-Field Amplifiers. (n.d.). Microwave and RF Vacuum Electronic Power Sources,629–658. doi:10.1017/9780511979231.017

[62] I. Tahir, A. Dexter and R. Carter, "Noise Performance of Frequency- and Phase-LockedMagnetrons Operated as Current-Controlled Oscillators, IEEE Transactions on ElectronDevices, vol.52, No.9, pp.2096–2103, 2005.

[63] T. Mitani, N. Shinohara, and H. Matsumoto, Development of a Pulse-Driven Phase-Controlled Magnetron, Proc. 8th IEEE International Vacuum Electronics ConferenceIVEC 2007, pp.425–426, 2007.

[64] C. Liu, H. Huang, Z. Liu, F. Huo, and K. Huang, Experimental Study on MicrowavePower Combining Based on Injection-Locked 15-kW S-Band Continuous-Wave Mag-netrons, IEEE Transactions on Plasma Science, vol.44, no.8, pp.1291–1297, August2016.

[65] H. Matsumoto, K. Hashimoto, N. Shinohara, and T. Mitani, Experimental equipmentsfor microwave power transmission in Kyoto University, Proc. 4th International Confer-ence on Solar Power from Space (SPS 04), Spain, pp.131–138, June 2004.

[66] N. Shinohara and H. Matsumoto, Phased Array Technology with Phase and AmplitudeControlled Magnetron for Microwave Power Transmission, Proc. 4th InternationalConference on Solar Power from Space (SPS 04), Spain, pp.117–124, June 2004.

[67] N. Shinohara, T. Mitani, and H. Matsumoto, Development of phase and amplitudecontrolled magnetron, Proc. 6th International Vacuum Electronics Conference IVEC2005, pp.61–65, Netherlands, April 2005.

[68] A. Nagahama, T. Mitani, N. Shinohara, N. Tsuji, K. Fukuda, Y. Kanari and K. Yonemoto,Study on a Microwave Power Transmi tting System for Mars Observation Airplane,IEEE MTT-S International Microwave Workshop Series on Innovative Wireless PowerTransmission: Technologies, Systems, and Applications (IMWS-IWPT), IWPT3-5,Proceedings, pp.63–66, Uji, Japan, May 2011.

140 REFERENCES

[69] B. Zhang, W. Jiang, Y. Yang, C. Yu, K. Huang and C. Liu, Experimental study on anS-band near-field microwave magnetron power transmission system on hundred-wattlevel, International Journal of Electronics, vol.102, No.11, pp.1818–1830, 2015.

[70] I. Tahir, A. Dexter and R. Carter, Frequency and Phase Modulation Performance ofan Injection-Locked CW Magnetron, IEEE Transactions on Electron Devices, vol.53,No.7, pp.1721–1729, July 2006.

[71] R. Adler, A Study of Locking Phenomena in Oscillators, Proceedings of I.R.E andWaves and Electrons, vol.34, pp.351–357, 1946.

[72] N. Shinohara, H. Matsumoto and K. Hashimoto, Solar Power Station/Satellite (SPS)with Phase Controlled Magnetrons, IEICE Trans.Electron, vol.E86-C, No.8, pp.1550–1555, August 2003.

[73] N. Tesla, "Transmission of energy without wires," Sci Am (Suppl), vol. 57, pp. 237-2601904.

[74] P. E. Glaser, "Power from the sun: Its future," Science, vol. 162, no.3856, pp. 857–861,Nov. 1968, DOI: 10.1126/science.162.3856.857

[75] W. C. Brown, T. Moreno, "Microwave power generation", IEEE Spectrum vol. 1, Issue:10, pp.77-81, Oct. 1964, DOI: 10.1109/MSPEC.1964.6501192

[76] W. C. Brown, "The History of the Development of the Rectenna", Johnson Space CenterSolar Power Satellite Microwave Transmission and Reception in NASA, pp.271-280,1980

[77] W. C. Brown, "The History of Power Transmission by Radio Waves," IEEE Transactionson Microwave Theory and Techniques, vol.32, Issue: 9, pp.1230-1242.Sep 1984, DOI:10.1109/TMTT.1984.1132833

[78] Dickinson R. M, "Evaluation of a microwave high-power reception-conversion array forwireless power transmission," NASA STI/Recon Technical Report NASA-CR-145625,1975

[79] N. Shinohara, "Beam Control Technologies with a High-Efficiency Phased Array forMicrowave Power Transmission in Japan," Proceedings of the IEEE, vol. 101, no. 6, pp.1448-1463, June 2013, DOI: 10.1109/JPROC.2013.2253062

[80] N. Kaya, H. Matsumoto, R. Akiba, "Rocket experiment METS microwave energytransmission in space," Space Power, vol. 11, no. 1&2, pp. 267-274, 1993.

[81] JAXA, "Ground demonstration testing of microwave wireless power transmission,"[Online]. Available: http://www.kenkai.jaxa.jp/eng/research/ssps/150301.html [Ac-cessed: 25- Feb- 2020]

[82] H. Matsumoto, K. Hashimoto, N. Shinohara, and T. Mitani, "Experimental equipmentfor microwave power transmission in Kyoto University," in Proc. 4th InternationalConference on Solar Power from Space (SPS ’04), pp.131–138, Spain, 2004.

REFERENCES 141

[83] B. Zhang, W. Jiang, Y. Yang, et al. "Experimental study on an S-band near-fieldmicrowave magnetron power transmission system on hundred-watt level". Interna-tional Journal of Electronics, vol. 102, no. 11, pp. 1818-1830, Jan 2015, DOI:10.1080/00207217.2014.996784

[84] C. Chen, K. Huang, Y. Yang. "Microwave transmitting system based on four-waymaster–slave injection-locked magnetrons and horn arrays with suppressed sidelobes",IEEE Transactions on Microwave Theory and Techniques, vol. 66, no. 5, pp. 2416-2424,May 2018, DOI: 10.1109/TMTT.2018.2790924

[85] B. Yang, T. Mitani, N. Shinohara, "Experimental Study on a 5.8 GHz Power-VariablePhase-Controlled Magnetron", IEICE Transactions on Electronics, Vol.E100-C, No. 10,pp. 901-907, Oct. 2017, DOI: 10.1587/transele.E100.C.901

[86] R. Adler, "A Study of Locking Phenomena in Oscillators," Proceedings of I.R.Eand Waves and Electrons, Vol.34, pp. 351–357, 1946, 10. DOI: 10.1109/434 JR-PROC.1946.229930.

[87] J. Benford, "Space Applications of High-Power Microwaves," in IEEE Transactions onPlasma Science, vol. 36, no. 3, pp. 569-581, June 2008. DOI: 10.1109/TPS.2008.923760

[88] "Gaussian Beam Optics." CVI Laser Optics, IDEX Optics & Photonics.

[89] Bandres, Miguel A., and Julio C. Gutiérrez-Vega. "Ince–gaussian beams." Optics letters29.2 (2004): 144-146.

[90] Yang, Yang, et al. "A 5.8 GHz circularly polarized rectenna with harmonic suppres-sion and rectenna array for wireless power transfer." IEEE Antennas and Wirelesspropagation letters 17.7 (2018): 1276-1280. DOI:10.1109/lawp.2018.2842105

[91] Yang, Yang, et al. "A circularly polarized rectenna array based on substrate integratedwaveguide structure with harmonic suppression." IEEE Antennas and Wireless Propa-gation Letters 17.4 (2018): 684-688. DOI:10.1109/lawp.2018.2811860

[92] I. Tahir, A. Dexter and R. Carter, "Noise Performance of Frequency-and Phase-LockedMagnetrons Operated as Current-Controlled Oscillators," IEEE Transactions on Elec-tron Devices, Vol.52, No.9, pp. 2096–2103, 2005, DOI: 10.1109/TED.2005.854276.

[93] T. Mitani, N. Shinohara, and H. Matsumoto, "Development of a Pulse-Driven Phase-Controlled Magnetron," Proc. 8th IEEE International Vacuum Electronics ConferenceIVEC 2007, pp. 425–426, 2007.

[94] H. Matsumoto, K. Hashimoto, N. Shinohara, and T. Mitani, "Experimental equip-ment for microwave power transmission in Kyoto University," Proc. 4th InternationalConference on Solar Power from Space (SPS ’04), Spain, pp. 131–138, June 2004.

[95] N. Shinohara and H. Matsumoto, "Phased Array Technology with Phase and AmplitudeControlled Magnetron for Microwave Power Transmission," Proc. 4th InternationalConference on Solar Power from Space (SPS ’04), Spain, pp. 117–124, June 2004.

142 REFERENCES

[96] N. Shinohara, T. Mitani, and H. Matsumoto, "Development of phase and amplitudecontrolled magnetron," Proc. 6th International Vacuum Electronics Conference IVEC2005, pp. 61–65, Netherlands, April 2005.

[97] C. Liu, H. Huang, Z. Liu, F. Huo, and K. Huang, "Experimental Study on MicrowavePower Combining Based on Injection-Locked 15-kW S-Band Continuous-Wave Mag-netrons," IEEE Transactions on Plasma Science, Vol.44, no. 8, pp. 1291–1297, August2016, DOI: 10.1109/TPS.2016.2565564.

[98] I. Tahir, A. Dexter and R. Carter, "Frequency and Phase Modulation Performance ofan Injection-Locked CW Magnetron," IEEE Transactions on Electron Devices, Vol.53,No.7, pp. 1721–1729, 2006, DOI: 10.1109/TED.2006.876268.

[99] B. Yang, T. Mitani, N. Shinohara, "Experimental Study on a 5.8 GHz Power-VariablePhase-Controlled Magnetron," IEICE Transactions on Electronics, Vol.E100-C, No.10,pp. 901-907, Oct. 2017, DOI: 10.1587/transele.E100.C.901.

[100] Y. Kobayshi, Y. Yoshida, Z.-I. Yamamoto, and S. Kawasaki, "S-band GaNon Si based 1KW-class SSPA system for space wireless applications", IEICETransactions on Electronics, Vol.E96-C, No.10, pp. 1140-1146, Oct. 2016, DOI:10.1587/transele.E96.C.1245.

[101] N. Hasegawa, N. Shinohara, S. Kawasaki, "A 7.1 GHz 170W Solid-State Power Ampli-fier with 20-Way Combiner for Space Applications", IEICE Transactions on Electronics,Vol.E99-C, No.10, pp. 1245-1253, Oct. 2016, DOI: 10.1587/transele.E99.C.1140.

[102] T. Yamasaki, Y. Kittaka, H. Minamide, K. Yamauchi, S. Miwa, S. Goto, M. Nakayama,M. Kono, and N. Yoshida, "A 68% efficiency C-band 100W GaN power amplifier forspace applications," IEEE MTT-S Int. Conf., Anaheim, CA, pp. 1384–1387, May 2010.

[103] I. Takenaka, K. Ishikura, K. Asano, S. Takahashi, Y. Murase, Y. Ando, H. Takahashi,and C. Sasaoka, "High-efficiency and high power microwave amplifier using GaN-on-SiFET with improved high-temperature operation characteristics," IEEE Transactionson Microwave Theory and Techniques, Vol.62, no.3, pp. 502–512, Jan. 2014, DOI:10.1109/TMTT.2014.2298381.

[104] N.-S. Cheng, P. Jia, D.B. Rensch, and R.A. York, "A 120-W X-band spatially combinedsolid-state amplifier," IEEE Transactions on Microwave Theory and Techniques, Vol.47,no. 12, pp. 2557–2561, Dec. 1999, DOI: 10.1109/22.809006.

[105] R. Adler, "A Study of Locking Phenomena in Oscillators," Proceedings of I.R.E andWaves and Electrons, Vol.34, pp. 351–357, 1946, DOI: 10.1109/JRPROC.1946.229930.

[106] W. C. Brown, "The high signal to noise ratio of the microwave oven magnetron andevidence of a negative feedback loop to control it," Proc.1st Int. Workshop Cross-FieldDevices, Aug. 1995, pp. 178-187.

[107] W. C. Brown, "Update on the Solar Power Satellite transmitter design," Space Power(ISSN 0883-6272), Vol. 6, 1986, pp. 123-135.

REFERENCES 143

[108] T. Mitani, N. Shinohara, H. Matsumoto, "Development of a Pulse-Driven Phase-Controlled Magnetron", Proc. International Vacuum Electronics Conference IVEC2007, Kitakyushu, 2007.

[109] S. Schneider, F. Hegger, N. Hochgeschwender, R. Dwiputra, A. Mori-arty, J. Berghofer,G. K. Kraetzschmar, "Design and development of a benchmarking testbed for theFactory of the Future", 2015 IEEE 20th Conference on Emerging Technologies &Factory Automation (ETFA), pp. 1-7, 2015.

[110] N. Shinohara, H. Matsumoto, K. Hashimoto, "Phase-controlled magnetron develop-ment for SPORTS: Space power radio transmission system," Radio Science Bulletin,vol. 2004, No. 310, pp. 29-35, 2004

[111] C. Liu, H. Huang, Z. Liu, F. Huo, and K. Huang, "Experimental Study on MicrowavePower Combining Based on Injection-Locked 15-kW S-Band Continuous-Wave Mag-netrons," IEEE Transactions on Plasma Science, vol.44, No. 8, pp.1291–1297, August2016.

[112] I. Tahir, A. Dexter and R. Carter, "Frequency and Phase Modulation Performance ofan Injection-Locked CW Magnetron," IEEE Transactions on Electron Devices, vol. 53,No. 7, pp. 1721-1729, 2006

[113] B. Yang, T. Mitani, N. Shinohara, "Experimental study on frequency modulation of aninjection-locked magnetron based on full wave voltage doubler," IVEC2018, Monterey,CA, USA, 2018

[114] J.Kojima, N.Shinohara, T.Mitani, T.Hashimoto, N.Kishi, H.Toyomura and A.Okazaki,"Development of highly efficient microwave wireless charging system for electric vehi-cle," Technical Report of IEICE Information and Communication Engineers SPS2007-16 (2008-03)

Publication list

Journal papers1. Bo Yang, Tomohiko Mitani, and Naoki Shinohara, “Experimental Study on a 5.8 GHz

Power-Variable Phase-Controlled Magnetron”, IEICE Trans. Electron, Vol.E100-C,No.10, pp.901-907, 2017, doi: 10.1587/transele.E100.C.901.

2. Bo Yang, Tomohiko Mitani, and Naoki Shinohara, “Evaluation of the ModulationPerformance of Injection-Locked Continuous-Wave Magnetrons”. IEEE Trans. ED,vol.66, no.1, pp.709-715, 2019, doi: 10.1109/TED.2018.2877204.

3. Ce Wang, Bo Yang, Seishiro Kojima, and Naoki Shinohara, “The Application ofGHz Band Charge Pump Rectifier and Rectenna Array for Satellite Internal Wire-less System”, Cambridge J. Wireless Power Transfer, pp.190-195, 2019.11, doi:10.1017/wpt.2019.13.

4. Bo Yang, Xiaojie Chen, Jie Chu, Tomohiko Mitani, and Naoki Shinohara, “A 5.8GHz Phased Array System Using Power-Variable Phase-Controlled Magnetrons forWireless Power Transfer”. IEEE Trans. MTT, vol.68, no.11, 2020, doi:10.1109/TMTT.2020.3007187.

5. Xiaojie Chen, Bo Yang, Naoki Shinohara, and Changjun Li, “Modeling and Experi-ments of an Injection-locked Magnetron with Various Load Reflection Levels”, IEEETrans. ED, vol.67, no.9, pp.3802-3808, 2020.9, doi: 10.1109/TED.2020.3009901.

6. Xiaojie Chen, Bo Yang, Naoki Shinohara, and Changjun Li, “A High-efficiency Mi-crowave Power Combining System based on Frequency-tuning Injection-LockedMagnetrons”, IEEE Trans. ED,vol.67, no.10, pp.4447-4452, 2020, doi: 10.1109/TED.2020.3013510.

International conference1. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “Study on a 5.8GHz Power-Variable

Phase-Controlled Magnetron for wireless power transfer”, 2016 Asian Wireless PowerTransfer Workshop (AWPT), Chengdu, 2016.12.16-18.

2. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “Development of a 5.8 GHz Power-Variable Phase-Controlled Magnetron”, 18th International Vacuum Electronics Confer-ence (IVEC), London, 2017.4.24-26.

146 PUBLICATION LIST

3. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “Study on Phase Controlled Mag-netron for Phase Modulation”, 2017 Thailand-Japan Microwave (TJMW), Bangkok,2017.6.14-16.

4. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “Experimental Study on FrequencyModulation of an Injection-Locked Magnetron Based on Full Wave Voltage Doubler”,19 th International Vacuum Electronics Conference (IVEC), Monterey, 2018.4.24-26.

5. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “Study on a 5.8 GHz Injection-lockedMagnetron for Transferring Data”, 31st International Vacuum Nanoelectronics Confer-ence (IVNC), Kyoto, 2018.7.9-13.

6. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “Injection-Locked CW Magnetron for awirelessly-powered TV”, 20 th International Vacuum Electronics Conference (IVEC),Busan, 2019.4.30-5.1.(keynote)

7. Bo Yang, Tomohiko Mitani, Naoki Shinohara, “A 5.8GHz Magnetron Phased ArraySystem”, 2019 Asia Wireless Power Transfer Workshop (AWPT), Xian, 2019.10.31-11.2.

Awards1. IEEE MTT-S Kansai Chapter WTC (Wakate Technical Committee) Presentation Award,

for “Study on a 5.8GHz Power-Variable Phase-Controlled Magnetron”, 9th KansaiMicrowave Meeting for Young Engineers, 2016.7.2.

2. 2016 Asia Wireless Power Transfer Workshop Student Paper Competition First Prize,for Bo Yang, Tomohiko Mitani, and Naoki Shinohara, “Study on a 5.8GHz Power-Variable Phase-Controlled Magnetron for Wireless Power Transfer”, 2016.12.16-18.

3. 2017 Thailand-Japan Microwave Presentation Encouragement Award, for Bo Yang,Tomohiko Mitani, and Naoki Shinohara, “Study on Phase Controlled Magnetron forPhase Modulation”, 2017.6.14-17.

4. 2017 Thailand-Japan Microwave Student Design Competition Winner, for KoutaOkazaki, Takashi Hirakawa, Bo Yang, 2017.6.14-17.

5. Best MHz Rectenna Award for Ce Wang, Seshiro Kojima, Bo Yang, Daichi Nishio,and Kono Tomoki, IEICE Wireless Power Transfer Technical Committee RectennaContest (in Japanese), 2017.9.12.

6. 2017 Best Student Presantation Award of IEICE Elecgtron Device Technical Com-mittee, for Bo Yang, Tomohiko Mitania, and Naoki Shinohara, “Study on a Phasecontrolled Magnetron based on Full Wave Voltage Doubler (in Japanese)”, 2017.10.26-27.

7. 2019 Asia Wireless Power Transfer Workshop Best Student Award, for Bo Yang,Tomohiko Mitani, and Naoki Shinohara, “A 5.8GHz Magnetron Phased Array System”,2019.10.31-11.2.

PUBLICATION LIST 147

8. 2019 Asia Wireless Power Transfer Workshop WiPoT Award, for Bo Yang, Tomo-hiko Mitani, and Naoki Shinohara, “A 5.8GHz Magnetron Phased Array System”,2019.10.31-11.2.

Patent1. Bo Yang, Tomohiko Mitani, Naoki Shinohara, Microwave Transmission Device, Japan

Patent 2019-80147.

Fellowship1. Japan Society for the Promotion of Science (JSPS) Research Fellowship for Young

Scientists (DC2), 19J12459.

Title High Power Microwave Wireless Power Transmission ...

Documents

Transcript of Title High Power Microwave Wireless Power Transmission ...