Communication Channel Measurement and Modeling for Inductive Wireless Power Transfer

Abstract: Wireless Power Transfer (WPT) technology has experienced phenomenal success in the application of wireless charging of smartphones. With the Qi Specification becoming the dominant standard, wireless charging's convenience goes beyond cable-free charging to facilitate interoperability between chargers and devices from different vendors. For wireless charging to operate correctly, the power transmitter (charger) and the receiver (user device) should communicate with each other. In the current release of the Qi Specification (Rel 1.2.4), the communication is performed in the same frequency band as the power transfer (in-band communication). Therefore, the properties of the in-band channel determine the quality of the communication link. However, since WPT has only become hugely popular in recent years, the in-band channel has not yet attracted a lot of attention. In this thesis, the focus is to measure and characterize the in-band channel for scenarios of interest to wireless charging. To enable the channel measurement, a custom-designed setup was made, allowing simultaneous measurements of currents and voltages in the transmitter and receiver. Using this setup, the channel has been characterized as a two-port network using Z-parameters. Observations regarding the Z-parameters, as well as the power transfer efficiency, in various situations are presented and discussed. The first harmonic approximation (FHA) based channel model was applied and extended for the analysis of the results. Major findings of the thesis include: 1) the amplitude of the imaginary part of the Z-parameters is larger than the real part by one to two orders of magnitude in the frequency of interest (80 to 500 kHz), setting a high accuracy requirement of the measurement system, 2) Power transfer ratio behavior is dependent heavily on WPT circuit parameters interacting with the channel parameters, pointing to opportunities for optimization, 3) The effects of ferrites and foreign objects can be included by extending the simple FHA based channel model with additional coil(s) to model eddy current induced inductance change and resistive losses.