Condition-Based Maintenance Algorithms for Power Converters Using Wide-Bandgap Devices
Patrick Wang with adviser A. Banerjee
This research project focuses on the prediction of MOSFET failures of silicon-based power electronics building blocks and is a collaboration with Dr. John Donnal at the United States Naval Academy. The research can be broken down to two major parts: Design and construction of a SiC-based resonant power converter and design of a signature-recognizing circuitry to detect aging in SiC MOSFETs. Shown in Figure 18 is the sensing circuit which estimates the leakage current of the SiC MOSFET and determines the condition of the device under test.
Figure 19 is the most up-to-date PCB which has the sensing circuit integrated into the converter circuit. This prototype aims to demonstrate that the proposed in-situ monitoring method works independently of the converter operating conditions (i.e., dc voltage, load current, duty ratio) and reliably monitors the devices during real-time operation. This work was funded by the Office of Naval Research MVDC Risk Reduction program under Grant Number N0001418WX01706.