Mitigation of The Influence of Parasitic Elements in Wide-Bandgap Power Converters

Abstract

This work undertakes a detailed investigation of the Mitigation of The Influence of Parasitic Elements in Wide-Bandgap Power Converters. It primarily considers techniques applied to Silicon Carbide (SiC) Metal-Oxide-Semiconductor Field-Effect Transistors (MOSFETs) in low voltage (<600V) power converters. It is shown that parasitic circuit elements have a broad range of negative impacts on the performance of Wide Band Gap (wide band gap) power converters. Analysis of an experimental set-up is performed to create an accurate simulation model which is then validated against experimental results. A parametric analysis of the parasitic Printed Circuit Board (PCB) elements is performed highlighting the sensitivity of the converter’s performance to these. State-of-the-art techniques for mitigating the influence of the parasitic elements are reviewed and a method for incorporating a small inductor into the converter output for mitigation of the load parasitics is proposed. Considerable analysis is dedicated to the identification of the temporal source of frequency domain characteristics of power converter waveforms. The technique of successive differentials is demonstrated as a tool for analysing time domain waveforms and identifying key noise generating features. The largest portion of this work proposes smoothed waveform transitions as an important technique in the mitigation of the influence of the parasitic elements of the converter. They are first defined, and a range of analytical tools considered. Following this they are shown to be able to provide significant performance improvements in terms of the interactions with parasitic elements of the converter. Finally, methods for the practical realisation of waveform smoothing are explored, showing that the predicted performance improvements can be realised. Additional work is performed on the creation of gate drive systems with low complexity that realise the suggested performance improvements.

Date of Award	19 Mar 2019
Original language	English
Awarding Institution	University of Bristol
Supervisor	Xibo Yuan (Supervisor)