Numerical modeling and analysis of power electronic devices

Hsien-Sheng Edward Hsing, Purdue University

Abstract

The purpose of the present study is to analyze recently developed power devices made of different materials with proposed empirical models in an advanced device simulator. In this thesis, a survey of all kinds of power device structures and operation mechanisms is given. The trade-off between performance requirements is also listed. To overcome material characteristic limitations, silicon carbide was seriously considered to replace silicon used in power devices due to its superior characteristics. Besides silicon carbide, other materials that have been proposed, e.g., diamond, gallium-arsenide, and gallium-nitride, are also described in this thesis. Based on current technology development and material properties, 6H-SiC is the most promising candidate for the power device because of its high energy bandgap, high thermal conductivity, and high carrier saturation velocity. Models based on experimental data for all kinds of 6H-SiC characteristics are proposed in this thesis. It is believed the models listed here are the most complete ones as compared with available information. To simulate power devices, the existing device simulator, ADEPT, has been modified to cope with concerns, i.e., extremely large device structure with huge region sizing difference, large input voltage step for high voltage application, self-heating effect, multi-dimensional case study, and so on. A data-recycling method and two-step iteration approach has been implemented and a heat conduction equation was included in ADEPT. By using the design of experiment (DOE) concept, the trade-off between driving and blocking capability inside 6H-SiC: UMOSFET power device is shown. The punch-through mechanism, instead of the impact ionization mechanism as in silicon, is the key role to decide the blocking capability in the 6H-SiC UMOS device. From simulation results we know that the channel length is the most significant factor to be adjusted in optimal device design.

Degree

Ph.D.

Advisors

Gray, Purdue University.

Subject Area

Electrical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS