Behavioural SiC IGBT Modelling Using Non-Linear Voltage and Current Dependent Capacitances
Date Issued
January 1, 2023
DOI
10.1109/DMC58182.2023.10412584
Abstract
This paper presents a behavioural silicon carbide (SiC) IGBT model that utilizes voltage and current dependent capacitances to simulate its switching characteristics, and a voltage dependent current source to simulate the static characteristics. The non-linear capacitances are extracted from dynamic Current-Voltage (IV) measurements, eliminating the need for non-standard Capacitance-Voltage (C-V) characterization methods under high voltage and high current. The accuracy of the compact model is compared with previously validated numerical Technology Computer Aided Design (TCAD) simulation results across a wide range of operational conditions. The model performance is demonstrated by accurately predicting the unique characteristics of a 27kV SiC IGBT, including dV/dt, dI/dt and losses, while significantly reducing the simulation time by 4-5 orders of magnitude. Additionally, the model convergence is tested using a buck converter topology with non-ideal parasitic elements.