Abstract
SiC MOSFET analytical models demonstrate higher computational efficiency and simpler implementation than physical/behavioral models. However, existing analytical models fail to properly account for Vgs-dependent Cgs/Cgd dynamics, causing significant errors in switching waveform predictions and switching characteristic parameters (dids/dt, dvds/dt and tdelay). Current SiC MOSFET analytical models implement Cgs and Cgd characterization through either lookup tables or parameterized mathematical expressions, where certain parameters require additional experimental data. This paper presents an accurate analytical model for SiC MOSFETs. This model provides physics-based expressions for Cgs(Vgs,Vds) and Cgd(Vgs,Vds), derived from charge-potential theory and simplified to capture the dominant capacitance dynamics in each switching Stage. Parameter extraction is performed exclusively using datasheet information. Experimental verification shows <10% loss prediction error, compared with 15-45% in existing analytical models. The model's universality has been validated using C3M0016120D, enabling converter optimization, dead-time control and snubber design.
Funding Information
  • National Natural Science Foundation of China (U24A20155)
  • science and technology innovation Program of Hunan Province (2024RC1036)
  • Yuelushan Center for Industrial Innovation (2025YCII0133)