EMI prediction based on datasheet parameters for hard-switched Si, SiC, and GaN MOSFETs

Abstract

Wide bandgap (WBG) semiconductors such as Silicon Carbide (SiC) and Gallium Nitride (GaN) enable improved power converter efficiency due to better material characteristics. However, their faster switching dynamics introduce electromagnetic interference (EMI) challenges requiring early-stage design consideration. Conventional EMI prediction approaches rely on trapezoidal waveform approximations that underestimate EMI and fail to capture nonlinear device characteristics, body diode reverse recovery, and circuit resonances. These limitations prevent accurate EMI characterization during the design phase, potentially leading to costly design iterations. The paper proposes a comprehensive analytical model for EMI prediction in hard-switched MOSFETs using only readily available datasheet information. This methodology, applicable to Silicon (Si), SiC, and GaN MOSFETs, employs segmented linear analysis with piecewise approximations to model transconductance and junction capacitances. Time domain switching waveforms calculated with this technique can be post-processed using Fast Fourier Transformation (FFT). This solution provides comprehensive EMI prediction and filter design capability during the early design phase, enabling designers to address EMI challenges before hardware implementation.