A Self-Consistent Approach Based on Bayesian Deconvolution for Trapping Time Constant Analysis: A Demonstration to Analyze ΔV_THTransients in p-GaN Gate Power HEMTs

Characterizing the trapping time constant is crucial to understanding the root cause of the instability. The conventional approach to analyzing the trap-related transient characteristics mostly utilizes the derivative method to extrapolate the trapping time constant, which may have information loss due to an unobvious trapping response. This article presents a novel Bayesian deconvolution methodology to accurately extract and capture time constants. The proposed technique is applied to determine trapping time constants and activation energies in p-GaN gate power high electron mobility transistors (HEMTs) under positive gate bias stress. The proposed method is versatile and suitable for a broad spectrum of electronic devices experiencing the trapping-related transient phenomena.