Reflectometry Based SOH Estimation Scheme for a SiC Buck Converter Having Branched Network

Abstract

A reflectometry based non-interfering method to identify the degradation of a buck converter built from SiC MOSFET has been proposed in this paper, and spread spectrum time domain reflectometry (SSTDR) has been applied between gate and source terminal to achieve the in-situ state of health (SOH) estimation. A controlled aging station was used to execute the power cycling method to age the SiC MOSFET. SSTDR reflections were recorded for both the healthy and the aged buck converter, and these recorded values were compared to determine the level of degradation. Through the experiment, it was apparent that a single measurement is sufficient to estimate the degradation level associated to the SiC MOSFET in a live buck converter having complex/branch network. The outcome of this research also proves that the SSTDR based technique can identify the aging of other components in the system. Since the gate terminal always stays at lower potential and easier to access, this technique brings an added advantage to the existing condition monitoring schemes.