Fault diagnosis of a ZVS dc-dc converter based on dc-link current pulse shapes

Abstract

The dc-dc converter is a critical component in a hybrid electric vehicle since it supplies power to an electronic control unit, as well as chassis electric components such as power windows, wipers, etc. In this paper, a low-cost diagnostic method for MOSFET faults in a zero-voltage-switching dc-dc converter is proposed. The proposed method utilizes the dc-link current patterns as the signatures of faults of MOSFETs. A presignal processing circuit consists of a peak detector and an integrator circuit. The ratio of peak-to-integral values, which is similar to the crest factor, is useful for diagnosis.