Switch Fault Diagnosis in Single Switch DC-DC Converters

In many applications, the dependability of the DC-DC converter's operation is essential. Thus, fault-tolerant converters are desired. Prior to fault tolerance, fast fault diagnosis is required. To solve this problem, this paper proposes a simple and new fault signature-based detection technique that can identify open-circuit and short-circuit switch faults in DC-DC converters. The suggested method for fault diagnosis is based on signal processing. It is demonstrated here that diode current contains significant information to identify switch faults. Fault detection is achieved by processing diode current as well as gate switching state information. The switch fault diagnosis technique utilizes the same sensors used for control, obviating the need for additional sensors. The fault detection is obtained in less than one switching time period. The obtained simulation and hardware results validate the suggested technique.