Short-Circuit Detection Methods for Wide- Bandgap Semiconductor Devices

Abstract

Wide-bandgap (WBG) semiconductors, typified by silicon carbide (SiC) and gallium nitride (GaN), have revolutionized power electronics due to their exceptional switching speed, thermal conductivity, and voltageblocking capabilities. These advantages enable the development of high-efficiency, miniaturized power converters critical for renewable energy integration, electric vehicles, and aerospace applications. Current short-circuit detection approaches for SiC/GaN wide bandgap (WBG) devices face three critical drawbacks. Single-sensor detection methods are limited by application scenarios and prone to interference. Multi-sensor fusion methods require complex implementation despite balancing speed and robustness. Additionally, the existing detection system lacks compatibility with soft-switching topologies in WBG power converters. This study focuses on short-circuit detection in wide bandgap (WBG) semiconductor devices. A hierarchical classification system for short-circuit detection methods is established, and the transient characteristics of WBG device shortcircuit faults are thoroughly analyzed. The optimization effect of electromagnetic modeling on detection circuits is quantitatively evaluated. A multi-sensor fusion detection strategy balancing speed and robustness is proposed, and application-oriented selection principles for detection methods are clarified. This paper systematically reviews state-of-the-art short-circuit detection methodologies tailored for SiC MOSFETs and GaN HEMTs, categorizing them into single-sensor approaches and multi-sensor fusion strategies (e.g., di/dt + Vgs hybrid method). Core performance metrics—detection speed (ranging from sub-100 ns to over 400 ns), sensitivity to fault signatures, and robustness against electromagnetic interference —are comprehensively analyzed. Critical challenges, including parasitic inductance/capacitance interference, false triggering caused by switching transients, and compatibility with fast-switching dynamics, are discussed in detail. Additionally, the integration of electromagnetic modeling tools (e.g., HFSS, Q3D) for optimizing detection circuit layouts is evaluated, with studies confirming that precise modeling of >100 MHz parasitics is essential for ensuring reliable detection. Finally, current research gaps and future directions are outlined, providing a valuable academic reference for advancing WBG power system protection technologies.