Search

The human-body-model (HBM) electrostatic discharge (ESD) robustness can be directly evaluated through the ESD simulator designed for product qualification or indirectly by the transmission line pulse (TLP) I–V system. Although the TLP test can be used to estimate drain-to-source HBM ESD robustness of Si mosfets, it causes misleading estimation for GaN HEMTs. To understand this miscorrelation, the behaviors of 100-V Schottky-type p-GaN gate HEMTs and Si mosfets under HBM ESD tests are characterized and compared in this work, especially the rise time (tR) of terminal voltages in the devices. In drain-to-source ESD of the Si mosfet and gate-side ESD of the GaN HEMT, the high leakage current and the device voltage clamping result in short tR (∼10 ns) and strong ESD robustness. In most other HBM ESD tests, the two devices behave like variable capacitors with long tR (>100 ns) due to the insignificant leakage current. Moreover, the breakdown voltage of GaN HEMTs is dependent on the voltage stress time. Since the appropriate HBM–TLP correlation is based on the assumption of short tR and weak time dependence of device breakdown, the TLP I–V test is not suitable for HBM ESD robustness evaluation of GaN HEMTs. Instead, a more accurate and simpler estimation of HBM ESD robustness can be realized by a versatile pulse I–V system with adjustable tR.