XDXu Du

Xu Du

*This is an algorithmically generated author record based on publicly available information.

Xu Du has a total of 6 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Evaluation and Establishment of an Empirical Model for the Dynamic Ron Effect of GaN Power Device in Hard Switching Mode

    by Shaoyu Sun, Xu Du, Ling Xia, Wengang Wu

    In recent years, the rapid popularization of GaN HEMT devices in power applications has made the traditional GaN static SPICE model unable to meet the requirements of high-voltage and high-frequency circuit design, due to the dynamic ON-resistance ( R $_{\text{on}}$ ) effect. In this article, we evaluate the dynamic R $_{\text{on}}$ effect of a commercial 100-V Schottky-type p-GaN HEMT in the hard switching mode and propose an empirical dynamic model. The pulse test results show that the gate-to-drain voltage stress is the main cause of the dynamic R $_{\text{on}}$ effect. The peak electric field located at the AlGaN barrier layer and channel of the device will cause threshold voltage positive shift and hot electron effect. These two physical mechanisms that cause the dynamic R $_{\text{on}}$ effect are verified by the TCAD simulation. Meanwhile, we calculated the contribution of the two physical mechanisms to the dynamic R $_{\text{on}}$ under different operation conditions. Based on the test and analysis results, the equivalent circuit is added to the static SPICE model of the device. The proposed empirical model could well simulate the dynamic R $_{\text{on}}$ characteristic and switching behavior of the device in hard switching mode.


  • Third-Quadrant Conduction Induced Dynamic Ron Effect in GaN Gate-Injection-Transistor

    by Shaoyu Sun, Xu Du, Ling Xia, Qi Qin, Wengang Wu

    The deterioration of dynamic ON-resistance ( R $_{\text{on}}$ ) and the mechanism of commercial GaN-based gate injection transistors in the third and the first quadrants is studied. In our test condition, the device exhibits an obvious dynamic R $_{\text{on}}$ effect in the half-bridge test circuit under high current conditions that behaves well in the hard-switching test circuit. According to the stress-to-measurement test, we consider that the origin of the dynamic R $_{\text{on}}$ in the half-bridge test circuit is related to the semi-ON-state of the device in the third-quadrant conduction. In this state, a peak of the electric field is located at the source edge of the gate, which will cause the traps located in the buffer or the surface to capture electrons. In the ON-state, the degree of suppression of dynamic R $_{\text{on}}$ by gate injection current ( I $_{\text{G}}$ ) is related to the device’s operating quadrants. Based on the TCAD simulation, we consider that this is because the drain voltage of the device affects the distribution of gate injection holes under different operating quadrants.


  • Reverse Current Stress Induced Dynamic Ron of GaN HEMTs in Soft-Switching Mode

    by Shaoyu Sun, Jiaoqi Zhang, Xu Du, Ling Xia, Wengang Wu

    The utilization of gallium nitride (GaN) in power electronics has been validated for mass production through its successful implementation in power charger applications. Eliminating technical roadblocks is crucial for the widespread adoption of GaN, and one such roadblock is the dynamic ON-resistance (Ron) effect that can lead to energy losses and reliability issues. While this effect has been improved in successive generations of commercial GaN devices, this letter argues that it still warrants attention at present. In particular, in soft-switching half-bridge circuits, certain operations can induce the dynamic Ron effect on devices that behaves well in hard-switching circuits. In our test conditions, the dynamic Ron effect is more likely to occur under high reverse current stress in soft switching mode. We believe that the dynamic Roneffect in the device is related to the presence of traps in the GaN buffer, and we use TCAD simulation to explain this phenomenon. Our findings and postulations regarding the root cause of the problem could aid in the optimization of GaN power devices.