瞬态(计算机编程)
拓扑(电路)
电感
电压
脉冲功率
物理
噪音(视频)
功率(物理)
发电机(电路理论)
脉冲发生器
电气工程
材料科学
计算机科学
极限(数学)
瞬态响应
电容器
香料
节点(物理)
脉搏(音乐)
脉冲宽度调制
电压尖峰
磁通量
电磁脉冲
马克思发生器
电流隔离
控制理论(社会学)
流离失所(心理学)
电阻抗
复合数
作者
Zihan Yuan,Yanyi Wang,Yixiang Wang,Xuandong Liu
摘要
All-solid-state Marx generators have shown broad application prospects in fields such as dielectric barrier discharge, biomedical engineering, and environmental treatment. When pursuing nanosecond-scale fast-front pulses, the extremely high voltage change rate (dv/dt) can readily induce Miller false turn-on in switching devices. Meanwhile, conventional isolated driving schemes with robust noise immunity typically rely on bulky independent power supplies, which severely limit the development of highly compact pulsed power systems. To address these challenges, this paper proposes a compact composite magnetic isolation drive circuit that eliminates the need for an independent high-voltage isolated power supply. The proposed topology integrates a passive negative-voltage bootstrap network with an active Miller clamp (AMC) branch. Through the establishment of a transient mathematical model and ablation experiments, the synergistic immunity mechanism of the composite drive is revealed: the steady-state negative bias generated by bootstrapping, together with the low-impedance discharge path provided by the AMC, physically blocks the path through which high-frequency displacement currents could otherwise elevate the gate potential. Quantitative parasitic parameter scanning shows that even with a severe gate stray inductance of 50 nH, the proposed scheme maintains the transient gate voltage within the safe operating area of the device. Based on the above investigation, a 20-stage SSMG experimental prototype was developed. Experimental results demonstrate that the system operates stably under a wide load range (2-20 kΩ) at a repetition frequency of 1 kHz, successfully outputting fast-front high-voltage pulse sequences with an amplitude of 20 kV and a rise time of 22.9 ns.
科研通智能强力驱动
Strongly Powered by AbleSci AI