A parametric SPICE model for the simulation of spark gap switches

Abstract

The use of SPICE-based software for the simulation of pulsed power systems—even large complex systems—has become commonplace in the pulsed power community. This is in contrast to earlier work in the field that relied on specially developed simulation codes such as Sandia’s Screamer or the Navy Research Lab’s Bertha, which natively incorporated models for common pulsed power components such as spark gap switches. Unlike these programs, SPICE programs provide a simple and familiar user interface and wide availability. However, SPICE programs do not include realistic models for key pulsed power circuit devices—including the spark gap switch. While simple switch models do exist in SPICE programs, these can only crudely approximate the behavior of a spark gap. This effort focuses on developing an SPICE circuit model for a gas-filled spark gap switch that is physically realistic while being simple enough to permit simulations to run in reasonable times on typical personal computers. Detailed information is provided for implementation in two common versions of SPICE: LTspice and Orcad PSPICE. Adaptation to other SPICE programs is possible with minimal modification. The model is intended as a design tool that uses physical parameters as inputs to connect it directly to the development of useable pulsed power systems. Data collected from the operation of a high-pressure pulsed-charged switch and a complete 12-stage Marx generator have been used to demonstrate the implementation and accuracy of the model over a wide range of parameters.

Development and Test of a 500-kV Compact Marx Generator Operating at 100-Hz PRF

“…difference between the electrodes exceeds the static breakdown voltage calculated by Paschen's law [25]. In our generator, the closing of an SGS from 2 to 13 occurs when an impulse voltage is applied, for which the breakdown value is higher than the dc value.…”

“…The circuit model of the SGS is based on the work reported in [25]. The adopted physical model predicts the behavior of an SGS using the values of the gas pressure and gap length as inputs.…”