Sub-nanosecond point-plane gas breakdown

Gas Breakdown in the sub-nanosecond regime is of interest for plasma limiters, ultra-wideband-RADAR sources, and switching in high-speed pulsed-power applications. Traditionally, gas breakdown is studied using conventional spark-gaps with point-plane geometry. This paper covers modifications done to a conventional spark-gap to increase the gap's performance. It also covers gas breakdown of the modified point-plane spark-gap, at different pressures of argon and a point radius less than 0.5 microns (µm). Both polarities are also included in the experimentation.

All experiments used a coaxial-transmission-line spark-gap to investigate gas breakdown in the sub-nanosecond regime. Previous experiments used a conventional coaxial-transmission-line spark-gap for gathering data. The conventional gap, with point-plane geometry, is integrated into the inner-conductor of a coaxial transmission-line and suffers from major impedance mismatches as well as lumped capacitances. Pulses produced for previous experiments had a 600 picosecond (ps) risetime and the reflected pulse-risetime was smeared to 1.37 nanoseconds (ns) by the conventional spark-gap. To improve system performance, a conical-section spark-gap was implemented. This new conical spark-gap only lengthened the incoming 500 ps pulse risetime to 760ps.

Gas breakdown in the sub-nanosecond regime was done with point-plane geometry. Breakdown times were tested in multiple pressures of argon of 0.5, 5, 10, 25, 50, 100, 150, 300, and 600 torr. Results with the new conical spark-gap show that gas breakdown happens faster with increased pressure and has a breakdown-time limit at 300 torr.

Finally, this paper contains a brief description of corona-discharge experiments in point-plane geometry of the conventional spark-gap. The point radius used was 10 µm with a point-plane distance of 0.5, 1.0, 1.5, 2.0, and 2.5 cm and larger corona discharges were observed for the shorter gaps. All of the experiments were done in dry air at atmospheric pressure. However, the current from the discharges was too small to record. No differences were noticed between single shots and 1 kHz rep-rated shots.