Fast electrical and optical diagnostics for the early phase of dielectric surface flashover

Virtually all concepts for the use of electrical energy at high voltages utilize solid dielectrics as electrical insulators, for which dielectric surface flashover is the primary factor limiting the magnitude of the applicable voltage. The voltage limitations imposed by dielectric surface flashover are particularly onerous because the physical mechanisms leading to flashover are far from being fully understood. With a new experimental setup, emphasizing high speed and high resolution electrical and optical diagnostic methods, we attempt to elucidate the physical mechanisms leading to surface flashover. The setup consists of a cable discharge, operating at up to 50 kV DC, or 120 kV pulsed via a triggered spark gap, and is connected to a micro-torr vacuum discharge gap. Currently, the system is operated in the de self-breakdown mode. The geometry of all interconnecting lines and of the discharge chamber is coaxial and the impedances are closely matched to the charging line. Measurement of. pre-flashover currents, in the range of milliamperes to amperes, is accomplished via a transmission line type current sensor with a sensitivity of 100 m VI A and a risetime of less than 1 ns. Current sensor output during flashover is limited to the dynamic range of a standard oscilloscope using a switching-diode voltage clipping technique, with negligible pulse shape distortion. Voltage measurements across the flashover gap are performed with a standard capacitive voltage divider. High amplification photomultiplier tubes measure pre-flashover self luminosity and provide spatial resolution of optical phenomena. Soft X-ray emission is measured and correlated to current and luminosity signals. Different materials (Pyrex, PMMA, Lexan, alumina, and boron-nitride) have been investigated. In addition, de magnetic fields of 100 mT have been applied, using permanent magnets outside of the vacuum chamber. The results show the existence of a linearly increasing current and a corresponding linearly increasing low amplitude luminosity before flashover occurs, which can be explained by the standard model (saturated electron avalanche and electron stimulated outgassing). Externally applied magnetic fields alter these pre-flashover phenomena drastically. Magnetic field effects have been observed which increase or decrease the flashover voltage, depending on the direction of the ExB drift. The change of risetime of the pre-flashover current due to application of the magnetic field supports the saturated secondary electron avalanche model.