Influence of Statistical Time Lags upon Electric Breakdown in Electronegative Gases

Abstract

Measurement of the electric strength of SF6 and a number of other electronegative gases has provided direct evidence that an unusually long statistical time lag is associated with the breakdown process in such gases under certain experimental conditions. For example, time lags in excess of one minute have been reported for the breakdown of SF6. The reason for the difference in breakdown behavior of such gases, when compared with the behavior of the more common gases, is not well understood. The intent in writing this paper was not merely to verify the fact that the time required for breakdown is completely random, although evidence of this plays an important role in the theories and equations developed here. The paper represents an attempt to provide an explanation of the role that these statistical time lags can often play in yielding misleading experimental values of the electric breakdown voltage, when measured under certain standard test conditions. A number of cases are discussed and evaluated by combining accurate experimental data with theories involving the occurance of random events. It is concluded that the ``apparent'' value of the electric breakdown voltage in electronegative gases ought to be strongly dependent upon the mean statistical time lag, the nature of voltage application, and the electrode configuration. The experimentally observed appearance of field emission from the cathode surface at voltages considerably in excess of the threshold value, can be used to derive theoretical relationships which indicate that rather significant ``apparent'' deviations from Paschen's law are to be expected under certain experimental conditions. The latter inconsistency and the method of avoiding it are described in the paper.