Flashover in vacuum tubes

Despite a controversy that is as old as the model the secondary electron cascade for explaining the flashover still abounds. However, advanced understanding of charge trapping and associated energetic phenomena in insulators open new perspectives. The logic of pulse power experiments was to measure a flashover strength under short voltage pulses (ns, μs). However, the dielectric strength is not a material intrinsic parameter because it depends on many parameters (temperature, vibrations, field design at the triple junction, insulator surface). The logic of a DC experiment is to relate the flashover occurrence with monitored charge localisation. It is shown that charge injection allows explanation of the various results and particularly the emissions that are related to losses. The injection may come from the cathode emission in vacuum or from the cathode-insulator interface. Losses are localised on zones where reinforced field or defect concentrations produce a high trap charge density. DC experiments show that the flashover is a quasi periodic phenomenon. When the flashover frequency is normalised to the dose that has passed through the specimen it seems that the electron injection (at the electrode-insulator interfaces or under irradiation) and trap characteristics determine the frequency. All of these results allow proposal of a new flashover mechanism based on the existence of a self-sustained charge relaxation wave. The interest of this model reaches beyond the flashover domain