Pulse shortening in high power microwave sources

Observations show that the ubiquitous pulse shortening problem is due to the formation of plasma, electron streaming, RF breakdown and beam disruption in the devices. We review recent experiments in terms of these mechanisms. Linear beam devices exhibit all these mechanisms; in relativistic magnetrons the dominant mechanism is resonance destruction by cathode plasma, probably from water contamination of the surface. We call for the introduction of improved surface conditioning, cathodes which do not produce plasmas and increased effort on measurements of the RF and plasma properties of HPM sources.