Microwave window breakdown experiments and simulations on the UM/L-3 relativistic magnetron

Experiments have been performed on the UM/L-3 (6-vane, $L$-band) relativistic magnetron to test a new microwave window configuration designed to limit vacuum side breakdown. In the baseline case, acrylic microwave windows were mounted between three of the waveguide coupling cavities in the anode block vacuum housing and the output waveguides. Each of the six 3 cm deep coupling cavities is separated from its corresponding anode cavity by a 1.75 cm wide aperture. In the baseline case, vacuum side window breakdown was observed to initiate at single waveguide output powers close to 20 MW. In the new window configuration, three Air Force Research Laboratory-designed, vacuum-rated directional coupler waveguide segments were mounted between the coupling cavities and the microwave windows. The inclusion of the vacuum side power couplers moved the microwave windows an additional 30 cm away from the anode apertures. Additionally, the Lucite microwave windows were replaced with polycarbonate windows and the microwave window mounts were redesigned to better maintain waveguide continuity in the region around the microwave windows. No vacuum side window breakdown was observed in the new window configuration at single waveguide output powers of $120+MW$ (a factor of 3 increase in measured microwave pulse duration and factor of 3 increase in measured peak power over the baseline case). Simulations were performed to investigate likely causes for the window breakdown in the original configuration. Results from these simulations have shown that in the original configuration, at typical operating voltage and magnetic field ranges, electrons emitted from the anode block microwave apertures strike the windows with a mean kinetic energy of 33 keV with a standard deviation of 14 keV. Calculations performed using electron impact angle and energy data predict a first generation secondary electron yield of 65% of the primary electron population. The effects of the primary aperture electron impacts, combined with multiplication of the secondary populations, were determined to be the likely causes of the poor microwave window performance in the original configuration.