Search

To investigate multipactor, we designed a testbed for S-band frequencies with geometries akin to the standard WR-284 waveguide geometry. Narrowing the waveguide height from WR-284 dimensions to 5.5 mm for a suitable frequency-gap product left the waveguide cutoff frequency unaffected and the RF signal propagating in the dominant TE10 mode. A coaxial magnetron provides the test input power at 2.85 GHz with a peak power output of 4 MW and a 3.5 μs pulse width, and an RF solid-state source using gallium nitride high-electron-mobility transistors delivers a pulse width of 100 μs with a test input power of 2 kW for comparison of threshold power. Local (electron multiplier tube) and global (phase/power) diagnostic methods are implemented to detect multipactor within the test setup. At power levels tested (MW) and a 5.5 mm gap, low multipactor orders (N = 1) are observed. Prior numerical studies demonstrated that structures with grooves in the waveguide broadside wall reduced multipactor susceptibility. Such grooves run along the direction of propagation to avoid continuous impedance mismatching. The efficacy of this mitigation technique was experimentally evaluated. Differences over a smooth broadside wall are noticeable, however limited.