A Fast and Accurate Coarse Method for Multipactor Threshold Prediction of RF Filters Under Modulated Signal Excitation

Multipactor is a key high-power effect limiting the system performance for onboard satellite hardware. Although modern particle simulators admit arbitrary geometries and signals as inputs, their practical use is often limited to continuous-wave (CW) excitations. Unfortunately, the multipactor analysis for input-modulated signals normally leads to prohibitively large CPU times, as signal lengths are very large compared to the electron population’s evolution time. The Coarse Method is an elegant way of overcoming this limitation, providing a good estimate of the multipactor threshold in reduced CPU times. However, if the input signal is not preprocessed before being analyzed, the method is unable to account for the frequency dependence as it operates with electron dynamics information extracted at a single frequency, leading to biased predictions for narrowband samples as filters. This article proposes an extension to the original Coarse Method implementation by considering the sample response and the modulated signal spectral distribution to account for the frequency dependence. The resulting method is suitable for estimating the multipactor threshold of narrowband samples excited by modulated signals, while keeping the benefits in terms of simplicity, efficiency, and generality of the Coarse Method. The proposed approach is benchmarked against laboratory measurement results, as well as particle simulators and legacy Coarse Method predictions, revealing the advantages of the novel technique and its range of applications.