Multipactor has become a keylimiting factor of the final performance of satellite communication systems, due to the increase in power levels and/or operating frequency bands. As a result, the critical components of these systems must meet demanding multipactor specifications which should be considered during the design process. This paper describes the different techniques available to predict the multipactor threshold power for radio frequency (RF) and microwave passive hardware under continuous wave (CW) excitation, from cumbersome particle simulations to fast approximate methods based on circuit models. All these techniques have been described and compared together for the first time, including also a detailed description of the configuration issues of commercial particle simulators required to obtain accurate multipactor threshold predictions. The techniques are applied to both wideband and narrowband application examples. The predictions have been compared with measured thresholds of manufactured samples obtained with a novel multipactor test bed, thus allowing to highlight the advantages and limitations of each technique and particle simulator. From this paper, it will be possible to choose the most suitable procedure (and an appropriate simulator, if needed) to obtain multipactor threshold prediction of passive hardware.

High power radio frequency (RF) breakdown testing is a subject of great relevance in the space industry, due to the increasing need of higher transmission power and smaller devices. This work presents a novel RF breakdown detection system, which monitors the same parameters as the microwave nulling system but with several advantages. Where microwave nulling-a de facto standard in RF breakdown testing-is narrowband and requires continuous tuning to keep its sensitivity, the proposed technique is broadband and maintains its performance for any RF signal. On top of that, defining the detection threshold is cumbersome due to the lack of an international standardized criterion. Small responses may appear in the detection system during the test and, sometimes, it is not possible to determine if these are an actual RF breakdown or random noise. This new detection system uses a larger analysis bandwidth, thus reducing the cases in which a small response is difficult to be classified. The proposed detection method represents a major step forward in high power testing as it runs without human intervention, warning the operator or decreasing the RF power automatically much faster than any human operator.

No abstract