Parameter Extraction of Power Entry Filters Using Probabilistic Approach

This paper proposes a probabilistic approach to investigate uncertainty of parasitic components on the performance of power entry filters, so-called electromagnetic interference (EMI) filters. Since its value is unknown, the effect of its degradation can only be observed by output measurement data. Based on an approximate equation of insertion loss (see CISPR 17) and experimental data, approximation values can be extracted. This approximation approach bases on a simple mathematical equation leading to huge errors. To improve accuracy, a detailed mathematical presentation of an accurate high-frequency model has to be proposed. Based on that presentation, Monte-Carlo simulations have been applied to extract precisely its parameters. A probabilistic approach can also predict possible degradation of parasitic components. Based on statistical results, the mitigation disturbance and parameter extraction can be used to clarify the disturbance property of a typical power entry filter.