Abstract
EMC spectrum analyzers are popularly used for electromagnetic interference (EMI) measurement in power electronics systems. Depending on the specifications of EMI standards, the EMI measurement could be very time-consuming. Conventionally, the fast Fourier transform (FFT) is used to derive the EMI spectrum from the measured time domain waveforms. However, these results may not agree with the measurement results from spectrum analyzers, and sometimes the difference could be significant. In this paper, a technique to quickly and accurately predict and analyze the EMI spectrum from time-domain waveforms is proposed. The technique is developed based on the spectrum analyzer's operating principle and the requirements of EMI standards. The EMI spectra of three modulation schemes are also analyzed. Theoretical analysis, simulations and experiments were all conducted. The predicted peak, quasi-peak and average EMI matches the measured EMI in whole conductive frequency range. The developed technique can accurately predict EMI using much shorter time than conventional EMC spectrum analyzers and it saves cost of expensive spectrum analyzers.
Funding Information
  • CRRC Zhuzhou Institute Co., Ltd.

This publication has 20 references indexed in Scilit: