A Study of the Influence of ACMains Impedance and FrequencyDomain EMI Evaluation Methods on EMI Measurements

One of the problems in EMI (electromagnetic interference) is poor reproducibility in measurements. This report is divided into three parts. In Part I, this paper shows that the real conducted noise can't be measured by a LISN (line-impedance-stabilization network) and a 50 ?>-input-impedance instrument, as specified by regulations such as those of CISPR (International Special Committee on Radio Interference). This is based on the generation mechanism for conducted noise on AC mains, generated from the AC-operated EUT (equipment under test). In Part II, it is pointed out that one of the causes of poor reproducibility in radiated-emission measurements is the difference among the line impedances of the AC mains. A VHF/UHF LISN (called Maeda's LISN) is introduced as a method to solve the poor reproducibility. In Part III, it is clarified that in evaluating the noise (including both conducted and radiated noise) represented in the frequency domain, the frequency spectra of noise, measured at different times for the same EUT, are different. This is shown by proposing a model to analyze the measured results obtained by the measurement method specified by the regulations.