Statistical Evaluation of Measurement Accuracy in Full Time-Domain EMI Measurement Systems

FFT-based measuring receivers acquire time-domain data and calculate the spectrum of the interference according to the voltage indication levels of the standard weighting detectors using complex functions and algorithms that are susceptible to heuristic, programming, or numerical errors. Therefore, it is necessary to perform an exhaustive evaluation for validating not only the algorithms in the digital processing core of the instrument but also, the hardware-software interaction under a diversity of conditions and settings. This requires executing the measuring receiver calibration for a variety of representative instrument settings. Such validation approach is beyond the scope of the baseline parameters verification defined in CISPR 16-1-1, however, it allows detecting and correcting problems that otherwise would be neglected. Through the case study of a Full Time-Domain Electromagnetic Interference measurement system for CISPR bands A and B, this paper shows how statistical analysis of the measurement errors can be used to guarantee the accuracy of FFT-based measuring receivers.