Analysis of sensitivity to the main parameters involved in the digital implementation of the UIE flickermeter

Abstract

The IEC 61000-4-15 standard describes the analog implementation of a flickermeter up to the output of block 4. The digital implementation of a flickermeter is accepted provided that it satisfies the accuracy specifications both for instantaneous flicker sensation per unit and for the short-term flicker severity or Pst. Nowadays, most of the commercial flickermeters are partially or totally digital. It is noticed that even though they comply with the accuracy specifications of the standard, they could yield maximum measurement differences of 30% in field tests according to some authors. In this paper the influence of the main parameters involved in the digital implementation of the UIE flickermeter has been analysed, both for instantaneous flicker sensation and for the Pst. The flickermeter has been implemented using the MATLAB programming language. The testing voltage signals can be analytically generated or read from field test record files. Four parameters has been analysed in order to quantify their influence in instantaneous flicker sensation: signal sampling rate; number of resolution bits in quantization; processing sampling rate; and influence of the high pass filter for demodulation. In relation to the flicker severity measurement accuracy, besides aforementioned parameters, the influence of the number of classes and of the type of classification in block 5 has been analysed.