Sample Rate Converter As a Means of Reducing Measurment Error of the Voltage Spectrum by Application of FFT

Abstract

Nowdays, measurements of different power quality parameters are based on spectrum measurements by application of Fast Fourier Transform (FFT). At the same time, in order to eliminate the "spectrum leakage" effect, measurement time should be a multiple of the input signal period. Frequency of real power grids is variable. For successful FFT application, the number of processed samples must be equal to a power of two or four. It calls for periodic measurement of the frequency of the input signal and oversampling. A sample rate converter (SRC) is applied to perform resampling. In this article a method of spectrum measurement and apparent power measurement, based on the application of FFT and SRC is described. Measurement errors of these values, due to inaccurate sampling frequency adjustment are analyzed. A SRC based on the Lagrange interpolator is considered. The Lagrange interpolator is implemented on the Farrow structure (Farrow Filter). Analytical expressions are obtained for estimating the error of the magnitude response and phase response of digital SRC of the order from zero to fourth.