Effects of window functions explained by signals typical to power electronics

A lot of questions arising in connection with power quality can best be answered in the frequency domain, that is after performing a fast Fourier transformation (FFT) of recorded or simulated time-dependent signals. Modern measuring equipment and modern simulation programs make it easy to perform such a FFT analysis of a time-dependent signal. This FFT analysis leads to a discrete spectrum containing the so-called harmonics of the time-dependent signal. It is well known that the result of such a FFT is only then easy to interpret if the time period used for the analysis matches the period of the time-dependent signal to be analysed. Especially in case of measured signals this often presents a problem. One way to solve this problem is the application of so-called window functions to the time-dependent signal prior to the FFT analysis. However, care has to be taken when applying window functions to some signals typical to power engineering. This paper shows, that severe misinterpretations are possible, and gives guidance for the proper application of window functions in power engineering applications and the correct interpretation of the results.