Hello everyone,
I am doing an impedance measurement on plasma machines for several load conditions based on an impedance spectrum. Those are supplied with pulse voltages and currents to create the plasma. I measured for several loading conditions the pulse voltage and the pulsed current. To get the impedance spectra I need basically four things the amplitude spectra (magnitude) of the voltage and current pulse and the phase spectra of the voltage and the current pulse.
For the plausibility check of the results the whole VI is tested with the csv data, coming from an ideal square wave. For the FFT I am using the Express VI: "Spectral Measurements". Before the Express VI I am reading in the csv data to arrays and resampling the array to a number 2^n of samples per periode of the incoming signal. After the VI I am just filtering the peaks from the spectra out and calculating the impedance spectra. The problem is that the Express VI delivers good and accurate results for the magnitude but not for the phase. The phase should be for the first harmonic 90deg, for the third harmonic -90deg, for the fifth 90deg, for the seventh -90deg and so on. I calculated the phases and the magnitude based on the Fourier transformation formulas (001.jpg)
In my case in the VI it stays at 90deg. I tested it also with a triangle signal and also got wrong phases. So has anyone experience how to get correct values for the phase using the express VI Spectral Measurements? To test the VI for both filepahts the attached file (1kHz_Rechteck_ideal.csv) has to be entered, for the Period time the value 1000 [µsec] has to be entered (due to T=1/1khz=1000µsec) and for the resampling of the incoming signal 2^n a value for n between 5 and 15 has to be entered.
Best regards
Alexander
