Frequency of a Voltage Waveform

[cwildmann]

I need to analyze the frequency response of a generator set to look for dips, rises and recovery times from load changes.  We are acquiring the voltage and current waveforms at 10kHz and recording them directly to a TDMS file.  We need to calculate the freguency of the voltage waveform based on the time domain and end up with a file with 10kHz data points that is equivalent to the voltage and current waveforms for analysis.  How can I go about doing this in DIAdem?

Many thanks in advance for your help,

Chris Wildmann

[Pesmith8]

Chris

To get frequency information out of voltage/current waveform, would depend on what the voltage or current waveforms looked like.  The NI guys may have some other tricks up their sleeves, they usually do.

Options  as I see it. 

1)  Time between Zero crossings and update the frequency from periods measured. (most accurate if is zero crossing.)

2)  Look at the slope between points and calculate a very rough frequency estimate.  (Very fast but not very accurate)

Would seem that any FFT done would not relate at all to time domain.

Paul

[Brad_Turpin]

Hi Chris,

If I understand your questions correctly, I think a Joint-Time-Frequency analysis might give you the visibility of the locations of these events and the frequency response to them.  DIAdem calls this an "Time Interval FFT".  You could also consider using the below posted example to make quick joint time frequency plots, but you'll need to extract a Time channel from your waveform data first because that example still doesn't work with waveforms.

http://www.ni.com/white-paper/3549/en/

Brad Turpin

DIAdem Product Support Engineer

National Instruments