Min & Max frequency of a waveform in LV 7.1

[Coq_Rouge]

I have worked a lot with peak detection and find it with some fine tuning to be quite accurate. At least for a relatively stable and noise free signal like yours seam to be. As I understand you do a peak detection in order to find peaks and then calculate the frequency from the time interval between the peaks. Your problem as I understand from the word document you posted first. Is that the peaks are more accurate detected, or artificial peak detections occur.  I you want I can do some test if you send me some data. The simplest way to do this is to copy the chart with data to a new VI then select make current values default. I do not have 7.1 but 7.0 and 8.6 so perhaps save in Labview 7.0 

[lmd2]

 so you think the approach can work (zero crossings sounded like it may have possibilities)

 here is the graph with data saved as default

good trick, this is smaller than the tab delimited text file

when I use the Peak Detector.vi  and input a subset of the data (between the cursors) say 4 cycles; depending on the actual window the peak detector might return 4 peaks or 5 or 7(?) At 60KHz the peaks should be about 1000 samples apart, most are but the 'extras'  are only 23 or 71 (or whatever) samples from a true peak. I've tried different values for the 'width' input parameter of the Peak Detector.vi, from 3 to 11 - no real difference that I noticed

thanx 

[Coq_Rouge]

I see your problem and with some fine tuning I think I can do something.

[lmd2]

thanx for the update, fine-tuning how? there  isn't much configuration

I really appreciate you following through with this,  

[Coq_Rouge]

I think I am into something is this correct acording to the data you sendt me?

[lmd2]

that would seem to be what I am experiencing - certainly not a smooth response

is that true instantaneous frequency, or is there some sampling error or something?

[Coq_Rouge · Accepted solution]

I think the error is in the main part introduced  during the sampling process. It can be smoothed somewhat, without reducing the peak value to much. But this is the best I can get. With current approach

[lmd2]

if I am not being rude, can you explain how you generated these plots. Seems to be instantaneous frequency versus time. very cool

also can you speculate as to what is causing the error in the sampling process?

and thanx again for all your help

lmd2 

[Coq_Rouge]

It was not very hard. Here is the code. But now you ove me some beer if I come to your place sometime 😉 I shall speculate during the night and see what I can come up with

[lmd2]

not sure of your physical location, but if you find yourself around Manhattan or Long Island, I will buy you several beers!!