LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

analyzing signal

The zero hertz component is due to the offset in your signal, not the power supply.  You can get rid of it by subtracting the mean of the data values from the signal before calculating the spectrum.

 

If you signal and its spectrum are clean enough, just use the peak detector function.  Loop through the data.  After finding the first peak, remove it and then find the peak value in the remaining portion.  Repeat until you have as many peaks as you want.  If there is much noise or spurious signals, spectral leakage, or windowing effects, this may not be entirely accurate.  That is one reason it is often useful to look at the data with a discriminating eye before trying to automate the analysis.

 

Please post some data.

 

Lynn

0 Kudos
Message 11 of 16
(889 Views)

We choose to use waveform peak detection VI. The input signal is from the result of FFT.  For the output, we use 'location' to get the peak frequency from the graph. we decide to use DBL VI to get a single value of the peak. But there is an error because of array element conflict. what should we do?

0 Kudos
Message 12 of 16
(873 Views)

Please post your code, showing what you have tried and what  is not working as you expect.  Include some typical data.

 

Lynn

0 Kudos
Message 13 of 16
(860 Views)

we already get the suitable DBL..but the problem now, we set in the 'simulate signal' to 50 Hz..so supposedly we get peak freq at 50Hz..but we only  get the value at 5Hz..when we change to 80Hz, it will display 8Hz.. so what's the problem?

0 Kudos
Message 14 of 16
(829 Views)

I do not see the problem you are describing.  When I run your VI with the signal at 80 Hz, the spectrum graphs show 80 Hz.

 

The peak detector on the plain FFT does not find peaks due to the small size of the peaks (~1E-13).  This seems to be due to spectral leakage or some mathematical quirk.  If you change the size of the FFT to anything other than the sample length (for example 128), the peaks have a reasonable amplitude.  The FFT does not have any timing information so the frequencies have to be calculated separately.

 

Lynn

0 Kudos
Message 15 of 16
(818 Views)

here is our updated VI..

 

the new problem is for the array comparison part, we want to compare the array of frequency graph and array of frequency mosquito with respect to the 'i'..but we don't know how to make a loop so that the freq. graph [i] will be compared with freq.mosquito [i]..

 

tqSmiley Wink

0 Kudos
Message 16 of 16
(802 Views)