Hi guys, Nyquist theorem – greater than two times the highest frequency component in the signal. My tests last for 50 milliseconds sampled at 100 kHz. How can I determine the optimum sampling rate that I can use for my tests in order to not lose the peaks? I am thinking to perform the test with different sampling frequencies and check whether the amplitude vs time is affected. When I plot the FFT, it shows approximately 2 kHz stable frequency (it becomes zero in the x axis). Shall I perform the same test with the the sampling rate as the only variable and check the amplitude?

LabVIEW

Nyquist theorem - Determination of Sampling Rate - FFT

Brad,

I have looked at your data and the specs for the PXI-4331. With that device you probably do not need any additional anti-aliasing filter. Most DAQ devices do not have such filters built in which was one reason I kept mentioning that.

It is impossible to tell whether you have any power line frequency noise in your data. The raw data does not have any visible indication of a signal with 20 ms period, although the smallest signal that would likely be detectable by "eye" is about 20 or 30 on the amplitude scale. A signal that big would only be 20-30 dB below the peak signal. With the amount of data in your dataset the frequency resolution is ~83 Hz, so any 50 Hz compnent is mixed in with the DC component. This explains why removing the offset has very little effect. There is quite a bit of energy at frequencies below 83 Hz in the data.

The baseline noise level is about 30-40 dB below the peak signal. This seems unusually noisy for a load cell.

Try acquiring data for 1 second with no impact on the load cell. This should allow you to separate the 50 Hz component and may give a better idea of the noise levels. Then try sampling again for 1 second with a sampling rate of 10 kHz and compare the results. If the noise is thermal noise, then the slower sampling rate and narrower anti-alias filter should reduce the noise by about 10 dB.

Lynn