Sampling speed problems with the Sound Pressure settings of the NI 9234

[Andrew_Sio]

Hi.

I have a problem with the sampling speed of the NI 9234 board whenever it is used with the Sound Pressure option.

In one of my experiments, I was sampling a 100 Hz signal at Fs = 50 kHz, and when I traced the FFT of that signal, I realized that the peaks were slightly swiped to the left in comparison to the signals acquired with other systems. After some tinkering, I concluded that the signal would be perfect if the sampling speed was 51160 Hz.

To confirm my suspicions, I acquired the signal of a 1 kHz calibrator at 94 dB and the conclusions were the same: there was a difference between the programmed sampling speed and the real sampling speed, with the only silver lining being a proper performance of the DAQ when the sampling speed was set to 51160 Hz.

Does anyone have any clue of why this is happening?

Thanks in advance.

[PsyenceFact · Accepted solution]

The NI 9234 supports a fairly limited number of sampling frequencies.  The spec says that the avaliable frequencies are given by

f_s = (f_m / 256) / N

where f_m is the frequency of the internal master timebase (13.1072 MHz) and N is an integer from 1 to 31.  I believe it's also possible to use an external timebase.

If you are requesting 50 kHz, I think it will give you the next highest available frequency, which is 51.2 kHz (N = 1), i.e. nearly 2.5 % out.  When you change your request to 51160 Hz, you still get 51.2 kHz but this is now less than 0.1 % out.

You can use the DAQmx Timing property node to read the actual Sample Clock Rate acheived after you have set it.

Andy

[Andrew_Sio]

That makes a lot of sense! Thank you for the quick and accurate answer!

Andrew