9233 sample rates

[E Murphy]

Hello everyone:

I´m using a 9233 USB device. It has antialias filters. I tried the device with a VB.NET program that I´m using with a DAQPad 6015. The problem is that certain sample rates give me incorrect frequencies in my spectrums. For example, using 6400 samples per second gives incorrect results, also 15,000 and 20,000 S/s. No problem with the DAQPad 6015. Could it be that the setting of the antialias filters is only at certain frequencies? I would like to use 2560, 6400, 12800 and 25600 S/s to get an adequate full scale spectrum (1000, 2500, 5000 and 10000 Hz, for 800 lines)

Thanks for the information

Regards

Eduardo Murphy

[Travis_W]

Valid sample rates for the 9233 are the following:

Fs = 50kHz / N, where N is an integer 2 . . . 25.
There are also the valid sample rates of 50kS/s and 33kS/s

If you set the sample rate to be something other than these, the actual sample rate gets coerced up to the next valid value.
The antialias filters are then scaled based on the sample rate.  For Fs<25k, the stopband is at Fs x (.45).

Make sure the inputs to your FFT routine account for the actual sample rate of the 9233.

Hope this helps,
Travis

[E Murphy]

Travis:

Thank you, it sure helps.

[duncan carter]

The  expression for the rates is  Fs = 100 kHz/N, where N is an integer  rangeing from 2 to 50.

You can get the actual rate by using:
double actualrate;

DAQmxGetTimingAttribute (taskHandle,DAQmx_SampClk_Rate,&actualrate);

after using:

error = DAQmxCfgSampClkTiming(taskHandle,"",rate,DAQmx_Val_Rising,DAQmx_Val_FiniteSamps,sampsPerChan);  

[E Murphy]

Hello Duncan:

Thank you for your reply. Why is this information not included in the user booklet? Several years ago (10 or more) NI did a better job with their manuals. Now you have to dig everywhere.

Regards

Eduardo Murphy

[duncan carter]

I don't know why this isn't listed in the 9233 manual or specs or sample code; it is/was for the E-series hardware.   I'm in process of writing a 9233 driver for my company's vibration software and I would have realized that this occurred sometimes in the next week but I'm happy that you asked the question so that I could make the needed changes in my software today.  I've written similar drivers for the E-series hardware using the older Nidaq and already have routines to handle this problem so that its not a difficult one for me.

[Travis_W]

Duncan,

Thanks for clarifying the valid sample rate equation.  I like yours a lot better.

NI has identified this lack of documentation as an issue.  We're currently working on getting this info into the specs for the USB-9233.  In the meantime, there should be a KnowledgeBase entry going live in a few days that details the valid sample rates.

Thanks,
Travis