Strange sampling with PCI-6034E

[golubovski]

Hello,

I have a strange sampling issue with a PCI-6034E.How can I obtain allowable sampling rates for this device? I already have an experience that NI-9234 does not report an error if I try to sample with less than its min Fs (1.6kS/s) - it coerces to 1.6kS/s!!!

With Fs=20kS/s and ReadSamples=400 (chunks of samples per channel per read) TDMS says the correct dt=0.00005s yet the waveform has the length of scans (Fs/chunk) instead of desired samples..?!?

If I put -1 for smaples_to_read (DAQmx read.VI) the length seems OK!

I am trying to find out the Fs values that PCI-6034E can sample with but there is no such info anywhere!

Can anyone explain this?

Thanks in advance,

[Joel-P]

Hello golubovski,

Here is a KnowledgeBase article which describes what the possible sampling frequencies are depending on the type of device.

http://digital.ni.com/public.nsf/websearch/4BBE1409700F6CE686256E9200652F6B

The allowable sampling rates with the NI-9234 is also documented in a KnowledgeBase here.

http://digital.ni.com/public.nsf/allkb/593cc07f76b1405a862570de005f6836

I am slightly confused, regarding the samples to read. My current understanding is that you are using finite sampling with a Rate of 20,000 and a Samples per Channel of 400 (samples to read). 

The dt should be the reciprocal of the sampling frequency, and this appears to be correct as 1/20,000 = .00005

You mentioned that with a sampling frequency of 20kS/s and a read of 400 samples per read that the waveform has a length of (20k/400) = 50 samples. Is this correct that you are only reading 50 samples?

Or is the length of scan that you are referring to the total time of the waveform. The length of time that the waveform takes should equal Samples per Channel (400) / Sampling Rate (20,000) which would translate to  0.02 seconds.

Setting the DAQmx read VI to -1 for Samples to Read will cause the DAQmx read vi to read all of the data available on the buffer. In Finite Sampling this should be the Samples per Channel.

I hope this helps,

Joel