2 parallel data acquisition loops

[tintin_99]

Hi. I use 6363 OEM daq and I would like to acquire 2 digital channels at different sampling rates ( one 2MS/s and 10kS/s)

So I tried to use 2 parallel data acquisition loop but when I run one of them works fine but the other one gives me this error and stops

Error -50103 occurred at DAQmx Start Task.vi:2

Possible reason(s):

NI Platform Services: The specified resource is reserved. The operation could not be completed as specified.

Could you please let me know how can I acquire bothe signals at differnt sampling rate with an example

Thanks

[Dennis_Knutson]

You can't. This error is common and has been discussed numerous times. Since you have a single clock for the one digital resource, you have to acquire all channels in a single task and at the same rate.

[tintin_99]

Thank you Denis. I have to acquire 6 digital signals. The sampling rate of 5 of them can be as low as 1000S/s but one of them should be atleast 1MS/s and then I need to log digital waveforms into a TDMS file. The problem I have is if I use 1MS/s for all of them then I make a very big TDMS file while 80 per cent of the samples are really not needed . I finally want to read the data from TDMS and graph the waveform and as I said I will waste a lot of memory if I use 1MS/s for the other 5 signals. Can you help me on this. How can I optimize this process?

Thanks

[ikonen]

You'll have to collect them all at the highest sampling rate you need and then downsample the other channels in software before you save them.  If you sample at 1MS/s and only need 1kS/s just pull out every thousandth element into a new array and update the waveform dT (1000 times larger).  I think there's a built-in Waveform processing function that will do both steps in one function, but I'm not looking a copy of LV at the moment.  It's not particularly hard to code manually if I'm wrong.

[Yamaeda]

Yes, there's a downsample VI, if it just decimates the array or average the parts i dont know, though it can be rather easily tested. If downsampling i feel it'd be good if you could average the downsampled part. Assuming alot of redundant samples it might be moot as differences would be small, but in e.g. a 2:1 or 3:1 downsample with frequencies nearing the Nyquist(?) frequency it would do some difference.

/Y