Data acquisition problem at 1.5M sampling rate

[msarun]

Hai friends,

               i am using 6133 simultaneous data acquisition PXI card in RT Controller PXI 8106, i have to acquire data from this card for 10 seconds at 1.5M sampling rate(1.5MS/s). while giving these much data as a samples to read per channel in finite sample data acquisition, the on board memory of this PXI card is not sufficient. but if i am going to continuous data acquisition, i am unable to store the data in file at this rate, i have tried producer and consumer algorithm, in that case also i am unable to write all the data....

During my testing, i didn't receive any error, but i can't see any data in waveform array. what is the reason for this?

      i have to acquire 10 seconds of data at 1.5M sampling rate, so what should i do?  ... pls help me,

[Dennis_Knutson]

DON'T rename a bmp file to get around the ban on posting them. Do you think someone won't notice or do you really think that changing the extension will change the format?

[msarun]

sorry, i dont know about that...

[Dennis_Knutson]

Open the image in paint and save as png. Attach that.

[msarun]

Please find the attached png file.

[Dennis_Knutson]

I don't have any experience with the RT controller but my suggestion would be to first get acquisition working before worrying about saving the data. Try using continuous mode and instead of setting the number of samples to be greater than the sample rate, specify -1 and place the DAQmx Read inside a loop. Stop the loop when the number of samples acquired is equal to 10 seconds worth of data.

An example such as Cont Acq&Graph Voltage-Int Clk would be a good place to start.

Message Edited by Dennis Knutson on 03-07-2010 04:16 PM

[AndrewMc]

---

msarun wrote:

the on board memory of this PXI card is not sufficient. but if i am going to continuous data acquisition, i am unable to store the data in file at this rate, i have tried producer and consumer algorithm, in that case also i am unable to write all the data....

---

The onboard memory should not be any problem here.  The device streams data to an intermediate buffer, which is then scaled and placed into a LabVIEW buffer during read.  By specifying the "samples per channel" on Configure Timing, the buffer should be about that size (about 30 MB), at which point, keeping up with the data should be irrelevant.  That is, all of the data would be acquired into memory and the device should not depend on your calling read.

To be honest, I don't know what the issue is here.  DAQmx Read should never return without either an error or data (unless you read 0 samples).  One small consideration in your example is that you should call Create Task before your loop and wire that in (rather than have an uninitialized shift register).  That shouldn't be the problem in this case, but just keep that in mind.

One idea here would be to install DAQmx 9.1.  Then, you can simply select the new "Logging" tab in the DAQ Assistant and check "Enable Logging".  If you do this, not only will you get the absolute best performance possible streaming to disk, but as well, this would be rather easy.  (Note that the "Configure Logging" VI is available in DAQmx 9.0.2, but just is not available throught the DAQ Assistant.)

If you want to use "Configure Logging" programmatically, check out: http://decibel.ni.com/content/docs/DOC-7659

If you use the DAQ Assistant, here's how you would do that: