cDAQ 9172 - NI 9234 buffer size

[José Alberto]

Hi everyone.

I am trying to get acceleration from 4 channels using a NI9234 and a cDAQ9172. I was using the DAQ assitant to get the signals from the four sensors, and then store the data in a text file using the save ascii/lvm function.

The problem that I am finding is that although in the graph the signal is represented perfectly, the data stored is not the whole measurement. I mean that I am measuring 5-10 minutes with a frequency of around 4KHz (in the graph it's perfect), but what is stored doesn't correspond to that. What I have seen is that I am storing the first 100000 samples (approximately), and then again these samples, but the first column (the time) it's being stored well. So it's like if I were storing the first part of the test several times, like a periodical signal.

I have done the same measurement with signal express and it stores perfectly the data, with no repeated samples. So I guess that I am doing something bad, maybe exceeding the buffer size, but I can't find it.

I am attaching an image of that part of the vi. The other part is just preparing the configuration (sucha as time to measure and sensitivities) so I think it's not important.

[sendia]

Hello Jose,

First of all, the NI-9234 cannot use any Acq. Frequency but must satisfy the specification of page 16 (http://www.ni.com/pdf/manuals/374238c.pdf). If the frequency you are using is different, the module will approach to the closer frequency.

Check the Save.vi configuration to make sure you are not appending data with old acquisition.

In this case, it is better to attach the file instead of a screenshot to check the configuration.

Regards.

[José Alberto]

THanks for your reply. I know about the frequencies. I was using 1651.61 Hz. The problem that I am finding is that it stores 100000 data, and then stores the same data again until it reaches the number of samples that I have introduced.

That's kind of weird because the time scale is normal. I wasn't using appending.

 I attach the vi but  it uses another subvis and it's just to create text files. The problem comes with the daqassistant and save to ascii.

Thank you very much

[sendia]

Do you see how data is changing in the ”Acceleration Graph” (I mean, after the fist pack)? or the data represented in the indicator is a periodic information too?

Depending on your answer the problem can be in the customized DAQAssistant.vi (attach it to check your configuration) on in the “save to ASCII.vi”

[José Alberto]

Yes that's the point. In the graph I see the data perfectly, not only the first pack, but when I use the data stored and I try to represent it with gnuplot or matlab it only represent the 100000 first pack but repeated.

I think the problem is in save.vi but I can't find where the problem is. I am attaching the vi.

Again thank you very much.

[José Alberto]

I have more test to figure out where the problem comes.

I have generated with signal express a project that gets the data from the sensors and store them into a text file. I have introduced a sample rate of 25.6KHz, and 2.56MS to read (100seconds). I have performed a measuremente and it works perfectly. I have represented the data with gnu plot and it's fine.

Then I have generated labview diagram with the generate code tool of signal express. It's suppossed to have the same functionality that signal express project. I have run the labview vi. And it has stored the data. I have represented it with gnu plot, and I have seen that again the same problem. The data is stored in apackages of 100000 samples, and this pack is repeated continously.

I'm posting the different results with signal express and with labview.