DAQ data sampling discontinuous

[shh2745]

Dear Sir:

I use DAQ 6023E to sampling the sinusoid power signal (60HZ) to my program. My setting sampling rate is 1920 samples per second, which means 32 samples per cycle. When I test the DAQ function, I found a strange result. I use DAQmax function setting to read DAQ sampling data with 32 samples for N channels N samples continuously without any delay. Every time I would read 32 samples sinusoid power signal, which is one cycle data, I should read one full complete sinusoid signal. However, I found that in 32 samples data, there are some unexpected data jump or loss during 32 samples. Is there any one have the same issue? Is it hardware issue or software issue? I just attached my test result. The red color data in this test result is the unexpected data sampling. The test condition is to record both the raw data and RMS data for each 32 samples. Thanks for your help.

Fred

[JoeSavage]

In order for us to help you figure out what the problem is, you'll need to give us more details about how you're programming DAQmx to acquire the data. You don't need to post your code; just a brief description of the function calls should help.

--Joe

[shh2745]

Dear Sir:

The test program is attached. After I read the sampling data from DAQ 6023E, I calculate the RMS value and do some calculation. Recently, I test my developed program with different DAQ, PCI MIO 16E4, and it works well. Therefore, is it possible that the DAQ 6023E have some problem with continuously sampling? Anyhow, thanks for your patient to help me figure out what's wrong with my application.

Best Regards.

Fred

[Marni]

Hello Fred. Thank you for contacting National Instruments. Please try the following:

1. Connect a Simple Error Handler.vi after the Clear Task.vi. Do you get any errors?

2. Do you get incorrect data with the PCI-6023E with only one channel? If not, how many channels does it take to get the error?

3. Look at your data with a waveform graph? Do you see any spikes?

Please let me know if you have any questions. Have a great day!

Marni S.
National Instruments

[shh2745]

Dear Sir:

As the test result I attached above, all the channels that I test have discontinuous data jump, which means spike at the same time. As I mentioned before, when I change to other DAQ, this situation did not happen. And there is no error message for DAQ. So I am really curious about this situation. Thanks for your response.

Fred.

[rocket scientist]

I have had a similar problem and I found the answer to be even stranger -- it was some kind of plotting bug which occured at teh boundaries of the data blocks. The plotting problem even persisted after I wrote the data to file and re-plotted it! I finally convinced myself that there was no missing data by carfully looking at the numbers and actually listening to the waveform. Its a perfect waveform--the plot just looked bad.

[shh2745]

Hi,

Thanks for you sharing your experience. I still have some doubt that when I record the data sampling from DAQ to file, and check the data in file, I really found the discontinuous data in the file in 32 samples, you can see the result I recorded above, test result.xls. Therefore, the daq sampling in the setting block window is not continuous. This probably is not the plot issue, but the hardware issue. I try to test with other DAQ and this problem did not show up. I suspect that hardware probably have some problem, but I still work on it try to find the real reason of this issues. Welcome for every one to share your experience. Thanks a lot.

Best Regards.

Fred

[Marni]

Dear Fred,

Does this happen with only one channel? If not, how many channels makes this occur?

Marni S.
National Instruments

[shh2745]

Dear Sir:

Finally, I figured out what's wrong with this kind DAQ data sampling discontinuous issue. The reasons is that while I am running my program, I need to keep a lot of data in process. When I process the sampling data, I have to record the data if some event trigger, therefore, I just keep to accumulate the data before I write them into file. After I reduce the size of accumulate data, this issue just gone. It seems like the memory is the main problem. I just think that can I use direct memory access to process my data? If yes, how? Welcome for everyone's suggestion.

Thanks.

Fred

[Marni]

Dear Fred,

The NI DAQ boards default to using DMA. The low cost E Series board PCI-6023E only has one DMA channel, where the PCI MIO-16E-4 has 3 DMA channels. From your previous posts it appears that you are only doing analog input, which would only use one DMA channel. Are you doing analog output or using the digital lines as well? Have a great day!

Marni S.
National Instruments