LabVIEW
Continuous data acquisition and saving in new files
Solved!
There is nothing attached.
Obviously, you probably need to show us your VI. There are many ways to run out of memory. Once we see the code, we can probably tell.
Hi kapil.
you don't need to answer 4 times in a row with the same answer, we all can see your message as it's a public board!
Your VI shows some mis-conceptions!
- You read 500k samples from DAQmxRead, with a samplerate of 50kS/s. The timeout is set to 10s: this may fit, but there is a chance it will not fit! The usual recommendation is to read just 1/10 of the samplerate, in your case that would be to read 5k samples at 50kS/s samplerate!
- Your charts are set to display 1024 elements in their history buffer. Each element right now is a waveform of 500kS, which needs 500k*8Byte=4MB! So keeping 1024 waveforms in the buffer requires 400MB of history buffer for each chart!
- Trying to plot (upto= 1024*500k samples = 500M samples in a chart of just ~150 pixels width is just ridiculous! It doesn't make any sense!
You need to use smaller waveforms, less history buffer and still a lot of data decimation when using such tiny charts… - Why do you need two ExpressVI functions (MergeSignal, FronDDT) to create an array of waveforms? What about functions in the array palette, like BuildArray or InitArray???
- Why do you need to open and close the TDMS file in each iteration? Open/close just once before/after the loop!
- Why don't you use the TDMS-saving feature directly built into the DAQmx driver? (Ok, you want to create a new file for each capture, but you also could employ creating groups in your TDMS file instead!)
