DAQ AO to AI time delay

[IoInBlack]

Hallo,

I'm working on an academic control project, where I need to read (ni9215) 3 or 4 channels very fast,

process the data and write them back (ni 9263) via usb-chassis 9172. To check the accuracy and time

delays I tried a loop from AO to AI (only one channel) and generated a step, saved the timestamp of the

step and checked, when the step was readed in the written measurement log. I doesn't see andy difference

in using two VI's to read and to write or combining them in one VI, so I added images of two VI's.

While comparing the timestamps of the sampled step and the log file, i noticed time delays of about 200ms

from output to input. So my VI's are times too slow for my control purpose... anybody knows, what to try?

In this VI I read data with 500hz, 10 samples to read and 10 samples/channel. The read timeout is set to

default, but if I set it to zero, strange (negative) times are written to the measurement log... I tried

to change from N samples acquisition to single sample acq, but it didn't speed up the VI.

Thanks for answering!

[Andrea_Perin]

Hi IoInNBlack,

first of all thank you for posting in this forum.

I think you should synchronize your input/output operation, otherwise I don´t understand how you can take the timestamps and consider them

reliable for your evaluation. I suggest you to modify the LV example here attached with timestamps and file I/O to build up your test and see what results you get then. Consider also that  20 ms are necessary for the acquisition of the samples and then you have add the time necessary to manage the file I/O.

Best regards.

Andrea P.

Message Edited by pxchellone on 07-06-2009 09:23 AM