Timer drift in Labview 8.20 application for a PDR 2000

[md12]

I am taking readings in Labview 8.20 from a MKS PDR 2000 pressure transducer. The Labview interface they provided allows me to set the reading buffer (using 3 or 4 ms right now) and the sampling time (set at every 20 s). I've noticed a timer drift as the clock goes on. For example, it will record at 0, 20, 40, 60 s, etc, but then eventually it will get to like 341, 361, 381 s. By the end of a 5 day run the drift is pretty significant. I don't even know if what it's recording is accurate or if there is some systematic error in the counter.

Has anyone had any problems like this or have a suggestion for how to test or fix it?

[Esteban_B]

Hi md12,

Is this instrument connected to the computer USB, Is the program that you have Using NI-VISA to communicate with the device? Could you post a link with more information about this device?

We need to understand first where the time information is coming from and then try to determine the cause of the drift.

The program that you have is an executable or a VI?

Regards,

[Michael_78]

Hi,

Over long periods of time you have to be very careful about choosing a timing method.  WaitmS and the elapsed time express vi will definitely loose time.  The Express vi will reset when it is called if it needs too, but this will always be some time after the time has actually elapsed.  You Could try using the wait for next mS multiple vi, or poll the Tick Count mS vi or a timestamp and look for a time difference being exceeded.  I have used the timestamp VI and look for the seconds to change to log data at 1 second, the application had to run for months on end and I found this to be the most reliable timing method.

Michael.