How can I prevent LabView from crashing after about 1080 iterations of my VI

[UW Soft Matter Physicist]

I have designed a vi that runs in a for loop. It takes data from the BNC-2110 (using the AI Config and Single Scan VIs), and then calls a VI to adjust the measurements to a temperature. After this it displays an output on a plot (per input channel). After this program runs for a while (not always the same amount of time), LabView will close and a Windows (2000), application error box will pop-up. The box will simply state that an error has occured and an error log is being recorded.


Here is a copy of the VI that I am using. It may take about 18 minutes for the error to occur (or 8 sec. if you remove the stopwatch). I think that everything is in order (on the programmin
g end). It is a short program and if anybody can find the problem, I would be very happy.


Any help is appreciated, thanks.


Tyrone

PS - I have redesigned this VI after some feedback but the problem still persists.

PPS - It is a fresh install on a newly configured machine.

[DJ_001]

Hi Tyrone,

I'm not sure if you can initialize multiple tasks for one device. If you
wire all the error clusters together and watch for an error, I think that
you'll find that the NI-DAQ VIs will report a failure due to multiple
initializations. Maybe you should try something like what I've attached.
One very inportant point is to always use the error cluster...

Hope this helps.

Denis.

P.S. I haven't tested this VI.





[Attachment Instrument.vi, see below]

[Mark A. Hanning-Lee]

UW Soft Matter Physicist wrote:
>
> I have designed a vi that runs in a for loop. It takes data from the
> BNC-2110 (using the AI Config and Single Scan VIs), and then calls a
> VI to adjust the measurements to a temperature. After this it
> displays an output on a plot (per input channel). After this program
> runs for a while (not always the same amount of time), LabView will
> close and a Windows (2000), application error box will pop-up. The
> box will simply state that an error has occured and an error log is
> being recorded.
>
> Here is a copy of the VI that I am using. It may take about 18
> minutes for the error to occur (or 8 sec. if you remove the
> stopwatch). I think that everything is in order (on the programming
> end). It is a s
hort program and if anybody can find the problem, I
> would be very happy.

Probably best posted to NI's free tech support via www.ni.com/support.

The BNC-2110 is a BNC accessory, not a board.
When you post to tech support, also give them the name of the board
used.

Mark

[mikeporter]

After looking at your code there are several things to try. First, implement some event handling that collects and reports any errors that occur. If you need more specific help on how to do this, contact me directly. I have VIs that can help. Second, try rewritting the code to not utilize the waveform datatype. It's pretty new and it's very possible you found a problem with it. Third, you never deinitialize the DAQ hardware. While this shouldn't cause the problem you are experiencing, but it's a good habit to get into.

Finally, while your application is running look at Windows' memory usage, do you see anything happening to it over time? Like it slowly creeping upwards?

The VIs I have posted implement the first two suggestions in a pretty basic way. T
he error handling in particular should be done better, but this will at least allow you to see if anything is happening...

Mike

mporter@arielcorp.com