Fieldpoint Channels not working when deployed to RT Target

[Scott93535]

Yes they are the same.  I'm not sure that this is a driver issue since one vi works and toggles the outputs correctly while another does not work.  The only reason I can think of is the fact that the outputs are toggled in a sub vi where the test program called them directly.  Its really the only thing different.

Scott

[Wayne.C]

Scott,

The fact that it is a subvi shouldn't make a difference.  Can you post the vi?

[Scott93535]

Here are the vi's.  Kool_APU and StartSeq are the two vi's I am having trouble with.  Kool_APU is the main program and StartSeq is the subvi that is not working properly.  Kool Test is the test vi that I ran which had everything working correctly.

Scott

[Wayne.C]

Scott,

I've modified your StartSeq a smidge.  You'll see that I've added frames to the sequence just for the time delays.  I got to thinking about the order of execution in your vi.  The way it was drawn, I'm not sure how Labview would decide what to do first.  It maybe that you are encountering some sort of race condition.  That might explain the difference between running on PC and on cFP.  Worth a try.

Try the attached vi and see if anything changes.

[Scott93535]

OK, I'll give it a try.  I also made another mod getting rid of the sequence structure all together except for the time delays.

Thanks

Scott

[Scott93535]

Ok, I discovered some things about Fieldpoint that does not seem to be intuitively obvious.

1) The vis I was running had two points in the execution there were experiencing problems.  The only thing they had in common were sequence statements and they were the onyl place the sequence statements were being used.

2) After eliminating the sequence statements in favor of a while loop case statement combination almost all of the problems went away.

3) Issues that remained were timing problems.  It looks like the labview running on the Fieldpoint controller has issues with "busy" statements, i.e. a case with a single time delay or I/O point toggle works fine but if you put both functions into the same case the controller has trouble determining which one to execute and in what order. 

4) My program has one last This evening I am going to try to break down the last portion of my code that the controller may find ambiguous and see it the last timing issue is resolved. 

Scott

[Wayne.C]

Scott,

I think the potential for this problem exists everywhere.   I always use a separate frame in a sequence for a delay.   That way I know exactly the order of execution.