GPIB hangs for a couple seconds then resumes

[M.Post]

I have an issue on one of my test sets where the GPIB is hanging for a few seconds and then resumes.  This happens intermittently and only on one of my 7 test sets (all of them identical).  I have 6 instruments connected with GPIB and all of the instruments hang during this time.  I'm running WinXP Pro and LabVIEW 8.6.1 (I know, it's a little out dated).  NI_VISA version is 4.4.  I've been using this test set for years and now all of the sudden, I'm getting this issue.  I've swapped PC, did not fix it.  I've swapped several instruments thinking it could have been one of them, but that did not fix it either.  I've run NI Spy and it confirms that there are no GPIB calls made during this freezing issue.  I'm at a lose, any suggestions?

[cstorey]

Are you using PCI-GPIB or USB GPIB?  Did you replace those?

[M.Post]

Ah, good question.  I'm using the PCI-GPIB card.

[M.Post]

I have some updates and a correction.  The NI-VISA driver is actually 4.4.1.  Also it's on WinXP (SP3).  I did swap out the PCI-GPIB card with one spare we had but got the same hanging issue.

[cstorey]

No software changes at all?  M$ .NET or other installed?  (At least you don't have Windows updates to wory about!)

The last time I had issues with a PCI-GPIB card and WinXP it was because the machine RAM was failing and it was page swapping to disk so often that the resources weren't there for any other operations.  New RAM fixed all issues.  But you tried a new PC?  So you have tried all of these?

- New PC + New PC-GPIB + all old instruments

- New PC + OLD PC-GPIB + all new instruments

- OLD PC + new PC-GPIB + all new instruments

What about cables?  Did you check cables haven't been crimped or bent?  Sometimes big racks get pushed back against the wall and a cable with get bent too far.  I'd inspect the cables and connections at all instruments.

Good luck!

[M.Post]

The combinations of changes that I've done are as follows: Note that when I say New PC, this is referring to another PC from another Test set which is just as "old' as the broken one but does not exhibit the pausing issue.

-OLD PC + new PC-GPIB + all old instruments (This did not fix it)

-New PC with it's own PC-GPIB + all old instruments (This did not fix it)

-OLD PC + OLD PC-GPIB + some new instruments, the ones I thought were suspect (This did not fix it)

As I looked into this further I did notice differences with the Test sets. The broken one is running XP (SP3) while the other good ones are running XP (SP2). Not sure if this is a big deal but I do plan to revert to SP2 when time permits.  Also the broken one has GPIB driver version 2.6.0f0 while the other good ones are running either 2.5.2f3 or 2.5.2f6.  Having said all this, I did install GPIB version 2.5.2f3 on the bad one but that did not fix the problem.

Here's the funny thing, when I have all 6 instruments turned on the issue is fixed.  When I have the one instrument that we hardly ever use turned off, I get the pausing issue.  Now before we declare this solved, the other good Test sets run fine with this particular instrument turned off.  Also this bad Test set used to work fine with this same instrument turned off.  This is really strange.

The last thing I have not done is check the condition of the GPIB cables for kinks and damage and what not.  Also the particular cable connection sequence between instruments has not been gathered either.

[cstorey]

Good job debugging!

Its suspicious that when that one instrument is off you see the error but not when its turned on. 

I suspect a cable issue more than ever now.  Bad cable somewhere, or the cable lengths between instruments are too long, or the total length too long.  I know there are guidelines for the total length and what fraction of instruments need to be powered on to maintain signal integrity...but you'd have to review GPIB specs for those however.

Good luck with the final solution.  My advice is document it all now, so you don't have to do it again in 5-10years!