PXI signal routing for synchronization

[bigg_jon]

Hello,

Today has not been a productive day.  Originally I had a problem with my PXI chassis not being recognized correctly.  See this message http://forums.ni.com/ni/board/message?board.id=270&message.id=3822.

Now, my devices are recognized correctly and I am trying to synchronize the AI and AO on a PXI-6259.  I do this by specifying AO/SampleClock as the source for the timing VI for the AI task.  No errors are generated, but it appears that the SampleClock doesn't make it to the AI timing engine.  No samples are acquired.

I opened up the Multi-Function Synch AI-AO.vi example which uses the AI StartTrigger as the StartTrigger for AO.  This works.  However, if I try to use the AO StartTrigger signal as the StartTrigger for AI, AI never gets started.  I've made sure that the AI Start Task VI executes before AO Start Task.

I guess what I'm getting at is that certain signals seem to route fine, while others don't route, but don't generate an error.  According to the routing table in MAX, both the AI/SampleClock <-> AO/SampleClock and AI/StartTrigger<->AO/StartTrigger are indirect routes through the TriggerBus.  Currently with my testing, only the AI/StartTrigger signal is making it the the AO/StartTrigger system while the other signals aren't being routed.  Is this normal?  Is my PXI backplane somewhat fried?

Any help would be great.

Thanks,

Jon

[bigg_jon]

Some more information:

There are four different synchronization scenarios I tried

     1.  Using AO/SampleClock as the SampleClock for the AI Task

     2.  Using AI/SampleClock as the SampleClock for the AO Task

     3.  Using AO/StartTrigger as the StartTrigger for the AI Task

     4.  Using AI/StartTrigger as the StartTrigger for the AO Task

Scenario 4 is the only one that worked as expected.  With Scenario 1 and 3, it appears that the AI task never starts.  With Scenario 2, it appears that the AO Task never starts.

My setup is:

LabVIEW 7.1.1

PXI-1042 Chassis using PXI-8310 Laptop Control Interface

Slot2: PXI-6602

Slot3: PXI-6259 (the device I'm trying to use)

Slot4: PXI-8420

Slot5: PXI-8431

Slot6&7: PXI-5401

Slot8: PXI-4072

I just upgraded to DAQmx 8.5 and VISA 4.1 to see if that helps, but it doesn't.  I'm still having the same problem.  At this point I'm thinking that something broke in the PXI backplane, but if anyone has any ideas or has seen this before, please let me know.

Thanks,

Jon

[Matt_A]

Hello Jon,

I would tend to doubt that there is a problem with your PXI backplane because, as I understand it, you are trying to route signals on the same device. If you are sharing signals on the same device, and there is a direct route shown in MAX (as pictured below) then the signals are routed directly on the device and do not go through the PXI backplane.

What I think is more likely is that the device routes may not be entirely configured correctly. I have modified the Multi-Function-Sync AI-AO.vi example that you referenced earlier and it works on my system. An image of the changes is attached below. Note that you have to change both the order in which the tasks start and the start trigger terminal mapping. These changes are circled in the diagram.

Can you modify the example as shown below and run it on your system? I hope this helps.

Message Edited by Matt A on 04-10-200701:35 PM

Message Edited by Matt A on 04-10-2007 01:38 PM

[bigg_jon]

Thanks for the info Matt.

This is interesting.  I hadn't noticed that AI/StartTrigger -> AO/StartTrigger was a direct route.

According to my system, AO/StartTrigger-> AI/StartTrigger is indirect through the TriggerBus as well as AI/SampleClock <-> AO/SampleClock.

Again, I believe this points to something being dead in the backplane as the AI/StartTrigger to AO/StartTrigger routing worked while none of the others did.

I'll see if I can steal a chassis from some other projects to try out.

Thanks again,

Jon

[bigg_jon]

Well, I tried another chassis and all four synchronization methods mentioned above worked.  I believe it's time to send the chassis back for an evaluation.

Jon

[Matt_A]

Hello John,

It does sound like there is a problem related to routing the indirect signals, but I don't know if this would necessarily indicate a damaged backplane. Are you routing any other signals through the backplane besides the triggering & timing signals for the NI PXI-6259? If you are attempting to use the same line for multiple routes this could cause a problem. I can't think of how the PXI backplane could be damaged so that it would affect a route between different signals on the same device, but I suppose that could be the issue. Let us know how things work out.

[bigg_jon]

Hi Matt,

I haven't been trying to route any other signals over the bus.  The only thing that I changed when this problem occurred was the addition of the PXI-5401 arb to the chassis.  I decide to remove it to see if that somehow affected the issue, but ran into some trouble with MAX recognizing DAQmx devices.  So I tried the fix for the MAX data directory listed in the KB somewhere (sorry, I don't have the article handy) and it seemed to work.  The devices appeared.  Now this morning, I fire up the laptop, and again the devices are not visible in MAX.  I tried a repair installation of DAQmx 8.5...no dice.  Then I tried uninstalling and reinstalling DAQmx.  Now MAX crashes every time I open it.  So it looks like I completely blew up my software installation.

Now I'm in the process of uninstalling/reinstalling MAX.  Hopefully this will make the devices visible again.  I'll keep this post updated with an progress (forward or backward) that I might make.

Thanks for your help,

Jon

[bigg_jon]

Well it appears that you were correct, Matt, that it wasn't the chassis.  I have gotten the routing to work again.  I'm not sure what exactly corrected it because I have a few things changed.

1.  I reinstalled MAX, DAQmx, VISA to get the devices recognized again.

2.  I removed the PXI-5401 from the chassis.

I'm not sure if the 5401 could some how reserve the trigger bus.  The problem began when I added it and now it is ok without it.  But I also pretty much did a full driver reinstall.

Anyway, I'm a little gun-shy about re-adding the 5401 card because all my problems started with it.  Is there any known reason why it could cause some problems?

Thanks,

Jon