change in PC hardware results in Error -200018???

[bp99]

Dear LabView helpers,

I'm stuck with a very counterintuitive hardware problem and I'm hoping that maybe somebody has seen something similar before:

We have a PXI-6259 in a PXIe-1073 chassis.

The analog output works perfectly fine at 1.2MHz when connected to a Dell Precision T3610.

However, when the same chassis and the same card are connected to a Dell Precision T5810, the analog output only works to a maximum of 800kHz. When I try anything faster than that I get "Error -200018:  DAC conversion attempted before data to be converted was available."

I have tried it on two different T5810 and it shows the same behaviour on both.

I've also used two different sets of PXIe-1073 chassis and 6259 card to confirm that it is a more systematic issue.

The T5810 has recently superseded the T3610, but according to Dell the two models supposedly use the very same chipset and the PCI port functionality should be exactly the same on both models.

The test conditions are as similar as possible on all the PCs, ie. they all run Windows 7 Professional 64bit with NI Device Drivers 2014.08. The analog output frequency was assessed using the test panel in NI-MAX 14.0.0f0.

Has anyone observed anything like it before or has any other ideas of what I could try to overcome the issue?

Is there anything obvious that I have missed?

Thanks for your help.

Best,

B.

[JÞB]

From your discription my first guess would be to see what BIOS changes are involved between the two machines.  Dell would be able to tell you more on that.

[bp99]

OK, so apparently the chipsets aren't actually identical, the old one is Intel C600, the new one is Intel C610. We're still talking to Dell to find out more differences between the two systems.

On recommendation from NI I've also updated the BIOS, but to no avail.

If anyone has any further suggestions, that would be greatly appreciated.

[VintageWillis]

Have you tried it in any other PCI-e slots on the T5810? Also what size is the slots avalible? e.g. xN

James

[AlexM1989]

Thanks James,

I work with the original poster. We're using a x1 PCIe slot (the smallest type), but i've tried all the avaliable PCIe slots in the T5810. The PCI/PCIe bus arrangment looks identical to that in the T3610 and we've always used the PCIe bus we're using now. We have also tried it on two different T5810s.

Best,

Alex

[VintageWillis]

Ok what version of DAQmx are you running?

[Sam_Sharp]

I would speak with NI - I seem to recall that there might be some hardware incompatibility between their PCI cards and certain makes/models of PCs and I seem to remember Dell being mentioned. We've also had to have some of our industrial PCs upgraded with a new BIOS to work with the PCI DAQ cards.

[ssomnath]

We are getting the same error code with a new Dell Precision T5810 connected to a PXI6124 in a PXI1073 chassis. I have tried all combinations possible and also think that it may have something to do with the computer (bios setting). I have also tried different values for the PCIe bus speed, generation of the slots, etc. but nothing has worked. Is there a solution to this problem?

[barriboy]

For the sake of posterity, I'll post a possible solution I found. Skip to the bold for the quick version.

We had an extremely similar problem to that detailed above, but this came after our switch from a Windows 7 machine to a Dell 3930 rackmount WS running Windows 10. We are using a very old PXI 6723 card, which has a PCI counterpart. The new system does not support PCI, so I have a PCI external chassis with a PCIe interface.

The problem was highly intermittent, happening in bursts spread out with several weeks in between them. It had something to do with data being received from other PXI cards, as we'd only get the problem when one of these other cards would trigger, but again, only intermittently. Eventually this became a problem we couldn't ignore, as one day it started happening every time that other card triggered, a necessary part of our data pipeline.

I have no idea what changed and caused the issue to become so much more pronounced so suddenly, as we've been in a period of relative downtime and have made no changes in weeks.

A hack solution that got us operational but I was unhappy with was to merely reduce the AO freq from 250 kHz to 200 kHz. However, I worried that this would create propagating errors or other problems that would be difficult to see, let alone diagnose.

I had a look at this guide which is about different hardware, but was clearly related. Neither of the middle conditions applied to me (periodic or programatic AO). The final section led me to this article detailing BIOS limits on PXI systems, and I used the Root Bus Detection Tool and saw that I was limited to 128 buses. I was skeptical that this was my problem, but I went into the BIOS to try to play with it.

Solution:

In the Dell BIOS I saw an option for a PCIe Link Speed that was defaulted to Auto. The BIOS informed me that the other options were Gen 2 and Gen 1. While auto will not force card to operate at higher clock speeds, it states rather vaguely that Gen 1 "changes the device capabilities register to only use the 1.1 version of the capabilities list. Adjusting this setting might improve compatibility with older PCIe add-in adapters."

I set to Gen 1, rebooted and moved back to my original AO speed of 250 kHz, and thus far I haven't had any problems. I will update if I run into problems again, especially as we stress test the system over the next few weeks.

One thing to note is that we run three monitors from a PCIe NVIDIA P620 GPU, and changing to Gen 1 has NOT caused any issues with that.

[RavensFan]

Thanks for posting such a well written message to a rather old thread.

That is just the kind of information that just might help someone else in the future if they have a similar problem.