DMM error 0xBFFA2003: Max Time exceeded before operation completed.

[Omar Castañeda]

I'm using a couple of NI-4060 PCI on an application. After several weeks running the error above started showing up.
All the measurements would be 0.0 after that.
Restarting the computer wouldn't help. The only workaround was to completely disconnect the AC power from the PS.

After discharging the PS and reconnecting the AC the multimeter would go back to be responsive.

I'm using version 2.3.1 of the DMM software.

Any ideas of what would be the problem?

[Claudia]

Hola Omar,

I have a PCI-4060 in my PC and could try to reproduce the issue you are seeing. I want to understand better what you are seeing, so I would appreciate if you could answer the following questions:

1. From your description it seems that you are getting the error above when you run your application right after boot up. That is, you boot your computer, you run your application, and the board returns 0.0 as readings and the error 0xBFFA2003. If after boot up you run your application and the board does not present this error, then your application runs fine. Is this correct?

2. Are you using a desktop PC (4 PCI slots or less) or are you using an Industrial PC (more than 4 PCI slots)? What are the characteristics of your PC (processor, motherboard, etc)?

3. What are the serial numbers of your boards? The serial number appears as a sticker on the board with a bar code.

Thanks,

Claudia L
DMM R&D
National Instruments

[Omar Castañeda]

I've seen this exact error (http://forums.ni.com/ni/board/message?board.id=80&message.id=308) somewhere else in the forums, but using a 4070.

1. Exactly. But it only has happened on a single DMM, the application requieres 2. The other 4060 works perfectly fine (as the "bad" one once I disconnected & discharged the PS then rebooted) and has not presented this error.

2. The PC is an Intel 875PBZ motherboard (3.2Ghz P4 processor) using a 2+1 redundant PS.

3. I've have to ask the customer. I currently do not have access to the equipment. It is on the client's site in another location.

Thanks,

Omar Castañeda

[Claudia]

Omar,

I think I have seen the behavior you are describing, but I first need to confirm my suspicions and that is why I need to get the serial numbers of your PCI-4060 boards.

Also, please ask your customer to run the following test. When they get the 0.0. reading and the error 0xBFFA2003, ask them to please go into MAX, right click on their PCI-4060, choose Properties and run the Test Resources. If this is what I have seen before, they will get the following error "The device is not responding to the first IRQ level". If this is the case, I have a fix for you.

I look forward to getting the serial numbers and the results of this last test.

Thanks,

Claudia L
DMM R&D
National Instruments

[Omar Castañeda]

The probably would take some time. The customer is already fetching the serial numbers, but the error only occurs once in a while. Currently the equipment has run for months (it runs 27/7) and the error has happened only a couple times.

I've already instructed the customer to call me if the error happens again so we can run the diagnostic you asked.

Anyway, I've found some of those "IRQ Level" errors before (We use 4050/4060/4070 in almost all our equipments). So if you know a fix for that please let me know.

Omar

[Claudia]

Omar,

The fix I am referring to is exclusive for the PCI-4060.

You could get the error "The device is not responding to the first IRQ level" when you run Test Resources in MAX with any of these boards (4050, 4060, 4070) under different circumstances, like if the driver is not properly installed, if there is an actual failure in the hardware, etc.

In your case, IF when you get readings of 0.0 and error 0xBFFA2003 in your application, YOU ALSO get the error "The device is not responding to the first IRQ level" when running Test Resources in MAX, then this could be pointing to an intermittent problem I saw once with PCI-4060s.

I can get useful information just with the serial numbers of your boards, but it would also help me to know the result of running Test Resources when your application has already shown the error.

Hope to hear soon from you,

Claudia

[Omar Castañeda]

Hello Claudia:

The serial numbers are:
17728061 (0x10E823D)
17661068 (0x10D7C8C)

The error hasn't occurred yet, I'll post back to you whenever it happens. The customer is already instructed on what they should do.

Thanks!

[Claudia]

Omar,

Recently I had a customer experiencing the same issue you are describing. They were using a new motherboard that supported Pentium IV processors and they were seeing intermittent failures with their PCI-4060 boards. The PCI-4060 boards would work fine in the system and sometimes they would fail after boot up not returning readings and failing the Test Resources in MAX with the error "The device is not responding to the first IRQ level", even though the board had worked previously in that same computer.

This intermittent issue was caused because the FPGA was not resetting appropriately. Some new PCs do not have deterministic control of the 5V and 3.3.V power lines in the backplane, and leave a level of voltage of a few hundreds of millivolts when powered down, which prevents the FPGA from resetting appropriately.

As you are also experiencing, they had some boards that worked fine, and some that did not. This is due to small variations in the chips themselves, and in some cases depends on the slot you install the board.

To correct this you need to send your boards for repair and reference in the problem description notes that you need to have ECO 34647 done on the board. I recommend you to send both boards for repair, to make sure you never see this failure in the board that has never shown this issue.

This is something that occurs rarely and in specific computers. Any PCI-4060 built after Nov-5-2004 already includes this fix, however your two PCI-4060 boards were built before then, even though they shipped at the end of Nov 2004.

Higher end NI Digital Multimeters (NI PXI/PCI-4070, NI PXI-4072 or NI PXI-4071) do not present this problem.

Please feel free to reply to this message if you have any questions,

Claudia L
DMM R&D
National Instruments

[Omar Castañeda]

Claudia:

Is this the fix (see attached picture)? The thing is we have a lot NI4060's on the field. And it would be really problematic for the customers to be waiting around for the fixed board.

[Claudia]

Omar,

That looks like the fix I am talking about. We could try arrange some solution for your customer so he can fix his boards without affecting his application. This would be a lot easier if we could contact each other directly.
Please feel free to call me at 1-800-433-3488 and ask to be routed to my extension.

Thanks,

Claudia Lorente
DMM R&D
National Instruments