Call to FPGA VI reboots computer

[Drewski]

Hey All,

I've developed a custom serial interface on a PXI7831R. Intermittently a call to the FPGA vi causes the computer to reboot. One instance of when this occurred recently. The FPGA vi is used to stimulate a product that is undergoing Environmental Stress Screening or ESS. Placed in an 'oven' that is controlled via GPIB, the product is powered on and tested 3 times at the end of a hot soak and again at the cold soak. Each hot and cold soak is called a cycle. The LabVIEW software ran for 4 cycles or 24 calls to the FPGA vi with out any problems. Then, on the 5th cycle, the software caused the PXI rack to reboot on the first test.

18 slot PXI Chassis
8187 PXI Controller, Windows 2000, LabVIEW 7.1, latest everything
2503 switch card
7831R FPGA card

There is a TTL level to 422 level voltage shifter chip from Maxim that uses 5 volts from the FPGA card. Testing shows that at most the chip draws 0.054 Amps, way below the supposed 1 Amp available (per the manual).

We do use interrupts and we tried the NiViPxiK.sys patch, but that caused a crash on every use, made things worse.

Any ideas?

Thanks,
-Andrew

[GValdes]

Hi, Andrew,

As I understand, your whole system is rebooting as you are interacting with your FPGA VI. I have a few questions that might help narrow down where the problem is:
- What version of NI-RIO do you have?
- What version of NI-VISA do you have?
- Do you get the same random reboots if you run a different program?
- Do you get the same reboots if you run a different program that doesn't use interrupts?
- Under what environmental conditions is your system operating (I'm specifically concerned about the temperature)?

Since you are using LabVIEW 7.1, I will recommend not using the NiViPxiK.sys patch; it was for a previous version of LabVIEW FPGA (1.0). Hopefully, you have a backup of the old file that you can recover.

Thanks for this info. I'll start researching as soon as I hear back from you.

Regards,

GValdes
National Instruments

[Drewski]

Hey G,

We're running NI-RIO 1.1, NI-VISA 3.2f1.
>Do you get the same random reboots if you run a different program?
No other software random reboots have occurred. As the equipment is highly visible in the company, any random reboot with any program would be immediately reported to me.
>Do you get the same reboots if you run a different program that doesn't use interrupts?
When I used a polling version of the FPGA software, I did not have any random reboots, however, I have not run the polling version anywhere close to the duration of time that the interrupt version has been running. I was having a performance issue during the FPGA code development cycle that I initially thought was related to polling verse interrupts, but was actually an unrelated memory allocation issue. I am under the impression that interrupts are faster, so after having switched the FPGA code to interrupts, I don't really want to switch back to polling.

As for your question on environmental conditions, I am not sure how to answer that question. The equipment is in a rack, the fan on the PXI chassis is set to auto. The other fans on the rack run at a constant speed. The area the rack operates in is normally high 60s to low 70s in degrees Fahrenheit. Humidity is higher than outside to lower ESD issues.

We did go back to the previous version of the NiViPxiK.sys file.

I also wouldn't call it a random reboot. The 7 documented cases have all occurred when the software is calling the FPGA code.

Update: The ESS program has run four times to completion since the initial post.

[GValdes]

Hi, Andrew,

It seems that you have well documented the case when your system reboots. Would you be willing to put your code here so I can try and replicate the hardware settings to see if I get the same behavior? If you are worried about publishing your code in this public forum, please call to National Instruments Technical Support so we can help determine what is the cause of this reboot. There shouldn't be a problem with the drivers or the hardware that you have, so I will be interested in seeing if something in the code is causing this.

Please let me know what you think.

Regards,

GValdes
National Instruments

[Drewski]

Thanks for the offer, but I'm not comfortable sending out the code. The purpose of the post was to see if other users were experiencing similar issues, and it looks like they are not. Haven't had another crash for two months now, and hundreds of calls to the FPGA have been made. I am going to stop worrying about it now.

When it does re-occur I'll pull the lever to start the process and paper work to allow the code to be sent out for support.

Thanks,
-Andrew