LabVIEW freezing while engine running

[PKP]

Hello, I have a very puzzling problem.

My LabVIEW 8.2 program for data acquisition freezes while a test engine is running and once the engine stops, the VI becomes active again.

The system consists of a Dell Dimension PC, (P4 with 768MB RAM), with 3 PCI cards installed. There is a PCI-6280 and two PCI-6024E cards, all connected to SCB-68s. The transducers (about 24 of them) include thermocouples and pressure transducers mostly connected in Differential mode, while about 4 are in single ended connections. The transducers are measuring the temperature and gas supply pressure of an engine and the surface temperatures.

The LabVEW 8.2 based program has 3 DAQ-Assistants to read from each card and writes all the data into a single file using another Write Measurement File VI.

The sampling rates tried were 10kS/s, 5kS/s, 2.5kS/s and finally 1kS/s. The program writes continuously for about 10 to 30 seconds while the engine is tested (24 channels x sampling rate x 30 s).

The PC with the DAQ cards is located in the test lab near the engine and is connected to a LAN socket via ethernet card. It is then remotely accessed via another PC in the control room, which is seperated by a wall from the engine room.

If the engine solenoid valves are run and the ignition is also turned on but no gas is supplied to the engine, the LabVIEW program runs fine and writes data, showing that the data throughput rate is not affecting the system.

As soon as the gas lines are opened and the valves now supply fuel, air, O2 to the engine and the engine starts firing, LabVIEW locks up. Once the engine stops, LabVIEW starts running again. The only difference being that the engine is being supplied with fuel, air and oxygen and combustion is taking place. If the gas lines are closed and the valves and the ignition fires also so that no combustion takes place, LabVIEW runs fine. So it is not a matter of solenoid or ignition noise. The pressures being measured are in the gas lines, so there is no sudden pressure spike occuring during engine run time.

We even have the PC sitting in a metal box with the entrances closed and it makes no difference.

When we disconnected the cables from the board to the SCB-68 boxes, with the PCI cards inside the PC and when the program is run with the engine running, it freezes as the engine starts and unfreezes once the engine stops. Then we removed the two PCI-6024E cards, leaving just the PCI-6280 inside the PC, but not connected to the SCB-68 and now when the engine is run, the program does not freeze up.

There must be a logical explanation that is eluding us, but it just is not evident right now.

Does anyone have any suggestions? Thank you very much for your time.

Philip

[Ima U]

Hi Philip,

Thank you for a very detailed problem description.  This is a strange issue indeed.  I noticed that the only time the PCI cards work when the engine is on is if the SCB-68 is disconnected and only one PCI card is in the computer.  I am going to explore this issue further to see where the problem stems from.  In the mean time I have a few questions for you.

1.        Does the entire computer lock up or is it just LabVIEW?

2.       Is LabVIEW pausing when the engine is on or do you lose data?  Is there a gap in the dt of your data points to show that LabVIEW is not pausing?

3.       Are the computer containing the DAQ cards and the engine on the same circuit breaker?  Could the engine be possibly drawing too much power?

4.       If you were to only put in one PCI-6024E and not connect it to the SCB-68, does this work fine just the way the PCI-6280 works when it is alone in the computer?

5.       Could vibrations from the engine be causing connections to loosen? 

[PKP]

Hello Ima,

Thank you for the reply. I think we may have finally solved this riddle.

The sound intensity inside the closed test room is over 140 db (our Radioshack db meter runs out at that point.) And if you enter the room with the engine running, you can feel the thud on your body from the sound waves in the air. The sound is so strong that it was shaking the cameras on tripods and articles on the tables.

Finally, after several different failed configurations, we moved the PC with the DAQ cards over to the control room, which is seperated from the test room by a concrete wall, and laid new extension cables over the wall for the sensors. Now the PC does not lock up. It is now clear that the loud vibrations were disturbing the PC and the PCI cards, causing Windows or LabVIEW to freeze up.

Now to answer your queries, sometimes the LabVIEW program would unfreeze and start to work again after the engine was turned off. Sometimes the computer would just give the blue screen of death. In both cases, we would not get data during the freeze.

As for the ckt. breakers, no, the PC with the DAQ cards was on a different breaker than the engine and its controls.

And we tried different permutations and combinations with the PCI cards (two 6024E and one 6280) and with and w/o the SCB-68 connected. In all cases, the PC would freeze with the engine running.

The PC with the DAQ had been installed in the test room on a small cart and was open. Interestingly, a 1042-Q chassis is also in the test room, but it is inside a 19" metal rack with doors. The 1042 chassis uses a fiber optic cable to connect to a PC in the control room. We never had a problem with noise freezing up or upsetting the data acquisition process with the 1042 chassis, although we have tremendous amounts of ignition and solenoid valve EMI in the signals. It does however show that the 1042 chassis is really rugged and vibration resistant.

Regards,

Philip

[AnalogKid2DigitalMan]

Good find! I was following this thread but could not think of any suggestions to troubleshoot.

You know, high intensity sounds can break windows (apparently the Microsoft kind too)

May I ask what kind of engine is it.

-AK2DM

[PKP]

It's a small jet engine for academic study. Small but loud.

[Ima U]

Hi Philip,

I am glad to hear that the problem is solved.  I agree, this was an interesting post.  Thank you for posting in detail the resolution.  I am excited to see that you are able to take advantage of how rugged the PXI systems are.  I wish you the best in your application.