Funny Voltage Reading

[ETULK]

Just asking if anyone has any advice for my current problem:  Within my oven I have three metal samples with a 3.00A constant current running across them.  The voltage is monitored across a crack in the samples and as the crack grows, the voltage increases (indicating cracking).  The voltages are all measured with a Keithley DMM wired up to Labview 8.5 

In general, when the temperature in the oven is increased, the voltage across every crack tip increases at similar rates.

In general, when I increase the current from the sources, it results in increase in voltage across every crack tip.

I also have one other oven that is running this program successfully.

One of my samples' voltage signal is showing rather odd behaviour, that, in my experience, is not associated with a loose wire.

When I increase the current, the voltage increases (as usual)

When the temperature increases, the voltage decreases at the rate that all the others increase.

When I reduce the current to~0, the voltage goes to a negative number

When I reverse the polarity of the power supply or the leads, I get a larger negative number than the corresponding reverse voltage.

In summary, it appears the voltage is reading correctly, but a) there is a negative offset and b) increase in temperature resistance appears to work in opposite to what would be expected.

I have measured this odd signal on multiple channels (ruling out dmm or computer programming error).  I have tried multiple power supplies.  I've been careful with the spot welds and wiring..

I was just wondering if anyone had any suggestions or comments (anything!)... and I can give more details.  

Thanks,

ETULK

[Drew_T]

ETULK,

This is indeed quite bizarre behavior, but I would guess that it is something relatively simple.  Either that or the laws of physics have changed.  

Anyways, I would double check that a custom scale has not been applied somewhere in the task, or the programming, or if custom scales are being used, that the values are still correct.

Are there differences in equipment or programming between the two ovens?  Have you tried switching the leads of the three samples?  If the problem follows one sample, the problem is with the leads/sample.  If the problem stays with the channel, it is more likely a DAQ hardware or programming problem.

Hope these troubleshooting steps help!  Let us know if they help narrow the issue or if you resolve the problem.