Why does my signal drift off then stabilize again?

[NotaTech]

I have six analog inputs to read.  They are all in the 0 to 5 Vdc range.  Five are from potentiometers, and I'm not having any trouble with them.  One is from an Omega thermocouple signal conditioner (DRF-TCK) which has a 0-10 Vdc output, but my temperatures will keep it in the 0-5 range.  The weird thing is, it will run stable for a minute or so at around 800mV then suddenly drift off to over 2V, then it wanders around and eventually stabilizes back to something reasonable.  When it is stable, it agrees exactly with my volt meter and when it is wandering it does not.  So the signal is good and the reading is in error.

[Captain_K]

Hi NotaTech,

A lot of the times, drift is due to input configuration to the DAQ device. Check out the Field Wiring and Noise Considerations for Analog Signals. There’s a lot of good information in there for how to wire up your analog inputs.

http://www.ni.com/white-paper/3344/en/

What NI hardware are you using and what pins are you using for your inputs?

[NotaTech]

Hi Captain_K,

As my screen name implies, Im not exactly qualified to be doing this.  So I really appreciate your assistance.

The board is a PCI-M10-16E-4

My potentiometers are all connected to the +5V terminal with the wiper on AI 1 through AI 5 (pins 33,65,30,28,60) and I have a cap between each wiper and ground.  The problem child (signal conditioner) is connected to AI 14 (pin 58) on the + side and AI SENSE (pin 62) on the - side.

I'm about halfway through the white paper you referenced, and so far a solution hasn't jumped out at me, but I may yet find one.

[OKors]

Is your signal noisy at all?
Use the differential ground when possible, not the AI Sense.
Re wire your inputs and report back. I have a feeling its not grounded properly

If you're bored, this has some pretty useful info.

http://www.ni.com/getting-started/set-up-hardware/data-acquisition/analog-voltage