Need help: Best field wiring configuration using NI 9215, 9263 and Celesco string Pot

[OKors]

I'm using a cDAQ-9172, NI 9215 and NI 9263 to acquire displacement data from a Celesco SP2 potentiometer-based cable extension transducer i.e. String Pot (following this wiring guide).

The 9263 is providing 10Vdc to the string pot and the signal out is being read by the 9215.
I have the generated DC signal and input scaled using Signal Express. So far the scaling is accurate and displacement measurements are spot on using a custom scale (linear slope)  generated from our Metrology lab.
My problem is, with the transducer at rest I'm seeing quite a bit of noise.

I realize there's some field wiring considerations and through trial and error I've been able to get the "cleanest" signal by jumpering the COM terminals of the two modules. The other issue is; for terminal configuration options, the 9263 only allows RSE, while the 9215 only allows Differential. This just compounds my confusion on how these two devices need to be coupled to the sensor.  

Based on the attached image I've created of my system wiring diagram can someone with a little more understanding of potential ground issues give me some advice?

If it helps, I also have NI 9201 and 9205 modules available to use. I just thought the 9215 was the most universal.

Thank you

And yes, I've read the white paper on field wiring but I'm still unsure of how to get the cleanest signal.

[OKors]

Heres a couple images of the noise. It appears to be AC coupled. One image is with the specimen at rest (zero load/displacement).
the other is with .25" of deflection.

PS. In my lab I also use 2 different SCXI systems, both of which output very clean stable signals. The cDAQ is the only one with the noise issue; more specifically the two modules in combination.  

Also all our "instrument" plugs have been modified to only fit the clean-power outlets which is a seperate circuit from the rest of the building. This is to insure that things like fans, etc dont get plugged into the power sources that are feeding our instrumentation.

In other words, the power coming into the chassis should not have any erroneous noise on it.

[AnalogKid2DigitalMan]

The frequency of the oscillation looks fairly low, not typical powerline interference.

In your image of the wiring- is the grey wire connected to AI- the shield? That should go to ground or possibly better yet- earth ground instead.

AI- should be connected to the COM of the AO module.

See if that helps.

-AK2DM

[OKors]

The grey wire is not a shield wire. It goes straight to the AI- terminal of the 9215. In fact something I didnt clarify which may lead to confusion in the drawing; the grey and black leads are tied together inside the housing of the string pot, so black and grey are common. There is no shield wire.
I've updated the drawing to show this a little more clearly.

I will try your suggestion by connecting the sensors AI- to the COM of the AO module.

[OKors]

I got to thinking about it and electrically, the AI- is connected to the COM of the AO module.
I still dont understand why my signal is so noisy and has this ocillation to it. Our SCXI system has a much cleaner signal from the same sensor.

[AnalogKid2DigitalMan]

Can you try a different power source instead of the AO module?

A 9V battery for example.

In looking at the specs of the 9263, the max current it can source is +/-1mA. You may be causing it to be a little unstable since 10V/10kohm = 1mA.

-AK2DM