Thermistor noise issue

[limata]

We have a thermistor in a bridge configuration for ultra low noise temperature measurement. Recently  we noticed that the thermistor temperature reading shows a periodic noise and  the noise signal has been related to the activity of data logger. When the data is logged, the mcu frequencies jump to 13 Mhz and some extra noise signal shows up. When the data logger is in idle mode ( 400khz? ), no extra noise is observed.  I have been modeled the thermistor as a temperature dependent variable resistor as a first order approximation and clearly this is not enough to fully explain this.
Anybody can give some insight about a semiconductor NTC thermistor. How the EMI can be coupled into the thermistor and generate noise?
Thanks

[pipe]

This would be much easier to answer if you posted a schematics, with all the details with regards to the power supply bypass, how the bridge is read with the logger, etc. Instinctively I'd say it's much more likely to be an issue with the MCU drawing more power from the rails, and injecting noise that way. Operational amplifiers aren't very good at rejecting high frequency noise. How is the bridge hooked up to the MCU?

[SArepairman]

are you using a digital isolator and separate supply rails? ultra low noise anything typically should not be connected to a MCU.

[ejeffrey]

Also, what is ultra-low noise?  Different people have different ideas of "low noise."  What about your sample rate?

[limata]

Thanks all for the replying.
The system can resolve temperature in sub mill degree and power supply is well isolated from the analog front end. We have ruled out the power issue as you mentioned above.
The data logger is a third party parts and logged in 1 Hz. Don't recall the sample rate off top of my head. I pretty much have narrowed down to the mcu high frequency "radiation" and this only happens when the all the electronics put into the Aluminum cylindrical enclosure. It seems " radiation" from mcu or other parts has been trapped inside this  Faraday cage and eventually it was picked by the thermistor . This noise can be suppressed by a small cap across the termistor input but we still don't understand the root cause of this. 

[T3sl4co1l]

AFAIK, you're looking at the wrong approach anyway; RTDs are better, and thermocouples slightly moreso, for extremely stable temperature measurements (mK to uK).  Which figures -- they're less sensitive, but less noisy; but the overall performance is better than a thermistor (which contributes the inherent noise of the ceramic material used).

But anyway, that doesn't account for changes.  Sounds like EMC.  Crappy front end?  No RF filtering?  Common mode noise generated by the logger?

Anything else connected to the logger, like a power supply?  A switching supply will kick out more noise when producing more power.

Simplest thing to try, place a cap at the input pin to GND, maybe 0.01uF.

Tim