PT100 reading circuit with high resolution

Status
Not open for further replies.

Dunke

New Member
Hi

Im designing a bigger regulation circuit including a PT100 temperature sensor. The readings from the PT100 have to be very precise, and I THINK Im looking for a ready-to-use solution with a constant current source, some sort of amplifier and a high resolution ADC. Are there such devices on the market or should I be looking for a whole other kind of solution? The regulation circuit has to maintain a constant temperature of +-5-10mK(at a span from -200 to +600C), so I think the resolution has to be at minimum 1mK.

Hope you can give me some advices

Regards
 
- First, I would use PT1000 (1/10 DIN 60751, class B to be more specific) because the temperature slope is bigger, so you get better resolution. And also the self heating is much lower, making the sensor more accurate also. Use very accurate constant current source, less than 1mA.

- 1 millikelvin resolution is very hard to achieve. Especially on that temperature range. In theory you would need at least a 23bit ADC. You can get 24bit ADCs, or you can use a lower resolution one and oversample.
Also the calibration of such a device is not easy nor cheap..
 
Last edited:
Thanks for the reply!

Im doing an internship at a calibration company(Teknological Institute), so it doesnt have to be cheap - it "just" has to be good! Right now we have a Fluke Hart 2100.
I was looking at 24 bit ADCs or maybe even 32 bit ADCs. Is it possible to get the needed resolution if I connect the ADC directly to the PT100 or should I put some kind of an amplifier in between the two? Do you know if it is possible to buy a package with current source and ADC made for the PT100 sensor? My plan was to take the ADC output into a µController and further on into a PC.
 

This is not an easy subject.. because of the accuracy and temperature range. Have you read this article: https://www.maximintegrated.com/app-notes/index.mvp/id/4875
EDIT: To get 24bit accuracy, your Signal to Noise Ratio needs to be at least -150dB.
 
Last edited:
Look at the MAX1454 although it can't be a ready solution of what your trying to achieve. e.g. YOU are developing cutting edge stuff. The MAX1454 might be a place to satrt though.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…