0.001degC resolution

[shantanu@india]

Hi,
I need to measure water temperature with that kind of resolution.
Range within 90 deg C.
Can Pt100 do the job?
What ADC can give stable readings?
Length of sensor cable would be 2-3 meters.
Thanks.

[RGV250]

Hi,
PT100 only has 0.1c resolution, when I googled "highest resolution temperature modules" I got this https://www.analog.com/media/en/news-marketing-collateral/product-highlight/TemperatureSensors.pdf
which states a resolution of 0.0078c so I am not sure if what you are trying to do is even achievable.

Bob

[TimB]

Look at the LTC2986 it will provide the resolution but since nobody will be able to verify it you can claim it is accurate to that.

[trastikata]

Do you need absolute accuracy 0.001 degree or relative accuracy 0.001 degree, or simply resolution of 0.001 degrees? Three very different things.

If it is only resolution - any 18+ bit ADC and analog sensor will provide it.

[shantanu@india]

Thanks Tim.....LTC2986 seems to be a good choice .
@trastikata....actually I need resolution , not absolute accuracy. It's the differential temperature across a centrifugal pump which is used for thermodynamic efficiency measurement.

[John Lawton]

If you need to detect differential temperature, how would you do that with two sensors that have different characteristics?

You could connect them in a bridge arrangement and then amplify the difference i.e. the bridge output. Having well matched sensors would be a requirement if you need to do this over a wide temperature range, otherwise just a bit of hardware trimming is required to give a nul at a given temperature.

John

[ken_k]

Hi Shantanu
I agree with John.
Bridges can be your best friend when dealing with differential temperatures. 4 x PT100's in a bridge can work very well.
For some applications the thermal inertia of the measuring devices and the thermal conductivity of the connecting leads can be a real issue.

I remember running PT100's in a fluid stream with their leads enclosed in tubing for some distance so the connecting leads were at the at same temperature as the PT100.

For some applications at higher differential temperatures two thermocouples can be wired in series such that their voltages cancel, a positive output indicates higher temperature on one side a negative voltage will indicate higher temperature on the other.

Some of the high gain zero drift opamps (< 4 microvolts offset) can give surprisingly good results.

I don't know if I have added anything you are not fully aware of.
Ken Kranz

[shantanu@india]

Thanks John and Ken.
Yes truely said.... absolute temperatures don't matter.
I've to think of connecting two thermocouples in series with polarity revered