Can a type K thermocouple be used with cryo?

[jonhswon]

Hi.

I can not understand why type K tc can not be used with cryogenic temperatures?

If for example calibrated with liquid nitrogen and ice, can it be used from 0 C to -200 C accurately?

I understand that there is a non-linear characteristic a lower temperatures, but then one could make a voltage-to-temp plot by calibrating against ice, liquid nitrogen, solid ice, etc.. and obtain a curve.

Does the inaccuracy arise simply because they are usually calibrated with ice and boiling water only, then linear interpolation/extrapolation line is drawn, OR is there something else that I am missing?

 

[f95toli]

You can use it, but thermocouples have a number of issues which means that they are typically not the preferred choice for cryogenic temperatures. Something like a standard Si diode sensor is nearly always a better choice.

See e.g.,
https://www.lakeshore.com/products/...yogenic-temperature-sensors/thermocouple-wire

 

[Baluncore]

Maybe the metals or the alloy junction are superconductors at cryogenic temperatures.

Does the Seebeck effect vanish, or the calibration change, in superconductive metals?

 

[f95toli]

Maybe the metals or the alloy junction are superconductors at cryogenic temperatures.

Does the Seebeck effect vanish, or the calibration change, in superconductive metals?

No, not at the temperatures mentioned in the OP (liquid nitrogen). There are very few metalsa/alloy that become superconducting even at liqduid helium temperatures (Nb, NbN, Pb and some alloys); there are none at liquid nitrogen temperatures.
For the type of metals that thermocouples are made you typically have to go to mK temperatures; and not all metals become superconducting even then.

 

[Dullard]

The Lakeshore link has an error (I think): The page claims a Type K sensitivity of 4.1 mV/K @ 20K. I think they meant 'uV.'