Convert Thermocouple Reading CJC Query

[Worle]

In Convert Thermocouple Reading (The one shipped with LabVIEW in the scaling pallet) they are converting the CJC temperature back to a thermocouple voltage, adding this voltage to the measured thermocouple voltage and converting back to a temperature.

Why is the measured thermocouple voltage not converted straight to a temperature to which the CJC temperature is added (like below)

I would have thought that calculating an equivalent thermocouple voltage for the CJC temperature and adding it to the measured voltage and reapplying the thermocouple polynomial would not be a mathematically correct thing to do.

What am I missing?

[Worle]

Upon reflection, is it because the published thermocouple curves are based upon a cold junction at 0°C?

This means that it wouldn't be correct to directly apply the thermocouple curve to the measured voltage if the thermocouples cold junction was at 25°C

I would appreciate it if any one could confirm this/correct me.

[Henrik_Volkers · Accepted solution]

Niatross is rigth, the TC EMF (voltage) is a nonlinear funktion of the dT and T (See Seebeck effect) along the TC cable .

nonlinear means that

f(a+b)=f(a)+f(b) is not valid for all a and b!

The polynoms give the voltage referenced to 0°C for a given TC pairing of different conductors.

To get the correct temperature one need the voltage with the CJC at 0°C. So if the CJC is not 0°C you need that missing voltage acording to the type of TC, f(T -> V_tc0)  add it to your TC voltage reading (compensate the CJC temperature to 0°C) and apply the f((V_tc0+V_tc)->T)