Themocouple calibration question

[gomerpyle]

One method I've found is putting the thermocouple in an ice bath, then boiling water and plotting actual temperature vs. measured TC temperature. A linear curve fit is then applied between the two reference points, and used in a DAQ program to 'correct' the TC measurement coming into the DAQ.

But then I've also read about just using a conversion equation from TC output voltage to temperature, with the specific polynomial coefficients being dictated by the type of TC. I have a limited knowledge of DAQs, but assuming a program like labview could measure either temperature or output voltage from a TC, couldn't you just use the output voltage with the conversion equation to get a reading of temp? If this is the case what would be the purpose of using a calibration method as described above?

 

[Navin A S]

The calibration method described above can provide more accurate temperature readings than using a conversion equation for the particular thermocouple type. This is because the conversion equation assumes perfect linearity between the temperature and the voltage output of the thermocouple, which is not always the case in practice. By calibrating the thermocouple using ice and boiling water, it is possible to account for any non-linearities in the thermocouple's response. The calibration data can then be used to more accurately convert the output voltage into a temperature value.