Thermocouple wire diameter

Question

I have four possible scenarios for making a type T thermocouple:

1. use a 24AWG copper wire and a 24AWG constantan wire
2. use a 20AWG copper wire and a 20 AWG constantan wire
3. use a 24AWG copper wire and a 20AWG constantan wire
4. use a 20AWG copper wire and a 24 AWG constantan wire

Based on the answer I received here, I'm under the assumption that I can mix and match different wire diameters (while holding wire material constant) without compromising the data since it is the Seebeck coefficient of the metal and not the resistance of the wire that determines how it operates as a thermocouple.

Will all 4 of these combinations give me accurate and similar data? If not, why?

Answer 1

Yes, they will all give similar results if the junctions are at the same temperatures.

Ideally you want to use thin wires to avoid the wires drawing too much heat from the junction, but also want the wires to be reasonably rugged and not too high in resistance. There is typically a straightforward trade-off between electrical and thermal conductivity (Wiedemann-Franz law). For example, we might use AWG8 (3.3mm diameter) type K wires for a kiln application because thinner ones won't last acceptably long.

Constantan has a much higher resistivity (and therefore a lower thermal conductivity) so I can see as you might, in some rare circumstance, want to mix a thicker Constantan wire with a thinner copper wire, however unless the wires are supplied with defined limits of error for use as a thermocouple you may lose on the accuracy.

Answer 2

You should consider the high thermal conductivity of copper. If you use a very thick copper wire, much heat is transfered away from the tip of the thermocouple. So the tip is colder than it should be to measure the temperature of your measurement point. You measure less temperature than the real temperature.