Question

How does a thermocouple work?

Answer 1

A thermocouple consists of two wires made of different metals connected at both ends. The other ends of each wire are connected to the terminals of a millivoltmeter. The first junction is heated and the temperature difference between the two junctions is determined based on the reading of the emf taken from the millivoltmeter.

Answer 2

Electrons in a metallic structure are delocalized, how when heat is applied at one junction, electrons are arranged to move in one direction, counterintuitive to the notion that as a result of increase in temperature, particles will move more rapidly and randomly. Why would we use two different metals if two different potential difference values are developed?

Answer 3

@Irishboy123

Answer 4

Seebeck coefficient varies with temperature. That means that it's not the same at 20 and 26°C. So what they do is they measure the relative Seebeck coeff. usually against platinum, for different temperatures. They use those measurements to calculate the temperature by checking the difference between the two relative Seebeck coeffs.

so V = Sab (T1-T2)
where
Sab is the difference of two relative Seebeck coeff. for materials used.
T is temp, one at the voltmeter, one at the measuring point
V is voltage.

$$S_{ab} = S_{aPt} - S_{bPt} = S_a - S_{Pt} - (S_b - S_{Pt}) = S_a - S_b - S_{Pt} + S_{Pt} = S_a - S_b$$

Because measuring the absolute Seebeck coefficient is almost near "impossible" due to various other effects that also can contribute to the generated voltage and affect measuraments.

Practical Temperature Measurement by Peter R. N. Childs google books p104 onwards

Answer 5

Why, when the wire is composed of one type of material, no voltage difference is recorded.

Answer 6

It is, but you can't relate that voltage nicely to temperature because apparently there are a lot of other factors involved. In that book that you can google it is mentioned that Peltier and Thompson effects influence the generated voltage.

Answer 7

So you can measure the voltage anytime there is a temperature gradient involved, but the law is not a nice "straight" line. Instead it is a different curve for every temperature you can measure.

Instead of measuring that curve for every temperature and then creating some kind of micro controller that comes with the wire and fits the different curves and then based on the best fit similarity to fits that were measured determines the temperature, they made a passive element that costs pennies apparently and doesn't involve any electronics that should be temp resistant etc etc etc

Answer 8

Turns out there's quite a lot of interesting stuff that happens, even with about half that book censored it's still worth a look.

I'd never figure it was this itneresting really :D