Question regarding thermocouples and a voltage terminal block

Only if the impedance of your meter causes significant IR voltage drop in the leads.

It may seem to be caused by the junction if you don't think about it much, but when you try to understand why a junction could cause thermocouple voltages there are no answers. What exactly is the physical effect? How does the junction know what the temperature at the other junction is? Where does the energy come from? In a 3-metal junction, where the thermocouple leads are separately spot welded to the part being monitored or where copper meter leads are attached, it is known that the third metal in the junction has no effect on the measurement - why is that?

Once you understand that it is the total temperature gradient between junctions that causes the voltage, the thermocouple actually makes sense. It is a heat engine which obeys all laws of thermodynamics; the energy which produces the voltage is driven by the conduction of heat energy from the hot end to the cold end. No temperature gradient in the third metal of the 3-metal junction means no voltage contribution.

When a temperature gradient exists in a metal most of the thermal energy transport (heat conduction) is done by movement of free electrons; higher energy electrons travel from the hot end to the cold end and lower energy electrons travel at the same rate from the cold end to the hot end. The higher energy electrons have a higher electrical resistance associated with their current flow from hot to cold than the lower energy electrons have associated with their identical current flow from cold to hot; therefore (in the absense of externally imposed current) the hot end is always more negative than the cold end in metals (or semiconductors where electrons are the majority carriers - opposite for holes). The magnitude of the effect is different for different materials, and it varies with temperature in a complex non-linear way which is different for different materials, but no problem, the results for metal pairs that thermocouples are made of have been tabulated for your convienience.

This has all been explained much better in the various references I have posted links to over the years, and I think anyone with any interest in thermocouples would benefit by reading them and/or additional reading on the subject.

I guess I just have a weakness for explanations that make physical sense :-).

Regards, Glen

That's true - but to the guy building the diff amp at the cold junction, the physics of how the wires work isn't really an issue.

I thought it was something similar to the way a battery or dissimilar rectifier might work, until I saw them making TCs from scratch by welding chromel wires to alumel wires. There's no "junction" in a (proper) weld - the metals merge and form an alloy, with a gradient.

Well, to the uninitiated, it doesn't care, because it's the "hotness" that causes the mumble mumble volts at the output! ;-)

The rest, I've snipped - you present a pretty good case - the only thing I was pointing out is that it _looks like_ it's coming from the junction - I left out the "things are sometimes not what they seem" part. ;-)

Cheers! Rich

Thermocouples 101

A length of wire, with each end at a different temperature, produces a voltage difference between the two ends.

Wires of different composition produce different voltages for the same temperatures.

The length of the wire, or the connections at the end, are irrelevent.

The way a thermocouple works is by taking two wires of different metals, connecting one end of each together, and measuring the voltage across the remaining ends. If you know the voltage, and the local temperature, you can calculate the remote temperature.

Example: If an Al wire is producing 10uV at the hot end, and a Cu wire is producing 7uV at the hot end, measuring across the cold ends results in 3uV (the difference).

So, it's not the junction that's doing the work, it's the whole length of thermocouple wire. Also, this is why you need to have the thermocouple wire reach back to the sensor board - the board needs to know the temperature of the cold end, using an extension cord puts the cold end too far from the board's temperature sensor.