Optimizing Thermocouple Connections for Accurate Temperature Measurement

[johng23]

I want to measure the temperature of a metal stage where I am holding a sample. What I have done is to clamp both wires from a thermocouple against the stage using a bolt and washer. Supposedly this is not the best practice but I don't see why this would be a problem.

I have read that you are supposed to weld the thermocouple wires together at one end, at the location you wish to probe. What is the purpose of this? It seems to me that as long as the two wires are at the same temperature, and are in electrical contact, the thermocouple will work.

Welding the wires will decrease the resistance, but if the other end of the thermocouple is an open circuit for measuring a voltage, there will be no current flow so the resistance will not give rise to any spurious voltage. If there is a temperature difference between the two wires, there may be an additional thermoelectric voltage. But as long as the two wires are close together there should not be a large temperature gradient.

In any case, is there some effect I am neglecting here?

 

[PhysicoRaj]

Both the ends should touch each other. In a thermocouple, There must be a closed circuit whose one end acts as the probe. It does not matter the two ENDS are at the same temperature, as there is only one end you should use to probe(the end formed by the junction of the two wires).

 

[Jobrag]

From what you've written you've clamped the two thermocouple elements together using potentialy two other metals. I think that you'll find that you get some odd readings due to the contact with extra metals.

 

[johng23]

So, it's been confirmed that one end has to be a junction, or else the thermocouple will not work well. But I still have no explanation for why this is so.

 

[jim hardy]

you should be okay with what you described.

suffice it to say
the wires should be in contact at the sensing end
and be connected to a high impedance voltmeter at the measuring end.
Microvolts you read at meter are in proportion to temperature difference between ends. So you need a thermometer alongside your voltmeter.

http://www.omega.com/thermocouples.html

Basically a wire that has a temperature gradient along it also has a small potential gradient.
Since the two wires are of different metals their voltage gradients are unequal so at measuring end you sense a potential difference .

Try it on your workbench.
Connect a piece or thermocouple extension wire to your voltmeter, twist end together and observe ~ zero.
Then place an ice cube on end, then heat it with a match and observe opposite polarities.

 

[Averagesupernova]

The wires don't really need to be connected physically to each other. Another conductor between will work.
-
Here is and example: At a place I used to work we used a K type thermocouple with the end just stuck down in the molten solder of a soldering pot. Of course this is a special situation. The solder is conductive to complete the circuit, and the individual wires were in very close proximity with each other.

 

[NascentOxygen]

The wires don't really need to be connected physically to each other. Another conductor between will work.

Isn't that simply a case of making the thermocouple junction from another metal? (And creating another junction using that intermediate metal plus your other metal of the original pair.)

Here is and example: At a place I used to work we used a K type thermocouple with the end just stuck down in the molten solder of a soldering pot. Of course this is a special situation. The solder is conductive to complete the circuit, and the individual wires were in very close proximity with each other.

You replaced one junction with two junctions. That is quite valid.