Type J Thermocouples- Does resistance change also ? I understand the mV change .

[Cydog]

Good Morning ,

Just wondering . I understand that millivolts
change when temperature rises . Does resistance
change also ?

Thanks in advance .

 

[danw]

The resistance of thermocouple wire is signficantly higher than the resistance of copper wire (except for Type T which is copper alloys).

The resistance of a 'good' T/C junction does not change over a temperature gradient, where good is the condition of a new thermocouple whose junction has not become polluted.

At high temperatures metallic ions migrate from the environment (the protection sheath or tube, contact surface or atmosphere) and pollute the chemistry of the junction, which increases the resistance of the junction.

When a junction gets polluted, it is no longer the same Type thermocouple as it was initially; it is now its own (unspecified) Type thermocouple (whose junction is some random chemistry of alloys). The bad thing is that the mV it generates at a given temperature is an unknown - the pollution of the junction causes the thermocouple to 'drift' because it is still a thermocouple, but there is no longer a fixed, known, established relationship between mV and temperature because the relationship is based on the two original dissimilar metals, not the chemistry of polluted junction.

There is no way to compensate for the drift error a polluted junction, but a polluted junction can be detected because of its increase in resistance.

Some vendors who provide direct connect thermocouple inputs on their AIs in transmitters, controllers or AI modules detect a drift situation by periodically measuring the junction's resistance by driving a current pulse through the T/C and measuring the resulting IR/voltage drop.

 

[GaryS]

One word of advice about working with thermocouples
If you already know this then is a reminder and it never hurts to refresh
always us the same type of thermocouple wire back to the controller or PLC input
the exception is if you use a compensated terminal block
you have o idea how many times I got a service calls because the installer ran copper thermostat wire from the thermocouple to the PLC.
Each junction will cause a temperature offset in the readings that changer with the change in junction temperature
the thermocouple input has a temperature compensation built in

 

[lfe]

At high temperatures metallic ions migrate from the environment (the protection sheath or tube, contact surface or atmosphere) and pollute the chemistry of the junction, which increases the resistance of the junction.

I have installed temperature probes with ceramic sheath and thermocouple type S or B (both with Platinum) up to 1600ºC. If the sheath is good quality, the thermocouple lasts a long time.

 

[Bit_Bucket_07]

Good Morning ,

Just wondering . I understand that millivolts
change when temperature rises . Does resistance
change also ?

Thanks in advance .

You mean "does the thermocouple's resistance change as the measured temperature changes". Right?

Good question. I feel like I should know the answer based upon some power theory, but it's escaping me at the moment. I'd think it would be difficult to measure the resistance of a circuit with an external DC voltage source already connected to it, and that's what the TC junction is.

Most metals do respond to temperature changes by changing their resistance. That's the principle behind RTD temperature sensors.

 

[GaryS]

To add to Bit Bucket a Thermocouple is a DC Source it's self.
the output voltage changes with the junction temperature
so measuring any resistance would be difficult at best.

 

[drbitboy]

Pushing a known current will yield a measured voltage, which will be offset by the temperature-dependent voltage of the TC junction, plus that of the leads.


Even if compensation for the offsets is problematic, a long-term trend of the result for a fixed set of measurement parameters (current, TC junction temperature, test setup, etc.) may still be useful even if the actual resistance value is not known.


Someone else already suggested this in an earlier post.


It's like measuring the temperature of a CPU: we cannot accurately measure the temperature of something we are actively cooling, but the trend over time is useful (e.g. the "measured" CPU temperature rises over the long term as heat sink thermal paste degrades or leaks away over time).

 

[GaryS]

Look this up
Peltier effect to create a heat flux at the junction of two different types of materials. A Peltier cooler, heater, or thermoelectric heat pump is a solid-state active heat pump which transfers heat from one side of the device to the other, with consumption of electrical energy, depending on the direction of the current.
So passing any current through a thermocouple junction will in effect create a heat pump
one side of the junction will heat up the other side will cool down
this is done all the time
https://www.thermoelectric.com/

Why would want to measure the resistance of a thermocouple that makes no sense
the mv output will tell you the temperature if there is 0 volts then the TC is open
you also need to understand cold junction compensation to get the most accurate reading

 

[danw]

When a thermocouple fails and breaks and becomes an open circuit, then that condition has historically been fairly easy to detect and indicate by driving the indicated reading/temperatuer upscale or downscale; the thermocouple 'break' function that every direct-connect AI has. The T/C break function doesn't need a good junction resistance measurement.

But there is indeed a valid reason for an AI to periodically measure and monitor the junction resistance to detect a drift condition. A drift condition is NOT evident because the thermocouple is still providing a mV signal and the AI is still interpreting that mV value according to whatever Type thermocouple the AI is programmed for. But the interpretation is not valid; the interpretation is providing a temperature value that is incorrect; it is not the true temperature.

Those AI's that do a resistance measurement use a flag or for an alarm that can be used alert the system/user to the fact that the state of thermocouple is suspect and that thermocouple should be replaced.

For those owners whose thermal process requires some degree of accuracy to provide a desired result, knowing that a thermocouple has drifted can be of value, because otherwise they're in the dark as to the validity of the accuracy of temperature values.