Maker Pro
Maker Pro

T Type Thermocouple for delta T measurement

Kamalakkannan

Jun 10, 2014
3
Joined
Jun 10, 2014
Messages
3
Hi,
I am working on an experiment in which I am expected to measure delta T values as low as 0.35.
The only resource available for the measurement is T Type thermocouple.
I would like to know if anyone knows how to wire up a T Type thermocouple to measure the delta T, what equations to use to get voltage to delta T. what would be the uncertainity etc.

Please share any ideas that you may have.

Thank you
Kamal
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,505
Joined
Jan 21, 2010
Messages
25,505
You should Google for "t type thermocouple differential".

I did that and found a really great pdf that I cant paste a link to.

Why don't you try it yourself.

As a general rule, you'll use two thermocouples wired to sense the difference in temperature.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
What do you mean by delta T. Do you wish to measure the temperature difference between two position or the temperature change at one position.

To measure the temperature difference of two positions, put a junction in each position and balance one voltage against the other.
To measure a temperature change, you will need a cold junction reference and some form of recording, pencil and paper, computer etc.

You can look up the sensitivity of a T type thermocouple and calculate the voltage you will get. For under 1K measurements, you will need a very accurate measurement system with low drift.

Edit The sensitivity is 39µV/K so 0.35K will give only 13.7µV.
 
Last edited:

Kamalakkannan

Jun 10, 2014
3
Joined
Jun 10, 2014
Messages
3
You should Google for "t type thermocouple differential".

I did that and found a really great pdf that I cant paste a link to.

Why don't you try it yourself.

As a general rule, you'll use two thermocouples wired to sense the difference in temperature.


Thank you for your reply. Some more questions before I actually start making the changes in my experimental set up.
Please correct me if I am wrong, I have been doing a lot of googling through the night and this is my take,

I understand that the two junctions will be at the two points,(like a M shape); the net result of this will be to read a voltage which would correspond to the differential temperature.
Now the wires that come out can no longer be connected to the data logger as temperature measurement because the two ends would be of the same material Cu, in the case of T type.
Also, we don't want the data logger's inbuilt algorithm to do any cold end compensation.
Thus the wires would be connected to the data logger as a voltage reading.

How do I proceed from here? Should I collect the voltage data and put it into an excel file and use the eight order polynomial to convert the voltage to temperature? And what would the accuracy of this reading be? What is the general method to arrive at the accuracy of such measurements?

Thanks in advance
Kamal
 

Kamalakkannan

Jun 10, 2014
3
Joined
Jun 10, 2014
Messages
3
What do you mean by delta T. Do you wish to measure the temperature difference between two position or the temperature change at one position.

To measure the temperature difference of two positions, put a junction in each position and balance one voltage against the other.
To measure a temperature change, you will need a cold junction reference and some form of recording, pencil and paper, computer etc.

You can look up the sensitivity of a T type thermocouple and calculate the voltage you will get. For under 1K measurements, you will need a very accurate measurement system with low drift.

Edit The sensitivity is 39µV/K so 0.35K will give only 13.7µV.



Thank you for your reply. Some more questions before I actually start making the changes in my experimental set up.
Please correct me if I am wrong, I have been doing a lot of googling through the night and this is my take,

I understand that the two junctions will be at the two points,(like a M shape); the net result of this will be to read a voltage which would correspond to the differential temperature.
Now the wires that come out can no longer be connected to the data logger as temperature measurement because the two ends would be of the same material Cu, in the case of T type.
Also, we don't want the data logger's inbuilt algorithm to do any cold end compensation.
Thus the wires would be connected to the data logger as a voltage reading.

How do I proceed from here? Should I collect the voltage data and put it into an excel file and use the eight order polynomial to convert the voltage to temperature? And what would the accuracy of this reading be? What is the general method to arrive at the accuracy of such measurements?

Thanks in advance
Kamal
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
You are correct that you do not need and should not have any cold junction compensation.

To get temperature just take voltage (µV) and divide by 39.

To get an idea of accuracy, measure the resistance of the thermocouple and find a resistor somewhere near. Pass a known current through it and measure the voltage across it. Calculate what the temperature corresponds to.
This depends on the thermocouple being accurate. You could check with a standard set up with the hot junction in steam at 373K and the cold junction in ice at 273K. You should get 3.9mV.
 
Top