Easiest way to calibrate an air-temp thermocouple?

[Howard]

I think the one I have is not rated for immersion, so ice/boil isn't going to work. Maybe if I attach it to something that gets submerged in ice water or boiling water?

It looks like this:

http://www.industrialzone.com/watlow-70xjsgc048a.html

 

[CycloWizard]

You could attach it to the surface of a metal object. Preferably the object would be copper or something else with high thermal conductivity to make sure the temperature the thermocouple sees is very close to the "true" temperature you're trying to use as a calibration.

That said, there are better ways to measure temperatures near room temperature. Is there a particular reason you want to use a thermocouple? Something like an LM-35 might be more straightforward, cheaper, and come pre-calibrated.

 

[Howard]

I have some heavy-wall aluminum square structural tube that might work.

The LM-35 does look attractive. Four problems, though:

1) I already have the thermocouple
2) I wish to use the sensor in environments up to 300 F
3) The sensor needs to be outputting to a PID controller
4) The sensor must have a fast response time

 

[Howard]

This is the controller I have.

http://sestos-hk.com/english/download/D1S-EN.pdf

 

[Howard]

I think I'll just heat up some mineral oil and test the output against my Thermapen.

 

[CycloWizard]

Yeah, the aluminum tube should do the trick. Just bolt your thermocouple to it and make sure you wait long enough for thermal equilibration.

 

[AD5MB]

Thermocouples are inherently calibrated. A Type J thermocouple generates X millivolts at Y temperature. there are published tables of these relationships

when you make a thermocouple, you twist together two wires that have a strong Seebeck effect - they generate voltage when twisted together and are subjected to temperature changes. this junction is the thermal junction.

what throws that off are the reference junctions. when you connect the thermocouple wires to copper wire you just created two more thermocouples, which react to ambient temperatures. Copper exhibits Seebeck effect.

a hundred years ago somebody determined that you could put the reference junctions in ice water filled cups - one per junction. this stabilized the reference junctions. then they determined tables of voltage versus temperature for various combinations of thermocouple wires.

the current way to deal with this is the Thermal Reference Junction Compensator. This is a device that acts as an interface between the thermocouple wire and the copper signal wire. It has a thermistor that measures ambient temperature and internal electronics which compensate for the effects of that ambient temperature on the reference junctions. it makes the reference junctions appear to be at ice water temperature.

when a Thermal Reference Junction Compensator AKA Ice Point is inserted in the circuit the millivolts / degree tables published by Nanmac and Omega are stone cold reliable

there are different Thermal Reference Junction Compensators for different types of thermocouple. a Type R compensator is useless with a Type K thermocouple

 

[Howard]

I don't doubt you, but my PID controller is reading way off with my type J.