Just wanted some thoughts on this: I have a customer who wants their type K thermocouples calibrated. I could see doing a function check to make sure there are no breaks in the wires but a thermocouple is, at its basics, just two wires. It either works or it doesn't and it should be obvious when it doesn't, you won't get a reading. I don't see a need for them to be on a calibration cycle. Any thoughts on this and/or the best way to state this to the customer?
Calibration of thermocouple
- Thread starter Zavtra
- Start date
Thermocouples absolutely drift as they get older.
You would want a triple point or heat well/dry block as the standard.
Thermocouple Drift
Thermocuople drift: the change in Seebeck coefficient of the two thermoelements due to physical and chemical modifications
www.msm.cam.ac.uk
You would want a triple point or heat well/dry block as the standard.
Having said that, we do not calibrate our thermocouples directly. We "calibrate" the oven/chart recorder as a system, but not the individual thermocouples.
When I have to keep adjusting temp offsets on an oven, that tells me either a heater bank or a thermocouple is starting to go. Happens often when you are running high temps for a long period.
While we do not see the value in calibrating thermocouples in our manufacturing plant, I could see the need for calibration in super accurate R&D labs. Remember, calibration is just a comparison to a known standard to report the error. Thermocouples do drift, and your customer wants to know by how much.
When I have to keep adjusting temp offsets on an oven, that tells me either a heater bank or a thermocouple is starting to go. Happens often when you are running high temps for a long period.
While we do not see the value in calibrating thermocouples in our manufacturing plant, I could see the need for calibration in super accurate R&D labs. Remember, calibration is just a comparison to a known standard to report the error. Thermocouples do drift, and your customer wants to know by how much.
Semoi
Involved In Discussions
My recommendation is: Ask your customer why he is interested in getting a calibrated thermocouple. Try to understand the reasoning. E.g. using the triple point sound clever, but is your customer most interested in around this temperature range? If so, does an equilibrium phase between water and ice in a large bucket (maybe even a Dewar cryostat) do the job as well? What measurement device are they planing to use (resolution, accuracy, precision)? Is the voltage drop due to the connection between the thermocoupler and the measurement device a significant error contribution?
In most applications the most accurate method is to fill a Dewar with water and ice and to measure the temperature of the equilibrium state using the final setup.
In most applications the most accurate method is to fill a Dewar with water and ice and to measure the temperature of the equilibrium state using the final setup.
Thermocouples are not " It either works or it doesn't and it should be obvious when it doesn't " items. They do age. They do drift, especially when exposed to high temperatures. They have a rated accuracy and often fail to meet those tolerances. The customer may be looking for the offset on a specific thermocouple to squeeze a little more accuracy out of it.
It all depends on how the thermocouple is being used. Serious users of thermocouples, like the heat treating industry, follow the guidelines of AMS2750G concerning use of thermocouples. They must be calibrated before use, you need to know their accuracy. After they are put into use they are not recalibrated and reused. In fact, they have limits on the time they are used, the temperatures they are used at, etc. and then they are disposed of, because thermocouples will change over time and use.
Similar threads
