Logging and Converting Thermocouple Data

[The Wolf of Walmart]

I've designed an oven and am trying to calibrate the PID controller. I have a J type thermocouple and am using the Analog Discovery 2 to log the millivolt reading at the terminals of the PID controller. I see in Waveforms that I am able to use functions on the thermocouple signal. I would like to make a function that converts the signal to a temperature so I can plot and compare the temperature to set point. 

 

The company that manufactures the thermocouple has a chart for emf/μV but I am not unsure how to turn that into a function. I contacted an applications engineer from the company and he was unable to help me. Searching online isn't turning anything up either. Is what I am trying to do possible? 

[JColvin]

Hi @The Wolf of Walmart,

It's been around 10 years since I got my chemical engineering degree in college, but I'll see what I can remember.

I'm not certain what you mean by chart for emf/uV (as in I don't know why it would be of any help to plot those two things against each other since the emf is normally measured in microvolts). Otherwise, J-Type thermocouples have a standard chart for the expected millivolts to degrees Celsius which you can view via the NIST website here: https://srdata.nist.gov/its90/download/type_j.tab.

As for how this table was created, thermocouples have their own summation formula that uses a specific set of coefficients that was derived many decades ago (thanks to how thermocouples are standardized with what elements are used to make them); this formula and the coefficients are also on the NIST website: https://srdata.nist.gov/its90/type_j/jcoefficients.html. You could recreate this function, though you would also have to invert the series formula to in terms of voltage to solve for temperature.

But realistically speaking, no matter how accurate you type out the equations, you are going to be bit by the absolute resolution and accuracy of the Analog Discovery 2, https://digilent.com/reference/test-and-measurement/analog-discovery-2/specifications#analog_inputs, as that accuracy differential is going to ruin any sort measurement.

If you are wanting to determine accurate temperature measurements from the thermocouple based off of the emf, I would recommend using some sort of dedicated external chip, such as this random one I found online, which has a voltage range specifically designed for measuring the sub-millivolt differences in J-Type thermocouples, and then linearly scales them up to a larger voltage that you can more easily measure; the datasheet for whatever module you use will then have a easy one variable linear formula to convert between measured voltage and corresponding temperature.

Thanks,
JColvin