Reading T thermocouple with a K meter

[Grant]

Is there a linear or set margin of error that can be predicted/calculated if using a meter that expects a K thermocouple with a T type thermocouple?

It seems "close".

My meter has an input for a K, but the guys at the roaster have a new scace thermofilter they lent me....but they come with a T.

Thoughts?

Grant

[HB]

I don't know a formula that would convert from Type T to K, but you could always test in boiling water. Or if nobody else has a better idea, I'll pull a couple shots with the Type set to K using type T thermocouple wire to give you a decent estimate for common brew temperatures like 200, 201, and 202F (the Fluke can flip back and forth between different thermocouple types so obviously it knows the relationship).

[another_jim]

Thermocouple equations of voltage to temperature are curved, and going between K and T will be too. You'll need to google the voltage to temperature curves, then do the algebra to do the conversion, then program it into a calculator. It's easier to buy the right TC.

[erics]

Here is a chart.



X axis is degrees F and Y axis is millivoltage output with cold junction at 32 F. I would say 5.8 to 6.2 degrees delta but as Jim said, the relationship between temp and voltage for thermocouples (while it may APPEAR linear) is NOT.

Eric S.

[jggall01]

What a great question! Like the others have said, it is a non-linear relationship and a direct, universal conversion is not possible.

A digital thermometer is essentially a voltmeter that reads the very small voltage produced along the thermocouple wires (in the 4 mV range, probably). Based on the non-linear thermocouple calibration curves (i.e. K, T, J, etc), the thermometer converts that millivolt reading into a temperature DIFFERENCE between the hot end (the probe) and the cold end (the termination at the thermometer).

So, to know the actual temperature at the probe, you need to know both the temperature difference (millivolts converted to degrees per calibration curve) and the ambient temperature at the so-called cold joint at the thermometer. It is the variation in ambient temperature that will defeat attempts to make a universal conversion table between K and T.

But, for a cold joint temperature of around 69F, the following values should give you at least an approximation. I developed these very quickly by feeding a T signal from my calibrator into a thermometer that was expecting K - kind of like what you are planning. The CJ temp varied between 68.5F and 69.2F as I was making the readings.

Table:

temp at               temp
hot joint             indication
of T t/c              on K thermometer

190.0                 195.9
195.0                 201.4
200.0                 207.0
205.0                 212.7
210.0                 218.3

As the ambient temperature at the thermometer moves away from 69F, these approximate values will have increasing error.

Jim

[jggall01]

[2xlp]

or you could go to radio shack and get a voltmeter that you can program K support into (or that supports k)