G
Glen Walpert
On a sunny day (Thu, 31 May 2012 15:36:26 -0400) it happened Jamie
<[email protected]>:
Hey, I think you could be wrong.
Not that I am a thermocouple genius, but I did study the subject when I
build by little thermocouple sensor thing, and wrote the soft.
Actually John was a great help suggesting the circuit,
and pointing out stupid errors in my first software release.
If you keep the cold-end sensing where it should be, *at the connector
where the thermocouple wires come in*,
with a sufficient low thermal constant, there should be zero errors.
Well maybe not quite zero but certainly very small, depending on the
accuracy of your CJC compensation. A $200 industrial thermocouple input
module will do this quite well, but a $50 Watlow temperature control
board with thermocouple input will not do a very good job at all, having
a poor CJC circuit located far from the thermocouple connection on the
same terminal block as the power and heater terminals, and management
might not see the value in switching. I am sure there are lots of other
cheap controllers on the market, suitable for +/- 5 C or so, and if you
stick to the $50 category, that is probably all you get.
People have been doing accurate thermocouple measurements in harsh
industrial environments like steel mills for about a century now. I find
it curious that it is still botched so often, even by Omega, one of the
largest manufacturers of thermocouples (some of their cheap TC meters and
controllers are quite poor).