Thermocouple CJC puzzle

You can calibrate away (or fudge away) some of such errors, which is probably what most people do.

you see when the CJC is in the vicinity of the screw terminals but no heroic measures are taken (measures which would likely preclude easy pluggability of the instrument guts, selectable sensor types and so on).

Of course the accuracy is specified after 30 minutes warmup and you can give the customer a trim to fiddle it a bit. Good enough for many.

Best regards, Spehro Pefhany

Thermocouple wire is a big thermal antenna to the outside world. Some alloys conduct heat better than others, which is yet more fun.

rential you see

ty

If you use paired wires (thermocouple zip cord) the temperatures should mat ch and a short bit of nickel-plated brass won't generate much series thermocou ple voltage because the brass in a screw-down connection is pretty certainly isothermal if it's only getting heat input from ambient temperature on the skinny thermocouple wire s. More important, is the nominal ( +/- 2.2C for type K) variance in thermocou ple alloys. When there's a trim, it's most useful when a replacement thermocouple comes from a fresh reel of wire (different alloy batch).

rential you see

ty

Thermocouples are usually made pretty crudely- little more than two similar diameter wires or ribbons welded together.

Considering Wiedemann?Franz and the fact we can usually tolerate a much higher resistance maybe there are better ways to make them.

Best regards, Spehro Pefhany

I discovered that relationship myself. My estimate was 150,000 k/w per ohm. That means that a small wire of any significant length, or a PCB trace, doesn't conduct heat very well.

Another R.O.T. is that one square of 1 oz copper on a PCB is about 500 uohm and 70 K/W per square. Lots of luck actually getting 1 oz when you order it.

But everybody knew that.