joining the ends of thermocouple wires

We purchased our own thermocouple welder. It was a little pricey (I think around $500) but it paid for itself in less than 6 months.

Charles Perry P.E.

The theromcouple welder is very simple. You should be able to design and build one yourself. In principal it works this way:

Connect the instrument side of the thermocouple wire to a large capacitor, perhaps 50 MFd, and then charge it to say 200 Volts. Then disconnect the charger.

At the sensing end, move the two thermocouple wires together until they touch. The resulting spark melts the wires together and, voila, a thermocouple junction!

Of course, do a good engineering job on your mechanism so you don't blind yourself or electrocute yourself.

The value of the capacitor and the voltage level can be determined by experiment to make a nice weld on the size and type of thermocouple wire you use.

Al

On Thu, 05 Jan 2006 21:02:53 GMT, Al Gave us:

Then, one should still create a calibration chart using an already calibrated standard. That way, one knows what one's offsets are at any given temperature.

In addition to what others have said, also keep in mind how the *other* end of the thermocouple wire is to be connected. Just connecting to a terminal board creates, in effect, two other thermocouples. The temperature at that junction needs to be known and adjusted for.

Some direct-reading thermocouple meters will do this for you internally. But if using 'generic' mvolt metering, the 'cold junction' temperature needs to be considered.

daestrom

The V vs T characteristic of a thermocouple is non-linear. There are ICs which do the cold junction compensation as well as conversion to linear mV per degree outputs (or digital out, if you prefer).