dknguyen

Does anyone know of any ill effects from using an analog multiplexer to connect multiple thermocouples to a cold junction compensation IC? Please assume that the temperature of the multiplexer and the CJC can be kept the same. I'm mainly wondering about issues such as the thermocouple voltage might go negative (beyond the supply of the mux) and whether or not a mux can cleanly pass such a low voltage signal.

http://www.analog.com/UploadedFiles/.../AD594_595.pdf
http://datasheets.maxim-ic.com/en/ds/MAX396-MAX397.pdf

Leftyretro

I worked several decades in a oil refinery and worked on several different TC multiplexer systems. So yes it has and can be done. Most of the ones I worked on used an 'up-scale' burn-out detection method, where if the TC sensor went open, the most common failure mode, it would measure full scale high. This was usually done by wiring a several megohm resistor from the + TC lead to a few volts positive source. This would drive the multiplexer's A/D to full positive scale.

The main challenge is noise and ground loop probelms. TCs are of course low voltage devices (10s of millivolts) so good common mode rejection specs are needed in the mux's sample and hold input circuit. However TCs are low impedenace sensors so even long distances can be measured if good design and construction methods are used.

Good luck

Lefty

__________________
Measurement changes behavior