Electronic – Understanding extra thermocouple voltage between computer and machinery

thermocouplevoltage measurement

I developed my own circuit board which reads temperature from a thermocouple. This board also plugs into a computer via USB and my software communicates with the firmware on the board. It's fairly simple.

This is being used to record temperature readings from coffee roasters (here is an example). Sometimes, a voltage develops between the roaster and computer. What happens is that this voltage can be transmitted through the TC wire to the circuitry and throw off the temperature reading (since TCs work on the principle that the potential changes between the two wires as a function of temperature). I've gotten around this by using ungrounded thermocouples exclusively, which isolate the actual TC wires from the outside world.

However, I'd really like to understand why this happens. I'm not an EE genius…I prototyped this board on an Arduino and paid a real EE to build and program the actual circuit. I'm simply confused about what is going on exactly, and how I can be sure to eliminate any stray potentials which can interfere with the temperature readings.

Best Answer

It sounds like you're describing a ground loop, also an obnoxious problem in audio applications.

Even though your computer and roaster are presumably grounded, "ground" is not perfect. There are all sorts of high impedance paths between various ground points, and all sorts of effects, from mechanical triboelectric to electrochemical to stay fields from other conductors will cause small, subtle variations in the voltage. In short, it's total chaos, so you have to deal with it.

In a coffee roaster I assume you have motors which are relatively noisy, and there might be an actual (very small) electrical current flowing out a vent, as all the rubbing and heat is bound to create charged particles.

Related Topic