Grounded Thermocouples

While I understand that the metal sheath is welded to the thermcouple itself, where does the ground connection appear within the voltage generated by the thermocouple?

To illustrate, if the voltage generated, at a given temperature is represented by a standard battery symbol is the ground connection at one of the two ends or somewhare in the middle?

There are lots of definitions of what a grounded thermocouple is I can't find a link which answers my question.

If the sheath is floats, like in the air, there's no effect of a grounded junction.

When the sheath electrically connects to an object, then there can a potential developed between that object and the whatever the TC connects to, presumably an analog input.

Since it appears at the junction on both T/C wires, it appears as "common mode" voltage, common to either AI connection.

Any of the tutorials on common mode voltage by National Instruments or anyone who makes signal isolators will typically show path of currents generated by the common mode.

The battery symbol is part and parcel of the "earth" between the object the T/C connects to and the AI. The earth can be helped in its generation of potential differences by motors that 'leak' to ground

Let's say we have a grounded tip thermocouple floating in air. Now, if you measure between the sheath and the positive lead will you get a positive voltage? If you measure between the sheath and the negative lead will you get a negative voltage? Or will the sheath be at zero relative to one and at some voltage relative to the other.

Or let's say you do have a common mode voltage. Wiil the positive lead be higher than that and the negative lead be lower than that? I'm trying to determine where the sheath's potential is relative to the two signal leads.

I think I see where the question lies. It took me a bit to wrap my brain around it as well.

You won't be reading anything relative to the two wires to each other since technically you are creating a new junction with each wire with the sheath.

Try thinking of it as a Wye wiring configuration. No matter where you put your leads on the ends of the wires (AB, BC, CA) you will only read the voltage across that segment. The voltage of the third segment will never come into play.



The sheath should be at near zero potential to each wire since the dissimilarity of the Stainless Steel sheath and each wire is so low. You would be hard pressed to measure it without a microvoltmeter.

However with that being said if the reading device (presumably a PLC) does not electrically isolate the inputs from ground, then a ground loop will be formed and voltage will flow causing the thermocouple to read incorrectly.

There is a voltage difference (Seebeck effect) at every junction of dissimilar metals. There will be a voltage difference at each thermocouple lead relative to the sheath. Looking at it this way, grounded isn't really grounded, but as far as noise is concerned, it is.
Even though the ground reference isn't exactly zero volts, it is a fixed reference that noise would need to overcome. Noise immunity is simply Ohms to ground. You can ignore the thermocouple (Seebeck) voltage. Raise the noise immunity by increasing the size of thermocouple wires to reduce lead resistance.
Best info is from Omega.

Perhaps an actual example can help you.

I've got a vacuum furnace where a thermocouple bead is welded directly to a heating element which forms a grounded junction T/C where the ground is at an above Earth potential. The heating element is 12 volt, 3000 amps, AC. The thermocouple is roughly in the middle of the heating element. If I measure with a multi-meter from the positive lead to earth ground I read 6V AC. If I read negative lead to earth ground I also read 6V AC. The grounded junction potential is present on both leads - its the common mode voltage. I can connect the two T/C leads to any instrument that can reject the 6VAC common mode voltage and read the temperature of the heating element and keep the element from approaching its melting temperature. While there are many instruments that can handle 6VAC or more common mode voltage, there are some that do it better than others, so you may need to be prepared to try a couple of different instruments. A 1771-NBTC card is pretty good at it. A 1756-IT6I not so much (despite what the specs say). Some of the T/C isolators that are on the market are even better.

The battery symbol shows the difference in potential between the ground at the hot end and the ground that the analog input uses.

OK, Danw answered my actual question with his last illustration. The sheath/ground connection, with its common mode voltage, can be considered as halfway through the gradient developed by the thermocouple itself. It isn't at either end. Thank you.

The sheath ground connection will never be in the middle, as this is impossible to manufature, maybe on one side, but it should be loose, btw be aware of ground loops, the meter end should be both loose from ground.
like you said the common mode. measure the current and you are sure no ground loop is presnt.