Distances between Thermocouples-PLC


Thread Starter

Alberto Maldonado

Dear List:

I have 6 Thermocouples type K to a PLC. 13 Meters of Extension wire and 40 Meters of cupper wire to PLC (Maximun distance). The PLC receives the mV
signals with an analogue card for thermocouples. Is there any problem with this signal in this distance ? (there are 6 motors 3HP everyone but 3 meters is the nearest motor of the thermocouples (the extension wires not run through these motors and the cupper wires less)).

Thank You for your attention.

Alberto Maldonado
Mail:[email protected]
I've used type J thermocouples with extensions of over 100' with no problem.

What's with the copper wire? Extension wire must be used all the way from the thermocouple to the cold junction unless you're using a
temperature transmitter or something. Connecting copper wire to thermocouple extension wire creates a junction of dissimilar metals that
will act as a thermocouple, producing voltage that will either add to or subtract from the signal from the real thermocouple.

Also, cold junction compensation must be implemented at the analog input terminals. An input configured for standard mV won't do that and
won't work...it has to be configured for your specific thermocouple type so that the PLC can apply cold junction compensation.

Steve Myres, PE

If you will be running the thermocouple signal the whole distance, I think you have to use thermocouple extension wire and not switch to
generic copper wire.

I would recommend not using the thermocouple module with the PLC, though. I would suggest using 4-20ma transmitters at the thermocouple, and run generic shielded cable from the TC to the PLC, with the shield grounded at the PLC end only. If you keep the signals isolated from the motor power leads, you shouldn't have noise problems. Routing is especially important if the motors are on VFD's, as the leads can radiate high frequency noise.

Steve Myres, PE
[email protected]

Curt Wuollet

Distance might not be a problem but the extra junctions at dissimilar
temperatures might well be.



Jorge Diaz Zamora

One customer installed several thermocouples type J and K and connected to PLC. Thermocouples were manufactured in USA and Cables in Germany, both
the same type. We had 25 =BAC app. of difference between the point of measurement and the PLC terminal. We did all the instrumentation practice without result. Finally we used a cable and thermocouple manufactured in the same country. The measure was ok. I suggest always to use temperature transmitter, you reduce this kind of problem, or grounding, noise, compensation, etc, etc. problems.


Jorge Diaz Zamora
[email protected]
In an oil refinery application I have successfully operated thermocouple control at a distance of 400 to 500 Meters between the thermocouple and the mv/I converter. The connecting wire is T/C extension lead of the same material as the thermocouple. The multipair extension cable is individually shielded with an overall shield, run in a dedicated cable tray, removed by at least 2 Meters from any HV motor leads.

I concur with the other responders that the use of copper instrument cable should be restricted to 4-20 ma level signals.

With the installation as you describe it, shielded T/C extension wire should give you adequate results.

John Beck
Control System Specialist, Retired.

Bob Peterson

My take on this.

1. This short of a distance is not a big deal for T/Cs. But you MUST use T/C extension wire, and T/C connectors where you connect the extension wires to the T/C. Wire nuts are not a good choice. Wire the T/C or T/C extension wire direct to the T/C input card. terminal blocks are nice for ease of wiring but unless you use T/C terminal blocks it will add in another junction, and another place where the connection can fail. People seem to use the smallest gauge possible of T/C wire, maybe to save money, but a smaller wire is more fragile.

2. Ungrounded T/Cs are best. Fewer ground loop problems.

3. For longer distances, or where you are forced to use grounded T/Cs, put an isolatedT/C transmitter in close to the T/C and feed a 4-20 mA signal back. You will probably save yourself a lot of grief. Longer distance being defined by how much it might cost you to change it if what you do does not work.

4. As far as shielding the T/C cables, I would suggest this is really futile. The T/C signal is so small, and shielding is imperfect. The seperate conduit you must run it in is one heck of a shield, compared to a braid and foil shield. Even in a seperate conduit follow good practices and keep them well away from conduits with high level voltages in them, particularly VFDs.

Bob Peterson