How Well Do You Really Know Thermocouples

[proof]

How does two TC in parallel act?

Using similar lengths (or similar resistance), they give an average reading.

 

[danw]

I disagree with your statement that heating both junctions to 200 or three hundred gives no volts. They do give volts, but they cancel each other out, so there is no current flow. You had the right conclusion, but for the wrong reason.

You might disagree, but a thermocouple's EMF is generated because of the temperature gradient - the difference in temperature from one end to the other. It is not the absolute temperature that generates EMF, it is the difference in temperature.

Measuring the voltage across a thermocouple with a DVM is also a misleading test. The DVM lead to on wire on the thermocouple creates a junction, and the other DVM lead to the other side also creates a thermocouple. So now you are not measuring the thermocouple voltages, but a new voltage from the three junctions. This voltage would be created from the difference between temperatures in the three junctions.

If this were true, then thermocouple thermometry would not be practical. A thermocouple analog input circuit uses conventional electronic materials (copper, brass, nickel plating) as conductors to sense the thermocouple's millivolt output, (optionally) amplify it and feed it to an A/D. So any conventional thermocouple transmitter, controller, handheld meter, or AI card would be subject to the limitation you cite.

However, the thermocouple Law of Intermediate Metals applies, which says that intermediate junctions (of the same metal composition) do not affect the T/C EMF output if the intermediate junction measurement points are isothermal, that is, at the same temperature.

The issue with measuring a thermocouple's mV output is trying to relate the mV value to the temperature listed in a thermocouple mv/temp table because the cold junction/icepoint compensation is not taken into account.

 

[danw]

Using similar lengths (or similar resistance), they give an average reading.

Yes, resistance needs to be taken into account:

Some times swamping resistors are used to compensate for different length runs of thermocouple wire.

 

[cncsparky]

Thanks for this informative post!

We have several ovens using J or K type TC's for PV into Honeywell UDC controllers. At some point, a PLC5 TC input was installed in order to get oven temps into a database. Using TC wire, the junction on the UDC was jumped to the PLC input so that both the UDC and PLC gets the same TC input.

Is this a good practice? What about CJC? Should just one device have CJC, or both? It seems to work ok and I've always wondered about it. Just doesn't seem to be best practice to me. Some later installs used a retransmit option on the UDC to spit out 4-20ma to the PLC for data acquisition.

 

[danw]

Is this a good practice?

It's been done for decades in the heat treat biz. If doing so produces the same temperature value on both devices, then it's valid.

There can be problems when paralleling a T/C to two devices:
- common mode ground loops when the thermocouple is grounded
- the thermocouple burn-out circuit on one device creating noise in the signal on the 'other device'
but if those problems are not evident, it works.

Other industries that use temperature transmitters choke when they hear about paralleling to a 2nd device, but that's their problem.

What about CJC? Should just one device have CJC, or both? It seems to work ok and I've always wondered about it. Just doesn't seem to be best practice to me.

Yes, each device, the UDC and the PLC5 definitely need their own individual CJC active and enabled. The CJ measures the terminal connection block temperature, so any CJ function should not have any effect on the other device.

Some later installs used a retransmit option on the UDC to spit out 4-20ma to the PLC for data acquisition.

That works, too.

 

[NetNathan]

Something to remember on jumping TCs across different devices...
Some devices actually send a voltage pulse across the TC input to check for open circuit.
This can cause the other device to see the pulse and jump a few degrees and go back.
I have found this on a lot of process recorders, but not control instruments.