We are using AB controllogix in one of our project. during commissioning we found that two of the temperature elements (thermocouple K type) drops down to 100 degree all of a sudden. we checked the voltage at the PLC end of the thermocouple cable and found that the voltage corresponds to 900-950 degree. To verify if its the issue with the channel or not, we connected it on another card, but there also we found the value as 100 degree.

we checked the range that we have given and it is 0-1200.

Its was working perfectly then all of a sudden the value dropped to zero; then after some hours it started showing the right value again then after some hours it dropped down. (this is not happening frequently that is what its confusing us).

could anyone please get the reason why its happening like this....

 

Somewhere between the hot junction and the analog input there's a short circuit in the extension lead wire or on a terminal block where the thermocouple connects that connects the thermocouple (+) lead to the thermocouple (-) lead.

It could be extension wire whose insulation was damaged/stripped in a pull through conduit. Heat expansion/contraction moves the wires closer to each other or to a common conductor, like conduit, creating a false junction.

 

If you are dealing with 950C as opposed to 950F, it could be a problem with the thermowell or the T/C element.

Heavy gauge T/C at those temperatures to avoid premature and intermittent failures. Special sheath materials and insulation are also required.

 

A drop to 100 is indicative of a short creating a false junction.

A drop to zero is indicative of a burn-out, open circuit condition with the burn-out direction set to 'low'.

Not sure if you have both situations or not.

 

but if its the problem with cable or element, would i get milli volt corresponding to 950 or 900 on the other end of the cable? we checked the millivolt at the other end of the cable, removing it off from the plc terminal.

> A drop to 100 is indicative of a short creating a false junction.

> A drop to zero is indicative of a burn-out, open circuit condition
> with the burn-out direction set to 'low'.

> Not sure if you have both situations or not.

 

If you get a high mV reading across the thermocouple wires, then

- moving the wires to make the reading changed the possible shorting problem

- the analog input has a problem such that it cannot deal with the mV signal

- there is a ground loop of varying intensity, but with a polarity that drives the thermocouple signal in a negative direction. Sometimes to zero, sometimes to 100.

A floating battery powered indicator would not 'see' the ground loop common mode and report only the differential voltage. A powered AI will see the ground loop common mode voltage with a corresponding offset.

The junction of heavy gauge T/C wire usually contacts the protection tube, which can be at the ground potential of the structure it is fastened to, which can be at a different ground potential than the analog input.

 

I agree with the suggestion of a short between the + and - wires somewhere. It could even be a defective T/C with a short inside of it I suppose.

The idea of common voltage throwing off the mV reading at the input card is plausible. I recall reading that T/C wire is more susceptible to EMI from large electrical equipment such as motors, generators and transformers. You might consider adding a 4-20ma transmitter closer to the T/C area if possible for a more reliable setup.

The random dropping sounds like it could be due to shorting of wires which can change with temperature, or condensation, moisture getting into one of the terminal boxes. But if EMI from electrical equipment turning on/off is the cause of the fault, then that is plausible as well.

 

If your getting any kind of moisture or water that can cause wrong readings also.

 

if grounded t/c's are being used, ground loops between the thermowell locations and its ground path and that of the PLC typically causes erratic behavior.

many sites use ungrounded thermocouples for this reason.

the potential differences between the field and the instrument/control grounding can be significant.

good luck