Temperature/thermocouple issues

Hi all,

Just wanted to get some ides here on this brainbuster. We have a machine running a GE 90-30 CPU 350. The machine has been operating without much problems for the past 9 years and hasn'r had any program changes. Up until about a year ago we have had an issue that cropts up about every couple months or so with the machine running 24/7. Their are two gas fired ovens on the system with a process thermocouple in each. the analog card controls the burners. It seems that every so often the thermocouple data freezes in the PLC ( sometimes one oven and sometimes both) which in turn either sends the CV to 100% which causes the hi-temp limits in the oven to trip out or low and doesn't allow it to go up to temp. Either way to correct we need to power down the PLC and power back up. At this point we have tried new thermocouples, a new thermoucouple input card and replaced the PLC backplane. We are at our wits end right now as to what could cause this and what to try next. If anyone has any thoughts please let me know. Thanks

Just a guess.
It may be that once in a while you are getting some type of interference on the thermocouple, maybe EMF or RF that causes the Thermocouple module to lose communications with the CPU. This could cause a problem with the PID loop in the PLC.
You should check the CPU for faults in both the CPU and the I/O Fault table. This should point you in the correct direction.
Also verify that the Thermocouple modules and CPU have the latest firmware in them.
If you do not have Proficy ME or other programming software get PACsAnalyzer to collect the fault table information.

What type of PID is used and how is it configured?
Also logic could be added to either reset the module before it causes a problem or to let someone know that the problem is present.

At this point we have tried new thermocouples, a new thermoucouple input card and replaced the PLC backplane.

Click to expand...

Consider adding shielded thermocouple extension wire from the TCs to the PLC. Thermocouple signals are very low-voltage DC signals, and are susceptible to induced noise from Alternating Current circuits in the area.

1) In addition to Russ's and Lancie's advice, if you are using grounded thermocouples, consider using ungrounded thermocouples because the easiest way to introduce crippling noise into a system is through a ground loop.

2) Long shot, but I've lived through 2 episodes personally with uP devices (and have heard of several related anecdotal stories) where the firmware hung-up the system because of a long term timer that timee out and caused subsequent problems. Obviously, there's nothing you can do about it, beyond checking to see if a manufacturer's firmware update fixed the "T/C 24-7 timeout bug".

Until you get to the source of the problem, perhaps you need to add some sort of failsafe logic in the PLC. If you state when the problem occurs, the input freezes, then does that mean the value is at the same number?
If so, put a timer that samples the TC input and places the previous TC input into an image register . If the TC input and the image value are the same value for say 10 seconds, then you can assume your input card has locked up and use an alarm bit to shugtdown heating to the oven.

I am assuming a cold oven is better to have than one that has no temperature control.