allen bradley 1762-it4 - HELP NEEDED

have a issue that i can not figure out what to do to correct. we have hundreds of machines that have molds with J type thermocouples that then go to a stand alone temperature controller such as a Fuji. Management now wants to replace those and have the plc control the temperature. I have posted earlier and with some advice from folks on this site been able on a work bench to accomplish that. I can heat a aluminum mode to the desired temp and maintain it.
Now here is the problem, the thermocouple is in a pre made housing that screws into the mold. Everything was working perfect until i happened to place my hand on one of the molds and was watching the input from the thermocoupler into the plc into a scale with parameters instruction and i saw that value start to jump around. on the display it went from 85 degrees which i had heated it to and maintained it jump around from +- 20 to 30 degrees. i have changed the filter on the input card to 10 hz. As a test i then took the 110vac that is feeding the power to the plc and connected a jumper from the ground and touched the side of the mold. The temperature reading really jumped around to as high as 300 degrees, i removed the ground wire and it went back to normal. i got some information from Ron on this site about using a ungrounded thermocouple but with the way these are mounted into the mold they touch the sides of the mold and the molds are attached to framing of the machine which will be grounded.
I hate to ask so many questions but i have called ab tech support and my local ab rep's control expert and they do not have a answer.
Any suggestions, etc. would be very greatly appreciated.
THANKS

it does appear that the thermocouple i have is the grounded type since if i touch the mold it causes it to start jumping around on the input value when i monitor with the plc.
Since these are hundreds of machines already in place i do not see changing out all the thermocouples as a option.
is there a way around this to prevent noise from being such as issue?

You have a ground loop, often referred to as common mode noise. The ground of your PLC where the T/C input connects is at a different ground potential than your mold. If you are changing the value by touching the mold, it sounds like your mold isn't grounded at all.

There are several potential ways to solve, from using ungrounded T/C's, grounding the molds at the same ground as the PLC, decoupling the T/C signal (creating an ungrounded T/C essentially).

Also, I assume you've wired the T/C correctly with the Cold Junction Compensator.

EDIT: Follow the guidelines in 2-8 and 2-9 and the grounding publication listed in the manual for the 1762-it4, especially regarding the shielding of one end only.

i think i have figured out what causes this. I am only using channel 1 of the thermocouple input card on the bench setup.
I had all 4 channels enabled in the configuration of the thermocouple card.
I disabled the channels that are not actually being used and it seems to have solved the problem.
This has really been a learning experience for me. Great advice and reference on this site has a been a life saver since the only other guy left who worked with me on this type of stuff was let go due to down sizing a couple of months ago.
Sometimes having someone else to bounce ideas off is a big help.
THANKS

Yes, always disable open channels when working with T/C and RTD modules. Glad you solved it relatively easily.

had me going for a little bit.
about 10 years ago my company hired a outside engineering firm to convert 300 of our machines over to use a automation direct plc and control temperature of molds. They never could get it working correctly. The company i work for spent over a half million dollars on this and a couple of years later we went back and converted the temperature control part of the machines over to use the fuji controllers.
now, same guy in charge again and here we go again, temperature control with the plc or else.
Thanks

About 10 years ago my company hired a outside engineering firm to convert 300 of our machines over to use a automation direct plc and control temperature of molds. They never could get it working correctly.

Click to expand...

Now the engineering company is working on Obamacare.