1794-IR8 failing in service, getting hot

Control logix CPU connected to Flex IO via Controlnet

The Flex io consists of 5 modules

IB10XOB6/A
IE8/B
IE4XOE2/B
IR8
IR8

The 2 IR8 modules get very hot and all inputs go to full scale. regardless of having a field connection or not.
The Module OK led show solid green (functioning normal)
When connected to the CPU via RSlogix the IO tree shows no issues.

Replacing the module allows the system to continue to run
This is a older installation so has been operation for some time (Years) there are 2 other Flex IO drops configured the same these are not having the same issue.

I have taken the faulted modules and installed them in a test rack and get the same results.

As this is a recurring issue we would like to understand what could cause these issues.

Several things come to mind....

A. Make sure the TB3 base wiring is correct, see the manual 1794-IR8 and double-check against the detail in chapter 2.

B. You have "noise" on your 24V dc supply to the modules.

EDIT : Possibly even AC on your field wiring to the RTDs !!

C. The modules are actually faulty, since they behave the same on your test rig, or they are getting damaged in the field setup, check A and B above.

daba said:

Several things come to mind....

A. Make sure the TB3 base wiring is correct, see the manual 1794-IR8 and double-check against the detail in chapter 2.

B. You have "noise" on your 24V dc supply to the modules.

EDIT : Possibly even AC on your field wiring to the RTDs !!

C. The modules are actually faulty, since they behave the same on your test rig, or they are getting damaged in the field setup, check A and B above.

Click to expand...

DABA thanks for your reply

A. wiring hasn't changed - Old installation. (will run ok for weeks then just fail, often after a maintenance shut down)
B. Will try to get a scope on the DC and see if any noise.
C. we have been through 8 modules in a couple of months.

Out of interest I did a temperature check on the modules in the test rack. Faulted module with no inputs surface temp 42 degrees C. Faulty module with test resistors installed 72 degrees C.

We are planning to send back to Rockwell for a possible Autopsy.