ControlNet Comm Loss

I have been having an issue with a system that's using ControlNet. I will randomly lose two inputs from NC stop pushbuttons in the field in two different panels ( The inputs are only lost for less than 1 second them come back). Losing these inputs stops the system. Sometimes diagnostics will show noise warnings but this won't happen when I lose the inputs. I have rescheduled the network a couple of times to no avail. Has anyone had an issue like this before?

Use a GSV to capture the connection status of the Module object for those I/O adapters and see if the connection drops at the time that you "lose" the inputs.

In general, when an I/O connection is broken the input data remains at the last state; discrete inputs that are true will remain true, not get set to zero.

If you are really seeing zeroes, focus on the actual voltage at the field terminal.

Oh, and welcome to the Forum !

The above is assuming you're using ControlLogix or CompactLogix.

What's the controller platform, and what type of I/O adapters are you using ?

I have ControlLogix, and point IO. The CL uses 1756-CNB a ACNR. The communication is between remote racks uses fiber and an RPA. I would have thought that if it was loss of communication that the input would not have changed. The inputs are in two different panels. They go off at the same time, I only catch them with latching bits. I have been troubleshooting this for hours. I'm not seeming to make any progress.

I would have thought that if it was loss of communication that the input would not have changed

Click to expand...

You are absolutely correct. Unless somebody has specifically written logic to detect the connection failure and write zeroes into the Input data tags, then the value of the Input data tag does not change when the I/O connection is broken.

Output data is different; the POINT module can be configured for what you want the output channel to do in the event of a communication loss. Usually it turns off digital outputs. So if the Input point is physically tied to something controlled by an output point, you might see the Input go to False upon a communication loss.. but only after the communications has been re-established.

I recommend that you get a voltmeter with a max/min capture feature out there on that input and see what the signal voltage is actually doing.

Ditto to what Ken R. said. Also, if one is available, a Fluke datalogging multimeter (like the model 189 or 289) is great at finding intermittent power issues. It can be set up to flag the exact date and time and also the duration of the power/voltage disruption. I have had faulty DC power supplies give me PLC input issues in the past.

It appears that while the data will remain the same while the connection is interrupted, when the connection is reestablished the 1756 IO moves a zero into the value for the first scan. AB had this posting in their knowledgebase. The next question is, where to start diagnosing the network issue. Do you have any ideas?

What's the Knowlegebase article ID that says that the re-established data connection has incorrect zero values ?

Have you used GSV instructions to conclusively prove that the logical connection is being broken and re-established ?

https://rockwellautomation.custhelp.com/app/answers/detail/a_id/21956

I have moved around 2 CNBs, rescheduled the network, updated firmware. I talked to Rockwell and they said it is unlikely anything other than cabling/connectors or taps that have gone bad. I'm bringing a tester in the morning as well as a few taps to replace. I have GSV on the CNBs.