SLC 5/05 Watchdog Timeout FAULT

We have 2 robot welding cells in the rear of our plant that have just started getting Watchdog timeout faults on the SLC 5/05 CPU's.

"0022: The User Watchdog scan time has been exceeded. In the processor under the Data Files S2 - Status, Scan Times, the following is labeled:
Maximum = the highest scan time recorded, I believe this only resets if you download a new program x1, not sure though it might reset when you go from pro to run.

Average = the average scan time recorded x1

Watchdog = the limit that is set to trigger the error x10"

These cells have been running nonstop for a few years without any problems of this sort.
The cells are connected only through a central M.H. robot that is interfaced through RIO to both cells.
The cells are connected to Robot controllers through RIO and to a WinME Panelview.

I did a search but found nothing similer to this.
Any thoughts?

ScottTheEngineer said:

We have 2 robot welding cells in the rear of our plant that have just started getting Watchdog timeout faults on the SLC 5/05 CPU's.

"0022: The User Watchdog scan time has been exceeded. In the processor under the Data Files S2 - Status, Scan Times, the following is labeled:
Maximum = the highest scan time recorded, I believe this only resets if you download a new program x1, not sure though it might reset when you go from pro to run.

Average = the average scan time recorded x1

Watchdog = the limit that is set to trigger the error x10"

These cells have been running nonstop for a few years without any problems of this sort.
The cells are connected only through a central M.H. robot that is interfaced through RIO to both cells.
The cells are connected to Robot controllers through RIO and to a WinME Panelview.

I did a search but found nothing similer to this.
Any thoughts?

Click to expand...

How often are the timeouts occuring? Do the timeouts occur while the cells are running, or also also when the cells are not running.

The fact that the code has been running a few years out almost certainly rules out a programming issue, unless there is something that was 'lurking in dark' all these years. Does the code have any conditional calls and any loops?

It is most likely a hardware issue, and I would start with replacing the RIO scanner as the M0/M1 file associated with RIO card affect scan time.

If you can't find out what is causing the problem, find out what is NOT causing the problem.

Ian

We found the problem. One of my fellow associates didn't like the way a portion of the program was runnning so he disabled part of it. The part he disabled was required to jump the scan out of a indirect addressing loop.

Therefore whenever a program fault of a certain type occured the processor would get stuck in the loop until it crashed.

ScottTheEngineer said:

We found the problem. One of my fellow associates didn't like the way a portion of the program was runnning so he disabled part of it. The part he disabled was required to jump the scan out of a indirect addressing loop.

Therefore whenever a program fault of a certain type occured the processor would get stuck in the loop until it crashed.

Click to expand...

Okay, that part answers it, so the code was NOT untouched for years. Its like when my guys go to a site and the process crashes after they leave and I tell them...

"of course it was working, all you done was go online, modify code, change wiring and other than that, nothing was done to make the PLC crash."

Ian