Communication loss fault handling

Hey guys. New to the forums. We have just finished a conversion from PLC5 to ControlLogix (using a 5571 controller) and are using ControlNet to communicate to one of our PLC5 racks and RIO to communicate with a second PLC5 rack. Now, I am trying to deal with fault handling if we loose communication to either of our PLC5 racks. I want to shut down the program and set outputs to 0 until communication is reset.

Using a Controller Fault Handler, I can reset the fault, but it stays in run mode. I was thinking about putting the controller into Rem Program Mode if a fault occurs and that way, once we get communication fixed, we just have to switch key to Run position and back to Remote and it will stay in Remote Run mode. I know I can get status info from the GSV instruction, but can't seem to set status info with SSV. Can anyone help with the logic to force it into program mode? Will that turn off the outputs on the CN and RIO racks?

Thanks and sorry, I'm new to this stuff.

I would choose not to force the CPU into program mode... That is kind of a last ditch effort, but if you must, you can do something like call a non-existent subroutine.

You can also just check off the box "Major Fault On Controller if Connection Fails While in Run Mode" on the adapter configuration in the IO tree.

Output modules (at least under ControlNet) have options for what to do if comms are lost, set those? I **think** the default action with RIO is that outputs go to a disabled state, but cant guarantee that.

For what it's worth, I have lines with dozens of ControlNet drops, and several with some CNet and RIO drops, and have never had a random failure as long as the cabling was installed properly.

Thanks, I will try that. Reason for program mode was so that, if a fault occurs and it goes to program mode, the logic stops and all we need to do is turn the key to get it back up and running instead of having to go get a computer and going online to clear the fault. Once the controller is faulted, both the fault handler and power up handler are useless, no?

Unfortunately, this is a CLX to PLC5 setup, so I don't have configuration for modules under ControlNet. All I see in IO tree under ControlNet is the 1756CN and the 1771ANCR, unless I have it setup wrong, but it works. For the RIO, I do have to put in each module, but I don't have output control on faults.

I do have the "Major Fault on Controller..." box checked for each module, which is what invokes the fault to begin with.

My main concern is that when I have tested loosing a rack and clearing the fault immediately, with fault handler, it stays in run mode and all my inputs/outputs stay at last state, which could mean motors running unsafely. Same thing happens if I choose to not trip the major fault as well.

Hopefully that subroutine call works and also kills my outputs.

Kevin, what logic do you have in the fault handler?
I bet that is why your processor continues to run.

rdrast, not sure how you call a non-existent routine. No matter how I try, RSL5000 throws an error saying 'unknown routine'. How do you call a routine that doesn't exist and have program verify that rung? Is there another way to programmatically put controller into program mode?

Mike, fault routine is just GSV for majorfaultrecord, two MOVs for code and type, then an SSV for majorfaultrecord. If you know of a way to stop the program and kill the outputs without letting it go to fault mode, I'm all ears. Thanks.

Kevin, you don't want to intentionally fault the processor.
A processor fault can be compared to a Blue Screen Of Death in Microsoft Windows
You don't want windows programs calling BSOD if there is a communications loss.

That is not how you handle system faults.

Create a MASTER_FAULT bit or something similar that turns on when there is a Comm Loss. Use the master fault bit to turn off all the outputs and anything else you want to disable.

Yes, I agree intentionally faulting the controller would be bad. I am trying to put it into program mode, not fault status. Is there a problem putting the controller into program mode?

I have verified that putting the CLX controller into program mode, it kills the outputs on both the CN rack and RIO rack.

I really like the idea of the major_fault, but I would have to include it in every output rung in my program, no? If I just disable all my JSR's using the major_fault bit, the outputs stay on.

You don't want to disable JSRs or do anything to stop the program.
You don't want to put the processor into Program Mode.
The program needs to keep running no matter what.

Yes, typically you would disable every output in the program with a "Fault" bit. Another way to do this is to use a physical fault relay to drop out the "Common" to the Output Card which will turn off all the outputs from the card.

This is typically how a EStop relay is used. Since you have a remote panel/RIO this may not be practical.

It's funny because that's what we ended up doing more or less, except I used the fault bit to control an output, which trips the MCR, which disconnects the live wire feeding the output cards. I liked the program mode idea because you has a visual cue, being the output card status LEDs, so I've put a pilot light on the reset button that indicates when the outputs have power.

Thanks for your help. And sorry for all the questions.

You're welcome. No need to be sorry. The purpose of this forum is to answer everyone's questions. Best of luck.