Hi,

Last night we had faced a serious problem at our paper machine one of our DCS Controller AC800M had lost communication and HMI's are Crossed for 4 to 5 seconds. After that HMI's are restored Plant had stopped. While trouble shooting we found there is no OPC Error. When we physically look at the Controller it was Ok, no alarm is shown on it. I had the Controller LOg if someone wants to see it.

If some one had faced any such problem kindly help me.

 

Sorry Ali, but there is no magic solution from what you have told us...

Define "Lost Communication" - Do you mean MMS comms to other controllers, serial link failures, or that your HMI Showed OPC Connection Errors?

What does the controller log say happened? CPU Crash? multiple I/O Failures due to power loss? Something else?

Have you got Profibus? Foundation Fieldbus? Modulebus IO?

Did you have a momentary power loss to some of the plant? Is that what stopped your machine?

What part of the paper machine stopped (drives, wires, stock prep, presses etc ??) and which part of the process was this AC800M controlling?

What errors and SOE did the other AC800M controllers show?

Why did your troubleshooting show you had "no OPC Error" - Red & White crosses on the HMI mean you had an OPC failure somewhere - that could be Ethernet switches, hubs, OPC server failover etc etc?

Have you sent the controller log to ABB? What did they say?

You can send the Controller log to the contact form on the web address below.

Rob
www[.]lymac.co.nz

 

The fact that the paper machine tripped indicates that the processor had faulted and 4 to 5 seconds was perhaps reboot time for the controller. Do you have a redundant controller setup for the AC800M?

It doesn't sound like an issue with the 800xA HMI, if the HMI went offline there is no reason for the plant to trip. The paper machine should continue to operate even if the HMI has crashed.

Concentrate on the controllers and power supplies.

Chris Jennings

 

Basically, it was press part that shut down completely, for which we have a separate controller and IO's.

and red cross appeared on particular objects on HMI's.

The log file shows that a particular set of IO's have gone down for a moment and then it became normal. But during that moment, it made the machine got down.

we have also checked its supply and communication cables, it was quite ok.

Still the problem is to be diagnosed.

we have sent this to ABB and waiting for their response.

Any more data you need can be sent to u.

 

Are you using Modulebus I/O ?

I've seen problems like yours in the past with the I/O-scanning time set to 0 (default). Set it to a fixed rate!

 

Yes. The CPU crashed. There are several warnings from the last controller start that this was likely to happen.

These are the important lines...<pre>
W 2010-11-28 09:48:59.978 Unit= _SWTargets CPU_17 CON 4004 P:RCUCableReverse in BkpStopped
W 2010-11-28 09:49:11.979 Unit= _SWTargets CPU_17 CON 4004 P:RCUCableOpen in BkpStopped
W 2010-11-28 09:49:13.980 Unit= _SWTargets CPU_17 CON 4004 P:RCUCableReverse in BkpStopped
W 2010-11-28 09:56:45.979 Unit= _SWTargets CPU_17 CON 4004 P:RCUCableOK in BkpStopped

Here's the start of the crash message ...

FATAL ERROR in \Br\Sv41_ru5\omega\workplace\Sv_41\source\redundantCPU\ModuleBus\MBTestMC.cpp[290]: 906100742 (0x36020006)</pre>
The first error messages are telling you there is a problem with the Redundancy Control Unit (RCU) not being terminated properly or an issue with the backup controller.

The crash seems to have occurred in the RedundantCPU code related to the Modulebus IO handling.

If your CPU's are redundant, check the RCU cable between them (the on on the bottom of the CPU). At the next opportunity test the redundancy failure and make sure the RCUCableOpen RCUCableReverse messages are eliminated.

If your CPU is not redundant you MUST terminate the RCU socket with the supplied terminator. Please check that it is fitted correctly and again at the next opportunity reset the controller and verify the errors are resolved. Otherwise you may need to replace the RCU cable, Terminator or worst case the CPU.

You should also follow up with ABB to ensure there is not another problem that I cannot see in the crash log.

Rob
www[.]lymac.co.nz