ControlLogix Processor Memory Cleared

Hi all, I have heard before that under certain fault conditions that Allen-Bradley processors will clear the program from their memory. I think this has just occured on one of the ControlLogix processors that we have here on site.

My first question is, how can I tell if the program was dumped because of some kind of fault or because the battery was dead? (There was a lightning strike at the plant that did cause the power to momentarily get bumped, but only for an instant). When I went to go online with the processor it remembered its name but had no configuration or program anymore.

The second part of the question is if there is any way to find out exactly what the fault was that caused the processor to dump its program? Is it possible that the memory got corrupted if a surge hit the processor during the lightning strike?

Thanks for your help!

matger said:

Hi all, I have heard before that under certain fault conditions that Allen-Bradley processors will clear the program from their memory. I think this has just occured on one of the ControlLogix processors that we have here on site.

My first question is, how can I tell if the program was dumped because of some kind of fault or because the battery was dead? (There was a lightning strike at the plant that did cause the power to momentarily get bumped, but only for an instant). When I went to go online with the processor it remembered its name but had no configuration or program anymore.

The second part of the question is if there is any way to find out exactly what the fault was that caused the processor to dump its program? Is it possible that the memory got corrupted if a surge hit the processor during the lightning strike?

Thanks for your help!

Click to expand...

A power surge will wipe out all the CompactLogix series CPUs in no time. I can vouch for it.
It only takes a split second and you're done.
Best case scenario, the program still there but corrupted, doesn't do you any good anyways so a download must be performed to ressurect it.
To answer your question, yes, it can become corrupted.
Flashcards/eproms can be a good option. On next power up, have the program to be loaded on the processor from the storage location.

Might be worth a call to A-B Tech Support - I recall several years ago hearing of a utility which sucks important info out of the Controller Serial Port that they can analyse.

If my memory correct it was called "IBeDead", or something similar, although I've searched the net and can't find it, or anything like.

I've had this happen several times to a 1769-L35E. A-B is looking into it very closely. Hopefully will have an answer shortly.

I had a Micrologix 1500 loose every thing last week in a new machine up in CT...the PLC is on a UPS, but something ran in on a t/c and put a pimple on the t/c input card voltage regulator chip, and that appears to have put enough load on the PS that it shut every thing down...AB is doing a check on all the cards and the PLC to see what happened...I'll let you know if they find anything cool.

David

:site:

matger said:

Hi all, I have heard before that under certain fault conditions that Allen-Bradley processors will clear the program from their memory. I think this has just occured on one of the ControlLogix processors that we have here on site.

My first question is, how can I tell if the program was dumped because of some kind of fault or because the battery was dead? (There was a lightning strike at the plant that did cause the power to momentarily get bumped, but only for an instant). When I went to go online with the processor it remembered its name but had no configuration or program anymore.

The second part of the question is if there is any way to find out exactly what the fault was that caused the processor to dump its program? Is it possible that the memory got corrupted if a surge hit the processor during the lightning strike?

Thanks for your help!

Click to expand...

What version firmware? There has been issues with certain versions that have not had power cycled and after XX months of continuous running loosing the memory

It used to happen with version 10 (or 11 - memory lapse here) firmware for the contrologix. If a tag was deleted online, the processor would clear its memory at some point in the future.
I remember that we had to rename tags instead of deleting them. on that particular site there were many tags starting with zzz....

We have had this problem with a 1756-L61.

It was not part of the recall group with bad Atmel chips, and was not a victim of an older firmware issue that could have caused the problem.

We added a compact flash card (1784-CF64) for a non-volatile memory backup and programmed it to restore from the card if it happens again.

Paul

RA have looked into my 1769-L35E memory erasure issue, and determined that it could be due to a missing end-cap or disconnected I/O on powerup.

It was found that during short durations of power down, a memory loss may occur. However, this only occurred when the 1769ECR (End Cap Terminator) was not installed and/or the I/O was configured but not attached and power cycle.

During the powerup sequence, the controller identifies the configured I/O and proceeds to verify the I/O via backplane communications. If the I/O does not match, the controller creates a “Recoverable Fault”, but should not cause a memory loss.

A firmware anomaly has been identified as the cause of the above stated memory loss problem. Operating the controller without the end cap termination and/or if the I/O configuration in the program does not match the physical I/O, may result in the controller going into a “NonRecoverable Fault” and the loss of memory.

It is important to note that this is not a hardware anomaly.

This firmware anomaly has been identified for resolution in the next release of software.

Click to expand...

This doesn't explain everything because its happened on systems with a good I/O connection and end-cap. Doesn't sound good. They can't reproduce it. Someone arsed up and lost the "dump" file in the processor which probably contained the evidence of what happened. This affected firmware 16.3

I'm having (had??) the same issue on a 1756-L62 running V15 firmware. It seems to occur 'out of the blue' in that I haven't been able to tie it to any specific action or state, other than the plc has always been in run mode when it happens. The common consensus is that the processoer is encountering an unrecognized error. It faults and wipes the memory as a protection mechanism.

Now, setting aside for a minute how the processor can detect an error it doesn't know exists, what good does wiping the memory do? The assumption must be that the memory is corrupt or the processor is failing in some unknown way. But presumably two possible things will happen after the fault. First, the user will clear the fault and put the plc back into run mode. If the original cause is still there the plc will just fault again. Or, second, the user will load in the original program and go on with life. This is what the user has to do after the program is wiped anyway.

Keith

I'm having (had??) the same issue on a 1756-L62 running V15 firmware. It seems to occur 'out of the blue' in that I haven't been able to tie it to any specific action or state, other than the plc has always been in run mode when it happens. The common consensus is that the processoer is encountering an unrecognized error. It faults and wipes the memory as a protection mechanism.

Same problem happened here last week with an L62 running version 15.4. It cleared the memory. We had to reload the program.
Processor had been running fine for 2 years now with no problems.
I may have to check into the compact flash cards.

I had to replace (2)1756-L61's due to the bad chips. The strange thing is that when it faulted, it was when the line was down (processor in run, just not making any product). It was part of the processor recall that Rockwell put out and have not had any more problems with the new processor. I have heard that they have expanded the recall to include other cards that have the Atmel chips.

killer said:

I've had this happen several times to a 1769-L35E. A-B is looking into it very closely. Hopefully will have an answer shortly.

Click to expand...

I've had a 1769-L32E and also a 1768-L43 do the same. Both after a spike though.

OkiePC said:

We have had this problem with a 1756-L61.

Click to expand...

Here also. The previous firmware/hardware versions never did it, but all of a sudden on certain "Unrecoverable Faults" it will dump memory. I've been told by the AB people (with little enthusiasm for their excuse) that the ControlLogix may do this under certain fault or corrupt comms data conditions. The only difference between the machines that do dump memory and the ones that don't at this stage seems to be SERCOS II comms and some Kinetix drives. We've sent in the error log files from CF cards in the PLCs but it's done little to get a result that satisfies me.

"They do it on purpose. It's to ensure the machine fails to a safe state." What a load.

killer said:

I've had this happen several times to a 1769-L35E. A-B is looking into it very closely. Hopefully will have an answer shortly.

Click to expand...

I supply hundreds of compact logix processors around the country and have this problem on a continuous basis. AB Tech Support thought they troubleshot it to a power supply, but after trying a new supply, the same problem occurs about 50% of the time. This is a problem that Rockwell must correct because power outages are a way of life. Also we have tried expensive line filters and Battery Backed UPS, but once the power is off for and extended time, the same problem occurs.