1756-L61 Memory Dump

nkinkelaar · Oct 27, 2009

Hello All,
I'm having a problem where a 1756-L61 is encountering a major fault (Power Up Fault). The system I'm working with has 3 1756 devicenet Scanner Cards, 1 1756 ENBT card, 3 1756-IB16 cards, 2 1756-OB16E cards, 2 1756-M16SE Sercos cards and a SST-PFB-CLX card.

Does anyone have any experience with this "power up fault". When it encounters this fault the L61 dumps all memory. Which from reading other posts I've found that the memory dump is probably intentional, for safety reasons. According to the customer the fault accurse while the machine is in operation. Yesterday I was able to get the system running again, it ran for about 8 hours then faulted and dumped memory again. Then it ran for 3 hours then faulted and dumped memory again. I haven't talked to the costumer yet today so I'm not sure how the machine did last night.

In the past several years I have commissioned 13 of these systems and this is the first time I've seen this problem. This particular machine has been running for about 3 years with no major faults.

Thanks

Chris S · Oct 27, 2009

Try montering your power to your powersupply, I have run into issues with spikes and low voltage causing the powersupply to fail early on the Contrologix systems. Mostly from over 2 years ago. They might have improved this issue now, but I'd start there. Also you might want to get a CF card to provide a backup on this processor. If you back it up and set it to reload on corrupt memory it will keep it running. But you will still not have solved your original problem. Good luck and let me know if you need more information.

rdrast · Oct 27, 2009

Have you checked your rev levels on the cards? Could you be hit with the Atmel chip problem?

That is one of the symptoms, crash, burn, complete program wipe.

I think the notes were: 40149 and 41204 (41204 has a scanning tool)

nkinkelaar · Oct 27, 2009

The L61 is on rev 16.2. We did get a batch of the bad Atmel chips a couple of years ago but this system was installed before we got the bad chips. Doesn't mean it's not possible. Currently we can't check the chip because the costumer is desperately trying to run off the current job hopefully in a day or so I can get them to check this.

After we get the problem resolved I'm definitely going to look into cf cards for back up. I was always under the impression that the battery in the L61 was suppose to cover this type of thing, I guess not. Thanks for the help. I will keep all updated

tvich · Oct 27, 2009

I'd second the Atmel chip issue. I know Rockwell even revised their original recall broadening the recall. Good luck.

Chris S · Oct 27, 2009

Yes the battery backup stores the memory but in a memory corruption case. I think that is the end result not the cause the CF backup will restore the program and your off and running again. But sounds like the other guys are giving you a good direction to head in.

Kent Hostetler · Oct 27, 2009

I third the Atmel chip issue. The decaying uptime for each successive run is exactly how mine behaved with the bad chips. First dump happened after approx six months in service, then got shorter, real fast!

This thread has a copy of the AB document. The chips were made starting October 2006, 3 years ago this month.

http://www.plctalk.net/qanda/showthread.php?t=36526

rdrast · Oct 28, 2009

Kent Hostetler said:
I third the Atmel chip issue. The decaying uptime for each successive run is exactly how mine behaved with the bad chips. First dump happened after approx six months in service, then got shorter, real fast!
/showthread.php?t=36526

That was exactly my symptom here as well... At first, I thought it was the CNB's 76.something day lockup, till I flashed the firmware...
Then it went from a couple months, to a month, to weekly until I finally tracked down one module that didn't show up originally on the diagnostic scan-tool for some reason.

Complete crash, full memory wipe.

nkinkelaar · Oct 29, 2009

Update......... The L61 faulted 3 times in one day(10-26). The last time it faulted the maintenance tech removed all of the cards from the chassis, reinserted them, cleared the fault, then reloaded the program. Now the machine has been running fine for 58 hours. Rather than paying for me to make a trip ( they are about 4 hours from me) they decided to try the "wait and see method".

I did talk to tech support, they were leaning more towards an incoming power issue than the Atmel chips. They say that if it were the chip the system would've had this problem from day one. They sent me a FDU utility file that they want me run on the controller. This is suppose to give them a log file containing fault details. Since I can't make the trip to run this file I don't think this will be an option.

Does anyone think it would be possible that the L61 wasn't seated correctly in the chassis, then when there tech reseated the card it took care of the problem? I guess time will tell. When it does crash again I'll probably be on the other side of the country and it will be a huge disaster!! Thanks for all of the input.

Nick

rdrast · Oct 29, 2009

nkinkelaar said:
Update......... The L61 faulted 3 times in one day(10-26). The last time it faulted the maintenance tech removed all of the cards from the chassis, reinserted them, cleared the fault, then reloaded the program. Now the machine has been running fine for 58 hours. Rather than paying for me to make a trip ( they are about 4 hours from me) they decided to try the "wait and see method".

Good luck. I hope process interruptions are inexpensive.

nkinkelaar said:
I did talk to tech support, they were leaning more towards an incoming power issue than the Atmel chips. They say that if it were the chip the system would've had this problem from day one. They sent me a FDU utility file that they want me run on the controller. This is suppose to give them a log file containing fault details. Since I can't make the trip to run this file I don't think this will be an option.

That is possible, but the "From day one" thing is complete BS. We didn't start seeing issues for almost a year after deployment, and then it became progressively worse.

nkinkelaar said:
Does anyone think it would be possible that the L61 wasn't seated correctly in the chassis, then when there tech reseated the card it took care of the problem? I guess time will tell. When it does crash again I'll probably be on the other side of the country and it will be a huge disaster!! Thanks for all of the input.

Nick

Very slight possibility, but if it wasn't seated correctly, you would have had many more issues.
More likely then 'not seating properly' would be a bad rack. A bad rack, or one that has something like conductive carbon bridging pins on the backplane can cause processor dumping, so that is something else to consider.

tvich · Oct 29, 2009

Second the 'complete BS' comment on the since day one. I've had history from a co-worker, where it happened a month or two into the project.

Kent Hostetler · Oct 29, 2009

Like I said, mine started after 6 months.

Did the tech inspect the controller for the serial #, date of mfg, and peek inside to check the Atmel chip per the PSA from AB when he had it out? This needs to be answered once and for all...

1756-L61 Memory Dump

nkinkelaar

Member

Chris S

Member

rdrast

Lifetime Supporting Member

nkinkelaar

Member

tvich

Member

Chris S

Member

Kent Hostetler

Member

rdrast

Lifetime Supporting Member

nkinkelaar

Member

rdrast

Lifetime Supporting Member

tvich

Member

Kent Hostetler

Member

Similar Topics