Is a major fault stored in the CPU after someone goes from RUN->PROG->RUN to reset

Is a major fault stored in the CPU after someone goes from RUN->PROG->RUN to reset

I recently had a CPU go into fault and a worker turned the key to PROG and then back to RUN, is the major fault still stored in the CPU's memory so that I can connect a laptop with RSLogix to it and retrieve the code, or has it been erased?

Which processor? Cat. #?

It should still be there. Use the structured view of the Processor Status and go to the errors tab, then hit F1 (Help) for an explanation of S:6 and S:54.

jkerekes said:

Which processor? Cat. #?

Click to expand...

SLC 5/05

OkiePC said:

It should still be there. Use the structured view of the Processor Status and go to the errors tab, then hit F1 (Help) for an explanation of S:6 and S:54.

Click to expand...

OK, thanks I'll give it a shot.

S:54 is available only for 5/03, 5/04 and 5/05 Series C processors.

Its a 1747-l553B/C series C FRN 3-9 64K

Did you go online with the processor to see if there is any error messages stored?

See below two links for the processor error messages... any help is appreciated

https://picasaweb.google.com/103780...hkey=Gv1sRgCL25vdfE8qrNHQ#5756168258215610290

https://picasaweb.google.com/103780...key=Gv1sRgCMecuJTJqMrz4wE#5756168140700815794

Here is the link to the zipped rss file, :

https://docs.google.com/open?id=0B9hqGjG3pF96MlpYOUZqVkJUWTQ

It is a program that uses a serial communications link between the allen bradley and another propriatery logic controller. The protocol is called genisys ( almost identical to rs-232) and the program is suppose to receive and transmit messages, and translate the messages into a bunch of physical inputs and outputs.

http://literature.rockwellautomation.com/idc/groups/literature/documents/rm/1747-rm001_-en-p.pdf
Page 16-7 explains and provides probable causes and solutions for the Watchdog Error you have experienced.

Can this error be automaticaly reset with software, e.g. fault routine or some other code? And how would this be done? At the moment I have to send someone out to manually turn the key to program and back, and then everything runs good from a week to several months before the processor faults again and we have to repeat this manual reset. This processor is used mainly as a non-vital indication panel, so i'd like for it to just reset by itself if this can done.

Chapter 16 of
http://literature.rockwellautomation.com/idc/groups/literature/documents/rm/1747-rm001_-en-p.pdf
deals with the controller faults handling.

However, I would first try increasing the Watchdog S:3(high byte) in small increments (maximum allowed is 2500 ms). It is probably a peer-to-peer communications issue since it happens randomly.

Something is probably going wrong in the code for a watchdog timeout to occur. I didn't read the whole program yet, but there are JMP/LBL pairs along with loop counters in all of the subroutines I read so far.

I didn't find anything that was a smoking gun yet, but there may be one in there somewhere.

By smoking gun, in this case, I mean conditions that could cause the rung with the backward JMP to never go false, in turn, causing a potentially infinite loop, which gets shut down by the watchdog timer and generates the fault code.

Backward JMPs are a red flag that there is a potential for a watchdog error, but with loop counting to enforce that they cannot become infinite loops, they should always work. What happens if one of those loop counters gets used in another routine or overwritten by an external device?

The power down file:rung numbers can be very useful too. It may not indicate exactly why the watchdog timer elapsed if there are multiple loops running "late" but if you find those values always park you smack in the middle of one of your loops, then it becomes your smoking gun...the shooter will be found nearby.

I had a chance to head out to the site yesterday and pulled the following info from the SLC500, any thoughts on what might be causing this?

https://picasaweb.google.com/103780...key=Gv1sRgCOGc6dTxo_u0kAE#5766592822980161154

Yes, Some indirect address is pointing to a element which which is larger than what exists in the data table.