SLC 5/05 Online Edit Error

I have three SLC 5/05 processors. One of them goes into fault EVERY time I try to do an online edit. The other two are fine. I doubt there's a hardware problem, because I have a spare 5/05 and have tried it with the same .rss file and it faults, too. Just be be 100% sure it isn't hardware-related, I swapped a processor with one of the two that don't fault and downloaded the .rss from the faulty processor - the new processor faults.

I tried a modification to RSLinx that was recommeded in this forum, but it didn't help.

Does anyone have any suggestions of what i should be looking for? The processor never faults, except when performing online edit.

I can upload a copy of the .rss if anyone wnats to take a look at it...

http://www.plctalk.net/qanda/showthread.php?t=48746&highlight=online+edit+resource

I started that thread awhile back and posted a link to it on the ab forum.

I still don't have a resolution.

I am not sure if the details of our situation are the same as yours. If the online edit status bits go to the unknown state, and the SLC loses communication 8 hours later, then we have the same issue. I have not found anyone yet who has reported those two specific things alike with mine.

We also can repeat this situation here with different processors, as long as they are running the same (although ever evolving) .rss file.

My counterpart, Mike, at first said he could make it happen by performing online edits quickly, but later recanted and now seems to think it's the total number of edits in a span of a few hours that causes the issue.

We have stopped doing online edits to prevent downtime, and now have to prepare our changes offline and download them prior to start up.

I need to start a new case number with a/b tech support and get some answers. I believe that it must be a problem with the SLC firmware or Ethernet daughtercard firmware but I don't have any way to prove that.

You may want to check the status of the online edit bits and see how your issue compares with mine. If we can get two or more people with the exact same faults, then perhaps we have a chance of getting a/b to fully investigate.

Paul

I don't want to try it tonight, because I'm 75 km away from the plant, and I'll have to drive in if I can't download after I fault it. I will try it on Tuesday morning and see if I can still connect to the processor over the DF1 port after it faults. Normally, I get an error that says I have to download the program to the processor, which automatically clears the fault. That makes it pretty tough to diagnose the problem if all the evidence is wiped away, doesn't it?

Another thing I plan to do is see if I can still read from the faulted processor with a MSG instruction. If I can, I can probably read the "S" registers to see what the fault is.

What if you unlatch the error bit at the end of the LAD2? This might keep the CPU running with out faulting at the end of the each scan.

Ranjith said:

What if you unlatch the error bit at the end of the LAD2? This might keep the CPU running with out faulting at the end of the each scan.

Click to expand...

You're right - it might, but it's not really a root-cause analysis. I'd really like to determine what is causing this, rather than figuring out how to make it go away.

My opinion is that it has to be a firmware issue that isn't playing nicely with the particular .RSS file that's running. Strangely, almost 75% of the rungs in the file have be edited recently. You'd think the odds would be on your side that you'd edit or delete the problem...

rguimond said:

You're right - it might, but it's not really a root-cause analysis. I'd really like to determine what is causing this, rather than figuring out how to make it go away.

My opinion is that it has to be a firmware issue that isn't playing nicely with the particular .RSS file that's running. Strangely, almost 75% of the rungs in the file have be edited recently. You'd think the odds would be on your side that you'd edit or delete the problem...

Click to expand...

Can you still communicate with the SLC5/05 after a fault? Do all three SLC5/05 Processors have identical Programs and I/O? By all means post the RSS File. Someone might see something.

You could latch a spare bit and move the Error Status Words to some N Words before you unlatch the Error Bit. That might give you some clues about what is going on. Of course this will only work if unlatching the Error Bit actually prevents the problem from occurring. If it does, then you can go Online if Communications are still possible and examine the N Words for clues.

Stu.....

I had to make a programming change yesterday, so I added an unlatch bit for S:5/0 (math overflow trap) at the bottom of every ladder. I noticed this instruction was used in the procesors that didn't fault. After I downloaded the new program, I tried online edit and it didn't fault! I've tried several times this morning and I haven't managed to fault it yet. It's tough to believe that a math overflow occurs during an online edit, but I'm cautiously optimistic that this is the cure.

Now, if I can only find the cause...

I've had this issue for years (three or four at least) with several SLCs in my plant. I agree with Okie - I put it down to an ENet or firmware issue vis a vis XP, XP Service Pack n, Linx (Classic), RSLogix. Or maybe phase of the moon, stock index, who knows! Fortunately, I rarely have to change logic in any of these machines and when I do I also prepare the new stuff offline, then download.

For a couple of those same processors I also have an annoying startup issue. These are both 5/05's with DNet IO, connected via ENet to PanelView+ HMIs and Compumotor 6K servo controllers. If the machines are turned off, at powerup the processor will fault because it failed to load memory from RAM within a timeout window. I forget the exact fault ID, but that's the gist of it.

[sarcasm]This is one of the more enjoyable errors[/sarcasm]because once this happens the PLC won't do anything until you download the program again. I've checked and changed grounding, power supplies, racks, etc. No change.

Then one day, by accident, we powered up with the PLC ENet disconnected. No problem! PLC was fine. Huh. Plugged the cable back in and cycled power. Bazinga! Failed to load memory. I tried this a few times, and found that it reliably failed with the ENet connected and was fine without.

I suspect the ENet startup is (1) occuring before the CPU tries to load memory and (2) taking so long that the memory watchdog times out. Again, seems like a firmware or ENet issue. Anybody else have this one?

Addohm said:

https://docs.google.com/document/d/1eeOP6-swbQq2IMRl95QaO-3zXtBwZoJV-AVw0UmMEKc/edit?usp=sharing

Click to expand...

Addohm, this is NOT the issue the 3-year old thread was discussing.

Their processor was faulting, not throwing RSLinx off comms.

I am having this same issue, in a SLC 5/04 processor.

It is connected over DH+ , but when I try to do online edit, it gets faulted.

Anyone have the solution of it yet.

Faulty slc

MAHAJAN247,

As rguimond stated, try unlatching the math overflow bit at the end of your program,

First, it is not good Forum practice to resurrect threads that are many years old. You can refer to them, or even link to them, but new discussions should get new threads.

Hagos, the problem described by Mahajan is almost certainly not related to the overflow bit because it correlates to online edits, not to overflows.

Mahajan, the problem is probably similar, but not the same, because your controller is an SLC-5/04 and the original controller was an SLC-5/05.

The problem may still be solved with an update to RSLinx Classic software, but for a different reason.

There was a specific driver-related problem with SLC-5/04 controllers and edits performed over DH+ several years ago.

If it were my system, I would upgrade RSLinx Classic to version 2.59, which was the most stable and recent of the 2.x versions and will be backward-compatible with a decade or more of RA software.

Ken Roach, thanks but I cant upgrade to RSlinx 2.59 because

[1] its RSlinx classic gateway, if I upgrade to 2.59 it will require a LIC Activation, my client is still using that Flopy disk activations.

[2] all other SLC are also communicating through same driver[1784 PKTXD DH+], only this one is having problem, so it should not be the reason.

After downloading logic through serial port I tried to store it to EPROM, but it says, its either corrupted or there is no EPROM.

My question is if the EPROM is corrupted also, why the processor could go to fault while doing any online edit.

Just check this out in case...

1747-l542c online edit problem

Regards,
George