Issues with PLC 5/40E Boot-Up - Need Suggestions

[Barbue]

Hello Everyone,

I am reaching out to this knowledgeable community with a peculiar problem I've been experiencing with a PLC 5/40E. This issue has been persisting despite several attempts at troubleshooting, and I'm hoping to find someone who has experienced a similar issue or can provide some guidance on how to resolve it.

Here's the problem: I'm unable to boot up the PLC consistently, even though we've changed the battery recently. An error message "OS Checksum Error at address 400DFE" is displayed while trying to boot it up. Interestingly, by applying pressure on the PLC towards the left during the boot-up process, it occasionally starts up normally. However, this isn't a reliable or a long-term solution.

We've tried this PLC on another rack and the problem persists, leading me to believe it's a problem with the PLC itself, not the rack.

Given that we have replaced the battery recently, it seems unlikely that the issue is related to battery failure. Additionally, when the PLC does manage to boot up, it seems to operate fine until the next power cycle.

Has anyone experienced similar issues? I'd really appreciate any troubleshooting suggestions or advice on what could be causing this issue. We're also considering potential alternatives if the current PLC cannot be repaired or made to function reliably.

Thanks in advance for your help!


P.S. See the photo attached (the STAT led blink red).

Best Regards,
Simon

 

[tarik1978]

Try cleaning the plc pins using an electronic contact cleaner, try also clearing the plc memory and reloading the program ( first backup the program).

 

[Barbue]

Try cleaning the plc pins using an electronic contact cleaner, try also clearing the plc memory and reloading the program ( first backup the program).

Thank you for your response and your suggestions.

We've already tried cleaning the PLC pins, but unfortunately, the problem persists. As for the memory clearing and reloading the program, we've also attempted this multiple times. The peculiar thing is that sometimes after booting up, it works fine. It's just not consistent and it's causing a real problem for us.

We've even tested with a new battery, given the potential issues around the memory being wiped due to a drained battery, but this hasn't improved the situation.

At this point, I feel like I've tried almost everything. The fact that the PLC will boot up normally when physically manipulated (pressed to the left) is particularly puzzling. Any other insights or suggestions you could offer would be very much appreciated.

Thank you in advance for your continued help with this issue.

 

[GaryS]

There may be an error in the code it only takes 1 bit to be changed
The checksum is generated using every bit in the memory even a variable word change can cause the error.
If the battery died, the original program may be corrupt
That why it’s always a good idea to upload and backup the program before pulling the battery or powering down when you have a low battery warning
Also you may want to check, there may be an EEPROM installed that holds the original program.
I would try upload and check the program for errors and download again I think the checksum is generated when the program is downloaded

 

[Ken Roach]

the PLC will boot up normally when physically manipulated (pressed to the left) is particularly puzzling.

You very probably have a cracked or corroded solder pad, chip pin, or other low-level conductor that has failed on the main board, probably in or around an EEPROM chip.

These can be difficult or impossible to troubleshoot or repair, especially on electronic devices that are no longer manufactured. You could certainly remove and examine any of the socketed devices, or examine the board very closely under magnification.

I believe that Rockwell still offers remanufacturing for these controllers, but at a hefty price. It may be time to replace it, or upgrade to a modern control system.

 

[Barbue]

Thank you both for your thoughtful responses.

To the first response, we have indeed considered the possibility of an error in the code. We've uploaded, checked, and downloaded the program again multiple times, but the problem still persists. It seems the issue is less about the program data and more about the PLC's physical condition or setup, given that it only boots up normally when physically manipulated.

Regarding the second response, we also suspect that there might be a physical issue such as a cracked or corroded solder pad, chip pin, or conductor. Close examination of the PLC and its components is definitely on our list, though as you've rightly pointed out, the challenge lies in pinpointing the exact location of the problem.

Upgrading to a more modern control system is also something we're considering, although the cost implication is something we need to look into.

I was wondering if you could provide further assistance on locating the EEPROM chip within the PLC (PLC 5/40E). I have been unable to find this information online. Any guidance would be greatly appreciated.

Your input has been very helpful and we appreciate your time and expertise in helping us diagnose this issue. We'll continue troubleshooting and keep your suggestions in mind. Any further insights would be most welcome.

 

[joseph_e2]

The memory module installation manual shows it just plugging into the front.

The quick start manual says a red steady PROC light means major fault:

• User RAM has checksum error
• Memory Module Error
• Internal diagnostics have failed

Fixes are:

• Clear memory and reload program
• Check backplane switch settings and/or insert correct memory module
• Power down, reseat processor and power up. Then, clear memory and reload program. Replace EEPROM with new program. Then, if necessary, replace the processor.

It sounds like you've done most of that already. I suspect you're going to have to replace some hardware from the secondary market. You might also try running it without the memory module to see what happens.

 

[Barbue]

Update with pictures

Hello Joseph, Ken, Gary. Thank you all for your insightful responses.

After closely examining the PCB under a microscope, I discovered a potentially damaged trace (please refer to the attached pictures). It's unclear whether it's only partially broken and unfortunately, I'm having a hard time identifying the chip to which it's connected.

As for the newly shared advice, it indeed seems that the steady red PROC light is indicating a major fault, as per the quick start manual. The suggested fixes, such as clearing memory and reloading the program, checking backplane switch settings, and possibly replacing the EEPROM with a new program, have mostly been attempted.

In consideration of the possible need to replace some hardware from the secondary market, I also plan on running the PLC without the memory module to observe the outcome (thanks Joseph).

I will keep you all updated with my progress and again, any further advice or observations are welcomed and greatly appreciated. Thanks for sticking with me on this one.

 

[tdoa]

Thats either a tantalum capacitor, or a diode, easy to test if its a diode or not with a multimeter.

You can just scrape the solder mask a bit before the via and solder a small jumper wire.

 

[GaryS]

I looked at the pictures and what from what I saw it may or may not be a broken trace on the board but usually with a broken trace the cover film is broken but I didn’t see that. Also I did not see any signs of a cold solder joint. I also saw no indication of the board damage.
Try soldering over the trace and reflow the solder on the solder connection.
you should look at this as opportunity
We know the PLC5 line has been obsolete and even rebuilt replacements are hard to fine. And I have found that third party repairs are questionable. I will only supply to a customer a factory rebuilt unit I have seen more than a few time where third party repairs / rebuilds failed
But consider an upgrade you have a working plc for a while. If you upgrade to a Logix processor and use the original rack as remote I/O You can avoid the downtime and write the new program in the Logix plc or do a conversion , then test it while you still have a working processor on your time table, not under the gun with a crashed system and the boss looking over your shoulder demanding a quick fix and lost revenue from the crashed system.
The hardest part is convincing the boss to move forward before the crash.
I have a client right now that has 4 machine that have obsolete plc’s but still running If any one of those do dawn their plane will have to shut down because of the lost production.
The lead time for the upgraded plc is over 6 weeks

 

[AutomaticInvest_4]

It might be possible that some contact are not working properly..

A possible reason would be some liquid (like water) spilling over some part of the Plc..

Even if the Plc was not plugged in while something like that, if it has some battery inside then the presence of battery creates a working circuit (so there is current inside) that is affected by liquids..

 

[AutomaticInvest_4]

Try to measure/test parts of the Plc to see which one doesn't work

With regards

 

[tdoa]

I looked at the pictures and what from what I saw it may or may not be a broken trace on the board but usually with a broken trace the cover film is broken but I didn’t see that. Also I did not see any signs of a cold solder joint. I also saw no indication of the board damage.
Try soldering over the trace and reflow the solder on the solder connection.
you should look at this as opportunity
We know the PLC5 line has been obsolete and even rebuilt replacements are hard to fine. And I have found that third party repairs are questionable. I will only supply to a customer a factory rebuilt unit I have seen more than a few time where third party repairs / rebuilds failed
But consider an upgrade you have a working plc for a while. If you upgrade to a Logix processor and use the original rack as remote I/O You can avoid the downtime and write the new program in the Logix plc or do a conversion , then test it while you still have a working processor on your time table, not under the gun with a crashed system and the boss looking over your shoulder demanding a quick fix and lost revenue from the crashed system.
The hardest part is convincing the boss to move forward before the crash.
I have a client right now that has 4 machine that have obsolete plc’s but still running If any one of those do dawn their plane will have to shut down because of the lost production.
The lead time for the upgraded plc is over 6 weeks

I wouldn't call lack of evidence of damage, the zone marked in red by OP seems to have been bent or stressed, PCB material doesn't turn white unless you damage it, at least I never saw such thing.
I would consider that board suspicious.

 

[JesperMP]

My vote goes to: Migrate to a new platform. Re-use nothing of the old system.
If you cannot wait until the new system gets implemented, get a replacement PLC5-40E CPU from a renowned reseller and with a guarantee. But don't delay starting to migrate to the new platform even if you do purchase a replacement CPU.

 

[joseph_e2]

I wouldn't call lack of evidence of damage, the zone marked in red by OP seems to have been bent or stressed, PCB material doesn't turn white unless you damage it, at least I never saw such thing.
I would consider that board suspicious.

My vote goes to: Migrate to a new platform. Re-use nothing of the old system.
If you cannot wait until the new system gets implemented, get a replacement PLC5-40E CPU from a renowned reseller and with a guarantee. But don't delay starting to migrate to the new platform even if you do purchase a replacement CPU.

I agree with both.
Step 1: emergency repair with a replacement 5-40E from a good reseller that will exchange it if it doesn't work. Immediately install & program it to make sure it at least powers up and controls everything properly through multiple power cycles.

Step 2: you're already on borrowed time with one foot in the grave. Come up with a replacement. I've seen a partial migration that used a ControlLogix chassis talking to the PLC5 chassis via RIO and it was an awful experience. Do yourself and your successors a favor and gut & re-do. If you need to do it incrementally for production reasons, go ahead, but understand that all of the boards in the PLC-5 chassis are obsolete, old, and could catastrophically fail at any moment.
If you're an AB house, consider ControlLogix and both 5370 and 5380 CompactLogix with in-rack and networked IO. It shouldn't be a tough migration as long as you can get the components. And as long as you don't have anything super crazy going on like motion control or BASIC modules...