We are having an issue where some tags across multiple PLCs having a Bad_failure tag quality, and do not receive a value until we restart the OPC UA module on the gateway.
Normally, we saw this when live edits were being done on a PLC (seemed like edits to one PLC impacted polling to all) but that would resolve itself shortly after the edit finished. Now it seems to be happening when there aren't edits being done and lasting much longer requiring the module restart. There's no logs to indicate an issue, nor do we lose connection to the PLCs, there is also no communication overload (msg class 3 is fine as well), and all our driven tag groups are still polling like regular so I'm not sure where to look next.
I've read into some issues on the forum about online edits on Allen Bradley PLCs and also re-browsing, but I'm not certain thats what we are experiencing.
Gateway version: v8.1.27
We have multiple 5069-L340ER and L3100ERS2 PLCs on v33 firmware.
v36 has been pretty solid from what I've seen. Just have to be careful of the change to several instructions that changed like
MOV -> MOVE
LES, LEQ, GEQ, GRT -> LT, LE, GE, GT
Possibly more, but those are the most frequent ones I use
I'd say avoid v38 for now as it has a nasty bug that the Finalize All Edits button will sometimes just freeze Studio 5000 completely, and the only workaround last I checked was to do the old school 3 button method of Accept, Test, Assemble.
Almost the most annoying "feature" AB has ever distributed. I just can not understand why this decision was made. Still infuriates me as I tend to type in my rung structures, and the addition/removal of charactres in those instructions battles my muscle memory.
IEC 61131. Everyone else in the industry follows a standard of instructions, AB has finally caught up with the industry. All this shows is that you don't use any other brands. For the average integrator that has to be fluent in 10+ PLC brands, this is brining AB up to the minimum standard industry wide.
Actually, I do use other brands, just not as often. I have programmed Eaton, Siemens, Mitsubishi, AutomationDirect, Omron, and others. I don't struggle with those, because its not what I do everyday. I work in a factory which has 150+ PLC's everyone of which is AB.
Thanks for the information, as I was unfamiliar with this being a move to IEC61131 (as that is not something I need to or care to stay up on). At least it sounds like there is a reasonable explanation.
Pretty sure this is a move by AB so that AI PLC code actually works on AB hardware. AI isn't going to automatically check for exceptions before suggesting ladder elements.
That is the scariest thing I have contimplated in some time. Bad enough that people are blindly trusting AI with software code, but someone taking AI generated PLC code and actually using it in the wild. Yikes!
Claude: How do I make a latching coils in ladder logic without latch bits. Make the output coil hold the safety contactor on in case the batch hasn't finished.
