SLC5/05 OPC Sporadic

amueller said:

I will update when I find out more, but I have put the IT folks on this one for now. 20+ connections to this PLC would not cause this, would it?

I've heard it has that many connections - is there a way to verify this in RSLinx or the HTTP interface?

Click to expand...

If you type in the IP address in a web browser of the SLC. The web browser will connect to it. Click on the diagnostics tag and you should see ethernet connections near the bottom. This will show you all the ethernet connections with the processor.

This PLC is Series C, which shows 4 connections. I would assume, if that were the case, then it would have to constantly stop/renew connections. I'm not seeing that, I don't think, but difficult to tell to be honest. I do see invalid packets and maybe that can also mean the PLC didn't have an available connection.

2300 is a lot for sure. There would be probably 2 nodes doing that. I think the Panelviews are significantly less in the order of 100 tags.

That Knowledgebase document's author should be handcuffed to a legal writer during a basic citation style lecture.

What it means is that all SLC-5/05 controllers, regardless of memory size, hardware Series, or firmware revision, have at least four TCP connections reserved for Incoming connections, and four reserved for Outgoing connections.

The number of connections that can be used as incoming or outgoing depends on the memory size, hardware series, and firmware revision.

Since yours is a Series C hardware it probably has v5 or later firmware (less than 15 years old):

1747-L551 controller supports 4+4+24 = 32 total connections.
1747-L552 controller supports 4+4+40 = 48 total connections
1747-L553 controller supports 4+4+56 = 64 total connections.

20+ connections to a single PLC is typically going to burden its capability to service communications. As folks have mentioned, there are some tricks to trading scantime for comms bandwidth, and of course there are ways to optimize the client devices.

The error counters are almost certainly not related to the amount of traffic unless you have an unmanaged (and ancient) Ethernet hub in the system. Those typically show Collisions, because they use collisions to arbitrate for bus access. The most basic switches prevent collisions. The error counters are very probably related to a damaged or defective port, cable, jack, or switch.

That does put my mind at ease on the connection side of things. No way there are 64 connections and this is an L553. I still have the IT team looking at these things and added cable length also since these could be long runs. The switches are all Cisco and I've heard there are no collisions.

One way... we sell L553's buy one and if it does not fix your issue send it back

https://www.plccable.com/search.php?search_query=l553&section=product

I still think you should keep a spare on the shelf if its a important process and if it runs 24/7 sounds like you dont want the downtime.

Maintenance has one - they are going to swap it during the Thanksgiving outage. Thanks, though, will keep that in mind for sure!

Please come back and let us know what happens after the swap and/or if you find the issue and the fix

Best of luck!