Logix 5000 Message Buffer Issues

[mylespetro]

Hey everyone,

I think I'm running into a bit of a bottleneck on the amount of messages executing on an L62 ControlLogix (V15.4), and I was hoping to get a little bit of insight on something that I had noticed about some of them and thought of a way that may be able to eliminate some messages and consolidate them into a single message. I noticed that there are 34 messages executing constantly talking to various PLCs in the plant, and I feel like that is choking up the buffer.

As for what I was looking at doing, our L62 (PLC1) is messaging a new L71 (PLC2) using 11 messages, let's call these PLC2MsgBlock1 thru PLC2MsgBlock11. Blocks 1 and 2 are a CIP Read and CIP Write respectively, but Blocks 3 thru 10 are Reads with the follwing structure (Block 11 is a small CIP Write):

Block 3

Source Element: N10[50]
Number Of Elements: 10
Destination Element: PLC2Data[50]

Block 4

Source Element: N10[150]
Number Of Elements: 10
Destination Element: PLC2Data[150]

...

Block 10

Source Element: N10[750]
Number Of Elements: 10
Destination Element: PLC2Data[750]

So there's 8 messages moving 10 words of data each, but I'm wondering if it would be more efficient to consolidate it all down to one message with the following:

Block X

Source Element: N10[50]
Number Of Elements: 724 (PLC2Data is an array of 775, so this should fill N10[50-775])
Destination Element: PLC2Data[50]

I'm just not sure if a single message of that many words would be more taxing than a bunch of smaller messages or not.

EDIT: Just to clarify, I'm pretty sure that consolidating Blocks 4-10 into Block 3 should reduce the number of messages to 27 from 34, and hopefully this will give the processor a bit of headroom.

 

[jstolaruk]

What communication modules are in the Controllogix racks with each processor? Choice of those modules will dictate the number of nodes/messages that can be used.

And yes, all of this data could be compacted into a single UDT and then use produced-consumed tags to transfer data.

 

[mylespetro]

What communication modules are in the Controllogix racks with each processor? Choice of those modules will dictate the number of nodes/messages that can be used.

And yes, all of this data could be compacted into a single UDT and then use produced-consumed tags to transfer data.

Both Ethernet/IP modules are 1756-EN2T cards. I tried putting all of the words into a single MSG instruction and now that block is erroring out, and I even tried it again with the words split evenly over two MSG instructions and I'm getting the same thing.

 

[jstolaruk]

Both Ethernet/IP modules are 1756-EN2T cards. I tried putting all of the words into a single MSG instruction and now that block is erroring out, and I even tried it again with the words split evenly over two MSG instructions and I'm getting the same thing.

Abandon that messaging technique and use produced-consumed tags. Here is the link to the manual:
https://literature.rockwellautomation.com/idc/groups/literature/documents/pm/1756-pm011_-en-p.pdf

 

[mylespetro]

Abandon that messaging technique and use produced-consumed tags. Here is the link to the manual:
https://literature.rockwellautomation.com/idc/groups/literature/documents/pm/1756-pm011_-en-p.pdf

Yes I was looking into this the other day. I'm assuming in PLC2 I could MOV or COP N10 in PLC2 (array of 775 INTs) into ProducedTag (INT[775]) and then create ConsumedTag (INT[775]) and connect them through the I/O tree of PLC1 and then MOV/COP ConsumedTag into PLC2Data (INT[775])?

 

[jstolaruk]

Yes I was looking into this the other day. I'm assuming in PLC2 I could MOV or COP N10 in PLC2 (array of 775 INTs) into ProducedTag (INT[775]) and then create ConsumedTag (INT[775]) and connect them through the I/O tree of PLC1 and then MOV/COP ConsumedTag into PLC2Data (INT[775])?

Looks correct but I would strongly suggest taking some more time and reading the manual.

 

[mylespetro]

Looks correct but I would strongly suggest taking some more time and reading the manual.

Taking a look now. I've explored this several times before but never had the opportunity to do it. Thanks for the tip.

 

[mylespetro]

Looks correct but I would strongly suggest taking some more time and reading the manual.

I've just realized that the Ethernet/IP module in the chassis with PLC2 is a 1756-EN2TR, which I don't believe is possible to add in the I/O tree of PLC1, as it's currently flashed to V15.02, where as far as I can tell, support for the EN2TR starts at V20.03 so I think I'm back to square one.

 

[jstolaruk]

I've just realized that the Ethernet/IP module in the chassis with PLC2 is a 1756-EN2TR, which I don't believe is possible to add in the I/O tree of PLC1, as it's currently flashed to V15.02, where as far as I can tell, support for the EN2TR starts at V20.03 so I think I'm back to square one.

Flash it to v20, v20.05 software is supported in Windows 10 also.

 

[mylespetro]

Flash it to v20, v20.05 software is supported in Windows 10 also.

I'll look into doing that at some point, they're trying to get started up here now so it's not a great time to do a flash on the processor. For the time being I put a 2-second self resetting timer in the program, and when the .ACC is between 0 and 1000, I have half of the messages executing, and when it's between 1001 and 2000 I have the other half executing. It seems to be working well for now, but it's just a band-aid and not a real solution. I would love to get every remote PLC communicating using produced and consumed tags if possible, so I'll look into getting that going next time the plant is down.

Thanks for your help, really appreciate it.