Slc 505 explicit Messaging interfears with MSG?

Join Date
Jan 2003
Location
Michigan
Posts
490
Hi All, Ive an SLC program that messages to several other slc 505s as well as a positrack system. The SLC explicit messages the positrack every .02 seconds. Ive noticed that my other messages can occasionally behave strangely. The message will be enabled, but it never times out, finishes, errors, nothing. The diagnostic file doesnt seem to indicate any problems either. It just sits there. Flipping the plc from run to program and back makes the messaging work again. It seems to be tied to the explicit messaging. I changed the explicit message so it it only does the explicit messaging when it actually needs the data, and so far it seems to have helped. So, my question is, can explicit messaging override normal MSG commands (ive never had to use explicit messaging before). Thanks

matt
 
Look at the number of connections the 5/05 has. I think you must have a diagnostic data file configured in order to see this.

I suspect that the SLC is using all its connections, since each unique IP address it communicates with (including RSLinx on your workstation) will take up a connection. Panelviews my take more than one.

Look under channel diagnostics/status.
 
It is likely that the MSG and EEM instructions were implemented without any buffer control and one or more of them are filling up the Message Buffers and the Message Queue when there is a delay or error on one of the Messages.

I have encountered dozens of instances where messaging instructions executed by timers, events, or self-repeating logic filled the 4 Buffers and the 10 Queue slots and left subsequent messages in the Enabled state, waiting for Queue space.

You mentioned that there is an Explicit Message triggered every 0.02 seconds... a typical SLC scantime is 10 milliseconds so this is trying to execute every other scan. The SLC only handles messaging during the "housekeeping" time after the program scan, or when SVC instructions are encountered.

The EEM instruction is handled by a slightly different firmware routine from the MSG instruction, but I have not encountered an instance where they "interfere" with one another.

Your first step toward improving your SLC's messaging performance is to summarize all the MSG instructions and examine how they are triggered. Then we can discuss the One/All Messages Processed Per Scan status bit, as well as the buffers and queue.
 
They do sit in the enabled state. So the buffers must be full then. There are many messages in this app, most are triggered only occasionally. In this case I have a MSG to another car (these are cars) that always runs (resets on the done, runs again) for collision avoidance. I have a 3 more messages to 2 other SLCs. A write and a read to one, and a write to another. These only trigger when the car is parked in front the the station (they repeat on done or error)(and there are 4 of these stations). Additionally, there are 3 wonderware stations all reading from the slc as well. Plus the explicit message as well. Total connections maxed out at 8, and since these are FRN 10s, they have more than enough left.

Edit : Also, I dont see the waiting for queue space bit on the message when they are sitting there. So it must be the buffer then, correct?

So, it must be the buffers. Ill see what I can do to open up buffer space. So the EEM doesnt use the same buffers and queue?

Edit : Anywhere I can read up on the SLCs buffer?

matt
 
Last edited:
Its the EEM causing it. Like I said in the original post, changing its operation helped. I cleaned up the rest of the messages now. If the EEM is running at .02, I can have 1 other MSG running (repeating on done or error). If I start more MSGs, they sit on enable. As soon as I put the EEM in slow poll (30 seconds) all the other messages start firing normally. Fortunately, I do not need the other messages and a fast EEM at the same time. Learn something new every day. I would like to read up on the queue and buffer some more if you know of any materials. Thanks.

matt
 
Ken and others have resolved extensive amounts of problems in relation to yours. Your best bet would be to search these forums with keywords like "msg", "buffer", "scan time", and of course "Ken Roach".
 
Well, its not really a problem anymore. It was more or less solved before I posted. Just wanted to clarify the "why".

Oh yea, Ken is awesome! I sure hope AB pays him for this!
 

Similar Topics

Hello, I inherited a control system one of my predecessors thought it was a good idea to put logic for cant optimization and Kinetix motion...
Replies
15
Views
3,509
Probably a silly question and asked a hundred times already, but i tried to do some searching here and couldn't find this question. I have 502's...
Replies
9
Views
2,964
I'm a mega noobie when it comes to PLCs and the such. I'm a network guy. We are currently experiencing an issue where the SLC immediately faults...
Replies
8
Views
1,873
Hi guys I hope your all safe and well , I have a question regarding a slc 505 on ethernet is it possible to add a ethernet remote I/O module...
Replies
5
Views
1,852
Ok I'm feeling really dumb right now. I was wanting to create an MG file in a SLC 5/05 processor. It is a 1747-L551C, OS501 Series C, FRN10...
Replies
3
Views
2,203
Back
Top Bottom