S7-300 OB86 and SF

Hello Nabuchos;

You seem to misunderstand the action of OB86 (and all other diagnostic OBs) in the S7 PLC.

It is called by the operating system of the CPU when it detects a fault in one of its main systems, in this case OB86 is called because a problem occured in a I/O rack. Other diagnostic OBs are called when specific faults are detected.

The diagnostic OB collects information on the fault and logs it in the diagnostic buffer, where it can be recovered by the user; il also allows the CPU to continue running by bypassing the fault (if it is not fatal). So the SF/BF LEDs are active while the CPU continues running; without the diagnostic OB, the CPU would go to STOP.

In the image you have shown there are 3 buttons you can use to get further details.

- Help on Event will give you generic help information on the type of glitch that was detected

- Save As.. will save the messages in the diagnostic buffer to a text file (you can open it with Notepad, for example) so it can be examined later; hint: send it to us as an attachment if you want us to look at the contents and help find your problem.

- Setting allows you to change the number of messages logged to the diagnostic buffer; you have 10 messages logged simultaneously at this time, I suggest you extend that number to 100 so it covers all events of interest to us. We see in your screenshot for example the message for one station recovering from a fault; we would need to see all messages related to the same station failing, a few miiliseconds before to see information on the fault itself. The text file with sufficient messages logged would be helpful.

So: you have faults on the Profibus-DP network (on 3 slaves it seems); the diagnostic buffer collects this data in messages we need to examine (and allows the CPU to keep running in spite of the fault), and the best way to see them is in the text file you can save it to, and post here for us to see.

Another way to have a graphic view of the faults in your system would be to open the hWConfig editor online (you show the offline view in your screenshot). Modules with faults will show with a red cross or bar over them, you can then open all these modules in the onlibne view and see the contents of each module's diagnostic buffer (all S7 intelligent modules have a buffer, not only the CPU).

Hope this helps,

Daniel Chartier

Hello Noabuchos;

After a quick look at your attachment, I think we can see a few things in your program:

- first, the diagnostic buffer does not see your OB86 in the CPU memory. Could you check if it is actually there? If not, open the Blocks folder of the CPU, create OB86 (if it is not present), and save/download it to the CPU (no program necessary inside at this time). Then run the program and reopen the diagnostic buffer.

- next, with only 10 messages we do not have enough information to follow the loss and recovery of your slaves. Specially since all 3 slaves are affected. Please try and change the settings of the buffer so more (up to 100 messages if possible) can be captured and examined.

Hope this helps,

Daniel Chartier

Hello Nabuchos;

Can you also check if the programmed actions are taking place, on station lost and returned?

If they are programmed in OB86 then you should have some kind of validation.

Hope this helps,

Daniel Chartier