PLC-5 question

Hello group, sorry about the length of this but I have a plc5/80 ethernet/IP with about 14 full rack rio's and 2 1/2 racks (an engineering nightmare) anyway, starting yesterday all of the MCC's would shut down all of the conveyor system associated with that processor and rio racks. Then within seconds the production folks are able to restart the system. When this happens I lose the comms on the network. So, I've ran over to the MCC with the processor (putting on my 70e gear) but by the time I get there everything is back to normal. I've noticed that on channel 1B I have very high retries on some of the racks. I'm not seeing any processor errors, faults or anything that would make me think that the processor or power supply is failing.
At this point I am wondering if this could be a network problem? Any help would be great.
Jeff

It sounds very much like a RIO network problem.

You can monitor the "Rack Fault" bits for each chassis to find out if the whole channel is failing or just specific racks.

The first thing to check is the termination resistors, then walk the daisy-chain and examine the connections.

If you can get a scope trace of the signal on the blue and clear conductors it will illustrate if noise or a short-circuit might be the problem.

Ken thanks for the reply. Good call on the resistors and network but could that cause an intermittent failure or hard failure? I have a Fluke 123 that I could use to scope the ethernet circuit, never tried that before but I could give it a try.
Jeff

I have a very similar RIO nightmare.

Are there anysoftware tools for mapping the RIO network?

I have a fully loade network with no prints or labeling for the rio. Have no idea what connects where.

peoplehouse i feel your pain.

Mr Noob, this place has been hacked up sooo bad from outside contractors. No prints, no descriptions, just a bunch of bad code with tons of MCR's and Latches. I'm just an electromech but it seems to me that this was some engineers first project. The maintenance supervisor who was supposed to keep the documents up to date has been demoted to a (nevermind)
I just talked to the guys on 2nd shift and now this is sounding more like an IT problem with one of switches that convert the fiber optic to ethernet. (not sure what it is actually called, I'm not a comms expert) but if anyone is interested I will post the results.
Jeff

Maybe I misunderstood when you said "lose comms on the network"; you mean when the I/O fails you also lose connectivity to the PLC5 over Ethernet ?

That sound less like a RIO adapter fault or even channel fault, and more like the whole PLC5 rebooted.

First thing I would do is write a logic trap using the First Scan bit to detect and timestamp a reboot of the controller.

Hi Ken, you understood perfectly. I would lose the ethernet connection on the processor and all 10 RIO's. When you say it sounds like the PLC rebooted that's exactly what it seemed to be doing.
I'm pretty sure we found the problem. A UPS in the cabinet that houses the ethernet switch would intermittently lose votage. That was causing the the switch to drop out and take down the network.
Now you have to tell me how to write a logic trap!