Ethernet Communications Issue

[amil30]

We have a cell that was done by an integrator. It has lots of intermittent comms issues that sometimes are minor and sometimes doesn't allow you to even ping the PLC compactLogix when connected to the LAN of the cell, but the cell can sometimes still function even in these instances.

Today I noticed there were 2 devices connected to a Balluff field network switch (BNI0089) that were not communicating even though they had link lights on. There were 4 other devices connected to this switch and they were all communicating properly. I tried power cycling the switch and it made it so all the devices stopped communicating.

I talked to some techs that had similar issues with this, and the ultimate fix is to power cycle the whole cell, but I can't do that at the moment or do i think it should be like that at all.

Any ideas?

 

[Ken Roach]

It could be bad wiring, it could be EMF interference, it could be mis-configuration of devices, or it could be bugs in firmware.

Baluff BNI0089 is an M12-connectorized unmanaged switch. What other devices in the workcell network have diagnostic features ?

How is the workcell isolated from the enterprise network; a dual-NIC HMI or workstation computer, or a router, or a NAT appliance, or something else ?

Intermittent network comms problems can be very tough to solve. Start by mapping out what's on the network, and figuring out which devices will be able to give you some insight into the normal conditions of the network and the disruptions.

 

[829]

Do the other things connected generally work? If so, have you checked for duplicate IP's in your network?

 

[amil30]

So let me reiterate what I mean by intermittent comms issues. Basically, when the cell is running, there are no problems, but when these issues occur, it's a whole cell issue and they don't fix themselves. It always requires a power cycle of the 24V devices. The next time it happens I was going to try and just do a power cycle of the main network devices and see if that fixes it.

Also, we have 4 identical cells and it happens periodically with all of them, so I kind of ruled out bad wiring.

As far as the network layout, we are using a 5069 CompactLogix w/ dual IP config so our 2nd port is connected to the enterprise network and the 1st port is connected to the LAN of the cell. There are no other outside connections to the cell.

I still have to look into the models of the other network devices, because I did see this Balluff device was unmanaged, so I was confused as to why power cycling an unmanaged switch would suddenly create problems on only that switch.

We have also checked for duplicate IPs with no luck there either

 

[cardosocea]

The fact that hits all cells randomly leads me to believe it'll be something common between them. So the enterprise network could be an issue, though from the rest of your description doesn't sound like it.

Do you have common network devices like the balluff switch that may be presenting the same problem? Like a firmware bug for example?

Where is this switch connected to?

 

[amil30]

The network is setup as a star topology and this switch comms is daisy chained from another device (probably IO device).

I haven't seen any other devices necessarily causing issues directly like interacting with this, but that doesn't mean they wouldn't.As for firmware that could be an issue but I have no idea what firmware was used by the integrator on these devices

I'm curious what you guys mean as the enterprise network causing an issue. I'm definitely not a network guy but if it's on separate networks entirely how would that cause issues with our local devices? We also have this type of setup with dual config in our plcs all over our company and have never really had this problem.