ControlNet Communication Errors

[jrsnydley]

Hello all,

I was setting up a network with a local chassis containing a 1756-L62 controller and a 1756-CNB ControlNet bridge and a remote chassis containing a 1756-CNB and a 1756-ENBT Ethernet bridge. Connected to this ENBT is a 1794-AENTR Flex I/O adapter and a 2711P-B7C4D8 PanelView Plus 6 700. A rough drawing of the network configuration is attached.

When I downloaded the program I was getting a Code 16#0317 Connection not scheduled error. I uploaded the ControlNet configuration from the scanner and I enabled edits and saved the network configuration. Now the error is gone in the controller's I/O configuration but I have no communication to my PanelView.

To say I'm no expert at networking would be an understatement. Looking at the network configuration does anybody see any possible issues?

 

[nhatsen]

a) Probably there's a problem with the shorcut in the PV application (FTV Studio -> RSLinx Enterprise -> Runtime tab).

b) Is it possible to move the ENBT module to the main chassis? I think this way you can avoid a bottleneck in the controlnet.

 

[jrsnydley]

Possible yes, just not as convenient. I double checked the connection in FactoryTalk View Studio. Everything appears to be fine.

 

[jrsnydley]

I think I found my answer on another thread. Ken Roach explained that there can only be one network connection between I/O devices and the CPU on Ethernet I/P and ControlNet architecture. So, I will have to run a little longer cable I suppose and put an ENBT in the local chassis.

 

[Ken Roach]

You've got it exactly right; for I/O connections, there can only be a 1756 backplane between the CPU and the "root network bridge module", in this case the 1756-CNB for the I/O devices on the ControlNet network.

When you add that 1756-ENBT, that's now a "root bridge module" for the EtherNet/IP network, so the ControlNet link can't be in the middle.

CIP unscheduled messages can traverse that link; the PanelView Plus should be able to connect to the CPU from that location.

But the 1794-AENT FLEX I/O adapter, because it's an I/O connection, cannot.

As I probably wrote on that older thread, the connection will *sometimes* work and Studio 5000 might not explicitly prohibit you from setting it up. But it won't be reliable because it uses buffers that could be reassigned for other requests and will break the I/O connection.

 

[jrsnydley]

I installed the ENBT module in the local chassis and all is working properly. I was getting the errors on the PanelView due to mapping the communications through our plant network's gateway module rather than the module that the PanelView was actually connected to.

 

[daba]

Now that you've Ethernet cabled back to the main chassis, can't you do away with ControlNet altogether and have everything communicating via Ethernet ??

Ethernet/IP is so much simpler, no need to schedule causing downtime if you need to add more I/O etc. .

 

[jrsnydley]

That is the long term plan but at least I have the cable run now!

 

[daba]

That is the long term plan but at least I have the cable run now!

...indeed, and all the hardware you need. I can't see how a ten-minute adjustment to the I/O configuration can be called a "long-term" plan.

Perhaps it is time to abandon that plan, and instigate a more immediate and definitely more "do-able" plan.

You can even duplicate the existing I/O on ethernet, leaving the controlnet configuration in-place, just inhibit the local CNB module. That will give you a roll-back if something untoward prevents the Ethernet connections working.

 

[jrsnydley]

Well there are two other remote chassis involved and one involves equipment that is running all week long. As soon as I get some equipment availability it will be done. So, I guess maybe long-term was not the appropriate phrasing.