ControlNet lifespan

[curlyandshemp]

We are quoting a job in a plant that has horrible ethernet issues. We are quoting IO to be on ControlNet, where a competitor is quoting EtherNet/IP. We guarantee that using ControlNet will work, where the competitor is telling the customer Rockwell are not going to support ControlNet in the future.

Is there any white papers or documents stating that ControlNet will be around? I recently heard at Automation Fair that ControlNet is not going anywhere, but i just need to find written proof.

 

[RussB]

Sorry, I can't answer your question. However, IMHO, Ethernet I/P is going top be around a lot longer than ControlNet.

Have you considered ProfiNet?

 

[curlyandshemp]

Sorry, I can't answer your question. However, IMHO, Ethernet I/P is going top be around a lot longer than ControlNet.

Have you considered ProfiNet?

That may be true, but this particular plant has horrible issues with Ethernet. This is more to do with the physical layer install, and unless the customer is willing to redo the entire etherent install, we are guarantee that 1756 ControlNet IO will work.

I have seen this type of scenerio before, we just finished a major plant rebuild last year. All our IO was on ControlNet. The plant could not get Ethernet to work reliably using Cat5e cabling, and as we were finished our part, the plant was redoing all Ethernet runs in Fiber Optics.

 

[dmargineau]

ControNet will be around for a while...at least until the moment EtherNet/IP will be a fully deterministic network...

Back to you particular issue though...From a developer's point of view, switching from a poorly implemented EtherNet/IP network to ControlNet might be the most efficient way to accomplish the "fix it" task, however, from an end user's one it might be the beginning of a nightmare...
The customer should be fully aware of what it takes in order to maintain and expand a functional ControNet network!
Any device replacement with an identical one carrying a different firmware revision, any node addition/modification or communications structure updating/editing, to name a few, will require a complete stop of the system for scheduling purposes.
You could prevent the above scenario by installing a redundant system or by choosing an unscheduled functionality (Logix platforms only!). The former will pretty much double-up the costs while the latter will not take advantage of the deterministic capability of the network.

IMHO, today, 12/17/2012, switching from EtherNet/IP to ControlNet is nothing but a quick "fix"; a lack of foresight customer might take it as a "pain killer" however, sooner or later, the "surgery" will have to take place...

 

[curlyandshemp]

ControNet will be around for a while...at least until the moment EtherNet/IP will be a fully deterministic network...

Back to you particular issue though...From a developer's point of view, switching from a poorly implemented EtherNet/IP network to ControlNet might be the most efficient way to accomplish the "fix it" task, however, from an end user's one it might be the beginning of a nightmare...
The customer should be fully aware of what it takes in order to maintain and expand a functional ControNet network!
Any device replacement with an identical one carrying a different firmware revision, any node addition/modification or communications structure updating/editing, to name a few, will require a complete stop of the system for scheduling purposes.
You could prevent the above scenario by installing a redundant system or by choosing an unscheduled functionality (Logix platforms only!). The former will pretty much double-up the costs while the latter will not take advantage of the deterministic capability of the network.

IMHO, today, 12/17/2012, switching from EtherNet/IP to ControlNet is nothing but a quick "fix"; a lack of foresight customer might take it as a "pain killer" however, sooner or later, the "surgery" will have to take place...

Actually a quick fix is what I need to do. This plant, as I have mentioned before, has a horrible Etherent install.
The plant started off as Quantum with Modbus+ and S908 RIO.
A previous integrator brought in a Contrologix rack with a 1756-MVI MB+ gateway.
Now Etherenet is in the picture with PV+ replacing failing panelmates.
Switches are popping up everywhere, netweork sprawl is occuring.
PowerFlex VFDs on EtherNet/IP are now added.
1756-racks on EtherNet/IP are now added
ASi Gateways on EtherNet/IP are now added
The plant etherenet is now at the point EtherNet/IP IO faults daily, the ASi gateway does not function, getting online with PLCs is a chore and update times on PV+ screens is slow.

The plant Etherent is a mess, and part of our solution is to get the critical 1756-IO on ControlNet. It is too late for the plant to go back to basics and plan the EtherNet/IP properly as per the Rockwell/Cisco document "Converged Plantwide Ethernet Design and Implementation Guide".

I personally beleive until all plant equipment Ethernet connections get away from the RJ45 and use Fibre or DIN connectors, Ethernet does not belong in an industrial environment. This plant is washdown with high humidity, high temperatures and caustic chemicals in the air.

 

[dmargineau]

I see...Well...ControlNet media is at least as finicky as the Ethernet one if not worse!
I guess you could implement several, parallel networks one of which could be ControNet for the critical I/O; you will have to schedule it and be aware that any functional mishap will bring the network down.
Ethernet communications are far more forgiving than CNet; a ControlNet Scanner will strictly enforce the NUT and if one data packet is missed a Fault will be declared; an EtherNet/IP bridge could "ignore" missing data packets and make up the diference by successfully resending the information within miliseconds.
Again, it's a customer's choice afterall; he will have to deal with it on a 24/7 basis.

 

[curlyandshemp]

I see...Well...ControlNet media is at least as finicky as the Ethernet one if not worse!
I guess you could implement several, parallel networks one of which could be ControNet for the critical I/O; you will have to schedule it and be aware that any functional mishap will bring the network down.
Ethernet communications are far more forgiving than CNet; a ControlNet Scanner will strictly enforce the NUT and if one data packet is missed a Fault will be declared; an EtherNet/IP bridge could "ignore" missing data packets and make up the diference by successfully resending the information within miliseconds.
Again, it's a customer's choice afterall; he will have to deal with it on a 24/7 basis.

Weve been using Cnet since 1999 and the only installs where there have been issues are the ones not using the proper crimping tool, and proper CNET cable.

I still firmly believe the RJ45 connector does not belong in an industrial hostile environment.

 

[brendan.buchan]

Need to know a bit more about the plant, but perhaps a few strategically placed managed switches and one or two fibre runs may solve their issues?

Ethernet/IP definitely seems like the way the industry is heading.. ControlNet, not so much.

 

[Timbert]

EtherNet vs. ControlNet aside.

I did find one document (from 2003) on the ODVA website that states "Meets minimum technology life of 25 years." See http://www.odva.org/Portals/0/Libra...t Numbered/CI_AGM_2003_ControlNet_Gallery.pdf

I also know there is FAQ by Rockwell asking is EtherNet going to replace ControlNet and the answer is basically they are complementary, one isn't going to replace the other.

In the short term, I don't think Rockwell walking away from ControlNet is really any more likely than them leaving any other fieldbus option high and dry. As to what the future holds, who knows for sure.

I did also note there is a number of ControlNet communications devices are placed on Rockwell's "silver series" status. They are mostly v1.25 devices and their replacements are just v1.5 devices.

 

[curlyandshemp]

EtherNet vs. ControlNet aside.

I did find one document (from 2003) on the ODVA website that states "Meets minimum technology life of 25 years." See http://www.odva.org/Portals/0/Libra...t Numbered/CI_AGM_2003_ControlNet_Gallery.pdf

I also know there is FAQ by Rockwell asking is EtherNet going to replace ControlNet and the answer is basically they are complementary, one isn't going to replace the other.

In the short term, I don't think Rockwell walking away from ControlNet is really any more likely than them leaving any other fieldbus option high and dry. As to what the future holds, who knows for sure.

I did also note there is a number of ControlNet communications devices are placed on Rockwell's "silver series" status. They are mostly v1.25 devices and their replacements are just v1.5 devices.

Thanks,
I found the same document. My OP was not about which is better EtherNet/IP or ControlNet as it appears this thread was going off in that direction.
The OP was mainly where is the proof our competitor told the customer that Rockwell are phasing out ControlNet.

I cannot find any such documentation.

 

[curlyandshemp]

Need to know a bit more about the plant, but perhaps a few strategically placed managed switches and one or two fibre runs may solve their issues?

Ethernet/IP definitely seems like the way the industry is heading.. ControlNet, not so much.

Agreed, EtherNet/IP will be the way of the future, but for existing plants, the Physical Layer of the network is the hold back. Still many plants utilize office environment network layouts and wonder why Ethernet problems arise.
I have noticed over the past 5 years or so at Automation Fair how predominate EtherNet/IP is becoming, but it seems only the past 2 or so years that the implemtation of Ethernet in industrial environments are front and center. This year at Automation Fair, Panduit were showcasing their Unified Infrastructure approach with Rockwell.