How to Chain EtherNet/IP Nodes

Most people are familiar with the EtherNet/IP architecture where some set of EtherNet/IP adapters are connected to one or more EtherNet/IP controllers (Scanners in EtherNet/IP terms). The reason for that is that this type of architecture is very common, but there are exceptions.

Sometimes we have levels of control. For example, look at how ABB tooling robots work in automotive plants. There is a PLC controller that messages the ABB robot using EtherNet/IP. That’s a typical Scanner/Adapter architecture. But if you look within the robot, you find another EtherNet/IP Scanner. That’s the Scanner that connects to the tool using EtherNet/IP IO messaging. In this architecture, there is an EtherNet/IP network within an EtherNet/IP network.

That’s typical of a lot of applications. A ControlLogix controller provides some overall recipe or parameters to a device using EtherNet/IP, and it has some kind of subnetwork to send instructions to one or more lower level devices. You have these kinds of scenarios in food and beverage, plastic injection molding, and many other industries.

Motor drives often use a similar architecture. There is a PLC that sends a motor drive a command (run, jog, stop, etc.) over EtherNet/IP. That motor drive is sometimes connected to other drives that it coordinates with to provide the desired machine motion.

Often, the primary motor drive is an EtherNet/IP Adapter to the main PLC controller and an EtherNet/IP Scanner to the secondary motor drives. That’s an almost identical application to the tooling robot example I gave above.

This architecture is also often used in overhead crane applications. The following drawing is of an Ellsten 2 ton single girder crane.

I don’t know the specifics of this particular crane, but often there are multiple motors in the trolley. Some of these trolleys lift massive weights and need more than one motor to propel them along the girder. Instead of having the PLC Controller message each of those drive motors, there is a master drive motor that communicates, sometimes with EtherNet/IP, to some number of slave motors.

I have never heard of this going three levels deep where the secondary drive motors control some tertiary set of motors, but in principle there is no reason that you couldn’t use that architecture.
EtherNet/IP is a flexible and widely used Ethernet application layer protocol. There are many different architectures for creating applications. The tooling robot and overheard crane applications are just two examples.