Fieldbuses Extend into the Ether

This technology would provide more information to the controller, and greatly reduce the cost and complexity of wiring field devices. But while companies joined organizations to begin developing these networks, other companies were exploring ways to use standard, commercial Ethernet as an industrial fieldbus.

The acceptance of Ethernet in manufacturing by many users, combined with the surge of interest by management for integrating detailed manufacturing data into business applications, spurred the fieldbus organizations into finding ways to extend their device level networks into the realm of commercial Ethernet.

The Fieldbus Foundation (www.fieldbus.org)使用商用现货以太网技术y, called Foundation Fieldbus HSE, at 100 megabits/second (Mbps) as a backbone to its Foundation Fieldbus H1 network. One reason for the popularity of commercial Ethernet TCP/IP (transmission control protocol/ Internet protocol) networks is that they are neutral to application protocols traveling over the wire. Fieldbus Foundation engineers can run the organization’s H1 protocol over Ethernet TCP/IP, enabling communication of many Fieldbus networks to controllers over the one backbone network.

Part of this implementation is the use of “Flexible Function Blocks” (FFBs) as part of the system programming tool to program communication. FFBs, which are application-specific, reside at the fieldbus user layer along with standard function blocks, and enable control strategies such as supervisory data acquisition, batch control, Programmable Logic Controller (PLC) sequencing, coordinated drive control and Input/Output (I/O) interfacing.

Similarly, technical committees of the Open DeviceNet Vendors Association (ODVA,www.odva.org) and ControlNet International (CI,www.controlnet.org) both needed a way to connect their networks to Ethernet. DeviceNet is a device level network limited to 32 devices on a network, while ControlNet was developed as a controller-to-controller network. Both use the Common Industrial Protocol (CIP). This is an application layer protocol that will run on an Ethernet TCP/IP message packet. CIP allows encapsulation of a DeviceNet or ControlNet message within a standard TCP/IP message, bringing these networks into standard Ethernet networking.

In a somewhat unfortunate choice of names, leading to some confusion, the creators called this technology EtherNet/IP (where IP in this case stands for Industrial Protocol). At 100 Mbps, integrators have used EtherNet/IP for I/O level messaging. Designers have integrated the IEEE 1588 standard of the Institute of Electrical and Electronics Engineers, which enables synchronizing of the internal clocks of the processors on the EtherNet/IP network in order to perform motion control messaging with no additional hardware required.

A slow fade

Profibus (www.profibus.com) developers also saw the need to connect its device network to Ethernet. ProfiNet also uses the standard Ethernet and TCP/IP stack to implement the solution. However, ProfiNet is also a component-based solution—in the sense of software components. Therefore, it uses the Microsoft Corp. technology of Component Object Model (COM) and Distributed COM (DCOM). These technologies are no longer supported by Microsoft, but proponents argue that there is so much of this technology in existence that it will not fade away for many years.

The components are instantiated through the use of function blocks. Unlike Foundation Fieldbus, ProfiNet uses function blocks defined by the IEC 61131-3 standard of the International Electrotechnical Commission. It is also using distributed function blocks as defined by IEC 61499 for network-to-network communication. ProfiNet uses a custom chip solution for its real-time implementation, and seems to be moving to a chip solution for standard ProfiNet as well.

Whether one solution or another is better is best left to individual users to decide. The clear conclusion is that Ethernet is the way to implement plant-wide communications.