Adding Ethernet protocol options

If your device has to be able to communicate using more than one industrial Ethernet protocol, the number of options is growing.

By Peter Welander, Control Engineering June 22, 2011

Let’s say you are a manufacturer and you have to build a group of machines that need Ethernet connectivity to transmit data and receive instructions. These could be analyzers, motor control centers, multivariable process sensors, controllers, or just about anything. What are your options?

If these machines are part of a normal product line and you expect to make them again and again, you’ll build the Ethernet interface directly into the main electronics. If these are for a smaller manufacturing lot, you may decide to use a commercial off-the-shelf (COTS) Ethernet module for the sake of expedience.

Now, to further complicate things, let’s say the specification calls for the device to communicate via one of the common industrial Ethernet protocols, such as EtherNet/IP, Profinet, EtherCAT, or others, but not one specifically. The customer (or your marketing department) wants the ability to do a final configuration at the plant during installation, and may even want to go so far as to be able to change to another after some period of time. What this means is that you cannot make the selection permanent at the factory. That is a little more challenging because it eliminates using ASICs (application specific integrated circuits) or other hardwired approaches.

“Most industrial communication can be described in terms of the 7-layer OSI model for communications,” says Tom Weingartner, vice president of marketing for Innovasic Semiconductor. “Typically, layers 1 and 2 are implemented in hardware and layers 3 through 7 are implemented in software. Ideally, you would like one hardware solution and simply change the software for each industrial communication protocol. However, physical connections and electrical characteristics require different hardware for layer 1 from one protocol to another, and real-time response may require different hardware from layer 2 and above.”

What are your options? There are more choices than you may realize.


A field programmable gate array (FPGA) contains a matrix of reconfigurable gate array logic circuitry that can be configured to create a hardware implementation of a software application. Sophisticated development tools from chip suppliers are enabling embedded control system designers to create and more easily adapt FPGA-based applications.

Unlike processors, FPGAs use dedicated hardware for processing logic and do not have an operating system. Because the processing paths are parallel, different operations do not have to compete for the same processing resources. That means speeds can be very fast, and multiple control loops can run on a single FPGA device at different rates.

FPGAs can also be reconfigured, allowing designers and users a high degree of flexibility. A real-life example of this is the FPGA RTEM (real-time Ethernet module) available from Softing. This is a COTS Ethernet connectivity module mentioned earlier. The company says that this unit is designed to offer a cost-effective board-level module for use with low-volume products or when a device manufacturer does not have the necessary R+D bandwidth to design a real-time Ethernet interface from scratch. It serves as a complete interface for connecting field devices to the Profinet, EtherNet/IP, and Modbus/TCP Ethernet protocols.

Softing’s module uses an Altera Cyclone FPGA as its heart. Altera characterizes this design as created for small form factor applications in wireless, wired, military, broadcast, industrial, consumer, and other communications applications. FPGAs use IP (intellectual property) blocks or cores to support the required communication protocol. Using pre-built IP cores helps simplify creating the configuration to implement complex or specialized features. The IP can be protected with encryption with volatile and nonvolatile keys.

They can be reused with different design or size devices, and can include:

  • Hardened memory controllers supporting Mobile DDR, LPDDR2 SDRAM, and 400-MHz DDR3 SDRAM
  • PCI Express Gen2 x1 with multifunction support and
  • Variable-precision digital signal processing (DSP) blocks.

In addition to Altera, FPGAs are available from a number of sources, including Xilinx. Ethernet modules are also available from National Instruments.

Configurable networking processors

In addition to FPGAs, there are processors that can be configured for specific functionality. Here are two examples.

Innovasic manufacturers a controller family called FIDO (flexible input, deterministic output). Simply stated, this is a CPU32+ processor with configurable I/O. The company characterizes this as moving specific real-time operating system (RTOS) functions into the silicon to provide real-time control capability and allow users to develop and debug their code faster than is possible on conventional microcontrollers. Complete solutions are available for industrial Ethernet protocols. Context switching, context management, scheduling, priority control, and memory protection are all built-in. For some applications, this capability eliminates the need for an RTOS altogether, or it may need only a small footprint RTOS.

The chip’s functionality or communication protocol can be changed by downloading new firmware over the Ethernet interface or through the controller’s debug port. Some typical applications are communication adapters, gateways, I/O modules, and communication interfaces in PACs/PLCs.

Hilscher offers its netX controller platform, which it says is a highly integrated network controller with a new system architecture optimized on communication and maximum data transfer. Each communication channel consists of three freely configurable ALUs (arithmetic and logic units), which can be configured with their command set and infrastructure to work with most fieldbus and real-time Ethernet systems.

With four configurable network communication channels, netX can support multiple communication protocols simultaneously. It can be used as a network co-processor with a standard dual-port memory interface or as a highly integrated single-chip solution for your custom control applications.

The central data switch connects via five data paths to the ARM CPU and the communication, graphic, and host controllers with the memory or the peripheral units. This allows the controllers to transmit their data in parallel. The controllers of the four communication channels are structured on two levels and are identical to each other, consisting of dedicated ALUs and special logic units that receive their protocol functions via microcode.

Making a selection

Given this range of options, how do you choose?

“With all of the advancements in FPGAs and processors becoming more configurable, the lines between these types of solutions are really getting blurred,” says Weingartner. “What you are seeing in the market right now is the FPGA guys providing HDL IP blocks so you don’t have to program the FPGA, and the processor guys providing software so you don’t have to program the processor. Which is the better approach? First and foremost is the solution that can meet your industrial Ethernet performance requirements. However, if FPGAs and standard processors all meet some minimum level of performance, that’s when you need to start looking at total ownership cost of the solution and its availability for an automation system’s product lifecycle.”

For more information, visit:

– Peter Welander is a content manager for Control Engineering, pwelander(at)

Control Engineering webcasts include more on Ethernet optimization.