What’s the Difference Between DeviceNet and ControlNet?

"What is the best network protocol?"

This single question can be a hotly debated topic between engineers, so this article will offer a surface-level look at some of the differences between DeviceNet and ControlNet. It will also explain why Ethernet/IP may dominate many applications, but may be difficult to entirely take over the industry.

As network technology grows and evolves, it is important to understand what's happening between the devices that deliver value to a production line. Start with getting to know what resources are currently available in different areas of a factory, such as power and network availability.

If something isn't broken, why fix it? An initial thought when working with legacy equipment or networks is to replace it with the latest and greatest. However, older equipment might not be able to communicate or even plug into modern networks. Which technology is best? Like many solutions, engineers find niche applications that result in using multiple technologies to provide the greatest value. Before throwing out old systems, make sure an upgrade is even necessary. There are solutions that bridge older technologies to newer devices and networks.

Positives and Negatives of DeviceNet and ControlNet?

In brief, some of the pros and cons can be simplified according to the following table.

What is DeviceNet?

Many companies just starting a digital transformation plan might only be looking for data to see where a connected device or advanced automation would offer the greatest ROI. This might require multiple sensors or simple devices that need two things: connection and power.

Figure 1. DeviceNet trunk/drop line topology. Image used courtesy of Rockwell Automation

DeviceNet offers both power and signal in a single 4-conductor cable to connect up to 64 devices, each with a specific node address. Power and signal in one cable can save on cost. Additionally, a trunk-and-drop line architecture makes integrating new devices relatively easy compared to Ethernet/IP, which often requires individual wires for each device. Additionally, DeviceNet’s architecture and individual node addresses work well for maintenance and troubleshooting.

DeviceNet is used to connect field-level compatible devices (like pushbuttons, limit switches, and some sensors), I/O modules, and motor drives to a PLC. As indicated in the name, DeviceNet is for device-level connections and limited sizes of data packets. In contrast, ControlNet facilitates communication between PLCs, drives, HMIs, and PCs, so the data size is much larger.

DeviceNet: Distance and Data Rates

Data from Rockwell Automation defines some of the specifications of the network. While limited on bandwidth and distance, DeviceNet still works well for connecting devices because many sensors don’t require a large bandwidth or a high network speed.

The cumulative drop line length, thick or thin, can not exceed the maximum length allowed for the data rate used.

This network is a quick and affordable way to connect simple IIoT devices and sensors on a relatively local network. Companies starting to connect processes and production lines can use DeviceNet, but need to keep in mind its digital transformation plans to scale. If distance or data rates become limiting, DeviceNet can work with other networks, such as Ethernet/IP. For example, Anybus X-Gateway connects DeviceNet devices or networks to many other Fieldbus or industrial Ethernet networks. 

DeviceNet can also work with ControlNet. Both ControlNet and DeviceNet are managed by the Open DeviceNet Vendor Association, known as ODVA. They have the same 7 Common Industrial Protocol (CIP) layers known as the Open System Interconnection (OSI) Model.

ControlNet, in contrast, can connect up to 99 devices and offers faster data rates. If DeviceNet is for companies looking for data to start a digital transformation and find where more connected solutions are needed, the network on which those time-sensitive and advanced automation devices could operate is ControlNet.  

What is ControlNet?

Advanced automation and IIoT solutions require more data, faster speeds, and more reliable handling of data. ControlNet's fast speeds make it valuable in applications such as vision systems, weld controls, motion control, robotics, and more. Deterministic, redundant, and time-sensitive applications might choose ControlNet over DeviceNet or Ethernet. 

ControlNet can use fiber optics, operate up to 1000 meters without a repeater, connect up to 99 devices, has data transfer rates of 5 Mbps, and offers multiple topologies. This network uses a logical token ring or token-passing BUS control network.

Figure 2. Control-level devices on a ControlNet network must be equipped with proper CN adapters. Image used courtesy of Rockwell Automation

The timing is set by the network update time (NUT), configurable between 2 and 100 ms. This timing is made up of three parts: scheduled, unscheduled, and guardband. Scheduled service time is used to communicate with each node for time-sensitive or priority messaging. If a node has no data to transfer, it will still send a short message to show it is still transmitting. Without this communication, a provision in the ControlNet protocol will regenerate the token and send it to the next node in the ring. 

Unscheduled service times can be used for non-time-critical messaging. This separation allows critical messaging to be sent over the network without interfering with time-sensitive controls and I/O data. The guardband helps ensure that the scheduled and unscheduled service times don’t interfere with one another.

ControlNet also uses Requested Packet Intervals (RPI) to determine how often I/O data will be transmitted during the scheduled NUT. In short, the NUT tells you how often the network or ring updates, and the RPI tells you how often a device is asked to transmit on the network.  

The Best Network

There are different tools for different solutions. Ethernet is widely adopted, easy to use, and offers the greatest bandwidth. However, cost, ease of integration, speed, and time-sensitive applications will keep DeviceNet and ControlNet around, if only in niche and legacy production, for a bit longer. More importantly, it proves that the engineering response to everything - whether what device is better, or what’s the best network - still remains as before: it depends. All of the design factors must be carefully considered before selecting the adapters and installing the equipment.

Revised: Original article published 6/2020.