The definitive guide to DeviceNet
When we talk about process automation protocols, many names pop into our heads. We have analog as the most common, of course. Then for digital protocols, HART, FOUNDATION Fieldbus, and PROFIBUS PA usually come up first. However, you can find other ways out there to integrate a field device to a control system that may suit your application.
Today we’ll talk about DeviceNet, a low-level application-layer protocol for industrial use. You can connect up to 64 nodes to a DeviceNet network. Each node can contain simple sensors or more complex gadgets such as a programmable logic controller (PLC).
DeviceNet uses the Controller Area Network (CAN) standard, a standard for intelligent devices to communicate with each other. Bosch created it in 1985 to replace automotive wiring. And in 1990, Allen-Bradley-Rockwell Automation created DeviceNet as an application layer protocol based on this standard. Allan-Bradley chose CAN as the DeviceNet’s physical layer for its robust physical layer, open tech, and cheap parts. This network also allows different topologies, cables, and more.
You’ll find DeviceNet most often used with simple sensors and actuators. However, you can find some more complex devices with it too, such as the Micro Motion 2400.
DeviceNet uses a trunkline-dropline topology (more on that later), where small lines connect devices to the main line. These small lines should not run more than six meters from the trunkline.
As a loop-powered gizmo, DeviceNet can supply power to many devices directly from its cables, depending on their power needs. It also supports various data exchanges such as cyclic, polled, and strobed. For setup, you can choose between multi-master or master-slave.
In your network, you can have up to 64 nodes, and end-to-end points will depend on your baud rate:
- 125 kilobytes per second (Kbps) – 500 meters
- 250 Kbps – 250 meters
- 500 Kbps – 100 meters
DeviceNet uses common industrial protocol (CIP) for its upper layers. The CIP follows the open systems interconnection (OSI) model in seven layers: physical, data link, network, transport, session, presentation, and application.
With DeviceNet, you can structure your network in many ways. However, you’ll need to know its limitations.
DeviceNet uses 24 volts of direct current and eight amps. With its loop power, you can run low-energy devices. However, if your device requires more juice than the network provides, you’ll need an external power supply. In that case, the network will only serve for communication.
The length of the network will depend on your network speed, as mentioned in the last section, and the type of cables you use. For instance, if you have a flat cable, the trunk distance for a baud rate of 125 Kbps will have to stay within 420 meters. So 250 Kbps needs 200 meters, and 500 will go to 75 meters. You can choose round thick cable, round thin cable, or others to change these numbers.
As you can see in this image, you can configure your droplines in many ways, as long as they remain thinner than your trunkline.
To set up a DeviceNet network, you’ll only need a couple of simple math equations. However, you definitely need to consider your cable type, device energy consumption, current, and other specs to do the math.
EDS stands for electronic datasheet. Like a physical datasheet, a device’s EDS file will give you all the relevant information you need on the device. So you have to have it in your configuration tool to know which device you have connected to the network and what parameters it has. You’ll find these files in many other field networks besides DeviceNet too. Let me give you an example.
Let’s say you want to set up a HART device. So your configuration tool needs the EDS file for that specific device. If your tool uses PROFIBUS for its protocol, then you need the EDS files for that device on that protocol. Does that make sense?
DeviceNet works the same way. Vendors must provide the proper files for their devices that use DeviceNet as their protocol. Often you’ll have to download this file from a vendor’s website, then upload it to your system. If you can’t find the file, then contact your local rep for help. This image shows you how it looks on the Rockwell Automation website.
The way you configure your DeviceNet devices can vary by factors such as the type of control system you have. However, we can give you an overview to point you in the right direction.
As mentioned before, you can have up to 64 nodes on your network. Each node has its own media access control identifier or MAC ID.
When you purchase a device for DeviceNet, it should have a MAC ID. After you set up the device, you also have to set the network where you plan to install it. You’ll need other data such as the class identifier, instance identifier, and attribute identifier as well.
After that, go to your control system and add the DeviceNet scanner. Then the software will browse the network, find all devices connected to the driver, and so on.
Remember to check the MAC IDs of your devices. If you find duplicates, you’ll have to change one of the addresses before you can finish your installation.
This video will show you how to configure Rockwell Automation’s DeviceNet:
You can use DeviceNet error codes either for corrective or preventive maintenance. You can buy specific tools for the DeviceNet network from a number of vendors, like NetAlert for DeviceNet, DeviceNet Detective 2, and others. This is probably the easiest way for most users.
Or you can use the good old multimeter to verify your network. This method will require more knowledge, but you can still do it. So if you found an error code and want to use the first method, then you can get the following:
- Errors – The tool can locate errors and show relevant data for easy analysis, such as minimum or maximum rate and total errors. You can also find and identify errors by nodes.
- Traffic – Studying traffic messages (bandwidth by errors, bus frames per second, etc.) can tell you if you have too many devices on one network. Then you can move them to other networks to reduce the data volume.
- Voltage – You should check this in the field where you have your device installed. Measuring the peak-to-peak value or minimum and maximum may point the way to the problem.
You can also gather more data with your DeviceNet analyzer tools or multimeter, like shield voltage, voltage differential between recessive and dominant, and so on. Somewhere in all this data, you should find your answer.
Can you use a PROFINET gateway with DeviceNet? Short answer: Yes! However, you’ll need a bit of fiddling to make it work.
Say you need to integrate more than one network into your control system. It happens quite often. You can choose to have one network in the field and another at the control system level. For that, you can use bridges or converters.
You can also have a clear connection between the control system and the field devices. For example, if you want PROFIBUS but have HART field devices, then you can use the HART over PROFIBUS concept. In this structure, you have a coupler converting the HART signal for the PROFIBUS. You can also set up devices remotely.
If you have a DeviceNet network in the field but want to connect it to a PROFIBUS control system, the market offers several options. This graphic shows how it works.
You can also check these options and read the specs to see how they might work for you:
- Anybus X-gateway – DeviceNet, PROFINET-IRT device
- Hilsher – Gateway DeviceNet to PROFINET IO devices
- ADF Web – HD67608 PROFINET/DeviceNet converter