Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 62026-3:2014 – DeviceNet Controller-Device Interface (CDI)
IEC 62026-3:2014 defines DeviceNet, a digital controller-device interface (CDI) designed for low-voltage switchgear and controlgear in industrial automation environments. DeviceNet is one of the most widely deployed industrial communication networks based on the Controller Area Network (CAN) technology, linking programmable controllers, sensors, actuators, and other automation devices over a single multi-drop cable. This standard specifies the physical layer, data link layer, application protocol, and device conformance test requirements to ensure interoperable, reliable communication in factory-floor and process-control applications.
Design Insight: DeviceNet uses a trunk-line/drop-line topology with built-in 24 V DC power distribution, allowing both data and power to share the same cable. This simplifies wiring and reduces installation costs compared to separate power and communication cabling.
1. Physical Layer and Network Topology
The DeviceNet physical layer is built upon CAN transceiver technology operating at a data rate of 125 kbit/s, 250 kbit/s, or 500 kbit/s, selectable per network segment. The standard defines a shielded, four-conductor cable carrying both signal (CAN_H, CAN_L) and power (V+, V-). Network segments are terminated with 120-ohm resistors at each end to minimize signal reflections.
| Parameter | Specification | Notes |
|---|---|---|
| Data Rate | 125 / 250 / 500 kbit/s | Set by network configuration |
| Max Trunk Length | 500 m (125 kbit/s) / 100 m (500 kbit/s) | Depends on data rate |
| Max Drop Length | 6 m per drop (at 500 kbit/s) | Cumulative drop length limited |
| Max Nodes | 64 nodes per network | Including master |
| Cable Type | Thick/Thin shielded twisted pair | Round or flat cable variants |
| Power Supply | 24 V DC ±1% | Shared on trunk cable |
The standard specifies two cable types: thick cable (for trunk lines up to 500 m) and thin cable (for drop connections up to 6 m). Each DeviceNet node includes a transceiver, a CAN controller, and a microprocessor running the DeviceNet application layer protocol. Node addressing is configured via hardware DIP switches or software, supporting up to 64 unique MAC IDs per network segment.
2. Communication Protocol and Object Model
DeviceNet employs the Common Industrial Protocol (CIP) at the application layer, providing a unified object-oriented data model. Every device is modeled as a collection of objects, each with specific attributes, services, and behaviors. The standard defines mandatory objects (Identity Object, Message Router, DeviceNet Object) and application-specific objects depending on the device type.
Design Insight: The object-oriented approach of DeviceNet/CIP allows seamless integration of devices from different manufacturers. Engineers should pay attention to the “Electronic Data Sheet” (EDS) files, which describe device parameters and I/O data formats in a standardized way, enabling configuration tools to auto-detect and configure devices without manual setup.
The protocol supports two types of communication: I/O messages (high-priority, time-critical data exchange using producer-consumer model) and Explicit messages (low-priority, request-response configuration and diagnostic access). The DeviceNet data link layer allocates CAN identifiers according to a predefined connection-ID scheme, ensuring that higher-priority I/O messages always win bus arbitration.
Important: DeviceNet power supply requirements are stringent: the supply must have a maximum rise time of 250 ms and peak-to-peak ripple below 250 mV. Violating these parameters can cause intermittent communication failures that are difficult to diagnose in the field.
3. Conformance Testing and Compliance
IEC 62026-3 defines comprehensive conformance test procedures covering electrical characteristics, EMC immunity, protocol behavior, and power supply performance. The standard requires that devices undergo testing in a prescribed sequence to verify correct operation under normal and fault conditions.
| Test Category | Key Tests | Acceptance Criteria |
|---|---|---|
| Power Supply | Rise time, ripple, current consumption | Rise time ≤250 ms, ripple ≤250 mV |
| Physical Layer | Driver output levels, receiver thresholds | Per CAN transceiver specs |
| Protocol | Duplicate MAC ID, explicit messaging | Valid response within timeout |
| EMC | Immunity to conducted/radiated noise | No communication loss |
The duplicate MAC ID detection protocol ensures that no two devices on the same network share an address. Upon power-up, each device transmits its claimed MAC ID and listens for a response from another device with the same ID. If a conflict is detected, the device enters a fault state rather than corrupting network traffic.
Critical: Bus length and drop length constraints must be respected to maintain signal integrity. Exceeding the maximum cumulative drop length can cause signal reflections, CRC errors, and intermittent node dropouts that are notoriously hard to trace in production environments.
Frequently Asked Questions
Q: Can DeviceNet and CANopen coexist on the same physical bus?
A: No. Although both use CAN as the physical layer, DeviceNet uses a different CAN identifier allocation scheme and application protocol (CIP) that is incompatible with CANopen. They cannot share a bus segment.
Q: What is the maximum number of devices on a DeviceNet network?
A: 64 nodes including the master. The actual practical limit is often lower due to power supply current constraints and cumulative drop length limitations.
Q: How do I troubleshoot intermittent DeviceNet communication failures?
A: Start by checking termination resistors (120-ohm at both ends), power supply ripple and voltage, cumulative drop length, and verify that all nodes have unique MAC IDs. A DeviceNet analyzer tool is essential for capturing bus traffic and identifying CRC errors or late collisions.
Q: Is DeviceNet still relevant with the rise of Ethernet/IP and PROFINET?
A: Yes. DeviceNet remains widely deployed in existing installations, particularly in automotive manufacturing, semiconductor fabrication, and material handling. Many facilities maintain DeviceNet networks alongside newer Ethernet-based systems, and replacement devices remain commercially available.
