Distributed Real-Time Systems

(TI-DRTS) Track 2
CAN-BUS
Introduction

Version 9.11.2009
Ref. VECTOR application note & Motorola note
 What is CAN?
Controller Area Network (CAN)
is a common, small area network
solution that supports distributed
product and distributed system
architectures
The CAN bus is used to
interconnect a network of
electronic nodes or modules
Typically, a two wire, twisted
pair cable is used for the network
interconnection

Ingenirhjskolen i rhus Slide 2

 CAN - Highlights

Is a high-integrity serial data communications

bus for real-time applications
Is event driven
Operates at data rates of up to 1 Mbits/s
Has excellent error detection capabilities
Was originally developed by Bosch for use in
cars
Is now being used in many other industrial
automation and control applications
Is an international standard: ISO 11898

Ingenirhjskolen i rhus Slide 3

 Why use CAN?

Ingenirhjskolen i rhus Slide 4

 CAN History

CAN first introduced by Bosch in 1986

Bosch published CAN specification version 2.0
in 1991
Official ISO CAN standard in 1993: ISO 11898
In 1999 57 million CAN controller chips sold
Estimated to 300 million CAN chips in 2003

Ingenirhjskolen i rhus Slide 5

 Field Buses in the
Automotive Industry

Ingenirhjskolen i rhus Slide 6

 Typical CAN Network

Ingenirhjskolen i rhus Slide 7

 CAN Standards
Although CAN was originally developed in Europe
by Robert Bosch for automotive applications, the
protocol has gained wide acceptance and has
become an open, international ISO standard
As a result, the Bosch CAN 2.0B specification has
become the de facto standard that new CAN chip
designs follow
CAN ISO Standardization:
ISO PRF 16845 : CAN Conformance Test Plan
ISO PRF 11898-1: CAN Transfer Layer
ISO PRF 11898-2: CAN High Speed Physical Layer
ISO DIS 11898-3: CAN Fault Tolerant Physical Layer
ISO DIS 11898-4: TTCAN Time Triggered CAN

Ingenirhjskolen i rhus Slide 8

 CAN 2.0 Structure of a CAN Node
ISO/OSI Reference Model

CAN Specification,
ISO 11898, deals only
with the Physical & Data
Link Layers for a CAN
network

Ingenirhjskolen i rhus Slide 9

 Three-Layer Reference Model

Application Process

Layer 7 Application Layer

Management
Layer 2 Data Link Layer
Layer 1 Physical Layer

Ingenirhjskolen i rhus Slide 10

 Key Reasons to Use CAN

Low connect cost

Low cost components
Growing number of CAN chips
Increasing knowledge base
Increasing integration service base
Wide variety of CAN-based products
Wide variety of Off-the-Shelf tools available
Potential lower wiring costs
Lower weight

Ingenirhjskolen i rhus Slide 11

 What Industries are using CAN?

Ingenirhjskolen i rhus Slide 12

 CAN Higher Layer Protocols (1)

CAN is used as the basis for several major

"7-layer" protocol developments such as:
CAL: CAN Application Layer
(CiA: Can In Automation)
CAN Kingdom (Kvasar)
CANopen (CiA: Can In Automation)
DeviceNet (Rockwell Automation, ODVA)
Volcano (Developed by Volvo)
SAE J1939 (Society of Automotive Engineers)
TTCAN: Time Triggered CAN (Bosch)

Ingenirhjskolen i rhus Slide 13

 CAN Higher Layer Protocols (2)

Each of these large protocol architectures

are essentially complete industry-specific
network solutions packaged to include
defined requirements for the:
physical layer,
address structure & message structure
conversation structure
data structure
application/network interface

Ingenirhjskolen i rhus Slide 14

 How does CAN operate?

CAN is a multiplexed serial communication channel

Can transfer up to 8 data bytes within a single message
For larger amounts of data, multiple messages are
commonly used
Most CAN based networks select a single bit rate
While communication bit rates may be as high as 1 MBit/s,
most implementations are 500Kbit/s or less
CAN supports data transfers between multiple peers
No master controller is needed to supervise the
network conversation

Ingenirhjskolen i rhus Slide 15

 Bit Rate versus Bus Length
Bit Rate Maximum Bus
(kBits/s) length (m)
1000 50
500 110
135 620
100 790
50 1640

A Rule of Thumb for bus length > 100 m:

Bit Rate (Mbit/s) * Lmax (m) <=60
[Ref: Etschberger]
Ingenirhjskolen i rhus Slide 16
 Bit Rate versus Bus Length

[Ref: Etschberger]
Ingenirhjskolen i rhus Slide 17
 CAN Identifiers

Labels the content (type) of a message

used by receivers to select a message
Used for arbitration & determines the
priority of the message
Low id.number = high priority

Ingenirhjskolen i rhus Slide 18

 CAN 2.0A vs CAN 2.0B

Ingenirhjskolen i rhus Slide 19

 CAN 2.0A Message Frame
CAN 2.0A (Standard Format)
11 bit message identifier (2048 different frames)
Transmits and receives only standard format messages

Ingenirhjskolen i rhus Slide 20

 CAN Message Frame

Ingenirhjskolen i rhus Slide 21

 CAN 2.0B Message Frame
CAN 2.0B (Extended Format)
Capable of receiving CAN 2.0A messages
29 bit message identifier (512 million frames)
11 bits for a CAN 2.0A message + 18 bits for a CAN 2.0B
message

Ingenirhjskolen i rhus Slide 22

 Two Communcation Types

Transmitter Receiver 1
1. Transmitter
Data Frame
initated
(ID=100, RTR Bit = 0) Receiver 2

2. Remote Transmission Request:

Remote Frame
(ID=105, RTR Bit = 1, Data Length=7)

Data Frame Receiver 7

(ID=105, RTR Bit = 0, Data Length= 7)

Ingenirhjskolen i rhus Slide 23

 Arbitration (1)

Carrier Sense, Multiple Access with

Collision Avoidance (CSMA/CA)
Method used to arbitrate and determine the
priority of messages
Uses enhanced capability of non-destructive
bitwise arbitration to provide collision
resolution

Ingenirhjskolen i rhus Slide 24

 Arbitration (2)

A station may send if the bus is free (carrier

sense)
Any message begins with a field for unique bus
arbitration containing the message ID
The station with the lowest ID is dominant
(D-Bit)
So the lowest ID has highest priority
Sending is not interfered since the propagation
on the bus is much smaller than a duration of a
bit

Ingenirhjskolen i rhus Slide 25

 Bitwise Arbitration
Any potential bus conflicts are resolved by bitwise
arbitration
Dominant state (logic 0) has precedence over a
recessive state (logic 1)

Competition for the bus is won by node 2 .

Nodes 0 and 1 automatically become receivers of
the message
Nodes 0 and 1 will re-transmit their messages when
the bus becomes available again
Ingenirhjskolen i rhus Slide 26
 Example of Bitwise Arbitration

Ingenirhjskolen i rhus Slide 27

 Qualities: Safe Collision and
Tx-feedback
The CAN-controller has 2 important features:
A collision do not destroy any message on the
bus
All Txs with recessive levels stops immediately and
changes to Rxs.
The Tx-node with highest priority wins the bus and sends
data
Every transmitted message is evaluated by each
receiving node, and if the received message is
damaged the Tx-node is alerted with a feedback
at dominant level, sent from the Rx-node

Ingenirhjskolen i rhus Slide 28

 Error Detection

CAN implements five error detection

mechanisms
Three at the message level
Cyclic Redundancy Checks (CRC)
Frame Checks
Acknowledgment Error Checks
Two at the bit level
Bit Monitoring
Bit Stuffing

Ingenirhjskolen i rhus Slide 29

 Cyclic Redundancy Check (CRC)
(Message Level)

The 15 bit CRC is computed by the

transmitter based on the message
content
All receivers that accept the message,
recalculates the CRC and compares
against the received CRC
If the two values do not match a CRC
error is flagged

Ingenirhjskolen i rhus Slide 30

 Frame Check (Message Level)
If a receiver detects an invalid bit in one of
these positions a Form Error (or Format
Error) will be flagged:
CRC Delimiter
ACK Delimiter
End of Frame Bit Field
Interframe Space (the 3 bit INTermission field and a possible
Bus Idle time)

Ingenirhjskolen i rhus Slide 31

 ACK Error Check (Message Level)

Each receiving node writes a dominant bit

into the ACK slot
If a transmitter determines that a message
has not been ACKnowledged then an ACK
Error is flagged.
ACK errors may occur because of
transmission errors because the ACK field
has been corrupted or there is no operational
receivers

Ingenirhjskolen i rhus Slide 32

 Bit Monitoring (Bit Level)

Each bit level (dominant or recessive)

on the bus is monitored by the
transmitting node
Bit monitoring is not performed during
arbitration or on the ACK Slot

Ingenirhjskolen i rhus Slide 33

 Bit Stuffing (Bit Level)

Bit stuffing is used to guarantee enough

edges in the NRZ bit stream to maintain
synchronization:
After five identical and consecutive bit levels have
been transmitted, the transmitter will
automatically inject (stuff) a bit of the opposite
polarity into the bit stream
Receivers of the message will automatically
delete (destuff) such bits
If any node detects six consecutive bits of the
same level, a stuff error is flagged

Ingenirhjskolen i rhus Slide 34

 Error Flag

If an error is detected by at least one

node
The node that detects the error will
immediately abort the transmission by
sending an Error Flag
An Error Flag consists of six dominant
bits
This violates the bit stuffing rule and all
other nodes respond by also transmitting
Error Flags

Ingenirhjskolen i rhus Slide 35

 What is needed to implement CAN?

Ingenirhjskolen i rhus Slide 36

 Requirements for a CAN Controller

Ingenirhjskolen i rhus Slide 37

 FullCAN vs BasicCAN Controller

FullCAN Controller:
Typically 16 message buffers, sometimes
more
Global and Dedicated Message Filtering
Masks
Dedicated H/W for Reducing CPU Workload
More Silicon => more cost
e.g. Powertrain

BasicCAN Controller:
1 or 2 Tx and Rx buffers
Minimal Filtering
More Software Intervention
Low cost
e.g. Car Body

Ingenirhjskolen i rhus Slide 38

 CANopen
CANopen: a standardized application for distributed
industrial automation systems
Based on CAN standard and CAL (Can Application
Layer)
In Europe the definitive standard for the
implementation of industrial CAN-based system
solutions
Standardized by CiA (CAN-in-Automation)
Devices profiles
e.g. digital/analog I/O modules, drives, encoders, MMI-units,
controllers
Two types of communication mechanisms:
unconfirmed transmission of data frames to transfer process
data
confirmed transmission of data (for configuration purpose)

Ingenirhjskolen i rhus Slide 39

 DeviceNet

DeviceNet developed by Rockwell

Automation in 1995
Main CAN automation technology in USA and
Asia
ODVA: Open DeviceNet Vendor Assocation
(>300 members)
DeviceNet is a connection-based
communication model (ConnectionId = CAN
identifier)
Two message types: explicit and I/O
messages
Max 64 nodes in a DeviceNet network

Ingenirhjskolen i rhus Slide 40

 Three Development Scenarios
based on a
based on a
based on a layer-2 standardized
standard
implementation application
application
profile
layer
Application
Application
Application
CANopen /
CAL/
DeviceNet
CAN Kingdom

CAN CAN CAN

Ingenirhjskolen i rhus Slide 41

 CAN Summary

CAN is designed for asynchronous

communication (event communication) with
little information contents (8 bytes)
Max 1MBit/s
Useful for soft real-time systems
Many microcontrollers comes with an
integrated CAN controller
A low-cost solution
New invention:
TTCAN a Time-Triggered CAN protocol

Ingenirhjskolen i rhus Slide 42

 References

[Etschberger]: Controller Area Network

basics, protocols, chips and applications,
by Konrad Etschberger, IXXAT Press, 2001
www.can.bosch.com
Contains specification documents
References
Links
ODVA: Open DeviceNet Vendors Association
www.odva.org
CiA: CAN in Automation
http://www.can-cia.org/

Ingenirhjskolen i rhus Slide 43