Transcript Freescale presentation template

MegaMeet 2007
MegaSquirt and MegaSquirt Logo are trademarks of BG Soflex, LLC.
CAN
Controller Area Network
Bowling & Grippo
MegaSquirt and MegaSquirt Logo are trademarks of BG Soflex, LLC.
Quick CAN Overview
MegaSquirt and MegaSquirt Logo are trademarks of BG Soflex, LLC.
Understanding CAN

Simple Serial Communications Protocol

Variable Bit Rate: 5 Kbps up to 1 Mbps

Un-limited node (limited to max. Electric load and time
delay)

Multi-Master (any node may initiate transmit any frame)

Multi-cast (any node may receive all messages)

CSMA/CD standard (Carrier Sense, Multiple Access &
Collision Detect)

Prioritization of messages via identifier

Fault confinement (determine permanent failures then
shut off node)

Automatic re-transmission of corrupted message
5 High level of error detection
Requirements of a CAN Controller

Simple user interface to CPU
–
–
–
Microcontroller
Access control & status registers
Access to buffers
Interrupt and error types
• Message filtering & buffering
– Store incoming & outgoing messages
– Only interrupt CPU w/ relevant messages
– Predictable Message Transmission
• Protocol handling
CAN
Transmit
Receive
Engine
– Error Detection
– Arbitration detection
– Bit monitoring/stuffing
Message
filtering +
buffering
CPU
Interface
• Physical layer interface
– Current & voltage control for bus
– Absorb transients
– Signal bus (line) faults & correct
CANH
CAN bus
CANL
6
Control
+ status
Physical
interface
TX
RX
H/W
Errors
CAN Terminology
7

CAN 2.0
- Robert Bosch CAN Specification, Revision 2.0

CAN 2.0 A
- Formerly CAN 1.2, limited to 11-bit identifiers

CAN 2.0 B
- Usually means complete protocol standard
- Including 11-bit and 29-bit identifiers
- Should read CAN 2.0 A/B

“Full-CAN”
buffers
standard
- Hardware implementation having at least 16 message
- Means nothing about compliance to CAN 2.0 A/B

MSCANxx
- Freescale Scalable CAN hardware implementation
(HC08, HC12, HCS12 families)

TOUCAN
- “Full-CAN” Freescale CAN hardware implementation
(M*CORE, PowerPC families)

FLEXCAN
- Similar to, but larger version of TouCAN
CAN Node/Message
Structure
MegaSquirt and MegaSquirt Logo are trademarks of BG Soflex, LLC.
CAN – Node Structure
CAN Protocol
Data Link Layer
Logical Link Control (LLC) sub-layer
Acceptance Filtering
Overload Notification
Recovery Management
OSI Layers
APPLICATION
PRESENTATION
Medium Access Control (MAC) sub-layer
Data Encapsulation/Decapsulation
Frame Coding (Bit
Stuffing/Unstuffing)
Medium Access Management
Error Detection/Signaling
Acknowledgement
Serialization/Dserialization
Physical Layer
Bit Encoding/Decoding
Bit Timing
Synchronization
9
SESSION
TRANSPORT
NETWORK
DATA LINK
PHYSICAL
OSEK COM/NM,
Vector Driver S/W
msCAN,TouCAN
CAN-B, CAN-C
Transceivers
CAN Message Transfer
10

Information sent in fixed format message frames.

Any node may start to transmit when bus is free (Bus Idle).

If two or more nodes start transmitting in same frame,
bus access conflict is resolved by bit-wise arbitration.

Highest priority message wins bus access.

Arbitration Field (identifier) determines the message priority.

Transmitting nodes which lose arbitration become receivers
and automatically re-transmit at next available time.

No data or time wasted, someone always wins.
CAN Message Transfer (Cont.)
11

All nodes check consistency of messages received and
will flag an inconsistent message to the entire network.

All receiving nodes acknowledge a valid message.

A message is received correctly by all nodes or no nodes.

All nodes apply Message Filtering to decide whether
to accept a message.

Any number of nodes can simultaneously receive and
accept a message (Multicast transmission).

ID - Labels message contents (no physical node addresses).
CAN – Message Types

Data Frame:
 (“Hi Everyone, I am sending data to Labeled X”)

Remote Frame:
 (“Hi Everyone, If you are Labeled Y; please send me
data”)

Error Frame:
 (“Everyone STOP!!!, Let’s try it again”)

Overload Frame:
 (“I am busy Labeled Z, please wait until I am free”)
12
CAN – Data Frame
recessive
IFS or SOF Arbitration Field Control Field
Bus Idle 1-bit 11-bits (Std ID)
6-bits
29-bits (Ext ID)
dominant
Data Field
CRC Field ACK
EOF
0-8 bytes
16-bits
7-bits
2-bits
IFS
Bit Stuffing
 A node is allowed to start transmitting a Data Frame after Inter-Frame
Space(IFS)
 The two complementary bus values are called dominant and recessive
 All receivers synchronize to leading edge of Start Of Frame (SOF)
 Arbitration Field is 11-bits for a Standard Format Data Frame (CAN2.0A/B)
 Arbitration Field is 29-bits for a Extended Format Data Frame (CAN2.0B only)
 Bit Stuffing
 whenever 5 consecutive bits of equal value are transmitted, 1 extra bit of
complementary value is automatically inserted into the bit stream: provides
edges for clock resynchronization. Receivers automatically de-stuff
13
CAN – Remote Frame
recessive
IFS or SOF
Bus Idle 1-bit
Arbitration Field
11-bits (Std ID)
29-bits (Ext ID)
Control Field
6-bits
CRC Field
ACK
EOF
16-bits
2-bits
7-bits
IFS
dominant
Bit Stuffing
 A node is allowed to start transmitting a Remote Transmission Request (RTR)
Frame after Inter-Frame Space (IFS)
 Similar to a Data Frame, but no Data Field
 Interpreted as a request for data that is associated with the Identifier in the
Arbitration Field
 Seldom used by automotive manufacture
14
CAN – Error Frame (Active)
recessive
Data Frame
Or
RTR Frame
Active Error
Flag
6-Bits
Echo Error
Flag
<= 6-Bits
Echo Delimiter
8-bits
Intermission
3-bits
dominant
Error Frame = Error Flag + Error Delimiter
2 types of Error Flag: (Active Error Flag and Passive Error Flag)
 Active Error Flags :
 Transmitted when error detected by an Error Active node
 Consists of 6 dominant bits (overwrite other bits on bus)
 Breaks rule of Bit Stuffing or destroy the ACK field or EOF field
 Consequently, all other nodes detect an error, and transmit own Error
Flag, (and no nodes receive the original message)
Number of dominant bits monitored on the bus can vary between 6 and 12.
15
CAN – Error Frame (Passive)
recessive
Data Frame
Or
RTR Frame
Passive Error
Flag
6 Bits
Echo Error
Flag
<= 6-Bits
Echo Delimiter
8-bits
Intermission
3-bits
dominant
 Passive Error Flags :
 Transmitted when error detected by an Error Passive node
 Consists of 6 dominant bits (do not overwrite other bits on bus)
 Complete when the Error Passive node has received 6
consecutive bits of equal value, beginning from start of Passive
Error Flag
 Passive Error Frame Delimiter consists of 8 recessive bits.
16
CAN – Overload Frame
recessive
EOF or
Error Delimiter or
Overload Delimiter
Overload
Flag
6-Bits
Echo Error
Flag
Overload Delimiter
8-bits
dominant
 Overload Frame transmitted when any of following are detected:
 Dominant bit during the 1st or 2nd bit of Intermission
 Dominant bit during the 7th bit of the end-of-frame (EOF) field in receive
frames.
 Dominant bit during the 8th (last) bit of Error Delimiter or Overload
Delimiter
 Receiver permitted to initiate an Overload Frame when not ready
to receive new data (not usually implemented)
 Maximum of 2 Consecutive Overload Frames may be generated
17
CAN – Standard Frame Format
18
CAN- Extended Frame Format
19
CAN PHY Transmission Standard
 Three automotive PHY standards:
 High Speed Differential Interfaces (ISO 11898-2, SAE J2284)
 Up to 1 Mbps differential
 Linear bus topology
 Silicon solutions to common mode noise issues
 Two-Wire Fault-tolerant interfaces (ISO 11898-3)
 Automatically switch to single wire reception when fault detected
 Limited to 125 kbps due to propagation delays
 Flexible topology: bus, star, ring
 Single Wire Interfaces (SAE J2411)
 Based on J1850-VPW technology with enhanced wakeup capability
 Limited to 33.33 kbps due to propagation delays
 Flexible topology: bus, star, ring
20
CAN Standard – High Speed





CAN-C (J2284)
Linear bus topology
500 kbps bit rate
30 m max bus length
Maximum of 16 nodes
CANH
3.5 V
2.5 V
CANL
1.5 V
recessive
dominant
recessive
Vcc
TX
ECU 2
RX
Stubs < 1 m
RT
120Ω
CANH
CANL
21
...
ECU n
ECU 1
RT
Gnd
Twisted-pair media
120Ω
CAN Standard – Fault Tolerant






Flexible bus topology
<=125 kbps bit rate
Auto-switch to Single-ended
Receive on Bus fault
40 m max bus length
Maximum of 32 nodes
Vcc
R
RX
ECU 1
ECU 2
Gnd
3.5 V
1.5 V
0V
recessive
L
TX
Vcc
CANH
CANL
recessive
dominant
...
ECU n
RH
Twisted-pair media
CANH
CANL
22
CAN Standard – Single Wire





Flexible bus topology
33.33 kbps bit rate
High Voltage Wakeup
40 m Max bus length
Max of 32 nodes
VBatt
Wakeup signal level
4.1V
CANH
0V
VBatt
recessive
Tx
dominant
recessive
Rx
ECU 2
RT
VBatt
ECU 1
CANH
23
LOG
...
Gnd
Twisted-pair media
ECU n
CAN Physical Interfaces
High Speed
Differential
Fault Tolerant
Differential
Single Wire
Number of
Bus Wires
2
2
1
(Ground Reference)
Maximum
Bus Speed
500 kbps
125 kbps
33.33 kbps
(Ltd by prop delay)
(Ltd by prop delay)
Bus
Topologies
Linear
Bus, Star, Ring
Bus, Star, Ring
Automotive
Standards
Documents
ISO 11898
SAE J2284
ISO - To Be Determined
SAE J2411
MC33989 (SBC),
PCA82C250
MC33388,
MC33889 (SBC Lite),
MC33389 (SBC),
PCA82C252
Parts
Available
24
CAN Bit Timing
BIT TIME
SYNC_SEG
PROP_SEG
PHASE_SEG1
Transmit Point
PHASE_SEG2
Sample Point
SYNC_SEG: The bit edge is expected to lie within this segment.
PROP_SEG: Allowance for physical delays.
PHASE_SEG1 & PHASE_SEG2: Define the position of the Sample Point. May
be adjusted to compensate for edge
phase errors.
SAMPLE POINT: The bus value at this point is taken as the value of the bit (if 3
samples per bit taken, this is position of 3rd sample).
25
CAN Bit Timing
BIT TIME = SYNC_SEG + PROP_SEG + PHASE_SEG1 + PHASE_SEG2
SYNC_SEG = Always 1 time quantum (CAN clock period)
PROP_SEG = 1- 8 time quanta (programmable)
PROP_SEG >= 2 x (bus propagation delay + input comparator delay +
output driver delay)
PHASE_SEG1 = 1- 8 time quanta (programmable)
PHASE_SEG2 = greater of
PHASE_SEG1 and INFORMATION PROCESSING TIME
… where INFORMATION PROCESSING TIME <= 2 time quanta
26
CAN Bit Timing Examples
Example 1
Desired Bit Rate = 1 Mbps
CAN Clock = 10 MHz
Delay of driver = 30 ns
Delay of receiver = 50 ns
Delay of bus line (40 m) = 220 ns
 Total propagation delay = 300 ns
Time quantum = 100 ns
Bit Time = 10 tq
SYNC_SEG = 1 tq
PROP_SEG = 6 tq (min)
PHASE_SEG1 = 1 tq
PHASE_SEG2 = 2 tq
SJW = 1 tq
Sample Point at 80% of bit time (Target for 1 Mbps should be around 80-85%)
27
CAN Bit Timing Examples
Example 2
Bit Rate = 125 kbps
CAN Clock = 2 MHz
Delay of driver = 50 ns
Delay of receiver = 70 ns
Delay of bus line (40 m) = 220 ns
Total propagation delay = 340 ns
Time quantum = 500ns
Bit Time = 16 tq
SYNC_SEG = 1 tq
PROP_SEG = 3 tq (min)
PHASE_SEG1 = 6 tq
PHASE_SEG2 = 6 tq
SJW = 4 tq
Sample Point at 75% of bit time (Target for 125kpbs should be around 75-80%)
28
CAN Synchronization

Hard Synchronization is performed at the leading edge of SOF:
SYNC_SEG begins with this edge.
All other recessive to dominant edges will be used for
Resynchronization.
(Exception: Transmitter will not resynchronize to edge with positive phase
error)
Sample Point
• If edge lies within:

SYNC_SEG
Phase Error = 0
• In Resynchronization,
PROP_SEG
PHASE_SEG1
Phase Error > 0
PHASE_SEG2
Phase Error < 0
(for next bit)
PHASE_SEG1 may be lengthened
or PHASE_SEG2 may be shortened
• Resynchronization Jump Width = maximum lengthening or shortening allowed
• Resynchronization Jump Width = 1 - 4 time quanta (programmable )
Maximum = time quanta in PHASE_SEG1
29
CAN Synchronization
Edge appears here...
Edge should be here...
Edge appears here...
Edge should be here...
r
CAN
bus d
Receiver
bit timing
S PS
P1
P2
S PS
P1
P2
S PS
P1
P2
S PS
P1
P2
S PS
P1
P2
Sample points
BIT 1
S = SYNC_SEG
PS = PROP_SEG
P1 = PHASE_SEG1
P2 = PHASE_SEG2
30
BIT 2
BIT 3
BIT 4
BIT 5
Bit 3
Bit 5
Negative Phase Error
Positive Phase Error
PHASE_SEG2 shortened
PHASE_SEG1 lengthened
CAN Oscillator Tolerance














31
Oscillator tolerance is set by the requirement to remain synchronized
between recessive to dominant edges.
| F - Fn |
dF = ————
Fn
Fn = nominal CAN clock freq.
F = actual CAN clock freq.
Requirement 1: correctly sample the first bit after sending
active error flag after a stuff error.
Min (tP_SEG1, tP_SEG2)
dF < — ——————————
2 x (13 x tBitTime - tP_SEG2)
Requirement 2: synchronize correctly in the stuffed part of the bit stream.
tSJW
dF < ——————————
2 x (10 x tBitTime)
CAN Oscillator Tolerance - Examples
 Example
1
1 Mbps
 Bit Time
=
10 tq
 SYNC_SEG =
1 tq
 PROP_SEG
=
6 tq
 PHASE_SEG1 =
1 tq
 PHASE_SEG2 =
2 tq
 SJW
= 1 tq
 Requirement
 Example
2
125 kbps
 Bit Time
=
16 tq
 SYNC_SEG
=
1 tq
 PROP_SEG
= 3 tq
 PHASE_SEG1 =
6 tq
 PHASE_SEG2 =
6 tq
 SJW
=
4 tq
1
 Requirement 1
 dF < 0.39% over 12.8-Bit
Times
(12.8  dF < 1.48% over 12.6-Bit Times
ms)
(101 ms)
 Requirement 2
 Requirement 2
 dF < 0.50% over 10.0-Bit
 dF < 1.25% over 10-Bit Times
Times
(10 ms)
(80 ms)
32
MSCAN Module
msCAN Receive / Transmit Engine
CPU Interface (Memory
Mapped I/O)
TX Buffer 0
Internal
Priority
Scheduling
Priority Register
TX Buffer 1
Global Identifier Filtering:
Priority Register
2 x 32-bit
TX Buffer 2
or 4 x 16-bit
Priority Register
or 8 x 8-bit
RX Buffer
RX Buffer
33
MSCAN12 Module

Implementation of the CAN protocol - Version CAN 2.0A/B
–
–
–

Double buffered receive storage system

Triple buffered transmit storage scheme with internal prioritization
using “local priority” concept
Flexible maskable identifier filters
Programmable wake-up functionality with integrated low-pass filter
Programmable loop-back mode supports self-test
Separate signaling and interrupt capabilities for all CAN receiver and
transmitter error states
Programmable clock source (PLL or oscillator)
Programmable link to on-chip timer module for time stamping or
network synchronization
Low-power sleep mode







34
Standard and extended data frames
0 - 8 bytes data length
programmable bit rate up to 1 MBit/s
MegaSquirt CAN Message
MegaSquirt and MegaSquirt Logo are trademarks of BG Soflex, LLC.
MegaSquirt-II CAN Message

Description of the CAN message format used in MS-II
–
–
–
–
–

36
CAN Bus target data rate of 500 Kb/s.
Target Packet period – 5ms maximum.
One Master Node – MS-II.
Multiple Peripheral Nodes.
Any node can transmit/receive to any other node.
MS-II uses Extended Frame Format (29 bits)
MegaSquirt-II CAN Message

MS-II CAN Message Format:
Variable Offset
ID28
ID27
ID26
ID25
ID24
ID23
ID22
Message Type
ID21
ID20
ID19
ID18
ID17
ID16
ID15
MsgSenderID
ID14
ID13 ID12
ID11
MsgDestID
ID10
ID9
ID8
VariableBlock
ID7
ID6
ID5
ID4
Spare
ID3
ID2
The CAN Message is an Extended Frame (29 bit) format containing:
– Variable Offset
– Message Type
– Message Sender ID
– Message Destination ID
– Variable Block ID
37
ID1
ID0
MegaSquirt-II CAN Variable Blocks

Variable Block ID:
–
–
–

Variable Offset:
–
–
–
38
Used to identify structures or tables in a device from or to
which it is desired to get/ send data
Occupies “Identifier 3” block in Extended Frame CAN Message
Block ID is 4 bits, allowing up to 16 blocks per device
Used to define offset into MS-II block (which is defined in
Variable_Block_ID)
Occupies “Identifier 0,1” blocks in Extended Frame CAN
Message
Offset is 11 bits, range up to 2048 bytes
MegaSquirt-II CAN Message Type

Message Type:
–

In “Identifier 1” position in Extended Frame
Number definitions:
0 = CMD: Tells destination to place Data Length
Register (DLR) data bytes in this message into the
variable block ID specified in message header,
starting at variable byte offset. Used to push data
from remote device into MS-II. This is a “push”
operation from a peripheral to the MS-II. There is no
acknowledge message returned.
39
MegaSquirt-II CAN Message Type

Message Type – Number definitions (Con’t)
1 = REQ: This message requests the destination
device to send back data from the Variable Block ID
in the message header, starting at the variable byte
offset. The DLR register is always set to 3: the first
payload data byte is the variable block to store the
return message (see below) ; the second and third
are the offset into this block and no. of bytes – this
data is reflected in the header of the returned data.
Sending this message from a peripheral to MS-II will
trigger a response message (type 2 = RSP).
40
MegaSquirt-II CAN Message Type

Message Type – Number definitions (Con’t)
2 = RSP: This is the response record to a REQ record
request.
3 = XSUB: This is for future use, it will allow the trigger
a subroutine/function on a destination device.
4 = BURN: This triggers the receiving device to issue
a flash program operation.
41
MegaSquirt-II CAN Sender/Dest ID

Sender and Destination ID:
–
–
–
42
All devices on the MS CAN network must have a
unique CAN ID number. This number defines the
Sender and Destination ID.
By default, the engine controller board (MS-II) is
defined as ID zero.
Each ID is defined ahead of time before joining the
CAN network.
Other MS-II CAN Parameters

Other CAN-specific bits that are included in message
header:
–
–
–

43
SRR = 1
IDE = 1
RTR = set to 1 if no data is being sent (same as DLR empty)
DLR is Data Length Register, defines how many bytes
are transmitted in CAN frame (up to 8)
Other MS-II CAN Parameters






44
CAN Data rate is 500 KBPS
16 Time Quanta (including Sync Seg)
Tseg1 = 13
Tseg2 = 2
Sync Jump Width (SJW) = 4
Note – often these numbers are entered in registers as
one less (I.e. 0 to n-1).
CAN Message Header Format
IDR0
bit 7
bit 6
ID28
ID27
bit 5
ID26
bit4
bit 3
bit 2
bit 1
bit 0
ID25
ID24
ID23
ID22
ID21
High 8 bits of Variable Offset
IDR1
ID20
ID19
ID18
RTR / SRR
Low 3 Variable Offset
IDR2
ID14
ID13
ID12
IDE
1
1
ID11
ID10
DeviceID (Sender)
IDR3
ID6
ID5
ID4
Variable Block ID
45
ID17
ID16
ID15
Message Type
ID9
ID8
ID7
DeviceID (Receiver)
ID3
ID2
ID1
Spare
ID0
RTR
1
MS-II Acceptance Filtering


46
A CAN network with multiple nodes will most likely
carry a lot of messages.
For any given message, each CAN node has to
determine if this message is intended for them:
–
They (node) can decode every CAN message and see if the
message was addresses to them – if so then proceed further,
else discard. –Or-
–
Use front-end acceptance filters which only allow messages
matching a user-defined bit pattern to pass to the MSCAN
module.
MS-II Acceptance Filtering

The acceptance filters compare each bit in each field,
and can either allow any bit value (0 or 1) to match, or
can be set to force a specific bit value match.

There are two registers for each bit match:

47

Acceptance filter register (AR name suffix) holds the bit
pattern to check against the message.

Acceptance mask register (MR name suffix) indicates whether
the corresponding bit position is actually checked (set to 0)
or ignored (set to 1).
The combination of the AR and MR registers indicate
the actual acceptance filter.
MS-II Acceptance Filtering
 Acceptance
 Mask
Register
 Resulting
Filter
0 0 0 1 1 1 1 0 (0x1E)
1 1 0 0 0 1 0 1 (0x3A)
X X 0 1 1 X 1 X
 Identifier
0 1 0 1 1 1 1 1
 Identifier
1 0 0 1 1 0 1 0
 Identifier
1 0 0 0 1 0 1 0
1
(accepted)
2
(accepted)
48
Register
3
(rejected)
MS-II Acceptance Filtering

49
Any peripheral on the MS-II CAN bus should
set the acceptance filter mask to filter on the
destination ID matching the peripheral mask.
MS-II Acceptance Filtering
Acceptance Filter and Mask for Destination ID =3
IDAR0
IDMR0
bit 7
bit 6
ID28
ID27
0
1
ID20
IDAR1
IDMR1
0
1
ID14
IDAR2
IDMR2
IDAR3
IDMR3
50
0
1
ID19
0
1
ID13
bit 5
ID26
0
1
ID18
bit4
bit 3
bit 2
bit 1
bit 0
ID25
ID24
ID23
ID22
ID21
0
1
0
1
0
1
0
1
0
1
ID17
ID16
ID15
0
1
0
1
0
1
ID8
ID7
1
0
1
0
ID0
RTR
0
1
0
1
RTR / SRR
0
1
1
1
ID12
ID11
0
1
0
1
0
1
0
1
ID6
ID5
ID4
ID3
0
1
0
1
0
1
0
1
IDE
1
1
ID10
0
0
ID2
0
1
ID9
0
0
ID1
0
1
MS-II CAN Example

Here is an example of a MS-II CAN peripheral:
–
–
–
51
Peripheral wants to obtain RPM from MS-II.
Peripheral ID is 1.
Peripheral wants to send a single packet (REQ = 1
Message Type) and receive RPM in return packet
(type = 2, RSP).
MS-II CAN Example
52
MS-II CAN Example
53
MS-II CAN Example
54