Debugging CAN, LIN and FlexRay Automotive Buses with an Oscilloscope

Introduction

Embedded systems applications are growing rapidly in the automotive industry. Many vehicle designs use CAN, LIN or FlexRay to communicate among electronic control units (ECUs), and between ECUs and sensors, actuators and displays. These buses are critical in providing real-time communications in and among important subsystems from braking systems to infotainment systems.

CAN, LIN, and FlexRay are relatively mature bus protocols and are designed to be robust and easy to integrate. Even so, communications can be impacted by noise, board layout, and power-up/down timing. Problems can include excessive bus errors and lock-ups. Unlike basic protocol analyzers, oscilloscopes equipped with protocol decoding, can be used to see both the decoded bus traffic, as well as signal quality.

Perhaps more importantly, oscilloscopes can be used to troubleshoot problems at the system level. Automobiles rely on extensive networks of sensors, actuators and displays, and many problems involve bus timing relative to I/O events or values. Oscilloscopes are well-suited for looking at I/O signals and bus transactions at the same instant.

This ability to see bus signals and decoded traffic at the electrical level and systems level make oscilloscopes the best choice for troubleshooting.

CAN/CAN FD

Controller Area Network (CAN) was originally developed in the 1980s by the Robert Bosch GmbH as a low cost communications bus between devices in electrically noisy environments. Mercedes-Benz became the first automobile manufacturer in 1992 to employ CAN in their vehicle systems. Today, every automotive manufacturer uses CAN controllers and networks to control a variety of Electronic Control Units (ECUs) in their automobiles. It is the primary bus used for engine timing controls, anti-lock braking systems and powertrain controls to name a few. And due to its electrical noise tolerance, minimal wiring, excellent error detection capabilities and high data transfer speeds, CAN expanded into other applications such as industrial control, marine, medical, aerospace, and more.

As vehicle networks have evolved to support more functions, a need has arisen to support faster data communication between nodes. This has led to the following CAN standards:

1. CAN FD is a higher speed version of CAN that can achieve a maximum data rate of 8 Mbps with a payload up to 64 bytes long compared to the maximum data rate of 1 Mbps and payload of 8 bytes for CAN. The first version of the CAN FD standard was released in 2012 but later updated into an ISO standard (ISO CAN FD) in 2015. The ISO version introduced additional safeguards to improve communication reliability. The original version is now known as non-ISO CAN FD and is not compatible with ISO CAN FD.
2. CAN XL is a new standard developed by CAN in Automation(CiA). This third generation of CAN meets requirements of E/E zone architectures for future vehicles. Its main usage will be as a backbone network to bridge the gap of bitrates and data between classical CAN/CAN FD and Ethernet 100BASE-T1. CAN XL supports data rate up to 20 Mb/s and data field length up to 1 to 2048 bytes.

HOW IT WORKS

CAN / CAN FD / CAN XL defined the data link layer of the ISO 7-layer model. While there are various physical layers specifications (include fault tolerant and single wire), ISO 11898 is the most prolific — used in most applications. It is a (differential) 2-wire interface running over a shielded twisted pair (STP), unshielded twisted pair (UTP) or ribbon cable. The data rate is defined by the user, with 1 Mb/s being the fastest for CAN, 8 Mb/s the fastest for CAN FD and 20 Mb/s the fastest for CAN XL. The lowest rate is 20 kb/s. All the nodes on the bus must communicate at the same bus speed. (There are configurations in which all the nodes change their bus speed to a higher rate for reflashing or diagnostics but then return back to their normal rate.) Maximum cable length depends on the data rate used. The maximum line length can be thousands of meters at low speeds; 40 meters at 1 Mb/s is typical. Termination resistors are used at each end of the cable.

Non Return to Zero (NRZ) bit encoding is used with bit stuffing to ensure compact messages with a minimum number of transitions and high noise immunity. The CAN bus interface uses an asynchronous transmission scheme where any node may begin transmitting anytime the bus is free. Messages are broadcast to all nodes on the network. In cases where multiple nodes initiate messages at the same time, CAN / CAN FD uses bitwise arbitration and CAN XL uses CSMA/CR (carrier sense multiple access/ collision resolution) to determine which message is higher priority. Messages can be one of four types:

Structure of Data and Remote Transmission Request Frames

The intermission (INT) field of three recessive bits indicates the bus is free. Bus Idle time may be any arbitrary length including zero.

SYMBOLIC CAN BUS DECODING

Symbolic decoding, as defined in DBC database files, can be used to display CAN, CAN FD and CAN XL bus traffic in an easy-to-read format. By loading a DBC database file into the scope, you can view CAN traffic as text, rather than raw hex values. The file syntax is standardized, but the specific encoding and translation is proprietary and closely held by manufacturers.

In this example, a few symbolic message and signal values are shown in the bus waveform and results table displays.

LIN

The Local Interconnect Network (LIN) bus was developed by the LIN consortium in 1999 as a lower cost alternative to the CAN bus for applications where the cost, versatility, and speed of CAN were excessive. These applications typically include communications between intelligent sensors and actuators such as window controls, door locks, rain sensors, windshield wiper controls, and climate control, to name a few.

However, due to its electrical noise tolerance, error detection capabilities, and high speed data transfer, CAN is still used today for engine timing controls, anti-lock braking systems, powertrain controls and more.

HOW IT WORKS

The LIN bus is a low-cost, single-wire implementation based on the Enhanced ISO9141 standard. LIN networks have a single master and one or more slaves. All messages are initiated by the master with only one slave responding to each message, so collision detection and arbitration capabilities are not needed as they are in CAN. Communication is based on UART/SCI with data being sent in eightbit bytes along with a start bit, stop bit and no parity. Data rates range from 1 kb/s to 20 kb/s. While this may sound slow, it is suitable for the intended applications and minimizes EMI.

The LIN bus is always in one of two states: active or sleep. When it’s active, all nodes on the bus are awake and listening for relevant bus commands. Nodes on the bus can be put to sleep by either the Master issuing a Sleep Frame or the bus going inactive for longer than a predetermined amount of time. The bus is then awakened by any node requesting a wakeup or by the master node issuing a break field.

As shown in Figure 10, LIN frames consist of two main parts, the header and the response. The header is sent by the master while the response is sent by the slave. The header and response each have subcomponents:

FlexRay

As cars get smarter and electronics find their way into more and more automotive applications, manufacturers are finding that existing automotive serial standards such as CAN and LIN do not have the speed, reliability, or redundancy required to address X-by-wire applications such as brake-by-wire or steer- by-wire. Today, these functions are dominated by mechanical and hydraulic systems. In the future they will be replaced by a network of sensors and highly reliable electronics that will not only lower the cost of the automobile, but also significantly increase passenger safety due to intelligent electronic based features such as anticipatory braking, collision avoidance, adaptive cruise control, etc.

HOW IT WORKS

FlexRay is a differential bus running over either a Shielded Twisted Pair (STP) or an Un-shielded Twisted Pair (UTP) at speeds up to 10 Mb/s, significantly faster than LIN’s 20 kb/s or CAN’s 1 Mb/s rates. FlexRay uses a dual channel architecture which has two major benefits. First, the two channels can be configured to provide redundant communication in safety critical applications such as X-by-wire to ensure the message gets through. Second, the two channels can be configured to send unique information on each at 10 Mb/s, giving an overall bus transfer rate of 20 Mb/s in less safety-critical applications.

FlexRay uses a time triggered protocol that incorporates the advantages of prior synchronous and asynchronous protocols via communication cycles that include both static and dynamic frames. Static frames are time slots of predetermined length allocated for each device on the bus to communicate during each cycle. Each device on the bus is also given a chance to communicate during each cycle via a Dynamic frame which can vary in length (and time). The FlexRay frame is made up of three major segments; the header segment, the payload segment, and the trailer segment. Each segment is made up of one or more components:

Dynamic frames have one additional component that follows the Trailer CRC called the Dynamic Trailing Sequence (DTS) that prevents premature channel idle detection by the bus receivers.

But how many of the events occurred? To find all of the bus events that meet a specific search criteria, you can use automated Wave Inspector search. The setup is similar to the bus trigger setup, allowing the oscilloscope to find and mark all of the specified bus events.

In the example above the oscilloscope was set up to trigger on a specific Frame ID. It captured and decoded approximately 80 FlexRay frames. Then Wave Inspector search was set up to go through the acquisition and and marked 49 occurrences within a range of data values. Each of the data values in the specified range is shown with the pink bracket icon. Using the arrow keys, you can instantly navigate between marked values.

And all of this was done with only 250,000 point record lengths. The 5 Series MSO, for example, can capture 62.5 Mpoints with standard record length, or 125 M points when equipped with extended record length.

DECODING MULTIPLE BUSES AT ONCE

A powerful feature of the oscilloscope is the ability to define and decode many serial buses simultaneously. Going back to our earlier example with CAN bus; now imagine that the window controls are operated by a LIN bus. When the driver presses the Passenger Window Down control, a message is initiated on a LIN bus in the driver door, passed through a central CAN gateway and then sent on to another LIN network in the passenger door. In this case, we can trigger on the relevant message on one of the buses and capture and decode all three buses simultaneously, making it exceptionally easy to view traffic as it goes from one bus to another through the system. In this example, the scope has triggered on the first LIN message and captured all three buses.