CAN bus in the car: the principle of operation and network design

The modern car is a complex computing complex in which dozens of electronic units must exchange data at a great speed. Previously, each sensor needed a separate wire to connect to the controller, which turned the wiring harnesses into a confusing maze occupying half the cabin. With the introduction of digital technologies, engineers managed to radically reduce the number of wires, combining all devices into a single network, called the “Smart Line” CAN tire.

This acronym stands for Controller Area Network, which means a network of controllers. It allows the engine control unit to instantly transmit data on the speed of rotation of the crankshaft to the dashboard, and the ABS unit to report on the wheel lock to the stabilization system. Without this technology, it would be impossible to imagine the work of any modern car, whether it is a budget. Lada Granta luxe Mercedes-Benz.

Understanding how this system works and functions is essential for every car owner who plans to diagnose or install additional equipment. Errors in the network often lead to “phantom” malfunctions, when serviceable nodes stop responding to commands due to data transmission problems. In this article, we will discuss in detail the network architecture, the physical principles of signal transmission, and methods for finding breaks or short circuits.

History and necessity of digital network implementation

Until the mid-1980s, the electric circuitry of the car was relatively simple and linear. Each switch was connected to the energy consumer by a separate wire, and complex systems such as fuel injection or ignition control used analog signals. However, the increase in the number of electronic assistants, the number of wires in the harnesses began to grow exponentially, increasing the weight of the machine and the likelihood of failure.

Company engineers Bosch The first to realize the deadlock of the way of increasing wiring, and in 1983 began to develop a protocol that would allow data to be transmitted over two wires for multiple devices at once. The result of their work was a standard that is now known all over the world. The main difference of the new system was the ability of any node to initiate data transfer, and not just wait for a survey from the main computer.

The introduction of technology has allowed not only to reduce the cross section of tourniquets, but also to increase the reliability of the entire system. If in the old wiring, a break of one wire disables a particular device, then in the digital network there is still the possibility of transferring data over a backup channel or redistributing functions between the remaining serviceable units. This has become critical for security systems where every millisecond matters.

Today, it is impossible to find a car built after 2000 that does not have this interface. Even in simple models, it is used to connect the dashboard with the engine control unit. In premium segments, the number of nodes connected to the network can reach seventy or more, creating a complex distributed computing environment.

Physical Structure and Types of CAN Buses

Physically, a network is a twisted pair of wires through which electrical signals are transmitted. To ensure noise protection, the wires are twisted together, which allows to compensate for external electromagnetic tips, inevitable in the conditions of the engine and generator. Depending on the data rate and the priority of messages, modern cars use several parallel lines.

High-speed line, often referred to as Drive CAN Power CAN, or Power CAN, combines critical nodes that require instant response. This includes engine, transmission, ABS and ESP control systems. The data transfer speed can reach 1 Mbps, which provides almost instantaneous response of security systems in case of an emergency.

A low-speed line, or Comfort CANIt is responsible for less critical functions: climate control, windows, central lock and interior lighting. The speed is lower, usually around 125 kbps, which is enough to comfortably control the auxiliary systems without the delays noticeable to the driver.

There is also a diagnostic bus that is activated only when the scanner is connected to the OBD-II connector. It allows technicians to access all control units for diagnostics and programming. The division into subnets is made so that overload in one system (for example, a failure in the windows) does not work critical engine control nodes.

Principle of data transfer and logic of work

The basis of the work is a differential method of signal transmission. This means that the data is not encoded by the absolute value of the voltage on one wire, but by the potential difference between two wires, which are traditionally called CAN-High and CAN-Low. At rest, the same voltage is maintained on both wires, usually around 2.5 volts.

When the transmitter starts sending a logical unit, it dilutes potentials: on the High wire, the voltage rises to 3.5-4.5 V, and on the Low wire it drops to 1.5-0.5 V. The receiver reads the difference and interprets it as a digital signal. This scheme provides high immunity: if the external tip affects both wires at the same time, the potential difference between them will not change, and the data will not be distorted.

The most important aspect is the arbitrage of access to the environment. Since there is no master control computer on the network, any block can start transferring at any time. If two blocks start transmitting simultaneously, priority is given to a message with a lower identifier. The lower priority system automatically shuts down and waits for the line to be released, ensuring that critical data is delivered without delay.

⚠️ Attention: When connecting third-party equipment (alarms, autosound) it is strictly impossible to crash into the wires of the CAN bus with the usual twisting. This disrupts the wave resistance of the line (should be strictly 120 ohms) and can lead to the “fall” of the entire network or the failure of expensive control units.

Each message on the network has a unique identifier, but not the address of the recipient. All blocks “hear” all messages, but only respond to those with the corresponding identifier. This makes it easy to add new devices to the network without reworking the entire architecture, as long as the new device “understands” the exchange protocol.

Diagnosis of faults and search for breaks

Detecting faults in digital communication lines often causes difficulties for beginners, as standard methods of “vertebrae” may not be informative enough. The most common problems are wire breaks, short circuits on the mass or between themselves, as well as failure of the terminal resistors. For high-quality diagnosis, an oscilloscope is necessary, although in some cases, you can do with a multimeter.

The first step in case of suspected network problems is a visual inspection of the connectors and measuring the voltage at rest. On a serviceable line, the multimeter should show approximately 2.5-2.6 volts relative to the mass on each of the two wires. If you see 0 volts or 12 volts, it indicates a short circuit or power break of the unit controlling the line.

Particular attention should be paid to termination. At the ends of the high-speed tire must necessarily be matching resistors with a nominal value of 120 ohms. When the ignition is turned off and the battery is disconnected, the resistance between the CAN-High and CAN-Low wires should be 60 Ohms (two parallel resistors of 120 Ohms). If the resistance is 120 ohms, one resistor is missing or burned, if infinity is a line break, or both resistors are faulty.

When using an oscilloscope, you should see clear rectangular pulses. If the signal shape is distorted, has gentle fronts or strong emissions, this indicates problems with shielding, the presence of severe interference or a malfunction of one of the transceivers in the connected blocks. Often, a “sick” block can “clog” the entire network, preventing the rest from working.

Table of parameters of standard communication lines

For ease of comparison of characteristics of different subnets used in modern cars, below is a summary table. It will help to understand why you can not connect diagnostic equipment to the comfort lines for controlling the engine.

The table shows that the single-wire version (LS-CAN or SW-CAN) is used less frequently and mainly for simple devices. In the event of a wire break in a dual-wire system, the low-speed network often goes into emergency mode using only one surviving wire, albeit with less noise immunity.

Impact of interference and protection methods

A car is an environment with extreme levels of electromagnetic interference. The generator, ignition coils, electric motors of windshield wipers and fans create powerful tips. That is why twisted pair is the standard: magnetic fields induced in one turn are compensated by fields in the adjacent turn, which has the opposite orientation.

However, if the harness wiring is damaged, the wires are woven or improperly laid near the high-voltage lines, interference can penetrate the network. This leads to errors in the protocol, which are fixed by control blocks. In response, the system may go into emergency mode, ignoring some commands or limiting functionality.