In industrial automation, automotive electronics, and embedded systems, the CAN (Controller Area Network) bus is one of the most commonly used communication methods. From simple sensor networks to complex vehicle control systems, CAN carries critical data transmission tasks.
However, in real-world projects, one question repeatedly arises:
What is the CAN bus max speed? Can all systems run at 1 Mbps? Why do some systems only use 250 kbps, or even lower?
In fact, the “maximum speed” of CAN is not a fixed value that can be directly applied. It is influenced by multiple factors such as communication distance, network structure, number of nodes, and electrical environment. Different types of CAN (such as High-Speed CAN, Low-Speed CAN, Single-Wire CAN) as well as the later-developed CAN FD also show significant differences in speed capability.
Therefore, to truly understand CAN Bus max speed, it is necessary not only to know the theoretical limits, but also to combine baud rate, bit rate, physical layer characteristics, and practical engineering constraints for comprehensive analysis.
CAN Speed Fundamentals: Baud Rate, Bit Rate, and Distance Limits
Baud Rate vs Bit Rate
In many CAN textbooks and application notes, “baud rate” and “bit rate” are often used interchangeably. Strictly speaking:
- Baud Rate: the number of signal state changes per second, unit: Baud
- Bit Rate: the number of binary bits transmitted per second, unit: bps (bits per second)
In CAN bus systems, NRZ (Non-Return to Zero) encoding is used, meaning one signal level change corresponds exactly to one bit. Therefore, in most CAN applications, the baud rate equals the bit rate. For example, a 500 kbps CAN bus also has a baud rate of 500 kBaud.
Why is this important? Because when you measure CAN_H and CAN_L with an oscilloscope, what you observe is the signal transition frequency, which is the baud rate. This helps you quickly determine whether the bus is operating at the expected speed.
Therefore, in CAN systems, when we discuss “CAN Bus Max Speed,” we usually refer to the bit rate (i.e., bits transmitted per second), such as the 1 Mbps limit of classical CAN.
Relationship Between CAN Speed and Transmission Distance
The upper limit of CAN speed is constrained by physical signal propagation delay and bit timing requirements. Simply put:
When a node sends a dominant bit (logic 0), the farthest node must detect it within the same bit time; otherwise, the arbitration mechanism will fail.
As a result, the higher the CAN speed, the shorter the allowable bus length.
The table below shows the typical relationship between CAN distance and speed recommended by ISO 11898 (actual values may vary depending on cable quality, node count, and electromagnetic noise)
| Bit Rate | Max Bus Length (Typical) | Description |
| 1 Mbit/s | 20–40 m | High speed, short distance, commonly used in powertrain systems |
| 800 kbit/s | 40–50 m | Industrial high-speed scenarios |
| 500 kbit/s | 100–110 m | Common in automotive and industrial systems |
| 250 kbit/s | 250–500 m | Suitable for longer-distance industrial or agricultural equipment |
| 125 kbit/s | ~500 m | Long-distance communication with priority on reliability |
| 50 kbit/s | ~1000 m | Industrial fieldbus applications |
| 10 kbit/s | ~5000 m | Ultra-long distance, such as building automation or mining |
From the table, it can be seen that 1 Mbps is only suitable for short-distance scenarios, and in most industrial or automotive systems, the actual operating speed is often lower than the theoretical maximum.
Related Article: Can Bus Maximum Length and How to Extend CAN Bus Distance
CAN Speed Categories
Under the CAN 2.0 protocol framework, different physical layer standards result in very different speed ranges and characteristics. The following introduces High-Speed CAN, Low-Speed CAN, and Single-Wire CAN.
High-Speed CAN Bus
The most common CAN physical layer type is High-Speed CAN, defined by ISO 11898-2.
High-Speed CAN uses a twisted-pair structure, either shielded or unshielded, and communicates via differential voltage. Its communication speed typically ranges from 40 kbit/s to 1 Mbit/s, which is why CAN’s maximum speed is often referred to as 1 Mbit/s.
The bus includes two signal lines:
- CAN_H (CAN High)
- CAN_L (CAN Low)
In High-Speed CAN networks, stub length must be strictly controlled. For example, at around 500 kbit/s, the stub length should generally not exceed 30 cm. If the wiring distance is long, a reference ground (GND) should be introduced to improve signal stability.
High-Speed CAN requires 120 Ω termination resistors at both ends of the bus to match the characteristic impedance and reduce signal reflections.
Voltage characteristics:
- Recessive level: CAN_H ≈ 2.5 V, CAN_L ≈ 2.5 V
- Dominant level: CAN_H ≈ 3.5 V, CAN_L ≈ 1.5 V
Differential signal:
- Vdiff = CAN_H – CAN_L
In practical systems, gateways can be used to connect High-Speed CAN with Low-Speed CAN networks, enabling communication between different subsystems.
Low-Speed CAN Bus
Low-Speed CAN is defined by ISO 11898-3 and is also known as Fault-Tolerant CAN.
It also uses a dual-wire structure, but with lower communication speeds, typically ranging from 40 kbit/s to 125 kbit/s.
Unlike High-Speed CAN, Low-Speed CAN does not use termination resistors at the ends of the bus. Instead, each node has its own termination.
Voltage ranges:
- CAN_H: approximately 0.2 V to 3.6 V
- CAN_L: approximately 1.8 V to 4.8 V
The key difference is not only speed, but also voltage range and fault tolerance. Due to its wider voltage range and lower speed, Low-Speed CAN can continue operating even when certain faults occur, such as:
- Open circuit on CAN_H or CAN_L
- Short circuit to power
- Short circuit to ground
- Short circuit between CAN_H and CAN_L
This makes it especially suitable for systems requiring high reliability but lower real-time performance, such as body electronics.
Single-Wire CAN (SWC)
Single-Wire CAN (SWC), defined by SAE J2411, uses a single unshielded wire as the communication medium, with voltage referenced to ground in the range of approximately 0 V to 4.1 V.
This technology was originally developed by General Motors (GM), also known as GMLAN, as a replacement for SAE J1850. It is mainly used in automotive comfort systems, such as:
- Central door locking
- Power windows
- Body control modules
In SWC networks, each transceiver is directly connected to battery voltage and integrates termination internally.
Data rates:
- Normal communication mode: ~33.33 kbit/s
- High-speed diagnostic mode: ~83.33 kbit/s
The network typically supports up to 32 nodes and includes a Selective Sleep mode to reduce power consumption.
CAN FD: Upgrade in Speed and Bandwidth
CAN was originally developed by Bosch to address communication and wiring complexity among multiple ECUs in vehicles. As system data volume increased, the bandwidth limitations of classical CAN became evident, leading to the introduction of CAN FD based on ISO 11898-1:2015.
CAN FD (Flexible Data Rate) is an upgraded version of Classical CAN (CAN 2.0). Its key improvements include breaking the 1 Mbps speed limit and significantly increasing payload capacity. Traditional CAN supports up to 8 bytes per frame, while CAN FD extends this to 0–64 bytes and increases the data phase speed to 2–8 Mbps, greatly improving communication efficiency.
A key feature of CAN FD is the separation of communication into two phases:
- Arbitration phase: same speed as Classical CAN to ensure compatibility
- Data phase: higher bit rate to improve throughput
Overall, Classical CAN is suitable for low-bandwidth, high-node-count scenarios, while CAN FD is better suited for high-data-volume, high real-time applications, representing an important evolution for next-generation automotive and industrial communication.
CAN Bus Maximum Speed Summary
Overall, CAN system maximum speed can be summarized at different levels:
- Classical CAN: up to 1 Mbps (High-Speed CAN)
- Low-Speed CAN: up to ~125 kbps
- Single-Wire CAN: ~33.3 kbps
- CAN FD: up to 2–8 Mbps
However, in real applications, the achievable speed depends on bus length, number of nodes, wiring quality, and electromagnetic environment. Therefore, CAN Bus Max Speed is more of a theoretical upper limit rather than a default operating value in engineering.
FAQ
Q: Is 8 Mbps in CAN FD the effective data rate?
No. 8 Mbps refers to the raw bit rate in the data phase. When accounting for frame overhead such as SOF, ID, CRC, and ACK, the effective payload rate is about 5–6 Mbps. However, this is still much higher than the 1 Mbps of Classical CAN.
Q: Why is the maximum length of High-Speed CAN only about 40 meters?
To ensure proper arbitration at 1 Mbps, the signal must propagate to the farthest node and return within approximately one-quarter of the bit time. Signal propagation in copper is about 5 ns/m, so a 40-meter round trip is about 400 ns, which is already close to the limit within a 1 µs bit time.
Q: Is higher CAN speed always better?
Not necessarily. Higher speeds impose stricter requirements on wiring, termination, and electromagnetic conditions, and significantly reduce communication distance. In many real-world projects, 500 kbps or 250 kbps is more commonly used because it provides better stability.
Conclusion
CAN bus max speed may seem like a simple parameter, but it is actually a system-level result determined by physical layer characteristics, electrical properties, communication timing, and network structure.
From the 1 Mbps limit of Classical CAN, to different physical layer implementations, and to the performance improvements of CAN FD, the development of CAN technology has always focused on one core challenge: balancing speed, distance, and reliability.
In practical engineering, rather than pursuing the highest speed, it is more important to choose communication parameters based on application requirements. Truly stable and reliable systems usually operate not at the maximum speed, but at the most appropriate speed range.
