The CAN-bus family has three main members: high-speed CAN, fault-tolerant CAN , and single-wire CAN. Fault-tolerant CAN, also known as low-speed CAN, differs from the most commonly used high-speed CAN in several ways. This article will share our understanding of fault-tolerant CAN.
I. The Origin of Fault-Tolerant CAN
In 1986, Bosch proposed the CAN bus concept at the SAE (Society of Automotive Engineers) conference, marking the initial development of the CAN bus in the automotive electronics industry. Subsequently, in 1987, Intel launched the first CAN controller chip, the 82526, igniting the spark for CAN bus development. Six years later, the international CAN standards ISO 11898/ISO 11519 were released, and the CAN bus began its widespread adoption in the communications field.
Figure 1. CAN bus application industries
ISO 11898 is the standard for high-speed CAN, while ISO 11519 is the standard for low-speed CAN. Initially, both the data link layer and physical layer of high-speed CAN were specified in standard ISO 11898. Later, it was split into ISO 11898-1 (only covering the data link layer) and ISO 11898-2 (only covering the physical layer). Standard ISO 11519-2-1994 was replaced by ISO 11898-3-2006 in 2006. This means that products conforming to standard ISO 11898-3 also support products conforming to standard ISO 11519-2.
Figure 2. Development History of the CAN Standard
II. Differences and similarities between fault-tolerant CAN and high-speed CAN
Like high-speed CAN, fault-tolerant CAN also uses differential twisted-pair cable for transmission, consisting of three wires: CAN_H, CAN_L, and GND. In demanding industrial applications, dedicated shielded twisted-pair cable and necessary protection circuitry are also required. As shown in Figure 3, taking the OSI 7-layer communication model as an example, the CAN bus standard actually specifies some physical layer, transport layer, and all data link layer rules, while the application layer, presentation layer, session layer, and network layer are not specified. High-speed CAN and fault-tolerant CAN have the same data link layer content; therefore, they are indistinguishable in bit transmission timing, bit arbitration, error handling, checksums, and frame structure.
Figure 3. Standardization of Fault-Tolerant CAN and High-Speed CAN
The difference lies in the physical layer definitions. Figure 4 compares the electrical signal data of ISO 11898 and ISO 11519-2. As shown in the figure, the maximum communication rate of high-speed CAN is 1 Mbps, while that of fault-tolerant CAN is 125 Kbps. Furthermore, high-speed CAN theoretically has a larger number of connected nodes than fault-tolerant CAN. The biggest difference between the two at the physical layer lies in the voltage levels of CAN_H and CAN_L when they are explicit or implicit. Therefore, fault-tolerant CAN devices cannot communicate directly with standard high-speed CAN devices; a CANBridge1054 adapter board must be added.
Figure 4 Comparison of electrical signal data between fault-tolerant CAN and high-speed CAN
III. Advantages and disadvantages of fault-tolerant CAN
Although fault-tolerant CAN has a lower communication rate and fewer carrying nodes, it has irreplaceable advantages. Based on the level signal data in Figure 4, we plotted the signal waveforms of both when they are working normally. As can be seen from Figure 5, the levels of CAN_H and CAN_L change by as much as 2.25V when there are explicit or implicit changes, while the level change of high-speed CAN is only 1V. This means that fault-tolerant CAN has higher anti-interference capability than high-speed CAN.
Figure 5 Comparison of fault-tolerant CAN and high-speed CAN signal waveforms
In addition, fault-tolerant CAN can ensure normal communication even when a short circuit or open circuit occurs in CAN_H or CAN_L. The fault-tolerant CAN transceiver automatically identifies the bus status and adjusts the receiver's receiving mode accordingly, which is the origin of the name "fault-tolerant CAN". Figure 6 shows the adjustment of the transmit and receive states of the fault-tolerant CAN transceiver under different conditions.
Figure 6. Fault-tolerant CAN multi-mode operating status
Note 1: 75μA pull-down current source function; Note 2: 75μA pull-up current source function
IV. Fault-tolerant CAN application circuit
Figure 7 shows the classic CTM1054T fault-tolerant CAN transceiver module as an example. It adopts a potting process and has extremely low electromagnetic radiation and high electromagnetic interference resistance. It fully complies with the ISO11898-3 standard, and the maximum number of nodes in a single network reaches 32.
Figure 7. Fault-tolerant CAN transceiver module CTM1054T
The fault-tolerant CAN node circuit design differs from the common high-speed CAN node design, and attention needs to be paid to the connection of the terminating resistor. In general, the module is connected to the power supply, and the port is connected to the CAN controller and the CAN network bus. The terminating resistors RTH and RTL are connected to CANH and CANL respectively, as shown in Figure 8.
Figure 8 Classic Fault-Tolerant CAN Node Circuit Design
A typical CAN-bus network is shown in Figure 9, with a maximum communication distance of 1km. If more nodes or longer communication distances are required, it can be expanded using devices such as CAN repeaters.
Fault-tolerant CAN transceivers exhibit optimal system performance when the total terminating resistance is 100Ω. When configuring the terminating resistance of a fault-tolerant CAN bus, first determine the number of nodes in the entire network. Each transceiver provides a portion of the total 100Ω terminating resistance. It is not required that each transceiver have the same terminating resistance, but the total terminating resistance should be 100Ω. For example, if there are 5 fault-tolerant CAN nodes on the bus, the 10 resistors connected to the network should all have a resistance value of 500Ω; if there are 10 fault-tolerant CAN nodes on the bus, the 20 resistors connected to the network should all have a resistance value of 1000Ω. Due to this unique configuration of fault-tolerant CAN terminating resistances, as long as the number of nodes is determined, star, tree, or other bus topologies can be used freely according to requirements.
Figure 9 Fault-tolerant CAN bus network topology
Fault-tolerant CAN is well-suited for low-speed, high-reliability industrial applications, and can adapt to various complex bus topologies when the number of nodes is fixed.
