Abstract: Due to the limitations of CAN transceivers, a CAN bus network can only have a maximum of 110 CAN nodes or a communication distance of 10km. Therefore, when the required number of nodes and distance in the CAN bus network exceeds these limitations, the CAN bus network must be expanded. Currently, the common expansion method is to add CAN bus repeaters. While this method is convenient and simple, it increases the system cost. Here, we propose a CAN bus network expansion scheme based on the ADG663 analog switch, which can connect two or more CAN bus networks without the need for CAN repeaters or other devices. This paper introduces the specific expansion method and hardware and software design. Keywords: CAN bus; analog switch; bus expansion 1. Introduction CAN (Controller Area Network) bus belongs to the fieldbus category. It is a serial data communication protocol developed by Bosch in the early 1980s to solve the data exchange between numerous control and test instruments in modern automobiles. Since Bosch introduced the CAN bus, it has been favored for its practicality, reliability, and economy, and has made significant progress. The CAN bus is currently the only fieldbus with an international standard, enabling fully distributed multi-machine systems. It employs non-destructive bus arbitration technology to meet diverse real-time requirements, with a maximum communication distance of 10km (transmission rate of 5Kb/s) and a maximum communication rate of 1Mb/s (transmission distance of 40m). It can support up to 110 nodes, using twisted-pair or fiber optic cables as the transmission medium. Messages use a short frame structure with CRC checksum and other error detection measures, resulting in an extremely low data error rate and extremely high reliability. With its superior characteristics, low cost, high reliability, and flexible structure, the CAN bus is widely recognized as one of the most promising fieldbuses. Because of its many unparalleled features, the CAN bus is widely used in many applications. However, sometimes, due to specific project requirements, it is necessary to connect more nodes or extend the communication distance over a longer distance, necessitating the expansion of the CAN bus network. Currently, a common method for expanding the CAN bus is to connect two CAN bus networks using CAN repeaters, such as Zhou Ligong's CANrep-A/B type intelligent fully isolated CAN repeater, CAN-3202 intelligent CAN bus two-way repeater bridge, XYCANR2 dual-port CAN opto-isolated repeater, ADAM-4515 CAN repeater, and WT406-CAN CAN bus repeater module. These CAN repeaters all use microcontrollers to store and forward data between the two CAN networks separately, thus achieving connection and bidirectional data transmission. Expanding the CAN bus network using repeaters requires adding corresponding microcontrollers: a CAN controller and a CAN driver. This increases system cost and engineering expenses. Based on these issues, we designed a CAN bus network expansion scheme based on the ADG663 analog switch. Utilizing the time-division multiplexing characteristic of analog switches, we time-division multiplex the signal from the microcontroller via the CAN controller to two CAN drivers located on different CAN networks. This allows the microcontroller to be time-divisionally connected to two CAN networks, enabling data exchange between them and achieving repeater-free expansion of the CAN bus network. As shown in Figure 1, connecting microcontrollers at adjacent network locations to two CAN networks via analog switches in a time-division multiplexing manner not only expands the network but also fulfills the measurement and control tasks of the local node, thus eliminating the need for repeaters and reducing system costs. Based on the characteristics of CAN communication, to switch between the two CAN signals, we use the independent four-channel controllable analog switch ADG663. The switching of the microcontroller between the two CAN bus networks is achieved by programming its control pins. 2. Introduction to ADG663 The ADG663 is an integrated CMOS switching device manufactured by Analog Devices (ADI). It contains four independent selectable analog switch channels, which can be easily controlled by the control terminal. These channels have very low on-resistance and a wide signal input range, enabling precise analog signal switching. The entire device is fabricated using Analog Devices' advanced linearly compatible CMOS (LC2MOS) process, featuring low leakage current, ultra-low power consumption, and minimal charge accumulation during high-speed operation. In particular, two of the four channels of the ADG663 are high-level conductions, while the other two are low-level conductions. This not only simplifies the control signals for the analog switches but also makes the CAN bus transmit/receive switching more synchronized. The functional block diagram of the ADG663 is shown in Figure 2. IN1～IN4 are the control signal terminals, S1～S4 are the input signals, and D1～D4 are the corresponding output signals. 3. CAN Bus Network Expansion Based on ADG663 3.1 Hardware Design of the Expansion Circuit In the CAN bus network, the microcontroller used for detection is the new PIC18F458 chip from Microchip Technology, an 8-bit CMOS microcontroller with an internal Harvard bus architecture. This allows all instructions to be single-byte and single-cycled, improving the CPU's instruction execution speed and thus increasing the microcontroller's operating speed. For the CAN driver, we chose the MCP2551 CAN bus driver chip from Microchip Technology. It is fully compatible with the ISO 11898 standard, has a maximum speed of 1Mb/s, and offers better electromagnetic radiation and electromagnetic interference immunity than the 82C250. The principle of using the ADG663 to expand the CAN bus network is shown in Figure 3. Due to the characteristics of the ADG663, the microcontroller can switch between two CAN drivers using only one I/O line, while simultaneously ensuring the microprocessor's synchronous transmission and reception for each CAN driver and that the microprocessor is always connected to the bus. It also improves system stability during analog switch switching. 3.2 Software Programming of the Expansion Circuit Based on the plug-and-play interface characteristics of the CAN bus, this expansion method only requires changing the programming of the relay node, without modifying the other nodes. The program for the relay node is written to enable it to perform its own measurement and control tasks while simultaneously storing and forwarding data from both sides. To accomplish these two tasks simultaneously, the microcontroller's operating frequency must be much higher than the CAN bus bit rate, ensuring that the microprocessor has sufficient time to store and forward data from both networks. Since the microprocessor uses analog switches to switch between two CAN networks, theoretically, some data loss on the bus is inevitable. This necessitates that the measurement nodes appropriately retransmit their transmitted data to ensure complete data exchange. For relay nodes, a timer is added to the original measurement node program to determine the switching frequency of the analog switches. Simultaneously, when the microcontroller connects to a network, it not only sends out its own detection information, control signals, and all information stored when connected to the other network, but also interrupts the reception of information from other nodes on that network and stores it for transmission when the analog switches switch. The number of node information retransmissions is determined by a combination of the microprocessor's operating frequency and the CAN bus's transmission rate. Choosing an appropriate number ensures that there is not too much redundant information on the bus, and that the microprocessor does not lose too much information during the analog switch switching process. 4. Conclusion The CAN bus network expansion scheme based on the ADG663 analog multiplier discussed in this paper eliminates the need for a CAN repeater, simplifies system hardware connections, reduces system costs, and provides a simple and effective implementation method for CAN bus network expansion. However, simplifying the hardware inevitably increases the complexity of the system software and imposes certain limitations on the system communication rate. Therefore, this scheme is suitable for situations where the data transmission volume is not large and the rate requirement is not very high. References 1 Li Gang, Lin Ling. Modern Measurement and Control Circuits [M]. Beijing: Higher Education Press. 2004 2 Zou Jijun, Rao Yuntao. CAN Repeater Design and Application, Electronic Technology Application, 2003. 29 (8): 39～41 3 Lin Ling, Hong Quan, Li Gang. Application of ADUM1201 in CAN Bus Communication System. Electronic Products World. 2005. 4: 98～100 4 Rao Yuntao. Fieldbus CAN Principle and Application Technology [M]. Beijing: Beijing University of Aeronautics and Astronautics Press. 2003 5 Li Xuehai. Practical Tutorial of PIC Microcontroller [M]. Beijing: Beijing University of Aeronautics and Astronautics Press. 2004 6 Liu Heping. PIC18XXX Microcontroller Principle and Interface Programming [M]. Beijing: Beijing University of Aeronautics and Astronautics Press. 2004