Design method and implementation of CAN bus remote transmission reliability

The CAN transceiver SN65HVD251 connects a resistor in parallel between the CANH and CANL output pins as the terminal resistor of the CAN bus. When this node is used as the terminal node of the CAN bus, close the jumper JP1 to make the terminal resistor work. The terminal resistor value R6 is equal to the characteristic impedance of the transmission cable. The general value of 120Ω is discussed in detail in the literature, which solves the influence of the mismatch of the near and far end impedance. The Rs pin of SN65HVD25l is the slope resistor input pin, which can change the working mode of the transceiver. Resistors R2 and R3 are connected in series on CANH and CANL to limit the current, and then a group of pull-up and pull-down resistors R4 and R5 are used to effectively suppress the interference of reflected waves. When the bus is kept in a high-impedance state, the receiving end always receives the "1" level, which increases the amplitude of the signal and reduces the bit error rate. In addition, a pair of transient diodes Dl and D2 in opposite directions are connected in parallel between CANH and NCANL to prevent lightning strikes and transient interference on other buses.

3 Overall system design

Based on the above devices, a reliable CAN bus remote control system network platform is built. This system consists of a host CAN node connected to the host PC through a USB interface. The main node is connected to the following functional nodes using a bus, as shown in Figure 3. The host CAN node is mainly used to send remote control broadcast commands, collect data from all nodes, and upload it to the host computer software for identification, classification and statistics. It realizes bus listening, network monitoring and host computer interface functions. The bottom node controls the bottom device in the system, sends 8-byte data CAN bus messages containing node information, listens to the network broadcast instructions of the host node, and adjusts the node function.

Figure 3 CAN bus control system multi-machine test platform

4 Experimental analysis

4.1 Analysis of communication results at different kilometres

Connect the system bus to a simulated 1km to 5km remote network. In order to better analyze the reliability of the CAN bus and enable the oscilloscope to better observe the message waveform, connect the two ends of the oscilloscope CH1 to the 0km distance from the master node, and connect the two ends of CH2 to the 5km distance from the master node, as shown in Figure 4. In this way, the communication message waveforms of the near end (CH1) and the far end (CH2) of the 5km communication relative to the host CAN node can be observed.

The waveform tested by CH1 is at the top, and the waveform tested by CH2 is at the bottom. The waveform marked as 1 at the CH1 end is the message sent by the master node, 2 is the message received by the bottom node at the CH2 end, 4 is the data message sent by the bottom node, and 3 is the data received by the host CAN node. 1 and 2, 3 and 4 are called a group of messages. The last bit of each frame of data is the response bit. There is a time gap between every two frames of messages, one of which is the time for the host CAN node and the upper PC to process data, and the other is the time for the bottom measurement node to process data.

After observation, the amplitude of message 1 sent by the near end is attenuated after traveling 5 kilometers to message 2 received by the far end; similarly, message 3 received by the near end is also attenuated based on the amplitude of message 4 received by the far end. By testing the waveforms of communication from 1 km to 4 km respectively, it can be found that the longer the communication distance, the greater the amplitude attenuation.

When other conditions remain unchanged, experiments were conducted for 1 km to 5 km respectively. It was found that the change of long-distance communication distance would have an impact on the message transmission rate, but it was very small. The obtained data is tabulated as shown in Table 1

As shown in Table 1, the transmission rate is the highest at 1 km, with 13.2972 frames transmitted per second, that is, one frame of data is transmitted in 0.0752 seconds. The so-called one frame is actually sent once and received once, which is actually 2 frames for the CAN bus. As the transmission distance increases, the transmission rate tends to decrease slightly, indicating that there is a certain network delay in long-distance transmission, but it has little effect at low baud rates.

4.2 Communication results at different test points with the same mileage

Next, taking the communication distance of 5 kilometers as an example, we observe that the two ends of CH1 are connected to the test point at 0 kilometers, and the two ends of CH2 are connected to the test points at 1 kilometers, 2 kilometers, 3 kilometers, 4 kilometers, and 5 kilometers. It can be seen that the amplitude of the message waveform has changed accordingly. After 1 kilometer of attenuation, the amplitude of the same group of messages is reduced by about 0.2V; the communication at a distance of 2 kilometers will cause the amplitude of the same group of messages to change by about 0.4V; similarly, the amplitude of the same group of messages transmitted at 3 kilometers, 4 kilometers, and 5 kilometers will be attenuated by 0.6V, 0.8V, and 1V respectively. Therefore, it can be concluded that the amplitude of the message signal will attenuate by 0.2V every time the same group of messages communicates at a distance of 1 kilometer.

4.3 Influence of the operating voltage of CAN transceiver SN65FIVD251

During the experiment, it was observed that the size of the SN65HVD251 working voltage VCC has a great influence on the transmission distance. After a large number of experiments, the critical voltage value of VCC for successful communication at a distance of 1-5 kilometers (accurate to 0.1V) was obtained. The so-called critical voltage value is the minimum value for normal data transmission within a certain distance. As shown in Table 2.

From the table, it can be concluded that the premise of ensuring successful communication over a distance of 1 km is that the voltage at the VCC terminal is greater than or equal to 3.6V. The higher the voltage at the VCC terminal, the longer the communication distance can be. In the 1-5 km experiment, the voltage at the VCC terminal increases by about 0.3V for every additional kilometer. The maximum VCC cannot be higher than the maximum operating voltage of SN65HVD251, which is 7V.

The long-distance communication distance has a relatively large impact on the amplitude of the message signal, which is attenuated by about 0.2V per kilometer. At the same time, the input voltage of the CAN transceiver SN65HVD251 has a certain impact on the long-distance communication distance. Ensuring high voltage input within the normal voltage range can increase the long-distance communication distance of the system. Every 0.3V increase in power supply voltage can extend 1 kilometer, and an increase of 1 kilometer will cause a loss of 0.2V, and the remaining 0.1V is consumed by the driver chip.

5 Performance summary of CAN bus remote control network

After the working voltage of the driver chip and the transmission baud rate are determined, the transmission distance of the CAN bus is mainly determined by the following two factors: (1) the level difference between the implicit voltage of the response bit of the transmitter and the level difference when the receiver converts the implicit voltage into the explicit level and then transmits it to the transmitter; (2) the phase change of the bit when the response bit sent by the transmitter is confirmed by the receiver and then sent back to the transmitter. The level difference of the former is 0.6V, and the latter cannot lag behind half of the time of each bit. The 0.6V level difference is much larger than the 100mv difference between RS485 and RS422 to identify "1" and "0". This means that under the same transmission conditions, RS485 has a longer transmission distance than the CAN bus. Similarly, RS485 and RS422 are susceptible to interference due to their small thresholds. In addition, other performances of the CAN bus are superior to RS485 and RS422, such as CRC hardening, multi-master communication mechanism, and multiple layers of hardened upper layer protocols. The bit error rate of RS485 is 10-7, and the bit error rate of the CAN bus can reach 2×10-11. Therefore, the following methods can be used to improve the reliability of remote transmission:

(1) Increase the operating voltage of the driver chip.
(2) Reduce the baud rate of transmission to reduce the impact of phase lag.
(3) Use thicker twisted pair cables to reduce the resistance of the communication wires, thereby reducing transmission losses.
(4) Use two driver chips in parallel to reduce the internal resistance of the driver chip and increase the driving current, that is, reduce the internal loss of 0.1V.
(5) Use twisted pair cables with smaller distributed capacitance to reduce the impact of distributed capacitance on the synchronization phase.

In general, the CAN bus control system designed in this paper has achieved good results in terms of reliability and other performance indicators, and it is responsible for data acquisition and communication in the slope monitoring system of Laxiwa Hydropower Station.