This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

SN65HVD257: Redundancy in CAN Architecture v2

I recently went through the material on redundant CAN topology provided by your team, accessible through the following link: [Link: https://www.ti.com/lit/ug/sllu172/sllu172.pdf?ts=1713461095781&ref_url=https%253A%252F%252Fwww.google.com%252F]. However, I have a few questions and uncertainties regarding the content.

Firstly, I encountered an issue with this topology where one port of node loses connection with the network, causing the transceiver to become unterminated (60R). This situation leads to an unattenuated signal, and the transition between dominant and recessive signals becomes distorted. Consequently, the disconnected bus "destroy" the signal from the good port in AND port. Are there any alternatives or solutions to mitigate this problem effectively? See image bellow (RXP = BUS1, RXS = BUS2)

Secondly, concerning the XOR logic of the two RX ports, is there a way to determine which signal has failed in cases where this bit would remain high, excluding scenarios where the bus remains dominant? (Example: In the case of imagem above)

Lastly, I'm interested in exploring further material from your resources regarding this topic. Could you please direct me to additional materials, such as other topologies or strategies, that I can consult for a more comprehensive understanding?

Thank you for your attention to these inquiries. I look forward to your prompt response.

Best regards, José Augusto Arbugeri.




  • Hi José,

    Firstly, I encountered an issue with this topology where one port of node loses connection with the network, causing the transceiver to become unterminated (60R). This situation leads to an unattenuated signal, and the transition between dominant and recessive signals becomes distorted. Consequently, the disconnected bus "destroy" the signal from the good port in AND port. Are there any alternatives or solutions to mitigate this problem effectively? See image bellow (RXP = BUS1, RXS = BUS2)

    According to ISO11898, placing termination on a node should be avoided since the bus lines lose termination if the node is disconnected from the bus. The termination need to be placed on each end of the bus. 

    Secondly, concerning the XOR logic of the two RX ports, is there a way to determine which signal has failed in cases where this bit would remain high, excluding scenarios where the bus remains dominant? (Example: In the case of imagem above)

    Through transmitting and receiving together, transmitting and receiving individually (through using the silent mode to use only one branch at a time for transmission during diagnostic mode), you should be able to figure out which bus has the error, and where on the bus it is.

    Lastly, I'm interested in exploring further material from your resources regarding this topic. Could you please direct me to additional materials, such as other topologies or strategies, that I can consult for a more comprehensive understanding?

    Unfortunately, we don't have more materials about redundant CAN network. I included a PPT file and you can find some other CAN related materials on slide 25. Hope that could be helpful.

    Industrial CAN Portfolio Overview.pptx

    Regards,

    Sean

  • Dear Sean,

    Thank you very much for your prompt response and clarification on the questions I raised regarding the redundant CAN topology.

    Upon further reflection, I realize my previous communication may have been unclear regarding certain aspects of the redundant CAN topology.

    1) In the network setup, terminations of 120R are indeed positioned at the ends. However, I encountered an issue when an input of a node is disconnected, as illustrated in the figure below (Node 3m Bus 1). The signal becomes corrupted due to the fault of termination at that particular node port. I also understand that the silent function is designed to resolve this issue. Is this a recognized challenge inherent in this topology?

    2) Regarding the XOR logic, my understanding now is that we cannot pinpoint the faulty node directly. We must silence each node individually and examine the network. Is this correct? In the material you provided, on page 20, there's a mention of: "Together with s/w routine using silent mode, the open bus may be determined." Could you clarify what "s/w routine" refers to in this context?

    3) I appreciate the materials you shared, and I look forward to your insights on these queries.

    Thank you once again for your assistance.

    Best regards,

    José Arbugeri

  • José,

    1. Understood. Yes, the silent mode and the RXDs XORed are designed to detect an open connection. Or you can try partial termination, which means put a large Ohm termination (for example, 2.6K) on each node, and keep the total termination of the bus to be ~60 Ohm. In this case if a node is disconnected there is still some termination on the node. But you can do this only if you know exactly the number of nodes you are going to put on the bus.

    2. This is correct. And "s/w routine" means software routine.

    Regards,

    Sean

  • Thanks Sean.