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.

SN65HVD234: SN65HVD234

Part Number: SN65HVD234

Is there a minimum termination resistance value (across CANH and CANL) to guarantee that the R output will correctly duplicate the data sent on the D input?  We have a problem with our power control software, which depends on the data received on the R pin being the same as the data transmitted on the D pin.  The R pin data will be corrupted if the CAN bus is not terminated, but we do not want to include a 120 ohm resistor inside our equipment.  Experimentally, about 2K ohms seems to do the trick, but I wonder if there is a minimum value I can rely on.

  • Hi Michael,

    Termination is used to allow the bus to return to a recessive state after being driven dominant as well as to dampen any reflections on the transmission line. The value for this is typically 120-Ohms at both ends of the cable. This resistor doesn't have to be included on the node itself if the cable is terminated. Could this be a solution?

    The value of the resistor is based off the characteristic impedance of the cable (120-Ohm typical) to reduce reflections. Using larger values will lengthen the fall time of dominant to recessive transitions (limiting data rate) and potentially allow reflections to impact the data.

    More on CAN termination can be found on section 11.2.1 on the datasheet.

    Let me know if this helps.

    Regards,
    Eric
  • Hi Eric,

    This problem actually originated in final test. At the end of this test, the operating unit is disconnected from its harness, which removes power from the unit. This signals the unit's microprocessor to shut down. Unfortunately, this process also disconnects the 120 ohm load on the CAN, and corrupts the CAN data returning to the micro. In this case, it does not shut down. Experimentally, we put an R-box across CANH and CANL, and found that about 2K was all you needed to make receive data match the transmitted data. The unit shuts down properly with this load resistance. It was suggested that, if the value was large enough, we could include this resistor in the product, since it would have minimal effect on the external load resistance. Is there a guaranteed value for minimum termination resistance?

    Thanks.
  • Hi Michael,

    2k-Ohm should work fine to allow the device to operate when isolated from the bus without significantly impacting connected performance. The maximum value of this resistor depends on the data rate the device operates at as well as the capacitance on the line (negligible when isolated), so I can't guarantee a specific value.

    When you reconnect the device to the bus, is the bus terminated as well? If this is the case, the bus termination should dampen signal reflections.
    What is the range of resistors your able to include on the device?

    Regards,
    Eric
  • Hi Eric,

    Thanks for the response. We would like to include a resistor that is small enough to terminate the bus during final test, but not small enough to over-terminate the bus in a normal vehicle installation which could have 1 or 2 terminating resistors. I suspect that 2K will do the job, but I have only tried a couple of units. We have experimentally found that the CAN xcvr will drive resistances down to about 11 ohms, so 2K wouldn't cause an issue in parallel with even a 60 ohm load. I was just hoping for a definitive minimum value.
  • Hi Michael,

    Unfortunately, there's not a concrete answer to this. Most CAN devices only specify their output amplitudes with the standard 60-Ohm loading from termination. The smaller resistance you add in parallel with this, the more loading the bus experiences and the harder it will be for the device to maintain output amplitudes. The only metric that I can really relate to this is the number of allowed nodes specified for this device, but this doesn't translate directly to parallel resistance. I estimate this gives a maximum allowable load of around 53.3-Ohms. Some newer devices are specified to at load conditions of 50-Ohms as a reference as well.

    This would be more a specific engineering choice regarding the acceptable amplitude margin in your system. 2k-Ohm in parallel with your termination does land within both of the above specifications, so if this works experimentally, it sounds like a good fit for the system.
    If you're still concerned about the value or would like to optimize the performance, I would recommend more experimentation with different values and testing with an oscilloscope connected to the bus to observe what amplitudes differences occur in each case.

    Apologies for not having a more definitive answer. I hope this helps you towards your final solution.

    Regards,
    Eric
  • Eric,

    No problems. I appreciate the responses.  Thanks!