• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 32 subscribers
  • Views 1108 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

3V3 CANbus & fault tolerance (faux single wire CAN)

Joshua Siegel
Prodigy 110 points
Other Parts Discussed in Thread: SN65HVD232, SN65HVDA541-Q1, SN65HVD256

I am creating a new design for a CAN data logger. Our old device is 5V and uses a Microchip MCP2551 (5V). Due to an architecture change, I would like to eliminate the 5V supply and use 3V3 instead.

I have worked with the TI SN65HVD232 in the past and would like to use this device again. However, I wish to keep a special function of the 2551: the ability to ground CANL and use the transceiver in single-wire mode. I tested this with the SN65HVD232 and received BUS OFF errors any time I took the module out of sleep.

Are there any (SOIC-8) 3V3 transceivers by TI that would allow this function? Alternatively, would anyone be able to suggest a "conversion" circuit that could be switched on / off with a jumper with minimal BOM impact? The maximum data rate in SW-mode is 33.3kbps, and 125kbps in dual-wire mode.

Thank you!

  • 0
    TI__Mastermind 20580 points

    Hello Joshua,

     

    Could you provide more information on the network topology you are trying to make work in this "fault" condition?  Another thing you can try is to use serial resistance instead of a hard short. If we understand more what the loading on the network looks like we may have some other options.

    -- Scott 

     

  • 0 in reply to Scott Monroe
    Prodigy 110 points

    Most commonly, I connect to a "GM-LAN" single-wire network. This is, I believe, SAE J2411 / GMW 3089. It's a single wire at 33.3kbit/s, up to 32 nodes but the number may vary. This network is not necessarily linear, and allows for selective node sleep. 

  • 0 in reply to Joshua Siegel
    TI__Mastermind 20580 points

    Based on trying to use a High Speed CAN transceiver (differential pair, centered on Vcc/2) in this sort of mis-use case as a single wire CAN transceiver, single ended referenced to GND it will be very difficult.  Even using a 5V CAN transceiver you are not guaranteed this will work in all cases they way I read the single wire specs.  For single wire the receiver input threshold is defined as 2.0 to 2.2V normal and high speed mode and 1.6V to 2.2V for "low battery".  High Speed CAN transceivers only guarantee a 1.5V minimum differential voltage which is essentially what you are using when you ground CANL.  Since you are now using this single ended vs differential the 3.3V devices have a lot less headroom to the 2.0V minimum on the single ended line vs the voltage into the transceiver.  If you can increase the Rload on the single ended view it might be enough to allow the voltage to build up enough that you would functionally see it as a high vs low on the single ended CAN.   Since these two CAN physical layers are not specified to be interchangeable there is now way to guarantee any operation, even with the 5V CAN case, but under some "nominal" conditions it seems reasonable it would work most of the time until the supply voltages, temperature or process shift. 

     

    If you want to stay with a modern 5V transceiver but need 3.3V I/O then the SN65HVD256 or the SN65HVDA541-Q1 should be interesting options that sort of bridge these two worlds.

    -- Scott