• State Resolved
  • Replies 2 replies
  • Subscribers 34 subscribers
  • Views 96 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TCAN4550: GPO1 & RST

AP
Intellectual 490 points

Part Number: TCAN4550

Hello,

plan is to use the TCAN4550 with external CAN transceiver (i.e. ISO1044B), having the RXD output of the CAN transceiver connected to the TCAN4550 GPO1 pin (which becomes the RXD_INT_CAN input in test mode with test_mode_config = 1).

However, after power-up (before any TCAN4550 programming via SPI) the GPO1 pin defaults out as the M_CAN_INT 1 (active low) interrupt. So this would mean that the GPO1 output and the RXD output of the CAN transceiver may drive against each other.

Questions:

  1. By default, is GPO1 a push-pull output or an open drain output?
  2. Is the GPO1 output disabled while RST = High level?
  3. Any suggestions?

Thanks

 

  • 0
    TI__Mastermind 36191 points

    Hello AP,

    AP said:
    By default, is GPO1 a push-pull output or an open drain output?

    GPIO1 is by default push-pull unlike GPO2 which is by default open-drain.

    AP said:
    Is the GPO1 output disabled while RST = High level?

    No, I measured the GPIO1 pin voltage on a TCAN4550EVM after power up and it is at the VIO level (3.3V in my test).  This is expected based on a non-interrupt condition for this pin.  I then continued to measure it while pressing the RESET pushbutton to make RST=High, and the GPIO1 pin remained at VIO (3.3V).

    The VIO supply pin is used as the voltage reference to the digital IO signals.  During a Reset Event, the digital core is reset so it will not provide a signal to activate the GPIO1 interrupt signal and the digital IO will remain at a constant level.

    AP said:
    Any suggestions?

    I'll first note that using an external transceiver with a TCAN4550 in test mode is somewhat common and this is the first time this question or concern has been asked.  The startup voltage contention between the GPIO1 pin and the RXD pin of the external transceiver has not resulted in any previous problems that have been brought to our attention.  While a momentary contention is possible, I don't know how serious of an issue this is.

    To brainstorm some options:

    - A schottkey diode could be placed between the RXD pin and GPIO1 pin to block any current flowing out of the GPIO1 pin.  This would result in a voltage level drop on the RXD signal, but the VIH level requirement is 0.7*VIO (or 2.31V if VIO=3.3V which is almost 1V which is greater than a diode's drop.  

    - A voltage sequencing technique could be implemented to delay the transceiver power up until after the TCAN4550 has been configured into test mode.

    - A buffer with an enable pin could be placed in-line that could allow the MCU to isolate the RXD and GPIO1 pins.

    Regards,

    Jonathan

  • 0 in reply to Jonathan Nerger
    Intellectual 490 points

    ... OK, thanks.

    I think we will try with a say 1K or so resistor right at the GPIO1 pin which hopefully becomes kind of transparent when operating as an input (in test mode).