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.

SN65LBC184: Deactivated output driver drives anyway?

Other Parts Discussed in Thread: SN65LBC184, ISO1050

Hi all,

I use the SN65LBC184 as a RS485 transceiver in a multi-point RS485 bus configuration.

According to the data sheet, table "driver function table", the output driver should be in "Z" state if the pin DE is at LOW level.

Now I measure that the output driver drive perfect levels (1V / 3.5V) even if the pin DE is stable at GND level, measured with open i/o pins (no load).
This changes the voltage level at the bus line making the communication instable.

Where might be my fault, what might be wrong?

/cu
Jens

  • Hi Jens,

    You are right that if the DE pin voltage is below the low-level threshold (0.8 V maximum) that the bus pins should be in a high-impedance state. When you are observing the normal drive levels, are they toggling or at a static level? If they are toggling, do they track the signal on the "D" pin? Are any other transceivers sharing the bus during this measurement? What sort of components (if any) are placed on the bus lines (e.g., for termination, filtering, etc.)?

    Max
  • Hi Max,

    Thanks for the quick answer.

    The output pins of the LBC184 will stuck at fixed levels, regardless what the "D" pin is applied to, Hi or Low, as expected.
    When I do the measurements the two output pins are completely open, "hanging" in the free air.
    Might there a few kOhm load required? If yes, what would you recommend at least?

    I have to couple the RS485 output pins with a CAN transveiver ISO1050 (TI too), expecting that CAN will operate properly when setting the RS485 driver into idle state. The other bus member may be both types, LBC128 and/or ISO1050.

    The issue is, that the Z-state LBC128 shifts the CAN recesive idle levels of 2.5V into a dominant level of 1.9/2.8V, setting the CAN bus into "dead" state.
    Following the datasheet of the ISO1050, the output driver is capable to sink/source 70mAmps, which should be powerfull enough to handle a "Z"-state output.

    Many thanks for any further idea.

    Jens

  • Hi Jens,

    This could be a byproduct of the LBC184's failsafe biasing scheme, which is intended to ensure that the receiver input is biased to a valid differential level in the event of an open-bus failure. If that is the case, adding a load should help to reduce the differential offset that you are seeing. Having a load is beneficial for the CAN transceiver as well, since the dominant-to-recessive edge is typically undriven (due to the expected 60-Ohm differential loading).

    I don't have LBC184 units readily available for testing since it is an older device, but I have ordered some samples from our distribution center. They should arrive this week, and then I will be able to verify whether or not this behavior is expected.

    Regards,
    Max
  • Hi Max,

    Could you already do some testing?

    Regards,
    Jens
  • Hi Jens,

    Sorry - I received the units earlier this week but didn't make it into the lab until today.

    I confirmed the idea that the failsafe biasing establishes a large differential voltage across an unterminated (high impedance) bus. With no load at all and the LBC184 driver disabled, I measured a voltage at "A" of about 3.5 V and a voltage at "B" of about 1 V. When a 60-Ohm differential load was introduced, the "A" and "B" voltages were both about 2 V.

    It seems the best solution to this issue would be to add a load, then. Would this be possible in your application?

    Best regards,
    Max
  • Hi Max,

    Many thanks for the investigations.
    As suggested, I will add a load, to solve it.

    Regards,
    Jens