• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 33 subscribers
  • Views 384 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

SN65HVD82: Receive line driven low when chip unpowered

Hayden Robinson
Prodigy 10 points
Part Number: SN65HVD82


Hi,

I have a design in which I parallel 8 SN65HVD82 RS485 channels as a multiplexer. 

All ~RE (receive enable) and DE (drive enable) lines are shorted together, and all driven at the same time from a single signal for half-duplex operation. D (driver data) lines are also all connected together and driven as one.

The R (receive data) lines are all passed through AND gates such that if any 1 channel goes low, the total AND gate goes low and the Microcontroller sees the incoming data. 

Relevant schematic snippets shown below.



My problem is that each transceiver is individually fused, and when one fuse pops and de-powers the transceiver, that transceivers receive line is pulled low which propagates through the AND gates, disabling reception of data from any of the other 8 channels.

My questions are:
a) Are there any transceivers with open-drain R (receive) lines / don't assert the R line while powered-off?
b) For this transceiver, when it is powered-off does the ~RE still control the output of the R line?

  • 0
    TI__Guru 74961 points
    Hayden Robinson said:
    a) Are there any transceivers with open-drain R (receive) lines / don't assert the R line while powered-off?

    As far as I know, we don't have any open drain Routput RS485 devices. If you did want to make an Routput that was open drain, you could probably put some discrete parts or an open drain level translator in front of the output. 

    The Routput pin should be high impedance if there is no Vcc. 

    Hayden Robinson said:
    b) For this transceiver, when it is powered-off does the ~RE still control the output of the R line?

    All the logic and outputs to the device do not function when the Vcc is absent. 

    -Bobby

  • 0 in reply to BOBBY
    Prodigy 10 points

    > The Routput pin should be high impedance if there is no Vcc. 

    This is the crux of my problem - I'm seeing that this is being pulled-low when I remove Vcc, despite external 10K pull-ups.

  • 0
    Prodigy 10 points

    Update to this. I've tried 3 boards now, all different channels on my boards, and can confirm that the R line pulls low when VCC loses power.

    With the board unpowered I've confirmed the 10K pull-up from the R line to my 5V supply rail (other side of polyfuse). A normally powered-on chip outputs 5V on the R line.

    When I've shorted the VCC side of the fuse to ground (and confirmed 0V at VCC pin) then monitored the R signal (pin 1) its sitting at 0.53V. This voltage remains unaffected by supplying 5V or 0V to the ~RE pin.

  • 0 in reply to Hayden Robinson
    TI__Guru 74961 points

    Hi Hayden,

    Are you sure it's actively driving to GND? My guess is there may be some kind of leakage current on the net pulling to GND. Are you able to place a pull up resistor of about 4.7k to the R pin during the issue and verify on an o-scope if it still holds to GND or if it pulls up closer to Vcc through the 4.7k resistor?

    -Bobby

  • 0 in reply to BOBBY
    Prodigy 10 points

    Hi Bobby, Will try later this week, I've been pulled onto another project in the short-term.

    I find this explanation unlikely though - In my images posted yesterday you can see the entirity of the net - it goes straight from the IC to a desoldered resistor, with just a 10K pull-up present on the line. I specifically desoldered the series resistor in the top-right to rule-out any external leakage current. 

    5V to 0.53V with a 10K pullup implies ~0.45mA leakage current.  This is far beyond the -10 to 10 uA specified in the datasheet.

    Never-the-less I'll see what the oscilloscope and 4K7 shows and report back sometime later this week.

  • 0 in reply to Hayden Robinson
    TI__Guru 74961 points

    Can you make sure the 4.7k reistor is connected to an independent  5V power rail that is active (ON)?

    I can try to order a sample of this device and see if I can put it on a board and see if I can get the output stuck with Vcc held to GND.

    I suspect this may take a while though for me to get the break out board and sample units. (maybe next week for the sample units and 2 weeks for the break out board).

    -Bobby