• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 230 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.

SN65HVD72: 24V overload on A/B signals

Jiri Naprstek2
Prodigy 100 points
Part Number: SN65HVD72

Dear all, we have got a design with SN65HVD72 and extra external world protection (similar to what is suggested in UG with 10R series resistors). Unfortunately, due to miswiring the device, in some cases the bus signals (A or B or both) were connected to 24V for several seconds.

Inspecting such devices the input TVS (CDSOT..) was damaged and the series resistors (10R) burned (0603, general purpose rating). Later bench top testing revealed that it looks like the transceiver itself works, however we are not sure about its health as the trasceiver must have experienced excursions beyond absolute maximum ratings.

Due to the silicon starvation issues, we would like to repair such devices, for which I'd like to ask you - in such situation, is it possible, that the transceiver would have sent something nasty to the supply line and/or to the data/enable signals? We are happy to replace the transceiver as a suspect part, however the 3V3 supplies number of other components and data signals are connected to an onboard MCU (DE/RE directly, TX has got series 100R and RX over series 1k).

The devices _seem_ working well (after replacing the TVS and passives), however we are concerned about latent damages that may shorten the device lifetime unpredictably.

  • 0
    TI__Mastermind 47328 points

    Hi Jiri,

    Any time you operate outside the absolute maximum conditions of a device you jeopardize the life expectancy of the device. Just because the device seems to work perfectly now, doesn't mean that the device will operate the same tomorrow. I understand that it would be a pain to receive a replacement, but that is what I suggest for good reliability for your system.

    From datasheet:

    Since inputs A and B were operated above the maximum voltage of 16.5V (24V for several seconds), I would not trust the reliability of the device. 

    As for signals that could have appeared to other connected devices, there is not a defined way for me to tell the exact behavior of the device if operated above abs. max conditions. 

    Regards,

    -Tyler

  • 0 in reply to Tyler Townsend
    Prodigy 100 points

    Hi Tyler,

       thanks for your answer, just to confirm - my aim is not to save the transceiver itself, for me it is gone even if it seems working. The crucial point of my question was, whether (with your insider knowledge of the silicon structure) you could elaborate, if such type of abuse may lead into shoot-through that would bring the voltage to data pins and/or supply pin? (please also note the series resistors that limited the input current, 24V was applier to the connector, not to the chip pins)

    The whole RS485 front end would get replaced for sure. My only aim at the moment it to assess the risk of keeping the microcontroller, which is now impossible to purchase and replace.

    Thank you,

    Jiri

  • +1 in reply to Jiri Naprstek2
    TI__Mastermind 47328 points

    Jiri,

    I would suspect that protection circuitry (TVS diodes, series resistors...) did their job by burning up before applying harmful conditions to the SN65HVD72, since the transceiver still seems to function normally. That being said, if the transistor actually saw conditions above absolute max, our devices are designed to fail before contributing harmful conditions downstream to other important devices such as your MCU. Generally, when a device fails internally due to exceeding current limits, silicon in the IC burns up leaving less conductive material behind than what was there originally. In the most extreme cases, voltage could appear on different pins than intended due to arcing over a dielectric, but since we are dealing with 24V, I don't think this would be a problem. 

    Regards,

    -Tyler

  • 0 in reply to Tyler Townsend
    Prodigy 100 points

    Hi Tyler, thanks a lot, that was exactly I was after!

    Much appreciated,

    Jiri