• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 1518 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.

RM48L952: VCCIO Short Circuit

Will Treharne
Intellectual 270 points
Part Number: RM48L952

Hello,

We have experienced several failures of PCBs which use the RM48L952 which we have so far been unable to explain. After the failures occur, the RM48 is overloading the 3V3 rail - we were able to isolate the VCCIO supply to the RM48, after which the TPS65381A power supply IC recovers. The failures appear to occur some time (hours or days) after connecting the board containing the RM48 to an I/O interface board, but we have not been able to find anything wrong with this interface board. All I/O voltages look sensible and we haven't found anything that could cause an over-voltage on the RM48 pins. The only thing we can think of is that at some point the logic on one of the GIO, N2HET or I2C lines running between the boards might have been in a conflicted state, but without much evidence.

Is anyone able to offer some advice on how to determine in what way the IC has been damaged or how to go about finding the root cause?

Many thanks

  • 0
    TI__Guru**** 196286 points
    Hello Will,

    The failure occurs after connecting the I/O interface board to the RM48 main board. If the I/O interface board is not attached, there is no error or failure after running the same test code on RM48, is my understanding correct?

    It would be helpful if you post the schematic of the I/O interface board.
  • 0 in reply to QJ Wang
    Intellectual 270 points

    Hello,

    Yes, your understanding is correct. We don't see any problems before we connect the I/O interface board. After the failure occurs the I/O board is still functional, but the RM48 board is damaged and doesn't recover after the I/O board is removed.

    The I/O board has interfaces for 24V digital inputs and outputs like the following:

    There's also two Microchip MCP23017 I2C I/O expanders connected to more of the above interfaces to give some additional I/O. In total there are around 30 inputs and 30 outputs.

    Have you got any tips to help us narrow down what part of the RM48 has been damaged?

    Many thanks

  • 0 in reply to Will Treharne
    TI__Guru**** 196286 points
    Hello Will,

    The circuitry looks good to me. After failure, how about the power supplies to MCU core and I/O (1.2V, and 3.3V) ? is the 3.3V on I/O card from the MCU board? is it possible that the 3.3V on I/O card was shorted to DGND or the DGND was shorted to other power supply (>3.3V)?
  • 0 in reply to QJ Wang
    Intellectual 270 points

    Hello,

    The 1.2V and 5V supplies look fine. It's just the 3.3V supply to RM48 being pulled down.There are some 0R links on the board which allow us to disconnect the 3.3V supply from the microcontroller, after which the 3.3V supply recovers. All the supplies are on the same board as the MCU.

    I don't think it's possible that supply was shorted to DGND or DGND to any other supply as everything functions correctly for some time before the failure occurs.

    Before the I/O board is connected the VADREFHI and VADREFLO are both floating. When the I/O board is connected, VADREFLO gets grounded, but VADREFHI is left floating. Is likely to be an issue?

    Many thanks,

    Will

  • 0 in reply to Will Treharne
    TI__Guru**** 196286 points
    Hello Will,

    It is ok to leave the VADREFHI floating and connect VADREFLO to GND. After the 0ohm resistor on 3.3V power supply to MCU is removed, the 3.3V comes back to normal. The 3.3V on MCU may be shorted to GND.
  • 0 in reply to QJ Wang
    Intellectual 270 points
    Hi,

    Thanks for the confirmation on the ADC reference inputs.

    After the MCU failure occurs, the 3.3V supply on the MCU appears to be internally shorted to ground. We're confident that there aren't any external shorts.

    Having determined that the short is caused by the MCU, are there any external measurements we could take to help us figure out more specifically what on the MCU has failed (e.g. which pins or peripheral have failed)? If there is, it might help us narrow down the root cause.

    Many thanks
  • 0 in reply to Will Treharne
    TI__Guru**** 196286 points
    Hello Will,

    No, I can not guess which pin or ball (BGA package) is shorted to Ground. You need to check all the signals routed to the connector for I/O board.