• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 84 subscribers
  • Views 554 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.

TMS320F28075: VDDIO short to VSS and GPIO58 short to VSS.

Howard Zou
Genius 12365 points
Part Number: TMS320F28075


Hi, my customer is using F28075 and find several devices broken recently.

5 devices are VDDIO short to VSS. 2 devices are GPIO58 short to VSS.

The schematic is shown below.

Three GPIOs and GND on board 1 and board 2 are connected to one DIP switch with about 70cm long wire.

The problem happens after the wire extended from less than 10cm to about 70cm. So I guess the problem is caused by the long cable.

It's strange that they just put the board there with 3.3V powered, main power stage(buck) not powered, and the device failed without any touch.

1. What may cause the device fail? Are GPIO58 short to VSS and VDDIO short to VSS caused by the same reason?

2. How to solve this problem?

  • 0
    TI__Guru 52040 points
    I have notified the team and someone will help you shortly.
  • 0 in reply to Nima Eskandari
    TI__Genius 12365 points
    Thanks. Waiting for your reply.
  • 0 in reply to Howard Zou
    TI__Guru* 95315 points
    Howard,
    With the limited information you have provided, it is very hard to come up with a theory as to what could have happened. It is likely Electrical OverStress (EOS).

    Is it correct that when the cable length was 10 cm, no damage occurred no matter how many times you powered up the system and the damage occurred only when the cable length was increased to 70 cm?

    How are the two boards powered up? Is there a possibility that some GPIO pin is driven before the MCU is powered up?

    Are the GPIO pins input or output? Is there a possibility of contention?
  • 0 in reply to Hareesh Janakiraman
    TI__Genius 12365 points

    Hareesh,
    It's very rare for the damage occur when the cable length was 10cm. And the damage occurred 100% percent when the cable increased to 70cm.
    The two boards are powered through isolated flyback converter.
    But shown in the picture, GND1 and GND2 are connected together through two 50ohm resistor so they are finally not isolated.
    I will check if the GPIO pin is driven before the MCU is powered up. Will this cause the device break?
    The GPIOs are input from DIP switch.
    What do you mean by contention?

  • 0 in reply to Howard Zou
    TI__Guru* 95315 points
    The FA report confirmed my prediction of EOS damage. In addition to GPIO58 that you mentioned, it also identified failures in GPIO55 & GPIO59. I also see a mention of the same failure in 2016, so there is some history behind this.

    Contention is when 2 connected pins drive a complimentary signal.
  • 0
    TI__Genius 12365 points
    It's EOS issue and the root cause haven't been identified.
    Will post back if we find the root cause.
  • 0 in reply to Hareesh Janakiraman
    TI__Genius 12365 points

    Hareesh,

    As shown in the picture, 3 GPIOs are connected to DIP switch and only GPIO58 doesn't have 150ohm input resistor.

    And the FA report, 1pc GPIO55-59 are shorted, 2pcs GPIO58 only is shorted. So I guess GPIO58 the cause for other pin shorted.

    Is it possible that adding the input resistor to GPIO58 will help?

    Will isolation between DIP switch and the 3GPIOs help solve this problem?

  • 0 in reply to Howard Zou
    TI__Guru* 95315 points
    Input resistor to GPIO58 might help. Not sure what kind of isolation you are talking about for the DIP switch. It is very hard to theorize without knowing the operation sequence. I don’t think you answered the question of whether a GPIO pin is excited before power-up.
  • 0 in reply to Hareesh Janakiraman
    TI__Genius 12365 points
    Hareesh,
    GPIO pin is excited before power-up because they can not make sure that the two 3.3V for two F28075 are powered up the same time.
    If F28075 a is powered up earlier, it will excite the GPIO on F28075 b before b is powered up.
    But the interesting thing is that the break never happens during the power up, it happens when the two boards are powered up, and work for a while.