• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 64 subscribers
  • Views 614 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TPS23754: Shutting down after 350ms

Mark Nohr
Intellectual 570 points
Part Number: TPS23754
Other Parts Discussed in Thread: TPS65400,

What appears to be the negotiation phase happens and the supply starts to regulate then turns off. 

The mystery is that there are other boards from the same lot that function properly. About 20% have this issue.

Design follows PMP5236 RevA including the layout (see images below)

See attached pdf schematic. 

Measurements taken from board with no load/ no regulation beyond the 12V output. Also tried attaching a 100ohm resistive load to the 12V.

The other phenomena is that the 12V output does not rise to 12V but regulates below @5V before falling to 0 at the 350 ms time. 

Yellow trace is voltage and blue trace is current (10 ohm resistor)

SSN_Multi_Serial_Local_Pwr.pdf

View from top

View from bottom

Top view Red top traces

Bottom view Blue -Bottom traces

  • 0
    TI__Genius 9230 points
    Hi Mark,

    Thanks for posting. I'll take a look at your schematic & post something tomorrow when I'm back in office.

    Regards,
    Tom Amlee
    PoE PD Apps Engineer
  • 0 in reply to Thomas Amlee
    TI__Genius 9230 points
    Hi Mark,

    Are you meeting the maintain power signature current of 10mA at the input? Does it remove input voltage with load?

    Also, do you have a waveform for the behavior of Vout, and a waveform for Vc?

    Thanks,
    Tom Amlee
    PoE PD Apps Engineer
  • 0 in reply to Thomas Amlee
    Intellectual 570 points

    I'm not sure what you mean by "meeting the maintain power signature current of 10mA at the input" - are you asking if there is a 10mA load always present?

    Does it remove input voltage with load? -  if so the answer is YES

    I will capture the Waveforms for Vout and Vc

  • 0 in reply to Thomas Amlee
    Intellectual 570 points

    Removed IC U28 to remove all loads except the bulk and decoupling capacitors.

    Added 1K Resistor on top of C63 for 1K Load which should present a 12mA resistive load.

    Pink line is VC

    Blue Line is Vout/12V

    Zoomed out to longer timeframe to show output ramp down. 

  • 0 in reply to Mark Nohr
    Intellectual 570 points

    I decided to take two trace shots. One of a working board and one of a the failing board (with the downstream supplies removed and the resistive load installed)

    Here is the working board 

    And here is the same settings with the non working board.

  • 0 in reply to Mark Nohr
    TI__Genius 9230 points

    Hi Mark,

    On the non-working boards, could you check to see how much capacitance & if there is some leakage on the Vc pin? The 22uF you have on the schematic is true to the original PMP5236, but there may an assembly error.

    Also, I've reviewed your schematic & have not found any issues. As you've noted above, be sure that the TPS65400 buck converter is disconnected, and that C65 & C166 input bulk capacitors are depopulated during debugging.


    Thanks,
    Tom Amlee
    PoE PD Apps Engineer

  • 0 in reply to Thomas Amlee
    Intellectual 570 points
    Ok I'll check.
    Please recall that I'm having this 'yield' issue on other designs that were built from a different stock of components.
    I also found it surprising the voltage on the VC pin was peaking at ~55V since the rating on the 22uF cap is 16V. Is this peak expected?

    Beyond this, is there something else I should looking at?
    Is it significant that the output voltage rises gradually almost immediately in the non working board but it rises sharply after 80ms in the working unit?
  • 0 in reply to Mark Nohr
    Intellectual 570 points
    So I found that there was an open via causing a disconnect C77+TPS23754 and D11+C70+C71
    Adding a rework jumper between the two isolated segments of the VC net resolved the issue.

    I checked the other designs that are having similar symptoms and they do not have the same problem so I will continue going through those to root cause that issue.