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.

LP87564-Q1: Buck output is getting off during long thermal testing

Part Number: LP87564-Q1

Hello TI experts,

We are using LP875640RNFTQ1 PMIC for the voltage and current requirement fulfillment in our design. We are very close to the production and doing thermal validation for the same.

We are facing some strange behavior from the PMIC at 80C ambient in thermal chamber. After some time when PMIC temperature is around 100C, some of the Buck outputs suddenly stopped working.

This is random, any of the Buck out of 4 or GPIO will be low and there is no request for the same from the Host side. Here are the details of the design.

Input voltage 3.3V

Buck 0 :- 1.1V @2A current (max)

Buck 1 :- 1.1V @1.5A current (max)

Buck 2 :-  1.07V @2.5A current (max)

Buck 3 :- 1.8V @ 1.5A current (max)

Please suggest us where and how to debug?

Snap of the design for your reference.

  • Hi Sevender,

    • Does ABA swap have the same result?
    • Do you have or can you get any scope captures of the outputs and inputs for me, including the enables?
    • Are you able to access the GUI for this device during this time?
    • Does this occur when load is present and not present?
    • Does this happen below or above this temperature?

    I may have more questions going forward, but the above would be helpful.

    Thanks,
    Field

  • Hi Field Tolson,

    Thank you for the response. Here is the reply for your questions.

    • Does ABA swap have the same result?--> Did not get What actually ABA swap means here?
    • Do you have or can you get any scope captures of the outputs and inputs for me, including the enables?-->Sure, I will try to capture as this is very random and all the things can't be capture in a single frame.
    • Are you able to access the GUI for this device during this time? --> We are testing our actual product, so we are not using any GUI. We can read registers of the PMIC through I2C only.
    • Does this occur when load is present and not present?--> We are testing with load only as the final use case.
    • Does this happen below or above this temperature?---> Below the 85C ambient we have not witnessed this issue yet.

    I hope you got some basic idea about my setup. Please guide.

  • Hi Sevender,

    ABA swap means moving the unit to a different board to see if the problem still occurs and trying a different unit on the board that showed the issue. This is just a common practice to isolate the behavior to the IC or the board.

    Field is out of office today but will be back tomorrow. I think the scope captures will be helpful for him to provide further guidance.

    Best regards,

    Matt

  • Hi Sevender,

    Did not get What actually ABA swap means here?

    As Matt mentioned, an ABA swap is taking a working device on another board/EVM and moving it to the board in question, and placing the device on the board in question on a working board/EVM and testing these to see if either or fail. This can help to tell whether it is the device or the board that is causing an issue.

    Sure, I will try to capture as this is very random and all the things can't be capture in a single frame.

    These would be much appreciated and are needed. The device has an absolute maximum rating of 150°C, with a recommended operating condition of 140°C for the junction and 125°C for ambient, and you are seeing this at a lower temperature. My idea is that something may be happening to an input, perhaps the enables, but may be others. This would also help to see what is transpiring. 

    We are testing our actual product, so we are not using any GUI. We can read registers of the PMIC through I2C only.

    Understood. The GUI has some interrupt and status bits that could be helpful or provide some clarity. It also has the functionality to read the registers and direct register access. There are some registers starting at 0x1A and onwards that may help to be read during an occasion if possible, in addition to the scope captures.

    We are testing with load only as the final use case.

    This is to separate the device and the load to see which one may be causing this. If load is connected and this is occurring, it's possible that the load side is affecting, driving something low, or causing something to happen. If the load is disconnected and this is occurring then we may be able to narrow this closer to the device instead of the load.

    Thanks,
    Field