• State TI Thinks Resolved
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 128 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TPS65219: TPS6521904 LDO1 output issue

neil chen
Prodigy 140 points
Part Number: TPS65219
Other Parts Discussed in Thread: AM623

Tool/software:

Hi 

We refer to the circuit of EVM SK-62P-B1, which uses the TPS6521904 PMIC and is paired with the AM6234 CPU. According to the EVM test results, when booting from an SD card, the AM623's GPIO0_31 should output a "low" to the PMIC's VSEL_SD pin, causing the LDO1 output voltage VDDSHV_SDIO to switch from 3.3V to 1.8V.

However, when we use the EVM SD card to boot our own board, AM623's GPIO0_31 is not pulled to "low", but to a strange half-level position, making the LDO1 output voltage become about 2.5V.

We've conducted our own experiments: disconnecting the AM6234 from the PMIC's VSEL_SD pin. Simply pulling up the PMIC's VSEL_SD pin produces 3.3V, but pulling it down drops it to 2.8V. This seems to be a major issue. The relevant circuit diagram is shown below. Is there anything we can fix? Please help.

  

Thanks.

  • 0
    TI__Genius 9477 points

    Hi, 

    Regarding the TPS6521904 PMIC, 

    LDO1 is set to to act in bypass mode, so in order for LDO1 to effectively output 1.8V when VSEL is pulled low, PVIN_LDO1 needs to be changed to a 1.8 source (VCC_1V8 from Buck2).

    Or, using I2C, set the mode to LDO mode by setting bit 6 (LDO1_BYP_CONFIG) of register 0x07 (LDO1_VOUT) to '0' instead of '1'.

    Regarding the AM623 GPIO not outputting low, I would recommend creating a separate thread under the AM623 part number, so that their device experts may help troubleshoot this issue.

    Best Regards, 
    Sarah

  • 0 in reply to Sarah Lao
    Prodigy 140 points

    Hi 

    But the EVM PVIN_LDO1 use 3V3_MAIN, not 1.8V. We did not use I2C to program the EVM, but the voltage can be changed.

    Your description seems a bit different from actual usage. Or is there something we are not doing right?
    I'll ask about the AM623 part again.

    Thanks.

  • 0 in reply to neil chen
    TI__Genius 9477 points

    Hi neil, 

    Apologies for the confusion, you are correct.
    The bypass feature on LDO1 should enable the dynamic voltage change from 3.3V to 1.8V without the need to change PVIN_LDO1.

    I may need some more information to figure out why exactly this isn't outputting correctly,

    Do you know what the voltage level on VSEL_SD pin is when it is at this "half-level"?

    And then when you were manually pulling this pin low in your own testing, was this being connected to GND, or to what voltage level?

    And I understand you did not use I2C to program, but are you able to use I2C to provide a readback of the PMIC registers?

    Best Regards, 
    Sarah

  • 0 in reply to Sarah Lao
    Prodigy 140 points

    Hi Sarah

    "Do you know what the voltage level on VSEL_SD pin is when it is at this "half-level"?"
    About 2.5V

    "And then when you were manually pulling this pin low in your own testing, was this being connected to GND, or to what voltage level?"

    connected to GND (pull down 10K ohm)

    "And I understand you did not use I2C to program, but are you able to use I2C to provide a readback of the PMIC registers?"

    I don't have the tools at hand at the moment, will this be necessary?

    Thanks.

  • 0 in reply to neil chen
    TI__Genius 9477 points

    Hi neil, 

    Thanks for the information.

    That is fine, I2C may not be necessary yet. 

    To clarify, the voltage at the VSEL_SD pin is at 2.5V from VSEL_SD_SOC, and then you are also seeing 2.5V at the LDO1 output?

    This is strange, the  2.5V should still register as a high level input for the VSEL pin. In which case the LDO1 output should have still been 3.3V.

    Please confirm voltage measured at VSEL_SD pin for when it is grounded as well.

    Can you also share if there is any additional load/capacitance connected to the LDO1 output, VDDSHV_SDIO?

    Best Regards, 
    Sarah

  • 0 in reply to Sarah Lao
    Prodigy 140 points

    Hi Sarah

    The VDDSHV_SDIO only connection to AM623 VDDSHV5 & SD data pull up, no connection to other places.

    Also, please ask, the SD power supply VDD of EVM seems to have control timing, will this matter?

    And the VSEL_SD & VDDSHV_SIO state is as follows (In power on)

    Thanks.

  • 0 in reply to neil chen
    TI__Genius 9477 points

    Hi neil, 

    Thanks for the confirmation. 

    What do you mean by control timing?

    Have you tried replacing the PMIC on the board with another TPS6521904 PMIC to see if this issue persists?

    If this doesn't yield anything, I would like to try some I2C commands to further debug, if you are able to have this setup. 

    Best Regards, 
    Sarah

  • 0 in reply to Sarah Lao
    Prodigy 140 points

    Hi Sarah

    "What do you mean by control timing?"

    Because EVB controls 3.3V on the SD through the load switch, I want to say whether it requires timing control?

    "Have you tried replacing the PMIC on the board with another TPS6521904 PMIC to see if this issue persists?"
    We have 10 boards, but each one is in this situation, so I think it has nothing to do with PMIC

    "If this doesn't yield anything, I would like to try some I2C commands to further debug, if you are able to have this setup. "

    I need to prepare

    Thanks.

  • 0 in reply to neil chen
    TI__Genius 9477 points

    Hi neil,

    The SD / MMC1 of the AM623 may have some timing requirements. This would be more of a question for the AM62x team.
    I believe the AM62x datasheet has some details on the MMC1/MMC2 timing requirements, starting around page 193:




    VSEL pin does not have any timing requirements

    Let me know when you have the I2C setup ready.

    Thank you, 
    Sarah