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.

AM5716: Issue UHS-I of SD card

Part Number: AM5716

Hi,

I have a question related below thread.

https://e2e.ti.com/support/processors/f/791/t/884770

my customer is developing their own board. Should they have added the ability to power cycle their boards using PORz / soft-reset to support UHS-I? Or any other work around? We would appreciate it if you would tell us any idea.

Best Regards,

M.Ohhashi

  • When supporting UHS-I card (dual voltage), there can be cases where the controller (AM571x) and the card get out-of-sync in regards to voltage.  Both default to 3.3V, and the controller sends commands to the card when switching to 1.8V.  If the card is removed, the controller can detect the change in card detect signal, and set supply back to default (3.3v) for when the card is re-inserted.  The issue is when a reset event occurs to the AM571x processor.  This would likely cause the controllers power to revert back to default (3.3v), but the card has not reset mechanism - thus would stay at 1.8v.  The only way to reset the card is to eject/re-insert.

    An alternate way to perform this 'card reset' is to cycle the power to the card, which can be done load switch or similar, which can be controlled via the controller.

  • Hi Robert-san,

    Thank you for your reply. My customer will reset the card by restarting power supply of the card.

    They will control power supply by PMIC (TPS659162RGZR). So please tell us about the following.

    1. Should the card power supply voltage be changed in the order of + 1.8V → 0V → + 3.3V?
    Can you tell us the sequence of changing the voltage (such as flags that need to be checked) and the flow?

    2. When changing the above voltage with a PMIC, there is a period when the power supply (VDDSHV8) of the AM5716 SD block becomes 0V. Is there any problem in use? Are settings such as SD block registers initialized?

    3. There may be cases where the CPU is reset during the UHS-1 switching sequence, but they want to determine if this is the case. Please tell us where to check to see if it is in that state.

    Best Regards,

    M.Ohhashi

  • 1. The card supply voltage does not change with IO voltage.  Card supply voltage is 3.3v, and remains 3.3V even when IO's are configured for 1.8V.  The SD card spec should provide the commands and sequence for switching from 3.3v IO to 1.8V IO.  (Don't confuse Card Supply voltage with SD IO voltage)

    2.  VDDSHV8 is the SD IO voltage.  This should not be 0v.  It is either 1.8V or 3.3V.  By default, the supply is 3.3V.  When sending commands to transition the card from IO from 3.3 to 1.8v, the PMIC can also change the its regulator from 3.3V to 1.8V.  If the card is reset (such that it defaults back to 3.3v IO, then the PMIC too will need to be transitioned back to 3.3v IO). 

    3. Not understanding question.  Do you want to read the card status to determine if it is in UHS-1 state or not?

  • Robert-san,

    Thank you for your reply.

    2. We would like to confirm. Since VDDSHV8 never goes to 0V, does it mean that the SD card block is not initialized and remains as it is?

    3. We are sorry. This is our misunderstanding. We thought that the CPU would reset because there was a moment when the voltage from the PMIC went to 0V.

    Best Regards,

    M.Ohhashi

  • 2.  The only way to reset the SD card is to power cycle the card (either by ejecting it or by cycling the power to the card (via on-board logic).  VDDSHV8 is the IO power supply to the processor, it is NOT power to the SD card.

    3.  I don't understand your question.  Please repeat what information you are requestiong.

    Thanks.