• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 100 subscribers
  • Views 252 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.

AM6548: SD Card Boot Issue

Matthew McKee
Intellectual 926 points
Part Number: AM6548

Hello,

I am following up on a post from Paul Eaves about an SD card boot issue affecting the AM65x EVM. In his post, he states that a quick fix was found that involved removing capacitor C541 and installing a 360 Ohm shunt resistor across capacitor C180. This fix doesn't appear to be fully noted in the latest errata documentation, although advisory i2026 does suggest a workaround that would be the equivalent of installing a 360 Ohm shunt resistor across capacitor C180. Was the removal of capacitor C541 deemed unnecessary? Any updates on the SD card boot issue would be appreciated.

Thanks,

Matt McKee

  • 0
    TI__Guru 60925 points

    The reason for adding a shunt resistor across C180 is described in advisory i2026. This is being done to work around a device issue. The removal of the capacitor C541 addresses a different issue related to the EVM. Therefore, it was not described in the AM65x Silicon Errata.

     

    Removing capacitor C541 reduces a voltage transient on the MMC PHY bias supply when sourced from SDIO LDO and the IO supply switches from 3.3V to 1.8V. The power switch in SDIO LDO initially connects the MMC PHY IO supply to 3.3V during initialization, but switches it to the bias supply (1.8V LDO output) if the attached card supports one of the 1.8V data transfer modes. The 3.3V charge stored on all of the IO supply capacitors will be shared with the 1.8V charge stored on all of the bias supply capacitors when the power switch changes from 3.3V to 1.8V. This creates a positive voltage transient on the bias supply and the magnitude of the transient is a function of the total capacitance of each power rail. We determined the total IO supply capacitance was too large relative to the total bias supply capacitance, and C541 wasn’t necessary. So we removed this capacitor to reduce the magnitude of the transient.

     

    I have not checked the EVM documents to see if an errata has been created with an advisory describing this issue. If not, I suspect the EVM team has this task in their list of things that need to be done.

     

    Regards,

    Paul