No announcement yet.

recovery after error

  • Filter
  • Time
  • Show
Clear All
new posts

  • #16

    Yeah, that's why I was asking what sort of action would convince the video module that the card had been removed and re-inserted, because that's doing the trick, and I was hoping it could've (or had) been done in the module's firmware. Nothing else happens when the uSD card is removed...the module continues to run, doesn't reboot again, etc. It's just that init succeeds the next time through.

    I've been doing a fair amount of reading about SD cards, and see no mention of low-voltage shut-down/lock-out anywhere.
    More soon.
    cked="false" Priority="22" SemiHidden="false" UnhideWhenUsed="false" QFormat="true" Name="Strong"/> ame="Colorful Shading Accent 1"/> UnhideWhenUsed="false" Name="Light Grid Accent 4"/> ity="67" SemiHidden="false" UnhideWhenUsed="false" Name="Medium Grid 1 Accent 6"/>


    • #17

      Haven't read that document very much, but what I'm thinking the problem is may lie around the workings of OCR (search for it in the doc) and the various gotchas associated with it.

      You can find the definitive document by googling Simplified Physical Layer Spec.


      • #18

        I looked through a bunch of the OCR material, and I'm still not seeing anything in the spec that would allow the card to "lock up" until it's ejected. But I'm definitely familiar with how hairy it can be to cover all of the SPI bases, particularly if you've never seen the condition in question (which is why I was saying how it could probably be re-created there using a bench power supply).

        Looking into other options to try to get the board to do the same sort of reset it does when it first sees power (short of exposing a normally-closed reset switch to my users...not good)...there's no software way to do that, right?

        Back to your voltage sag theory, like I said, the module is downstream of a boost converter that I can pretty much guarantee isn't slipping to where a 3.3V regulator won't work. We'll look again, though. We soldered over/around the reverse-battery diode on one of these, just to rule out marginal voltage even further...that was no help.

        Back to head-scratching, here.