Using SD card while in IDE

Hello, all!

I’ve been chomping at the bit to get into machine vision with OpenMV, but until now, school and work have prevented me from poking around too much. I’ve just downloaded the Mac IDE and started looking at examples to begin to familiarize myself with the board’s capabilities. The basics worked just fine, but when I got to scripts which needed SD cards, I started having trouble. I have an SD card, but loading it into the board makes it completely unresponsive to the IDE, blinking blue and yellow. I imagine this is because the board is trying to find the script, so I tried adding a script requiring an SD card (the basic differencing script is the one I used) to the SD card, but its behavior doesn’t change. I also tried to run it without an SD card for kicks and got the error I expected about failing to write the requested bytes.

Any advice here? I’m really looking forward to the tutorials, because I am woefully unprepared for this machine vision jargon! Time to read up! If you have any sources for learning this topics, feel free to drop those here, too.


Hi, you shouldn’t be encounter in such problems. Um, try reformatting you as card for fat32. If the SD card is NTFS then the OpenMV cam can’t read it. Also, use a less than 2gb SD card if possible.

Hi, there really seems to be a new issue with microSD card handling (for me not IDE specific):

  1. With “old” firmware
Micro Python v1.4.4-189-g4f6880b on 2016-05-03; OPENMV2 with STM32F427

powering up with inserted microSD card everything looks fine:

>>> import os
>>> os.getcwd()
  1. With new (well, 2 days old) firmware binary (
MicroPython v1.8-4338-g200e05f on 2016-08-14; OPENMV2 with STM32F427

powering up with inserted microSD card (the same tested working card as in case 1) leads to white led and “stuck system” (can’t get REPL in terminal emulation).
Leaving sd card away is “o.k.”:

>>> import os
>>> os.getcwd()

Tested by reflashing back and forth (would you prefer github issue?).


Yes, please make one. Ibrahim needs to get on this. We’ve been moving things around in the firmware to free up RAM and he’s trying to remove the flash buffer (16KB) so that we can have a separate JPEG buffer for IDE streaming.