Hi,

Does anyone have suggestions on how to get a board out of bootloader mode beyond just the reset button?

I have a feather that I have been using for the last couple months with very few issues. A couple days ago after uploading some code to the board, I got no response in the serial terminal after pressing the reset button.

After much debugging it appears the board is stuck in bootloader mode (I can’t verify with the blue led since I had to remove it due to a previous issue of it not turning off - even though it would exit bootloader mode). A few notes on it.

  1. I’ll disconnect all power to the board
  2. When I try plug it back in I can run the zephyr-tools newtmgr -c .... upload command and it will successfully upload to the board (no need to put it into bootloader)
  3. I have tried 5 different applications, including many of the basic sample applications
  4. Even after uploading the AT Client and disconnecting and reconnecting the board to the serial, it won’t display any output
  5. I have uploaded a program that I know connects to the hologram network automatically. After uploading the code and resetting the device, it doesn’t show up as connected in the hologram dashboard (e.g. the code doesn’t run).

Unfortunately I have no blue led anymore to test a simple blinky application. Mentioned in a previous post, but the blue LED wouldn’t turn off even though it would still exit bootloader mode (not sure if it’s related).

We have about 5 of these boards across our developers and they haven’t found any issues with our test code or device overlays. Might just a specific board issue.

@callumctaylor interesting issue.

Do you have your board plugged into anything? Do you have a multimeter? Can you measure at the mode (MD) pin, and then also at the mode button to see what the voltage levels are? (Measure against GND)

The polarity of the LED also changed in v5 so you need to update your device overlay if you want your LED back. (That is if you are using V5)

Also, what version is your board? The way the devices partition changed across different SDKs so if you have an older boot loader, it will not take newer versions of the SDK (at least, not without some finagling)

All the boards we have are V5. The board is by itself and have tried powering through VBUS and VBAT independently and combined. It draws about 10mA at 3.7V from my power supply when I power through VBAT. I can see it dip down and jump back up to 10mA when I press the reset button. I also checked with a multimeter and reset pin is properly going to ground when button is pressed.

From testing against GND
Mode pin: 0.638V
Mode button: 0.638V (on mode/wake pin - tested with VSYS at 5V and 3.7V)
RST pin: 2.34V
RST button: 2.34V
Does this all look right to you?

The updated overlay didn’t fix the blue LED problem (it worked for all other boards we had), so I think we might have physically damaged this specific board. Really appreciate the support and looks like it might just be an isolated issue.

    Yea 10mA sounds a bit much unless you’re using the cellular modem part of the chip.

    callumctaylor The updated overlay didn’t fix the blue LED problem (it worked for all other boards we had), so I think we might have physically damaged this specific board. Really appreciate the support and looks like it might just be an isolated issue.

    Hmm yea interesting.. sounds like an isolated issue to me.

    callumctaylor Mode pin: 0.638V
    Mode button: 0.638V (on mode/wake pin - tested with VSYS at 5V and 3.7V)

    Speaking particularly to this, there’s an internal pull-up on P0.12 that should be keeping the pin high. Can you check pin 3 of Q4?

    By default the MD header pin should be close to 0V while P0.12 on the other side of the MOSFET should be close to 3.3V

        Will check the pin when back in the lab, but new board is working perfectly with same code so I think we might have just damaged the old board.

        Cheers again for the help.

        Terms and Conditions | Privacy Policy