I've got a brand new B404x that doesn't seem to ever actually turn on fully. The yellow charge indicator flashes continuously very fast but no status light ever comes on. The 3.3V pin is 0V as well regardless of if I ground the EN pin or not.
I initially sent this through the email support, but was told to post here instead.
I had the same thing happen to a Boron404x recentl. Was testing some code involving sleep mode and it would not reset, etc. tried CLI and not luck the led was off mostly and would briefly flash white when I tried rest or tried to enter DFU mode. The charging light was flashing very quickly. That was all I got. Thought it was bricked. Hit on the idea of powering it with the battery only using my LiPo battery on the normal charge port. Once I did that and then plugged in the usb cable it stopped as the battery was already fully charged. Used the buttons to get it in DFU mode with CLI then updated and loaded Tinker. All now working. Don’t know if it will work for you, but worked for me. Figured something to do with charging firmaware had gone wrong, but don’t really know. I had entered some dodgy sleep code for testing……
If you have code that immediately is going into sleep on wake you can also hold down the MODE button while plugging in USB power. This will allow you to go into safe mode or DFU mode when you can't press the normal button sequence fast enough at boot.
As for the original problem, if there is no 3.3V output on 3V3, it's probably a hardware problem, most likely with the 3.3V regulator.
If you power by USB or battery, the pull-up on EN should be 3.6V to 5V depending on your power source. If that's high and 3V3 is off, then it's almost certainly a regulator issue.
Sounds like this one has a regulator issue then. Is there a warranty that we can work through? I'm not sure what its device id is since it has never been powered up and communicating over USB.
VUSB: 4.96V
EN: 4.26V
3V3: 0.0V
I have a friend I support with IoT devices that had this same thing happen. He put the Boron into deep sleep / woke it up over and over again, and this problem eventually happened.
He tried everything but only when connecting a LiPo battery did the Boron finally boot.
I have a theory I’m trying out: I think the BQ24195 charger IC gets confused during a sleep/wakeup and stops outputting power to the rest of the Boron.
I think it’s necessary (and as per the datasheet) to inform the BQ24195 that there isn’t a battery connected before putting the Boron into a deep sleep.
I’m trying to reproduce the problem first, and then I’ll add the code to tell the BQ chip there’s no battery before trying to produce the problem again. I will report back in a week or so.
I personally can only get this behavior if I provide a leakage path via a GPIO pin during USB power removal and re-insertion. But even then, the reset button will reboot the Boron. The latch-up is clearly a known effect. My friend swears he had nothing but a switch connected to the Boron during this problem. This shall remain a mystery.
That said, I'm curious as to why the yellow charging LED flashes at a very fast rate during this lockup period. After looking at the Boron schematics and the BQ24195 datasheet, I see that the Boron designer chose to use SYS as the yellow LED's power rail rather than REGN as per the TI datasheet. Is there a good reason for this?
Also, even with when the Boron does boot it flashes the yellow LED for a brief period. Is there some sort of oscillation going on here?
Lastly, to satisfy my own curiosity I'd love to look at the source boot code that is not re-programmable and that does the initial Boron booting. Maybe there is some sort of adverse loop it can enter when interacting with the BQ24195 chip which my friend swears was remedied by connecting a LiPo battery to it.
My friend's problem with the Boron not booting (and flashing charger light yellow rapidly) turned out to be power related.
He was using 3 x 1.5V Alkaline batteries to power the VUSB pin. The voltage had dropped to 4.4V total after much usage and the Boron refused to boot (with the yellow charger light blinking fast).
I've easily reproduced it with a lab supply. Mystery solved... (Occum's razor intact).