NRF52832 Bricked, nrfjprog Segmentation fault

[lbe]

Hi everyone,

I am facing the weirdest problem here. 

I am working on a custom board using the NRF52832 which I was able to program without issues. I have been working with the NRF52 for a couple of here now, so I have a decent understanding of the whole environment. I use Segger Embedded Studio and occasionally nrfjprog command line.

The problem I have encountered is that I was modifying the current firmware we use on our board, and at a given point I was not able to communicate with the board anymore. On Segger Embedded Studio I got the error message "Failed to connect to target. No idcode detected" which is normally given when the board is not powered. Weird, but ok I though there was something wrong, so I moved to command line.

If I try to erase the whole chip with nrfjprog -e I got a "Segmentation fault. Core dumped" error. Same message if I try with a nrf52832 --recovery command.

The command line tools and segger embedded studio work fine with the development kits and with other custom boards (not of the same family of the broken one), so I thought there was something wrong with the board I am using, so i just took a second one. I programmed the firmware and I got the same problem.

Now, before I brick all our stock, is there any way I can recover the bricked boards?

Thanks for your help

[Kenneth]

Hi,

Is there any way we can recreate this here? E.g. can you share the hex file so I may program it to an nRF52-DK?

Best regards,
Kenneth

[lbe]

Hello, I don't think this will do any good, because if I program the NRF52-DK or another board (a completely different layout, design and PCB) the chip does not get bricked, so I can in any case erase the memory and reflash with another firmware.

At this point, my guess is that as soon as I give power to the bricked board, the firmware inside starts doing something weird during the configuration of some pins....but can this entirely block the programmer from communicating with the NRF52?

[Kenneth]

It's the first I have seen of this, let me check with the team internally if they may have any suggestion.

[lbe]

It's the first time for me as well....so any suggestion would be appreciated. 

[Kenneth]

May be some corner case, can you run:

nrfjprog --version
nrfjprog --recover --log

[lbe]

Hello, I was able to fix the issue and to reprogram the board. In an old thread (sorry I lost the link...), I found someone complaining of a bricked NRF52....and a forum user for suggesting that maybe the firmware was doing weird things immediately after starting. I thought this could be the case, so what I did was to connect the programmer just an instant before the real programming started, so that the board itself was not powered before. After a few trials, I was able to time the connection correctly, so that I was able to flash a clean firmware and recover the NRF52. 

I still don't understand why the --recover of nrfjprog was not able to do its work, but at least I was able to recover the bricked board.

Thansk for your suggestions.