First time user trying to install OpenWRT with poor luck here.
I got my hands on a bricked C7 v2. When it's powered on only power, status and 5GHz lights are constantly on but nothing else happens. No Wifi networks will show up. LAN light starts to blink (evenly) when a cable is connected between LAN and a PC but I get no network connectivity. Holding reset does nothing.
Tried TFTP but same thing, nothing happens on the PC side.
Connected with serial and this is what I get in console:
Normal power on:
U-Boot 1.1.4 (Dec 17 2013 - 16:37:27)
ap135 - Scorpion 1.0DRAM:
sri
Scorpion 1.0
ath_ddr_initial_config(178): (32bit) ddr2 init
tap = 0x00000003
Tap (low, high) = (0xaa55aa55, 0x0)
Tap values = (0x8, 0x8, 0x8, 0x8)
128 MB
then nothing.
Power on with reset/wps button held down:
U-Boot 1.1.4 (Dec 17 2013 - 16:37:27)
ap135 - Scorpion 1.0DRAM:
sri
Scorpion 1.0
ath_ddr_initial_config(178): (32bit) ddr2 init
tap = 0x00000003
Tap (low, high) = (0xaa55aa55, 0x0)
Tap values = (0x8, 0x8, 0x8, 0x8)
4 MB
Beyond repair, maybe…, maybe not - but it's certainly not done with push-button tftp or serial console access (raising the question if it's still economically feasible to try, despite an unknown outcome).
In the 'best' case the bootloader is shot, this means you would have to recover via an spi-nor writer (desolder flash chip, recover flash contents, hope that the calibration data is still intact (it likely isn't) and extract ART - assemble new image (OEM bootloader + u-boot-env + firmware + ART) and flash, solder back into the board, hope for the best).
What would worry me, are basically two things:
parts of the bootloader are obviously still present (starting with U-Boot 1.1.4 (Dec 17 2013 - 16:37:27)), so it's either just partially clobbered, or…
the detected RAM size differs between your attempts (128 MB vs 4 MB), hinting at a more sinister issue - suggesting that the bootloader might even still be intact, but fails to properly initialize the hardware (RAM) <-- hardware damage.
Impossible to say for sure, if you have soldering equipment and a spi-nor writer it might be worth trying, if you'd have to go shopping first, I'd save my money and get a working specimen instead. Chances of recovery are under 50% (significantly, thinking about ART in case of option 1 - and option 2 would be fatal).
Thank you for taking the time to write such a detailed reply.
I was afraid of this and also thought it was strange the reported RAM was not the same between tries. I don't have a spi-nor writer and I agree it's definitely not worth going for shopping as a new router is not exactly expensive. Even if I had the equipment at hand the process seems to be quite complicated for me so I'm just going to save my money and time for something else.