(July 2024)
Recovering my non-booting MBAir 2012
Steve-Jobs-approved. Or at least Steve-Wozniak-and-McGyver-approved :-)
Just before my company got acqui-hired back in 2012, my cofounders decided we deserved excellent HW as a farewell gift. With their 8GB of memory and 256GB SSDs, MBAirs were considered beefy machines; and they were so lightweight! Arguably the best laptop at the time.
I obviously installed Linux on it immediately. And back then, that was no easy feat - driver issues aplenty... But I persevered, and basically never used OSX.
12 years later, it still works. I periodically boot it up and pacman -Syu to keep Arch updated.
I consider it my fallback machine.
One of those days
Saturday, yesterday. I wake up and think to myself I'll start the day watching Big Clive reverse engineer some electronics.
I try booting my daily-driver laptop... one that runs much faster than the MBAir, and also consumes far less power.
...but get nothing. No power.
I decide to look into it later, and just boot my fallback MBAir from 2012 in the meantime...
...only to find out that the MBAir doesn't power up either!
Sigh.
Did I get hit by an EMP overnight, or something?
I plug the MBAir to the charger; no difference.
Search engines point me to articles suggesting resetting the SMC. Various key combinations are supposed to be pressed during powerup...
Tried them all. No joy.
So instead of watching Big Clive tinker with electronics, I instead have two non-working laptops.
EXCELLENT. Sorry Big Clive, looks like I'll be the one doing the electronics tinkering today.
Which laptop should I investigate first?
Well, the MBAir from 2012 - due to its age - has leaked schematics all over the web. My daily driver laptop, like almost everything else out there, is a black box. Obvious choice: use the schematic, Luke.
So I Google for it, and search inside it for anything related to the power button.
SMC and power button
I see; there's an SMC_ONOFF_L signal that feeds the SMC's "power button" input. OK, maybe my laptop's power button isn't working, and doesn't drive the signal through...
It's a _L (active low) signal, so can I temporarily short it to ground to see if the laptop boots?
Searching for the SMC_ONOFF_L signal...
Bypassing
Aha! There are OMIT-ed "Debug" resistors, that would be used during debugging to pull the signal down to ground. All I need to do, is temporariy short to ground, and my ancient relic may, in fact, boot...
Where is that signal accessible?
Openboardview to the rescue, and... there we are:
The engineer's way to boot up
I short the two solder blobs...
... and the laptop powers up.
The Steve-Jobs-approved aesthetically pleasing solution
OK, then.
I pick up some enamelled wire that I salvaged from the dead corpse of a Chinese death-power-supply-adapter (yes, I am that kind of guy). I solder to the two points, and secure the enamelled wires on the motherboard with Kapton tape.
Then I do the unthinkable, and drill a hole on the side of my MBAir, to pass the two enamelled wires through.
The ensemble is completed with an SMD button that is stuck on the side of the laptop with double-sided tape.
I am pretty sure Steve Jobs's ghost is vomiting...
...but I am also sure Steve Wozniak and McGyver would approve :-)
Steve-Jobs-approved. Or at least Steve-Wozniak-and-McGyver-approved :-)
| Index | CV | Updated: Sun Jul 28 10:48:14 2024 |