Bose SoundDock series 1 USB power mod
Here's yet another piece of kit I stumbled upon while dumpster diving: A Bose SoundDock complete with its power supply (but missing the remote :( ). A first test with my iPod 3G showed that it basically worked flawlessly, except for a little loose joint at the supply connector. This was easily fixed with a little careful shaping of the single contacts though, and now I own a nice little sound system and saved it from the landfill for now. But there's this little ''but''...: the iPod won't charge off the dock! Since I'm not very familiar with bitten apple technology, I had to google for a little enlightenment. It turns out that the first generations of i-Pid/Pad/Pud's (which this dock was made for) used to receive their charge via FireWire only, and not USB, like it's usual nowadays. So I'd have to look why there were no 5V there.
The SoundDock is powered through a SMPS with 2 x 18V @ 1A each, and so 18V was the voltage at the FireWire-Power-Pin. Inside are two linear regulators, which supply the digital logic with some 3.3V and 1.xV. No 5V there at all. Bummer. Okay, so I'd have to pull that out of the hat somewhere. Since power-hungry state-of-the-art battery-powered equipment easily sucks 2A out of the charger, I wouldn't get away with just another linear regulator, so a switch-mode controller had to find its place inside the case. I didn't have anything suitable at hand, so I ordered a LM2576 in DDPAK case, together with a beefy schottky-diode and a 1000uF/6.3V cap. They didn't stock any suitable coils, so I used a 47uH salvaged from the digital audio section of a dead TV set. The regulator came as a fixed 5V part to keep external parts count low, so I only needed coil, diode and cap – and that's it.
make some creative use of the existing space...
For a good thermal connection (and electrical conn' to GND, of course) I soldered the controller directly to the metal can. The other parts were then fitted within the existing space and wired accordingly. The orange wire connects the input directly to +18V from the supply connector, since there were no +18V inside the can. Coil and cap could've been a little smaller, since the lid doesn't fit snugly anymore, but there's enough room in the case anyway.
So there's our 5V supply. Now to get them to the iPod-Connector, which is connected to the mainboard via a 24-pin flex. I planned to use one or two of that pins for my 5V, but I wasn't able to separate them from the original traces due to some buried vias, so I ultimately soldered an extra wire between the two boards.
'5V is not enough'
I separated a little pad from the GND layer to solder the 5V wire onto. The real problem however was to run a wire to the iPod-Connector, since the needed Pin 8 wasn't connected to any trace on the board. So I had to solder my tiny 34 gauge directly to the connector pin, with a solder tip which was as wide as three of those pins :( . Wasn't easy, but in the end it worked out okay :) .
So I put everything back together and.. it didn't work :( . Damnation. Checked everything again, everything was working like it was supposed to, so what the hell...? To figure out what was wrong, I connected the iPod to my lab supply with an USB cable, and there it was: The pod sucked the allowed maximum of 100mA, but that was not enough to charge it! So I'd have to resemble a correct charger to convince the iPod that it's allowed to pull some more amps. pinouts.ru has all the needed information, so I set up two voltage dividers at the USB-data pins, consisting of 33k/22k and 33k/47k between 5V and GND. Finding some space for four little chips was easy and soldering the wires to the existing vias (which didn't connect to anything else, btw) wasn't hard, either. And that was it:
battery is getting charged!
While googling around I found out that Bose offers a replacement board, but that's rather expensive, in the range of $100. And where's the fun in that? ;)
Four and a half years later...
After a couple of years in use the SoundDock started to crap out again. There were some terrible noises emerging, especially audible when the battery was charging and no music playing, up until the day when not even the battery was charged anymore. All signs pointing at a faulty switcher, so I revisited that thing.
Cheapo-Teapo on the bulge
The culprit was found immediately - by eye as well as by nose. The smell of cheap electrolyte in the morning... Since I didn't have any suitable caps at hand, I simply threw in a 2-buck, 2-amp DC-DC-module and everything's back up and running. Didn't bother to remove the old stuff...