I recently acquired two 12" iBook G3s, a 500Mhz, translucent IceBook and the first Opaque, 700Mhz (16MB VRAM model) SnowBook. They were both marked as "untested, good for parts only" and pretty scrappy looking in the photos, covered with stickers and grime, but I cleaned them up and they almost look as good as new. This is a testament to the polycarbonate plastic Apple used on the white G3 iBooks, it is tough and ages really well - I always found the iBook G4s to be made of a more crack-prone material and let's not forget the pre-Unibody MacBooks which all happened to crack and split pretty easily, even when they were brand new. On both models, the screens are bright, the hinges are tight and internally they are both in great condition. The 500Mhz model exhibited no problems whatsoever and is a perfect collectable. The previous owner just had no way of powering it, so sold as-is for cheap and I got lucky. The included battery (G4 gray-foot model) also reported only 2 cycles and a charge capacity of 4400+mAh, It holds it's charge for nearly 6 hours of use. The 700Mhz model however has been a real head scratcher... The saga goes a little something like this... (it's long, so my apologies in advance) On first boot, we got a chime, but no display. I expected this might be the case because of the seller's vague description and the well documented logic board recall for the GPU solder failing on this series. My assessment was confirmed by gently squeezing the case with my fingers where the HDD sits (left of the trackpad) and powering on while keeping the pressure applied. I was greeted with a bright display and a flashing question mark. I dismantled and installed a pre-loaded hard drive (40GB HDD taken from my TiBook, which now has an SSD). While I had all the layers of the iBook dismantled, I completely removed the logic board from the frame for a thorough inspection, clean and thermal pad replacement. I also noticed a tiny amount of corrosive build up near the PMU capacitor and cleaned it off with IPA and a toothbrush. To alleviate the GPU failure, I replaced the thermal pad on the GPU with a 2mm (thicker) pad and also packed an extra pad between the EMI shield and the outer (bottom) casing, essentially using the flexibility of the plastic bottom casing to apply pressure from below. All back together for Boot time and we have success! The iBook booted A OK, full GPU usage, Tiger is installed and all seems A OK. I used the iBook to do some writing that night, then shut it down and powered everything off. The next morning I went to switch on the iBook to collect the notes I had taken and I was stumped when the power button would not respond at all. I tried resetting the PMU as directed by Apple's PMU reset support page by removing the battery and AC power, then holding down Ctrl-Opt-Shift and pressing power, wait 5 seconds, then plug in AC and press power. This didn't do anything. I tried combinations of this - battery installed with no AC, AC with no battery, hold key combo for 5 seconds, 10 seconds, 30 seconds, wait longer, etc etc and all appeared to show no sign of life... Reading about PMU corruption on the iBook, I followed the advice of disconnecting the battery and AC and leaving it for a few days to discharge completely. This technique resulted in being able to boot again a couple of days later, which to me, confirmed it was that suspect PMU capacitor at C91 causing a problem. I dismantled it again and heated up the soldering iron to remove the 5.5v 0.33uF H-type terminal coin capacitor and replaced with a new one (I had spares from doing this on my Blueberry iBook - min purchase was 10x from HK). Happy with the quick replacement I re-assembled and tried to boot. However, we had nothing again. I dismantled again and found a resistor near the PMU capacitor had been victim to the slight corrosion causing the solder joint at one end to have been eaten off by the leaky cap. Not really thinking it through, I removed the resistor and just jumped it with a solder blob... In the process I lost the speck of an SMD resistor in the carpet (forever) and my record keeping photos didn't make it clear what the rating on the resistor was... I re-assembled and to no surprise, it refused to boot. Realizing I had messed up and the resistance is likely a requirement of the circuit to boot, I asked @LightBulbFun if he could help me figure out which resistor it could have been. Using Dez's iBook schematics, we figured it was a 470k resistor, and using this handy SMD decoder, I found this would require an SMD with the marking "474". I hunted around on the board and the only 474 SMD I could find was at the Broadcom Ethernet controller.. So, I set about de-soldering this and re-soldering it at the PMU location. All looked good and tested as expected with the meter, but still no boot... damn. I went to my local electronics store to buy an 8pk of 470K long wire-type metal resistors and replaced the re-used SMD with the wire resistor at the PMU circuit and then again at the broadcom controller location. At this point in time I had the logic board and DC-in board standalone and just connected power to the DC-in board and attempted to jump-start by shorting the power-on pads/pins on the board while headphones were connected to the audio out. There was a chime!! I pulled the power out and promptly started to re-assemble the machine. Once it was all back together again, I plugged it in and feeling confident I had repaired the issue, I hit the power button and.... nothing. No life. Feeling defeated on this board, I went onto PowerBook Medic and found a replacement 700Mhz logic board for $27 plus shipping and put my order in. When the board arrived, I installed it only to find it wasn't an exact match, but the model prior, which used the 16MB Mobility Radeon  and not the 16MB Mobility Radeon 7500. I'm okay with that, and at least the earlier board didn't exhibit the GPU failure issues which plagued those with the 7500 GPUs. Installation of the replacement board went smooth. I noticed (even slighter) corrosion on the PMU capacitor on the replacement board, just a tiny build up of rust on the cell itself, so I proactively soldered the cap off and installed a new one before re-assembling. All back together and it powered on A OK. The replacement board worked and everything is humming along OK. It also seems to run substantially cooler than the board with the Mobility Radeon 7500. This might be due to the lower spec 7000 producing less heat, or it could have been that the poor GPU solder joints on the 7500 model were producing more heat than they should have. Happy with the replacement, I shut the iBook off and went on to something else. When I went to boot it next... Nothing. no life. WTF!! So back to the drawing board... If the replacement logic board wasn't fixing the non-boot issue, it had to be something else... Throughout all of the trial and error, I had tried two different iBook 12" batteries and 4 different power adapters of both 45W and 65W varieties. Figuring the DC-in board must be the culprit, I found one on TheBookyard.com for cheap and had it shipped from the UK. One thing I found with the existing DC-in board is that it would cause the power adapter light to be somewhere between orange and green (like a lighter orange) and then switch to the usual deep orange (amber) when charging the battery, but it was never just green as you'd expect to see when no battery is installed. While waiting for the replacement DC-in board, I found the iBook would boot up after a day or two of no power, then the moment I shut down it would refuse to boot again. Reboot/Restart however seemed to be fine. When the replacement DC-in board arrived I promptly installed it and found the power adapter would show green as expected. Because of this, I was fairly confident that there must have been a resistance short or something on the prior DC-in board causing the PMU to crash and fail to boot until the main board completely drained of power. All seemed to be OK with the replacement DC-in board and logic board. It handled the first shutdown on battery power OK. It booted up again fine, then I tried with AC and the non-boot issue returned. I let it sit for an hour and tried to boot up and all was okay again. I repeated the shut down on battery and again it was happy, so I shut down a second time and left it for the night. The next day, It refused to switch on after being left overnight with the battery installed. I disconnected the battery and let it sit for an hour and no good. I let it sit for the day and then tried powering on OK again.. So... The TL;DR is the PMU seems to just crash on shut down, even after replacing the logic board, the PMU capacitor, the DC-in board, the battery and the power adapter.. Occasionally, it won't crash if I shut down from battery power, but I can't store it in shutdown state with the battery installed or it crashes the PMU (again). Any thoughts on how to resolve this short of trying another logic board? Are the MAXIM PMU ICs prone to regular corruption like this? I don't recall having this issue with my 800Mhz 12" iBook G3 model back in the day. I love these little iBooks, but damn.. this has been a real frustration. I am just hoping someone out there can say; "Yes, there was an issue with the PMU on the iBook G3, this is how to properly fix it...." The only reliable option seems to be to just leave it in sleep and never shut it off!