iBook G4 video chip problem, can it be whipped?

[macclassic]

I got drawn into trying to fix my nephew's iBook G4 recently, and perhaps rashly bought a new HD first and formatted it and loaded the OS with Firewire from a working iBook..... and thought I'd fixed the no boot problem by re-soldering pins 1 & 28 on the so called V reg chip (plus three very dry pins on the power adapter board) and got the chime and the spinning gear followed by a bluescreen but at which it stopped.

Repeated attempts to get this machine to boot. using every keystroke under the sun, then gradually go downhill with time, as if as the chip warms up a connection gets worse, so I'm thinking of trying to re-solder the remaining pins once I work out how to make a shield to let me to solder just one pin at a time, with a plan B of taking the chip off the board splaying the pins out to make them easier to solder and running wires between the chip and the pads on the board!

But my question really is why does this chip prevent the machine from booting?, and can anyone give me a breakdown of the bootup sequence in OSX ? as I really want to win this one having stripped the thing down about three times.

 

[ClassicHasClass]

Try booting it in verbose mode. Maybe the problem is software. Hold down Cmd-V immediately after you turn it on.

 

[macclassic]

I got into Verbose once or twice but don't think it got to the end without freezing.

I also checked the Hard Disk with Disk Utilities from another iBook with Firewire and it found no problem - but I've just had a thought!! is it possible to set the mounted Firewire Hard Disk as a startup disk on the working iBook just to prove that it can boot.

 

[Byrd]

Found the same as you - resoldering the faulty vreg chip works great ... for about 30 minutes, then it just gets worse and worse. It can't be whipped. Have you tried scraping back the PCB traces to get a better solder contact - recalling the main culprit was one of the legs on the corner.

JB

 

[macclassic]

THANKS, it's good to have ones experience of a problem confirmed, you feel less on your own, and yes I did scrape the traces back on the corner pins and thought I'd done a good job considering the scale of the pins.

I'm seriously considering de-soldering the chip from the board to site it elsewhere and to bend the row of pins alternately up and down to create enough space to solder wires onto them and back to the traces, maybe also in a staggered fashion.

Did your soldering do the job?

 

[Byrd]

There is little to lose, so I'd give it a go Spent ages too fixing it up but suspect it's not just the vreg chip delamination but other components and the multi-layered PCB itself, which has developed small stress freactures over time. Funnily enough I've come across a few non-Apple laptops with similar issues lately (when you flex the board, it boots).

My soldering did result in a 100% working iBook - for 30 minutes. Then 15, then 10. Repeat next day for same result.

 

[macclassic]

I'll have a go anyway, but if only there was a way of locating stress fractures on PCB's whilst flexing them under power, like a super sensitive infra red camera that might show the odd hot spot, or by some other means which would indicate a point of electrical disturbance?

I think this problem could be caused by people passing iBook's around held just by one corner. 8-o

 

[Byrd]

Not worth the effort when you could probably pickup a "faulty" G4 iBook for < $20. It seems the longer the issue is left unrepaired, the worse it becomes.

I don't like using a heat gun on PCBs - it rarely works. The heat spread is too widespread, and the BGA (balls of solder) when melted make the component "slippery" and it moves around - put it out by the tiniest degree, and you've stuffed it even more, beyond repair.

 

[macclassic]

Yes, I could end up doing that after one more go, as I like fixing things but an understanding of how something works is needed to diagnose and pinpoint what's gone wrong with something, and you are effectively locked out from so much with modern PCB's by dint of scale, and, if you read the manufacturers fault finding steps which almost instantly fall to the replacement of whole items then the reality of "one trip manufacturing techniques" and the haunting phrase "mean time before failure" starts to spring to mind.