Question about PowerBook Wallstreet II Graphics failure

[beachycove]

My old faithful Wallstreet II (266mhz) has been languishing with a failing hard drive and a failing display (a rectangle of artifacts covering about 1/3 of the screen), so it being the Christmas holidays and me having some time on my hands, I decided today to do some work on the machine today.

I have replaced the 40GB drive with a 30GB that I had on hand, and am currently imaging the old hard drive for installation on the new (transplanted it into a Pismo, booted in FireWire Target mode, and am doing the imaging on a G4). I hope it continues to spin long enough to recover all my old files, but nearly there, so here's hoping.

The graphics problem is a bit more involved. The initial assumption was that the trouble would most likely lie in the data cable (the earlier Wallstreets were notorious for it) , but as I had a parts Wallstreet body here, I started out by transplanting the problematic screen to it, before I started disassembling things more thoroughly, only to discover that the screen and therefore the date cables are just fine. Works perfectly, and as the parts main body is in good shape, problem solved.

There is still, however, the question of the failed machine, where the trouble must lie on the logic board.

So here's my question: is it more or less given in these circumstances that the graphics chip is fried, or is there a simpler solution possible, namely, that there are capacitors to be serviced on the machine? Can failing capacitors cause graphics artifacts? Anyone know?

 

[MattB]

I haven’t really heard of capacitor problems on PowerBook G3s of any type. Does it have a failed PRAM battery? Machines of that age tend to have leaky PRAM batteries and I’ve pulled them out of every machine of that vintage that I own.

 

[beachycove]

I’ll have to take the case apart and examine the board to see if there is any obvious sign of capacitor trouble. I do seem to recall that there are electrolytics in the Wallstreet — there are on the charge board, at any rate — so I will disassemble the thing in the next wee while to see what can be seen. My plan is minimally to recap a charge board, refashion a backup battery (I  have had the requisite six Panasonic VL2330s for ages, just haven’t gotten around to doing anything with them), and transplant the two into the working unit.

My Wallstreet is one of my favourite machines; I relied on it heavily for years, so it would be good to set the machine up properly for another few years of existence. I find it handy, for instance, for downloading floppy images and making physical disks. It makes a great Hotline client in that sense, and for writing, it has an exceptionally good keyboard and a screen as good as any from the period.

On a less hopeful note, that failing drive I posted about crashed with an awful scraping noise before the disk imaging could be completed, so there was a loss there. It was partly duplicated on the Pismo, however, so not a complete disaster. 

 

[Franklinstein]

The VRAM on those is in an unfortunate location under the keyboard and can often be found with crumbs and other junk stuck to the leads and discrete components nearby, and the display header lives near the itty bitty fan, which also can have a bunch of dirt around it. Maybe a good cleaning will sort it out?

Otherwise those logic boards are very very dense and it's difficult to identify where things go. You can try applying pressure to various points to see if that helps (broken solder joints) but I don't have any other suggestions; it's hard to fix those things without just replacing the whole assembly. Maybe if someone posts some component-level diagnostics data somewhere it would be helpful but in the meantime if it's not obviously exploded it's difficult to locate some of the faults on those boards.