Alright, I put in a new IRF9Z30 into Q20, and it seemed to make a little difference; smoothed voltage out a little, brought it up to maybe 4.2V. So I reflowed the connections at Q25, Q25, R24, and R25 since they were looking pretty scuzzy, and it seemed to improve things even more dramatically - DISK_+5 is now reading about 4.7V, and pretty steady. (I also rechecked R161, as I was thinking of swapping it for the same rated component, but it was suddenly reading 970k, so
that's obviously not an issue any more...)
So I figured I'd give it a shot. I attached everything there on the bench and it *will* boot to the HDD, but after some labouring. Takes a few tries to get spinning. Once it's spinning though, it seems to operate normally, and restart normally. So that's somewhat promising.
However, it's clearly still not getting enough power. 4.7V isn't enough, so there's a drag somewhere. I just don't know where. Now I'm looking at Q24 or perhaps the OP20 at U7M... somewhere in that SCSI power enable logic circuit (see attached image)... maybe take it all the way back to C25/C26, pull those up check for crap... maybe this is all just parasitic. I just don't know what the voltages are supposed to be anywhere along that chain.
There are other, perhaps (un)related issues:
- there's a static/noise in the speaker, like there's an open ground somewhere or something. If I don't have the HDD connected, the noise becomes a quiet squeal when the floppy icon appears (when the voltage on the Sony IC's drops from 12 down to 4.x). I happened to discover that touching certain pins with a lead makes it louder or change: pin 1 of Q20, pin 2 of Q16, and
especially pin 3 of the OP20 IC right above the part number on the Hybrid. I reflowed the pins on that OP20, but it didn't change anything. Maybe this is simply because I have the board on my bench. I'll see if it changes once I put it in the case. I suppose it could also be residual electrolytic somewhere.
- Trying to load the Battery control panel gives me an error --- "The application "unknown" has unexpectedly quit, because an error of type 2 occurred" (see attached image) --- is that a memory address issue??? The app starts - the little bar shows up for a second - then quits. Every time. The lightning bolt never appears. Other control panels work fine. I fear this may be related to the Hybrid.
At this point, it seems sensible to just to go around and reflow anything suspicious. I also understand that reflowing the components on the Hybrid is recommended, so I guess that's on the menu.
