Centris 610 slow or no chime, related to U46

I bought a pretty rough Centris 610 off Fleabay recently, went ahead and recapped it (as you do) and replaced the power supply, since the one it came with just clicked when you pressed the power button. I did have some corrosion on the board from cap leakage, but I've definitely seen worse (any '040 powered PowerBook, for example). I'm having some weird issues. Sometimes it won't chime for a considerable amount of time (upwards of a minute or two), then boots as expected. Sometimes it chimes right away and boots normally. Sometimes it doesn't chime at all and won't boot.

I decided to dive into this, since the full-fat 040 I have on the way to complete this thing's transformation into an A/UX machine is nearly here. I made sure all of the caps are well soldered and in the right direction (I'd expect it not to boot ever if they were reversed, but I know they're not reversed for sure now). Then I started poking around (literally poking, as in pressing on chips while the board had power) and stumbled upon this:


It seems that when (a seemingly random amount of) pressure is applied to the IC at location U46, I can get it to chime.

I attempted to reflow the existing solder joints with hot air (on the theory that one of them was just loose) but that didn't change anything. I removed the chip (carefully) with hot air and some tweezers, checked the traces underneath and verified none are broken, and re-soldered it (I'm not great at soldering SMD components, but I do know I didn't short out any legs in this process). It's worth pointing out that the video was taken after I re-soldered it, and the system did have video (obviously with the ? disk icon, because no drives were connected) at least twice.

What does this IC do? Is it plausible that it's the culprit, or is it merely a symptom? Are replacements available, or is this something I'll need to try and scavenge off a differently-broken board? I don't have an oscilloscope to test the actual IC with, so my options are a bit limited for confirming its inputs/outputs. Anybody have an idea how I can proceed?
 
Update: My skills at SMD soldering continue to not surprise me. On a wild thought, I desoldered that IC and spent a bit more time cleaning up the pads, and tinned them just a hair more than I had originally. Re-soldered with hot air, and now it's happily chiming every single boot. Now the test is "can I get it to install OS 8.1?" at least until the full-fat 040 arrives and I can shove an A/UX image on one of the two SCSI emulators I possess.
 

Forrest

Well-known member
Great job. I see the problem IC had a leaking capacitor on both ends. The old cap juice is nearly invisible, and takes some effort to remove completely.
 
Great job. I see the problem IC had a leaking capacitor on both ends. The old cap juice is nearly invisible, and takes some effort to remove completely.
Actually, on these, Apple went with nice reliable tantalum caps for every other value besides 16v 47uF. The only cap I replaced near the problem IC is the cap between it and the power LED connector towards the PDS slot.
 
Top