C62 is a 150uf 35v with its negative foot on the -12v rails but it goes to... dunno
regulation feedback?
C63 is a 100uf 35v on the -12v rail
ax man had these Sprague 502D 150uf 35v caps with a date stamp of 1988 but searching for a data sheet brings up catalogs from the 70s on bitsavers
you know what they're probably fine
i expect to just run it on a bluescsi, unclear what the -+12v will even do on this machine
ethernet phy? serial? audio amp?
weird
one meters at 184pf and the other three up over 200pf
why kill it when you can overkill it
salvage replacement caps are still a tad bigger than the originals
as they say on roadkill:
don't get it right, just get it running 🫡
beware, the fan sandwiches in here
think i made it work
caps soldered in, and checked for shorts
think i can call this section of the board done
its a shiny work of art 🥹
time for bed, guess that's where i leave it for the week
okay C19 is a 27uf 35v that's related to the board with a Mitsubishi M51995P Switching Regulator Controller
indulging in some reverse engineering, i'm fairly certain it corresponds to Cvcc in the datasheet
critical to regulator startup, its probably not going to chooch without it
the value doesn't seem hugely critical, a 15uf + 10uf from '87/'88 gets us close enough 🫡
i have a match for C21 on the card, a 3.3uf 50v, which is involved in the current limiting circuit
more capacitor shenanigans for C67, another 27uf 35v on a card with... a sample and hold chip?
dunno what that does, the basic operation of a switcher is easy to understand, but how they get multiple positive and negative rails out of it still escapes me
with #MARCHintosh coming to an end i'm eager to find out if i'm wasting time and money on a lost cause
i scrub this with IPA a bunch and try to solder it, this doesn't go well
IPA doesn't seem to clean up the capacitor juice very well, it seems to make solder not stick and turns in to a brown sludge, it takes a few rounds of hot flux to clear it enough to get a connection and puts off that fishy stink 🐟
okay time hook it to a sadly insufficient 10W 12v halogen load and some instrumentation and SEND IT!!
lets get some use out of my oscilloscope
its outputting all the right voltages, yellow trace is +5v and blue trace is +12v, we have a ~90khz switching rate and ~150mv ripple on the 5v which i believe is typical for a computer supply
just gonna let it bake for a while
it's almost time to play everyone's favorite game:
WILL IT BONG??
GEM is truly truly outrageous
48kRAM but with parity