The other day I picked up a Macintosh 512k with a broken monitor for $5
thinking that it'd be a fun project to hack on (yes, another thing to
add to my pile of things to hack on, just what I need :)).
The power/sweep board had a couple of obviously dead capacitors and a
few cracked solder joints, which I fixed up and lo and behold, the
screen came back to life and I did a little dance. There was still a
bit of jitter in the picture from time to time, and jiggling one of the
connectors revealed another dry joint so I powered it down and prepared
to fix the other joint. First point of business, I discharged the CRT.
To the main chassis. This, as I have now discovered, is not what you
are supposed to do to discharge the CRT unless you want to destroy the
logic board.
No matter what certain books and websites say, IMHO you should always
have some kind of current limiting resistor when discharging a CRT.
Otherwise, as you unfortunately discovered, the can be voltages induced
in the strangest placed that will blow semiconductors. I normally use my
EHT probe, it's got a internal resistance of 800M, and will discharge a
CRT in a couple of seconds.
The corerect point to discrhage to in just about all machines/monitors is
the outer aquadag coutaing. After all, what you're discharging is the
capacitor formed by the inner and outer aquadag coatings on th CRT flare.
In quite a number of machines, that outer coating is connected to wire on
the same connector as the CRT base wires. which means it takes a somewhat
circuitous route to chassis. Hence, alas, your problems
I now have a working monitor but a fried logic board;
on powerup, the
normally short boot tone is long and drawn out, as if the machine were
running at a tiny fraction of its normal speed. Which I suppose is
What's odd is that it's running at all..
My first thorught is you've lost some DTAck/ generation, and the
processor is extending every bus cycle until it times out. That would
certainly slow things down a lot.
I've never been inside a Mac 512, but I've worked on a Mac+. The problem
is thst most of the logic is in HAL chips (Hard Array Logic -- mask
programmed gate arrays with the saem architecture as a PAL). And AFAIK
there's no way to read out the 'fuse map' from a HAL (certainly I've
never managed it). So I don't have the logic equations, which makes
troubleshooting, and indeed repair, a little hard.
I thinkl I would start wih the one chip you certainly know, the 68000.
Look to see if the bus cycles (AS/ width, etc) look reasonable.
-tony