On 4/13/15 10:20 AM, tony duell wrote:
So I had the
Ridge 32 running again (when I went to debug the -12V
supply it had started working again, so I'll take that for now, I
suppose). After running for an hour the SMD drive (a Fujitsu 2312) in
it failed. At power up, the drive spins up and when it attempts to load
the heads, there's a loud *clunk* and the drive then quickly spins
down. (If anyone has any bright ideas, let me know...)
There is a Fujitsu 2312
service manual on bitsavers. According to that, once
the spindle is up to speed the heads should move in at high speed until
the outer guard band is detected, then slow down and end up locking on
cylinder 0. Sounds (!) as though your drive is running the heads at high speed
into the inner end stop, realising something is wrong, and shutting down.
Apparently there are status LEDs on the drive, what do they show?
I'll have to
remove the drive and run it where I can observe them all.
It'll be a few days, but I'll let you know.
Since the drive doesn't make head-crash noises (does it?) it's entirely possible
the HDA is fine and the fault is on one of the PCBs.
The drive doesn't appear
to have crashed, aside from the clunk at load
time it sounds fine (and sounded fine at the point it died).
Fujitsu used a number of
ASICs in these drives, but it appers there are still things you can check and
repair. I would start with the power amplifier that drives the positioner (voice
coil), check the transistors there first. Then see what the input to that stage is
doing when it attempts to load the heads (does it seem to be trying to ram
them into the end stop?) and work back from there.
The drive has a transparent
cover so I might actually be able to *watch*
it happen, but it's a matter of getting the drive into a position where
I can run it and observe...
Thanks,
Josh
-tony