I'm messing around with some junk harddrives. I figured making one into
a kinetic sculpture of something would be a nice distraction from my
real projects. In that vein I've been taking their cases off etc. all
the usual stuff. 

In playing with the voice coil actuators I've noticed that the movement
of the heads seems to be a velocity linearly perportional to the current
through the coil. The drive obviously uses feedback from the encoded
data on the platters to position the heads accurately. I'd like to be
able to have a PIC chip do some sort of PWM to position the heads and
make pretty patterns. So this means I'll need to just have my PIC
blindly guess and run the heads to the limits every so often to
"recalibrate" the location right?

Just thought I'd run it by you guys, I can't think of anything better,
and I'd rather not get into trying to actually talk to the drive
controller, makes for high-pin count pics == expensive project. :)


Some amusing things I've found too:

One of these old hd's had a socketed SMT flash chip... Looked to be SOIC
style with 50mil lead pitch. Also looked to be very fidgedy and
expensive, all for a 40mb harddrive!

Another one was badly designed to say the least... The voice coil setup
was designed so that nothing was preventing the heads from "falling off"
the edge of the platters! I tried disengaging the head lock, which was
nothing more than a very tiny permenent magnet kinda holding the heads
in the parking position, and giving the unit a sharp torque, and the
heads fell right off the platter, destroying the drive! Even worse was
the drive looked like it was designed for laptops as it was very thin,
half-inch height. I'm sure some users had that sequence of events,
dropping you laptop might just be enough to do it. I later tried putting
the heads back on the platters, and simply applied some 5V to the voice
coils, same result.

I can just imagine some firmware designer cackling with glee as he
implemented a secret "jump off cliff" command that did nothing more than
disable the servo-feedback mechanism.

-- 
pete@petertodd.ca http://www.petertodd.ca
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist