As to anti-static precautions that
Vanguard mentioned I always use the wrist-strap when getting close
to electronics.
If the wrist strap is attached to the computer case, make sure to
leave the power cord connected since that's how you get grounded.
Then unplug the 20/24-pin PSU connector to the motherboard to ensure
the 5V standby line isn't still powering the PS-ON circuitry on the
motherboard.
I removed the platter when it turned up that it isn't spinning at
all to look for any indications of the reason for that. However,
I still don't know why the motor (platter removed) wasn't moving at
all in
case of one of the drives and in case of the other was making small
jumpy movements.
The motor is very weak. It has very little torque. It is meant to
keep the platters spinning and sucks in much more current when having
to first start up (which it does slowly and why it sounds like a jet
engine winding up). If the platters don't spin then that surge
current remains high and damages the motor and/or the PCB logic. All
motors suck in more current when they first start turning (and
especially if they don't turn at all) but hard drive motors are very
weak. The motor in your dryer has far more startup torque in
comparison to the weight on the load on that motor than does a hard
drive motor which has very little startup torque.
Since the platters didn't spin, that's why I mentioned that a voltage
regulator chip (about the size of 2 pinheads) probably blew and the
motor will never spin up again. On the PCBs that I've worked on,
there are 3 voltage regulator chips usually at the backside of the PCB
(away from the connector end). They're so small that seeing one blown
out is hard by the naked eye. I saw it by using a magnifying glass
and noticing a bit of shiny substrate where there should've been a
ceramic covering, and then looking very carefully underneath to notice
a carbon mark.
This was a drive with some wanted data (the user, of course, never did
backups) so I found someone still selling the same brand and model of
the really old hard drive (which was pricey by comparison with other
current drives of larger capacity) that *looked* like it had the same
PCB on it. I swapped the PCB, the platters now spun up but I still
couldn't read from it. That's because the logic timing of the PCB has
to be tuned to the hardware drive mechanism. The user still has the
old hard drive but hasn't yet bothered sending it in to a recovery lab
to pay the $1500 because the data really isn't that valuable - yet.
She figures when she wins the lottery then she'll have the data
recovered.
Before dismantling the hard drives, and before they went dead (i.e.,
stopped spinning or for whatever reason they were no longer usable),
did you check how hot they were running when the platters used to
spin? Could be the bearing(s) got worn so there was a lot of heat.
Components expand with heat which can either loosen up or sieze up the
parts. I had one hard drive that go so hot that you could cook eggs
on it (it would actually burn your arm) but it would keep spinning.
When stopped and cooled down, the platters wouldn't spin up until I
gave it a sharp but light rap with a metal hammer (my knuckles were
too soft). That drive actually had a plasticky coating used to
prevent the heads from scrapping on the rust (oxide) on the platter.
Because the drive got so hot, when the platters stopped turning, the
heads would settle on the plasticky surface (I think its the epoxy
they use with the rust to adhere to the substrate). With the high
heat, tiny divits impossible to see would form due to the pressure of
the heads and the high heat making the surface coating more pliant.
The motor was too weak for its torque to get the heads out of the
divits unless I rapped the drive to giggle them out. It was a
portable drive and built for industrial use so the PCB regulators
didn't blow (you'd actually hear some clicking noises as the motor
tried to start turning and then cut off to cool down the motor from
the surge current, and retry this cycle a few times).
It was quite interesting though to find out, thanks to
Vanguard that
there are some low-level formatting marks that need to be lined up.
That is why no one anymore does a true low-level format of a hard
drive because it will render it useless. At one time, hard drives
came with a label of defective spots and users could do a low-level
format and reenter this data. The PCB and heads are too weak to
determine the bad spots reliably and why special equipment is used to
detect and mask them out. At some point maybe around 15 years ago
(I'm guessing), and due to too many users ****ing over their hard
drives with low-level formats, the HD makers decided to pre-low-level
format so now you only do high-level formatting. There was a
transition period, however, where many users still thought they could
do low-level formatting which resulted in dead drives. I recall using
SpinRite back then because it was probably the only tool around (or
that I knew) that could non-destructively low-level format a hard
drive. I wouldn't recommend anything else nowadays, but then
low-level formatting is rarely required (the actuator and heads don't
usually get that much out of alignment), and by then it is often
cheaper to grab the data and replace the drive. Here is a quote from
SpinRite:
"No software of any sort can truly low-level format today's modern
drives. The ability to low-level format hard drives was lost back in
the early 1990's when disc surfaces began incorporating factory
written "embedded servo data". If you have a very old drive that can
truly be low-level reformatted, SpinRite v5.0 will do that for you
(which all v6.0 owners are welcome to download and run anytime). But
this is only possible on very old non-servo based MFM and RLL drives
with capacities up to a few hundred megabytes."
When users speak of formatting a hard drive, they are doing a
high-level, logical format, not a low-level physical format. As I
recall (I haven't used SpinRite in over a decade), it was possible to
realign the heads/actuator if the worst case for one of the heads was
less than half the width of a track out of alignment. You couldn't
manually replace the platters and ever hope to get with that level of
alignment. Read:
http://www.pcguide.com/ref/hdd/geom/formatLow-c.html
http://www.pcguide.com/ref/hdd/geom/formatHigh-c.html
Once IDE showed up which moved the controller onto the PCB on the hard
drive, you couldn't low-level format anymore. With the really old
drives, low-level formating wiped out the defective sectors that the
manufacturer's equipment was best at detecting (and why the drives had
a table printed on it of what the manufacturer found). Read:
http://www.pcguide.com/ref/hdd/geom/formatHigh-c.html
You aren't using those old drives. You only do high-level
formatting - but that is performed on an assembled hard drive. Before
you get it, the manufacturer still has to do the low-level formatting.
You won't be able to disassemble and reassemble a hard drive so that
all the low-level formatting is still usable.
