Maker Pro
Maker Pro

Sources of info to retrieve data off failing (failed) hard drives.

S

Stellijer

Jan 1, 1970
0
JW said:
Much more likely would be the motor controller on the PCB.


That would probably be the best news since I could probably fix it by swapping
out the PCB; an easy fix.

Could that cause the motor to make a "skreee" sound ever 30-60 seconds or so,
too? It's surely not a head sound since the drive wasn't spinning yet. It's
like the motor's gone crazy.
 
S

Stellijer

Jan 1, 1970
0
Yes... cute.

However I'm trying to get as much information to avoid exactly that problem.

It's not that I'm EVER going to take this drive to a data-recovery company,
though - this is NOT an example in the article where there's a ton of valuable
information there. It's personally valuable to me to a degree but not to the
tune of $1000 or more per drive.
 
J

James Sweet

Jan 1, 1970
0
Stellijer said:
(plus the


That would probably be the best news since I could probably fix it by swapping
out the PCB; an easy fix.

Could that cause the motor to make a "skreee" sound ever 30-60 seconds or so,
too? It's surely not a head sound since the drive wasn't spinning yet. It's
like the motor's gone crazy.

That sounds like the bearings were failing, which probably overheated the
motor driver and burned out one channel. You could probably just replace the
motor driver IC, or swap boards.
 
J

James Sweet

Jan 1, 1970
0
JW said:
Have you tried spinning it in your hand back and forth on the plane of disk
rotation right after applying power? This method used to work for me, if it
was a "stiction" problem.

I haven't seen a drive suffer from stiction in probably 10 years, things
have changed since then, platters and heads are much different now.
 
B

BOB URZ

Jan 1, 1970
0
Stellijer said:
Anyone know of any great web sites or discussions boards for reviving hard
drives which have failed fir data retreival?

Well, i think its pretty much been established your motor is not
spinning up. If your adventurous, you might try to ID the motor platter
control chip on the PB board. Inspect it (and the whole board) for
any suspicious solder connections. Try to find the chip's pin out on line
and verify all the voltages are to it. Maybe you can try to hot-wire it
to jump-start it. You need to find the data sheet (if you can at all) and
see what fun you can do. There could even be a low value series resistor
that has opened. Probably a wild goose chance, but with a scope, DVM
and a pin out, you could amuse yourself for awhile.

Bob
 
A

Alan Peterman

Jan 1, 1970
0
I haven't seen a drive suffer from stiction in probably 10 years, things
have changed since then, platters and heads are much different now.

One of my clients got 3 systems in 1999 with Samsung SV0432A drives in them.
ALL three of those drives (made in 1999) failed in 2001-2002 due to stiction
problems. And on each drive I managed to spin them up by hand and recover the
data to a new drive. I'm still using one of them as a paperweight on my desk!
 
C

Chaos Master

Jan 1, 1970
0
Stellijer([email protected]) spoke said:
Tried that. No luck.

From a Brazilian hardware/electronics site:

If the problem is on the heads/motors, try opening the HD on a clean room (as
clean as possible), and "move" the heads to the right place, run IDE auto-detect
from BIOS, and cross fingers. This will not last too much, but it will be enough
to copy some data. (disclaimer: I've never tried that. If the HD quits working,
it was "doomed" already)

If the problem is on the logic part, try getting the circuit board from an
IDENTICAL hard drive. (Sometimes this works; I've seen this done)
 
S

Sam Goldwasser

Jan 1, 1970
0
Chaos Master said:
From a Brazilian hardware/electronics site:

If the problem is on the heads/motors, try opening the HD on a clean room (as
clean as possible), and "move" the heads to the right place, run IDE auto-detect
from BIOS, and cross fingers. This will not last too much, but it will be enough
to copy some data. (disclaimer: I've never tried that. If the HD quits working,
it was "doomed" already)

"Move" the heads to the right place??? There are several thousand
tracks on a modern hard drive. Does he expect you to guess at the
track location? Furthermore, any access sequence will start via the FAT
and directory structure (on unix, the Inodes or whatever), not the "right
place". :)

--- sam | Sci.Electronics.Repair FAQ Home Page: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Sam's Laser FAQ: http://www.repairfaq.org/sam/lasersam.htm
| Mirror Site Info: http://www.repairfaq.org/REPAIR/F_mirror.html

Important: The email address in this message header may no longer work. To
contact me, please use the following email address: [email protected].

The Feedback Form at repairfaq.org is temporarily broken. Thanks.
 
C

Chaos Master

Jan 1, 1970
0
Sam said:
"Move" the heads to the right place??? There are several thousand
tracks on a modern hard drive. Does he expect you to guess at the
track location? Furthermore, any access sequence will start via the FAT
and directory structure (on unix, the Inodes or whatever), not the "right
place". :)

I don't remember the exact text from the site. The site said about trying to
"force" movement on the drive. (I.e. hitting the drive' side with something
light, when it's trying to start up)
 
S

Stellijer

Jan 1, 1970
0
Chaos Master said:
From a Brazilian hardware/electronics site:

If the problem is on the heads/motors, try opening the HD on a clean room (as
clean as possible), and "move" the heads to the right place, run IDE auto-detect
from BIOS, and cross fingers. This will not last too much, but it will be enough
to copy some data. (disclaimer: I've never tried that. If the HD quits working,
it was "doomed" already)

If the problem is on the logic part, try getting the circuit board from an
IDENTICAL hard drive. (Sometimes this works; I've seen this done)


How would moving the heads help when the motor is what's not spinning up? I'm
curious how that could help.

As for a new PCB, I'm trying to obtain one. It surely would be the least risky
thing to try at the moment. I'm not sure if it's the problem or it's the motor
but I'd rather spend $50 on a used drive (which I still might be able to resell)
and try the circuit board first before trying to open the case.
 
S

Stellijer

Jan 1, 1970
0
Chaos Master said:
I don't remember the exact text from the site. The site said about trying to
"force" movement on the drive. (I.e. hitting the drive' side with something
light, when it's trying to start up)


Indeed. However twisting the hard drive as quickly as possible by hand and
tapping it lightly have had little effect. Such actions make the drive "jerk" a
little but not actually start.

It's probably either a dead motor or in the motor control. I'm not sure if
spinning the platters by hand at a decent speed would help at all but as I said
before, I'd rather try the new controller board first.
 
S

Stellijer

Jan 1, 1970
0
James Sweet said:
I haven't seen a drive suffer from stiction in probably 10 years, things
have changed since then, platters and heads are much different now.


Indeed. The platters in this drive seem to spin.

What do you believe might be wrong, given that it jerks a little as if trying to
start? Bad motor or bad controller?
 
S

Stellijer

Jan 1, 1970
0
Alan Peterman said:
One of my clients got 3 systems in 1999 with Samsung SV0432A drives in them.
ALL three of those drives (made in 1999) failed in 2001-2002 due to stiction
problems. And on each drive I managed to spin them up by hand and recover the
data to a new drive. I'm still using one of them as a paperweight on my desk!


What were the symptoms before the stiction occured? Would they jerk a little
trying to start?

Twisting this drive isn't helping. I'm not sure if manually spinning the
platter at a decent speed would help jump start it. I'm trying to find someone
who's had the same "occasional jerking" symptom that I have.
 
S

Stellijer

Jan 1, 1970
0
James Sweet said:
That sounds like the bearings were failing, which probably overheated the
motor driver and burned out one channel. You could probably just replace the
motor driver IC, or swap boards.


You feel that burned out a motor component ITSELF, or a component on the circuit
board? I'm not sure which one you mean. It would be INFINITELY better if a new
controller board would help get this started, without having to open the case.

How would you ever replace an IC on these boards? They look like you'd have to
solder them with microscoping equipment! Also, I would have no idea where to
get that same IC or know WHICH is the motor driver IC.
 
S

Stellijer

Jan 1, 1970
0
<see interspersed>


BOB URZ said:
Well, i think its pretty much been established your motor is not
spinning up.

Indeed. I just need to ascertain WHY it's not spinning up; bad controller or
bad motor. And if bad motor, could I spin it up manually one time...

If your adventurous, you might try to ID the motor platter
control chip on the PB board. Inspect it (and the whole board) for
any suspicious solder connections.

I'm adventurous but I try to pick and choose my battles! :)

I'm not sure WHICH would be the right chip, anyway. Sounds a lot easier to try
to obtain a used idential series drive off ebay.

Try to find the chip's pin out on line
and verify all the voltages are to it. Maybe you can try to hot-wire it
to jump-start it. You need to find the data sheet (if you can at all) and
see what fun you can do.

Indeed... sounds like a pain. I'm not sure how I would solder things that small
anyway.

There could even be a low value series resistor
that has opened. Probably a wild goose chance, but with a scope, DVM
and a pin out, you could amuse yourself for awhile.

So... you are of the belief that it's probably the motor CONTROLLER as opposed
to the motor itself?

I would hope it's the controller since it's external. I'm afraid of damage if I
have to replace the motor itself. It looks like you have to unfasten all the
platters and everything to get at the motor which exposes the platters to more
risk than I care for.
 
D

DaveC

Jan 1, 1970
0
I would hope it's the controller since it's external. I'm afraid of damage
if I have to replace the motor itself. It looks like you have to unfasten
all the platters and everything to get at the motor which exposes the
platters to more risk than I care for.

It's much more likely that the driver IC (on the PCB) or something else on
the PCB has failed rather than the motor. Usually the semiconductors are the
more-fragile component rather than the motor windings, in the motor drive
circuit, in general.

I've had good luck with changing the PCB on a drive (IBM).

It seems that contributors have covered all your options. You need to make a
chioce, now, and it seems that your only one left (reasonably) is to find a
similar drive and swap PCBs. (I don't believe it has to be the *exact* drive;
a series of HDAs (the mechanism) usually comes in several capacaties which
used the same PCB. A PCB from any of these should work.)

There is *no* possibility of replacing the motor if you don't have a real
clean-room environment. At least, I'd not consider it.

Good luck,
 
S

Stellijer

Jan 1, 1970
0
DaveC said:
It's much more likely that the driver IC (on the PCB) or something else on
the PCB has failed rather than the motor. Usually the semiconductors are the
more-fragile component rather than the motor windings, in the motor drive
circuit, in general.

I've had good luck with changing the PCB on a drive (IBM).

It seems that contributors have covered all your options. You need to make a
chioce, now, and it seems that your only one left (reasonably) is to find a
similar drive and swap PCBs. (I don't believe it has to be the *exact* drive;
a series of HDAs (the mechanism) usually comes in several capacaties which
used the same PCB. A PCB from any of these should work.)

There is *no* possibility of replacing the motor if you don't have a real
clean-room environment. At least, I'd not consider it.


Yes, I believe the options have been covered; I'm just trying to gain as much
knowledge as possible.

I'm also seeing if anyone has experienced the "jerking" situation where there
are these little pulses from the motor. That would go far towards IDing the
problem.

As for choices; my next avenue of action is clear, unless there is any more
information which comes out in the mean time. That's to obtain a used drive and
try swapping out the PCB. It's surely the least invasive solution. I'm keeping
my eye out on eBay. There are plenty of drives of that MODEL, but finding the
same SERIES may take a little while. As for series - I know two 80GB drives of
WD make had notably different PCBs. MAYBE some would be interchangable but not
knowing, my best option would be to hold out for the same series. I would
imagine one would appear in no more than a few weeks.

Indeed, swapping the motor out would be the LAST resort. Who knows; if it's the
ONLY thing left to me at some future point, I may even find someone who has a
clean room in his/her office I may get to use at some point. Anything is
possible.

In the mean time I'll keep learning as much as possible; especially when it
comes to the other drive I spoke of.

I'm just a little surprised there aren't more documents or websites that people
have found valuable for this kind of information.

Thanks for the added suggestions.
 
I

Isaac Wingfield

Jan 1, 1970
0
--snip--
How would you ever replace an IC on these boards? They look like you'd have
to solder them with microscoping equipment! Also, I would have no idea where to
get that same IC or know WHICH is the motor driver IC.

Getting the IC may be a problem, but replacing it with "ordinary" tools
is only moderately difficult (assuming you can *see* the pins; if it's a
ball-grid job, you're out of luck):

1) With a sharp hobby knife, cut through all the legs of the bad chip,
right next to the body of the chip. Take care not to cut any traces on
the board. The body may have a bit of stuff on the bottom to attach it;
you may have to lever it up a bit to get it off. MAKE SURE you've cut
all the legs before you do this.

2) Using a "solder sucker", you can melt the solder and "suck" the
individual pins off the board. Work quickly. Make sure none just move
somewhere else and get overlooked.

3)If the pads still have "bumps" of solder, use some solder wick to
clean things. Clean the flux off with acetone. Put a little more solder
on two opposite corner pads.

4) Orient the replacement chip (you *did* notice which way the broken
one went, right?) and *making sure all the pins are directly above their
appropriate pads*, tack down those two corner pads.

5) Re-check to make sure that no leg is bridging adjacent pads.

6) Using fine solder, just run a small iron around the chip, covering
all pins with solder. Make sure every pin is attached to its pad. DO NOT
WORRY about solder bridges at this time.Working quickly is more
important.

7) Once again use the solder wick, this time to remove all solder
bridges. Use a loupe, if you have one.

8) After you're confident that all legs are attached and there are no
bridges, clean off any remaining flux with acetone.

Isaac
 
I

Isaac Wingfield

Jan 1, 1970
0
Stellijer said:
How would moving the heads help when the motor is what's not spinning up?
I'm curious how that could help.

A while back, some manufacturers (Sony was one) had compatibility
problems between disk material and head material -- sometimes, when the
drive was turned off for a while, they would stick together (really,
cold-weld, I think). The motor starting torque was insufficient to break
the bond, but a good whack could do it.

I once had such a disk, but even whacking its did no good. Finally I
really gave it a good shot. It started spinning but it was making an
awful racket. Now that it was obviously no good, I took it apart, to
find the offending head *still attached to the disk*. My final whack had
ripped it right off its supporting arm.

Isaac
 
S

Sunny

Jan 1, 1970
0
Isaac said:
--snip--




Getting the IC may be a problem, but replacing it with "ordinary" tools
is only moderately difficult (assuming you can *see* the pins; if it's a
ball-grid job, you're out of luck):

1) With a sharp hobby knife, cut through all the legs of the bad chip,
right next to the body of the chip. Take care not to cut any traces on
the board. The body may have a bit of stuff on the bottom to attach it;
you may have to lever it up a bit to get it off. MAKE SURE you've cut
all the legs before you do this.

Carve up the chip with a hobby knife? Ugh!

The proper way to remove surface mount chips is with chipquik and a
strip of stainless steel like the ones dentists wrap around errant
teeth. Working on one side of the chip at a time, flux the pins
generously, set the soldering station to 350F, coat the pins with
chipquik and keep it moving while very gently sliding the stainless
strip between board and pins. Let it cool, pull the strip out, and
repeat for remaining side(s).

This technique allows you to remove chips intact, which is often the
only way to obtain a replacement, and eliminates the risk of damage to
the board. It also leaves the site nice and clean.
 
Top