Total harddrive failure...options

[dg]

I don't see any reason for that either. Semiconductor damage is likely,
but the heads flat on an air-cushion generated by the spindle spinning.

I agree, and if I was just NOT going to pay anybody for attempted recovery I
would examine the circuit board and try to see if something is visibly
burned. There is a chance, I don't have any guess at probability, that only
1 component that is (possibly near the power input) burned up. Maybe
something as simple as a surface mount resistor-many devices do use
resistors as a cheap current limiting fuse more or less. I have not
repaired hard drives in this manner but I have repaired many other
electronics devices that had only 1 bad component damaged from power
problems, usually a semiconductor of some sort just as Arno Wagner
mentioned. If you replace one burned component and it works, sweet! If you
never fix it, you aren't out much except the data you gambled with-couldn't
have been THAT important if it wasn't backed up.

--Dan

 

[Al Smith]

Another trick i read about, and tried without success, is freezing the drive

for an hour or so, and then putting it quickly in as a slave. They say when
this works you have about 5 minutes to get what you gotta get. The theory
is the freeze shrinks the metel and the stuck spindle is released until
expansion due to friction heat sticks it again.

Avery

A better way of doing it, in my opinion, would be to get a plastic
sandwich bag, fill it with ice cubes, and sit it on top of your
defective hard drive inside your computer case. Wait fifteen
minutes, try turning on your computer and see if the drive spins
up. That way, if it works, the ice will keep it cold for a
sustained length of time. Dry ice would also work, and would be
colder. That's the route I'd go, anyway, rather than the
refrigerator route.

 

[Zvi Netiv]

I agree, and if I was just NOT going to pay anybody for attempted recovery I
would examine the circuit board and try to see if something is visibly
burned. There is a chance, I don't have any guess at probability, that only
1 component that is (possibly near the power input) burned up.

Unlikely, given the vast damage caused to multiple components such as power
supply, mobo, hard drives, and CD drives.

Maybe
something as simple as a surface mount resistor-many devices do use
resistors as a cheap current limiting fuse more or less.

The damage is typical to that caused by lightning. The spike leading edge must
have been very steep (and high) in order to break through that many components
simultaneously, before any protection mechanism could intervene. Surge
protection components are ineffective against such impulses. Most chances are
that the defective component(s) will show no visible sign.

I have not
repaired hard drives in this manner but I have repaired many other
electronics devices that had only 1 bad component damaged from power
problems, usually a semiconductor of some sort just as Arno Wagner
mentioned. If you replace one burned component and it works, sweet! If you
never fix it, you aren't out much except the data you gambled with-couldn't
have been THAT important if it wasn't backed up.

Could be worth trying to replace (swap) the electronics board of the HD.

Regards, Zvi

 

[John Doe]

Unlikely, given the vast damage caused to multiple components
such as power supply, mobo, hard drives, and CD drives.


The damage is typical to that caused by lightning. The spike
leading edge must have been very steep (and high) in order to
break through that many components simultaneously, before any
protection mechanism could intervene. Surge protection
components are ineffective against such impulses. Most chances
are that the defective component(s) will show no visible sign.

Whatever. Reply authors cannot prove or disprove your assertions.




....

 

[Zvi Netiv]

Whatever. Reply authors cannot prove or disprove your assertions.

Just my experience in FMEA, an engineering discipline I was in charge of
sometime during my career (FMEA - Failure Mode and Effect Analysis).

Regards, Zvi

 

[John Doe]

Just my experience in FMEA, an engineering discipline I was in
charge of sometime during my career (FMEA - Failure Mode and
Effect Analysis).

In fact, lightning can easily cause visible damage, you don't know
whether it did or not. Your argument is pointless, especially when
directed at reply authors.

 

[Ron Reaugh]

Power spike THAT bad will have crashed your drive
head bigtime.

Nonsense, there is no electromechanical actuator the sends the head towards
the disk surface save a squib and plastic under the drive.

 

[Folkert Rienstra]

He's already tried that.


That's usually something quite basic, a defective power connector
in the original system. The metal tunnels the pins go into can open
up over time and not make good contact. If that is the case, you
dont need to put the hard drive in another system, just try one of
the other power connectors, like off one of the optical drives etc.

 

[Arno Wagner]

The damage is typical to that caused by lightning. The spike
leading edge must have been very steep (and high) in order to break
through that many components simultaneously, before any protection
mechanism could intervene. Surge protection components are
ineffective against such impulses.

That is untrue. Well designed surge protection mechanisms are faster
than the pulses can push enough energy into other components to damage
them. The principle is that: The surge protection device takes most
of the energy that comes through the line, thereby possible being
destroyed. The protected components are slower to take in the energy
and survive. The amount of energy you can push through, e.g., the
transformer in a PSU, is lmited, since it has to go through components
that can transfer only so much before failing .

Example: A standard transil protector diode can take 100A for 10us
without suffering damage. A standard spark-based surge protector can
take in the range of 10.000A for 10us. A standard metal-oxyde surge
protection resistor can take 100A for 10us. Reaction times for all these
devices are in the nanosecond range. If applied correctly all these
can be used to sucessfully protect a computer from any type of surges
that can come through a power outlet and will not set the house on fire
anyways.

The problem here is that many cheap surge-protection devices are not
well-designed and that many cheap PSUs do not even have them or only
have far too small ones. The spark-based protection device costs
something like 2 USD, fgor example. That is obviously too much for
ceaply designed electronics.

Arno