Can a bad mobo physically damage a HD ?

[- Bob -]

I know a bad mobo can corrupt data... but can it do anything to
physically damage a hard drive? Or, from the other side of the
question - if I find bad sectors on a HD, does that mean the HD went
bad on it's own or could the mobo have caused it ?

Thanks,

 

[Rod Speed]

I know a bad mobo can corrupt data... but can it
do anything to physically damage a hard drive?
Nope.

Or, from the other side of the question - if I find bad sectors
on a HD, does that mean the HD went bad on it's own

Not necessarily, it can be due to running much too hot,
and some HDs cant handle bad power during writes
properly and can produce a few bad sectors that way.

or could the mobo have caused it ?

Nope.

 

[mike]

Nope.

Hmmmm...What if the mobo +12 shorts to the drive +5 or data lines or
.....????

 

[Rod Speed]

Rod Speed wrote

Hmmmm...What if the mobo +12 shorts to the drive +5

Cant happen.

or data lines or ....????

Could certainly get molested by the fairys at the bottom of the garden tho.

 

[Arno Wagner]

I know a bad mobo can corrupt data... but can it do anything to
physically damage a hard drive? Or, from the other side of the
question - if I find bad sectors on a HD, does that mean the HD went
bad on it's own or could the mobo have caused it ?

The mainboard canu usually not cause it. The only component that
could inject enough noise into the power signals is the CPU voltage
regulators. But with a good PSU, you would very likely get a lot
of crashes or complete CPU nonfunction in addition. There is no
way the mainboard can cause bad sectors over the data connection.

Typically bad secors are the HDDs own problem. In some cases a bad
PSU is to blame. If it is only a few bad sectors in the SMART
stats (''reallocated sector count''), it may not indicate a problem
at all.

Arno

 

[Arno Wagner]

Hmmmm...What if the mobo +12 shorts to the drive +5 or data lines or
....????

Then you should get some very suspicuous smoke clouds in addition
to the bad sectors or at least a complete, permanet system failure.

+12V is nowere near the data lines and unless heavily damaged such a
short is not possible. Even if it happened, it would not cause
bad sectors, but a burnt-out interface chip, i.e. the HDD would
become unresponsive.

Arno

 

[Aidan Karley]

I know a bad mobo can corrupt data... but can it do anything to
physically damage a hard drive?

I've seen a motherboard that would reliably fry any IDE device
that was connected to it's secondary IDE channel.
Once we'd become certain that it was the motherboard which was
doing it, it got snapped in two in a vise.

Whether the physical frying of the drive's controller card
induced corruption in the hard drive itself, we didn't try to find out.

 

[Arno Wagner]

I've seen a motherboard that would reliably fry any IDE device
that was connected to it's secondary IDE channel.
Once we'd become certain that it was the motherboard which was
doing it, it got snapped in two in a vise.

Whether the physical frying of the drive's controller card
induced corruption in the hard drive itself, we didn't try to find out.

That would actually be extremely unlikely. Frying the interface is easy.
Just put overvoltage into it. Causing defect sectors is hard.

Arno

 

[zappo]

I've seen a motherboard that would reliably fry any IDE device
that was connected to it's secondary IDE channel.

Frying an IDE drive isnt the same thing as corrupting the data on it.

Once we'd become certain that it was the motherboard which was
doing it, it got snapped in two in a vise.

Whether the physical frying of the drive's controller card induced
corruption in the hard drive itself, we didn't try to find out.

It doesnt.

 

[Aidan Karley]

That would actually be extremely unlikely. Frying the interface is easy.
Just put overvoltage into it. Causing defect sectors is hard.

We just stopped experimenting once we'd got as far as "plug a drive
into this channel" = "****ed drive", then solved the problem by not doing
that. The cost of shoving another mo-bo in form the stripper pile was less
than the hassle of trying to figure out the problem any further.
If I recall, the situation was that we were trying to get a roomful
of working computers ready for a client's training course tomorrow, which
my mate's boss had only told him (the mate) about at 4pm on the day before.
Nice boss, long since sacked. What did I get out of the job? A 6-pack of
beer, a pizza and about 10kg of reference books that client left behind in
the classroom. And a couple of the hard drives that hadn't got ****ed.
All of which reminiscing reminds me to get on with writing the best
man's speech for my mate's wedding.

 

[Folkert Rienstra]

That would actually be extremely unlikely. Frying the interface is easy.
Just put overvoltage into it.

Causing defect sectors is hard.

Very simple indeed if the contoller never finished the writes it started.
Or writes the wrong ECC.

However, detecting them with a fried interface card, who cares. Idjut.