Strange Screws

[Folkert Rienstra]

Examine that filter carefully

Do it yourself and find out that you are talking through your ass.

and you will find that its primary function is to filter the tiny
amount of air moving through the pressure-equalization hole.

Nope. Older IBM drives didn't even have filters for that. Just a long-
ish serpentine duct cut out in the exterior case and covered with self
adhesive plastic, with a small hole at both ends, one external, one internal.

and that there is no mechanism by which all or any significant portion
of the air circulating inside the capsule can be made to pass through it.

It doesn't need to.

The filter he talks about is sitting in the airstream in a corner of the
drive. Particals are swept along the sides of the platter cavity and
driven into the corner where the filter sits, where they get caught.
All it needs is enough air to get through for the particals not to take
another route.

It dies as soon as something hard enough to scratch the platter or head and
small enough to get wedged between them finds its way into that space.

Which should be microscopic.

In the real world people have tried this, and the drives typically died in
anywhere from a few hours to a few weeks.

Guess he's not in the real world then. You of course are.
You of course are of the school of hard knocks.
Unfortunately one went to the head.

Now look how silly you are.

 

[Folkert Rienstra]

<snip>

the Math-CS dept at my alma-mater had pretty poster depicting the magnified
size of various particles hair, dust, skin cell and smoke particles next to the
disk heads and the cushion of air that head float on and if i remember correctly

the smoke particle would barely squeeze between the head and platter

So it won't go there in the first place.
The hurricane that rages there will blow it away before it even reaches there.

guessing you were dust lucky with the drive surgery

Pun intended?

 

[Folkert Rienstra]

"J. Clarke" (e-mail address removed)> wrote in message news:[email protected]...

Stan Blazejewski wrote:

[and snip again]

<snip>

the Math-CS dept at my alma-mater had pretty poster depicting the
magnified size of various particles hair, dust, skin cell and smoke
particles next to the disk heads and the cushion of air that head float
on and if i remember correctly the smoke particle would barely squeeze
between the head and platter

guessing you were dust lucky with the drive surgery

Not me, the only time I've opened drives either they were already dead or
they were old drives being used as show-and-tells when teaching a class.
Never expected them to actually run afterwards. But if you will google
the archive for this newsgroup I think you'll find some reports from people

who have done this to see how long the drive would run.

Would run opened, continuously all the time.

One thing that is really bad news is fingerprints. I've seen a fingerprint
rip the head off and toss it across the room (somebody decided to power
up a show-and-tell after it had been passed around the classroom).

If it's _thicker_ than the gap and not stuck down and not too massive it
just gets pushed aside.

At low speed, which means it must be in the landing zone.
Anywhere else it is catapulted off the platter long before
the platters reach full speed and before the heads are released.

If it's brittle though it may shatter when it hits the head

It has to hit it first for which it has to be stuck solid and then released
at high velocity and hit the head (small chance) or not released at all and
sitting in the landing zone where forces are less and is swept away at low
speed or is practically glued to the surface somewhere else and even sticks
at full velocity and the head hits it eventually during a seek or R/W.
Now what are the chances of that, (bar finger prints).

and make smaller particles.

Which in that case fly away and end up in the filter.

If it's hard and massive enough then it can chip or deform the head.

A chance of one in millions.

If it's an insect you get ichor on the platter and the result is similar to
fingerprints.

An insect will never make it to the spinning platter.

Bear in mind that that poster dates from the days of
removable disk packs, when drives were _not_ sealed.

Disk packs are still here today (relatively).

Few minicomputers were installed in special rooms and even mainframe shops
generally didn't have laminar-flow positive pressure clean rooms with airlocks.

Nonetheless a standard operation was
for some guy in a suit and tie (if it was a corporate shop) or tie-dyed tee
shirt and jeans and hair down to his ankles (if it was a university shop)
to change disk packs and they ran a good long time despite such treatment.
That's the reason for the posters, to remind people of why that area was
supposed to be kept clean.

And this little anecdote is to show what?

 

[Arno Wagner]

Perhaps the drive already *is* dead.

Don't overestimate clean rooms - they contain 100 particles per cubic
meter as opposed to an "average" room containing 600 particles. A
"clean" "average" room will contain far less than the 600 particles.

For what it's worth, I've had a drive running non-stop for over a week
without its cover (platters exposed) and haven't had any hiccups. This
hype about "clean rooms" is a load of drivel.

Interessting.

Arno

 

[Folkert Rienstra]

Every single time I've ever had a hard drive clicking it was caused by a
failure of the drive,

So either you have a pathetically inadequate
small sample or you are killing all your drives.

I've never even heard of it

So you obviously should refrain from commen-
ting as if you are the resident expert on this.

caused by those other issues, with the exception being
a couple of early very hot running 10K rpm drives.

As if that can't happen to IDE drives.

Bad drive is 99% the reason.

In your case.
You are known as a 'pathetically inadequate sample'.

 

[James Sweet]

Interessting.

Interesting but the drive is running on borrowed time. Perhaps you
should store all your critical data on it and see how long it continues
to operate like that.

 

[Folkert Rienstra]

The exact same thing happened to me. I was copying stuff off one of my old
Seagate HDD, and there is one file that XP can't read, saying ECC error. I
ran Seatools <http://www.seagate.com/support/seatools/> on the drive and it
identified 2 bad sectors with full diagnostic. I was able to get the file
off the drive by having Seatools force a remapping of the bad sectors.

The remapped sectors are zeroed, so you're getting the file damaged, but it
is better than not getting anything at all. The good thing is Seatools
tries to identify and tell you which file is affected (although in short 8.3
name only), so you can decide if you want to risk it or not.

As for whether it will get worse, it depends on what caused the error. If
it was just a transient glitch that caused the drive to make a bad write, it
could be that it will develop no more error afterwards. If, say, the drive
electronics is failing, you'll see more and more bad sectors (thus more
clicking when you access previously-okay files).

HDDs are cheap enough nowadays that I wouldn't risk my data on such a
drive, but YMMV.

Strange how you have no such concerns with risking your data on a new drive
without checking your powersupply/supply of power first.

 

[Stephen Lee -- post replies please]

It only clicks if I try to access the bad partition, and even then not
always . I can read the good partitions, but I'm told the clicking
will get worse.

The exact same thing happened to me. I was copying stuff off one of my old
Seagate HDD, and there is one file that XP can't read, saying ECC error. I
ran Seatools <http://www.seagate.com/support/seatools/> on the drive and it
identified 2 bad sectors with full diagnostic. I was able to get the file
off the drive by having Seatools force a remapping of the bad sectors.

The remapped sectors are zeroed, so you're getting the file damaged, but it
is better than not getting anything at all. The good thing is Seatools
tries to identify and tell you which file is affected (although in short 8.3
name only), so you can decide if you want to risk it or not.

As for whether it will get worse, it depends on what caused the error. If
it was just a transient glitch that caused the drive to make a bad write, it
could be that it will develop no more error afterwards. If, say, the drive
electronics is failing, you'll see more and more bad sectors (thus more
clicking when you access previously-okay files). HDDs are cheap enough
nowadays that I wouldn't risk my data on such a drive, but YMMV.

Stephen

 

[srwolcott]

The Seagate external hard drives are in a case that holds th
2.5" drive. The case screws are Torx Plus (5 point) an
are available at the site Buffalo Bill showed. The bit siz
looks to be an IP6 but don't quote me on that since I haven't gotte
a bit yet.
There are other Torx Plus tools available at the Wiha site here
http://www.wihatools.com/365_IPser.ht

Normal hard drives do use a standard Torx and not the Plus version.
have opened an old drive just to look inside and if there's any dus
or humidity around it will mess up the platters.
I opened mine to get it going long enough to get some files off of i
myself rather than paying someone else to recover it.

Good luck