Cause of Hard Disk Failure
Albert
A question came up on a different NG about whether the constant
changing, rearranging, formatting or otherwise manipulating data to
and from their hard drive wears out the hard disk surface.
Does the pick up on the hard disk actually touch the surface of the
disk and if so is there any wearing of the surface as in vinyl records
and or audio and videotapes?
Generally what is the usual cause of failure was a hard disk?
Albert
Albert
wrote:
I apologize, I just googled it and got a real good explanation.
http://en.wikipedia.org/wiki/Hard_disk_failure
Albert
Paul
Before Hitachi sold off their (IBM acquired) disk division,
they used to have web pages concerning hard drive technology.
They provided information (in the "IBM way") that the other
disk companies don't provide.
That information is now lost, as the web pages in question
were guarded with a "norobots" and consequently were not
archived by archive.org .
One thing they were experimenting with, was "contact" style
disk recording. They allowed the clearance to drop to zero,
between the head and the platter, while the platter was
spinning.
They were experimenting with lubricants (presumably dry lubricants)
which could extend head life enough to make this possible.
In their experiments, the head could "fly" at zero height
for one month (30 days), before the head was destroyed. So
the lifespan at the point they were publishing the results,
wasn't sufficient to introduce the feature.
Flying height is getting very close to zero, but is
still a finite distance, and the heads still don't touch
while the disk idles.
One innovation on disks, is variable flying height. Modern
high capacity drives, use a different flying height on reads
versus writes. The head contains a heater. The heater is
activated, whenever they want the head to "swell" and get
closer to the media. This allows whatever mechanisms they
have for tracking flying height, to perhaps maintain the
head assembly at some distance, while dialing in a bit
of differential height using thermal expansion properties.
I've read references to them being able to monitor
flying height. Whether this is just for lab purposes,
or whether the drive actively uses that capability, I
haven't found any reference material on that. Maybe they
could use capacitance between head and platter, to
measure such a thing. They didn't state what the tech was
to allow doing that, merely that they had the ability to
measure what flying height was achieved. Flying height
can be implied from rotational velocity, head lift factor
(wing lift), head load (downforce via spring loading) and
so on. But being able to measure it, allows even closer
tolerances to be involved.
The Hitachi (hgst.com) web pages provided some insight
as to how the stuff works, but without detailing everything.
For example, on a Hitachi web page that described the
solid lubricant on the surface of the platter, I was able
to find a patent application with a much greater level
of detail of the chemical composition. Still, I wouldn't
have been able to *find* that patent, if I didn't have
the name of the chemical compound to search with.
This is what you get now, from that web site...
If you know of a web page cache other than archive.org,
you can see if there is still a copy of this around.
http://www.hgst.com/hdd/research/storage/as/index.html
"Sorry, but that page does not exist!"
HTH,
Paul