I spent a couple of hours reading here and I wish I had found this board earlier.
Well that said a Samsung drive of mine suffered a head crash and first professional told me the surface is irrevocably damaged. There seem to be concentric circles which is no good news.

By no means I am educated in this matter but I was surprised to read so many different opinions here. Some said even a barely visible microscratch can make the data unrecoverable while there is this famous youtube clip of this guy claiming to recovered all data from this drive: https://www.youtube.com/watch?v=n1OcTKDjO2g

I read some more about the additional issues that come along with head crash and scratching like debris. I even checked out the Youtube clip featuring the Gillware burnish thing. It gave me a good idea that data recovery is not that easy but I still cannot understand what the problem with scratched surface is. Of course besides the fact that the scratched data is gone, why is it so hard to read data "simply" before and after the scratch. I am fully aware that there must be a reason and it is not simple. But I do not understand what the particular problem is.
Maybe there are several but what are the main issues. I would really like to understand the problem why a micro scratch can kill an entire drive.

If anyone wants to give some information on this, I would really appreciate it.

Heads die on a scratch.
HDD can not deduct the firmware, if the scratch in the service area.

Thanks for your answer. Do you mean the heads die physically due to bumpiness of the surface or logically because the head does not know how to handle the scratched area?

The service area thing is very interesting, I didn´t know that the hard disk has its own software on the platters as well. I wonder how this makes sense and not some dram module in the body ad least for emergency recovery so the service area must not be taken care of.

The head needs to pass over the scratch. In doing so it will encounter an obstruction that will look like Mt Everest.

http://www.tekbasics.com/dirtysecrets_f ... d-gap2.jpg

_________________
A backup a day keeps DR away.

Ok. I though as the scratch goes into the platter, the head might be able to hover over it. It seems not. If that´s the case this would mean that the more outwards the scratch is the lesser data is recoverable?

Wo what is missing is something that can read the inner parts? Some external head that can travel across the entire platters and and pass scratches that are present.

That diagram demonstrates how tough the heads are. They're a battering ram utterly destroying anything in their path! And they jump over the small stuff like fingerprints.

_________________
On a clear disk you can seek forever.

The average distance between the molecules of a gas at STP is 3.3nm. Doesn't this mean that the air bearing in current drives is only one molecule thick?

http://www.google.com/search?q=cube+roo ... r%29+in+nm

This 1999 WD document states that the fly height of that time was 1.2 microinches. That works out to about 300 Angstroms or 30nm. It appears that today's drives fly 10 times lower (3nm or 30A).

Fly Height Monitor Improves Hard Drive Reliability (4/99):
http://www.wdc.com/wdproducts/library/o ... 850123.pdf

The atomic radius of oxygen and nitrogen is around 0.6 Angstroms, which means that a gap of 3nm corresponds to only 25 atoms.

https://en.wikipedia.org/wiki/Atomic_radius

_________________
A backup a day keeps DR away.

Yes I have read that somewhere that the flying height decreased continusously. And that the manufacturer are starting to get to the limit due to air turbulences. This is why they are using Helium now which allows less height again. If I remember correctly.

It is hard to believe that a height like this can be controlled in a stable way. It is fascinating and scary at the same time. Precious data 25 atoms away from the head crash. Very scary.