How Hard Drives Work - The Importance Of Platter Alignment 

Thanks

is it possible to Re-low level or layout new servo tracks to get the drive working again in the event this happens? Factory has to do it somehow... Not data recovery wise, just use the drive again from blank.

Can we 'reset' the drive? Like use the platters as is now? Or is ready for the trashbin?

Great demonstration; a little too late for a little too many curious people. Will be linking this to clients who come in expecting miracles. Thank you.

What about drives with only 1 platter? Does it result in the same effect?

I can see a way to regain the alignment, but it would require an automated iterative approach moving the top disk relative to the bottom one. Better not to lose the alignment. How do you do that?

Does not organisations with enormous funds like the NSA have the resources to individually read the platters and match them digitally to make the data readable again? I really think they do

After the platters have been moved out of alignment, does that mean the drive can never be used again and is totally scrap, or can it be used again if you're prepared to sacrifice the data on it by repartitioning and formating it? In other words, is it just the data that's gone for good, or is it the entire drive and data that's gone for good? ...or for BAD as the case would be.

is that the case for hdd that has only a single platter? I swapped a single platter WD500 onto a donor and got clocking sound and it was only a single platter

Um, why the hell is this a problem? Seems like a huge oversight. The platters should have some type of internal alignment data that cannot be overwritten that allows the heads to determine the out of alignment and then realign in software if necessary. E.g., if you have platters that are out of alignment by N bits(which will be what happens due to rotational changes(as I don't think track disalignment can happen)) then the heads can still read the bits. If the alignment error is known(which can be done by having a built in alignment sequence) the the heads can read the data and the software and realign the data by shifting. E.g., if H1 1 2 3 4 5 6 ... H2 4 5 6 7 8 9 ... Then the the data can be realigned by simple shifting in the software by shifting H2 by 3. All this requires is that there is some type of alignment data on the platters(this could be physical or if written to and never overwritten, e.g., a single track of a random bit pattern or some bit sequence that doesn't repeat)). This would only have the issue of slowing down the reading of data since the heads have to read data before and after but this may not be as big an issue since most of the time the heads read more data than necessary anyways(the worst case is if all the heads are out of maximum alignment requiring excessive reading). It doesn't matter if the data is scattered around the drive because the alignment be a function of theta(the angle of rotation) for each platter. It may be that one has to also know the unaligned theta for each track (r*theta) but this probably could be calculated or a radial track of alignment could be used(wasting some space). Of course there is no reason to do this since drives are not expected to have their platters swapped but it would make alignment for data recovery a non-issue. Even just having a little alignment data on the platters could help significantly for data recovery and it possibly could be done by software if software could read and write a special alignment track before disk initialization. Another way that might work is for each platter to have a grove(s) cut in it that align to the spindle in only 1 way... but not sure if this would be or could be done accurately enough. It also seems that, possibly, one could realign in software. After all, it's just a shifting problem and it is unlikely all but one combination works to produce valid data. This would require a contiguous file known ahead of time put across the platters that could be used to reconstruct for alignment discovery. It just seems to me that the manufacturers offer no way to actually align platters easily(or maybe they do?) when it is really not that difficult of a problem and a little extra can go a long way(making something impossible possible).

Well it can be fixed by super-duper high-tech companies that can read the patterns by them self outside of the case in high tech equipment, but yea they take money for it and heck a lot more then you, which is not that bad if I might say.