Hard Drive Failure *Cries* - Engineers to the rescue?

[cowboybob]

Off topic, here, but I can't help myself...

Mike.

Sub Navy, eh? I was an ET on Fast Attack (Deep Divin', Fast runnin', Hunter/Killers) Sub tenders (two, both a few miles from here), 67'-74'. Got to play with the SubRocs (with the occasional nuke tip - tweaked the electronics). Those puppies NEVER missed in test listen-for/hunt-for launches.

Back on topic...

Our data storage (once we upgraded away from the punch cards) was 2400' (I think that was the length) tapes. We'd unreel the used ones at sea off the fantail using a pencil as an axle. The goal was to accommplish a complete unreeling before the pencil was eaten through and broke.

Not to mention all the other crap we used to deep six.

 

[Megamox]

Hey guys,

Thanks so much for comments, they're all helpful as usual which is what i've come to expect from this forum! Believe me I have learned my lesson and as soon as I work out the logistics of my newest back up routine, hopefully I won't be caught out again. Sounds like a few of you have had some pretty similar experiences if not worse, I guess data loss or corruption comes to us all eventually? Those USB thumb drives do seem indeed to be getting bigger and it might be good for an immediate back up solution. Although it's never happened to me personally, I have heard these flash drives can go bad too? I saw the comment on cooking the drives in the oven, that's definitely a new one to me, how did you think that one up! I believe the newest technology of these hard drives rely on fluidic bearings instead of the earlier ball bearing type. Perhaps the freezing or heating is what gets this fluid going again. Who knows?

I honestly didn't realise that hard drive warranties were so low, 1 year for some?! That's crazy. I've read that there's even a class action happening against one hard drive manufacture which seemed to have let hard drives go out with a known fault and although fixed later with a firmware update, still affected quite a few customers. The hard drive I've bought to replace this dead seagate is actually a Samsung, I had no idea they're both the same company (at least for HD manufacture)! The samsung arrived this morning and has the word 'momentus' on it, which is what my dead seagate is labelled as. I think the universe is conspiring against me. Well I wont 'write' seagate off for the moment for one bad experience, but we'll see. Says it was manufactured 12/12 so at least it's new? Not sure if that's good or bad?

Never thought of the cloud experience for back ups before, I've always worried about sharing my data with people I dont really know or i'll never meet. Maybe this is the future though, I mean.. all of my hotmail/gmail email information is stored on a server somewhere in the world and we're all used to that by now. I'm sure bill gates is an avid reader of the forum and if so bill, free cloud storage hook ups for the electrotech community members please?!

Some of the stories have been quite entertaining, especially the corporate ones about how data loss affects big business. The military ones too, especially as I can imagine how much data integrity means in that context. The CSI one made me laugh, good idea though if you could work out how to do it! It is a little worrying to hear of the failure rate of optical media such as CD drives though. I'll usually copy important stuff over to a CD/DVD and typically use Nero to burn it + verification. Then if it's really important I'll run the end result through a comparison program called 'Beyond compare'. In case any of you have never heard of it, it compares any two drives/files/folders for differences and highlights them, pretty good and well worth the money! I keep all my CD/DVD's in the spindles they came in, never thought about it before, but is the generally what you all do with them too? Didnt we (rather, NASA) send out the voyager spacecraft with information recorded on a gold disc? When it finally meets an alien in a few thousand years, will the data still be accessible? What if the hard drive controlling the space ship crashes, does it have a back up? Does THAT have a back up? I cant say for sure but I'm sure I even heard once Nasa used to send spacecraft off using the Z80? No way! So cool.

Finally just a quick update. Put the dead drive into a USB caddy and physically twisted and hand spun the drive with the power on and guess what. No more beeping. But.. now it's clicking! Great. Perhaps the twisting free'd the head from the platter which could have meant it was previously a stiction issue and now the head being free, is damaged. Hopefully whatever killed the drive is what caused the damage and not my mad wrist twisting technique. Good news though is the motor spins up fine, gives off a lovely high pitched whine when it gets going as it usually used to. So I guess the heads are bad and need to be swapped.... Am I Crazy enough to do this myself? Hmm... It would be fun...

Recent quote from Seagate after explaining the new developments results in a price reduction strangely. From £1000 for a possible seized motor (which is no longer the case), to £450 to a head swap. Who knows, if I keep messing with it, it may go down even more. However all bets are off if I do ever open it up apparantly, the charge rockets from that point.

In order to aid my next decision I've got myself a faulty drive off ebay of the same make and model as the dead seagate and so when it arrives, I'll open it up and have a tinker with it to see if im eventually skilled enough to do a head replacement in a quick enough time (hopefully under a minute) to minimise any airbourne debris landing on the platter. Probably impossible or very unlikely, maybe the cold shower idea would help! Be interested to hear if anyone's actually opened one of these things and has got a positive outcome. I've also got myself one of those ball squirty things that shoot air out of the nozzle when you squeeze them (think they're also used to clean camera lenses too) which may also help during the process to clean as i go, as well as a can of compressed air. Not sure if I can use that on the platter yet though.

Any more suggestions welcome!

Megamox

PS. Here's a 3 minute video of how the professionals perform a head swap and I guess what I'm hopefully deciding whether I'd like to attempt eventually. Seems straightforward and I'm sure I could get by without that special plastic tool comb (it's used to make sure the heads never touch). I've seen other videos where other companies slide a folded up bit of paper between the head arms to keep them separated instead. Dont have that green magnet sucker tool thing either, or even know what it's called.

https://www.youtube.com/watch?v=wVcMC8uoNYk&list=UL

Possible chance of success? Let me know what you all think! I'll also look into seeing if I can obtain access to my old university's clean room (Class 100) which does allow external visitors. I'll politely explain what I'd like to do and maybe they'll take pity on me. Not sure how they'd feel about me setting up shop on one of their benches though, even if it was for 5 minutes.

(Sorry about the long post, the thread is moving quickly!)

 

[atferrari]

The best of that video is that it makes you believe that it IS easy!

 

[Megamox]

Probably horribly complicated in real life I'm sure! Cant wait to practise on the faulty one that's coming next week. I'm sure it'll put me in my place accordingly At least I'll end up with a nice pair of stripey gloves I can tell stories about at parties

Megamox

 

[kubeek]

Possible chance of success?

I don´t mean to put you off, but the chances are very slim, especially with newer drives. I can´t find the video, but it was showing a guy that actually repairs disks and the importance of setting the screws of the cover plate! to the right torque. He loosened the screwes and tightened them just by guess without even removing the cover, and the drive started clicking and didn´t work at all. Then he used a torque screwdriver to tighten them properly to spec, and voila it started working again. So I think even if you manage to get the heads out without damaging the platters (which is not easy or even possible with some drives, I took apart a bunch of them in job as a way to dispose of the data) and then put in the new head you will have very little chance of getting it to work. If you value your data, send it to some recovery company instead of fiddling with it, you might damage the disk beyond recovery.

 

[Megamox]

That's alright, I'm sure you're completely right! I do think I've seen the video you mention and I do remember thinking how significant the tighteness of the torque screws are. I'm not sure if I could record the tightness of each individual screw as I remove it, possibly marking the start and end positions of each screw as I undo it + number of revolutions? Something else to think about. Strangely in that video the guy mentions that torque screw tightness is important, but then in every other hard drive head/platter swap video i've seen it's not taken into account. My drive was built in 2011, I'm not sure if that means it's relatively new or old. I'd really love to have the data back of course, but I'd love to see how possible it was to get it back myself even more. It's a strange compulsion I know! Here's the video I think you're referring to, the guy does sound like he knows what he's talking about. It's from Gillware Tech:

https://www.youtube.com/watch?v=5CAn22TiVV0&list=UL

Feel like I've watched so many of these now, they're becoming like old friends

Megamox

 

[kubeek]

Yes that is the video I meant. But imagine you are not just swapping the cover, you are swapping the head assembly, so the torque, flatness of the surface etc. will play an even more important role. Usually when I was taking the heads out, there was no way to prevent them to touch each other, because they are parked in a plastic thing that keeps them apart, so without the comb tool you won´t get them out in the proper shape. The second thing I noticed was that allmost allways I had to remove the platters first to be able to get the heads out. Of course if you do that the platters will loose alignement and you are srewed. This video also shows why swapping the boards doesn´t work anymore https://www.youtube.com/watch?v=ICR-xw1FYlU&list=ULThwvsGbqk9c

 

[Mike odom]

crap being the key word here... yeah, our system had paper tape also, we used it to transfer guidance system data in and out (when receiving and shipping guidance systems)... our sonar had mag tapes, but those and the tapes from our IDAS (integrated data aquisition system) went to John Hopkins for analysis after each patrol.

 

[Megamox]

Good point about the head assembly! I was thinking that the faulty hard drive I've bought which is due to arrive next week will have it's own little plastic spacer which parks the heads in that drive (I believe it's called a ramp?). Anyway I can remove it and use it to space the good heads while I'm transplanting them, what do you think? I hope the heads are removable, I wouldn't even attempt a multi platter swap, not without the special tools. I've heard alignment needs to stay within micrometers or the data is lost. Here's a funny video of a guy who attempted making his own platter transplant tool, quite hillariously, with a tin of sweet corn.

https://www.youtube.com/watch?v=wgqOcEAdSas&list=UL

I laughed but secretly thought, how ingenious!

Megamox

 

[cowboybob]

...Finally just a quick update. Put the dead drive into a USB caddy and physically twisted and hand spun the drive with the power on and guess what. No more beeping. But.. now it's clicking!...

The clicking might have been the heads' servo re-aligning itself. Was the USB caddy providing a data link as well? If so, was there any comms or recognition of the drive at all?

In the past I've used DOS level Fdisk.exe to at least see if the drive is working well enough to allow a read of the partition tables' data (since that's the first info the mounting BIOS encounters), even though Windows saw nothing on the drive or even the drive itself.

Careful though - Fdisk can be very unforgiving. It does NOT allow "hot" partition table alterations. Just use it to look.

 

[Megamox]

It's more of an actual cable hook up than a caddy to be honest, but I had it connected to the USB port of the computer and a power supply at one end, and at the other it powered the SATA drive and a connection to the DATA. It spins up, clicks about once a second for about 10 seconds, then spins down. Doesnt show up in windows but I could have sworn that on the first attempt windows popped up a message saying your new hardware is available to use. When I checked though, nothing still showing up in Windows nor in the Windows Drive Management. Havent seen that message since. I doubt Fdisk would work on a drive that didn't show up would it? How would I go about doing this? CMD - Fdisk.exe D:\ ?

Thanks,
Megamox

 

[kubeek]

you would need to properly boot to DOS to be able to use fdisk, windows won´t let you touch the hardware. Or you can boot from a flash drive / CD with Parted and use the linux version of fdisk.

 

[Megamox]

I do believe I have a Live CD of linux somewhere which I used to partition my memory cards with, I think it was called GParted. I'll go dig it out. So the drive may show up on that but not on windows? Strange. How would Fdisk be used to fix the drive and recover data if it does detect it? I'm such a windows geek, linux not so much

Megamox

 

[kubeek]

I doubt that fdisk will help you much with recovering the data, but it can help you with one of the most basic tests - whether the drive can read at least the partition table that is stored in the first block of data.

 

[Timescope]

Did you clean these contacts as I suggested earlier? A thin layer of oxide makes a good insulator. Use a pencil eraser to clean them.

Timescope

 

[Megamox]

Completely forgot to mention, I did! I used some rubbing alcohol actually and it seemed to help. Maybe that contributed to things, thanks! Cant seem to find that Live Linux CD at the moment... but I'm still on the hunt. Cant remember what I had planned doing this week but a hard drive failure seems to have over shadowed everything else. At this point in time I know more about hard drives, platters, heads, spindles and stiction than I ever cared to

Megamox

 

[cowboybob]

...I doubt Fdisk would work on a drive that didn't show up would it? How would I go about doing this? CMD - Fdisk.exe D:\ ?

Thanks,
Megamox

Maybe.

Boot up your machine witout the iffy drive. Navigate yourself, using the "Command Prompt" accessory to the drive that has fdisk on it.

Hook up the iffy drive. Data lines first, then power. As quickly as possible, after the disk has spun up, run Fdisk from whatever location it is on. The syntax at the command prompt would be:

fdisk d:

or whatever is the next available letter after all the other drive letter designations already attached to the computer).

Don't hesitate to use different timing scenerios to cover your bets.

If it works, do not alter any partitions. That will make the data utterly unavailable without some serious Sector Editor level apps.

This will at least confirm R/W head/platter, first cylinder condition. If successful, you could then use other apps for data retrieval.

 

[()blivion]

I don't recommend listening to Scott Holewinski or any of his videos.

Some of the things he says are true to be sure, but some of them are fishy at best. You'll also notice that he has the comments for all his videos set to be shown only after his approval. This is because he doesn't allow any views that conflict with his own, fact or not.

Case in point, *THIS* Video, where he tells you that you can't swap boards with new drives, even if they are right next to each other in serial number and such. Although it is true that they have calibration data, you should easily be able to recover that data and move it to the new board. For example, use the manufacturer firmware tool to rip the old firmware off the board, and then put it on the new board.

I mean, if what he says is true and it can't be done, then how the heck is he doing it... Magic???

I have made several comments to the same tune on that particular video and I believe he has blocked them all. I also mentioned the flaw in his logic that head calibration data could be stored on the platter, which obviously could only ever be read by a properly calibrated head, which is a direct contradiction. Even if not for that, the argument doesn't make sense from the start as switching boards doesn't switch the data on the platters, so there should be no problem retrieving the calibration data. Other than of course the first problem.

Finally, there are plenty of YouTube videos out there doing exactly what he is saying you can't. Are they all lying, and only he is telling you the truth???

I don't think so, It's more likely technological propaganda. IMO He's a salesman using YouTube to try to make you buy his product. It's his business after all, he can't possibly be impartial to DIY.

 

[WTP Pepper]

Another word of warning found through experience at work and home.
If you back up to DVD or CD and keep them in their original jewel cases where there is no contact with the disk surface then you are OK.

I discovered problems when they were stored in those nice large folders for a few years. The chemicals in the plastic react with the CD surface that is effectively in permanent contact with the disk surface over the years.

I have several movie DVDs that won't play having being stored in such folders, whereas those stored in their original jewel cases are fine. Both look immaculate on visual inspection.

 

[()blivion]

If the OP is absolutely determined to disassemble his drive, and genuinely wants a chance of recovering his data, he needs a clean environment.

Rather than a clean room, I would recommend that the OP try and get/make a "clean box". It's a much smaller area, so it will be cheaper to make and easier to control. It's much more practical to DIY than a full clean room. I would imagine that one would build such things with an inbuilt real HEPA filter of U15 class or better. Probably followed by an electrostatic precipitator to further enhance particle capture. Another possibility is the use of a tank of inert gas to fill the chamber, such as welding gas run through a high grade HEPA filter. This would purge the air and (hopefully) replace it with something having no particles. One would need to make sure the gas was vented outside ones work environment. Otherwise one may die of asphyxiation, which wouldn't be very productive /)

The box would be made of a smooth material, most likely acrylic plastic or similar. Rough or irregularly textured surfaces would collect particles. The box would also be built to minimize sharp inside corners, so as to make for less places dust could collect. The bottom of the box would have an anti-static mat of semi soft material connected to ground. This is to minimize the possibility of ESD damage caused by the electrostatic precipitator if one was used. The front would have two gloves, made of anti-static/anti-particulate material, with the edges of the gloves sealed to the box. The top front, at an angle, would be a viewing window that opened and latched down air tight. Inside would be an air squirter for removing dust, and all the tools needed to take drives apart and put them back together. It may also be beneficial to have installed either a desiccant, or a water dish to control the humidity.

To use the box, you would open the window, insert pre-cleaned drives, latch the window, and start the HEPA filter or turn on the gas. With the gas/air filter running, you would de-dust all your tools, the drives, and the internal environment as best as you could with the air squirter. You would want to let the box environment "cure" for several hours, agitating/squirting all the contents periodically to release trapped particles. At this point it should be safe to open the hard drive, and do some work on it. Though I would still recommend working relatively fast.

Here are air specs for a typical hard drive clean room, in parts per cubic foot. I would try to exceed these specs if possible.

< 5000 particles smaller than 200nm
< 250 particles smaller than 300nm
< 50 particles smaller than 500nm
Must have no particles any larger.


Conclusion.
If the drive in fact needed to be opened, and if you did all of the above, you *MIGHT* *MAYBE* make the drive operational long enough to get the data off of it. Hard drives during their life wear down and emit microscopic bits of metal that they have to deal with if they are to survive long term. To combat this, most drives have an internal capture/filter mechanism to clean up a bit. This filter is "powered" by the spinning of the drive. Because of this, it is possible, if you minimize contaminants greatly, for this filter to pick up the stragglers enough that the drive will be able to function. The chances of any of this happening are pretty slim though. And in the end, such a clean box would still probably cost you as much as any recovery service.