Platter transfer - my story & guidance request!

[MrFuzz]

Greetings!

A few months ago, my dad's Seagate ST910021AS in his works T60P gave up. The drive refused to spin up even though we could hear the heads moving back and forth. My dad did not really mind loosing the drive but he did have a few things in there he would have loved to be able to recover. Alas - a simple controller swap did not do the trick and we've concluded that the motor seized.

I managed to source the same drive from another T60P and had the brilliant idea of doing a platter swap. Lucky for my dad - he happens to work in a semi-conductor manufacturing company and has access to a class-1 clean room (that's 1 PPM) - so having a clean and safe environment to do the platter swap was the easy part. Unfortunately our happiness soon turned into gloom and doom when I discovered this drive had twin platters and not a single platter as I originally thought. Crap!

Well - my dad decided to give it a shot anyway and using some rather crafty methods, he swapped the platters and had them aligned as close to the original state as possible... or so we thought.

Once the drive was assembled, we can hear it spin up however it does not initialize; we can hear the heads trying to seek the relative starting points on both platters but I guess they are ever so slightly out of alignment and eventually the drive just shuts down.

Here's where I seek your guidance, dear Gurus: what can I do to try and recover the data from the hard drive? I figured I could reflash the firmware on the controller and trick the drive into thinking that it is a brand new drive, then use some sort of a low-diagnostic tool to read the raw data off the drive... unfortunately I do not have any further ideas... so please - share your knowledge!

Thanks in advance!

[Cleanroom]

Doing a platter swap on a multiple platter drive is no easy task. If the platters were moved independently even a micro meter, you are SOL. the platters need to come out as a single stack and there are tools for that, but tools are only a small part of it. but I am assuming you did not have any of those tools. Platter swaps have zero tolerance for alignment oddities

Platters swaps are the most difficult type of data recovery and I would say only a select few people/companies out there can do them with confidence.

At this point I would just say a eulogy for your data and move on. I doubt you will find any data recovery firm to take on this case with the current information you provided, since chance of recovery is pretty much nil.

as for flashing new firmware, That will have zero impact on y our results.

[Nick_CT]

Your alignment is gone, so is your data.

[drc]

The drive refused to spin up even though we could hear the heads moving back and forth.

Heads were moving even though the drive was not spinning? Extremely unlikely, though if true this would have most likely damaged the media surface and also the heads.

I figured I could reflash the firmware on the controller and trick the drive into thinking that it is a brand new drive, then use some sort of a low-diagnostic tool to read the raw data off the drive

This is not how hard drives work.

what can I do to try and recover the data from the hard drive?

What can you do at this point? Depending on how bad it is, an unusually skilled pro may be able to still get something, but I can tell you now that it would be quite expensive.

[lcoughey]

If your data wasn't worth the $1000 for a pro to recover in the first place, I'm guessing that it likely ins't worth the potentially ten times that amount in the state it is in now...if recoverable at all.

[unknown]

I don't think it will be recoverable. ( At least total recovery )
Even a partial recovery will cost $$$$

[CK]

Unfortunately this is yet another case of misdiagnosis by someone without the correct knowledge. A platter swap is the LAST thing that should ever be done.

This could have been recoverable at a reasonable cost had it been handled properly before a DIY effort was initiated.

[MrFuzz]

Ah well - no big deal. The data on it was not crucial at all; more like "would have liked to have" but not "omg end of the world."

Well - thanks for the input everyone.

[pcimage]

Unfortunately this is yet another case of misdiagnosis by someone without the correct knowledge. A platter swap is the LAST thing that should ever be done.

This could have been recoverable at a reasonable cost had it been handled properly before a DIY effort was initiated.

Yep, becoming more and more common

What's even more worrying is the amount of "data recovery shops" popping up and advertising like buggery on Google. All bigging up their "cleanest cleanroom in the world" Big on image, small on skills

[MrFuzz]

Unfortunately this is yet another case of misdiagnosis by someone without the correct knowledge. A platter swap is the LAST thing that should ever be done.

This could have been recoverable at a reasonable cost had it been handled properly before a DIY effort was initiated.

Yep, becoming more and more common

What's even more worrying is the amount of "data recovery shops" popping up and advertising like buggery on Google. All bigging up their "cleanest cleanroom in the world" Big on image, small on skills

Well to be honest we never admitted to be geniuses at data recovery but my pops does have access to a class1 clean room... oh well no big deal.

[BlackST]

Well to be honest we never admitted to be geniuses at data recovery but my pops does have access to a class1 clean room... oh well no big deal.

The problem is that a pro can even do without "class 0.0001 clean room" maintaining a high probability of success, can do without many tools , but just because of being a pro.
All of us "seasoned" people have made emergency recoveries on "improbable" conditions, at least one time in a life... what it takes ? Know our stuff. Plain and simple.
But even the highest chances of a successful work can be destroyed in seconds by these kind of attempts we see...

[Nick_CT]

I fully agree with BlackST and pcimage's comments, a simple DR fix can turn into a complex or even non-recoverable job thanks to some happy-go-lucky DIY attempt. In the OP's defence though, he did say that the data wasn't crucial and he wasn't worried if he lost it or not. It sounds like he was just having some inquisitive fun and trying to figure things out for himself, so in this case it wasn't a big event and fair enough. Unfortunately that's not usually the case considering the regular disastrous consequences when important data is involved.

[pcimage]

I fully agree with BlackST and pcimage's comments, a simple DR fix can turn into a complex or even non-recoverable job thanks to some happy-go-lucky DIY attempt. In the OP's defence though, he did say that the data wasn't crucial and he wasn't worried if he lost it or not. It sounds like he was just having some inquisitive fun and trying to figure things out for himself, so in this case it wasn't a big event and fair enough. Unfortunately that's not usually the case considering the regular disastrous consequences when important data is involved.

Yes, exactly so. In this case the data was not important so the OP had every right to "have a go" at DIY.

What annoys me are the "mate who knows about computers" who has a stab at recovering valuable data with mis-informed diagnosis's, local PC shops tinkering with drives and these "fly by night" recovery outfits who seems to offer nothing but "class 0.0001 cleanroom" and so on...

[DrFaustus]

Lesson learned, i wonder if it was a seized motor, since he mentioned he could hear the heads wiggling. may be a case of stiction ?

[gtd4242]

The heads won't move if the spindle is seized, and they definitely won't move if stiction is the problem. No level of cleanliness would have helped your case. A skilled data recovery tech would have better luck recovering the data in the middle of a sand storm, than you would have attempting it in the most clean of clean rooms. It's not a knock against you, it's just that there is more to this than most people realize, hence the reason it costs so much. A clean room is not a make a break factor when it comes to recovering data.

[Michael.Reilly]

So here's a random question, because I never understood it despite being a data recovery professional. (Strictly no clean room work for me thus far though.)

Why is alignment a big deal? Are we talking alignment on the platters twistwise or North, South, East, West?

I can understand why NSEW alignment will be a big deal, especially when the spindle and the platters have more play than strictly needed (because then the tracks fall too far off of circular and the drive can't keep track.) But if it's twistwise it doesn't make any sense to me, because doesn't each head acquire servo independently?

[Nick_CT]

So you are a 'data recovery professional' but you've never done any clean room work? Hmmm.

Do a bit of research on how hard drives work and then you'll understand why keeping multiple platters aligned is necessary.
http://en.wikipedia.org/wiki/Cylinder-h ... #Cylinders

[Michael.Reilly]

There's quite a bit you can do with a DDI, a PC-3K and some good old fashioned software know how. We don't do clean room work at our facility for various reasons.

As far as "doing a bit of research to find out how hard drives work" yes, I know how hard drives work, when hard drives switch heads they must acquire servo data for that head. So (theoretically) it shouldn't matter how you twist the drive as long as each platter is close enough to its previous center to acquire servo on all heads (well enough to read data.) (Because it doesn't care what's happening on the other platters)

[Nick_CT]

Indeed, there's a lot you can do with DDI and PC3K and some good knowledge and basic tools. However, quite often physical intervention is required and when you have to swap out a bad head stack there's really not any other option, other than disabling dead heads if possible, but that's not going to give you a full recovery obviously. I'm guessing you outsource cleanroom work then.

Here's an article that should answer your question regarding platter alignment:
http://hddscan.com/doc/HDD_Tracks_and_Zones.html

[Michael.Reilly]

Yeah, we outsource clean room work.

The only thing I read in that article that I didn't already know is that sometimes tracks aren't concentric, which surprised me. I assume those areas where the tracks intersect are mapped out by the p-list. But there's nothing there that would theoretically prevent you from twisting the platters independently. Is it just a matter of practicality then?