DIY Data recovery from 'dead' drives

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
This is pretty early (not attempting anything until next summer - this will be a slow topic maybe only periodically responded to til then) but I have a box full of 'dead' drives from years ago especially from the last time my video archiving project went to hell. They arent worth the money for professional recovery, but I dont mind attempting even advanced technical procedures as a "you lose nothing" last chance.

I'm expecting to build a small 'clean room' or at least clean box for the purposes of drive disassembly and operation - I don't have directions for that yet but I assume it's either out there or even something improvised will be superior to just doing it open air in a normal room. But I was thinking at least HEPA filtration (not sure what it cost to go up to ULPA) pushing into an overpressure box with a plexiglass top and some attached gloves on the side that you reach through and the cleaned tools already in there...

This is in part because i'd like to learn/sounds like nerd fun to be able to do such a thing as well - I already consider the data hopelessly lost, if it's actually not that's gravy. Might be a useful skill to barter with friends in the future too. Does anyone have any suggested starting points for information?
 

pricklypunter

Well-Known Member
Nov 10, 2015
1,708
515
113
Canada
Start gathering up a wee collection of good controllers for the disks you'll be working with. It's a crap shoot obviously, but a fair percentage of disks fail because of poor motor drive and other issues on the controller board, rather than anything mechanical in the disk. Sometimes a sharp tap on the side of a disk will get a stuck motor spinning again, but don't go crazy, you don't want the heads leaving the landing zone as a result. Also, make sure you have the means to transfer any recoverable data on hand, so that you do not have to physically disturb a disk that decides to play. That means having those interfaces available inside whatever micro environment you are planning to work in. Avoid silly stuff like placing disks in a bag and putting them in the freezer. Instead, use a can of freezer spray to give the controller board a blast for those times when either a de-laminated via or bad solder joint is the issue. Lastly, forget about ever putting any recovered disk back into operation, it's simply too risky :)
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
if it works out I will send you my Seagate 3TB crap drive I am still waiting for Seagate to pay up in the class action lawsuit...lol...
 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,140
594
113
New York City
www.glaver.org
I'm expecting to build a small 'clean room' or at least clean box for the purposes of drive disassembly and operation - I don't have directions for that yet but I assume it's either out there or even something improvised will be superior to just doing it open air in a normal room. But I was thinking at least HEPA filtration (not sure what it cost to go up to ULPA) pushing into an overpressure box with a plexiglass top and some attached gloves on the side that you reach through and the cleaned tools already in there...
You will probably give up on this idea the first time you open a drive in your hot-box and it sheds particulate contaminants all over your clean room. Manufacturers usually include an inspection port which they can open to inspect the interior without risking contamination. On WD drives, it is often a foil oval on the side. Most other drives have a plastic "cork", usually underneath the label with the model / serial / etc.

The manufacturer refurb process generally goes like this:

1) Test entire drive - if "no problem found", move to end of rework line, before final inspection
2) If test from 1 fails, determine if problem is on logic board or inside sealed HDA. If on logic board, replace board and go back to 1.
3) If problem is inside sealed HDA, determine if HDA needs to be opened or if problem can be fixed by rewriting servo / data areas. If problem can be fixed without opening, fix and go back to 1.
4) If we get here, evaluate whether sufficient stock of reworked drives exist to send out replacement and keep this drive on a pile of "to be reworked, maybe".
5) If we have insufficient stock, determine whether a reworked newer and/or larger-capacity model can be sent out instead.
6) If drives must go to HDA rework, open inspection port and categorize failure type. Sort into piles of decreasing interest in rework (head crash with loose particulate goes on "hope we can write this off as a loss and we don't have to fix it" pile, bad spindle motor or head pre-amp goes on pile for clean room re-work.
 

Patrick

Administrator
Staff member
Dec 21, 2010
12,511
5,792
113
@Terry Kennedy you forgot that after 1 some percentage goes off the line to get SMART data reset so grey market HDD vendors claim new drives.
 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,140
594
113
New York City
www.glaver.org
@Terry Kennedy you forgot that after 1 some percentage goes off the line to get SMART data reset so grey market HDD vendors claim new drives.
I don't know if that's a diversion out of manufacturer's RMA areas or if it is companies that buy up batches of drives from off-lease systems and then slap their own labels on them and mis-represent them as new. I've only ever run into one of those drives, and it was in a PC someone bought from "Microsoft's largest OEM refurbishment partner" (presumable for cheap Windows licenses) and also had bad memory, which they asked me to fix. The drive and memory went to the recycler.

OTOH, the manufacturers are at least partially complicit in that whole thing - since the commands to wipe the SMART data are vendor-specific, there's no reason the manufacturers could not add a "SMART tampered with" attribute that would flag any drives that had the SMART data reset in the field.
 

Patrick

Administrator
Staff member
Dec 21, 2010
12,511
5,792
113
Some of the grey market guys are the 3rd parties contracted in the returns process.

So after the first inspection "oh these drives are terrible mechanical failures" -> erase SMART and sell.
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
Some of the grey market guys are the 3rd parties contracted in the returns process.

So after the first inspection "oh these drives are terrible mechanical failures" -> erase SMART and sell.
This has FRAUD written all over it!
 

funkywizard

mmm.... bandwidth.
Jan 15, 2017
848
402
63
USA
ioflood.com
I would say that "sticking them in the freezer in a bag" has worked plenty of times for me. Especially on Seagate drives. 15 minutes should be plenty. Sometimes that's all the drive needs (works fine after that for as long as you need to copy your data), and sometimes the drive is good for 15 minutes to an hour before it needs another freeze treatment. So be prepared to copy your data quickly if it does work, and use a copy method where it's easy to start again where you left off.
 
You will probably give up on this idea the first time you open a drive in your hot-box and it sheds particulate contaminants all over your clean room. Manufacturers usually include an inspection port which they can open to inspect the interior without risking contamination. On WD drives, it is often a foil oval on the side. Most other drives have a plastic "cork", usually underneath the label with the model / serial / etc.
I'm not looking to refurbish drives, just recover data. I would plan for the clean box to itself be cleaned right before and after work. The minimum i'd like to do is a logic board swap - I probably have a few of just those. I have some Seagate 3tb's that didn't make any funny sounds, just stopped responding. I'd like to see if a logic board swap might bring any of them back to life. I'm curious what the next step at data recovery places normally is and whether that might be replicated at all, in any fashion, even if less carefully than they do.

I'm an inherent tinkerer and have nothing to lose, why dissuade me..?
 

vl1969

Active Member
Feb 5, 2014
634
76
28
I would say that "sticking them in the freezer in a bag" has worked plenty of times for me. Especially on Seagate drives. 15 minutes should be plenty. Sometimes that's all the drive needs (works fine after that for as long as you need to copy your data), and sometimes the drive is good for 15 minutes to an hour before it needs another freeze treatment. So be prepared to copy your data quickly if it does work, and use a copy method where it's easy to start again where you left off.
Definitely done that several times, some times it works some times it does not.
also come across drives that had to be positioned a certain way for them to work. like on one of the sides or on the top edge, or upside down, controller up. even after freezing.

had a drive once that had to hard spin by hand (set it on a flat surface controller down and give a good vigorous spin/shake ) for it to work.

never tried to open up the drive to recover data. with all that is required to build a clean room it may be cheaper to find a data recovery service, if it is just one time deal.
 

vl1969

Active Member
Feb 5, 2014
634
76
28
I'm not looking to refurbish drives, just recover data. I would plan for the clean box to itself be cleaned right before and after work. The minimum i'd like to do is a logic board swap - I probably have a few of just those. I have some Seagate 3tb's that didn't make any funny sounds, just stopped responding. I'd like to see if a logic board swap might bring any of them back to life. I'm curious what the next step at data recovery places normally is and whether that might be replicated at all, in any fashion, even if less carefully than they do.

I'm an inherent tinkerer and have nothing to lose, why dissuade me..?
you do not need a clean room for logic board swap. the board is outside the drive and can be replaced at any time.
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
you do not need a clean room for logic board swap. the board is outside the drive and can be replaced at any time.
How do I know if this is the case?

I have a Seagate 3tb, the infamous one... Is is worth it to spend like $50 to buy one that are like on sale some times and swap out the board? anyone knows the reason why most of these Seagate 3tb failed?

I haven't tried the freezer yet. but how long do you leave it in? do you worry about condensation when you take it out?
 

vl1969

Active Member
Feb 5, 2014
634
76
28
problem is that the drives MUST be the same for controller swap. you might spend your 50$ and get the drive that is similar but not the same. I have 5 WD drives at home that are all 2TB and similar models but board are different on 3 of them. they do not fit each other.

ok, freezer only works if you can hear the drive spinning up when connected to power.
and only if the issue is not the controller board.

they good ,more or less safe, way to do it is put the drive into ZIP bag or a good think plastic bag and seal it up as much as possible. (not a problem with Zip bags) .
put into freezer for about 15-20 min. BE PREPARED TO RECOVER THE DATA :)
have the PC you will use to recover the data up and running, and the program that you will use open as well.
FYI, if the actual data is intact you may only need to make drive accessible so prepare the folder for data recovery beforehand as well and have it open.

after that as soon as you take it out of freezer, connect the drive to power and bus. I guess SATA,
if PC finds the drive and mounts it simply copy the data into recovery folder.
if PC finds the drive but can not mount, go into the program for recovery and see if it can access the disk and do scan on it. one of the method might work .

also, if you have one of those freeze packs you can put the disk on it(inside the zip bag) during recovery if it works.

this are the steps to try before moving onto buying controller.

have reread your post and see that you said that data is not worth the professional recovery.
let me tell you right off the bat, if the data is not worth professional recovery, it is not worth spending time on clean room. there is nothing inside the drive that will let you recover the data. the only reason recovery specialist open the drive it because they have special and expensive electronic gadgets that let them read the platters directly, bypassing the the whole drive mechanism and circuitry if needed. you do not have that.
you can only replace the controller that is outside the casing. and for that you only need an antistatic mat to make sure you do not fry the board with static charge.
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
drive would spin up, connect to sata it would show the folder structures and file names, but trying to copy you get a weird high pitch noise and it would just fail to copy a lot of the files. i were able to recover about 10-20% back when it first failed like 2 yrs ago.. haven't tried again recently
 

William

Well-Known Member
May 7, 2015
789
252
63
66
There are a couple of programs I have used to recover data from dead drives... note: they still need to spin and some what work.

ARAX Disk Doctor Data Recovery
EASEUS Data Recovery Wizard Professional
Active File Recovery

If I remember correct ARAX worked the best.
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
well, like i said, it's 3tb drive so the data copying would take a long long time even if you can copy it out. so do you refreeze again after 20 min? i am still concerned about the condensation that happens when the drive taken out of the freezer. the zip bag might keep the moisture out while in freezer but when you remove it from the freezer, wouldn't the cold temperature of the drive cause condensation from the hotter air in the room?
 

vl1969

Active Member
Feb 5, 2014
634
76
28
well, like i said, it's 3tb drive so the data copying would take a long long time even if you can copy it out. so do you refreeze again after 20 min? i am still concerned about the condensation that happens when the drive taken out of the freezer. the zip bag might keep the moisture out while in freezer but when you remove it from the freezer, wouldn't the cold temperature of the drive cause condensation from the hotter air in the room?
true, but than that is why I suggested using a freezer pack under the drive to keep it cool.
I had recover 2TB once with that technique.
in my office we had a client that send out a POS terminal as not working. the backup was old so we needed the data from the terminal.
so I put it in the freezer, and than used the bag of ice under the drive in a closed up cardboard box to keep it running. data recovery was 6 hours. whole day. we had put a double up gallon size zip bag full of ice on the bottom of small shipping box put the drive in a freezer in quart size zip bag for 30 , than take it out, open the bag and connected the drive to PC. close the bag as much as possible (it was a PATA disk so wide PATA able ) and fit it in the box. closed the box and run recovery app. I believe at the time it was an R-STUDIO or something like that. it took 6 hours to get the data out.

however in your case it might be pointless as it may be the drive head issue, the wining noise you hear when trying to read the drive maybe the reading head hitting the platers.
 
you do not need a clean room for logic board swap. the board is outside the drive and can be replaced at any time.
I know, but I want to know more about what the next step is anyways. :) I'm one of these people that has to do things myself.

How do I know if this is the case?
I have a Seagate 3tb, the infamous one... Is is worth it to spend like $50 to buy one that are like on sale some times and swap out the board? anyone knows the reason why most of these Seagate 3tb failed?

I haven't tried the freezer yet. but how long do you leave it in? do you worry about condensation when you take it out?
Probably because they all came after that flooding that wiped out half of all drive manufacturing or something in 2010 or 2011 that tsunami so quality took a dive.


have reread your post and see that you said that data is not worth the professional recovery.
let me tell you right off the bat, if the data is not worth professional recovery, it is not worth spending time on clean room. there is nothing inside the drive that will let you recover the data.

you can only replace the controller that is outside the casing. and for that you only need an antistatic mat to make sure you do not fry the board with static charge.
Well #1 we all have our hobbies. :) #2 I dont get paid professional rates in my daily life as i'm a starving film student so the amount of hours I have to work (probably 100+) to pay a professional to look at the drive is not worth it. I'm pretty sure I could put together a cleanbox for alot less than that and might do so just out of curiosity.

That being said, what about swapping parts of the mechanism like the head? I'm the guy that when I see something like circuit board component level repair, or some guy tearing apart a car cylinder head, I say "oh hey I could do that". It's just precision work, not brain surgery. So I swap a head and maybe it scratches it, I bet it still read 90% of the remaining drive. Or 50%. That's 50% more data to get back than I had access to before.

I'd like to see video of the next step processes before writing things off, or to understand what those processes even are. I'd heard years ago of other people successfully doing this themself, I just dont remember the details because it was beyond my needs at the time.

There are a couple of programs I have used to recover data from dead drives... note: they still need to spin and some what work..
The best i've used is SpinRite, I dont have any experience with the other three but if anyone does i'd love to hear a comparison. That brought one drive back from the dead for me from where I dont even think it would see the MBR anymore (but at least knew a drive was connected, so it directly accessed something somehow) to where it fully booted to windows again and i've NEVER heard a drive make as wierd sounds as that one did, during and after the recovery... lasted long enough to copy 80% of files off with 80,000 bad sectors for the rest, but again the problem is that it has to show up to the system first.

My problem is I have several of these Seagate 3tb drives that just went dead without a word - they don't show up as a connected drive at all - not even "I can see the MBR and it's blank" just "nothing is attached" instead. I can hear them spin up but they don't recognize so it's not stiction. A logic board swap will be the first attempt when I get to it. I bought a bunch of matching drives at the time so I think I should be able to find one.

I actually wonder if one of the drives just got "overly full" (meaning 100% capacity) and then had like some sectors die somewhere, attempted to do some kind of on disk management of sector reallocation, and had nowhere left to put even an updated MBR. It was on a task downloading until full until it just shut down and I say this because that drive was already going thru sector reallocation stuff before it got that full.
 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,140
594
113
New York City
www.glaver.org
My problem is I have several of these Seagate 3tb drives that just went dead without a word - they don't show up as a connected drive at all - not even "I can see the MBR and it's blank" just "nothing is attached" instead. I can hear them spin up but they don't recognize so it's not stiction. A logic board swap will be the first attempt when I get to it. I bought a bunch of matching drives at the time so I think I should be able to find one.
Seagate drives [at least used to] store most of their firmware on the drive platters itself - the small amount in flash memory on the drive logic board is just enough to get the drive to spin up, move the heads to the dedicated firmware area, and read the rest of the firmware. If something goes wrong during that step, the host won't see the drive.
I actually wonder if one of the drives just got "overly full" (meaning 100% capacity) and then had like some sectors die somewhere, attempted to do some kind of on disk management of sector reallocation, and had nowhere left to put even an updated MBR. It was on a task downloading until full until it just shut down and I say this because that drive was already going thru sector reallocation stuff before it got that full.
Reallocated sectors come out of dedicated spare area(s) on the drive, not out of the user-accessible space.