Other people's hard drive recovery stories

Here are some stories sent to me by other people.

Back in the days of 8" 10MB drives

I recall, way back in the day when 8" hard drives were hot stuff, a guy
I used to work with had a Quantum 8" drive (I think it was 10 MB, but
it might have been bigger) that went kablooie. At work we had some
just like it, so the boss let him pull the same trick with one of our
drives. And just like yours, he got his data back.

Thanks for the interesting story!

--Fred Smith

Fujitsu 8.4gig drive

Did same thing with a Fujitsu 8.4 gig drive. Lucky for me I had purchased a couple at a time for different systems so that I didn't have to go out and buy another one. Although its hard to tell from your photos, but on the Fujitsu drives they have those micro flat cables that had to be removed and reinserted, a real pain.

Glad it worked for you.


Success when drives didn't even have the same capacities

I did the same swaperoo as you on a Quantum fireball hard drive. the new board didn't even match the old board's storage, but it controlled the drive just fine.

check out the aCCount here:

Even having the same firmware is no guarantee

I used to work for Western Digital support (up til last Feb.)

You were actually extremely lucky in the end to get it
working. The major reason it's so hard to successfully switch logic boards
on a drive is because these days with the drives being so miniaturized you
can no longer use absolute positioning for the head stack and any even tiny
variation in temperature results in expansion/contraction of the
platters/heads/actuators that moves them way out of alignment. As a result
of this, when the drives are put together at the factory and having their
firmware write and burn-in testing done, each drive is somewhat individually
tuned to itself to be able to compensate for any minor manufacturing
defects, bad sectors, and expansion of the drive itself. And therefore when
you try to switch logic boards on drives, that's why it's SO difficult to
get it to work. As you say in your web page, even if you DO get a board
with exactly the same model and firmware, there is still a big chance that
it won't work at all in the end anyway.


Created his own cleanroom in his bathroom!

Well, I've got a story for you...

So I had a server with 2, 20 gig, Maxtor hard disks in it in a software
mirror. I was running RedHat Linux 6.2 with a custom kernel. Having
recently moved my office, I decided to clean up the office. I took the
Rsync mirror (an old hodgepodge machine with 5 disks striped together to
be enough space for the data) and started rebuilding it. Having just
purchased 2 new spare Matrox disks of the same model as the ones in the
server I decided to use them as the data partition and use one of the
striped disks as the system.

So, here I was with my rsync machine half installed on one of the disks
that was the data partition of it and I got a message on my pager.
Stopping for a second to read the page I immediately felt my stomach
collapse. One of the disks had fallen out of the mirror set on the
production server. So, I immediately checked the status. Thankfully
Linux had started recovering the mirror for me.

So, now I'm thinking... Wow this is working grrr..... uhhh. It was at
that point that I realized the mirror recovery was synchronizing
/dev/hdg1 to /dev/hdg1. This you might understand is a bad thing.

After the panic set it I started the discovery process. It turned out
that /dev/hdg1 was completely destroyed, the filesystem was just
trashed. It had copied data from the middle of the disk to the
beginning of the disk. Also the other disk in the mirror, /dev/hde was
in fact dead. Unfortunately it would in fact spin up, it just thunked
the whole time. The dreaded sound you hate to hear.

I tried what you tried to swap the controller but the controllers were
all 4 different. In fact the since Maxtor had just purchased Quantum,
opening up the new disks (remember they were all the model numbers)
revealed that I had actually received 30 gig Quantum drives. I tried
copying data, recovering data, even using bit scrapers, nothing worked.

Trying to restore from tapes didn't work out because as it turned out,
the cabinet that backup server got set next to wasn't a cabinet. It was
an electrical transformer for the air conditioner system.

After many attempts at a lot of things, I ended up finding a
description on how to make a clean room in your own house and that's
what I ended up doing. So after all that, here it is.

I took all my tools, a bottle of bleach, some lint free rags, and sealed
myself into the bathroom. After thoroughly cleaning the bathroom with
bleach, I turned on the shower and ran it for about an hour. This
created a room full of steam. At this point you wait another half hour
for all the steam to naturally subside. Take the bleach and lint free
rags and clean off the porcelain counter. At this point I took the disk
that had crashed and the disk that had corrupted itself and pulled them
apart. I took the disk platters from the disks and swapped them.

Don't try this at home folks, if I knew what I know now I wouldn't have
done it. However, it worked. I copied the data and recovered from the
failure with a lot more knowledge about drives, backups, restoration and
operating a fault tolerant system. I also replaced the mirror set with
a hardware based mirror set on SCSI disks.

--Chris Tooley

Blew two drives - found replacement on ebay after bidding war

I had the EXACT same problem, except that it was *TWO* drives.

I had two 40GB 7200 Seagate Barracuda ATA IV drives mirrored for redundancy.
My power supply blew, and *BOTH* drives toasted the circuit cards. The
damaged component was up at the top part of the circuit card (where wraps
around the motor and comes to a point) the leads were blackened and the IC
was melted. Clearly there was a power surge. I don't know if the drive
caused it, or it was a result of the power supply dying. It was a 400 Watt
Powmax PS.

Because the drives were mirrored, I hadn't done any backups on them in a
LONG time, and there was some seriously important stuff on them.

I went the same route as you, bought a replacement drive, only to find it
didn't work. Same story, different firmware, different circuit card, etc...

After searching EBay for weeks, asking every seller to give me the firmware
and configuration codes, I finally found someone with the same drive. Ended
up getting into a bidding war with an asshole and ended up paying higher
than retail for it. Ugh.

This can't be a coincidence. The same problem, with two different kinds of
Barracuda IV drives; it must be a flaw in Seagate's design. It's not
limited to a certain run either, since mine was manufactured in mid-2002,
yours was manufactured in early 2003 based on the date code.

I too am going to try to get a warranty replacement.

--Name withheld

Freezer trick worked

eheh, good story...
actually I'm one of the guys who tried that hdd-in-a-freezer trick too and
it worked! the disk made funny noices and the bios said "no disk here, man".
listening to the repetetive noise I could tell it was a mechanical failure.
so I put the disk in a plastic box and left in the fridge for some time.
then I took it out and quickly connected (you don't want it to reach the
room temperature) to my pc. the bios welcomed the drive and i recovered
90Megs of data, i.e. everything stored on that disk. the explanation to that
was that the low temperature causes the mechanical parts of the disk to
shrink and everyting is "in its place" till the disk warms up. thus you gain
some time to get the data back...

I wonder if there's any data recovery method involving boiling the disk...

--doga dogu

Tried the oven!


Extremely cool hard drive data recovery.

I've done both a refrigerator/freezer and an oven. The freezer worked once, I don't recall about the oven. But I've never taken the cover off to swap PCBs.

Great job.

-- Roy Zider

Refuses to buy new drive unless he or someone else buys another

A friend of mine did exactly the same thing you describe here, and he
was able to get a warranty replacement of the dead drive.

Now he refuses to buy a new hard drive unless either he or someone
else buys exactly the same drive at the same time.

--Riad Wahby

Another success

I was just reading your dead hard drive story, and chuckling to myself. I
had a friend of mines drive die one day after a bad storm (we think serious
power pulls killed it) -- but with information we needed for work on the
drives we took out its sister drive and exchanged boards to get the data off
the dead drive -- worked perfectly. Almost scary when it happens like that.

--Eric Montgomery

Didn't work

I did the same thing on my 80 GB maxtor but it didnt help (yes, I had the
right PCB firmware! I asked maxtor what I should have and they helped me).
The motor would spin down after some seconds.

But at least I tried.
Nowadays, I backup everything. Before I didnt backup everything.

--Peter Magnusson

Opened it up and got platters spinning by hand - dead hardware liberates you!

You're an inspiration to those of use who refuse to let hardware die!
I once disassembled a drive that wouldn't spin up, reattached it to the
system and, by gently touching the outer edge of the platters, was
able to get it to get the apparently stuck drive spin up. I quickly
copied the data that was important to me to a second drive that was
already in the system and then let the poor thing die a, somewhat,
normal death. Just shows that you should never give up easily.

I've been working with hardware since I was old enough
to hold a soldiering iron. I've heard of (and done) all kinds of unconventional
things to coax an otherwise garbage bound piece of equipment to work again.

When you reach that point where you simply can't make things any worse it
really liberates you to try things you'd ordinarily never do.

--James Kimble, Loveland Ohio

Older boards/drives more interchangeable

I used to repair hard drives for a living. I have swapped boards many
times to recover data. In the old days this was much easier as the boards
were pretty much 100% exchangeable..

Nowadays not only are there many different board revisions and firmware revisions
but many drive manufacturers run a Servo Calibarition routine during manufacturing of the drive.
This process writes information about that specific drive into eeprom on thedrive. If
this calibaration info is too different from the drive you are using it can cause the board not
to work with the drive.

Fun Stuff.


On a tangent - getting scratched CDs to work

Great Idea about recovering the data from broken HDDs. I had similar experiences but with the CD-ROMs. One of my friend gave me few CDs and I kept it in the sack without the covers. I was totally lost about the it and after few days, all the CDs were having scratched and cannot be detected.

Now I was in a trouble, as the valuable data on the CD's cannot be recovered. But I was knowing the information about CD-ROMs that the actual data is written on the other side (silver coat below label) and not on the clear surface. I thought to give a try and polish the clear surface by applying some oil. So I applied the some oil and rubbed in circular motion for some time and then wiped out the excess oil. And wow, the CD got detected and the first thing I did was to get all the data to my HDD.

Not to forget to remove the excess oil, other wise it may harm the CD drive.




Got a similar story? me.

Other stories
About Adsense