LaCie bigdisk 1Tb firewire/usb successful recovery story

[Burner]

I'm the unhappy owner of an external LaCie 1TB Bigdisk Firewire/USB, these are well known for blowing the power supply after a short time. The symptoms is that the drive start clicking and fail to mount when started. If i send the drive back for warenty repair, i'd likely loose all data on it, so i decided to take it apart and recover my data. After searching for a while i did not find any useful information, so i decided write some. The volume consists of 2 SATA 500Gb drives on raid0 with a HFS+ (MAC) filesystem on it. First i tried to power the drives from my workstation(PC)'s power supply but that did not help, so took the drives out and connected them directly to the workstation's SATA controller.
The rest i will write as a step by step guide to recover these drives, i hope this article can help others.
This guide should work for MAC as well as PC.

Note: The guide is written from memory, i'm not able to test it here is i'm not at my workstation at the time of writing.

1. Download Ubuntu linux live CD http://www.ubuntu.com/getubuntu/download and burn to a CD
2. Shut down the system and place both drives in your computer and connect them.
3. Boot the computer from the Ubuntu CD, you dont have to install anything, all "installations" will be done in memory only and lost after reboot.
4. Open a Terminal "Applications->Accessories->Terminal" and install a few tools:

Ubuntu:$> sudo aptitude install hfsplus hfsprogs hfsutils mdadm testdisk

5. You will need to find the right device files for the two drives.
This will output all known drives and partitions on the computer, two drives we are looking for have Unknown Partition.
Ubuntu:$> sudo sfdisk -d
# partition table of /dev/sdathe out
unit: sectors

/dev/sda1 : start= 2048, size=122880000, Id= 7, bootable
/dev/sda2 : start=122882048, size= 3906056, Id=82
...
...
# partition table of /dev/sdc
Unknown partition
....
...
# partition table of /dev/sdd
Unknown partition
...
...

Ok, our drives is in this case /dev/sdc and /dev/sdd, this will be different on your system.

6. Build the raid md0 from the two drives (none of this write anything to the drives, its perfectly safe)

Ubuntu:$> sudo mdadm -B -l 0 -n 2 --chunk=256 /dev/md0 /dev/sdc /dev/sdd
This output nothing if nothing goes wrong, now check of the kernel found partitions.
Ubuntu:$> ls /dev/md0*
/dev/md0 /dev/md0p1 /dev/md0p2 /dev/md0p3 /dev/md0p4 /dev/md0p5 /dev/md0p6

If the output show 6 partitions like this, then everything is ok, if only /dev/md0 and /dev/md0p1 is shown, then the drive order is wrong and you will need to switch them around in the raid, to do this, stop md0 "sudo -S /dev/md0" and build the array again with the two drives swiched around like this ... "/dev/md0 /dev/sdd /dev/sdc"

If you dont see all 6 partitions now, you should stop and get professional recovery help

7. Mount the drive.

Ubuntu:$> sudo mkdir -p /media/bigdisk
Ubuntu:$> sudo mount -t hfs /dev/md0p6 /media/bigdisk

Congratulations, you should now be able to access all your data in /media/bigdisk, go ahead and recover your precious data
Cheers
//Burner

** Never underestimate the bandwidth of a truck full of tapes **

[tamy_beatles]

Hello, our Lacie Bigdisk 1Tb triple interface has stopped being recognisable by our computer. We've tried on other computers but we were unlucky and also tried a new Lacie power supply but no luck either. As you said, if we send the hard drive to Lacie we would loose our data and we don't want that. On the other hand, if we send it to recover the data it would be very expensive, so we would like to try your technique but have an important questions:

We were told that the two 500gb disks inside the case are joined by the Lacie circuit so we were told that if we take them out the data cannot be read cos they are not more joined to this circuit. I don't know much about this subject, but could you explain to me please how these two disks are joined and how do you recreat the RAID 0 (meaning, the union between the 2 disks) when you plug them directly to your PC to read the data?

Your help will be very appreciated.

Thank you.

Tamara.

[ppumkin]

You dont need to worry about unplugging the hard drives from the Lacie Box. But do mark which one is Left (usually raid drive 0) and which one is right(usually raid drive 1) for reference, in case it matters later.

Disassemble the unit taking extra care while handling your important data.

The circuits inside are the same as any other IDE/SATA interface running a special RAID chip. As long as you know which RAID (0 or 1) or maybe higher you are running and which version of OS is running inside your box.

Burner used a safe way to recreate a standard raid 0, allowing him to access his drives again.

These techniques can vary and you need some experience in Linux. I have not seen any windows based software based raid emulation like he has done. You will need some good guide to help you out.

Maybe burner could tell how he knew what to do.

Just remeber if you going to attempt to do this, you must remember not to re-initialise or re-build the drives. This will 'format' them and you will loose your data.

[enderst]

step 1.
IMAGE THE DRIVES

[tamy_beatles]

Hi guys, thanks for your replies. One of our IT people is coming today to check the HDD and we'll pass him all your advice.

Another query. On our computer we have the Stellar Phoenix Data Recovery (Windows) installed but when we scann out PC it did not recognise the Lacie hard drive either(of course, the Lacie HDD was plugged into the PC) Any ideas why this could happen?

Thanks again!

Tamara.

[seaccentaigu]

Hi,

tamy_beatles and burner (& any others...)

Does the initial problem looked like this ? Anyone knows about this problem (this is my video by the way):

http://www.youtube.com/watch?v=u4sfcBRaLpA

Lacie told me to send the drive for data recovery, the data recovery company told me that this might be a short-circuit problem.

I am trapped in choosing to send the drive for data recovery and replacement OR dismantling it to try to recover the files and loose the warranty.

I absolutely want the files that are on the drive, they represent 10 years of precious sound recordings that I was in the process of backing up.


thanks a lot to everybody who can help me
I hope I will be able to help some in return


Sébastien,
Montréal
QUÉBEC

[jono-ats]

The LaCie power supplies fail in many cases. You can order a replacement from LaCie, or find a local shop to break open yours, and replace the electrolytic capacitors. Maybe 70-80% chance this will fix it.

Jono

[seaccentaigu]

Thank you Jono,

Unfortunately, I tried a new power supply with exactly the same serial at my local shop and it was giving the same error signal (green light flashing as in the video).

I suspect the power supply to have caused a failure on the board of the drive.


Sébastien

[gonzillaaa]

I have issues with the same drive I've tried following the instructions above but my two drives from the lacie do not switch on when I plug them... I tried other enclosures and a separate sata-usb cable. I know it doesn't sound good especially since the lacie died with a "burnt" smell.

Any suggestions greatly appreciated.

[fzabkar]

I know it doesn't sound good especially since the lacie died with a "burnt" smell.

That sounds like a PCB fault, in which case a DIY repair should be possible.

Can you upload detailed photos of the component sides of each PCB?

[gonzillaaa]

I have actually taken the hard drives out of the lacie casing and PCB, the PCB is burnt underneath (A few capacitors are black) I am triying to connect the harddrives using a standard sata to usb cable, I'm using these:

http://www.amazon.co.uk/USB-SATA-IDE-CABLE-ADAPTER/dp/B003Q97FIW/ref=sr_1_3?ie=UTF8&s=electronics&qid=1282384699&sr=8-3


the drives don't start up at all, they do not spin... I am running ubuntu on a paralleles virtual machine inside os x neither os x or ubuntu sees the drives when I plug them in.

[fzabkar]

Can we see detailed photos of the HDD PCBs?

[gonzillaaa]

fzabkar thank for your question it actually made me look at he hd pcb and realised it also has a burnt capacitor. I assumed all along the damage was on the lacie device pcb.

you can see both HD here http://www.flickr.com/photos/gonzillaaa ... 81/detail/

Does anyone know the value of those capacitors? is it work changing them? how do I know if there is more damage on the PCB that is not visible?

What other option do I have? can I buy the same HD and swap the actual disc? how hard is it?

[fzabkar]

The part that has failed is a 5V TVS (Transient Voltage Suppression) diode. It normally sacrifices itself by going short circuit to protect the rest of the logic. There is also a zero ohm link (000) to the right and below. This acts like a fuse. If there is no other damage, the fix is to remove the diode with flush cutters, and replace the 000 link with a wire link or 2A smt fuse or picofuse. There is a vacant location directly above the link that you can use for your fuse.

The other diode (BUX marking code) is the 12V TVS diode. The X125F/16V part is a polyswitch. It should measure close to zero ohms.

If you wish to replace the 5V diode, you can substitute an SMAJ5.0A from Farnell, Mouser, Digikey. Make sure the stripe on the diode lines up with the + on the board.

My notes should help you identify the parts:
http://www.users.on.net/~fzabkar/HDD/HDD_ICs.txt
http://www.users.on.net/~fzabkar/HDD/TVS_diodes.txt

Good luck.

[gonzillaaa]

@fzabkar thank you for your detailed response, it sounds a bit daunting, perhaps beyond my skill level with a soldering iron. I'm trying to get the same HD on ebay etc.. and just swap the pcb so I can image them and do the whole mdadm dance.

if that doesn't work out I'll definitely have a look at your instructions.


Thank you.

[netw3rker]

@fzabkar thank you for your detailed response, it sounds a bit daunting, perhaps beyond my skill level with a soldering iron. I'm trying to get the same HD on ebay etc.. and just swap the pcb so I can image them and do the whole mdadm dance.

if that doesn't work out I'll definitely have a look at your instructions.


Thank you.

Pretty much every post on this forum related to these drives says you can't simply do a pcb swap as the ROM chip has to be transfered to new board. The side-cutter solution is probably easiest since it doesn't require the use of a soldering iron.

Good luck!

[gonzillaaa]

The side-cutter solution is probably easiest since it doesn't require the use of a soldering iron.

@netw3rker thanks for the pointer I did not know about the ROM.

I undertood something different from @fzabkar:

1 - I must removed the 5V TVS Diode ( I am assuming this is the bigger that looks more burnt on my PCB)
2 - replace the 000 link (above it) with a wire link or 2A smt fuse or picofuse (soldered I suppose) I checked farnell for these but there are a lot of options, not sure which one to get.
3 - the 12V TVS diode ( I am not sure where is this one on the PCB) Should I change/remove it as well?


Thanks for your help.

[fzabkar]

Only change or remove the 12V diode (BUX) if it is shorted, ie if it measures close to zero ohms. It is located near the edge of the board, near the mounting screw.

The 5V diode is near the Samsung SDRAM. Both diodes have a "+" on the PCB near their striped ends.

You can obtain a fuse from Farnell.

Littelfuse Surface Mount Fuses, N = 2A, S = 4A:
http://www.farnell.com/datasheets/48294.pdf

This photo clip should help you identify the parts:
http://www.users.on.net/~fzabkar/HDD/Lacie_TVS.jpg

[gonzillaaa]

@fzabkar thanks for the cristal clear instructions, I checked my 12v diodes today and they are fine. I just bought the fuses and replacement 5v diodes in case i want to keep using the drives.

Hopefully nothing else got damaged, I'll report back on how I get on.


Thanks again.

[gonzillaaa]

I soldered the fuses and voila! the disks fire up and mount good news. Now on the recovery of data front not so good news. I have followed the steps above and get unexpected results.

Code:

sudo sfdisk -d
[sudo] password for gonzillaaa:
# partition table of /dev/sda
unit: sectors

/dev/sda1 : start=       63, size=128648457, Id=83, bootable
/dev/sda2 : start=128648520, size=  5558490, Id= 5
/dev/sda3 : start=        0, size=        0, Id= 0
/dev/sda4 : start=        0, size=        0, Id= 0
/dev/sda5 : start=128648583, size=  5558427, Id=82
Warning: extended partition does not start at a cylinder boundary.
DOS and Linux will interpret the contents differently.
# partition table of /dev/sdc
unit: sectors

/dev/sdc1 : start=       63, size=  2008062, Id= 5
/dev/sdc2 : start=  2008125, size=974759940, Id=83
/dev/sdc3 : start=        0, size=        0, Id= 0
/dev/sdc4 : start=        0, size=        0, Id= 0
/dev/sdc5 : start=      126, size=   256914, Id=82
/dev/sdc6 : start=   257103, size=    16002, Id=83
/dev/sdc7 : start=   273168, size=    16002, Id=83
/dev/sdc8 : start=   289233, size=   256977, Id=83
/dev/sdc9 : start=   546273, size=  1461852, Id=83

sfdisk: ERROR: sector 0 does not have an msdos signature
/dev/sdd: unrecognized partition table type
No partitions found


The error above does not look good. Then when i try with mdadm:

Code:

gonzillaaa@ubuntu:~$ sudo mdadm -B -l 0 -n 2 --chunk=256 /dev/md0 /dev/sdc /dev/sdd
mdadm: Cannot open /dev/sdc: Device or resource busy
gonzillaaa@ubuntu:~$ sudo mdadm -S /dev/md0
mdadm: stopped /dev/md0
gonzillaaa@ubuntu:~$ sudo mdadm -A /dev/md0
mdadm: /dev/md0 assembled from 1 drive - not enough to start the array.


Finally just in case I try to mount and I get:

Code:

gonzillaaa@ubuntu:~$ sudo mount -t hfs /dev/md0 /media/bigdisk/
mount: wrong fs type, bad option, bad superblock on /dev/md0,
       missing codepage or helper program, or other error
       (could this be the IDE device where you in fact use
       ide-scsi so that sr0 or sda or so is needed?)
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

gonzillaaa@ubuntu:~$ dmesg | tail
[  699.561273] md: array md0 already has disks!
[  907.353665] md: array md0 already has disks!
[  931.569486] md: md0 stopped.
[  931.569502] md: unbind<sdb>
[  931.569643] md: export_rdev(sdb)
[  934.191469] md: md0 stopped.
[  973.182849] md: md0 stopped.
[  978.376954] md: md0 still in use.
[  978.389472] md: bind<sdd>
[ 1109.026855] hfs: can't find a HFS filesystem on dev md0.


What I seem to understand is that one of the drives is missing a filesystem, is there a way to install one without losing any data?

Thanks.