MK6034GSX PCB Swap

[Reg]

Hello,

I have a dead PCB on a Toshiba 60GB hard drive that I need to recover some data from if possible.

The drive does not power on when plugged into power. I found burn marks on the pins where the SATA cable connects to the board.

I have read that you have to move the cpu to a donor board in order for it to work. Is there virtually no chance of getting duplicate drive and the pcb working without swapping?

I have attached a jpeg of the pcb. Can you tell me what is the cpu that needs to be swapped to the good pcb and how close of a model number I need to get?

http://i850.photobucket.com/albums/ab66 ... CN0259.jpg

Here is the drive info:

MK6034GSX
HDD2D35 H ZK01 T

Thank you very much,
Please let me know if you need anymore information to help me.

Reg

[fzabkar]

There is a fuse, F1, and a TVS (?) diode, D1, near the SATA power connector.

If the fuse is open, bridge it. If the diode is shorted, remove it.

Alternatively, you can purchase a 2A smt fuse from Farnell, Mouser, Digikey.

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

If there is no other damage, then the drive should work OK, but it will no longer have any overvoltage protection. Be certain that your PSU is good.

Before powering up, you may like to measure the resistance between the +5V and ground pins at the power connector, in order to confirm that there are no other shorts or excessively low resistances due to other components.

BTW, unless there is another chip on the reverse side of the board, the "adaptive" information would be stored internally within the Marvell CPU (big "M" logo).

[BlackST]

There is a fuse, F1, and a TVS (?) diode, D1, near the SATA power connector.

If the fuse is open, bridge it.

This is to ensure proper destruction of something else in case there's an initial damage on the ...######... as common people have of course at their disposal current limited dual output power supply, usually... (see schematic) ??

Just like when the fuse blows on the TV / power supply due to something elese beyond the fuse and average joe takes some aluminum cooking foil and wraps the fuse , replace it... and BOOOM... - Alternatively : "let's try to make this TH in conduction state FORCEFULLY" ...

DIY (D for Do / Destroy) at your own risk.

[Reg]

So for the fuse, I can bridge it by dabbing a small amount of solder on it, correct?

If that doesn't work, removing the diode on D1 could also allow the drive to work long enough to recover the data if it is still there?

The reason it blew is because someone spilled liquid all over the macbook and they tried to power it on! Duh! Not very smart (and it wasn't me, they brought it to me to fix it).

Thanks for all the help!

[BlackST]

Quote :

So for the fuse, I can bridge it by dabbing a small amount of solder on it, correct?

NO.

[drc]

"given an existing problem, it may be better not to do something, or even to do nothing, than to risk causing more harm than good."

[Reg]

Any other suggestions, not doing anything really isn't an option.

[dobrevjetser]

Doing nothing is better than destroiing the data.

[Reg]

How about testing the fuse first?

If I use a volt meter can I test it by putting putting the red on one side and the black on the other side? Should that tell me if it is dead?

[dobrevjetser]

For a fuse polarity makes No difference. Put your meter on ohms and it should indicate something near to zero.
If more than a few ohms fuse is bad.

[BlackST]

If you don't know how to handle a multimeter and measure a fuse, how do you want to put hands on a pcb with SMD components ?

[ppumkin]

How about testing the fuse first?

If I use a volt meter can I test it by putting putting the red on one side and the black on the other side? Should that tell me if it is dead?

If its a German multi meter and testing the fuse is bad - it will tell you "kaputen schlaussen"

[msurgeon]

First, decide how important the data is. If it's not life-or-death data then try bridging the fuse as suggested. Testing a fuse with a multimeter is probably explained in the first couple of pages of the meter manual. Understand, though, that if it's not the fuse - and if you do no further damage by shorting the fuse- that the recovery will not be cheap because it involves de-soldering the large Marvell chip from the PCB and re-soldering it onto an identical board. The process is not done by all DR shops so inquire specifically if the place you choose has the capability of BGA soldering. If the Marvell chip is damaged then there is no hope of recovery.

[BlackST]

Quote :

If the Marvell chip is damaged then there is no hope of recovery.

Not true.

[msurgeon]

Quote :

If the Marvell chip is damaged then there is no hope of recovery.

Not true.

Well, if you can recover from this situation then you've unearthed a valuable bit of information that, although you won't share it, I'm sure the Nobel Prize committee might be interested to know the next time they hand out the prize for data recovery. And don't tell me that you would just try to keep searching for a board that had close enough adaptives. I don't consider that a practical solution.

[ppumkin]

BlackST is up for a noble prize Nice work!

[Reg]

Well the fuse tested good. So no other suggestion before BGA soldering?

[fzabkar]

How much resistance do you measure between fuse and ground?

[fzabkar]

BlackST is up for a noble prize :D Nice work!

Did I miss something, or is merely claiming that you can do something proof enough that you can do it?

[Reg]

Almost none. It started at about .08 and went down from there. Tested another similar board and got the same results.

Also, had nothing to loose so tried with it connected to power and had 10 volts running thru the fuse on both sides.