Hello everyone
first of all, i'm sorry if my sentences will be confusing or not well written but my english is a little rusty so...

My story:

three days ago, suddenly, my IBM DeskStar 120GXP 7200 40Gb Cache 2Mb 3.5" P-ATA UDMA100 ATA5 IC35L040AVVN07-0 has stopped working. the bios of two different motherboards can't no longer recognize it, and the disk doesn't spin at all...
i've some old photos on it and it would be great to recover them, so i've started searching for a solution, in a limbo between big money-no stress-option of data rapair centers and small money-big stress-option of PCB swap...800-1000€ vs 20€...i've started thinking for the pcb swap (c'mon don't say that memories are priceless and i'm a heartless-nerdly-scrooge, i'm only a student!)...after "some little" researches i've spotted the right pcb, MLC, P/N, first two lines of the white label, all corresponding...but now...i find out that isn't enough...i'll probably (almost for sure, but i don't want to think to that now) need to swap the NVRAM chip....well. i think i can find where the chip is on the board, but the work....mmm...i've got good manual skills ( ) and i've already done some soldering works but this pcb is too small!

so here comes the questions:
do I really need to swap the nvram?
what could happen to the drive if i first try the swap off the entire pcb?
do i risk to compromise forever the data?
i'm going to lose both the pcbs?
are there any chances to get my data back?

thanks in advance

giorgio-italy

DIY attempts can spoil the condition of the drive further. So if data is important, save the drive till you have the money for a pro recovery.

Firstly , how or what made the drive to fail ?
Try to see if you can rectify the original PCB(post a pic ).

Depending upon the above you can think about using the donor pcb.
(if the original pcb is not burnt you can try the donor to see if your drive spins up).

If you are not sure, better take pro help than take the risk.

the disk simply didn't started again after the last shut down. no response, nothing. no handling or either voltage problems...simply it didn't boot again.
ok i understand that pcb swapping can "corrupt" the disk integrity, but how?
if i try and it doesn't work and then i swap back the pcb again?

you can try the PCB and check if it spins up and detects properly in BIOS.

i've been searching for TVS diodes on this pcb but the only diodes i've found are on the other side of the board.
named d1 and d2, one of them is slightly different, like "shiny"...
anyway, this is the front of the pcb.

https://www.dropbox.com/s/iuniin22kki2fx6/IMG_0668.JPG
https://www.dropbox.com/s/tg0rjy42rzzckqb/IMG_0669.JPG


the nvram is the LC66 chip?

little update,
a recovery center has just replied me saying that they could do the work for about 350euros...
not bad...i've already replied them that i've got an almost identical drive (i think that they will perform the pcb swap+nvram before opening the plates) asking for a little discount. "c'mon dude i'm avoiding you the research of a compatible pcb...easy money for you"
no seriously, what do you think a data recovery center first do to recover a drive?
diagnosis of course...and then? i think that in any case they need a compatible drive, for the pcb swap or the plates swap..or not?

most DR companies will have the boards needed to do the recovery already on hand. So providing them a part may not actually help them at all.

As for research.. they already know what they need.

since they said they can do it, then they already have what is needed most likely.

maybe but how?
i've found something like 20 different kinds of pcbs for the same hdd model...and it is 12 years old...
does exist a "pcb maximus alpha" wich can fit all the models? or simply easy reprogrammable chips mounted of custom pcbs? it's interesting...

i'm still evaluating what to do...if perform an nvram swap or not...and if yes, i need to find the most finest italian soldering artist.

Eh ma chissà quanto costa...
Yes but who knows how much it's gonna cost...

Years of experience gives the DR professionals knowledge of what is needed for the repair. That is why we are professionals

un po' ti aspettavo...tipo l'apparizione della madonna

I was waiting for you...any advice in particular to what to do here in our beloved italian boot?



i understand and i'm very fascinated by the work behind the data rapair. that's why i'm thinking on "how".

i think i'll go for the pcb swap, someone could confirm me what i've to desolder-resolder?

other question...if i try first a "brutal" pcb swap what could happen?
for example:
if i try first to swap pcb and i find out that it doesn't work, can i simply put back everything and try to desolder-resolder the nvram on the pcb?

Che a seconda del problema f**ti l'altro PCB o il drive. Ma magari no. Ecco perche' la diagnosi va fatta prima.

That you screw donor PCB or the drive itself depending on problem. But maybe no. That's why need proper diagnose.



Forse si (vedi sopra) ma magari a quel punto e' troppo tardi.

Maybe yes (see above) but at that point it could be too late.

so it's recommended swapping the nvram before the pcb?
anyway i'm not sure on wich is the correct chip!!

well,
thanks Splidit, i've searched on the net and i've find out that the nvram usually is an 8 leg component with a code wich starts with "25...".
the low profile chip next to the LC66 has a 25xxxxxx number on it...
i'm a little confused

25xxx is programmable EEPROM. This is not NVRAM. The info from Splidit is right

Little Update:
i've found a compatible pcb and i've swapped the NVRAM chip, the disk is alive!
i've recovered the files i needed, a few gbs with an external ide to usb cable, powering the hdd with a standard atx psu (the external little thing they gave me with the cable wasn't stable enough).
so, i can tell that in this case, searching for a donor pcb, with the same:

-p/n number, on the disk label
-MLC Number, on the disk label
-first two line of the pcb number, on the pcb label, above the ide connector
-same assembly place (made in china in my case)
-max 30 days between the production date

in my particular case
-the motor controller was different,
-the main controller has the same first line number 47P2010 but the second numerical line was different for the last third digit (original was: 1J06T02MLQ donor was: 1J06T02QLQ)
-the other big component, i think is the memory, but i'm not sure, is different:

DONOR:__________ORIGINAL:
IBM______________IBM
TAIWAN__________TAIWAN
SPRCBM BB_______SPRCBM BB
50G7143__________50G7143
1F10076TQT_______1J14002TQT

WELL,
i don't really know if i was lucky (for the first time ever) but i can say that, IN MY CASE, the pcb swap, with nvram swap, has worked.
thanks everyone for the help and the suggestions

bye
giorgio

Good job.
And definitely lucky..