TOSHIBA MK3252GSX HDD2H01 B UK01 S

[PrinceOfAbyss]

Guys I made some theoretical progress.

First of all, I found the datasheet of my EEPROM. It's this one. So, according to the datasheet, I must have cut off the VCC and HOLD# pins.

Supposing that the EEPROM has an on-chip pullup, I think there will be no problem to leave the HOLD# pin disconnected. The only problem is the VCC, which obviously can't be disconnected. Also, from what I know, most of the digital inputs of such chips have a couple of internal protective diodes, one for the VCC and one for the GND. Now, if I connect a 3.3V+0.7V=4.0V to one of those input gates, that would trigger the diode to the VCC, and the VCC itself would present a current of 4.0V-0.7V=3.3V which is the operating current of the chip. A gate like this on my chip is W#.

All this (if successful) will help me read the content of my EEPROM. Then, I can easily re-write it to an empty EEPROM, and install it on my PCB.

I already ordered this EEPROM SPI programmer, and a couple of these adapters to help me experimentise safely with my EEPROM.

Now this leaves me only to find a couple of good empty compatible EEPROM chips to buy. Do you have any recommendations for me?

Guru BlackST, any help please? (Forgive me the rest of you if I mention BlackST in particular, it's just that I think he's much into electronics )

[msurgeon]

I applaud your persistence in this endeavor. But what possible advice is there left to offer you in this matter? You are now spending money on equipment which could have better gone toward recovering your data in the first place; it wouldn't have been as expensive as you seem to think. By the time you're done with this science experiment (which, in my humble opinion, really has little chance of success) you may well have expended more time, energy and money than it would have cost for professional recovery. However, you seem to have an excess of time on your hands so I wish you success.

[BlackST]

What help can I give ? I have everything, but HERE ... I don't go "by trial and error" ....

[PrinceOfAbyss]

OK, let me just ask for a confirmation, since I mostly researched a compatible EEPROM as well...

So, is any of the following two compatible with my EEPROM, in terms of copying the content of my EEPROM to one of them?

Atmel AT25DF041A

SST SST25VF016B

[BlackST]

First you have to read YOUR serial prom. If you can, of course

Then you can go on.

Just curious : all this hassle must have a reason. WHY ALL THIS ?

[PrinceOfAbyss]

By the time you're done with this science experiment (which, in my humble opinion, really has little chance of success) you may well have expended more time, energy and money than it would have cost for professional recovery. However, you seem to have an excess of time on your hands so I wish you success.

Yea, I have so much time on my hands because I am 20 years old, live alone in a deserted island, eating coconuts and drinking sea water...

Well, no Sir. I am 31 years old, have a day job, a wife, and a small daughter, and I do this research in the expense of my night sleep, just because I can't afford the 1000+ EUR that the pro I contacted asked to do the heads swap (when I hadn't even put a single finger on the disk to do it - or, try it - myself).

And since you care so much about my wallet, let me inform you about what I've spent so far. Well, two donor disks costing 70USD each, the SPI programmer that cost me 27.50EUR inc. shipping, a tweezers with extra fine tips that cost me 5EUR, and 4 SOIC adapters that cost me 17EUR inc. shipping. To sum it up, 140USD ~ 120EUR + 27.50EUR + 5EUR + 17EUR = 169.5EUR. So answer me this: What pro would ask for that money, not to recover the data.... just to take a look at the disk...?

What help can I give ? I have everything, but HERE ... I don't go "by trial and error" ....

You don't go by trial and error because you are a pro my dear BlackST. Of course you've gone by that way a long time ago when you were learning it yourself, but now you have the knowledge to play safe. I'm no pro however, and unfortunately I can't pay you (though it would have saved me a LOT of time if I could). So, I'm down to try whatever I can with my limited knowledge, very limited hardware resources, and also average finances.

Just curious : all this hassle must have a reason. WHY ALL THIS ?

Because I can't accept that I'll loose about 40 pics of my daughter that I don't have in any backup, all vacation pics of me and my family, some very important PHP script programming I've done for a forum I administer, a dozen of movies, a lot of music, etc...

All that is written on those two damn platters and I can't have access to them... It's something that I just can't accept. And because I can't pay you to do it for me. Simple, isn't it?

[BlackST]

To be precise : I had access to pro equipment when I was still "learning" : I didn't start doing anything until I had the exactly same pro equipment I was familiar with.
No errors were tolerated when I was learning, as I were under other people's responsibility, now I tolerate no errors even from myself now : when I am not sure, I simply don't do it or refuse to do it (happens few times, but STILL HAPPENS). If I don't have the necessary premises, I get them first, then when I am totally sure I go - not before. Sometimes you don't have 2nd chances.
I think (but I have no other experience so maybe I can be wrong) this was a great way to build up self-confidence - no, I was not scared, I took it as a challenge.
Maybe it's excessive, but we say "better to be afraid then cry later" , especially on irreplaceable data (memories, work etc.).
This is what I think.

[fzabkar]

Also, from what I know, most of the digital inputs of such chips have a couple of internal protective diodes, one for the VCC and one for the GND. Now, if I connect a 3.3V+0.7V=4.0V to one of those input gates, that would trigger the diode to the VCC, and the VCC itself would present a current of 4.0V-0.7V=3.3V which is the operating current of the chip. A gate like this on my chip is W#.

All this (if successful) will help me read the content of my EEPROM. Then, I can easily re-write it to an empty EEPROM, and install it on my PCB.

That will be a stroke of genius if it works, and I hope very much that it does. However, you are very much on your own now. Most of the DR guys appear to come from a software background, and most DR work is done from behind a keyboard.

Before you try your "hack", I suggest that you try the same thing with a working EEPROM. Also, when substituting EEPROMs, be aware that each device has unique Manufacture and
Device Identification bytes. This may or may not be significant.

[jono-ats]

Most of the DR guys appear to come from a software background, and most DR work is done from behind a keyboard.

Not everyone.

I suspect a lot of DR guys are quiet because they saw a photo of the remains of the EEPROM. There are significant challenges to getting good connections onto this device. Maybe it can be done . . .

I have to give the OP an A for attitude.

[dobrevjetser]

Well, i must honestly say i had such a situation when i was a student and made my thesis.
I had a circuit where for some reason the Vcc to one of the IC's was not connected, but it WORKED
The reason was like you said that the IC was fed via one of the clamping diodes, so your idea is not so insane as one could imagine.
Anyway, if you blow the eeprom up, then you're really in big trouble, but if you succeed i'll say 'chapeau'

Good luck,

Dobre

[BlackST]

I have tried with a common 25P05 with my universal programmer/tester (HiLo Systems) not to supply VCC and it doesn't work, maybe with different programmer can be.

If not, I remember some time ago I made something for a desperate case (broken pins) and I had to work under microscope with chemicals + microbit to expose some contact area. A microprobe was used to make a momentary contact (their use is at chip fabrication/test, a little bit difficult to find and somewhat expensive...) and it was a pain in the ass and balls to make everything work, but finally I got the data. It was another SPI rom 25xx.

And now, sorry to ruin the day : if you know what to do, you can do even without YOUR SPI flash. But this is another story.

[PrinceOfAbyss]

The second donor disk is here. I'll remove the EEPROM later today...

I still have to wait for the programmer and SOIC adapter but it's a start... and definitely the hardest part for me to successfully desolder this little 5mm x 6mm thing...

Oh, BlackST, I wonder why you couldn't perform my proposed way to read your EEPROM. The diodes are there, I've checked it... Maybe you missed something with the extra power? As I said, you have to give it 4V, not the 3.3V that the programmer feeds it with!

[BlackST]

I don't have the time to tinker with programmers and Vcc - In any case I even don't need original flash in case of emergency, it takes a lot more of work but who cares ? (if and only if it is rewarded later).

[Odiferous]

"I also have one good and one bad news:

The good: The heads swap was performed flawlessly. I really did a pro job even though I'm not a pro..."


How do you know it was performed flawlessly?

I doubt it very, very much indeed. I spent almost two years practising head swaps before letting myself loose on a live client drive.

I don't for one second believe you managed a flawless head swap - no way.

[Alexii]

"I also have one good and one bad news:

The good: The heads swap was performed flawlessly. I really did a pro job even though I'm not a pro..."


How do you know it was performed flawlessly?

I doubt it very, very much indeed. I spent almost two years practising head swaps before letting myself loose on a live client drive.

I don't for one second believe you managed a flawless head swap - no way.

My very 1st head transplant was a success. And the funny part - it was the dreaded WD BB. Toshibas , like it was said above, are forgiving. Its a good chance the swap was ok. What i cant belive , is that it took u 2 years of moving heads to get comfortable.

[PrinceOfAbyss]

What i cant belive , is that it took u 2 years of moving heads to get comfortable.

Maybe he suffers Parkinson...

[PrinceOfAbyss]

Hello to all again.

Did you miss me? I'm not dead, I didn't give up, I was here waiting patiently for some hardware I had to order to arrive.

Basically, the programmer I initially bought was not appropriate for my chip (EEPROM programmer, for an SPI Flash chip), so I had to buy this one: True-USB PRO GQ-4X Willem Programmer.

Anyway, the programmer arrived from Canada yesterday, and I immediately tried to read the donor's chip (an ST 25P40V6). It read it like a breeze, rapidly, and without a problem…

Then, I tried to read my patient chip (a Spansion S25FL040A006) using the method of the external power supply I mentioned somewhere above, but didn't work. I initially tried 3.5V. Then I increased the Volts incrementally up to 4.1V. The chip however showed no sign of life.

Then, I had an idea. Since the method of the external power lies on the existence of a diode between the W# and Vcc pins, I could check for that diode on the donor's chip and get a conclusion for my patient chip as well (since they are compatible). Let me remind you here that on my patient chip I had checked with a multimeter and found out that there is a diode between W# and GND, so I assumed that there had to be a diode between W# and Vcc as well. So, I checked the donor's chip, and to my great disappointment there is no diode between W# and Vcc, although there is one between W# and GND. I fell in one of those very rare occasions, since 99% of the times, if there is a diode between a digital gate and GND, there is another diode between that gate and Vcc as well.

Anyway, since the diode was not there I stopped trying the method with the external power, and had a brainstorming for other options I had available... One of those that came on my mind is this: Since I destroyed the two pins of the chip (Vcc and Hold#), and even scratched the corner of the chip to try to find the traces of the wires that connect to the chip's core (but didn't see them obviously due to their microscopic size), if I could use a sort of liquid electrically conductive material on that little corner, and put a wire there, I could replicate the missing pads. My question was whether there would be a problem to short Vcc and Hold#, but the chip's datasheet proved handy on this... So, according to it, Hold#:

Pauses any serial communication with the device without deselecting it. When driven low, SO is at high impedance, and all input at SI and SCK are ignored. Requires that CS# also be driven low.

After Once I read that, I realized there would be no problem at all to short those two pins. So, I bought a bottle of that fantastic material that is used to fill in the gaps of the rear windshield's defogger, that is obviously electrically conductive. So, I desoldered my patient chip from the SOIC adapter I had it installed, and cut out the top right corner of the adapter's PCB for better manipulation of the area, then soldered the chip back, and glued a piece of kinar wire using instant glue on the chip's top. Once it was secured there, I pilled the wire's insulator on that end and brought it as close as possible to the vertical side of the chip, where the traces should be. Then, applied two-three hands of the conductive material, making sure it covers the whole vertical area of the broken corner, and once it healed, I tried to read my chip with the programmer... And it read it... Guys, it READ it! I still have that smile on my face! I'm really so happy that I got over this difficult part.

Now, the process is more or less known to me... I'll write the contents of my patient chip to the donor's chip, solder the donor's chip back on the donor's PCB and read my data (assuming I performed a flawless head transplant, which I did anyway)...

You can see some images of my experiment below:



















Of course, I'll keep you informed about the rest of this saga.

[PrinceOfAbyss]

Now the difficult, and one million dollar question goes to BlackST.

I can't take out of my mind your phrase: "if you know what to do, you can do even without YOUR SPI flash. But this is another story."

Since, I now have my SPI Flash and its original contents, and you will not reveal anything really useful, and I'm not even in the business either, could you tell me, in a PM of course, about your procedure to read an HD with another chip? This is just out of curiosity, and because I love learning things, not because of anything else. And of course if you decide to share it with me, I promise to keep it for me.

Actually, I used a Hex comparison tool to try to spot the differences between the two files, but there were many of them. Actually, some small parts was totally different, some others where there on both files but offset by some lines below or above one another, etc... I'm really curious about this...

One obvious area is the one mentioning the serial numbers of the disks... Don't tell me it's THAT easy, that is to change the serial number of the donor with the patient's one, and it'd work...

[BlackST]

No comment.

Last edited by BlackST on July 23rd, 2010, 13:29, edited 1 time in total.

[PrinceOfAbyss]

Something like what? An exchange of knowledge for solving my question? If that's the case, me being a programmer, I can teach you some Visual C# .NET...