Hello,

I would like to know if a hard drive Serial Number is programmed into U12 chip (ROM) for Caviar drives (like WD400 and WD800, etc.)

After ROM dump to file for examination with hex editor, I can locate firmware version, but no data regarding serial number, either as plain ASCII text or encoded as hex bytes.

Thank you for any info in advance.

I think you'll find that they're in the SA. Download some firmware resources and scan the SA modules.

_________________
A backup a day keeps DR away.

Yes, but SA is damaged, not readable at all, because of defective heads. I have several donors for heads and access to HEPA booth, but I would like to ensure that the ROM chip is native for this hard disk.

Long story short, the owner tried to "fix" the problem by just replacing PCBs, obviously without success because of adaptives. After "researching" on the web, he just found out that the ROM chip (U12) must be replaced too, so he started to soldering/desoldering chip U12 available on every 2060-701292-000 PCB he got. So, now, I have a bunch of ROM chips, a bunch of PCBs, and a bunch of donors drives.

So, the objetive now is to match the patient hard disk (which is plenty identified because of their serial number printed on the case label), to a unique ROM chip (the original). After thoroughly analyzing the ROM dumps for Caviar drives the serial number is not embedded into the ROM (at least, not in ASCII or encoded in Hex-nibbles). It could be obscured or encoded in some other ways?

Any suggestions or ideas?

Thank you in advance.

these are not easy when it comes to head replacement, so i would recommend contacting a pro if the data is important. Actually, it is not the head replacement itself that's hard, rather getting the data out of it afterwards.
ROM does not contain SN.

_________________
Adatmentés - Data recovery

@jcarlosmor, please provide the following:

1/ Date of manufacture on labels of each HDD, patient and donors
2/ Date code on each PCB (WWYY, week/year).

(See "2407" at top left of Samsung DRAM in this photo: https://sep.yimg.com/ay/yhst-14437584971410/91710981-6.gif)

3/ Date code on DRAM ("731" in photo, YWW = 2007, week 31)
4/ All markings on each ROM IC.

By matching the date codes, we will hopefully be able to deduce which parts belong together.

_________________
A backup a day keeps DR away.

@fzabkar

Yes!!! I was thinking that route since datasheets for U12 chips (ST and Atmel) let easy idenfitication of manufacturing lots. Unfortunately, we did a very "serious" task to obtain almost similar donors, so we have all of the parts within a few weeks or months (pre-date to manufacturing). We even have two donors of the same date printed on label case.

In the process of learning more in-depth, I had used your great ROM analysis tool to check every ROM chip we have. Then, using LZHUF to decompressing PCMBlocks to obtain "real code and data". However, in none of the several decompressed PCMBlocks are embedded data regarding serial numbers.

I am starting to believe that the term "copies of the ROM modules from SA" is not a real copy ( I mean for 0Dh, 0Ah, etc.). For instance, a ROM module only contains the string "S/N: XYZ.......", and that string it is copied to SA, but then, the SA it is filled with the real serial number according to manufacturing stage for every drive.

So, my plan now is, at first, inspect every ROM to ensure they have the same firmware version, and then compare in binary if all necesary PCMBlocks are the same data ( I expect not to encounter minor revisions changes). Then after sorting the more appropiate ROMs, I would need to insert in Hex editor the necessary ROM modules 02, 0D, 03, 47, etc. Module 0A would also serve to sort the ROMs since the headmap must match the unit in question. Finally, after extensive test with a full-working HSA I would need to sort every adaptive set data until avoiding head-clicking noises.

Any ideas?

Thank you in advance.