I'm also looking for tips on dealing with ROYL series drives when the PCB ha failed. When I've been able to access the embedded ROM it has been straight forward enough, but I am still struggling on drives where the PCB has been very badly damaged and beyond repair.

I see on the PCB revisions of the early ROYL series drives (e.g. WD5000AAKS) the MCU surface mount has pins down the side. I have had some success in removing the entire chip and remounting it on to an appropriate pcb.

Newer ROYL drives (e.g. WD6400AAKS) have no pins down the side of the MCU, consequently de-soldering and remounting is much more difficult.

I was wondering if anyone had any tips on how best to deal with this problem?

Many thanks,
John

If you can access the SA then you can rebuild the ROM from SA copies of ROM modules

_________________
You don't have to backup all of your files, just the ones you want to keep.

Hi John

MCU with no pins visible = B.G.A. (Ball Grid Array) and requires specialist (de)soldering tools to remove and replace. Even an X-ray machine to view solder balls to ensure soldering is ok!

If the Marvell SoC (system-on-chip) which contains the ROM is damaged itself then the ROM content must be rebuilt (information can be taken from target drive's system area using a PCB programmed with a special firmware), or in some cases a donor PCB that has similar adaptive configuration.

If the Marvell SoC is not damaged but the normal ATA commands cannot be used (e.g. PCB is U/S) then I suspect JTAG is the only way. If we had the datasheet it would identify the JTAG pins but since they are brought out to pads by WD it's not hard to work out which ones they are. However a Boundary Scan file is needed to take the hard work out of discovering internal chip resources.

SB

I've not quite got the skills or equipment to deal with BGA soldering yet, something I am going to have to consider over the next few months I think.

I've got the patient drive spinning with a good match donor. I have SA access at the moment and as I am typing this message the firmware modules are being backed up (very slowly). I guess I will try to regenerate the ROM using these. However, one of things I have noticed in the firmware backup is the high number of errors. I am not sure if this is because I am using a donor PCB or that the drive is in "slow speed," or that there is genuine corruption in the SA. Either way I am a little concerned that when I come to regenerate the ROM I will not have the appropriate working firmware modules to do so?

Which modules are critical for ROM regeneration?

Thanks for the help,
John

_________________
CDR - Manchester Data Recovery Services
0161 408 4857
http://www.cheadledatarecovery.co.uk/

Sent you a PM.

_________________
You don't have to backup all of your files, just the ones you want to keep.

Just to let you know that I was successfully able to access the SA using a donor pcb and then regenerate the ROM using several of the modules. I did note that before the ROM regeneration the ATA module was showing as corrupt, but it checking out as OK after the ROM regeneration.

BGA soldering can wait for another day.

I was working on a WD6400AAKS-00A7B0 / 2060-701537-003 REV A - If you have any firmware requests please ask.

John

_________________
CDR - Manchester Data Recovery Services
0161 408 4857
http://www.cheadledatarecovery.co.uk/

@cheadledatarecovery

Did you find a way to restore broken ROM in Marvell 88i6745n or atlast way to boot from
U12 external EEPROM 24p20?

good work stuff

_dex_, please contact me