Hello all,

This being my first post I apologize if I brake any rules,

Due to a falty Y splitter cable I accidentally powered 10 HDDs with reversed polarity (the 12V and 5V lines).
BTW: they were in a RAID6 but the array was shut down cleanly

Upon checking all of them in an extenal enclosure known to be good none was recognized by the os (linux). Some dis spin out some did not.

I understood that modern drives have protection on the PCB.
Q: do the WD30EFRX have such protection?

I have the opurtunity to replace some of the PCBs with some from the same batch.
Q: I am worried if the new board would mess up the data!

I would like to start up troubleshooting but would need hints where to find more info or which threads/posts to start up reading
Obviously the schematic of the PCB would be good but I do not think there is one.
Besides if all my efforts would fail I have to get asistance from some data recovery experts and do not want to make things worse

Thanks

TVS Diode FAQ:
http://www.hddoracle.com/viewtopic.php?f=100&t=86

Catastrophic failures in Western Digital PCBs:
http://www.hddoracle.com/viewtopic.php?f=59&t=1119

_________________
A backup a day keeps DR away.

That's a very expensive mistake. Even a quick look down the board will show it's also one of the most common topics posted. You only need 8 working drives to rebuild so at least that's something.

How many drives aren't spinning up, have you checked the PCB's sometimes the more serious damage is very obvious.

_________________
Data Recovery Services in the UK.
https://www.usbrecovery.co.uk/

@fzabkar really appreciate your links. after reading them i have to investigate all 10 drives. i have some pictures if needed but sofar there are at least 2 different pcb. i looked at only 2 drives til now. both have the D3 in short and at least one has the R67 around 250ohm instead of 0ohm. found the model of a diode from stmicro SMA6J5.0A marking 6au

Apologize for double posting if the mod does not cach it before. was tiping on my phone and hit the 'submit' button in middle of sentance.

@fzabkar
Thanks for the hints. I am in the process of evaluating the problems: i.e. D3 and R67, damages, behaviour of each drive (spins or not, usb adapter lights on/off/blinking). Will come back with details and a couple of pictures and probably a lot of questions. Btw I found the external ROM is a winbond chip, so just in case I will end up with donor pcb will be replaced or replicated

@lardman
Expensive? not only the value of the hardware (the mobe does not boot as well) but mostly the content. Yes I know I need only 8 out of 10 but after such disaster if I can have them all running the better. I was in the process of moving everything on new drives and wanted to do too much (new sas3 raid card, some SSD for boot a fresh os, optimizing the alignament of RAID/LVM/FS block/extent/chunk sizes, etc). As mentioned abouve I still have a lot to do. will be back with details including how many spin or not but withoud replaceing the damage components...

Hi.

I would number each HDD, it's PCB and not power them until analysis.

1) remove the PCI (never open the HDD itself!). Take photos from both sides of the PCB with enough resolution to read components and see cracks/damaged components

2) I've checked your PCB modem and it's an external ROM. I would desolder each one and save it's data with a cheap EEPROM reader (such as ch341a with the correct voltage adapter). Do not mix them! Number each one.

3) solder ROM back, use a multimeter to check if is there any shorts on 5v and gnd and 12v and gnd. Cut a SATA power cable to make measurements easier and check pinout on Google.

If no short is found and there are not other visible damaged components on the board I would measure diodes, and resistors (check the low ohm current sensores near the motor driver).

If a short is found try to desolder the TV's diode. And measure again. If short is gone maybe ok to power (again, inspect the above itens too). I would buy a new tvs and put it on the board before power up just to always have protection.

do not use the same y cable.. before testing use an old HDD and check if the PSU and motherboard are Okay.

For the boards that fried you can buy used replacement and solder the original bios chip in them.

If the bios/ROM chip is gone you are in deep water here.

Try not to make things worse

@ercorreia unfortunately as per my first post I already power thrm up using an external enclosure. Thanks for the tip on CH341A. Was about to build one based on one of the microcontroller boards I have. But that involved some extra work mostly writing software. BTW there is a CLI for linux to use the CH341A.
- there ar no hard shorts (nothing below cca 600ohm)
- don't have a good board to compare
- same cable 'duh'. threw them out. probably from now on will double check every cable I buy
- psu and mobo are on the list but for now not use them until fully cheched.
- waiting for replacements PCBs. They are coming from overseas will be here in a month or so.
- even if only one is working I can use it to read the data from each drive and transfer it to a new media. then get to the next drive and replace the bios etc,,,

Out of the 10, 3 spin and other do not. All had the 5V TVS shorted and the 0ohm resistor showing at least 250ohm instead of 0ohm

Sofar I managed to remove the TVS from 2 drives (one that spins and one that did not) replaced the 0Ohm resistor and tried them in the external usb enclosure.
Results were negative:
- drive that did not spin continued to exhibit the same behaviour
- drive that spun:
-- starts as normal
-- a couple of head moves (from the soud of it)
-- after about 10s the moteor shuts down.

Still not recognized by the os.

I am waiting for parts (flash, resistors and TVS diodes)
It looks like the TVS diodes (D3) are
- SMA6J5.0A marking 6au (maybe 6UA have to check again)
- SMA6F5.0A marking SUA (package SMA flat)
The 0ohm resistors (R67) are
- 1206 and 0805 as size

As mentioned befor I can post piuctures of the boards

Any other hints of finding the issues on the boards would be appreciated.

Thanks

Linux command line to read/write using ch341a = flashrom.org

I'm conservative so I would wait untill new boards come in. Meanwhile I would desolder and backup the "BIOS" IC (SPI Flash chip).

For the 0ohm resistors you can solder a wire. Make sure top markings are really 0ohm and not 0,something ohms.

For the TVSs you can remove then.

This 10s spin + normal head moving sounds + turn off can be just the HDD powering down to save energy IDLE.


If you really can not wait use this one that is spinning up and replace the 0ohm resistores for a short circuit.

Can you send pictures of the boards^}

Those that spin - are they recognised by the bios and do they show the correct metrics or not? If so, leave them alone for now.

Spinning then stopping for these is normally a sign of internal problems- How many of those do you have?

It sounds like you have the right idea and the technical knowledge, you can pull the roms, fzabkar has written a WD tool to check basic integrity but there's little you can do if the problems are internal.

Do you have enough slack space to image the drives rather than working live?

_________________
Data Recovery Services in the UK.
https://www.usbrecovery.co.uk/

Still waiting for parts and tools. After receiving them I can try some hi res photos.

@lardman None is recognized on a external USB. I am a bit reluctant to stick them in a motherboard.
Thanks for the info about normal behaviour.
Q: where can i find the tool you mentioned (written by fzabkar)

@ercorreia Found a couple of other projects involving ch341a (SNANDer and CH341A-Pro both on github) I believe flashrom is the more mature.
I received a CH341A programmer and tried on a hdd board i sacrificed (not one of the 10) from a WD Re 2TB just because it had diode and jumbpers to harvest and the flash was from the same company although a bit bigger Pm25LD040 instead of Pm25LD020 although some of my EFRX drives may have both types.
Replaceing with a bigger flash will be ok I assume. Same family just 2x bigger. Skimming throug both datasheets it seemed to me that they are identical except for the capacity.

@
CH341 needs level shifters or mods for 3.3V
But OTOH I had tested on the WD Re 2TB board and it seem to work ok. I mean I used a SOIC8 clip to perform read/erase/write/verify in system without desoldering the flash. Anyway I prefer to use level shifter on the damaged boards.

Ordered a preheating platform and found also very low temperature (138F) solder wire which I intend to use such to reduce overheating risks if reading InSystem will fail.

Questions:
- Any idea how to test the heads and/or preamp (I suppose there is one inside the drive)
- Does the flash contains all the FW for the controller? Data for the Marvell uC not found (sad)
- what other data is contained in that flash?
- Info about the SMOOTH ic?

The folowing is a very generic question:
- Assuming you do not care about the data would a drive be usable just by replacing the PCB with the same type PCB w/out swapping the flash.


Thank you guys

My WD ROM verification tool:

http://www.users.on.net/~fzabkar/FreeBasic_W32/WD/wdROMv17.exe
http://www.users.on.net/~fzabkar/FreeBasic_W32/WD/wdROMv17.bas

Here are two primers which I hope will explain why simply swapping PCBs most probably will not work.

The hard drive -- a computer-within-a-computer:

http://www.hddoracle.com/viewtopic.php?f=56&t=2600

"Adaptives" -- why a straight PCB swap doesn't work in modern hard drives:

http://www.hddoracle.com/viewtopic.php?f=56&t=2600&p=20704#p19090

_________________
A backup a day keeps DR away.