WD drive mounts but stops responding

[JGAN]

I'm having the exact same issue as was described in this thread here: https://forum.hddguru.com/viewtopic.php?f=1&t=32501

My drive is a WD20NMVW-11W68S0. In USB mode, it is detected by the computer and mounts, but after a few seconds there is a click and then I can no longer access the drive.

I removed U14, disconnected the USB-to-SATA bridge IC, and soldered on a SATA port; and it is recognized by the computer but I still am unable to get it to stay alive long enough to ddrescue it.

I don't want to attempt slow fix until I can get it to stay alive for a predictable amount of time -- unless someone advises differently. I will see if I can dump module 01 and 02 though.

What is the best way to proceed from here? How can I determine which areas of the drive are likely causing the issue?

[JGAN]

From http://www.hddoracle.com/viewtopic.php?f=117&t=1737:

Sometimes it's possible on some drives to disable one head (or more) on RAM to allow for proper drive working operations (with the other available heads) and to clone the drive.

Can someone elaborate on if there is a way to do this using a conventional PC? Or will it require hardware assisted cloning tools?

[MindMergepk]

i suspect a faulty/weak head which is common on these passport drives.
ddrescue did't work at all ?

[JGAN]

No, drive completely stops responding.

[F_ARG]

As MindMergepk said it could be a head problem. Do you have LBA number where DDRescue hangs? It is always on the same place?

If your problem is a weak or faulty head you may need a hardware assisted clonning tool. And if you want all your info you may be need to change heads too.

Best!

F

[JGAN]

ddrescue will hang as soon as the drive stops responding, so sometimes it doesn't even make it far enough to start.

[F_ARG]

From http://www.hddoracle.com/viewtopic.php?f=117&t=1737:

Sometimes it's possible on some drives to disable one head (or more) on RAM to allow for proper drive working operations (with the other available heads) and to clone the drive.

Can someone elaborate on if there is a way to do this using a conventional PC? Or will it require hardware assisted cloning tools?

I use to do it using hardware assisted cloning tools (PC3K). I think there is no easy way to do it with a regular PC.

On the orher hand, i think you need to do slow responding fix (based on what you described) and see if it there is not another problem behind slow responding error, but I recommend that if you really need you data you better go to a pro.


Best!

F

[JGAN]

From http://www.hddoracle.com/viewtopic.php?f=117&t=1737:

Sometimes it's possible on some drives to disable one head (or more) on RAM to allow for proper drive working operations (with the other available heads) and to clone the drive.

Can someone elaborate on if there is a way to do this using a conventional PC? Or will it require hardware assisted cloning tools?

That procedure/guide was done using hardware assisted HRT-DRE.

It have one HARDWARE PCI CARD. It's NOT software only.



All of those tools are hardware based like PC-3000, MRT Pro, DDI4, even Salvation Data - Data Compass have hardware on it.

I think disabling USB was a BAD MOVE because of POSSIBLE ENCRYPTION.

I would attempt to do the "Slow Fix" by USB as described on you mentioned thread with HDDSuperScript tool.

The drive is encrypted but reallymine will decrypt it with no problem if I can get an image. Worst comes to worst, it is possible to return the PCB to its original state.

I will try to dump mod 02 and 32 to see if it looks like slow fix needs to be applied. However, the main issue I am having is that I cannot get the drive to respond long enough to access it and read/write to the SA. Does anybody have suggestions for a workaround?

[maximus]

First, if the data is important then you should seek professional recovery and not try yourself. With that being said...

Now that the drive is SATA and it is seen by computer, but then quickly becomes totally unresponsive and does not come back without a power cycle, it is likely going into device fault. If it was just the slow issue it should still read very slowly in between long hangs. If this is the case, the next question would be, is the device fault happening due to the slow issue, or is it because after the drive becomes visible to the OS, a sector is read that is causing the fault. If it is because of the slow issue, you are probably all done as you would have to catch the drive very quickly after power up trying to apply the slow fix before it faulted.

If it is because of a sector being read by the OS (and trust me, both Windows and Linux will try to read some sectors sometimes for unknown reasons), you may be able to at least see if this is the case. HDDSuperTool has a direct mode you can try. You must follow the instructions in the manual for hiding (disabling) the drive, and the BIOS must be in IDE mode. You might get lucky and be able to connect to it long enough to apply the slow fix to see if that helps. But I will admit, it is a long shot, and it is very likely that you will have a difficult time recovering this drive by yourself.

[JGAN]

I tried to use hddsupertool but the direct IO DMA is only available in the paid version... Which has not been released

[JGAN]

-direct returns

Code:

hddsupertool 1.7-1.6 20160515
Overwriting (null) with new indentation...
Cannot open (null) for writing (Bad address).
Total program run time = 0 seconds


EDIT: Nevermind got it to work

[JGAN]

The drive stayed alive long enough to dump mod 02 and 32 in direct mode, but I think it died after I tried reading the modules. I checked them out, 02 looks like it needs cleaning but 32 looks clean. Attached is what I found.

Thoughts?

Attachments

dump.zip

(223.96 KiB) Downloaded 479 times

[maximus]

Code:

Data record #27:
61a: 00 01 02 01 80 0c 00 00 20 03 00 00 00 00 32 00 ........ .....2.
62a: 00 00 64 00 00 00 1e 00 00 00 05 d6 06 c0 12 00 ..d.............
63a: 00 19 00 00 00 01 01 09 03 00 00 00 00 10 00 04 ................
64a: 10 00 10 01 00 00 00 00 32 00 00 2c 01 ........2..,.


Offset 2 of record 27 has a value of 02, which can be an indication of the slow issue. Patch mod02 and then power cycle the drive. Don't worry about mod32, it does look clean.

[JGAN]

I will try to patch mod 02.

Question, is it possible to do a selective LBA SMART scan while using direct IO?

[JGAN]

WOW. So looks like this was a variation of slow fix. Patched mod 02 and it has stayed on longer than I have ever seen it before, so now ddrescue can go to town!!

Attachments

[maximus]

Glad to see that it is able to image now! The curious part is why you had to use direct mode in hddsupertool to get it to respond long enough to do anything like apply the slow fix. The reading of sectors by the OS must have been doing something to the drive, maybe aggravating the slow issue to the point of lock up.

[JGAN]

Yup, that's all I can think of too. Anyways, I took splidits advice and converted it back to USB as decrypting 2tb would take days, and it worked out perfectly! Thanks everyone!!