WD Elements 1 TB that spins but is not detected (or rarely)

[SOSdonnees]

Hi folks,

I have in hands a faulty WD Elements 1TB drive (P/N: WDBUZG0010BBK), which is as follow:

- The drive spins and there is no special noise.
- The diode on the enclosure lights well.
- The USB socket is not weak and seem a priori correctly soldered
I also checked after having opened the enclosure.
- In Windows, the disk manager was unable to see the drive.
- In Linux, "lsblk" was only once able to detect the drive
and "fdisk -l" was also only once able to list a NTFS partition.
It took about 20 to 30 seconds for "fdisk -l" to answer.
Unfortunately, this was possible only once. The disk remains now undetected.
I believe the diode was blinking when the Linux commands worked.
- I tested with several USB 3.0 wires. All were very well attached to the USB socket (no weak contact at this location).

I would assume some weakness with the PCB.

There seem not being power issues with the USB, but communication issues.

I heard about issues in the SATA to USB bridge for those drives and hesitate between
a) connecting wires to the SATA connectors
and
b) replacing the whole circuit and transfer the chip.

The drive itself is a WD10JMVW-11AJGS2 from 19 Feb 2014.

Are there known typical issues with this model?

Thank you.

[pcimage]

99% PCB is OK.

Almost certainly some internal media issues, and probably some firmware problem(s)

Not DIY

[data-medics]

It could be a simple matter of bad sectors in the beginning of the drive causing it to become unresponsive when the filesystem is accessed.

Find another drive with the exact same LBA size. Then try cloning using ddrescue in Linux as this guide will walk you through: http://www.data-medics.com/forum/viewto ... f=21&t=133

Then after you clone all the way to the end (important because there's an encryption key at the end) use the bridge on the new drive (necessary to decrypt the data).

Last edited by data-medics on February 21st, 2015, 12:04, edited 1 time in total.

[data-medics]

And when you ddrescue it you have to be using a SATA connection, not a USB adapter.

And you need to match the LBA size of the drive as detected via SATA. The USB bridge will mask the true LBA to keep a hidden area at the end used by it.

[fzabkar]

@SOSdonnees, once you establish a SATA connection, you can apply the "slow fix".

viewtopic.php?f=1&t=29187&start=20

http://www.alexsoft.org/viewtopic.php?t ... 4345#p4345

Alternatively, you can "deal with slow responding" by purchasing a one-month licence for WDMarvel (US$10).

[MindMergepk]

@SOSdonnees, once you establish a SATA connection, you can apply the "slow fix".

viewtopic.php?f=1&t=29187&start=20

http://www.alexsoft.org/viewtopic.php?t ... 4345#p4345

Alternatively, you can "deal with slow responding" by purchasing a one-month licence for WDMarvel (US$10).

+1
not always but some time you don't need to convert USB to SATA in order to apply slow fix.

[SOSdonnees]

Thank you.

@pcimage: Could you explain makes you think to a firmware problem?
The strange thing is that the drive was once responding correctly.
Should this not suggest some electrical problem?

@data-medics: I'm a big fan of ddrescue myself, but I simply cannot image the drive in this case as it is undetected. Only one time it was detected, then no more.
As concerning encryption, I believe there should not be any for this drive model (Reference:http://forum.hddguru.com/viewtopic.php?f=1&t=27240).

@fzabkar: Thank you for mentioning the "slow fix" as well as WD Marvel.
It is now 15$ per month and no more 10$.

[fzabkar]

As concerning encryption, I believe there should not be any for this drive model

That's right. Essentials models are encrypted, Elements are not.

[SOSdonnees]

Well, this drive really behaves strangely.

After some days, I observed that the electrical contact was weak and I had to slightly bend the USB connector downwards to have the drive correctly powered on.

I assumed that the data transmission could also have weak contact and decided to melt the weld points with the hope to remove possible cracks in the weld.
This solved the weak power and the drive now spins correctly without that I have to bend the connector.

Yesterday evening, I could test the drive after this repair and the spindle was making the well known repetitive "click of death" (which was not before).

The very strange is that today morning, I tried again connecting the drive.
The click of death had disappeared and it could initialize!
There was often some kind of "bip" after about 1 minute, something like if the spindle lose power and parked.
I had to reconnect the drive, to make it visible again.

The drive always required about 20 to 30 seconds to initialise and to answer to the "fdisk -l" command. Then the LED was blinking as the drive was reading data.

I could list partition with "fdisk -l" and even start cloning, but a very low speed of 3 Kb/s (The drive is 1 TB ...).

Because it would have taken months to clone the whole drive so slowly, I tried to mount the partition in Linux, with the hope to save first the most important files.
Linux tried to fix some errors in the NTFS files system, but this was too slow and failed because the drive disconnected before the end or the fix.

Now, the drive does not initialize anymore. I tested on many versions of Windows (XP, Vista, Seven), in several USB sockets, with several wires, as well as in Linux.

I assumed some possible bad contact and selected those from my USB wires which had the most protruding connectors.

I also cleaned the spindle contacts below the PCB and make the motor contacts more protruding.

The LED lights but does not blink.

HD Sentinel detected some WD drive, but was unable to access the SMART data.

The hypothesis I keep in mind are :
a) week contacts or weak electrical component
b) a weak component inside
c) firmware corruption

"c)" appears me less probable, because the drive also behaved the same before and this did not prevented it to work again today morning.

I have the exact same drive model in hands (incl. Firmware and DCM)*, working, but hesitate about the strategy. (* I'm aware that buying a SATA replacement board would have been much better for the diagnosis.)

- replace the USB connector first ?
- swap the U12 chip?
There is also a second windbond chip and there seem being a controversy if it must be transfered too. This second chip is very close from other components and hence hard to unsolder / resolder.
- briefly open the drive in a clean environment to look inside how it behaves.
(I don't have a clean bench yet.)

Thanks.

[MindMergepk]

I could list partition with "fdisk -l" and even start cloning, but a very low speed of 3 Kb/s (The drive is 1 TB ...).

as already mentioned, you need to apply slow fix.
I can assist on a TV connection if OP is willing to.
If all heads are ok, you should be able to clone the drive.

good luck

[SOSdonnees]

Thank Mindmergepk, but as described lower in my message, the drive does not initialize anymore. So I cannot see how could I apply the "slow fix" when the drive is not detected.

As I have the exact same drive, I wonder if it is worth trying to swap the two windbond BIOS chips.
The fact is that if is fails, I will have destroyed a working hard drive for nothing.
Also, as the other PCB is in USB too, I'm not sure if I will then be able to apply the "slow fix". (Using a SATA compatible circuit would have been better.)

[MindMergepk]

changing pcb will change nothing, the problem is somewhere else.
I am sorry but there was a fair chance to apply the slow fix but it is gone now I feel.
if by any means drive is just detected in windows and not initialize, i think my trick will still work to apply slow fix, but it require drive to be detected atleast.
did you try to connect to usb 2.0 instead of 3.0 port ? it might help !