HDD GURU FORUMS

Hello again,

I have a WD Blue 2TB 3D NAND Sata SSD [WDS200T2B0A-00SM50] drive. This drive has faced a short when I used the wrong PSU cables and I was able to repair the HDD with the help of fzabkar. However, for the SSD, the data recovery place claim to be unable to recover the data on the drive. I just picked it up today and figured I might as well do what I can here too.

I attached the picture of the internals of the SSD front and back. It's hard to tell if they even did anything on this drive, and I'm not sure if the controller itself was caked in flux or something..looks a bit off. I hope these people didn't butcher my drive..





Thanks again.

Thats common wd/sandisk hardware which I see almost daily in my shop. Any data recovery engineer with experience should be able to easily troubleshoot this. Its so common we have all those power management components instock. It may not be as simple as solving your hard drive but if you own a multimeter and are good with electronics you can troubleshoot the basics to make sure. Otherwise your welcome to send it to me or I recommend you find someone.

Can you tell us the markings on U5 in the second photo (near the SATA power connector)? I expect this is a MOSFET, in which case it would be the next component to be clobbered after the TVS diode.

The diode is the one with the Littelfuse logo (LF + AG + OE). Is this diode shorted?

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

Can you tell us the markings on the 9040VMxxx chip in the first photo? It looks cooked.

Can you measure the resistance between each of the inductors and ground (L7, L8, L9, L10)? L7 is the larger component under L10.

fzabkar wrote:Can you tell us the markings on U5 in the second photo (near the SATA power connector)? I expect this is a MOSFET, in which case it would be the next component to be clobbered after the TVS diode.

It's extremely hard to read for me, but I think it says 1D Wxx (The X's are characters I cannot read)

fzabkar wrote:The diode is the one with the Littelfuse logo (LF + AG + OE). Is this diode shorted?

I tested with my multimeter, with the black probe on the line side and got no beeps. Switching to OHMS mode, and it shows 6.0ohms but slowly went up over time exceeding 8 ohms

fzabkar wrote:Can you tell us the markings on the 9040VMxxx chip in the first photo? It looks cooked.

It looks like leftover flux or something, though the gunk feels pretty solid. The markings on this are as follows:
9040VW509
U000Z2
170345

fzabkar wrote:Can you measure the resistance between each of the inductors and ground (L7, L8, L9, L10)? L7 is the larger component under L10.

[/quote]
L7 - All inductors have solid beeps
L8 - All inductors have solid beeps
L9 - One inductor is not beeping on either side and it's marked in the picture below
L10 - All inductors have solid beeps

The black dot in the image is where I put my black probe for ground and I used the Red probe on all of the inductors on both sides. Let me know if I missed anything..

Beeps are not conclusive. Some meters will beep if the resistance is 100 ohms. Instead, use the 200 ohms range.

If those capacitors really are measuring 0 ohms to ground on both sides, then that's a short. Moreover, the shorts are most likely due to shorted ICs. This means that the controller, NANDs and SDRAM would be suspect. The only conclusive way to narrow down the source of the short would be to remove the PMIC (9040VW509).

1D is a MOSFET. It can be bypassed by connecting Drain to Source. I believe it functions as an ideal diode to prevent backfeed into the 5V supply.

BTW, the top of the PMIC looks cooked. The oozy stuff around the edge is some kind of epoxy glue. My suspicion is that the overvoltage has punched through the PMIC and damaged the downstream ICs. If this is the case, then your data recovery prospects look grim.

I'm surprised it's been labelled as unrecoverable without having taken the pmic off to actually check. Cant you pull the thermal pad off the controller so we can see if that's let the magic smoke out.

fzabkar wrote:Beeps are not conclusive. Some meters will beep if the resistance is 100 ohms. Instead, use the 200 ohms range.

If those capacitors really are measuring 0 ohms to ground on both sides, then that's a short. Moreover, the shorts are most likely due to shorted ICs. This means that the controller, NANDs and SDRAM would be suspect. The only conclusive way to narrow down the source of the short would be to remove the PMIC (9040VW509).

I measured the inductors to see if they read anything and anything that I circled green in the other picture are measuring 0. The only one not measuring 0 would be the one I circled in red which goes over 50.

fzabkar wrote:1D is a MOSFET. It can be bypassed by connecting Drain to Source. I believe it functions as an ideal diode to prevent backfeed into the 5V supply.

As I'm not super familiar with this process, you might have to explain in layman's terms here.

fzabkar wrote:BTW, the top of the PMIC looks cooked. The oozy stuff around the edge is some kind of epoxy glue. My suspicion is that the overvoltage has punched through the PMIC and damaged the downstream ICs. If this is the case, then your data recovery prospects look grim.

If this is an epoxy glue, how could I remove the PMIC at that point? I hope this data recovery center didn't sabotage the unit..

Lardman wrote:I'm surprised it's been labelled as unrecoverable without having taken the pmic off to actually check. Cant you pull the thermal pad off the controller so we can see if that's let the magic smoke out.

I removed that white pad and took another picture of the board


I'm very much still a novice on this stuff, would recovery be possible at this point or would things be too complex that I should look at a second opinion by a different but more reputable recovery place? The company I initially went with was Secure Data Recovery which I feel like was a mistake.

Just to be absolutely clear, if the resistance between the ends of each capacitor is 0.0 ohms, then the corresponding voltage is shorted. Some meters will display an "O" or "OL" when they have overranged on a particular scale. That's something else.

Removing the PMIC requires a hot air station and some skill. It's not a typical DIY job. You need to find a tech with the requisite experience. This person does not need to be a data recovery tech.

Blackbird337 wrote:If this is an epoxy glue, how could I remove the PMIC at that point? I hope this data recovery center didn't sabotage the unit..

It's just goop to secure the chips from ping off, it's normal. Although a PITA to remove. I'd certainly get a second opinion, above confirming the PMIC has blown repair and recovery at this scale really isn't DIY work unless you have a microscope.