Overvolted Crucial BX100 SSD

[Neowarex]

Hi,

My friend inadvertently used the wrong SATA cable (EVGA GQ 1000W cable on an EVGA G2 1000W PSU) and seems to have killed his drive. The odd thing is that he said it worked at first and then after a shutdown and then a startup again the SSD with his Windows and some important files as well as a WD Black 2TB had died.

The WD Black 2TB had visible burn marks on the "Smooth" Motor Controller but was fully functional after a swap with an identical PCB (no mod (bios?) chip switch was needed).

However the SSD doesn't have any burn marks or any other visible damage except some heating (uncomfortable/burning after a few seconds) on the section closest to the SATA power plug.

This section (left side mostly) has several hot parts (not extremely hot but painful to to the touch.

I am attaching some images and I was wondering if there is a way to revive this drive long enough to recover any data?

I have checked some of the pins on the J4 and except pins 1 and 5 the rest are 1.796-1.807 volts. However I don't know much what to check or if it's fixable to just recover the data?

I tried finding the PCB schematics online but I couldn't find anything. The 2 diodes I could find (D13 and not sure about D12 (if it's the right one) but seem to measure OL on one direction and give a number on the other).

PCB Brand/Model

Any help would be very much appreciated.

Complete view of the PCB

Reverse

Thank you,


Neo

[fzabkar]

There appear to be 5 DC-DC converters. I would measure the resistance between Ground and the output of each converter. That should tell you if the surge has clobbered any of the big ICs (SM2246 controller, SDRAM, NAND).

I would have marked the test points for you, but your photos are too poor to work with.

BTW, it would appear that your SSD has no protection devices.

[Neowarex]

Hi,

Thanks a lot for getting back to me. I'd like to test them but I honestly don't know what kind of part I need to be looking for the DC-DC converters.

I have taken more closeup pictures, do those help? If not, can you tell me which section of the PCB should I take a picture of? I tried finding how a DC-DC converter looks online but haven't been able to find out what it is on this PCB.

Thanks,

Neo

Attachments

[fzabkar]

Measure the resistances between ground and each of V1 - V5.

Attachments

regs3.jpg (104.55 KiB) Viewed 18044 times

[Neowarex]

Hello again,

I have measured the values as per your request:

V1 (both separately): 1.814 kΩ
V2 (both separately): 1.99 kΩ
V3: 0.005 Ω
V4: 16.6 Ω
V5: 32.2 Ω

I am afraid the chip 1R0's values of 0.005 Ω is bad...

What do you recommend?

Thanks a lot for your help again Frank,

Neo

[fzabkar]

It looks like the surge may have punched through the "ABE" DC-DC converter and damaged the load. One or both of V4 and V5 look suspect, although the Vcore resistance is normally low, so that may account for one of the readings.

I would measure the resistances of the small capacitors near the large ICs. That should tell you which IC is potentially responsible for your 0.005 ohm reading. (That's an accurate meter ???)

[Neowarex]

Yes, the meter is accurate, I'll check the capacitors near all the ICs, do I need to look for anything in particular?

Thanks again,

Neo

[fzabkar]

We need to determine which ICs are responsible for the low readings. Those tiny capacitors bypass the supply voltages at those ICs. Therefore, if you find a capacitor that reads 0.005 ohm, then the adjacent IC is one of the possible culprits.

[Neowarex]

I have measured a lot of capacitors and all the highlighted ones measure 0.005 Ω on one side and 0.002 Ω from the other (not sure which side to measure from but both sides are at 0.005-7 Ω and below).

Edit: Highlighed C95 in the image as a capacitor with low resistance.

Attachments

Corrected file

[fzabkar]

You have found another short, this time at the +5V input. That means that any of the DC-DC converters could be shorted.

It still remains to find the major IC which is powered from V3. Measure the tiny caps (BCnn, CDnn, CXnn) near the SDRAM, NAND, controller. When measuring a capacitor, just measure the resistance between its two terminals.

[Neowarex]

I have measured everything (even the miniscule almost invisible caps like C1/C2/C3..) and highlighted the ones that give a reading of 000.5 Ω or less (my mistake for the previous readings, the readings above I had mentioned as being 0.005 Ω were actually 000.5 Ω, for some reason the multimeter shows the reading as such).

I have found a few that are 016.0 Ω and a few that are 0.34.0 Ω would you like me to highlight those as well?

Sorry to have taken so long, I measured some 2-3 times to make sure.

Thank you again,


Neo

Attachments

[fzabkar]

It would appear that U84 provides the +3.3V supply for the NAND array.

I had assumed that U14 received its input from +5V, but that is inconsistent with your measurements. Instead it would appear that U14 gets its input from the +3.3V supply. I now suspect that U14 generates a second supply for the NAND, probably 1.8V.

Therefore it would appear that the short is due to U84 or U14 or the NANDs themselves. A shorted capacitor is unlikely, IMHO, given the nature of the overvoltage event. Such events usually damage the semiconductor components rather than the passives.

In short, you would need a rework station to proceed, and even then you may reach a dead end.

[Neowarex]

I think I understand that it should pretty much go to data recovery for any hope of recovering his data.

Thank you so much and I really appreciate all your help and I think I learned how to measure whether a capacitor is good or bad (or at least if there are issues around it) with your guidance.

Thanks a lot again,


Neo

[fzabkar]

Very few data recovery shops have component level electronic repair skills. You need to be extremely careful in your choice.

If it were my data, I would engage a component level repair technician to remove the DC-DC converters and then retest the circuit. If after removing U14 and U84 the short persists, then the NANDs are dead and your data will be unrecoverable.

[Neowarex]

I understand. I have transmitted your messages to my friend, it will be up to him to decide what are the next steps. Are the DC-DC converters something I can remove with a soldering iron (or heatgun)?

Thank you,

Neo

[fzabkar]

You need practice. It's not realistically a job for a novice.

If it helps, you, or your tech, can buy the "BE" converters here:

https://www.mouser.com/Search/Refine?Keyword=mp2158

[Neowarex]

O ok, I better leave that to professionals with proper tools. I'll post an update here depending on if he goes through with it and the fate of his data.

Thanks so much again,

Neo