Hi,
I have a Samsung 860 Evo 1TB SATA drive (model MZ7LH1T0HMLU) which is not seen by the BIOS. I was hot plugging some test HDDs to an eSata connector, and while doing a read test on another drive, the Samsung went offline. No damage was done to the other drive, but the 860 Evo does not appear now at all. The usual diagnosis steps (try different cable, try different PC, etc) did not help. Drive is not recognized in anything.
Upon disassembling, I saw a board marked as MZ7LH1T0HML0. The layout is very close if not the same as every other teardown you can see online - for ex. https://tpucdn.com/ssd-specs/images/d/12-pcb-front.jpg. Visual inspection of the components did not show any damage.

With the power off, I measured all the SMD capacitors and none of them looked shorted (none had continuity between the two ends, all measured to have some resistance).

With power on (using a usb2.0 to SATA external connector I had at hand), I was reluctant to do much measurements due to how small the components are. One important thing that stood out was two components getting hot: one was the DRAM chip, which got very warm to the touch, but I suppose this is normal. The other was the e-fuse right next to the DRAM (small rectangular 10-pin chip) which was BURNING hot and I could not touch it for more than a second. Due to the extreme heat given off by the e-fuse, I'm thinking that it is the culprit.

I read in another thread that it is possible to just bridge the voltage input and output pins of the e-fuse to get the drive working "with some caveats", if this works the first thing I'd do is a full backup, so if it can give 2-3 more hours of life to the drive, that would be all I need.

So, my questions are:
- what else can and should I do to get a more accurate diagnosis?
- could there be something else broken, or is it most likely the e-fuse? I did not measure any shorts on any of the caps.
- if it is the e-fuse, should I remove it before bridging the Vin and Vout pins, or can I leave it in place and bridge it that way? Obviously I'd route the bridge cable so it goes around the chip, to avoid the heat.

My last backup of this drive is over a year old (yes, I'm stupid), so I'd really like it if it was possible to restore the data. It's not terribly important data, but restoring it would be very annoying and take a while.

Attached pics of the board serial and the e-fuse, they are crappy phone pics, I'll make some proper pics soon.

Without power to the drive, check the power rails for a short to ground. Both the main vcc to pre and post the fuse, then check the individual rails from the pmic to see if any of those are shorted. If your fuse is shorted it will need to be removed and either replaced or jumped.

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

Yeah, 5V and Ground on the SATA power pins seem to be shorted. The other power input lines look fine. Vout on the e-fuse is not shorted, just the three Vin lines.

I tried to take some detailed photos, but it is very difficult to do so due to the small size of the board. Here they are:
https://postimg.cc/gallery/bqFMb2G

The PMIC is the RB990, right? It is very difficult to test since it is close to other components and there's less space than where my multimeter tip can fit. Pin 2 seems to be grounded but I could just be accidentally touching the ground end of the cap (?) right next to it. Pins 6 and 7 seem to connect to ground, but according to the pinout these are meant to be grounded anyway, and pin 2 is meant to be NC, so it doesn't look like there's a problem there.

There seems to be a very small cap (?) that seems to be problematic, see the attached picture. It has continuity but measures 20 ohms resistance. But one end seems to be connected to the +5V straight from the SATA pins, so this could be a consequence of the short, not the cause. The via on the other end of the cap might be leading to the e-fuse Vin on the other side (extremely hard to tell as not all vias line up, but it looks like the other end is the slightly copper colored via on the e-fuse pin 2).

pmic is s2fpS05A01 - big square chip shown in your first post second image surrounded by inductors. Check at the inductors for shorts. Don't worry about the cap you've highlighted it's connected to +5v VCC.

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

The inductors (the largest grey smd components) all show continuity, but isn't that normal?

They should have continuity from pad to pad yes, you need to check for a short to ground. Put your black probe on a screw hold and measure the resistance to ground from the inductors.

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

none of the inductors around the PMIC seem to show continuity to ground (tested by clamping the screw hole). Some of the caps did, but only on the ground end, which is normal.

In that case you should be good to remove the efuse and bridge vin and vout.

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

Yeah, that's what I was suspecting after seeing the 5v and ground shorting, but I didn't dare to start without a second opinion from the experts. I'll do that this evening. Probably will have to use the heat gun because the chip is too small to desolder otherwise, hopefully it won't dislodge any of the other components. I hate how they miniaturize things so much that they can't be repaired by hand, I'm too used to working on 90s tech...

Fingers crossed, and many thanks for the help!

You could drive a double-decker bus around that chip ! Just watch the heat on the sata connector. I have to do everything under the scope nowadays.

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

Wish I had that good driving skills. Actually, I don't have any driving skills. And I can barely see the pcb I soldered to, but fortunately there were 3 pins on each side, so I could afford to be less precise.

Had to clean up the leftover solder blob from the middle pad, which was grounded and ended up connected to the Vout pins - if, for no other reason, than because it looked like it tore the pins off the PCB... After double checking for shorts, then soldering both ends of that tiny bridge wire, then double checking for any shorts... and the drive works! Already put it back in its case and imaging it to my network array at ~300MB/s, plus a secondary offline backup that will act as a hot swap until I get a new drive. It's looking good!

This is twice I had to ask for help on this forum and twice I managed to recover my data. Thank you so much for the help, though I wish that next time I visit it will be under less stressful circumstances.

Hello,
My SSD Samsung 860 PRO 1TB (3 years old) also suddenly failed last Monday, out of nowhere (while in use), with the message "no bootable device."
Even with an external SSD reader, my SSD is no longer recognized.
I have taken some measurements, see the link below.
Are these continuities (short circuit) normal?

I have not observed any moisture damage or poor solder joints under the microscope.

Could it also be that only the E-fuse is the problem in my case?
Unfortunately, my last backup is already three months old.

https://drive.google.com/file/d/1MTajp_ ... YrM60/view

Thank you

We can't see anything from that.

Continuity is fine on inductors but bad on capacitors. Measure the resistance to ground on the v5 rail pre and post the fuse. Use a screw hole for ground.

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

Hello, below a new video-link with more measurements.
Does that help?
What do you suspect? Could it be the same issue?

Thank you

https://drive.google.com/file/d/1TyiJwj ... sp=sharing

I have a continuity post the e-fuse....
Would it make sense to desolder and short-circuit the E-Fuse in this case?

Thank you

We can't tell because we can't see where youre taking your measurements. If the VIN is grounded then yes, VOUT possibly but if it's the ground - No.

Take a hi res picture of the chip and mark the resistance to ground on each pin and post it.

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

See here (E-Fuse output has a short with GND, couldn't it be that the e-fuse mosfet makes a short to GND?):

https://drive.google.com/file/d/1ZBLFTn ... sp=sharing

That's short on vout. If you check the caps the other side of sdram (back of the pmic) they are probably also short. Hard to tell where the short is on the rail, power it up and see if anything (efuse, or caps on the rail) is getting hot. If nothing is remove the fuse and check again don't bridge anything yet.

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

The capacitors around the PMIC have a resistance of aprox. 350 ohms (measured to GND).
Nothing is getting hot....

So should I remove the efuse now?

IT's not the caps around the pmic they're on the output rails but the ones on the back fuse side of the board directly behind it.

Remove the fuse but put it safe somewhere and retest resistance at it's vout.

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