Question Samsung 850 PRO 128GB Won't Initialize

[choover567]

Hello, I have a 128GB Samsung 850 Pro SSD that went dark on me. Was my main OS drive and I have a couple documents that I'd like to get back if possible. I've tried different SATA, external SATA cable, and taken it to a nearby shop and they can't seem to get it to initialize or register to windows. When I am in the disk management window it shows up as "Disk 4" and when I initialize it says "The request failed due to a fatal device hardware error". There was another shop who said the data is recoverable but cost was too high.

I looked into some forums and saw that there is an e-fuse on this model (no TVS diode). Looking at the e-fuse chip when powered I have 5V at input and 5V at the output of the chip. This is where I'm not sure where to continue troubleshooting. I'm an EE by trade and if I had a schematic I'd be all up in it. Can solder, measure, etc. Just need direction! Any help out there?




This link sprung me down a road of troubleshooting to some degree: https://forums.tomshardware.com/thr...vs-diode-samsung-evo-850.3742180/page-2#posts.

Thank you!

 

[drivinfast247]

How important is the information contained on the drive? Is it worth the cost of data recovery?

 

[choover567]

How important is the information contained on the drive? Is it worth the cost of data recovery?

Not $3500, nope. Hoping I could tinker with it on my own. Considered buying another 850 pro and replacing the SMT components I can see to see if it wakes anything up.

 

[fzabkar]

To eliminate the possibility of a supply fault, measure the voltages at each of the 6 inductors around the A0M4KB PMIC, starting at pin #1 and working ACW. These constitute 6 step-down converters.

Opposite pin #27 is a 7th inductor. Measure the voltage at the cathode of the adjacent diode. This constitutes a boost converter and should be +12V.

Here is an example of an S2FPS05A01 PMIC from an 860 Evo m.2 SSD:

https://www.hddoracle.com/viewtopic.php?p=24246#p24246

If all is OK, the most likely culprit is firmware "panic" due to bad NAND.

To test whether the flash controller has basic sanity, short the 2 end pads at the nearby row of 7 pads. Do this before you power on the SSD, then release the short a few seconds after power-up. This should place the SSD in "safe mode" or "ROM mode". The SSD should then identify itself with its factory alias.

https://www.hddoracle.com/viewtopic.php?p=24266#p24266

Some safe mode test points and UART Tx/Rx pads here:

https://www.szsghdd.cn/3948.html

 

[choover567]

To eliminate the possibility of a supply fault, measure the voltages at each of the 6 inductors around the A0M4KB PMIC, starting at pin #1 and working ACW. These constitute 6 step-down converters.

Opposite pin #27 is a 7th inductor. Measure the voltage at the cathode of the adjacent diode. This constitutes a boost converter and should be +12V.

Here is an example of an S2FPS05A01 PMIC from an 860 Evo m.2 SSD:

https://www.hddoracle.com/viewtopic.php?p=24246#p24246

If all is OK, the most likely culprit is firmware "panic" due to bad NAND.

To test whether the flash controller has basic sanity, short the 2 end pads at the nearby row of 7 pads. Do this before you power on the SSD, then release the short a few seconds after power-up. This should place the SSD in "safe mode" or "ROM mode". The SSD should then identify itself with its factory alias.

https://www.hddoracle.com/viewtopic.php?p=24266#p24266

Some safe mode test points and UART Tx/Rx pads here:

https://www.szsghdd.cn/3948.html

1.06V
1.11V
1.26V
1.87V
1.82V
3.04V
4.97V
--
12.2V



Going to grab tweezers for a quick short.

 

[fzabkar]

Those voltages seem to be OK.

 

[choover567]

Shorted the two pins and plugged into a USB/Sata cable - nothing. Is there a software or something that I need to run to interact with the SSD when in "safe mode"?

 

[fzabkar]

CrystalDiskInfo or GSmartControl should identify the "safe mode" model number, serial number and firmware version.

Something like this for an 840 Evo:

Model Number:       SAMSUNG SATA SSD
Serial Number:      00000000000000   
Firmware Revision:  ROMMEX17

Perhaps you need to leave the jumper in place until Windows boots, or until the drive appears in BIOS. I'm wondering if the drive is being reset before you get a chance to see its safe mode ID.

 

[choover567]

Doesn't seem to pick anything up using this Sabrent USB/sata cable. Maybe I'll try plugging into the MB sata port.

 

[choover567]

If it's not talking or being picked up at all, I figured power. If power is OK then I was curious if I should look at something along the SATA communication line path?

 

[fzabkar]

Unfortunately, I've never experimented with safe mode, so I don't know if your model does things differently. I suspect that the SATA data path must be OK, otherwise the drive wouldn't show up as "Disk 4". It does look like the firmware has panicked.

 

[fzabkar]

This is what the pros do:

https://forum.acelab.eu.com/viewtopic.php?t=9147

 

[choover567]

It looks like I can get bios and the PC to recognize something, but not the right thing. It shows up in bios as "Samsung SATA SSD (415mb)". In GSmartControl it shows up as a "SAMSUNG SATA SSD" with a ~435mb capacity.

 

[fzabkar]

There is one more IC to test, an "RB966" load switch. Measure its Vin and Vout. They should be identical.

TPS22966, Texas Instruments, 2-ch, 5.5V, 6A, 16-mOhm, load switch, marking RB966:
https://www.ti.com/lit/gpn/TPS22966-Q1

 

[fzabkar]

It looks like I can get bios and the PC to recognize something, but not the right thing. It shows up in bios as "Samsung SATA SSD (415mb)". In GSmartControl it shows up as a "SAMSUNG SATA SSD" with a ~435mb capacity.

That's safe mode. This means that the controller is not brain dead.

 

[choover567]

There is one more IC to test, an "RB966" load switch. Measure its Vin and Vout. They should be identical.

TPS22966, Texas Instruments, 2-ch, 5.5V, 6A, 16-mOhm, load switch, marking RB966:
https://www.ti.com/lit/gpn/TPS22966-Q1

Vin1 1.26V
Vo1 1.26V
Vin2 1.82V
Vo2 1.82V

 

[choover567]

That's safe mode. This means that the controller is not brain dead.

I can get it to show up with or without "safe mode", it seems - as long as I'm plugged directly into SATA port on motherboard.

 

[fzabkar]

Sorry, but I think your DIY adventure has come to an end. :-(

Thank you for making it easy to help you. I can't tell you how refreshing it is not having to explain to someone how to use a multimeter.

BTW, I'm an EE, too.

 

[choover567]

And now I can't get it to show up in bios or PC again. But I did for a second... *shrug*... oh well. Thanks for the help!

 

[fzabkar]

Crucial suggests that you leave an unresponsive SSD powered up for several hours to allow its firmware to repair its FTL (Flash Translation Layer). Do this with the SATA data cable disconnected. I don't recall if I've ever seen this work, though.

BTW, the professionals use PC3000 SSD, a tool developed by Russian researchers, aka hackers. They place the SSD in safe mode, then they upload special "loader" code. This code enables them to read the NANDs in raw mode. They then assemble a virtual file system from the NAND dumps and recover your data from that.

I would ask the data recovery pros at reddit for a recommendation (I hang out there, too).

https://ww.reddit.com/r/datarecovery/new/

 

[choover567]

Sorry to resurrect an old thread, but can anyone tell me about this guy - circled in the picture?

 

[fzabkar]

It looks like a 5V e-fuse. You should measure 5V at both capacitors. If not, then the e-fuse can be bypassed.