One of my SSD-drives have given up. The data on it is not important (got it on a backup-drive) but as for now I'm curious about what exactly happened to it, and if there is something that can be done to make it work again. Like I wrote, the data is not important so this is a "playground".

See the image. There are two LEDs near the top sticker (activity + power?), none of these two are lit. This gives me two clues:

1) the controller chip is dead
2) the power source (charge pumps) are dead
3) 1 + 2.

There is zero activity in /var/log/syslog when I put the drive in a SATA-cradle.

What can I measure on the boards to find more clues?

Hi! Just one thing I noticed is the main board has F2 written in the lower left corner. A fuse? But of course in real life it can not be so simple.

It won't help you fix it, but here is a datasheet for the SSD:

Hynix_mSATA_SSD_Datasheet__HFS064_128G3AMNM_(Rev1.0)%5B1%5D.pdf
http://cfile208.uf.daum.net/attach/1376 ... A8D204C214

If the fuse is OK, I could try to help you with some test points, but I would need a much clearer photo or scan of the components. At the very least you would need to tell me their markings.

_________________
A backup a day keeps DR away.

I'd be damned if it was that simple

Unfortunately it wasn't. In the top right corner there's F1 without any components near, and on the bottom side of the PCB there's F3/F4 just right under F1/F2.

I have however played around a bit more now and I can confirm that the board is powerless. Measuring with DMM straight on the SATA-power connector all is well (0V, 5V, 12V [3.3V unused]) but measuring every pin on the mainboard connector whilst powered up yields 0V. That is, every pin is 0V. There must be some sort of a break somewhere, but there is nothing obvious on the top sides. I really had hoped for a blown fuse, but there seems to be none. Only Cx/Rx/Ux on the power board. The two Ux are connected to the data-pins (one chip per data-pair) on the SATA-cable so they're not in charge of the power.

Maybe these are filters, ie inductors ???


According to the datasheet, your SSD runs off +3.3V only.

_________________
A backup a day keeps DR away.

Well what do you know, the datasheet I found said the opposite; 3.3 not used.

Inspecting my cradle, it doesn't provide 3.3 so I hooked up an ATX-psu with 3.3V-sata and the drive-LED starts to flash red!

Still an error ofcourse, but atleast now I know it wwas my cradle that fooled me. The drive still goes unnoticed in the logs but now - with power - I can atleast inspect it a bit more.

I would verify that all the low voltage supplies are present, but I expect that the SSD may have a firmware problem.

_________________
A backup a day keeps DR away.

There's definately a power problem present. Using a DMM measuring on the orange PSU-cable I see 2.4V, not 3.3V as expected, when the drive is connected. If I remove the drive, I measure 3.3V.

Could this indicate a short circuit of some sort?

Are you loading the PSU correctly? That is, do you have a sufficient load on each of the +5V and +12V rails? A partially loaded ATX PSU may not regulate properly.

_________________
A backup a day keeps DR away.

You know what? I forgot to load the PSU. Now I hooked three old, power-consuming, ATA-drives to it and the 3.3 is balanced. And I can reach the SATA-drive.

All this makes me feel bad for not seeing the obvious; a broken PSU. The drive doesn't even have any errors, it works like a charm.

Thanks for pointing me in the right direction!

Seems I celebrated to early.

After about 5 minutes of usage, the logs are filled with lines like these:



And at this point no access can be made to the drive until I power-cycle it. Then it can be read as if nothing had happened, until about 5 minutes is passed. Then it goes into an unreachable mode until I power-cycle it. The read-rate is about 1MB per second.

Does this indicate a firmware problem or something else?

Again, this drive is a playground for experience points only. The more I learn about SSD in general the better.

I would examine the drive's SMART report.

_________________
A backup a day keeps DR away.

Is the drive's user data fully accessible after one power cycle every single time?

_________________
Hard Disk Drive (HDD), Solid State Drive (SSD, SATA, NVMe, etc), USB Flash Drive and RAID Data Recovery Specialist in Massachusetts

@labtech: Yes.

In my little experience with this series, I want to say it may be a firmware issue. There could be an issue with one or more of the NANDs, but I am not certain.

Check SMART per suggestion above and that may clear things up a bit more.

_________________
Hard Disk Drive (HDD), Solid State Drive (SSD, SATA, NVMe, etc), USB Flash Drive and RAID Data Recovery Specialist in Massachusetts

Also, how was this drive used until failure? On 24/7? Or occasional power on and power off on weekends? And how long has it been in use since purchase?

_________________
Hard Disk Drive (HDD), Solid State Drive (SSD, SATA, NVMe, etc), USB Flash Drive and RAID Data Recovery Specialist in Massachusetts

Hello bos, i have also a samsung 840 Evo (not PRO), i was wondering if it was difficult to remove the SSD external case?
About your issue i would suggest you to test it under windows.

Never recommended to do this with drives that are prone to further file system damage as result of other underlying issues.

Although ultimately it is dependent on exact partitioning scheme, any drive connected to Win based machine Windows will try to talk to it. Dependent on the type of damage sustained on the drive, subsequent write operations could be taken place by Windows to the drive. If data recovery is of importance than connecting to Windows could cause further data loss and file structure damage.

Best route is to perform diagnostic on the drive in a read-only environment.

_________________
Hard Disk Drive (HDD), Solid State Drive (SSD, SATA, NVMe, etc), USB Flash Drive and RAID Data Recovery Specialist in Massachusetts

Hi

It was a rudimentary issue to remove the shell. No black magic required.

The SMART-logs said that Relocated Sector Count was off the charts, as well as Relocated Event Count. I don't know if this is the reason the drive works flawlessly for 5 minutes before hogging the communation to death.

Drive usage was normal office hours Mon-Fri with power saving mode (not hibernation) during off hours. I don't recall when the drive was bought, but if I'd take a guess I'd say a year. Not very much longer.