Dead Samsung EVO 850 250GB

[Adisa]

Dear readers,
I have this disc that gives no sign of life when it is plugged into a computer.
It does not show up in BIOS.

My guess (or at least home) is that there is a fuse somewhere that could be fixed.
Is there any pro out there who can point me in the right direction?

I would appreciate your help!
Thanks!

http://forumbilder.se/G82SA/img-4896-s
http://forumbilder.se/G82SA/img-4895-s

[fzabkar]

Measure the marked voltages.

PCB protection devices - Electronic Fuse:
http://www.hddoracle.com/viewtopic.php?f=100&t=1615&p=8872

Attachments

regs_2.jpg (44.91 KiB) Viewed 19765 times

regs_1.jpg (62.46 KiB) Viewed 19765 times

e-fuse_2.jpg (21.47 KiB) Viewed 19765 times

[fzabkar]

The 14-pin IC adjacent to V1 and V3 could be a TPS22966. I would check the two Vouts.

TPS22966, Texas Instruments, 5.5V, 6A, 16-mohm ON-Resistance Dual-Channel Load Switch, marking RB966:
http://www.ti.com/lit/ds/symlink/tps22966.pdf

Attachments

[Adisa]

Dear fzabkar,
Thanks for your help!

I have measured all points you marked for me:
V1 1.10Vdc
V2 3.05Vdc
V3 1.87Vdc
V4 1.81Vdc
V5 12.15Vdc
V6 1.10Vdc
V7 1.26Vdc

At the e-fuse:
Vin 5Vdc
Vout 5Vdc

14 PIN
Vout1 1,81Vdc
Vin1 1.81Vdc
Vout2 1.26Vdc
Vin2 1.26Vdc

Bias 5Vdc

http://forumbilder.se/G82V8/singleshot0000

Looks quite suspicious that the Vin and Vout are the same at the 14pin IC?

[sourcerer]

Please try to attach it to a Linux system and look at the output of "dmesg", whether you see anything there.
If you get messages like "COMRESET failed" after about 10-20 seconds, then the SSD hangs during initialisation.

[Adisa]

I've done the measurements, here are the results:
V1 1,10Vdc
V2 3,05Vdc
V3 1,87Vdc
V4 1,81Vdc
V5 12,15Vdc
V6 1,10Vdc
V7 1,26Vdc

At the e-fuse:
Vin 5Vdc
Vout 5Vdc

14 PIN
Vout1 1,81Vdc
Vin1 1.81Vdc
Vout2 1,26Vdc
Vin2 1,26Vdc
Bias 5Vdc

Looks a bit wierd that the Vin and Vout are the same on the 14pin IC, am I right?

Attachments

14pin.jpg (104.64 KiB) Viewed 19674 times

[Adisa]

Please try to attach it to a Linux system and look at the output of "dmesg", whether you see anything there.
If you get messages like "COMRESET failed" after about 10-20 seconds, then the SSD hangs during initialisation.

Hi Sourcerer,
Nice advice, its not completely dead afterall

attached a log

Attachments

dmesg.txt

(6.85 KiB) Downloaded 983 times

[fzabkar]

RB966 is a switch, so Vout = Vin. All the voltages look OK.

[fzabkar]

ISTM that the drive IDs correctly, but does not allow read access.

Code:

[  110.831727] scsi 8:0:0:0: Direct-Access     Samsung  SSD 850 EVO 250G B    PQ: 0 ANSI: 0
[  110.832233] sd 8:0:0:0: Attached scsi generic sg8 type 0
[  110.832787] sd 8:0:0:0: [sdi] 488397168 512-byte logical blocks: (250 GB/233 GiB)
[  110.833899] sd 8:0:0:0: [sdi] Write Protect is off
[  110.833903] sd 8:0:0:0: [sdi] Mode Sense: 3b 00 00 00
[  110.836165] sd 8:0:0:0: [sdi] No Caching mode page found
[  110.836171] sd 8:0:0:0: [sdi] Assuming drive cache: write through
...
[  296.954366] sd 8:0:0:0: [sdi] tag#0 FAILED Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK
[  296.954373] sd 8:0:0:0: [sdi] tag#0 CDB: Read(10) 28 00 00 00 00 00 00 00 08 00
[  296.954376] blk_update_request: I/O error, dev sdi, sector 0
[  296.954380] Buffer I/O error on dev sdi, logical block 0, async page read

Does smartctl (Linux) provide any information?

[Adisa]

ISTM that the drive IDs correctly, but does not allow read access.

Code:

[  110.831727] scsi 8:0:0:0: Direct-Access     Samsung  SSD 850 EVO 250G B    PQ: 0 ANSI: 0
[  110.832233] sd 8:0:0:0: Attached scsi generic sg8 type 0
[  110.832787] sd 8:0:0:0: [sdi] 488397168 512-byte logical blocks: (250 GB/233 GiB)
[  110.833899] sd 8:0:0:0: [sdi] Write Protect is off
[  110.833903] sd 8:0:0:0: [sdi] Mode Sense: 3b 00 00 00
[  110.836165] sd 8:0:0:0: [sdi] No Caching mode page found
[  110.836171] sd 8:0:0:0: [sdi] Assuming drive cache: write through
...
[  296.954366] sd 8:0:0:0: [sdi] tag#0 FAILED Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK
[  296.954373] sd 8:0:0:0: [sdi] tag#0 CDB: Read(10) 28 00 00 00 00 00 00 00 08 00
[  296.954376] blk_update_request: I/O error, dev sdi, sector 0
[  296.954380] Buffer I/O error on dev sdi, logical block 0, async page read

Does smartctl (Linux) provide any information?

Didn't have the SMART sw installed on this Linux machine and no internet for now.
I will install it and check the SMART as soon as I can.

[sourcerer]

Did you try to attach it through a USB-SATA interface? Is possible, please also try a native (PCI/...) SATA interface instead. (The USB2-SATA adapters usually hide some details due to limitations in the USB mass-storage protocol.)
Yes, it seems that the SSD starts but it has problems reading your data later on.
Has the SSD been unpowered for a longer period of time? (Months)

Does the SSD case look legitimate? The output from your logfile looks a bit strange to me, but I am not sure. I have seen a fake Samsung SSD in the wild already (which was sold by a reputable dealer). The indicators I found on that fake SSD were the following:
* A Samsung Logo was missing on a sticker on the SSD
* The serial number format was wrong
* A sticker was attached on the wrong side of the SSD, not covering the screwholes
* The firmware version it reported did not adhere to the Samsung Firmware number scheme
* Samsung Magician did not recognize it as a Samsung SSD

So I would suggest to also install Samsung Magician from the Samsung website, and have it analyze the SSD. (In addition to running hdparm -I and smartctl --all as fzabkar already suggested)
I think for Samsung Magician it might be also better to use a native SATA interface, without USB.

[Adisa]

Did you try to attach it through a USB-SATA interface? Is possible, please also try a native (PCI/...) SATA interface instead. (The USB2-SATA adapters usually hide some details due to limitations in the USB mass-storage protocol.)
Yes, it seems that the SSD starts but it has problems reading your data later on.
Has the SSD been unpowered for a longer period of time? (Months)

Does the SSD case look legitimate? The output from your logfile looks a bit strange to me, but I am not sure. I have seen a fake Samsung SSD in the wild already (which was sold by a reputable dealer). The indicators I found on that fake SSD were the following:
* A Samsung Logo was missing on a sticker on the SSD
* The serial number format was wrong
* A sticker was attached on the wrong side of the SSD, not covering the screwholes
* The firmware version it reported did not adhere to the Samsung Firmware number scheme
* Samsung Magician did not recognize it as a Samsung SSD

So I would suggest to also install Samsung Magician from the Samsung website, and have it analyze the SSD. (In addition to running hdparm -I and smartctl --all as fzabkar already suggested)
I think for Samsung Magician it might be also better to use a native SATA interface, without USB.

Yes, I used USB-SATA adapter. When I use native SATA I get this:

Code:

[  218.239847] ata5: exception Emask 0x10 SAct 0x0 SErr 0x4090000 action 0xe frozen
[  218.239853] ata5: irq_stat 0x00400040, connection status changed
[  218.239856] ata5: SError: { PHYRdyChg 10B8B DevExch }
[  218.239863] ata5: hard resetting link
[  218.962561] ata5: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[  223.962551] ata5.00: qc timeout (cmd 0xef)
[  223.962568] ata5.00: ACPI cmd ef/10:06:00:00:00:00 (SET FEATURES) failed (Emask=0x4 Stat=0x50 Err=0x01)
[  223.962575] ata5: hard resetting link
[  229.322554] ata5: link is slow to respond, please be patient (ready=0)
[  233.970554] ata5: COMRESET failed (errno=-16)
[  233.970565] ata5: hard resetting link
[  239.330549] ata5: link is slow to respond, please be patient (ready=0)
[  243.978545] ata5: COMRESET failed (errno=-16)
[  243.978555] ata5: hard resetting link
[  249.338545] ata5: link is slow to respond, please be patient (ready=0)
[  279.030535] ata5: COMRESET failed (errno=-16)
[  279.030547] ata5: limiting SATA link speed to 1.5 Gbps
[  279.030550] ata5: hard resetting link
[  284.054551] ata5: COMRESET failed (errno=-16)
[  284.054573] ata5: reset failed, giving up
[  284.054588] ata5: EH complete


There is the "COMRESET failed" you mentioned earlier.
So it hangs. Is it fixable?

It might have been unpowered for a longer time, I don't know but I could ask.
I really doubt this is a fake Samsung SSD. I have one of my own and it looks exactly the same.

I cant get any SMART attributes from the device. The terminal just freezes when I run the commands.

[DR-Kiev]

The main goal not clarified. Do you need data or just repair device? In first case it is possible to do, need to run it in tech-mode, connect to terminal (Rx-Tx) and download specific microcode. So far it is possible only with expensive tools like (pc3k ssd edition).

[Adisa]

The main goal not clarified. Do you need data or just repair device? In first case it is possible to do, need to run it in tech-mode, connect to terminal (Rx-Tx) and download specific microcode. So far it is possible only with expensive tools like (pc3k ssd edition).

Hi Kiev,
Yes, the main goal is to recovery the data!
Don't really care about the disc itself.
Since I don't have access to pc3k, I guess I'm screwed?

[sourcerer]

Hmm, ok, so it seems more like the USB bridge was too optimistic about the COMRESET/COMINIT handshake. So the firmware does not start correctly.
It seems very much like the problem I have with my EVO 840 250GB. You can read about my experiences with the EVO 840 here: http://www2.futureware.at/~philipp/ssd/ ... Manual.pdf
I guess that most of it applies to the EVO850 as well, but I haven't had the time to investigate the EVO850 deeper yet.
Since you do not have access to a PC3K and you want your data recovered, I would suggest that you send your SSD to a data recovery company that has access to a PC3K.
If you are interested in a DIY method, what I can currently offer is a method to dump the raw (encrypted) FLASH sectors. But I do not have a method to decrypt and reconstruct them yet.

[Adisa]

Thanks allot for your time, trying to help me out!
I will consider sending it to DR after all.

Take care!

[mr_spokk]

Thanks allot for your time, trying to help me out!
I will consider sending it to DR after all.

Take care!

We can have a look on it, I think at this stage it won't be that expensive.

Bosse