Hello,

I have a HDD like in title with data i'm trying to recover as many as it possible.

I have tried firstly to make post sector copy with:
-DMDE Windows
-DMDE Linux
-Acronis
-ddrescue linux

All attempts ending with first i/o error. For example it copy first 300 MB and then it meets first damaged file. From this moment disc hangs, and programs viewing i/o error for each sector.

In Wondows I see all the folder structure on disc and can exploring normally through disc, but when i'm trying to copy I can only copy some specific files. For example- I have a folder with 600 video files. I can copy first 1-349 files many times over and over without a problem. I copied this files for about 15x all ready to check if damaging progressing. No. Those files are still healthy, and many others. But when i try to copy file nr 350 (and many others single files) disc hangs and disappear after while. If i want to make disc appear again, normal restart is not enough. I need to turn off computer, switch off power supply, push and hold power button for a while, then turn on power supply and computer.... But make restarts to identify each corrupted file will take ages ...
I have tried also TERA copy with same result ..

I'm posting SMART:

1- before first attempts
2. after maaaaaany attempts (after 166 power on and 220h)

I know there is some BAD sectors allocated, but this parameter doesn't increase... So ...

Is there something more I can try ? I'm not pro in HDD repair. What is possibly causing I/O error and hanging HDD ?

Thanks for any response

You could try to switch off certain firmware settings via the drive's terminal port.

http://www.hddoracle.com/viewtopic.php?f=59&t=1842&p=10834#p10834

You will need to purchase a USB-TTL adapter (~US$5).

_________________
A backup a day keeps DR away.

Did this happen spontaneously or did the drive sustain some kind of external damage ? (Shock, power surge, wrath of God, angry girlfriend...)
With ddrescue, how much data did you recover ? (In GB or in %.)
Have you tried a read scan with HDTune or Victoria or HD Sentinel a similar tool ? It's not recommended at this point, because it can stress an already failing drive without recovering a single byte, but it may hint at what's going on. A better way of getting a similar visualization would be to analyze the ddrescue log with ddrescueview. You could also post the log here, someone may have a clue by looking at it.
If you can still explore the folder structure it means that key system files like the MFT are accessible and in good condition. You should attempt to recover that first, because without it the recovery becomes much more complicated. You can do that with ddru_ntfsbitmap, a command line tool (designed by a wizard who happens to be a member of this forum) included in the HDDLiveCD (provided by the very same wizard, and also containing ddrescue as well as his own HDDSuperClone, which may be helpful in your case as it's supposed to deal with hanging / freezing HDDs better than ddrescue), which quickly analyses NTFS partitions and creates “mapfiles”, restricting the ddrescue recovery to actually occupied sectors, with an option to create a second “mapfile” to specifically extract the sectors containing the MFT. With the MFT mapfile you can run a quick ddrescue pass (with the “-m” switch followed by the mapfile's name) which will secure the MFT.
Then you could try imaging in reverse (going backward) with the “-R” switch (large case R). Or if there's a particular area which contains a lot of bad sectors (which may be identified by analyzing the log file of the previous recovery attempt) you could completely bypass it for now by setting a specific input position with the “-i” switch (small case i). Or both. The “-d” switch (direct disk access) may help, although I'm not sure exactly in what kind of situation it improves the performance, or if there are drawbacks associated with this mode.
The good news indeed is that the drive's condition doesn't seem to have deteriorated in the process of those stressful recovery attempts. But there may be something else that is not visible in the SMART status.

As for Windows tools, although it's not an ideal environment for the recovery of problematic drives, a good free tool to copy individual files with possible read errors is Roadkil Unstoppable Copier. Something like TeraCopy is not adequate as it will abort the copy if there's a single error.

Ok, I have ordered USB-TTL adapter already.. But till I receive it, can you confirm what i should do, cause I've never done this..
So..

I found tutorial- http://www.hddoracle.com/viewtopic.php? ... 455&p=7530

And according to this i should:

1. connect it to the Seagate F3 arch drive like this : - viewtopic.php?f=6&t=193

2. Download, install and run HyperTerminal

3. Now power on your drive, wait for it to initialize and press CTRL+Z

4. If i will see "F3 T>" line, then type and enter one by one all commands below:

F"READ_SPARING_ENABLED",0,22
F"WRITE_SPARING_ENABLED",0,22
F"OFFLINE_SPARING_ENABLED",0,22
F"DAR_ENABLED",0,22
F"DISABLE_IDLE_ACTIVITY",1,22
F"BGMS_DISABLE_DATA_REFRESH",1,22
F"ABORT_PREFETCH",1,22
F"READ_LOOKAHEAD_DISABLED_ON_POWER_UP",1,22
F"READ_CACHING_DISABLED_ON_POWER_UP",1,22

F"RWRecoveryFlags",00,22
F"BGMSFlags",00,22
F"PerformanceFlags",043C,22
F"MediaCacheControl",00,22


?



Firstly- Thank you that you wanted to write so much
I will try those diagnose programs and post the results.
For now i can only say, that my brother say that one day he plug in this drive (it was portable USB drive) and it didn't appear under Windows. He swore me many times, that disc didn't fall. Only thing he does wrong is unsafely unplugging...

For now i can only post ddrescue screenshot after 13h of working .. So as I said. It recover first 142 MB and hang, so after few minutes of recovering I get only read errors all the time for 13h, so i decide to abort action ...

I'll humbly admit that my knowledge and experience are limited, but I know that some drives just freeze when an error is encountered, and require a power cycle for ddrescue to proceed successfully again. I've had a Toshiba drive (purchased new a few months prior) which misbehaved that way, I had to shut it down and start it up countless times to recover the non backed-up files on it (and I didn't even know about ddrescue at the time so I had to restart the copy of the aborted file with Robocopy). More recently I took care of a badly failing HGST drive which had a similar issue (and was clicking like crazy, not the infamous “click-of-death” but a frequent clicking which stopped eventually for long enough so some more bytes could be extracted again – and at the end I got all the non-system files in good shape except 20 which were corrupted). So if you see that nothing is read (do use the -P switch to get a preview), that the read speed stays at 0, and the “bad sector” size gets very high very quickly, hit CTRL+C, shut down the drive, wait a few seconds, start it up again, hit the “up” key and ENTER to re-launch the same command (or change some parameter, for instance if it hit a wall going forward ask it to go backward, or the other way around), and pray Joe Pesci...
HDDSuperClone is supposed to deal better with that kind of situation, although I don't know if those special features are available in the free version.
But, reading your first post again, it would seem like you already tried maaaaaany times to restart it during those attempts, so the above might be completely irrelevant...
Have you tried ddru_ntfsbitmap ? And have you tried imaging in reverse, starting from the very end (you have to find the exact size) and using the -R switch ?
It might require professional service anyway, so do all this at your own risk and be aware that every attempt can further deteriorate its condition. The TTL procedure should be safe though (but from what I read it could mess things up badly if the wrong command is sent).