Corrupt files on 2x WD-20EARS Caviar Green 2TB/Sharkoon case

[dlo2011]

Hello, and a happy New Year!
I'm new to this forum, visiting because of a very annoying problem I was hoping to find help with.

I have bought a Western Digital WD-20EARS (2TB) and placed in a Sharkoon external case, then copied a lot of pictures and other data on it, transferring by USB. At this point I used Teracopy to copy the files. After having made quite some modifications to files and adding others, Windows (Windows 7) started giving an error message that the disk might be corrupted. All the images also suddenly displayed horizontal lines on them, music files stopped in the middle and movies blinked when played. Checking the file system with Windows Scandisk makes scandisk exit without any notification before the process is completed and using chkdsk /f in the command prompt, a lot of orphaned files are shown, as well as index $I30 errors, ending with insufficient disk space to correct errors and and unknown error occurring at the end with an obscure reference which is always different.

I buy a second WD-20EARS, put it in another Sharkoon case, check what files I have been adding/modifying since I copied from the last backup, and started first by copying the backup to the new disk (using normal Windows Copy). Afterwards, I recovered the added/modified files using Active@ File Recovery. I check the files, they seem all to be ok. When re-establishing a sync with my backupdrive, all files are reread, the sync program gives an 'API' error, and suddenly, all the images again contain lines and other files are corrupted...

These file transfers and recoveries all take such a long time, so I really need some expert advice:
- do you think the casing is incompatible with the drives? it is a Sharkoon Rapid-Case 3.5" ESATA/USB2 which is connected by USB2; this is what a guy from Western Digital suggests
- is there a problem with both drives, physically? Both drives pass all possible tests with the Western Digital SMART/HD-checking utility
- is there a problem with a few files with a corrupted index, which then for some strange reason corrupt the index or readability of the other files? If so, can such a situation be fixed?

I don't think my files are really in danger, as I was able to do a complete recovery with Active@. The information is there, but I don't know how to get it out and keep it stable. Without an expert opinion, I fear I will destroy the drives at my next failed recovery attempt...

[zebong]

Try to contact this member

dobrevjetser

He is in your country.

[dmarques]

First of all, the drives should be imaged, to avoid any further risks. After that someone with knowledge must check your drives and understand what kind of problem are they facing.

[Vulcan]

Just a quick reply about one thing that you said:

I don't think my files are really in danger

I politely disagree. Hardware-related issues (as this appears to be) do not treat your data (photos, movies etc.) any differently to the filesystem metadata, and your story seems to say that the filesystem is getting corrupted. In one part, your story actually seems to say that file data was also being corrupted, since you describe images with lines on them etc. (at least when data was being read - although the exact details are not completely clear IMHO).

Therefore I don't see your evidence for believing that your files are not in danger. Your (apparently?) successful recovery one time, does not guarantee successful recovery the next time...

[dlo2011]

Thanks everybody. I'll contact dobrevjetser, he apparently has a data recovery company and is Dutch speaking; he might be able to help.
Imaging the drives is not really a bad suggestion. Originally, I was hoping not to have to spend that much time with the recovery, specially not falling back to the original problem after seemingly escaping from it! Unfortunately, I don't have that much space left, but I think I'll be able to manage that.
And Vulcan, I'm so sorry to have to say that I'm afraid you're right... The last thing I would want is lose everything being wreckless because convinced that it cannot be lost. I'll be careful, and I'm quite afraid my other data might get 'infected' by this problem in some way.

This is in any way a problem I've never seen before. It seems as if the entire drive contents is alright with the exception of a few corrupted files, which cannot easily be identified. At the time these files are read, an error occurs (in Windows: Windows cannot read from the specified file or disk); at that very moment, from that error message on, all the other files seem to be corrupted, although I confirm that they were intact after the copy process.
The lines that are shown are often different, occurring in about 1 on 3 pictures. Some contain only one horizontal line, others have more than one, others show a differently coloured square or an opaque square making the actual picture almost invisible. Sometimes, the right end of the picture appears on the left. It is as if that one reading error creates a faulty reading habit for the entire drive. Is this in any way possible?

[warnerr]

I am suspect of Sharkoon Rapid-Case's usb controllers to handle 2 terabyte drives. I suspect all would be well if connected to sata directly.

[fzabkar]

I would determine which USB-SATA bridge chip is used in the enclosure, and then search for known issues.

Look for the largest chip on the bridge board and post its part number. I know of at least one JMicron chip that has silent write errors, if the standard reference design is implemented.

[dlo2011]

Thanks again for all the help...
Western Digital Support didn't really help that much. Incompatibility with the case was their only suggestion.
Sharkoon Support was more helpful: they confirmed me the chipset is the Initio 1610, that no specific compatibility issues were reported; they later reported to be running H2Benchw on the combination Rapid-Case/WD20EARS.
I've also contacted dobrevjetsers data recovery company, who told me that the problem could be due to a MFT-issue. It could be possible that some of the files have invalid characters which seems to screw up something with the FRS in the MFT. I was looking on the internet for this, and I found that the diagnoses by chkdsk in those cases were similar. Has anyone heard about this?
So now I guess I should go look for the corrupt files that cause this issue (probably the same which are read at the time the file errors start occurring) and delete them someway... I read somewhere that the solution in that case was: backup the entire drive, format it, recopy the backup with exception of the originally corrupted files.
As the few corrupted files corrupt all the rest, does anyone has any experience with this kind of MFT-issue? How will I be able to find the files that have corrupted properties in the MFT in the first place?
Sorry if I'm not using the right terms here, I'm afraid I'm starting to give my own meaning to them already.

@warnerr: I can't imagine that would be true but then again... it's absolutely worth the shot to take out the drive and test it in my desktop computer at work. I'll report about that later.

[dlo2011]

Sharkoon reported that the combination of the Rapid-Case and the WD-20EARS did not give any problem with their test.
So I still have to test the disk connected by SATA, and try to find corrupted files...