Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I replaced a hard drive for a client, but they have a lot of family photos on the old one, so I'm trying my best to recover them. The drive powers up, but often doesn't appear in HDClone. Sometimes, however, it does appear and I'm able to follow through to the duplication process. Each and every time, the process stalls at sector 13,847,762 and the drive begins making a series of clicking noises.

The drive is a Seagate Barracuda 7200.7 160GB ST3160021A.

Does anyone have any suggestions? Does this sound like something a replacement circuit board might fix or is it probably a mechanical issue?

share|improve this question
2  
It's fairly likely that there's a defect at that location, and the drive gets caught in a loop trying to recover from it. It could be the defect is on a timing track rather than the data track. You need a utility that can read and write individual sectors, to see if you can get beyond it. If so, then writing that sector might allow a copy to proceed. –  Daniel R Hicks Jan 16 '12 at 1:03
    
What operating system are we dealing with, here? –  Mark Johnson Jan 16 '12 at 2:06
    
@Mark Johnson: I was using HDClone on Windows, but PhotoRec on Linux stalls on the same sector. I'll use any OS I have to. –  David Brown Jan 16 '12 at 2:39
    
Have you tried DD in linux yet? –  Scott Chamberlain Jan 16 '12 at 22:03

3 Answers 3

If it was just the one sector, I'd say it sounds like a simple media problem, a dead sector that hasn't been remapped. Since the drive isn't always visible, though, it sounds like the situation could be more serious, with possible mechanical problems. Check the SMART status on the drive, that will give you some indication of how bad things are. See How can I read my hard drive's SMART status in Windows 7? for details.

I don't think the bad sector indicates a problem with the electronics, but with the drive sometimes not responding (not visible), I guess that can't be ruled out. Controller swaps are sometimes possible, but it depends on the make and model of the drive, and that would be a last resort. I don't think it would solve the problem with the bad sector even if it was possible.

To get whatever data you can, try Unstoppable Copier, from Windows. It should give the defective sector up for dead and keep going, letting you at least get the files that can be salvaged. Then you can go back later and keep trying for the rest. If you don't mind booting up Linux, you can do the same thing with dd or some wrappers for it. See How can I get a specific large file off of a broken SSD? for more details on both approaches.

share|improve this answer

Most likely there are one or more bad blocks on the hard-drive. What you are hearing is the read-write head resetting its position as the drive controller attempts to pull a useable signal off the media. Each block includes error-correcting code which both detects corrupt data and, if a strong enough signal can be obtained, reconstructs the data.

Backup everything you can, cool down the drive by turning up the fans or cooling the environment, and keep trying to read the file. Often you can get the block read after many tries, at which point the firmware will re-write the data into a spare block and mark the bad one as off limits. This will need to be repeated for each bad block, so its helpful to isolate exactly which files are causing problems and work on them individually.

I wrote an article explaining this recover procedure in detail: Hard Drive Recovery

share|improve this answer
    
Small note - not each "block", but each sector, as the filesystem deals with either blocks or clusters (Windows) that are composed of the smallest unit on the hdd level - sectors. –  XXL Jan 20 '12 at 8:31

For a commercial software that may help you out I recommend SpinRite. It's a little expensive ($89.00) but it does the job well. If you can not get any free utilities to get it to work I would try it as a last resort. It will read the sector as much as it can then mark the sector bad. When it tries the "read as much as it can" its not just doing a normal read, it actually reading the raw data off of the head and doing probability statistics to figure out what is supposed to be there. Be prepared to wait a while, it will continue to try until it has exhausted all possible ways to read the disk, I have heard people have let it run for months as it cranks away at a bad sector (and get the data back).

A lot of people knock SpinRite for recovering the data in place instead of copying the data to another drive, but you must understand that SplinRite is not for restoring data like other data recovery tools. It is a tool to allow other tools like HD Clone and DD to work.


Followup to other question in original post:

Does this sound like something a replacement circuit board might fix or is it probably a mechanical issue?

No the issue is on the drive platter itself, replacing the circuit board will not fix it. If it was the circuit board the drive would not read at all, not go to a specific sector and stop.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.