Seagate ST3500320AS 7200.11 BSY successfully removed but cannot access

[spadge]

This thread should be used to separate my issue as suggested by BlouBul and is targeted especially @jaclaz

So here is a summary of what happened and what I did (wrong ):

I own a ST3500320AS (original SD15 Firmware, made in China) which had the BSY issue.

After the BSY occured I found out that there is a way unlocking the disk by using a terminal RS232 connection - so I followed the instructions unlocking the device from BSY and fixing LBA0 (isolating the pcb<->motor connection aso.).

Everything worked out fine - until now... Stupidly I updated to firmware SD1A after rebooting from unlocking the BSY and LBA0 issues.

Currently I'm having a drive which is accessible by the BIOS but not by the OS itself because in my opinion the OS sees/gets only corrupted data and therefore windows nor linux will start (at least not after a reasonable time).

Next thing I've tried is to access the drive using USB - this doesn't work either - system gets slowed down but I see at least the 3 partitions (ntfs) I've on this disk in windows explorer - but I'm unable to access anything further than that.

Ok my last try was to downgrade the firmware to AD14 and back again to SD1A - as I haven't found a SD15 update suitable for my drive (I've found a SD1A-34D.bin on hddguru but I was not able to flash it, the flash just didn't work/did any progress).

So now I'm stuck here and beg for some help. Maybe if someone has the right SD15 firmware I would try that or a good suggestion.

I lay everything now into all your mighty hands

[BlouBul]

Where did you get your original SD1A from? The very first version of SD1A caused similar problems. See if anything here helps. Latest version of SD1A is here.

It seems like the latest SD1A will fix the problem. http://www.tek-tips.com/viewthread.cfm?qid=1525155

Edited December 24, 2010 by BlouBul

[jaclaz]

It's strange that the drive doesn't work at all via SATA but it "works somewhat" via USB.

With the drive USB connected you can see the \\PhysicalDisk in Disk Manager and the three partitions?

It is possible that somehow the drive is still working but very, very slowly and when connected to SATA it simply "times out".

It should be possible (I seem to remember reading something about it) to "force" a firmware upgrade. I'll check and see if I can find some reference).

What does Seatools say about the drive? (try using the DOS version).

The "latest" SD1A *should* be this one:

http://seagate.custkb.com/seagate/crm/selfservice/news.jsp?DocId=207951

You should have ready a (say) 640 Gb hard disk so that if we find a way to access the disk, you can image it ASAP.

jaclaz

P.S.: Here is some reference:

http://forums.seagate.com/t5/Barracuda-XT-Barracuda-and/ST31500541AS-force-upgrade/m-p/79104

http://niallbest.com/seagate-2tb-st32000542as-cc35-firmware-upgrade/

of course you need to adapt the procedure and command line to the actual .LOD you have.

Edited December 24, 2010 by jaclaz

[spadge]

Where did you get your original SD1A from? The very first version of SD1A caused similar problems. See if anything here helps. Latest version of SD1A is here.

It seems like the latest SD1A will fix the problem. http://www.tek-tips.com/viewthread.cfm?qid=1525155

Hi BlouBul,

Hi jaclaz,

merry christmas to both of you...

I've had the latest firmware update ISO from seagate - I've heard of the problems of the first updates also, but I think the version I get should be fine

@jaclaz: I don't know if the drive would work using SATA but I haven' had enough patience wait until windows starts - broke the startup after about 15 min.

Using USB is some kind of similar to this but I managed to see my 3 partitions in windows explorer, but the drive stays busy and windows gets slowed down. After a while a message appears like "Drive E: is not formatted, would you like to do this now?

Is there a way to do a bitwise copy of the whole drive bypassing the OS, BIOS and maybe CRC or similar of the drive?

I will give it another try to update firmware to SD1A and will read through your links, thanks a lot to both of you, I will let you know as soon as I have done this - christmas is horrible time consuming

[jaclaz]

Is there a way to do a bitwise copy of the whole drive bypassing the OS, BIOS and maybe CRC or similar of the drive?

http://www.datarescue.com/photorescue/v3/drdd.htm

http://erwan.l.free.fr/clonedisk/

(but these of course WON'T bypass BIOS/OS - you will need a hardware imager - something in the several hundred dollars price range).

jaclaz

Edited December 26, 2010 by jaclaz

[spadge]

I've tried updating again to SD1A - using the latest firmware found on seagate's homepage.

Doesn't help either...

If I hotplug the drive to the mainboards SATA interface in windows, it's nearly the same than using USB: windows slows down, but I'm unable to see anything in drivemanager or in windows explorer (but I see all 3 partition letters in the "remove hardware dialog").

Note: after downgrading to AD14 the drive is not recognized in BIOS but accessible for a firmware update (SD1A) using the standard seagate update ISO.

Another curious thing running SD1A: MHDD does not recognize the harddisk, but the BIOS does.

If the drive is powered up running SD1A you can hear every 1 or 2 seconds the ususal clicking (short) as if the drive is accessing something (even connected only to power).

Going back to SD15 (the original version) is not possible because I haven't got the proper version - I think I will need SD15 for 2 discs - not for 3 or 4 discs as found on hddguru. Do you probably know a source for this? Would it make sense to ask seagte for it?

Next thing I would try is to access it using testdisk.

Do you have some more ideas?

[jaclaz]

Try on another machine (one to which you NEVER connected the hard disk in a USB enclosure).

It sounds like you have some "messed up" entries in the Registry. (which mind you seem like ADDITIONAL to the other problems you list).

Also, can you try repeating - from scratch - the 0 LBA fix?

jaclaz

[spadge]

Try on another machine (one to which you NEVER connected the hard disk in a USB enclosure).

It sounds like you have some "messed up" entries in the Registry. (which mind you seem like ADDITIONAL to the other problems you list).

Also, can you try repeating - from scratch - the 0 LBA fix?

jaclaz

Ok, I will try this - but note that I've also the same problems using various linux live/rescue cds.

I will repeat the LBA fix as you suggested. Shoudl I use AD14 or should I try using SD1A?

[spadge]

It crossed my mind that maybe something is wrong with the defective sector list - could this be possible?

[jaclaz]

It crossed my mind that maybe something is wrong with the defective sector list - could this be possible?

YES, it is possible (though UNlikely), but remember though that the only thing that may have gone bad is that you ALREADY wiped/reset the G-list, so the only tihing you can do (i.e. re-wipe re-reset it) won't change situation/behaviour.

You do not change the firmware to perform a LBA0 fix..

Try just doing it - leaving the firmware "as is" (whatever it is currently).

Then you may want to run, from DOS, the Seatools.

jaclaz

Edited December 26, 2010 by jaclaz

[spadge]

Hm, I think I've wiped the G-List. Is there a way to restore/regenerate it? The instructions I followed advice to do this

Currently I'm trying to get a dump of the whole drive using the Universal Boot Disk running the DOS rescue system. It just works but after severeal gigabyte I got a lot of read errors, >400 sectors - probably all in a row - could not be read properly.

After that I will try the LBA fix and then I'll run the seatools.

Thanks jaclaz for your support.

Will take some time now (getting a dump of the drive) as soon as I get further I will post...

[jaclaz]

Is there a way to restore/regenerate it?

NO, the "old" one is "lost forever", but running the manufacturer tool should make a new one.

Currently I'm trying to get a dump of the whole drive using the Universal Boot Disk running the DOS rescue system. It just works but after severeal gigabyte I got a lot of read errors, >400 sectors - probably all in a row - could not be read properly.

It is NOT a good idea to "insist" on a failed/failing HD area.

Which tool are you using?

You may want to try making partial chunks of image, skipping the problematic part(s).

(I don't think that any of the programs on UBCD have this possibility) - maybe Diskman

You should need something like ddrescue (Linux):

If you are "confined" to DOS, maybe this would be of use:

http://johnson.tmfc.net/dos/index.html

http://johnson.tmfc.net/dos/todisk.html

jaclaz

Edited December 27, 2010 by jaclaz

[spadge]

Is there a way to restore/regenerate it?

NO, the "old" one is "lost forever", but running the manufacturer tool should make a new one.

You are talking of Seagate's Seatools? Or what do you suggest? But regenerating them is somekind similar to a lowlevel format (losing all data) or am I wrong?

Or are there some AT commands to regenerate the G-list without destroying any data? DST short or long test ( AT Command NB or NC in Level 1) won't help here, right?

Currently I'm trying to get a dump of the whole drive using the Universal Boot Disk running the DOS rescue system. It just works but after severeal gigabyte I got a lot of read errors, >400 sectors - probably all in a row - could not be read properly.

It is NOT a good idea to "insist" on a failed/failing HD area.

Which tool are you using?

You may want to try making partial chunks of image, skipping the problematic part(s).

(I don't think that any of the programs on UBCD have this possibility) - maybe Diskman

I'm using "partition saving" tool of UBCD. Currently >176.000 unreadable sectors in a row, starting at sector 58133624 - maybe this is the spare sector area?

You should need something like ddrescue (Linux):

If you are "confined" to DOS, maybe this would be of use:

http://johnson.tmfc.net/dos/index.html

http://johnson.tmfc.net/dos/todisk.html

I don't prefer DOS to linux, was just the first thing I've tried due to the fact that dos does no try to mount/interpret the filesystem structure as this will result in a system slowdown.

Ok I will give ddrescue a try...

Thank you very much jaclaz,

spadge

[jaclaz]

I am not at all familiar with the Seagate tools, they may or may not be able to re-create the G-list. (and I have no idea idf this will work without "low--level" formatting the HD BTW "low-level" formatting doesn't exist anymore, at the most is "intermediate level" formatting)

176000 bad blocks in a row is well beyond "normal".

This kind of problem is often due to a head crash.

As said, insisting on that area may make things worse.

You can try changing the Partition ID in the MBR (as an example with MBRWIZ DOS: http://mbrwizard.com/download.php ) to an "unmountable" for windows partition ID, like - again as a example - ID 12, this way the mount manager should ignore the partition. (but we don't know if the "lock" or "freeze" in windows is connected with the actual "disk manager" or with "Mount manager" ).

If this trick works, DatarescueDD should be able to access the physicaldrive allright.

jaclaz

[spadge]

You can try changing the Partition ID in the MBR (as an example with MBRWIZ DOS: http://mbrwizard.com/download.php ) to an "unmountable" for windows partition ID, like - again as a example - ID 12, this way the mount manager should ignore the partition. (but we don't know if the "lock" or "freeze" in windows is connected with the actual "disk manager" or with "Mount manager" ).

If this trick works, DatarescueDD should be able to access the physicaldrive allright.

So far so good... After changing the partition id to 12 I was able to boot up windows and now DatarescueDD is running - we will see if it works better then the dos imaging tool.