Bizarre Seagate Case

Jared

Jared

Administrator
Staff member
So I've got this Seagate ST9750423AS that came it. it's been to two other data recovery places already who gave up on it. On arrival the heads are stuck to the platters (I'm assuming they put it back this way). After freeing it up the drive IDs but shows zero capacity. System file 28 seems to be bad on both copies however after doing a translator regen I can get sector access to some areas of the drive. Not the beginning few thousand sectors, but I can read the last sector and some good areas in the middle. So it doesn't seem like a typical translator issue where you can only read up to a certain point. It seems like across the drive I'll get a few thousand good sectors then a few thousand bad ones, it doesn't seem like typical bad sector patterns from media damage, and it's the same for all heads. All heads read fine in the good areas, then just can't read for several thousand bad sectors. Every sector will return an UNC error. Seems more like the translator shift points aren't all lining up, almost as if someone was trying to manually rebuild the translator but didn't work out all the fine details. What do you guys think? Need to stick a fork in it? Something else I'm not thinking of causing this?
 
Jared

Jared

Administrator
Staff member
OK, it gets even stranger but I think I've figured it out.

I notice that it seems too consistent the amount of sectors it reads and then can't read. So I decide to do a little quick math, and here's what I notice. The drive reads exactly 81,778 sectors then fails to read 27,415 sectors and repeats this pattern. The exact same number are read, then unreadable each time.

It doesn't seem to be linked to any one head according to the headmap in DE. However if you do the math 81778 is almost exactly three times 27415. This would make perfect sense if one head is failed out of four heads in the drive. I think the real problem is that DE isn't able to properly build the headmap, so it appears that all heads are reading (at least in some areas). This is probably what lead the other companies to give up on it.

I'll test in a few moments to see if DDI can build a proper headmap for it.
 
Jared

Jared

Administrator
Staff member
Update: DDI seems to have the same issue, it's like the head map generated doesn't actually line up with the real head map. Does anyone know what system files DE and DDI use to generate the head map?
 
J

jol

Member
27,415 sectors is too small to be for one head even at the end of drive, a fortiori at the beginnig
 
S

Sam

Member
Your question about which files are used to generate head map is a good question for Serge/Deepspar.
 
Jared

Jared

Administrator
Staff member
So here's the interesting way this case worked out. I left it imaging on DDI hitting very consistent blocks of unreadable sectors. At around the 10% mark it started only hitting some slow patches but not any bad ones. By 16% it was imaging full speed and hitting no bad blocks (before in DE it was hitting them everywhere). After imaging the remaining data, it went backwards on pass 2 and was now able to read almost all the previously unreadable sectors. Just finishing up pass 3 now and although it's slow it's cleaning up every last sector. Even read the MBR, which I'd tried at least 20 times to read in DE to no avail. Seems like this mystery problem just went away on it's own during imaging.
 
You must log in or register to reply here.
Top