MK3254GSY 2.5" Recovery

[Sam]

Toshiba 2.5" 4-head 2-platter from a laptop that was punched by an angry old man.

Top platter surface has visible media damage, no contamination on pillow, only top slider had a little contamination which I was able to clean off.

Same symptoms before & after slider cleaning: Drive makes a couple nasty knocks then comes ready after about 10 sec. Was able to make heads map with DDI.

If I try to start imaging in any random location, drive makes a couple nasty knocks, comes ready after 10sec, images quickly for about 2 seconds and starts clicking.

Same symptoms with new heads and all the other parts from the donor drive: magnets, flappers, little "guitar pick" looking slider that pivots into position between the platters.

Have had it spinning for at least several minutes with donor heads. No contamination built up on them and when placed back in donor, media test was perfect.

I knew this was a longshot when I started because as I understand, the SA for this is on the top platter. Can someone verify this? Assuming I had a utility to edit the head map, and assuming the SA is on the damaged top platter, wouldn't it be futile anyway?

Though I do not see any noticeable platter wobble, I'm willing to try a platter swap as a last-ditch effort.

Sam

 

[Jared]

I don't think platter swap is even something you should consider. If you have access to PC-3000, I'd suggest you image the drive using a software translator with solving for G-list damage enabled. That will minimize the accessing of the SA so the drive hopefully will remain stable.

 

[Sam]

Thanks for the helpful suggestion Jared! I haven't invested in a PC3K yet...I've been getting along well with a set of SD tools (which I modified to be more stable) for years and recently bought the DFL-Seagate tool. I've recently started to encounter cases where the PC3K would be more helpful, such as this one.

It wasn't looking like I was going to get any suggestions on this one but was actually going to follow up my own post today with an update.

I also gave the platter swap a low chance of improving the situation but I did it anyway. Sure sounded like something was out of alignment.

It didn't make things worse but afterwards I was able to clear G-list with the DDI. I immediately started imaging afterwards and with 3 heads enabled I was able to image ~50% of the drive.
Funny thing was, as the imaging progressed, about 25% was read-timeout yellow unreadable blocks, very evenly spaced throughout the process. Buuut the drive never actually had a read-timout--the yellow blocks appeared instantly, as quickly as each successfully read block--~9ms. Also, the DDI interface in the upper R corner showed "0 unread", meaning it never registered a read timeout. With a second pass, most of the yellow "read-timeout" blocks were successfully read.

Serge from Deepspar suggested that this odd behavior was because Techno-on mode was left on by the DDI after clearing the G-list, and that badly damaged drives will only read in Techno-on mode. This seemed to be the case here--I got all read-timeouts anywhere on the drive unless I cleared the G-list before starting imaging.

Jared--suppose I DID have a PC3k and approached this as you suggested--I assume after I put the software translator into place I would then need to image with Data Extractor--correct?

Sam

 

[Jared]

Yes, it would need to be read through DE. As far as I'm concerned PC-3000 w/o DE is just for refurbishing drives. DE makes it a complete package.

 

[Sam]

Tried a new head stack. Imaged for 5 minutes then wouldn't come ready again. I've finally given in--PC3K is on the way. I need a 3rd DDI anyway and the DE would work well as an alternative. Will post about how imaging goes using a virtual translator.

 

[Jared]

Tried a new head stack. Imaged for 5 minutes then wouldn't come ready again. I've finally given in--PC3K is on the way. I need a 3rd DDI anyway and the DE would work well as an alternative. Will post about how imaging goes using a virtual translator.

Welcome to the major leagues

 

[Sam]

Yeah, yeah. I should have done it a long time ago. Maybe I was challenging myself to see how long I could squeeeeeak by with my inferior tools :lol:

 

[Amarbir[CDR-Labs]]

Yeah, yeah. I should have done it a long time ago. Maybe I was challenging myself to see how long I could squeeeeeak by with my inferior tools :lol:

Hi Congrats again for pc3k ,
BTW PIO mode read does wonders on toshibas try it out sometimes

 

[Sam]

[glow=red]Amarbir, I owe you a Kingfisher[/glow]. This drive has been imaging on a port on my PC3k for 3 weeks and it's only read about 50 million sectors successfully. On top of that, it's got skipped sectors all over the place that don't come back after multiple passes or read-retries. A few days ago I switched over to PIO mode and that barely made things better. Also experimented with backwards imaging, block size, and read-lookahead. Had overall best results with read-lookahead disabled and a 16 sector block size. This produced maaaybe a 50kb/second read rate.

I first started imaging this drive on a Deepspar right before I got the PC3k. Today I pulled out the destination drive that had the partial Deepspar image on it and fired it up. I switched to PIO mode and that sucker has been imaging like a champ--it should only take a day or 2 till I can finally put this project behind me!
Looks like this is one case where the Deepspar has the edge over the DE.

 

[Amarbir[CDR-Labs]]

[glow=red]Amarbir, I owe you a Kingfisher[/glow]. This drive has been imaging on a port on my PC3k for 3 weeks and it's only read about 50 million sectors successfully. On top of that, it's got skipped sectors all over the place that don't come back after multiple passes or read-retries. A few days ago I switched over to PIO mode and that barely made things better. Also experimented with backwards imaging, block size, and read-lookahead. Had overall best results with read-lookahead disabled and a 16 sector block size. This produced maaaybe a 50kb/second read rate.

I first started imaging this drive on a Deepspar right before I got the PC3k. Today I pulled out the destination drive that had the partial Deepspar image on it and fired it up. I switched to PIO mode and that sucker has been imaging like a champ--it should only take a day or 2 till I can finally put this project behind me!
Looks like this is one case where the Deepspar has the edge over the DE.

Well,
Thats Nice To Hear If I Could Be Of Any Help ,When It Comers To Data Recoevery Of HDD's I am Now Not That Bad