Can anybody successfully replace heads in 500GB + drives?

[DRNJ]

Can anybody successfully and repeatable replace heads in 500GB + drives?

My guess is the tracks are so tightly spaced that tolerance conditions will be almost impossible to align all heads to the same cylinder.

If it were possible to mirror copy one disk surface at a time to a new drive I think it would be possible.

Does anybody make such a tool?

Hey Ace Lab, Salvation Data, Maysoft, if any of you can make such a tool you would become instant multi millionaires!

[darkforce1]

What Model of the hard disk you are talking about.


DF

[HDD_MASTER]

Most models are pretty simple to do.

Samsung, Seagate and Maxtors are my favourites. I have succesfully replaced many headstacks on these. Most of the time they required platterswaps first.

I find that the Seagates have great resiliance with headswaps and particularly platterswap.

The worst I find for headswaps are WD's. I think they are OK for platterswap but the heads are a bitch to match up.

[DRNJ]

That’s good to know about WD drives. I thought it was just me.

What do you think about Hitachi?
I have been working on a Hitachi 500GB Deskstar HDS725050KLA360 (5 platters & 10 heads) without much success.

Theoretically the actuator arm has a manufacturing build tolerance.
The track spacing keeps getting smaller and smaller.
If the build tolerance exceeds the track tolerance the only way to recover the drive is to bend actuator arms.

At what point will nobody be able to recover a drive? 500GB?, 1TB?, 20TB?

These 500GB drives seem to be failing at a greater rate than anything I have seen before.

Are hard drives going to hit a limit like they did with processor speed?

[jono-ats]

Most models are pretty simple to do.

Samsung, Seagate and Maxtors are my favourites. I have succesfully replaced many headstacks on these. Most of the time they required platterswaps first.

I find that the Seagates have great resiliance with headswaps and particularly platterswap.

The worst I find for headswaps are WD's. I think they are OK for platterswap but the heads are a bitch to match up.

If you had a Seagate 250 GB drive with a bad bearing (two platters), would you swap heads & platters or concentrate on fixing the bearing enough to spin the drive up?

Thanks in advance,

Jon

[hiljak]

Hello,

Move the platters, head stack, and magnets to donor.

[jono-ats]

Thank you.

[coffeebean]

Hello,

Move the platters, head stack, and magnets to donor.

Yes, that's an interesting point - swapping magnets as well.

I sometimes do and almost always don't swap magnets.
Did you find some variance in magnets to suggest swapping these as well ?
I have not found significance when comparing magnets, but maybe I miss something.

Interesting to hear you viewpoint

[hiljak]

Hello,

I've gotten in the habit of moving magnets along with the head stack. Not to say that there will always be variance in magnet sets, however, the swap obviously won't be any less sucessfull as a result of moving the magnets. Plus, it takes an extra five minutes to complete. With this in mind, it's better to complete surgery only once

[HDD_MASTER]

With regard to hitachi deskstars I dont think you should worry about tolerences so much but maybe look at the technique used in removing and replacing the dead and donor heads. The problems can sometimes arise with a mis-diagnostic of head failure.

I have performed several headswaps on Hitachi and IBM drives and include the 500gb drive in my tally. Swapping the heads is easy and recovery a doddle particularly with Data extractor/PC3K.

Some problems occur after a succesful headswap that show damaged g-list and sometimes p-list. Hard to know whether the problem existed originaly or not. Make sure you have no firmware issues and try to garuantee the diagnostic as true.

With regards to platterswaps on hitachis I have only ever got 2 working again. Some families of the drive appear to have great problems with magnetic centering of the platters to the motor spindle. To be honest I dont even bother doing them so much these days.

With regards to the Seagates I wouldnt bother trying to fic the bearings, its pretty much a lost cause. I have never swapped magnets though. It Has never made a difference to me and I still find the Seagates brilliant drives to work with
and love working with the 500 & 750Gb drives. Not done a 1TB yet but looking forwrd to it.

[hiljak]

"These 500GB drives seem to be failing at a greater rate than anything I have seen before."

In regards to this comment, it seems the more active heads in the drive, the greater the potential points of failure. They are indeed a bitch.

[Odiferous]

Hello,

Move the platters, head stack, and magnets to donor.

Yes, that's an interesting point - swapping magnets as well.

I sometimes do and almost always don't swap magnets.
Did you find some variance in magnets to suggest swapping these as well ?
I have not found significance when comparing magnets, but maybe I miss something.

Interesting to hear you viewpoint

Absolutely YES - the magnets DO make a difference.

It's practically impossible to make two identical magnets; considering that the drives are calibrated in the factory with their magnets already installed, if you then change to a magnet from a different drive, you will have problems.

(It took me a while to work this out - fortunately while I was still in "experimental" mode, rather than with live jobs.)

And another thing. When you replace the magnet - make sure there isn't a stray screw stuck to it.....

[elite-recovery]

to Coffeebean: Maybe it would be better for you to provide some insight as why not to replace the magnets? This is standard procedure.

[coffeebean]

Hi elite-recovery,

I do not suggest NOT to transfer magnets so i could not provide
any insight NOT to transfer magnets.

I just wanted to know if other members found some significance
in transferring magnets with heads and platters and Odiferous has
commented in favour - so i accept this experience.

[PalmerData]

Hi all,

Interesting discussion this magnet swapping business. From my understanding, the heads locate themselves mainly by the servo markers, hence why some heads bang so violently if all heads are damged or the servo markers are magnetically corrupted. Therefore this would imply that the magnet should not make any difference.

But I would have to admit that I have had a couple of drives that worked much better using the original magnets and as Hiljak rightly mentioned, it does not take very long to swap them.