Swapping Logic Boards......

[Dok]

Hi everyone!

I've got a 120GB WD that has crashed. The BIOS sees the drive but
cannnot access. The Data Lifeguard thing sees the drive but also
cannot access (various time-out and faulty cable errors, unknown
errors, etc). I tried RESQ but still cannot access. When I slave it
to a working drive the BIOS sometimes sees it but will get to Scandisk
and halt (Win98SE). If the BIOS does not see it, c: will boot but
sometimes I get a D: label, but cannot access ("Drive not ready"). I
was able to see the file system once in DOS from a rescue disk boot,
but had no other drive to transfer data to at the time.. I think
it's the PC board that is faulty, so I'm planning to swap a brand new
board to see if it helps.

Do I have to format/partition, etc the new drive to "prep" the board
before the swap or just swap fresh and go?

Thanks!

Dok

 

[Arno Wagner]

Hi everyone!

I've got a 120GB WD that has crashed. The BIOS sees the drive but
cannnot access. The Data Lifeguard thing sees the drive but also
cannot access (various time-out and faulty cable errors, unknown
errors, etc). I tried RESQ but still cannot access. When I slave it
to a working drive the BIOS sometimes sees it but will get to Scandisk
and halt (Win98SE). If the BIOS does not see it, c: will boot but
sometimes I get a D: label, but cannot access ("Drive not ready"). I
was able to see the file system once in DOS from a rescue disk boot,
but had no other drive to transfer data to at the time.. I think
it's the PC board that is faulty, so I'm planning to swap a brand new
board to see if it helps.

Do I have to format/partition, etc the new drive to "prep" the board
before the swap or just swap fresh and go?

Usually all the state is on the HDD, so no preparation. However
your failure mode sounds more like mechanical problems. Otherwise
the BIOS would likely not see the drive at all.

Arno

 

[Leo]

I'm planning to swap a brand new
board to see if it helps.

It will not help. The problem is most probably with one of heads.

Leonid

 

[Tom Scales]

I agree. I tried it just last week with a dead Maxtor 200GB. Didn't work.

Tom

 

[Dok]

I agree. I tried it just last week with a dead Maxtor 200GB. Didn't work.

Tom

Well, it's worth a shot anyway. Explain if you will, though, how does
a head malfunction?

Dok

 

[Dok]

Usually all the state is on the HDD, so no preparation. However
your failure mode sounds more like mechanical problems. Otherwise
the BIOS would likely not see the drive at all.

Arno

Sometimes the BIOS will not see the drive at all, either as solo
master, master to slave, or slave, which was why I surmised might be
the board, rightly or wrongly. We'll see in a few days when the new
drive arrives. Keep your fingers crossed.

Dok

 

[Folkert Rienstra]

It will not help.

You don't have any say in that.

The problem is most probably with one of heads.

The problem is intermittent so probably not.
So it may very well help if it's not in the HDA.

 

[Leo]

Hi everyone!

I've got a 120GB WD that has crashed. The BIOS sees the drive but
cannnot access. The Data Lifeguard thing sees the drive but also
cannot access (various time-out and faulty cable errors, unknown
errors, etc). I tried RESQ but still cannot access. When I slave it

It is typically for new WD. If you can see correct drive parameters in
BIOS Setup, it means, that drive can read firmware modules from disk.
Modules are duplicated on different disk surfaces, so can be read by
different heads. PCB is most likely good, because it can read at all.

to a working drive the BIOS sometimes sees it but will get to Scandisk
and halt (Win98SE). If the BIOS does not see it, c: will boot but
sometimes I get a D: label, but cannot access ("Drive not ready"). I

You can try to read different sectors by Norton Disk Editor for
example, sectors from LBA=0, from LBA=1000, 2000 etc. If one of drive
heads is damaged, the sectors are read by strips: bad area, good area,
bad area, etc. In bad area drive will be busy for a long time ("Drive
not ready"), may be some "clicking" sounds.

was able to see the file system once in DOS from a rescue disk boot,
but had no other drive to transfer data to at the time.. I think
it's the PC board that is faulty, so I'm planning to swap a brand new
board to see if it helps.

On my statistics, PCB malfunction at WD EB/BB/JB usually consist in
burning of L6278 chip (Spindle/VCM driver), thus the disk does not
spin. And at the symptoms described by you usually there is a
degradation of one of heads.

Do I have to format/partition, etc the new drive to "prep" the board
before the swap or just swap fresh and go?

No, it is not needed to make any preparation, but you need a PCB with
exactly same version and firmware revision, otherwise it is possible
to damage preamp chip into HDA.

PS. Just it is now copied WD400BB which had precisely same symptoms.
It was faulty head 0.

Leonid

 

[Dok]

It is typically for new WD. If you can see correct drive parameters in
BIOS Setup, it means, that drive can read firmware modules from disk.
Modules are duplicated on different disk surfaces, so can be read by
different heads. PCB is most likely good, because it can read at all.

But the BIOS recognition is intermittant. One one boot it finds it,
another boot it does not. Same thing if it's slave or master.

You can try to read different sectors by Norton Disk Editor for
example, sectors from LBA=0, from LBA=1000, 2000 etc. If one of drive
heads is damaged, the sectors are read by strips: bad area, good area,
bad area, etc. In bad area drive will be busy for a long time ("Drive
not ready"), may be some "clicking" sounds.

I tried R-studio using the bad HD as slave, but R-studio hangs when
trying to read the disk if found by the BIOS, and R-Studio will not
find it at all if not found by BIOS. I don't know how the Norton Disk
Editor works.

On my statistics, PCB malfunction at WD EB/BB/JB usually consist in
burning of L6278 chip (Spindle/VCM driver), thus the disk does not
spin. And at the symptoms described by you usually there is a
degradation of one of heads.


No, it is not needed to make any preparation, but you need a PCB with
exactly same version and firmware revision, otherwise it is possible
to damage preamp chip into HDA.

Hmm, hopefully they will match. How will I know about the version and
firmware, is it stamped on the PCB?

PS. Just it is now copied WD400BB which had precisely same symptoms.
It was faulty head 0.

How were you able to access ths disk?

BTW, how exactly does a head fail??


For the sake of my education, if we can say the PCB is good but the
head is bad, how would a data-recovery service get the data off the
disk?

Dok

 

[Leo]

But the BIOS recognition is intermittant. One one boot it finds it,
another boot it does not. Same thing if it's slave or master.

It means unstable reading. It is important for us, that reading all
the same is sometimes possible. For WD it usually means a good PCB.
But it is better to check up it by replacing, certainly.

I tried R-studio using the bad HD as slave, but R-studio hangs when
trying to read the disk if found by the BIOS, and R-Studio will not
find it at all if not found by BIOS. I don't know how the Norton Disk
Editor works.

To use R-studio, it is necessary to make a sector-by-sector copy of
HDD. And to work with a copy.

Hmm, hopefully they will match. How will I know about the version and
firmware, is it stamped on the PCB?

It is stamped on the HDA label in MDL: string (last 6 symbols). And
PCBs should be exactly identical.

How were you able to access ths disk?

I have disassembled a same HDD and have replaced the head assembly.

BTW, how exactly does a head fail??

I do not know authentically a physical mechanism of malfunction. The
useful signal from a head decreases. Sometimes it may be adjusted (MR
bias and other programmable parameters) but more often it is necessary
to replace the head assembly from good drive.

For the sake of my education, if we can say the PCB is good but the
head is bad, how would a data-recovery service get the data off the
disk?

By replacing of heads, for example. Clean room, special tools etc. I
am working in data recovery service ;-).

Leonid

 

[Folkert Rienstra]

That does make very little sense when user data isn't duplicated.
It doesn't make the drive more functional.
The number of errors to be reported by ATA drives is very limited.
Surely the ROM firmware alone can report them.

Not if it has an intermittent fault.

It means unstable reading.

Or it could be that the electronics is unstable and POST doesn't detect
it all the time but sometimes it does.

It is important for us, that reading all the same is sometimes possible.

That doesn't make sense.

For WD it usually means a good PCB.

Nor that.

 

[Dok]

It means unstable reading. It is important for us, that reading all
the same is sometimes possible. For WD it usually means a good PCB.
But it is better to check up it by replacing, certainly.

Well, I tried replacing the logic board with a new one. However, the
drive model and boards were not exactly the same (both WD1200BB but
different suffix string) and the boards had a different version stamp.
I tried it anyway. It powered up, but i got a lot of clicking sound
in the drive, so I quickly shut down, and put the new logic board back
onto the new drive. Fortunately everythign works OK, but my poor dead
drive is still unreadable.

I noticed that if you put it as slave on the first boot-up, the BIOS
sees it and Windows starts, but locks up at Scandisk. (I even got a
good SMART OK!) If I try to reboot after that the BIOS does not see
the drive at all.

To use R-studio, it is necessary to make a sector-by-sector copy of
HDD. And to work with a copy.


It is stamped on the HDA label in MDL: string (last 6 symbols). And
PCBs should be exactly identical.
(SNIP)


I have disassembled a same HDD and have replaced the head assembly.


I do not know authentically a physical mechanism of malfunction. The
useful signal from a head decreases. Sometimes it may be adjusted (MR
bias and other programmable parameters) but more often it is necessary
to replace the head assembly from good drive.


By replacing of heads, for example. Clean room, special tools etc. I
am working in data recovery service ;-).

Oh, well I figured it was that but I thought it was some super-secret
way. Special tools? The last (Maxtor) HD I opened used only a 7 or 8
Torx to unscrew the case and the ribbon interface to the ouside
electronics, and a screwdriver to unscrew the head. (Those magnets are
strong!!) What else is used? Is there any adjustment made to the head
once you screw it into place? Do the heads clear the platters easily?

Would you care to sell me one of those heads? Yes, I know about
the clean room and all, but it's worth a shot and it would be a cool
learning experience. Probably not that simple, right?

Dok

 

[Rod Speed]

Well, I tried replacing the logic board with a new one. However, the
drive model and boards were not exactly the same (both WD1200BB but
different suffix string) and the boards had a different version stamp.
I tried it anyway. It powered up, but i got a lot of clicking sound
in the drive, so I quickly shut down, and put the new logic board back
onto the new drive. Fortunately everythign works OK, but my poor dead
drive is still unreadable.

I noticed that if you put it as slave on the first boot-up, the BIOS
sees it and Windows starts, but locks up at Scandisk. (I even got a
good SMART OK!) If I try to reboot after that the BIOS does not see
the drive at all.

You have allowed for the fact that the WDs have a unique jumper
config when the drive is the only drive on the ribbon cable ?

 

[Leo]

Oh, well I figured it was that but I thought it was some super-secret
way. Special tools? The last (Maxtor) HD I opened used only a 7 or 8
Torx to unscrew the case and the ribbon interface to the ouside
electronics, and a screwdriver to unscrew the head. (Those magnets are
strong!!) What else is used? Is there any adjustment made to the head
once you screw it into place? Do the heads clear the platters easily?

What means "clear"? If on a disk there will be a particle of a dust,
its hit under a head most likely will lead to occurrence of a scratch
on a disk and damage of a head.

Would you care to sell me one of those heads? Yes, I know about
the clean room and all, but it's worth a shot and it would be a cool
learning experience. Probably not that simple, right?

The main thing - not simply to replace heads, and to make it so that
HDD worked after that. It happens not so simple.
Sorry, but I shall not give the instruction on DIY replacement of
heads. It not on my behalf, the data after that become non-recoverable
;-)

Leonid

 

[Rod Speed]

What means "clear"?

Move the heads completely off the platters so
they can be physically removed and replaced.

 

[Odie]

I noticed that if you put it as slave on the first boot-up, the BIOS
sees it and Windows starts, but locks up at Scandisk. (I even got a
good SMART OK!) If I try to reboot after that the BIOS does not see
the drive at all.

You need to be aware that Scandisk screws up more drives than it fixes -
especially if there is any fault with the drive.

Something else you might want to try manual settings in the BIOS for the
drive. I've known it to help on a number of occasions.

Odie

 

[Odie]

By replacing of heads, for example. Clean room, special tools etc. I
am working in data recovery service ;-).

Leonid

This is not allways possible - the heads of a number of drives can be
tracked all the way to a clip on the circuit board. It's almost easier
to move the platters across to a working drive.


Odie

 

[Dok]

(SNIP)

You have allowed for the fact that the WDs have a unique jumper
config when the drive is the only drive on the ribbon cable ?

If you are asking if I re-jumpered and/or recabled the drive as a
master or a slave, yes I did these things. In fact, the stamp on the
drive says to leave the jumper off if it is a master only, then
connect to the master connector on the cable. I wish it was that
easy...

Dok

 

[Dok]

Move the heads completely off the platters so
they can be physically removed and replaced.

Actually besides getting clearance to get the head off, I was asking
about the reinstalled head clearing the platter surface going back on,
i.e., any adjustment neded to avoid scratching the platters as the arm
swings back into place?

Dok

 

[Dok]

You need to be aware that Scandisk screws up more drives than it fixes -
especially if there is any fault with the drive.

Something else you might want to try manual settings in the BIOS for the
drive. I've known it to help on a number of occasions.

Odie

I tried this also according to the settings provided with the drive. No luck.

Dok