2GB Seagate drive dead; recovery options?

[Tom Hansen]

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

Upon plugging the drive into the computer with the burner, there was
no response from the drive. It doesn't make any sound at all. We've
tried it in several computers (including the original one), wiggled
the power plug, tried different jumper settings, etc, to no avail.
The chips on the logic board get quite warm after a minute or two, but
absolutely no sound or vibration comes from the drive.

We have considered swapping the logic board from another identical
drive. However, I'm concerned that there's a EPROM chip on the logic
board containing the map of bad sectors created during the original
low-level format that would obviously not match from the swappped
drive and cause data corruption. Is this true?

Any suggestions would be helpful and much appreciated.

Thanks!!

 

[J. Clarke]

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

Upon plugging the drive into the computer with the burner, there was
no response from the drive. It doesn't make any sound at all. We've
tried it in several computers (including the original one), wiggled
the power plug, tried different jumper settings, etc, to no avail.
The chips on the logic board get quite warm after a minute or two, but
absolutely no sound or vibration comes from the drive.

We have considered swapping the logic board from another identical
drive. However, I'm concerned that there's a EPROM chip on the logic
board containing the map of bad sectors created during the original
low-level format that would obviously not match from the swappped
drive and cause data corruption. Is this true?

Any suggestions would be helpful and much appreciated.

Suggestion for future reference--do "over the wire" transfer if
possible--that way you don't "jiggle" a working setup that might be fragile
due to age.

Suggestion for drive problem--the symptoms you describe are the classic ones
of a problem that is generally referred to as "stiction", the cause of
which has not been reliably reported (if you google it you'll find all
kinds of speculation but little hard data). The quick fix is to hold the
drive in your hand and rotate it "smartly" about the spindle axis several
times so that inertia causes the platters to move relative to the case.
You do NOT repeat ***NOT*** have to open the drive to do this. I put that
note in because usually at this point someone misinteprets the suggestion
in some manner that leads to a discussion of the advisability of opening
the drive.

The bad sector map is maintained on a reserved section of the drive
platters, not on an EPROM--that should not be an issue.

While it doesn't specifically mention your drive, you might also find
<http://www.heathkit-museum.com/ListArchive/msg00520.html> to be of
interest.

 

[Michael Hawes]

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

Upon plugging the drive into the computer with the burner, there was
no response from the drive. It doesn't make any sound at all. We've
tried it in several computers (including the original one), wiggled
the power plug, tried different jumper settings, etc, to no avail.
The chips on the logic board get quite warm after a minute or two, but
absolutely no sound or vibration comes from the drive.

We have considered swapping the logic board from another identical
drive. However, I'm concerned that there's a EPROM chip on the logic
board containing the map of bad sectors created during the original
low-level format that would obviously not match from the swappped
drive and cause data corruption. Is this true?

Any suggestions would be helpful and much appreciated.

Thanks!!

The platter(s) may have jammed. Try holding drive horizontal and level and
then rotating it to and fro around vertical axis. This often frees a drive
suffering from 'stiction' where the heads 'stick' to the platter. If this
fails you can try 'gentle' tapping of a corner against a hard surface, or
place in freezer for 30 min to cool the drive.
Mike.

 

[Folkert Rienstra]

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

Upon plugging the drive into the computer with the burner, there was
no response from the drive. It doesn't make any sound at all. We've
tried it in several computers (including the original one), wiggled
the power plug, tried different jumper settings, etc, to no avail.
The chips on the logic board get quite warm after a minute or two,
but absolutely no sound or vibration comes from the drive.

We have considered swapping the logic board from another identical
drive.

However, I'm concerned

Oh, why?

that there's a EPROM chip on the logic
board containing the map of bad sectors

That doesn't make sense. Why should the primary defect list be in
"a EPROM chip" when that is very inconvenient.

created during the original low-level format that would obviously
not match from the swappped drive and cause data corruption.

Yes, wouldn't it. So why on earth would you want to do that for.

Is this true?

Nope. As it wouldn't make sense at all.

 

[Folkert Rienstra]

Suggestion for future reference--do "over the wire" transfer if
possible--that way you don't "jiggle" a working setup that might be fragile
due to age.

Suggestion for drive problem--the symptoms you describe are the classic
ones of a problem that is generally referred to as "stiction",

Nonsense, he said "It doesn't make any sound at all".
With stiction you will hear distinct repeated startup noises from the
motor unit that stop after 3 or so tries.
The symptoms he described are the classic ones of a problem that is
generally referred to as "dead", not stiction.

the cause of which has not been reliably reported (if you google it you'll
find all kinds of speculation but little hard data). The quick fix is to hold
the drive in your hand and rotate it "smartly" about the spindle axis several
times so that inertia causes the platters to move relative to the case.

Actually that may make the heads plough into the platters if you do that
in the wrong direction.

You do NOT repeat ***NOT*** have to open the drive to do this. I put that
note in because usually at this point someone misinteprets the suggestion in
some manner that leads to a discussion of the advisability of opening the drive.

As if the heads ploughing into the platters is of little to no concern.

 

[Rick]

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

Upon plugging the drive into the computer with the burner, there was
no response from the drive. It doesn't make any sound at all. We've
tried it in several computers (including the original one), wiggled
the power plug, tried different jumper settings, etc, to no avail.
The chips on the logic board get quite warm after a minute or two, but
absolutely no sound or vibration comes from the drive.

We have considered swapping the logic board from another identical
drive. However, I'm concerned that there's a EPROM chip on the logic
board containing the map of bad sectors created during the original
low-level format that would obviously not match from the swappped
drive and cause data corruption. Is this true?

Any suggestions would be helpful and much appreciated.

Thanks!!

Before everyone jumps on the "stiction" bandwagon and you start
disassembling the drive I have to point out the obvious. It sounds
more likely that the drive isn't communicating with the controller.
There's a lot of reeeaaaallly obvious things here that the OP isn't
even mentioning, any of which could cause a drive to not spin up:

You *did* enter CMOS setup and let the BIOS autodetect the drive? On
all of the computers you tried? (Not mentioned at all.)

Is the BIOS detecting the drive? (Not mentioned at all.)

Any BIOS error messages regarding the drive on boot up? (Not
mentioned at all...)

Did you try putting the Seagate and the DVD drive on separate IDE
channels? A vintage 2Gig drive may not know how to play nice with a
DVD burner on the same IDE channel.

You *do* have the IDE cable connected with the correct orientation -
pin 1 to pin 1? On both the IDE controller and the drive? On older
cables and some controllers the connectors/cables are *not* keyed.

Did you pull the data sheets on *both* the Seagate drive and any
other drive you are connecting it with on the same IDE channel?
Jumpering the drive(s) isn't a guessing game. Use the wrong
master/slave settings and you can trash the data on one or both
drives in the process. Some drives have different jumper setting for
"master" or "master with slave present" and have to be jumpered
accordingly, depending on what you have connected - or not.

 

[Alexander Grigoriev]

If IDE cable is plugged backwards, Reset signal is always active and the
drive doesn't start. Check it.

 

[J. Clarke]

Before everyone jumps on the "stiction" bandwagon and you start
disassembling the drive I have to point out the obvious.

In what post did anybody recommend "disassembling the drive" to deal with
stiction?

It sounds
more likely that the drive isn't communicating with the controller.

Since the controller is part of the drive that is a very serious problem,
however why would the electronics suddenly fail?

There's a lot of reeeaaaallly obvious things here that the OP isn't
even mentioning, any of which could cause a drive to not spin up:

You *did* enter CMOS setup and let the BIOS autodetect the drive? On
all of the computers you tried? (Not mentioned at all.)

That doesn't usually prevent the drive from spinning up. It doesn't even
have to have a data cable connected in order to spin up.

Is the BIOS detecting the drive? (Not mentioned at all.)

Same concern.

Any BIOS error messages regarding the drive on boot up? (Not
mentioned at all...)

Did you try putting the Seagate and the DVD drive on separate IDE
channels? A vintage 2Gig drive may not know how to play nice with a
DVD burner on the same IDE channel.

Still shouldn't prevent the drive from spinning up.

You *do* have the IDE cable connected with the correct orientation -
pin 1 to pin 1? On both the IDE controller and the drive? On older
cables and some controllers the connectors/cables are *not* keyed.

This is the one thing you have mentioned that would give the symptoms he
describes, but it is unlikely that he made the same mistake on several
different machines.

Did you pull the data sheets on *both* the Seagate drive and any
other drive you are connecting it with on the same IDE channel?
Jumpering the drive(s) isn't a guessing game. Use the wrong
master/slave settings and you can trash the data on one or both
drives in the process. Some drives have different jumper setting for
"master" or "master with slave present" and have to be jumpered
accordingly, depending on what you have connected - or not.

Which still doesn't prevent the drive from spinning up.

 

[Folkert Rienstra]

Before everyone jumps on the "stiction" bandwagon and you start
disassembling the drive I have to point out the obvious.

Obvious eh?

It sounds
more likely that the drive isn't communicating with the controller.

And how does that prevent spin-up?

There's a lot of reeeaaaallly obvious things here that the OP isn't
even mentioning, any of which could cause a drive to not spin up:
Oh?


You *did* enter CMOS setup and let the BIOS autodetect the drive?
On all of the computers you tried? (Not mentioned at all.)

And how does that prevent spin-up?

Is the BIOS detecting the drive? (Not mentioned at all.)

And how does that prevent spin-up?
Bios won't see a(n older) drive until it has spun-up, not the other
way around.

Any BIOS error messages regarding the drive on boot up? (Not
mentioned at all...)

There won't be unless the drive initializes after spin-up.
Well surprise: It doesn't spin-up.

Did you try putting the Seagate and the DVD drive on separate IDE
channels?

A vintage 2Gig

Oh wow, you must be soo young.

drive may not know how to play nice with a DVD burner on the same
IDE channel.

Actually, it is more recent drives that don't play nice unless correctly
jumpered. However, usually the jumper is for slaves that don't play nice.

You *do* have the IDE cable connected with the correct orientation -
pin 1 to pin 1? On both the IDE controller and the drive? On older
cables and some controllers the connectors/cables are *not* keyed.

Did you pull the data sheets on *both* the Seagate drive and any
other drive you are connecting it with on the same IDE channel?

Jumpering the drive(s) isn't a guessing game.

With some brands it actually is.

Use the wrong master/slave settings and you can
trash the data on one or both drives in the process.

Never heard of that.

Some drives have different jumper setting for "master"
or "master with slave present" and have to be jumpered
accordingly, depending on what you have connected - or not.

Master with slave present isn't always what the words suggest.

 

[Tom Hansen]

OK. To summarize, here's what I think I have learned. Please,
anyone, correct me if I'm wrong.

1. The master defect table is NOT stored on the external logic board,
but rather internally somewhere. Therefore it is safe to swap the
external logic board with another identical model hard drive without
risking data loss due to the inevitable mismatch between the defect
tables of the two drives. [there is an identical model posted on eBay
right now for $5, it is tempting.]

2. It may or may not be a good idea to try to twist the drive on its
axis with a flick of the wrist to try and get the drive spinning.
(Note, we did already try this, but I am certain that we did not
exceed the 5 G's that is stated on the drive for operating limits.)

3. It may or may not be a problem called "stiction."

Also, some clarifications:

A. The drive makes no noise or vibration whatsoever upon having power
applied to it. When I power it up with it sitting on the palm of my
hand, I feel no movement or vibration at all, and of course I hear
nothing. Some of the chips on the logic board heat up, however.

B. This same exact behavior is noted no matter what the jumper
setting, and no matter whether the IDE cable is plugged in or not.

C. The drive is not recognized by the BIOS of any of the three
computers we have tried it in.

Thanks for everyone's input.

-Tom

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

[redundant details skipped]

 

[Rick]

In what post did anybody recommend "disassembling the drive" to deal with
stiction?


Since the controller is part of the drive that is a very serious problem,
however why would the electronics suddenly fail?


That doesn't usually prevent the drive from spinning up. It doesn't even
have to have a data cable connected in order to spin up.


Same concern.


Still shouldn't prevent the drive from spinning up.


This is the one thing you have mentioned that would give the symptoms he
describes, but it is unlikely that he made the same mistake on several
different machines.


Which still doesn't prevent the drive from spinning up.

Not from my experience. A drive can sit there dead in the water if
it isn't receiving a reset signal from the system. The more obvious
symptom being no head seek sounds on boot. Whether you can "feel"
the spindle motor functioning can be a subjective thing. A decent
drive shouldn't be felt to be vibrating all over the place. I have
one drive powerered up right now, not connected to an IDE channel -
dead silent - you couldn't tell from "feeling" the drive if the
spindle motor was working or not. That doesn't mean it has stiction,
it just has nothing to do.

 

[Rick]

OK. To summarize, here's what I think I have learned. Please,
anyone, correct me if I'm wrong.

1. The master defect table is NOT stored on the external logic board,
but rather internally somewhere. Therefore it is safe to swap the
external logic board with another identical model hard drive without
risking data loss due to the inevitable mismatch between the defect
tables of the two drives. [there is an identical model posted on eBay
right now for $5, it is tempting.]

2. It may or may not be a good idea to try to twist the drive on its
axis with a flick of the wrist to try and get the drive spinning.
(Note, we did already try this, but I am certain that we did not
exceed the 5 G's that is stated on the drive for operating limits.)

3. It may or may not be a problem called "stiction."

Also, some clarifications:

A. The drive makes no noise or vibration whatsoever upon having power
applied to it. When I power it up with it sitting on the palm of my
hand, I feel no movement or vibration at all, and of course I hear
nothing. Some of the chips on the logic board heat up, however.

B. This same exact behavior is noted no matter what the jumper
setting, and no matter whether the IDE cable is plugged in or not.

C. The drive is not recognized by the BIOS of any of the three
computers we have tried it in.

Thanks for everyone's input.

-Tom

I'd say "C" is your biggest cause for concern.

You may want to get the Seagate diagnostics and see what it has to
say. (Probably nothing new though, but from past experience it will
report a specific diagnostic error code for the drive.) To get the
program to (try to) access the drive you will have to manually
re-enter the CMOS parameters for the dead drive.

 

[Folkert Rienstra]

OK. To summarize, here's what I think I have learned. Please,
anyone, correct me if I'm wrong.

1. The master defect table is NOT stored on the external logic board,
but rather internally somewhere.

On the platters. It is read into controllers memory when the drive has
spun-up and done it's diagnostics as part of the "bring up" sequence.
Btw, this drive may still have sector IDs and not have tables at all.

Therefore it is safe to swap the external logic board with another
identical model hard drive without risking data loss due to the
inevitable mismatch between the defect tables of the two drives.

Actually, if it was that way it wouldn't be contained to the bad sectors.
Most all blocknumbers would probably be off by 1 or 2 or more.

[there is an identical model posted on eBay right now for $5, it is tempting.]

Yes, you can do that.
Hopefully the problem that killed your drive's controlboard is not in the HDA
itself so it is not killing the other controlboard as well. If the drive doesn't
immediately start to spin with the new board as well, cut the power immediately.

2. It may or may not be a good idea to try to twist the drive on its
axis with a flick of the wrist to try and get the drive spinning.

The drive needs to be under power and trying to spin the platters so that the
twisting is helping the momentum and the platters are prohibited to rotate in
the counterclockwise (i.e. abnormal) direction. You also need to find out what
the normal direction is and flick the casing in the opposite direction.

(Note, we did already try this, but I am certain that we did not
exceed the 5 G's that is stated on the drive for operating limits.)

3. It may or may not be a problem called "stiction."

Probably not. Maybe stiction killed it (although it shouldn't).

Also, some clarifications:

A. The drive makes no noise or vibration whatsoever upon having power
applied to it. When I power it up with it sitting on the palm of my
hand, I feel no movement or vibration at all, and of course I hear
nothing. Some of the chips on the logic board heat up, however.

B. This same exact behavior is noted no matter what the jumper
setting, and no matter whether the IDE cable is plugged in or not.

So in all likelyhood it is "dead".

C. The drive is not recognized by the BIOS of any of the three
computers we have tried it in.

Because it doesn't spin and therefor didn't complete it's bring up
sequence and therefor doesn't respond to anything.

Thanks for everyone's input.

-Tom

We have a 2GB ST32122A drive. It was working just fine until
yesterday. The user upgraded to a new computer, so we pulled the
drive in question out of the old computer in order to hook it to a
different machine to burn the data on it to CDs.

[redundant details skipped]

Any suggestions would be helpful and much appreciated.

Thanks!!

 

[Folkert Rienstra]

OK. To summarize, here's what I think I have learned. Please,
anyone, correct me if I'm wrong.

1. The master defect table is NOT stored on the external logic board,
but rather internally somewhere. Therefore it is safe to swap the
external logic board with another identical model hard drive without
risking data loss due to the inevitable mismatch between the defect
tables of the two drives. [there is an identical model posted on eBay
right now for $5, it is tempting.]

2. It may or may not be a good idea to try to twist the drive on its
axis with a flick of the wrist to try and get the drive spinning.
(Note, we did already try this, but I am certain that we did not
exceed the 5 G's that is stated on the drive for operating limits.)

3. It may or may not be a problem called "stiction."

Also, some clarifications:

A. The drive makes no noise or vibration whatsoever upon having power
applied to it. When I power it up with it sitting on the palm of my
hand, I feel no movement or vibration at all, and of course I hear
nothing. Some of the chips on the logic board heat up, however.

B. This same exact behavior is noted no matter what the jumper
setting, and no matter whether the IDE cable is plugged in or not.

C. The drive is not recognized by the BIOS of any of the three
computers we have tried it in.

Thanks for everyone's input.

-Tom

I'd say "C" is your biggest cause for concern.

You don't say.

You may want to get the Seagate diagnostics and see what it has to
say. (Probably nothing new though, but from past experience it will
report a specific diagnostic error code for the drive.)

Unless the diagnostic does something different that the bios
does, it won't see the drive and therefor it won't run at all.

To get the
program to (try to) access the drive you will have to manually
re-enter the CMOS parameters for the dead drive.

Nonsense. The parameters that it needs are on the drive itself
and if it can't get to that, which it doesn't nothing else will help.

The drive isn't spinning, get it to your skull.
No spinning, no access.

 

[Tom Hansen]

Well, I managed to get hold of another ST32122A drive. After
confirming that the new drive worked, I swapped the logic board in.
The problem drive with the new logic board spun right up and was fully
recognized by the BIOS.

That's it for the good news.

Now the drive experiences numerous read errors, and the disk
diagnostic programs I have run have come back with all sorts of
cryptic diagnostic codes. Before its abrupt death, the drive had no
errors.

I'm suspecting that there was something to my original theory, that
the defect table is stored on the logic board.

It looks like I'm going to have to spend several hundred dollars to
have the data recovered by one of those data recovery places.

Does anyone have a recommendation? One that charges by the gig would
be optimal, since it's only a two gig drive.

Thanks again everyone for your input.

 

[Folkert Rienstra]

Well, I managed to get hold of another ST32122A drive. After
confirming that the new drive worked, I swapped the logic board in.
The problem drive with the new logic board spun right up and was fully
recognized by the BIOS.

How about the software side of it? Partitions recognized?

That's it for the good news.

So far for stiction then.

Now the drive experiences numerous read errors,

Please define 'numerous'. 1 in 4, 1 in 10.000? Dispersed how?

and the disk diagnostic programs I have run have come back with all
sorts of cryptic diagnostic codes.

Such as?

Before its abrupt death, the drive had no errors.

But what killed the controller may have also damaged the/some
preamps that sit on each head beam.

I'm suspecting that there was something to my original theory, that
the defect table is stored on the logic board.

Oh? How so? Only 1 in 10.000 is a bad block.
Very good chance of them not being in the important data structures
Run HD Tach, let's see how the errors are dispersed.

It looks like I'm going to have to spend several hundred dollars to
have the data recovered by one of those data recovery places.

Not so quick.

 

[Tom Hansen]

Oh? How so? Only 1 in 10.000 is a bad block.
Very good chance of them not being in the important data structures
Run HD Tach, let's see how the errors are dispersed.

We have determined that any attempt to access beyond cylinder 830
results in a system hang, even with XP when the drive is hooked up as
a secondary drive. No diagnostic programs are able to read beyond
cylinder 830.

 

[Folkert Rienstra]

We have determined that any attempt to access beyond cylinder 830
results in a system hang, even with XP when the drive is hooked up as
a secondary drive. No diagnostic programs are able to read beyond
cylinder 830.

Uhuh, and how is that bad?
At 256 heads and 63 sectors that makes 6.4 GB.
Not bad for only a 2 GB drive.

Now if we had 64 heads we would need 1024 cylinders and your
drive starts to have (a) problem(s) at 80% into your drive.

So, how exactly does that correspond to your 'defect table' theory?
when you should see bad blocks every 5 MB or so?

Shall we start anew, and with some meaningful data this time?
And how about answering my other questions.

 

[Tom Hansen]

Uhuh, and how is that bad?
At 256 heads and 63 sectors that makes 6.4 GB.
Not bad for only a 2 GB drive.

256 heads?!!? That'd be one hell of a hard drive, at least 128
platters. It would probably be over a foot tall. It'd be really
fast, though.

Actually, the disk in question has 16 heads, 4092 cylinders, and 63
sectors.

 

[Folkert Rienstra]

256 heads?!!?

Yes. Ever heard of Int13? I suppose not.
That is the only geometry that software is to know about and that
is recorded in the partition tables. It is limited to 1024 256 63
The BIOS or driver can translate that into the default geometry
as reported by the drive (in your case probably 4092 16 63) or
in a socalled 'current translation' geometry of it's own choosing.
Any software that works through BIOS or driver will be oblivious
of this 'default' or 'current' geometry translation.

That'd be one hell of a hard drive, at least 128 platters.
It would probably be over a foot tall.

It'd be really fast, though.

You obviously have no idea.
It would be 128 times as slow as a single platter one.
The access time would be tremendously slow too, due to the sheer mass
of the headstack.

Actually, the disk in question has 16 heads, 4092 cylinders, and 63
sectors.

Uhuh, so you actually think your harddrive has 8 platters? Interesting.
Any idea how to fit 8 platters in a 1" hight drive chassis when there is
only room for 6?

The P-CHS (16 heads) translation as used on the IDE interface is just
as fake as the L-CHS (256 heads) translation as used by the BIOS-
or Driver-Program interface.
No drive really has 16 heads and 63 sectors per track.

Right. Now that we have 4092 cylinders, at 800+ cylinders your drive
only starts to have (a) problem(s) at 20% (400MB) into your drive.

So, I ask again, how exactly does that correspond to your 'defect table'
theory and not see some 80 bad blocks before that 400MB when you should
see bad blocks every 5 MB or so ?

You are not really interested in getting your harddrive to work, is it.
You don't answer any questions and make fun of something you obviously
don't have a clue about.

Are you by any chance related to Darren Harris?