23 Mar 2003, 10:46, JIM HOLSONBACK (1:123/140), wrote to MATT MC_CARTHY:

Hi JIM.
.........

 MM>> the ones marked as "bad" _may_ have been due to illegal 
 MM>> FAT entries, ie: pointing to sectors that do not exist on the 
 MM>> drive.  The "factory program" for that drive has the
ability to 
 MM>> zero out the entire drive, FAT included, as well as finding and 
........

 JH> I had apparent success with one of those yesterday.  Maxtor 
 JH> 2.1GB, salvaged from on its way to the "junkman".  Found with a 
 JH> sticker on it saying "run LL".  I dunno if our director tried that.

 JH> I used a Win98 rescue diskette - C: was not accessible, due to 
 JH> "invalid media".  I used FDISK to remove partition which
was on there,
 JH> make a new one, and tried FORMAT C: /u/s.  That stumbled repeatedly 
 JH> and badly over trying to recover bad clusters, and FORMAT finally 
 JH> gave up and terminated itself.

Yep, that's a _lot_ of fun after waiting hours hoping that the number
displayed will be the last one...

 JH> I then went to a bootable diskette with Maxtor's "PowerMax" 
 JH> utilities on it.  Ran the "Write Disk Pack (low level
format)" utility 
 JH> to write zeroes all over the drive.  Then ran "Advanced Test (Factory
 JH> Recertification)" - so far so good.  Then back to the Win98 
 JH> diskette - used FDISK again, and retried FORMAT - - this time with 
 JH> success - and NO bad clusters.  I then did a total of 15 passes with 
 JH> the "Burn-In test", with no problems found.  The Maxtor
software now 
 JH> says the drive is "certified error-free".  :-).  Final
test - ran 2 
 JH> passes with SCANDISK, found no bad clusters.

 JH> One thing I don't quite understand is how the first use if FDISK 
 JH> to make a new partition failed to do any good - -  maybe it didn't 
 JH> clear out the bad FAT?

FDISK is a strange program, whether by intent or by accident I can't
answer.  I've run FDISK about five times per day over the last 12 years,
(that's about 21,900 times), and I've seen LOTs of strange things,
especially when the drive parameters were not _identical_ to what had
previously been used, (ie: moving a drive from a machine with AMI BIOS to
one with AWARD BIOS, or to a machine that had a later version of the same
BIOS).

At least three times in that bunch of times, I've run FDISK, hit ESC to
exit, and gone for a cup of coffee, then came back to find the drive had
booted, and everything seemed to work on a drive that previously failed to
boot.  In those cases, FDISK obviously did _not_ clear the FAT, nor the
pointers to the BPB, yet it "fixed" whatever was wrong, where an
FDISK /MBR had no effect whatever.  Totally beyond my comprehension!

 JH> So maybe the only problem the drive had was from a virus attack?

Some of those are particularly tricky, as in the case of the
"MONKEY" virus.  FDISK will not fix a drive with MONKEY, as for
some reason, even though it clears the 'real' FAT, the code to point to the
encrypted MONKEY 'FAT' remains, and you still have a screwed up drive.  The
pointers to the BPB are gone, no files are visible, but FDISK will now tell
you that your 200Meg drive is now 35Meg, as it is counting only the free
space left after what the MONKEY FAT has marked as "used".

Somewhere around that time I got into the habit of zeroing the first
100,000 sectors on _any_ drive that was going into a customer's machine,
even the ones that came out of it (after a backup), whether it was for a
software upgrade or whatever.  With the larger drives now, I zero the first
1,000,000 sectors, as the 16-bit software I have doesn't correctly display
the sector numbers, and I'm not sure where the 'important stuff' ends.

 JH> Back to Wayne's drive -  he has now reported that it repeatedly 
 JH> spins up and spins down.  BIR from a question here or in the HDD 
 JH> conference some time back,  that is a symptom which can be caused by a 
 JH> bad controller board?

Sometimes.  The base cause is a failure of the drive electronics to
identify that the drive has spun up to full speed within a given time
frame.  The causes for that response are many.  It could be a loose power
cable connector, bad or cracked solder joint at the power socket, weak or
failing transistor array that drives the spin motor, weak or failing
feedback circuitry that 'reads' the drive RPM, shorted windings in the spin
motor, dried out grease in the motor bearings, dirt buildup on the heads
themselves preventing them from 'flying' causing excessive friction sliding
on the platters, or scratched platters disturbing the airflow and causing
head drag.

In Wayne's case, he hears the 'clicking', which indicates the drive comes
up to speed and unlatches the heads, then it shuts down.  This would likely
be overcurrent shutdown due to either bad spin circuitry, overspeed,
feedback error, or head drag.  Three out of four would be on the circuit
board, so that would be a fair option.

In bygone years when a 100Meg drive was worth $300, and we had several
identical models in stock, it was worth a shot, especially for recovering a
customer's data (no one ever backs up).

Now, with 80Gig drives worth about $100, and even finding two with the same
numbers on the PCB in the same shipment is rare, we almost never fool with
the boards any more.  My dealer always puts two drives (sometinmes three)
in every system, and sets up Windows to automatically 'backup' to the other
drives.  To have two (or three) drives fail at exactly the same time would
be rare indeed.  Aside from that, once he has the initial software fully
setup and working, he burns a copy on DVD and keeps it in the customer's
file.  That, plus the customer's data from the redundant drives, has kept
MANY customers out of hot water!   :-)


     Good luck...  M.

--- Msged/386 TE 06 (pre)