SpinRite is 50% crap (check www.grcsucks.com for a thorough explanation of it).
If a disk starts to fail there can be certain various problems on it. Usually it’s a combination of all.
As we all know an harddisk is made of electronics and metal components ;In basic :
The platters (this is were the data physically is. One platter has two sides which are both rewritable magnetic surfacess)
The servo motor (this is the positionmotor for the comb)
The platter motor (this is the motor that drives the platter around)
The comb (this is the array of heads for the harddisk)
The print circuit board (This is where the electronics are and the IDE/SCSI interface for the translation is. Also there is firmware here in one of the chips , the control of the servo motor , the power and much more)
Now , how does a crash take place ?
A dust particle has flown on top of the platter , setting a contact between the heads and the surface of the platter. Making a physical scratch… Since the disk spins quite a lot (7200 rpm) the scratch is a nice line in no time.
If the dust decides to move , another line and so forth… Disk is beyond repair.
Climatic conditions. Though this is very very very very unusual as we all might know from science class some metal components tend to shrink or grow in size if climatic conditions change. Also think about humidity. Usually have the hardisk in a normal dry place (20 degrees Celcuis , below 80% humidity) and there’s no problem.
Again … this hardly ever takes place since all harddisks are now made of metal hybrids.
Heat. If prints , electronic circuits, and stuff like that get heated they can perform differently. This can have various results. The best method is to cool it down as well. Are there other devices that generate heat in the neighbourhood ?
Control interface or malfunction of the operating system. Every harddisk has its own self check. If you connect a harddisk to the power … the spinning and tick tick sounds you here in the first seconds are also the Power On Self test of the disk. This could go wrong due to power fluctuations , combination of heat , humidity and so forth. (That’s why you need a power supply that can deliver a STABLE power even when more power is asked (for instance , during the spinning of the disks)
Recalibration of platters , heads , comb , servo motor and all. As we all know moving parts deteriorate over time. In harddisk technology a micrometer of position shift can give disasterous results.
Now , how to control these things ? Usually you can’t. But there is the option called S.M.A.R.T. that checks how many times a disk has to re-read the data and with nice cool looking formula’s the program decides if it’s gonna fail or not. If the SMART technology says the disk is gonna fail… it usually will within a week.
What’s also quite handy is that all harddisks have a spare data built in. This means that if a place on the disk is going faulty , the internal translation changes. The sector is relocated to another sector. This however is always on the end of the platter , so if it’s a sector well within the beginning of the platter , your harddisk will go slower. If this error is occured during dust or something and the dust particle changes position , you would have to reformat it over and over again. This is so annoying that buying a new harddisk can cost less than trying to repair an old 5 year old harddisk.
Another myth is that the interrior of a harddisk is vacuum. It’s NOT ! It’s dust free , but not vacuum. (Seen those little breathing holes some quantum and maxtor disks have ?)
So in your case i’d try maxllf.exe. Let the harddisk low level format. It will check all the data on the disk by writing it , reading it , rewriting it and rereading it. In this case all the data is overwritten.
Then partition the disk with FDISK. The fdisk will check the size as well.
Then do a FORMAT C: /AUTOTEST. The autotest option will check the sectors and clusters when they are formatted.