• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

how is TRUE low-level formatting done on a HDD?

Jerboy

Banned
Usually drive manufactures have a utility that does so called low-level formatting, but strictly within user data area.

How are the hard drives low level formatted with drive control, servo info and stuff? I ran into a hard drive testing equipment that was about the size of a typical desktop style PC with two sets of connectors for two hard drives at a surplus store. Is this what they use to do it?
 
My impression is that true low-level formatting has always been a SCSI-only thing, and that it is executed by the SCSI controller, making use of hardware that IDE drives do not have.
 
True LLF cannot be performed by a modern drive.

The servo data, cylinder and sectors are written onto the platters by a special machine called a servo-writer before they are assembled into drives.

The data density on a modern drive is so high that the heads in a hard drive cannot be positioned accurately except by using the data already on the platters as a guide. Clearly a drive which REQUIRES the LLF data even to operate, cannot perform an LLF on itself.

On much older drives, it was possible for the drive to perform an LLF because the heads could be positioned at an exact position without reference to the data on the platters - in those days, a potential failure mode of an HD, was for the heads to slip out of alignment with the platters - an LLF at that point would rewrite the cylinders and sector info in the new position, and the drive would work again.

A very few drives were manufactured with the ability to perform an LLF, but with hardware that could not reliably perform it. These could be damaged by this operation.

Virtually all drives, however, do not support the LLF commands - instead just deleting the user data, or more frequently, just ignoring them.

 
Not quite, guys. Mark is perfectly right on about the servo information stuff though. Yet still, there is some controller level formatting left.

Both SCSI and IDE define a formatting command. If you want a SCSI drive to format its media, you send it a FORMAT UNIT command and wait until it completes. If it's IDE, you need to send it as many FORMAT TRACK commands as it reports to have tracks.

On both interfaces, it's entirely up to the drive what it does internally. Some don't do anything, some test the surface and map out bad spots, some do classic formatting (rewriting track and/or sector headers, plus surface test). Today's SCSI drives typically at least do a surface scan to redirect the bad spots to internal spares, in order to appear defect-free to the host system. Many IDE drives, typically not having any spares, do nothing but return OK status.

And it has always been that way, with the exception of the really old MFM, RLL and ESDI attachments (and early SCSI drive controllers that also used MFM and RLL drives) when the controller was still a separate ISA card. There the controller absolutely had to format the separate drive unit to its needs.

regards, Peter
 
Back
Top