marking bad cyls on drive

[Mike Chambers]

i have this 20 MB IDE drive in my new tandy 1000tx, with 39 bad spots found with a program called HARDPREP. i go to mark them bad, as i thought this should make the drive ignore them. it doesn't make a difference. FDISK still wants to use all the cylinders. am i doing something wrong, or does marking them bad not do what i thought it does?

 

[modem7]

First, be aware of the distinction between track and cylinder (a track is the combination of a cylinder and a side). The manufacturers 'defect list' placed on drives is a list of defective tracks. And various tools that access the drive at the low level usually mark entire tracks as 'bad', not just one or more of the sectors in that track.

1. Low level formatter flags all of the sectors in a 'bad' track by setting the CRC in the sectors to 'invalid'.

For the early versions of DOS (where FDISK doesn't scan the drive), my understanding is:

2. FDISK runs per normal.
3. When FORMAT encounters the 'invalid CRC' sectors, it marks the associated block as 'bad' (bad block).

 

[Trixter]

While spinrite has its problems, one thing it is good at doing is testing for bad sectors and marking them as bad in the FAT, so I would recommend just doing that.

 

[Mike Chambers]

which version of spinrite would be best for an ol' 286 like the 1000tx? i did a full size partition on the disk, and ran format then even scandisk's surface scan and marked bad clusters but it still has problems.

maybe the drive is just shot beyond hope. i'm back to the multiple-partition scheme.