VOGONS


First post, by jesolo

User metadata
Rank l33t
Rank
l33t

I've recently picked up a Seagate ST-225 drive locally for about $10 US (including shipping).
I know that to low level format the drive, you can just use the DOS Debug command and enter "G=C800:5"

I just wanted to know what parameters (including drive characteristics) one should specify for the ST-225 drive when you want to low level format.

Based on what I've read on a forum or two, some people say that the default interleave of 3 is too low for an XT controller and that 6 is recommended. What interleave value would be a good value?
Regarding the disk characteristics:
According to a "hints" text file on the Seagate FTP site, the disk characteristics of the ST-225 drive is:
615 4 616 300 11 3 (ccc h rrr ppp ee o - as per Debug)

Are these the correct values? Some forums state the last one should be 5

However, I went with an interleave of 6 and used the above suggested values as per the "hints" text file I found on the Seagate FTP site.
After the formatting was completed the following message appeared:

"Do you want to format Bad Tracks Y/N"

I chose "N"

Reason for my choice is that according to one post:
"Because a Hard Disk is a physically manufactured item, it sometimes has physical defects. Each drive manufacturer tests their drive and gives you a bad track media list, when you purchase the drive. This bad track list is usually located on the top of the disk drive case.
Because of the addressing method used by DOS, marking an entire track bad results in more than one sector being marked as bad. Since DOS limits the number of defects that it accepts, a drive with excessive media defects can cause the FORMAT program to terminate with an error,

"TRACK 00 BAD - DRIVE UNUSABLE"

It is recommended to answer the question of formatting bad tracks as NO!!!"

Is the above statement correct or, is it still advisable to format the bad tracks as well?

After the above was completed, I followed the standard DOS FDISK process and formatted the drive under DOS.
It came back with a total of 10240 bytes (+/- 10 KB) in bad sectors. From what I gather it's not uncommon for old MFM drives to have a couple of bad sectors after formatting it within DOS as well.

Otherwise, it seems to be working fine (it's just very noisy, so much that I can't even hear the hard drive "work").

Any further advice would be appreciated.

PS: I'm using a Western Digital WD1002A-WX1 controller

Reply 1 of 3, by Matth79

User metadata
Rank Oldbie
Rank
Oldbie

You would normally "protect" the marked defect tracks from being used, if you do not, then the next format MAY exclude them, but a borderline marked defect may get through and be a source of unreliability.

The format program pictured her looks even better:
https://en.wikipedia.org/wiki/Disk_formatting … evel_format.jpg

Probes for optimum interleave!
Found it via archive.org...

For an XT class system, 6 is probably about right, but might be able to do tighter if it's an 8 bit controller in a faster system.

NB. Interleave too tight = reads 1 sector per revolution, or 17 revolutions per track
Interleave too loose = uses more revolutions than needed per track. At 6, it takes 6 to read the track

The old Norton DOS utilities could test and re-interleave while preserving data

Reply 2 of 3, by jesolo

User metadata
Rank l33t
Rank
l33t
Matth79 wrote:

You would normally "protect" the marked defect tracks from being used, if you do not, then the next format MAY exclude them, but a borderline marked defect may get through and be a source of unreliability.

Does that mean I have to choose "Y" or "N" to format bad sectors as well during the last stage of the low level format?

Reply 3 of 3, by Matth79

User metadata
Rank Oldbie
Rank
Oldbie

The advice I'm reading is different from the way it was at the peak of this kind of drive, when you DID choose to format bad tracks and use the marked track data - I remember photocopying the list off the hard drive. It was part of the procedure.
The track defect list is pretty broad brush though, it can only void the whole track.

To be honest, I can't really see why it's needed, when the next formatting stage will detect bad sectors - and I think some utilities may have offered the ability to test and re-activate sectors on the marked bad tracks.