3 ms·
Yes, the ultimate (unformatted) capacity of the medium is rougly 2 million bytes, and includes the sector headers as well as the gaps to account for rotational
by userbinator 6y ago
Yes, the ultimate (unformatted) capacity of the medium is rougly 2 million bytes, and includes the sector headers as well as the gaps to account for rotational speed variance. The calculation is as follows:
A 1.44M drive nominally writes at 500Kbit/s and spins at 300RPM, i.e. one track takes 200ms and 100Kbit are written. There are 80 tracks so 8000Kbit on one side, or 16000Kbit on two sides, corresponding to 2Mbit in total.
By writing "long sectors" that take up the whole track, omitting almost all the overhead and gaps, and using up to 82 tracks, the 2M utility I linked in another comment here manages to reach very close to that limit.
- VLM 6y agoIn the 90s there used to be a web page for linux describing all the possible physical formats and one interesting side effect was changes in average kilobytes/sec transfer rate. Obviously more tracks simply took longer to read more data in a linear fashion but there were weird interactions where long sectors increased average bandwidth, mostly. I'm old enough to manually set interleave parameters on hard disks in the early 80s, so that was interesting. I don't remember interleave optimizing on floppies but I may just forget.