One more concern:
Be very careful centering it!
If the cookie is off-center, relative to the hub, then tracks won't be in
the right positions. 3.5" disks are 135 tracks per inch, so an
imperfection of 1/5 mm is more than an entire track space!
Actually, that could be VERY useful!
<DUMB IDEA>
On a "normal" disk, if the hub is off-center, then two positions of each
track will have correct radial alignment, and the rest will be off.
If, for example, sector 2 reads correctly, then it should be close to
correct between sectors 6 and 7 of a 720K, or 11 on a 1.4M.
Actually, only PART of sector 2 will be correctly aligned, but the rest of
it should be close enough.
If another drive reads sector 3 correctly, instead, then it is out of
alignment. If a third drive reads sector 6 or 7 (between 2 and 3 on 1.4M)
then it is halfway between the other two drives.
You would not know which direction drive #2 or #3 are off, but by trial
and error, you could go each way, until it works, and then save the
information about which way it had been off.
So, it would be a VERY crude version of the Dysan Digital Diagnostic
diskette, at least in terms of radial alignment.
1.4M would be preferable, due to the larger number and physically shorter
sectors, although an odd number of sectors adds "half-spacing" on the
opposite side of the track.
You would need to try a known good (radial aligned) drive to determine
which sector (or range of sectors) is correctly positioned. Then need to
take another drive for which you are willing to change or corrupt the
alignment, and by trial and error find which direction it is off, to
record whether sectors before or after the "correct" sector(s) are too far
in or out. If you MEASURE how far you need to move it, then you could
calibrate it. Then, to get a crude alignment of a drive, just list
which sector(s) will read (median for ranges), and look up direction and
distance.
<EVEN DUMBER>
OB_ridiculous-extensions:
Make numerous off-center disks.
Create a consistent format for the calibration data.
For mass production, modify a drive to be able to move the hub clamp known
distances off center of the spindle. (could be set different for different
tracks, and calibration data could be written to the sectors themselves!)
Create software that checks which sectors are readable on a track,
accesses the appropriate calibration data, and
displays color graphic images for common drive types,
showing how to adjust radial alignment, and
interactively guide semi-skilled users through making the adjustments.
</EVEN DUMBER>
</DUMB IDEA>
How many things are wrong with this dumb idea?
--
Grumpy Ol' Fred [email protected]
