smdev@csustan.UUCP (Scott Hazen Mueller) (11/07/86)
In article <> gnu@hoptoad.uucp (John Gilmore) writes: > >I don't understand why nobody has built magnetic disks that spin at >a constant speed ... [t]his means that you might get 30,000 bytes per >track on the inside and 90,000 on the outside -- but who cares? It's been done before. I don't think that it is quite done by changing the clocking rate, but I do know of disk drives that vary the number of sectors per track in order to fit the most data on the disk. I believe that the lowly Commodore 64's disk drive is one such device. >John Gilmore \scott -- Scott Hazen Mueller lll-crg!csustan!smdev City of Turlock work: (209) 668-5590 -or- 5628 901 South Walnut Avenue home: (209) 527-1203 Turlock, CA 95380 <Insert pithy saying here...>
ken@rochester.ARPA (Comfy chair) (11/07/86)
Wait a minute. If we are talking about constant angular velocity (CAV) disks, then the surface moves faster under the head on the outer tracks. Then to fit more sectors in you'd have to write the bits faster, so you would still have to change the clock rate. Am I missing something here? Ken
jkg@gitpyr.gatech.EDU (Jim Greenlee) (11/07/86)
In article <22218@rochester.ARPA> ken@rochester.UUCP (Comfy chair) writes: >Wait a minute. If we are talking about constant angular velocity (CAV) >disks, then the surface moves faster under the head on the outer >tracks. Then to fit more sectors in you'd have to write the bits >faster, so you would still have to change the clock rate. It depends on the encoding scheme used. For single-density (FM-encoded) diskettes, the clock pulses are interleaved with the data pulses so all you have to do is sync on those and presto-bingo - no changes in clock rate are required. For double-density (MFM-encoded) diskettes, the clock pulses are derived from an external oscillator which is usually hooked into the drive motor somehow to generate a consistent clock pulse (this is typically done by means of a phase-locked loop). Most of the variable-rate drives with which I am familiar take a slightly different tack by varying the speed of the drive motor to increase the bit density on the outer tracks. The Victor 9000 was the first micro that I am aware of which used this technique - they hooked up an Intel 8048 to a bare drive motor and used it as the drive controller. It was smart enough to compensate for differences between drive rotational speed (the 'C' in CAV isn't always so from drive to drive) so you could move diskettes from machine to machine without worrying about having the data get garbaged. Variable-speed drives are obviously impractical for large hard disks because of the inertial forces built up by the spinning platters. I would think a variable clock rate would work for a disk with fixed media if the rotational speed could be tightly controlled or compensated for. Removable media are a different story, however, because the tiniest difference in rotational speed would make disks unreadable by any drive besides the one that wrote it. I would expect the hardware for such a system to be extremely complicated and cost-prohibitive. Jim Greenlee -- The Shadow Georgia Insitute of Technology, Atlanta Georgia, 30332 ...!{akgua,allegra,amd,hplabs,ihnp4,seismo,ut-ngp}!gatech!gitpyr!jkg