On Mon, Dec 19, 2011 at 3:25 PM, Al Kossow<aek at
bitsavers.org> wrote:
On 12/19/11 11:46 AM, jim s wrote:
I though these were either 2311 or 2315 IBM type
technology, but one could
use the embedded technology
They are actually IBM 5440 physical packaging, first
used on System/3.
The cylinder positioning uses an embedded servo, on track is hard sectored.
Interesting. I'm not surprised to learn the form factor came from an
IBM drive, but I'm now wondering how many other drives used the same
physical pack types and if they were all embedded servo or if any
could be formatted in the field (i.e. - if one has a degaussed pack,
_can_ it be re-used anywhere).
I don't have it anymore, but 20-ish years ago, someone gave me a drive
set that I _think_ was Plessy-badged, with one fixed platter and one
RL02-type top-loading pack. If I'm right in remembering it took an
IBM 5440-style pack, I think I'm also right in remembering that you
could *not* put a DEC RL01 or RL02 pack inside and expect it to work -
you had to use packs formatted for that drive. I don't think there
was a risk of damage, just a lack of successful functionality.
-ethan
Is this the drive that is about 20mb or so on the platter? CDC had a
drive that had one platter like this over 2 or 3 fixed and was 80mb
total. It used winchester heads, I think. since the platter on top was
spun by the same spindle as the fixed ones there had to be some
different method of finding the data for the removable packs. I never
looked at the pack for sector marks.
I vaguely recall one of the CDC drives having a physical index, but the
use of it in formatting data was optional. We had a controller that
would get on track when formatting and lay down the sectors with 0 data
payload, and when we would do our verification, we would first re-record
the format aligned with the index mark, then precess by 1/2 sector for
each retry. The dropouts were guaranteed to be permanent write splice
errors, so we didn't worry about anything but getting a good track laid
down if the first guess failed.
The controller interface and drive was SMD in that case.
Jim