Yes the disk rotates continuously, the spindle is driven by an AC motor
that runs as long as power is supplied. Since this is a single sided
drive the head is against one side of the diskette and there is a
pressure pad on the other side that presses the diskette against the
head. This pressure pad is normally held away from the diskette by the
head load solenoid and only dropped when the diskette is being
accessed. I can think of a couple things that might cause what you
observed. The first would be maladjustment or failure of the head load
solenoid that would result in the pressure pad always being in contact
with the diskette, the other is media degradation, the binder in the
media coating breaks down and the coating comes off in a gummy mess.
When double sided diskette drives came out IBM used a head carriage
design that lifted both heads away from the media when idle.
In the family of the 3660 supermarket system there was a controller that
was designed for small stores that did not support scanners and operated
off of a diskette instead of the usual 3651 store controller that ran
off a hard disk. When it was initially released it would store a
cashiers totals in the same location on the diskette, which would mean
that every time the cashier hit the total key on here register that spot
would be accessed. When first installed the customer started to
complain about diskettes life in the controller, so an investigation was
started and after doing some quick calculations it was determined that
the number of times the spot on the diskette where the totals where
stored far exceeded the specified maximum number of accesses in the time
between a new diskette being installed and it wearing out, The solution
was to alter the microcode of the controller to move the spot where
totals where stored around so that the same spots on the diskette where
not continuously pounded.
Paul.
On 2026-06-09 06:08, Peter Coghlan via cctalk wrote:
Paul Berger wrote:
Several years later I was a hardware support person in the Toronto
Lab and one of the rooms I looked after had a 3880 controller with a
couple 3380 DASD units and the first time I heard the 3880 start up I
thought I know that sound, so I looked inside and sure enough it
IMLed from a 33FD.
In the 1980s, the college where I studied and later worked had a 3880
managing
two strings of 3375 DASD with four units in each string. This
arrangement ran
for several years.
On one occasion, after a power failure, the 3880 failed to restart. I
later
heard from the people that maintained it that the floopy disk it imled
from
had either a ring of oxide worn off or the actual disk itself worn
through
due to prolonged head contact with the same track of the rotating disk.
The implication was that the disk had been rotating continuously even
though it
was only required at iml time. Was this the intended mode of
operation or was
the motor supposed to stop once iml was complete?
As far as I recall, the floppy drive was mounted vertically in the
3880 cabinet.
I managed to snag the switch panel from that 3880 when it was eventually
retired.
Regards,
Peter Coghlan.
