Re: Floppy disk density

[Eric Smith <eric-no-spam-for-me@brouhaha.com>]

Andy McFadden replied:
> I believe the 5.25" floppy drives for the Apple Lisa used some of these
> tricks media to get as much as possible onto the disks (1MB?).  Of course,
> the drives had dedicated CPUs...

mjmahon@aol.com (Michael J. Mahon) writes:
> Yes, that's my understanding of the Twiggy drives.

Just to be clear, the *drive* does not have a CPU.  The Lisa I/O card
has a CPU (6504, IIRC) to control a pair of Twiggy drives, which was
later used for a single 3.5-inch drive.  The original controller used
logic almost identical to the Disk II, though running at twice the
clock rate.  The newer I/O board used in the Lisa 2/10 and Macintosh XL
uses an IWM chip.  There is 1 KByte of RAM dual-ported between the
main CPU (68000) and the floppy controller CPU, though the 68000 must not
access it during floppy transfers.  The floppy controller CPU generates
an interrupt to the 68000 when a floppy transfer is completed.

The Twiggy drive has speed control implemented by shifting a speed
control byte into a shift register in the drive.  The disk is divided
into zones that are accessed at different speeds.

Unlike most double-sided floppy drives, the Twiggy had the top and bottom
heads on opposite sides of the spindle.  This was done so that each head
would be opposed by a pressure pad, like a single-sided drive, in order to
reduce head and media wear, rather than having the two heads directly
opposing.  The two heads use the same positioner, howeover, so that when
the top head is at the outermost track the bottom head is at the innermost,
and vice versa.  Crude ASCII art:


    ------- upper
    |       head
    |         v
    |      --====cutout====----spindle----====cutout====--    <--- floppy
    |                                      ^
    |                                    lower
    ------------------------------------ head
      actuator moves <------>


Thus when the motor speed is correct for one head, it is wrong for
the other, except possibly near the middle track.  Most systems with
double sided disks order the track access 0T, 0B, 1T, 1B, 2T, 2B, etc.
(T and B representing top and bottom), so that two tracks are accessible
without repositioning the head.  But on the Twiggy, it takes much longer
to change the motor speed than to seek one track, so the order is
0T, 1T, 2T, 3T, ... 44T, 45T, 45B, 44B, ... 2B, 1B, 0B.

The 400K 3.5-inch drive requires the controller to supply a PWM signal
for motor speed control.  The original Lisa I/O card did not provide
this, so the 3.5-inch drive cage assembly used in the Lisa 2 and Lisa 2/5,
and in upgraded original Lisas, has a small PC board called a "Lisa Lite
adapter" which provides the PWM.  The new Lisa I/O card used in the
Lisa 2/10 and Macintosh XL includes the PWM circuitry so no Lisa Lite
adapter is needed.

The later 800K drive does control its own motor speed, so it can be
used in a Lisa with the old I/O card with no Lisa Lite adapter, provided
that you wire a suitable cable.  Use of an 800K drive does require
different I/O processor firmware which was available from Sun Remarketing.

Eric