3 Oct
2024
3 Oct
'24
2:41 p.m.
On Thu, 3 Oct 2024, dwight via cctalk wrote:
A small laser interferomenter and a screw driver could
be used, once one determined the center of the track by magnetic material and a
microscope. Some what special equipment but not all that special, now days.
Years ago, I went to a Seagate building to help a friend with a servo writer problem. It
only took a few minutes so I never charged them. Of interest here was that they used a
stepper motor to control the position of the head. The problem was that they wanted to
turn the power off to the stepper motor to remove an stray magnetic fields.
They did it too soon while the motor was still in motion, causing the motor to not stop
at the desired location. Adding some delay solved that one. So, the point is that a
stepper was enough for a hard drive servo track.
Of other interest, they were using an AIM 6502, running figForth. The programmer was
obviously a previous BASIC programmer because he'd named vaiables with names like A,
B, ... , instead of meaningful names that would have made the code more readable.
Dwight
That delay is usually handle in software/firmware in the host machine.
After a step pulse, wait some time before another step, or other drive
access, so that the head finishes moving, and "settles".
On PC, for floppies, there is a parameter block pointed to by Int 1Eh.
"Over-clockers" like to cut that track to track time down to the lowext
that they THINK is adequate.
One (of several) reasons for the introduction of PC-DOS 2.10 was that the
PCJr used Qunetrak 142 drives that were too slow for PC-DOS 2.00, so they
had to increase the time. Sure enough, an over-clocker put out a patch to
shorten that extended time.
(-: might that be a use for the relays in the
"[cctalk] Might be antique computer parts" thread? :-)
(emoticon captioned for the humour impaired; there are better ways to delay)
--
Grumpy Ol' Fred cisin(a)xenosoft.com