I've never designed a guidance system before, so I am speculating... but it
seems there would be a couple ways to deal with this.
(1) The gyro-systems could have considerable moment of inertia and
considerable angular velocity which would combine to make a lot of angular
momentum. Then, when the telescope, antenna, whatever is rotated "against"
the gyro-system, the angular displacement of the gyro-system would be
"insignificant." Here we would have to define insignificant in accordance
with the positioning error we can tolerate. Anyway, if the relative moments
of inertia between the reference system and pointed object are different
enough we might be able to live with small angular errors (which could
certainly be calculated and corrected.)
(2) If we really want to keep the reference frame fixed with respect to a
particular set of stars, we can put photo-detectors on little telescopes
aimed at the stars, and when the reference frame begins to move off the
stars, the computers controlling the gyro-system can accelerate or
decelerate the appropriate gyros to keep the reference frame on the stars.
(3) Option (2) is cumbersome and no doubt tricky and expensive. Another
less cumbersome way to approach it would be to have massive high-speed gyros
for providing the reference frame, but have another smaller gyrocompass
system for positioning. The smaller gyro-system is "isolated" via gimbals
and has optical sensors to note whenever the reference frame begins to
rotate with respect to the gyrocompass. Then, as in (2), the massive
gyro-system has the individual gyros accelerated or decelerated as necessary
to realign the reference frame to the smaller gyrocompass.
At any rate, I think (3) is the way it's done. A gimbaled gyrocompass
maintains the desired orientation, and a more massive gyro system provides
the required "inertia" so that the telescope can be moved relative to the
frame without moving the frame too much. The frame can then be kept in
alignment or brought back into alignment with the gimbaled compass by
adjustment of the large motors.
Michael D. Edmiston, Ph.D. Phone/voice-mail: 419-358-3270
Professor of Chemistry & Physics FAX: 419-358-3323
Chairman, Science Department E-Mail edmiston@bluffton.edu
Bluffton College
280 West College Avenue
Bluffton, OH 45817