[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

RE: commutating antennas (was Re: Canted S-band antennas on satellite)

At 09:34 AM 5/3/2002, Ken Ernandes wrote:
>The technical challenge is having accurate, real-time knowledge of which
>side of the spacecraft if facing the Earth.  I've been pondering

I think this can be done easily.

Note that this is a spinning satellite, and I've proposed a low 
inclination, and spin axis roughly normal to the orbit plane.  That means 
that simple sun & earth sensors can be made by putting a photocell in a 
tube behind a slit pointing out one of the side faces.  The sun and earth 
spin by once per rotation.

The more difficult attitude determination problem occurs in NONSPINNING 
spacecraft, such as AO-40 in 3-axis stabilized mode.  (Damn near 
impossible, in the case of AO-40 because it doesn't have attitude sensors 
designed for 3-axis stabilized operation.)

We're only dealing with one dimension here.  I presume that the spin axis 
is fixed over the short term.

One approach is to only sense the position of the sun.  The software would 
note the time that the sun goes by the sun sensor, and lock to the spin 
rate.  Assuming a roughly circular orbit, if the software also knew the 
orbit period and mean anomaly, then it could trivially compute the 
direction of the earth.  The precision requirements here are not 
difficult.  Suppose we want to switch within 3 degrees of the correct 
angle.  If we allocate half of the error budget for the error in knowledge 
of orbit, then that's 1.5 degrees error in propagated mean anomaly.  That 
seems like it might be workable.

If one were willing to sense the edges of the lit earth as well, then need 
for precise knowledge of the orbit period and mean anomaly mostly goes 
away.  The software still needs to know a little bit about the orbit, 
because unfortunately the whole earth disk is seldom lit.  One edge of the 
illuminated region is a real edge of the earth, and the other is 
not.  Software needs to know enough to decide which edge of illuminated 
region is the real edge of the earth and which is not.  One could just 
sense the edge of the earth, and use a fixed offset from that as the 
direction of the center of the earth.  If orbit is noncircular, then a 
little calculation would be used to do something a little better than just 
applying a fixed offset.

Sun & earth sensors have operated on several AMSAT spinning spacecraft, so 
nothing new about the hardware required, right?

Is there something I've missed that you believe makes this difficult?

Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org