[Ground-station] Satellite program

[Phil Karn]

On 5/16/18 22:29, Bruce Perens wrote:
> Here is SATNOGS rotator V3: https://wiki.satnogs.org/SatNOGS_Rotator_v3
> The whole thing is built from extruder-printed parts and T-slot, and a
> few easily-acquired parts, so someone without machining capacity can put
> it together for a bit more than $220.
> It would be nice to have an ME to make another design or to improve on
> this one.

That's the one I saw. Because they have actual experience building this
thing and I don't, I won't presume that any idea I might come up with is
necessarily better than theirs. So it's certainly worth looking at and
then seeing if we can improve it.

Note that they mention X-Y pointing as an option. I think that's
something we'll definitely want to consider as well. Any two-axis
antenna pointing system necessarily has singularities; the usual az/el
system has one straight up that can cause azimuth slew rate problems
when a LEO passes nearly overhead.

The usual fix (and the one used by NASA's dishes intended for LEOs) is
to instead make both antenna axes horizontal. Now it is easy to cover
the area directly overhead, at the cost of introducing *two*
singularities at the horizon 180 degrees apart. (The az/el system also
has two singularities, but the second points straight down, which isn't
a problem in practice). On the other hand, satellites don't move very
rapidly at the horizon, and you often can't see/hear them anyway because
of terrain obstacles.

You can also avoid singularities by adding a third degree of freedom
(i.e., a third rotor perpendicular to the other two) that keeps any
singularities out of the way. This actually gives you back an azimuth
axis, so you get the best of both systems at the cost of a third rotor.

All this is closely related to the famous "gimbal lock" problem that got
much mention in the movie "Apollo 13".

Phil