[guide-user] Re: Scope tracking, ST8/9, etc.

Markus G. Kempf Jan 27, 2000

Maybe it is a matter of design philosophy. I would prefer one of the two
extreme positions. Either get high speed positions at 18.2Hz and feed them
to the existing servo code (almost no change at all for Mel) or get the
orbital elements of the satellite and do the position calculations in
the TCS. I'm shure, that is no problem for a pentium class machine.
No more timing problems, no jitter, no synchronization. To
get polynomial coefficients every second or minute and then compute the
new positions at 18.2Hz has no advantage IMHO. With the track rates in
sideral coordinates you still have to convert into track rates for the
mount. At its heart, Mel's system is a position controller.

> I would prefer rates rather than a higher frequency absolute position
> update. My system, I would imagine, is not the only one where a 'move'
> means something different than a 'track' command. Specifically, the
> 'move' means to startup, make the move, and most importantly, bring the
> scope to a safe stop with no overshoot. A 'track' on the otherhand
> means to continue moving at a rate (in my program case, I generate a
> rate based on positional information that is updated at the timer tick
> frequency).
>
You have to do that distinction because DOS is not a real-time
operating system. Also because of the way you have implemented it,
there is some amount of time every 55ms used for the user interface,
position updates and other stuff. It's my understanding, that the
mount does not move during this time. Therefore your 'Track' modus
is really a 'move' with an ideal motor (the rate is lower than the
max. start/stop frequency of the steppers). At the end of the period
the velocity is zero. To have true rate (velocity) control, you
would need a continously running thread. That's diffcult without a
true RTOS.
If this is wrong, please correct me.

Markus Kempf