[Wavytone]

John:

What you suggest is true if you were trying to achieve raw blind pointing accuracy using absolute encoders - in just the same manner as a theodolite, and it is partly why a decent theodolite with circles reading to 1 arc second was (and is) a very expensive beastie.

But you are also implicitly assuming this is the way the DSC software works, which is incorrect.

None of the DSC's rely on absolute encoders. That they all require calibration on reference stars gives away the fact that they all use incremental encoders (orders of magnitude cheaper) or are counting stepping motor "steps" (how all those GOTO mounts work which require the axes to remain locked after calibration - Celestron CPC, Meade LX, Lightswitch). Either way its the same approach. Even the mounts with encoders invariably have incremental ones - not absolute encoders.

The point is to calibrate on several reference stars, the more the better if they are spread across the sky and encircle the horizon. This deals with most of the unknown mechanical errors and mount flexure in much the way an auto guider does.

There's also an opportunity for some smart cookie to figure out a way to calibrate these encoders as-fitted to a mount, right down to the step. I'm thinking of some sort of dingbat that could be set up in a basement. Solve that and these would be as good as absolute encoders.



The only thing these can't solve - and which an auto guider does solve - is atmospheric refraction.

Suggest that instead of pontificating on hypotheticals based on false assumptions, get out and try a scope fitted with these, and see what it can do. Even better try the trick of using SkySafari at maximum zoom to track Saturn blind, while someone else observes through a scope at 300X or more. That is simply not possible with low-res encoders as the angular resolution and feedback frequency is too low. You're welcome to visit and have a decker through my Santel at the same time.