I suppose I should have been a little clearer. The closed loop includes encoders indeed -- one for each axis. This is the model: www.usdigital.com/products/encoders/incremental/kit/EC35
Here is a picture of how it is set up:

If you look closely, you can see an encoder on the same shaft as the motor. It was just easier to mount the encoders on the outside. I toyed with the idea of getting Renishaw encoders on the driven shafts attached to the scope, but ultimately decided backlash from the drive train would be manageable, since I'm using pulleys. So, I put these encoders on the shaft up the drive train. This essentially multiplies my quadrature count per revolution (40,000) by the gear ratio 28.8:1. This is indeed ~1 arcsecond resolution. Now, if I can match that resolution with calibrated accuracy, taking into account PEC, only time will tell.

In any case, it's a little late to be switching to steppers, and believe it or not, I'm doing it this way for a reason.

Eventually, I'll add field rotation to the focuser for astrophotography. Yes, you can take long exposure photographs with an Alt-Az mount -- you just need to account for field rotation. This is a prototype that I would like to turn into a commercial product at some point if everything works out. Right now, I'm making it for my father though.

I'll check out this "IND alignment subsystem." Indeed, it sounds like exactly what I need to get up and running. Thank you for all the suggestions so far.

Read More...