Having set targets and goals what follows is drafting the type of mount. I suppose this part will be pretty arbitrary...
Few types were casually taken into consideration with their pros and cons. Equatorial as GEM is ruled out as it would have to be huge and heavy - mainly driven by the overhang, big bearing, big shafts likely in steel to contain costs, last but not least the drive would require large worm gears. Too heavy and expensive.
So we shift our thoughts to Alt-az. Weight load is inherently more balanced leaving room to more humble materials. Only drawback that I can really see? Field rotation. Will come back to the de-rotator at a later post.
Drafting the scope we ended up in something I'd call a glorified Dobsonian. A Dob in its essence with few upgrades to turn it into a mount that can track for imaging. Big appeal is represented by the friction drive which perfectly fit this configuration.
In hindsight, the horseshoe set-up earns a special mention - it's very palatable as is equatorial and it would adopt very similar material/manufacturing processes as the glorified Dob.
How to turn a Dobson-like mount into a proper imager? What makes a mount an accurate tracking system? I reckon this boils down to 2 main elements - Low backlash and sound structure.
How to achieve that? friction drive, stiff yet light structure, autoguiding capability.
Quick math check to start with.
A perfect structure with 1micron (0.00004") error on the friction rollers equals a drift on 0.23arcseconds on the focal plane. Now, if you consider a sensor with a generic 10micron pixel, every pixel equals to 0.79arcseconds (at least for this mount)
Reality check: a commercially available ground bar or shaft has a run-out of 13 microns (ISO h6) - just to consider a relevant geometric error.
If you are following me in this generic and very simplified calculation, long focal length optical systems cannot perform accurate tracking based solely on geometrical accuracy. It's just not robust enough.
Hence we need autoguiding...how does a mount look like to achieve an effective autoguiding? Again low backlash via adopting friction drive paired to a stiff and light structure that promptly reacts to tracking adjustments. As you can see it's a whole package of traits that needs to be developed and implemented holistically.
If one aspect is missed or poorly executed the entire system is not capable.
Pics of parts I had a head start with:
- 2 driven azimuth roller assemblies . The third one will be the driving one.
- The wood preparation for the upper cage.