Fabrication and Tolerancing
Sorry, page under construction! :)
Let me just add quickly, in response to the number of people expressing a very valid concern, that centration and spacing tolerances are only about half as tight as a cass with the same f/x primary. And we're still ahead of the game, since the spherical primary has no pre-defined optical axis, primary tilt and decenter are essentially the same thing. Which leaves us with two main things to pay attention to: primary to triplet spacing, which must remain constant through thermal cycling to within 10 mils or so (a change of 1mm reduces the strehl ratio to 0.8 -- carbon-fiber struts or tubing material called for here), and centration of the upper ring assembly (triplet, diagonal, doublet and focusser as a unit), which means that "collimation" can be conveniently accomplished while looking through the EP, through the use of micrometer adjustments of this assembly (a decenter of 0.25mm again reduces the strehl to 0.8).
And thanks for all the nice emails! -MP