Telescope Reviews: 2008 big project home made 40X203 big binocular

Pcad: Does your name come from work you may do with CAD?

I believe that you will see the image rotation if you confine your attention to only the rays from the target which come from the second mirror. This may require construction of a model in which there is no physical roof. The roof exists as an imaginary line in space where the extensions of the planes of the mirrors intersect. Second surface acylic mirrors suffice. I built such a model. A door hinge may help.

The Amici roof mirror idea appeared to me in the mid '70s. But the optimization of positioning of mirrors in 3D space and choice among the additional methods available for an interpupillary mechanism are more complicated than with the monolithic block of a prism, particularly if one departs from the usual 90 deg. deviation of the line of sight( except in the extreme case of 180 degree deviation).

Luciano: Is that a computer graphics drawing, or drawing by hand? Very clear. But you have shown the simple case of a point target on the optical axis. The true situation must include the inclined rays from the top and the bottom of the objective which illuminate the edge of the field in the focal plane, which will be passed only if the mirrors are larger than the sizes required by the axial image point. How much larger depends upon how much vignetting one allows. Using mirrors , why not illuminate the edge of the field in a way not limited by mirror size ? Cost , size, and weight considerations require the acceptance of more vignetting in the case of an otherwise equivalent prism. You probably know this, because your drawing shows mirrors larger than required for the axial point. Does your program(?) give quantitative information for mirror size, angular and xyz location in 3-space, full illumination of the field, etc? A 2D drawing of the field rays, with the mirror planes shown as lines inclined to the axial ray by the angle of incidence, helps in optimizing. But how can one know the angle of incidence without prior 3D solid geometry 3-vector calculation? That it is 60 deg. for the 90 deg. deviation case has been revealed. There remain the problems of finding specific coordinates in 3D to locate the mirrors after minimizing the required sizes, avoiding physically impossible overlapping of the mirrors, and inclusion of 3D effects from the chosen method of changing the interpupillary distance, as well as changes of the mirror shape from elliptical to round, rectangular, truncated or not, mummy shape, tapered, untapered, thick, thin, the mirror mount, etc. I found that my intuition in 3D was fallible, and that it is best to rely on the vector calculations. Such tools are no doubt in CAD and optical system layout programs with graphics, so one might not need to study and program vector solid geometry( which is not very difficult if one knows trigonometry), which I found to be enjoyable and stimulating. The only calculus I programmed was a derivative to minimize the spacings of the mirrors, as functions of their edge shapes and presence/absence of taper.

The early Matsumoto 90 deg. deviation systems used a Pentax 67 accessory focuser , between the mirrors to vary the IPD, and then the user refocused each eyepiece, is this not correct? Does he still use the Pentax 67 accessory?