A Cylindrical-mirror telescope

[digby dart]

I saw this a couple of years ago on the internet. I came across it again tonight and thought someone else might find it interesting. :)

http://www.bruno.postle.net/neatstuff/telescope/

I find the idea fascinating. "A proposal for building a telescope using a number of cylindrical/parabolic mirrors instead of spherical/parabolic mirrors. Bruno Postle, February 2001."

[Bill Cooper]

Hi Digby,

I find this design fascinating. It has the ability to be scaled up fairly easily allowing one to build a large collecting area fairly cheaply, particulary if all you want is visual observation.

No one I have talked to has be able to spot a flaw in the optics of this system. Suppose the only way forward is to try and build one!

Cheers,

Bill Cooper

[Meru]

Wow looks cool. I'd love to see one being built up as well

[ZeroID]

Interesting possibilies for a binoscope design maybe ..still trying to 'spot the flaws'.
I actuually wonder if it is a lot harder to make an accurate flat surface mirror that could be curved somehow than grinding a solid surfaced parabola.
Or grind an accurate curved surface mirror ...?
???
I have a collection of various curved and flat front surfaced laser printer mirrors\optic elements from work. Must have a scull through them to see if the basic principle is feasible...

[bojan]

No way.
This is good for collecting light (heat), for sun furnace for example, or energy generators (because cylindrically shaped reflective sheet is cheaper than rotational paraboloid), but not for imaging of any significant field - too many distortions to deal with (astigmatism, to start with, then coma...).
Could be also useful for spectroscopy, where you do not need the image of an object... but not for imaging.

For imaging, you have to have the wave front dot disturbed (meaning, all distances from focus towards the entry aperture must be the same expressed in wavelenghts, down to 1/10 at least.
This can't be done with cylindrical system of mirrors in all orientations relative to optical axis.

[Wavytone]

Don't bother, it won't work for several reasons and the originator is so hopelessly ignorant of basic optics that I am lost as to where to start. It's on par with the mathematically illiterate trying to disprove Einsteins theory of relativity.

Cylindrical trough mirrors have their uses though, as solar collectors and for some radar applications where the beam is essentially swept over a plane. As Bojan points out they're also used in spectroscopy and in some scanning applications where there is a linear element in the path - either a grating (a bunch of lines), a prism, a narrow slit, or a linear CCD detector.

But what that article proposes is really really flawed in so many ways.

Firstly, he doesn't understand that a parabolic cylindrical mirror produces a line focus (not a point) from a distant point object (star or the sun). If you use two cylindrical mirrors with the axes at 90 degrees, what you will get is two lines crossed, not a point.

In each plane, what he shows is one-half of a cassegrain two mirror telescope. What he shows is afocal - no image plane and the rays are parallel, not converging to form an image. So IMHO he doesn't even understand how an image is formed.

He's also ignorant of the geometry concerning the angular magnification in a two-mirror telescope, and that by displacing the mirrors along the optical axis, while it may produce the same magnification in the vertical and horizontal planes, the angular magnifications will be different and there is not the slightest chance of this producing a coherent image.

I'll give you an easy analogy that most might understand. Buy two 8" Maksutov cassegrains from Intes Micro, one being an f/10. the other being f/12. Using a diamond saw, cut both in half along the optical axis.

Now glue the left half of the f/10 scope to the right half of the f/12 scope. Good luck trying to combine the beams into an image !