Camera advice

[xilman]

Originally Posted by diep

I have on paper (just mathematical paper) also a R-C design here where i chose the parameters such of the primary mirror that it can be grind as a parabole, as the difference between a hyperbole and parabole is far under 1 nm which is impossible to grind to that accuracy (the accuracy of the grinding is going to be factors factors worse than that of course).

I don't understand this point. A parabola has eccentricity of 1.0 and a circle of 0.0, with an ellipse somewhere between those two limits. A hyperbola has an eccentricity >1.0 with no upper limit. A hyperboloidal surface could, in principle, be very deep and have a very narrow aperture though that would be very difficult to produce to 100nm accuracy(0.2 \lambda in the visible spectrum). In practice, at the focal ratios commonly used, the discrepancy between a paraboloid, a sphere, and a hyperboloid is very significant. It's the primary cause of spherical aberration and coma or, equivalently, between a spherical and a paraboloidal mirror which respectively have zero of the opposite aberrations.

[diep]

Originally Posted by xilman

I don't understand this point. A parabola has eccentricity of 1.0 and a circle of 0.0, with an ellipse somewhere between those two limits. A hyperbola has an eccentricity >1.0 with no upper limit. A hyperboloidal surface could, in principle, be very deep and have a very narrow aperture though that would be very difficult to produce to 100nm accuracy(0.2 \lambda in the visible spectrum). In practice, at the focal ratios commonly used, the discrepancy between a paraboloid, a sphere, and a hyperboloid is very significant. It's the primary cause of spherical aberration and coma or, equivalently, between a spherical and a paraboloidal mirror which respectively have zero of the opposite aberrations.

https://en.wikipedia.org/wiki/Ritche...tien_telescope

Ritchey-Chretien uses for the primary mirror a hyperbolic primary mirror.

A hyperbolic mirror is very difficult to test during grinding.
A parabolic mirror is far easier to test during grinding.

If you take the parameters such of the hyperbolic primary mirror of R-C then there is no difference in practice with a parabolic mirror, if we take grinding accuracies into account.

good example is hubble telescope. The primary mirror on paper should have been a parabolic mirror, which just in theory is a hyperbolic mirror.

Yet maximum deviation for hubbles primary mirror, so on paper in theory (not what actually was delivered) versus a parabolic mirror is less than 0.1 nm when i did do the math for it using the public published numbers.

That's because primary mirror is what is it something like 2.40 meters diameter and it has a focus 4+ meters away.

So with a very simple test the primary could have been grind back then.

If you grind a primary mirror of just 40 CM or so, then there is a huge difference between a parabolic and hyperbolic primary mirror and it'll be way more complicated to setup testing for the primary mirror for testing during grinding.

If you take a mirror 1 meter diameter with a very long tube, then you again get back in the easier to grind area (apart from the more massive job for the primary) :)

Of course i ignore the secondary here - that's yet another story.