Posts Made By: Vladimir Sacek

Sensitivity of the Ronchi test in such setup is sufficient to show only large errors. For instance, with 100 lines per inch, f/6 and 3 to 4 pairs of lines intercepted, it is according to Suiter less than 1/2 wave. A good Barlow increases the sensitivity in proportion to its magnification factor; if the scope can't focus with it, you can always add a bit of extension.

>The mirror star tests slightly undercorrected (bright outer ring inside focus and vice versa.) I find this somewhat puzzling as the center is clearly too deep and the majority of the tests suggests that the edge is also low.<

Center portion is of little consequence. Mirror
area from ~0.5 radius to the edge contributes ~80% of the light , and determines the actual error.
Most of the graphs indicate that measured wavefront deviation at ~70% radius is lower than both 0.5 radius and the edge. That makes the outer 0.5 radius of the wavefront bent stronger than corresponding section of a perfect wavefront. Hence
symptoms of undercorrection.

Vlad

Hi William,

>just found Vlad's one-week-old (!) message in the >ATM List: (my interpretation) the coma will be >worse by 1/obstruction, astigmatism by 1/obs->squared.

It is even somewhat worse: it is not obstruction, but a minimum secondary size, determined by the width of the axial cone at secondary's location.
The reason is that any incident angle A at the primary becomes A/x after reflection from the secondary (either Gregorian or Cassegrain), "x" being the min secondary size in units of the aperture. So with x=0.25, all off-axis angles become four times larger after reflection from the secondary. That would make inherent coma of the imaging instrument four times worse, and its inherent astigmatism 16 times worse (which would
inevitably worsen field curvature too). The displaced aperture stop (which is for the imaging instrument approximately at the location of secondary mirror), would have some effect, but likely minor compared to that caused by the increase
in angles of incidence.

Vlad

It may not be colimation related.
Dominant miscollimation aberrations for all two-mirror telescopes are coma
(for tilt and decenter) and spherical
aberration (despace). There is always
a particular combination of tilt and decenter that results in no coma error, but astigmatism is so small in comparison that it would take gross miscollimation of this (unlikely) type to show any significant astigmatism.

Vlad

Hi Kevin,

The Ronchi is not really accurate at the r.o.c. of fast parabolas. The reason is that they suffer from significant spherical aberration with objects placed so close. In fact, a parabola at r.o.c. has as much of wavefront error s.a. as a comparable sphere at infinity focus (only of opposite sign). This causes significant distortion of Ronchi lines with even perfect parabolas, and relatively small additional deformation caused by comparatively much smaller deviation from perfect is often times hard to measure with sufficient accuracy.

This particular mirror, if perfect parabola, would have 2.1 wave p-v of overcorrection at the r.o.c. (from 0.89D/F^3, for D in mm). If you want to try to figure out possible errors from the Ronchigram, you can use one of available softwares at

http://www.users.bigpond.com/PJIFL/ronchi_software.html

If you go with Peter Smith's software (simple and user friendly), just remember that this parabola, if perfect, should have 2.1 wave of overcorrection
at the r.o.c. So if you get, for instance, 1.9 wave, that indicates figure error resulting in 0.2 wave of s.a. (undercorrection).

Still, probably more accurate result could be obtained with star testing the mirror.

Vlad

Classical Cassegrain has field advantage, but for an ultimate planetary scope what matters more is the mid-field quality. In this department, Dall-Kirkham is likely to have practical advantage, due to its lover sensitivity to misalignment.

Assuming an f/4.5/18 system, the DK would have 8.1 times smaller quality linear field than the CC. But this is still at the level of linear field of an f/9 system, with about 16mm diffraction limited field diameter (7.6 arcmin). With good tracking,
more than necessary.

On the other hand, as little as 1mm decenter of the secondary would induce 0.23 wave of center-field coma with the CC. That alone would push its Strehl down to 0.83. Add 1/10 wave from despace (about 6mm, with ~1/60 wave per mm), and the wavefront is just at the diffraction limited level.

At the same time, the DK would have 2.65 times smaller error from the decenter (1/11.5 wave, or ~0.97 Strehl), and four times smaller error from the despace - entirely negligible. A tilt error (to which both are equally sensitive) which would knock DK's Strehl down to diffraction limited level (about 1/14 wave RMS), would have the CC down to 0.66 Strehl (1/2.9 wave s.a. level) in the very middle of the field.

Vlad

Vic,

Why would you go with classical Gregorian, if aplantic one is not harder to make? If your secondary magnification is 6 and back focus 0.167 in units of primary's f.l., aplantic primary would have conic of -0.989 (vs. -1 for the classical), and aplantic secondary -0.54 (vs. -0.51 for the classical).

Vlad

Floyd Blue said:

OK, so we now know that figure and polish and PV and RMS and even Strehl are all different characteristics.
We are also informed that Wavefront error is not the most important of these attributes.

Although Scott already addressed this "concept", it won't hurt to do it again, just because those are real basics of the concept of optical quality.

There is an optical surface, and there is a wavefront produced by that surface. Any surface deviation from perfect will cause wavefront deviation. That includes gross figure errors, as well as medium, small and smallest local errors on the surface - anything and everything. This means that the Strehl as a measure of wavefront quality includes effect of any and every concievable surface imperfection, be it large, or microscopic, or anything in between.

It is easy to calculate Strehl for any type of optical surface;
difficult part, as Scott pointed out, is to have surface deviations accurately measured.

That brings us back to Floyd's question, that despite this mile-long thread still hasn't been answered to his satisfaction. I don't think he asked us whether his subjective perception of a new scope is worth the money - it seems to be rather obvious.
What he wants to know, I believe, is an *objective* difference between a mediocre 1/4 wave (of spherical aberration) level and near-perfect (according to the rating) mirror worth extra money.

The answer is: all depends. Assuming that the 0.997 Strehl of his 10" f/6 is true, and adds considerable extra cost vs. ~0.96 true Strehl mirror, the answer is: no, it isn't. The problem is, no one knows for sure what the actual mirror Strehl is, because it hasn't been adequately tested. Chances are, it is lower. So, in this air of uncertainty, going for the highest available rating at least increases the odds of getting mirror that is close to - or even somewhat better than - the level which can be called "sensibly perfect". And if it resulted in getting a mirror with a true Strehl of 0.95, or even 0.9, the answer is
resounding YES: it is worth the extra money vs. mirror with a 0.8 true Strehl (1/4 wave od spherical aberration level).

Vlad

Gene Griggs said:

I have a large ATM Maksutov Cassegrain project that's around f/6.5. Before I finish this out I'm curious if it's possible to use the same corrector and primary but just switch the secondary and make it a Mak Newt...

Has anyone ever tried this on a slow Mak Cass design -- or is it simply not going to work due to the correction of the meniscus? Note: The secondary is not a spot on the meniscus but an actual secondary mirror with no magnification.

Are you saying that your Mak-Cass is designed with flat secondary? That would be highly unusual, because either results in inaccessible focus, or requires >50% obstruction to make it accessible. Also, it makes hard to correct for coma, since corrector alone cannot generate enough coma to correct the opposite coma of the primary (unless the corrector is placed closer to primary's r.o.c., making the system long)

If the secondary is flat, then all the correction for spherical aberration is at the corrector/primary, and you could use any configuration with a flat secondary mirror you want. If (spherical) secondary is curved, then it takes part in system's spherical correction, which means that replacing it with a flat would result in a system with significant undercorrection.

Vlad

Darren Drake said:

I recently obtained a second hand 17.5 inch f/4.5 Coulter mirror which was tested to be 74% corrected from sphere to parabaloid... I've been wondering if the undercorrection is evenly spread does anyone know what the wave front rating would be?

The 74% corrected should mean the appropriate conic K=-0.74. The max surface error vs. parabola is given by 45D(K+1)/F^3, with "F" being the mirror F# and D the aperture in inches. This gives the max surface error of 2.25 waves.

This error doubles in the wavefront at the focus of either paraxial or marginal rays. However, at the best (diffraction) focus in between (0.707 zone focus), the error is only half of it, or ~1.1 wave.

Vlad