REFLECTORS FOR PLANETARY WORK

[Kaptain Klevtsov]

Just so I'm not misleading anybody then, as Ajohn seems to think that having the secondary mirror offset means that it won't be at 45 degrees to both optical axes (primary and focus tube). As per diagram

http://www.bbastrodesigns.com/OFFSET.JPG

The mirror in the diagram is at 45 degrees, but in fact it could be at any angle, except at right angles to the primary axis, and an offset would be necessary. From the diagram, at any point along the 'scope axis the cone of light extends the same distance up (in the diagram) above the aaxis as it does below the axis. Thats to do with it being a cone, nothing more. As the secondary mirror cuts the light cone at an angle, then the top (in the diagram) of the secondary mirror is further away from the primary mirror where the cone is of a smaller diameter. It follows therefore that the secondary mirror doesn't need to extend as far from the 'scope axis upwards as it does downwards in order to catch the cone of light.

Similarly, if you look at the diagram WRT the focus tube, the secondary needs to be more to the right (in the diagram) as the light cone is larger where the mirror insersects it.

The optical axis of an elliptical secondary mirror is not in the centre of the mirror.

If you were to consider the design of a 200mm Newtonian 'scope with a focal ratio of f/2 as an extreme example to illustrate the point, you'd get something like this.

(click to enlarge)

I trust that this clarifies the situation?

Kaptain Klevtsov

[beamish]

Works for me KK !

[KaStern_Former_Member]

Hello Kaptain Klevtsow,

I think that is right. Looks like in the Book "Star Testing Astronomical Telescopes" from R.H. Suiter.

Regards, Karsten

[Ajohn]

Looks like we will have to agree to disagree with this one. The reason I mentioned the not at 45 degree 2ndry was because of the meade scope and because it fitted in with my recollection of the s&t article. In some ways that doesn't matter because all sorts of arrangements are possible including one where the 2ndry is at 45 degrees. Maybe I didn't make that point clear.

My point about an off axis look to the main mirror in the 2ndry through the focuser still stands. Set like that the off axis field that the 2ndry captures will differ according to direction if the usual shaped ellipse is being used. In that case light will be missed on the other axis too. (I mean light that would be captured if the 2ndry was central.) I also stand by my point that the gain isn't worth while given the trouble it will cause in the adjustment of the spider even neglecting the missed light on the other axis. Would I argue with Suiter yes I'm afraid I would. I accepted the article too but then discovered the implications when I tried to set up the meade. That was a used scope that the seller had tried to set up with a laser collimator without fully being aware of what he was doing. He would have run into trouble using the usual eyeball method too. He was fairly open about the problem and I had spotted it anyway. I was just annoyed with meade. 8) I gained a free laser collimator too.

If anyone reading this has a solid none adjustable spider ideally a Vixen or some other manufacturer who isn't prone to variations in fashion it would be of interest to see if that is offset. As to me I'll stick to keeping it central if the 2ndry still has the normal shape. Maximum beam off the main mirror coverage for the minimum obstruction. If I find that the shape has changed or the scope will not set up like that it looks like some maths will be needed to sort our where it should be. The manufacturers should state a figure according to the scope really.

Also note the comment about SC's with a 13% central obstruction. I would advice working it back from 13% by area to by diameter to get the true figure. Think you will find it's around 30%. May even be more. There is a problem of that nature on most compound scopes and if the percentage area of the obstruction is given it's often based on area which is to say the least misleading.

John

[KaStern_Former_Member]

Hello Ajohn,

now I try to give an answer to your postings.

I do not write upon collimation, several good pages wich exlain newtonian collimation can be found in the internet.

I would like to say something to your other comments.

talk about the airy disc again. That's the little dot a star forms hopefully surrounded by nice round rings of light under high magnification. Lord Rayleigh came up with a limit for optical accuracy many many years ago but it's still valid today. He basically showed that to produce an acceptable image there must not be more tha 1/4 of the wave length of light error in the wavefront.

The 1/4 wavelength wavefront error criterium aplies for the error "spherical aberration".

This leeds to much bigger rms wavefront error than 1/4 wavelength astigmatism,

wich leeds to a much bigger rms wavefront error than 1/4 wavelength turned down edge,

wich leeds to a much bigger rms wavefront error than a small zone,

wich leeds to a much bigger rms wavefront error than a simple scratch.

So besides the maximum wavefront deviation it is important how big the part of the mirror surface is wich leeds to this error!

If it is a small zone the resulting rms wavefront error ist way smaller than if the whole surface is involved,

like it is when you have global spherical aberration.

In addition, if you have the rms (root mean square) wavefront error you can immediately calculate the Strehl ratio out of it.

So I much prefer the rms over the PtV (peak to valley) wavefront error.

In addition it is necessary to know the wavelength to wich th rms or the PtV is referring.

You can do an interferometrical measureing in different wavelength. In a pure reflector this is no problem to do,

the Strehl result will not differ (or only 1% or so due to the differing size of the airy disc in different wavelength.

The red airy disc is bigger than green, wich is bigger than blue), but since the correction for spherical aberration

in a refractor does vary over the spectrum of visible light, it is vital to know the exact wavelength

of the interferometrical repoport of arefractor!

Different wavelength:

I know that many of the interferometrical measuring is done with 532nm wavelength or 630nm wavelength.

So if someone states that there is 1/5 spherical aberration, how big is the wavefront error?

Of course it is important to know wich wavelength, and if the error is the survace error or the wavefront error.

Let`s say it is wavefront error. At 532 nm wavelength it is 106.4 nm wavefront error. Surface error is 53.2 nm.

1/5 wavelength wavefront error at 630nm is 126nm wavefront error, and 63nm surface error.

So please let us be very careful with the numbers...

The only major manufacturer that I'm aware of who quotes figure is Orion UK.

Orion UK produce research grade mirrors . They provide a full interferometrical test report with those mirrors!

I know that some of them were re-tested interferometrically and they met the specs again.

Of corse, they cost more than standart mirrors butwith the interferometrically test report

you can be sure to get a high quality mirror.

Lomo in Russia produce high-quality mirrors, and some other manufacturers too.

Those mirrors are no bargain, but the are reliably well made.

Yu can get well-made secondary mirrors too, some of them with interferometrical test report. For example:

http://www.fpi-protostar.com/quartz.htm

You are right that a good planetary newtonian reflector needs 2 good mirrors, but if it has a Strehl of 0.9

it can provide very sharp and contrasty views as long as it is well-collimated, well-cooled,

well baffled or flocked against stray light and when the seeing is well too.

My best planetary views ever were wit a 12" Nasmyth-Cassegrain with 30% central obstruction,

wich I could compare to a 10" f/6 Newt with 16% co and my 8"f/6 Newt with 19% co.

The 10" was second and my 8" only was third.

Aperture is an inportant criterium. With increasing aperture the airy disc size get smaller.

A 8" has an airy disc diameter of half the diameter of a 4" scope.

When you need 150x to see the airy disc of the 4" scope as a tiny disc,

you need 300x to see the airy disc as a mall disc.

This is why a 8" can have sufficiant contrast transfer at a spatial frequency where a 4" cannot transfer contrast anymore.

Soory, my english is not so good, so it is not easy for me to write these complicated things.

I hope some of tha what I wanted to explain is understandable.

If not, or if additional information is needed please ask me.

Regards, Karsten

[Ajohn]

Hi Karsten. Your English is very good. I have to use a web translator to look at German sites.

Not wishing to play ping pong in threads or display greater or lesser knowledge it's difficult to know how far to go with technical topics that will be read by many different people.

On the Rayleigh limit. The 1/4 wave error relates to a wavefront error and is equivalent to a surface error of of 1/8th wave on a mirror and 1/4 wave on a lens. I entirely agree that the location of the error has consequences too. My understanding is that the Rayleigh calculations where done on the basis of slope error. All good astronomical mirrors will (one hopes) be smooth and free from zones as you describe.

The trouble with the rms error specification is that it can mask large errors that will depending on there location have a dramatic effect on the image. The calculation is after all 1/2 the square root of the maximum measured value. You will also note that Orion UKs 1/6 wave peak to valley is also outside of the limit and says nothing about there size location or anything else. I'm not suggesting that they make bad mirrors just pointing out that they can meet that spec and still not meet the Rayleigh limit.

Measurements are generally made in green light. NASA etc though do make infa red mirrors from time to time. A Scottish optician (Waland) received a sever telling off for manufacturing one of these to that would have worked well in normal light. His book outlines the manufacture of a high resolution 61ins F4 mirror. The finished surface is relatively smooth with a few slow deviations of +/- 0.031 waves. That's following machine polishing before hand work. He also describes an F2 mirror for infa red where the green light error is +/0.23. These figures give you an idea what standard professional mirrors are manufactured to. It's an interesting book as it describes just how to go about making very fast mirrors.

On resolution I was explaining why even though the aery disc is the same size on all scopes of the same F ratio larger scopes can have better resolution - they have longer focal lengths so the image size is larger. The resulution limit calculated from the diameter of the scope is usually quoted and was also developed by Rayleigh. (138 arc seconds / diameter in mms ???) This is actually the resolving power when looking at stars. It's detectable. Not exactly a clean split more a double hump in the light curve.

Flats do need to be rather flat for precision work but have the advantage that they are close to the focus. I suspect that many telescopes are made with what is know as polished plate. That means that they tend to be flat but no one actually tests them. Most of the flats available on the atm market where like that and do give decent results.

I'm grateful for your suggestion on applying a central obstruction to an apo to see what it does. I can't wait to try it. Might turn out that I'm believing folklaw too and the true reason lies elsewhere. I have had assurance of a very capable atm (Amateur Telescope Maker) man that low 2ndry mirror obstructions do make a distinct difference to low contrast definition. He would definitely be using smooth 1/10 wave mirrors though. Your way I can see for myself. I'm also glad to see that you use % diameter and not area.

On testing you might like to take a look at oldham optics web site in the UK. He has a point in my opinion. The other way is entirely down to the accuracey of the referance surface.

John

[Macavity]

Also note the comment about SC's with a 13% central obstruction. I would advice working it back from 13% by area to by diameter to get the true figure. Think you will find it's around 30%. May even be more. There is a problem of that nature on most compound scopes and if the percentage area of the obstruction is given it's often based on area which is to say the least misleading.

In fairness, the original article DID quote the 13% with reference to the primary's diameter - Unfortunately I omitted that in my quote - Albeit sensing this to be the default method of specifying such things. [Oops!] To be strictly correct it was indeed (as I correctly stated) a Maksutov-Newtonian, and the "13%" does look impressively small on the picture in the article. A 7" F8 Intes Micro MN78 apparently...