Simple telescope design question

On Mon, 28 Jun 2004 20:07:17 -0700, Robert Maxwell Robinson
wrote:


Thanks, James! This puts the nail in the coffin.

To sum up the responses I've received, a pre-primary flat surface
isn't a good replacement for a small flat secondary. The flat would
need to be bigger than the primary (the problem I already knew about).
It would need to be quite flat over its entire surface, and there is
no known process for grinding something optically flat that approaches
the ease with which a parabolic mirror can be ground; the best idea
involves grinding three blanks against each other, which takes half
again as much work.

All of this leaves the idea of using a curved mirror instead of a flat
mirror, and that puts the question firmly into a different category.
If I keep going on this idea, I am certain I'll end up reinventing the
classical Cassegrain design, or something else that was discarded in
favor of the classical Cassegrain.

There are things called Tilted Component Telescopes (TCT), aka
Trischifspieger, Yolo and so on that have three curved, tilted
surfaces where the abberations produced by the tilts cancel out. A
Google search on some of these terms will come up with some really
strange boxes containing scopes. They avoid the problems with the
spider and central obstruction but at a cost in size and complexity.

Chris

Thanks, James! This puts the nail in the coffin.

To sum up the responses I've received, a pre-primary flat surface
isn't a good replacement for a small flat secondary. The flat would
need to be bigger than the primary (the problem I already knew about).
It would need to be quite flat over its entire surface, and there is
no known process for grinding something optically flat that approaches
the ease with which a parabolic mirror can be ground; the best idea
involves grinding three blanks against each other, which takes half
again as much work.

All of this leaves the idea of using a curved mirror instead of a flat
mirror, and that puts the question firmly into a different category.
If I keep going on this idea, I am certain I'll end up reinventing the
classical Cassegrain design, or something else that was discarded in
favor of the classical Cassegrain.

I was sure there was a good reason I haven't seen that design; turns
out there is.

--Max

On Tue, 29 Jun 2004, James Horn wrote:

|Robert Maxwell Robinson wrote:
|
| One comment I was going to make was that I don't think the large
| "flat" mirror would need to be nearly as flat as the flat secondary of
| a standard Newtonian, since the flat secondary is put at a place where
| the image is already highly magnified.
|
|Actually, it's far worse. The secondary in a Newtonian only needs to be
|accurate over an area as large as a point in the final image appears on
|it. So, for instance, my 2" diagonal on my 10" f/6.5 Dob needs to be 1/10
|wave (or whatever you're going for) accurate over each 1.25" area of its
|surface.
|
|A pre-primary flat has to be that accurate over the *entire* surface - or
|over 10" in my case. And do it after a hole has been put in it, with the
|change in stresses that yields. And do it at the front, unprotected from
|dew and thermal changes.
|
|It does eliminate spider (secondary mount) diffraction though.
|
|Best to you!
|
|Jim Horn
|

Robert Maxwell Robinson wrote:

One comment I was going to make was that I don't think the large
"flat" mirror would need to be nearly as flat as the flat secondary of
a standard Newtonian, since the flat secondary is put at a place where
the image is already highly magnified.

Actually, it's far worse. The secondary in a Newtonian only needs to be
accurate over an area as large as a point in the final image appears on
it. So, for instance, my 2" diagonal on my 10" f/6.5 Dob needs to be 1/10
wave (or whatever you're going for) accurate over each 1.25" area of its
surface.

A pre-primary flat has to be that accurate over the *entire* surface - or
over 10" in my case. And do it after a hole has been put in it, with the
change in stresses that yields. And do it at the front, unprotected from
dew and thermal changes.

It does eliminate spider (secondary mount) diffraction though.

Best to you!

Jim Horn

In message ngton.edu,
Robert Maxwell Robinson writes

Thanks for your help; I'll look at websites on solar scopes. The idea
that large, flat mirrors are harder to make than large parabolic
mirrors sounds *way* strange to me; I thought you practically started
with the one to make the other!

Rough flats are easy - window glass is roughly flat. But making optical
flats is another matter altogether. The classical method requires three
identical pieces and working each one against the others in a systematic
way to get them all exactly flat and polished. Its a lot of work. Most
ATMs buy their flat since they only want one, and it is tedious to do
well.

If you want to look at it another way. Making an optical flat is rather
like trying to make a mirror with a specified radius of curvature (R ~
infinity). ISTR For a lambda/8 flat of diameter D it comes roughly to
R10^7 x D

One comment I was going to make was that I don't think the large
"flat" mirror would need to be nearly as flat as the flat secondary of
a standard Newtonian, since the flat secondary is put at a place where
the image is already highly magnified.

That isn't how it works. The wavefront must be unmolested on the way
into the scope or the resulting image will not be diffraction limited.

At any rate, I'd like a second opinion about how much trouble the flat
mirror would cause in this design.

It can be done, but it would be more expensive than the standard method.

Typically it is used for some professional scopes that are too bulky or
fragile to be pointed at the sky. Try a Google search on siderostat or
heliostat. Some of the optical interferometry systems use flat mirrors
to collect the starlight to feed into the optical bench for combining.

Regards,
--
Martin Brown

In message ngton.edu,
Robert Maxwell Robinson writes

Thanks for your help; I'll look at websites on solar scopes. The idea
that large, flat mirrors are harder to make than large parabolic
mirrors sounds *way* strange to me; I thought you practically started
with the one to make the other!

Rough flats are easy - window glass is roughly flat. But making optical
flats is another matter altogether. The classical method requires three
identical pieces and working each one against the others in a systematic
way to get them all exactly flat and polished. Its a lot of work. Most
ATMs buy their flat since they only want one, and it is tedious to do
well.

If you want to look at it another way. Making an optical flat is rather
like trying to make a mirror with a specified radius of curvature (R ~
infinity). ISTR For a lambda/8 flat of diameter D it comes roughly to
R10^7 x D

One comment I was going to make was that I don't think the large
"flat" mirror would need to be nearly as flat as the flat secondary of
a standard Newtonian, since the flat secondary is put at a place where
the image is already highly magnified.

That isn't how it works. The wavefront must be unmolested on the way
into the scope or the resulting image will not be diffraction limited.

At any rate, I'd like a second opinion about how much trouble the flat
mirror would cause in this design.

It can be done, but it would be more expensive than the standard method.

Typically it is used for some professional scopes that are too bulky or
fragile to be pointed at the sky. Try a Google search on siderostat or
heliostat. Some of the optical interferometry systems use flat mirrors
to collect the starlight to feed into the optical bench for combining.

Regards,
--
Martin Brown

Thanks for your help; I'll look at websites on solar scopes. The idea
that large, flat mirrors are harder to make than large parabolic
mirrors sounds *way* strange to me; I thought you practically started
with the one to make the other!

One comment I was going to make was that I don't think the large
"flat" mirror would need to be nearly as flat as the flat secondary of
a standard Newtonian, since the flat secondary is put at a place where
the image is already highly magnified.

At any rate, I'd like a second opinion about how much trouble the flat
mirror would cause in this design.

On Mon, 28 Jun 2004, Chris Rowland wrote:

|I don't know much about this but as that doesn't stop anyone else;-)
|
Myself included.

|Large flat mirrors are much more difficult to make than parabolic
|mirrors. This is probably the principle objection to this idea.
|
|However scopes are made with this sort of arangement, particularly
|solar scopes.
|
|Another advantage is that the scope can be fixed and only the flat
|miror moved to aim at and track objects.
|
|Chris

I don't know much about this but as that doesn't stop anyone else;-)

Large flat mirrors are much more difficult to make than parabolic
mirrors. This is probably the principle objection to this idea.

However scopes are made with this sort of arangement, particularly
solar scopes.

Another advantage is that the scope can be fixed and only the flat
miror moved to aim at and track objects.

Chris

On Mon, 28 Jun 2004 15:08:37 -0700, Robert Maxwell Robinson
wrote:


Hi, I'm new to the group. I have been learning about telescope
designs for a month or so, and have a question that I haven't been
able to find the answer to; I thought one of you might like to answer
it.

My question is about a variant of a Newtonian reflector. A Newtonian
reflector has a parabolic primary and a flat secondary that is placed
on the optical axis some distance shy of the focal point. The light
reflected off the secondary goes to the eyepiece.

Would it be a Bad Idea to reverse the order of the two mirrors? The
flat elliptical mirror would have to grow to have the same diameter
(along it's _shorter_ axis) as the parabolic mirror, and would be
similar in position to what I think is called a Steering Mirror.
Light would hit the steering mirror, then the parabolic mirror, then
pass through a hole in the steering mirror and go directly into the
eyepiece, like this (only longer):

pppp......................S
ppp ... S
pp ... S
pp ... S
p ...
p.....................|= Eyepiece
p ...
pp ... S
pp ... S
ppp ... S
pppp...........S

I can't believe noone has considered this simple variant on a
Newtonian before; so does anyone know the name of this design? Also
I've never heard of one being constructed, so there must be some
significant problem with it. Can anyone tell me what it is?

The obvious fact of this design that makes it look worse than Isaac
Newton's design is the large, heavy flat mirror instead of a small,
light one. But here are the advantages I see that make me ask:

1. The only real collimation required is collimating the eyepiece to
line up with the optical axis of the parabolic mirror. If the
steering mirror is slightly out of alignment, you see a slightly
different portion of the sky, but nothing goes awry optically. In
a Newtonian, the diagonal has to be correctly aligned to bounce
light directly down the center of the mount for the eyepiece, and
then the eyepiece has to be correctly aligned along that same
axis.

2. Counterweights are often used to balance a Newtonian telescope,
because its weight is predominantly at one end. In this design
there is already weight at both ends, which should minimize the
need additional weights.

3. I believe steering mirrors are often used by owners of large
binoculars to put the eyepieces in a more convenient place, and to
reduce the amount of weight that has to be moved to steer the
field of view. The steering mirror in this design should provide
both of those advantages, but without being an extra optical
element that steals light as it is in other cases.

4. The "obstruction" is a hole rather than the back of a mirror. To
use the obstructed light in a Newtonian reflector, another
diagonal mirror would have to be used to divert the light before
it hits the secondary; and that mirror would grab some of the
light, and have to be aligned with the components that use the
otherwise wasted light. In this case, the light passes through
and can be viewed with an (on-axis!) viewfinder, or for digital
astrophotography it can be focused, collected with a second
CCD, and ultimately added back into the digital image. [Having
two detectors on the same optical axis might allow for some fancy
cross-comparison of off-axis light, for example from two different
optical designs, allowing both to be corrected into a superior
image...but I digress.]

5. The prime focus would be somewhere after the light had passed
through the hole in the steering mirror. This presents a golden
opportunity to use an iris to eliminate the farthest off-axis
light and enhance contrast when viewing the moon or other bright
objects, does it not?

5. A Maksutov or Schmidt corrector could still be used, placed
in the light path before the diagonal mirror and out of the way
of the light reflected from the parabolic mirror. If one is
willing to place the corrector even farther from the flat mirror,
I think a less curved corrector could be used; mightn't that make
them cheaper (at the expense of ending up with an even bulkier,
L-shaped telescope)?

All in all, it sounds like rather a good idea to me. So how wrong am
I?

Thanks,

Max Robinson
Seattle

(This is a rewording of a similar message I posted to alt.astronomy,
before I knew about this newsgroup. Apologies if you've read it
twice now).

Robert Maxwell Robinson wrote:
pppp......................S
ppp ... S
pp ... S
pp ... S
p ...
p.....................|= Eyepiece
p ...
pp ... S
pp ... S
ppp ... S
pppp...........S

Ooh. ASCII art. I like it. (Honestly.)

The obvious fact of this design that makes it look worse than Isaac
Newton's design is the large, heavy flat mirror instead of a small,
light one. But here are the advantages I see that make me ask:

It's not the large that concerns me; it's the flat. Much less expensive
to create a flat secondary a couple of inches across than one that's
40 percent again as large as the primary (and that's assuming you only
want 100 percent illumination at the center of the field of view).
Trouble creating the flat is why most Newtonian manufacturers would rather
use spiders than optical windows--especially in large sizes.

And it's not true that the secondary doesn't steal light. It most
certainly does--but instead of being an obstruction that steals the
light, it has a hole in it that steals the light. The light that would
have bounced off where the hole for the eyepiece goes is missing from
what gets sent on to the primary. That creates essentially the same
light loss and (more importantly) diffraction effects. You *could*, as
you suggest, put a finderscope (my recommendation) or a detector beneath
the hole, so to speak, but the diffraction effects are why you do not
want to just "add the second image" back in to the principal image.

Incidentally, using the finderscope reintroduces the precise alignment
requirement that was taken out of the collimation process. If you
change the orientation of the steering mirror, you will also remove any
alignment you had between where the main telescope and the finderscope
are pointing. (Unless you have some fancy mechanism for halving the
angle of steering.)

As far as the balance issues are concerned, if you were to mount it on
a German equatorial mount, say, the issue is not the weight distribution
along the length of the scope. That is corrected by putting the rings
(or whatever the support mechanism is) further up or down the telescope.
The counterweight is to balance out the scope around the fulcrum that
turns in right ascension--which will still be a problem with this design.

This can be solved for either design by putting the telescope in a
Dobsonian mount and using an equatorial platform.

5. The prime focus would be somewhere after the light had passed
through the hole in the steering mirror. This presents a golden
opportunity to use an iris to eliminate the farthest off-axis
light and enhance contrast when viewing the moon or other bright
objects, does it not?

How is that different from a set of baffles?

5. A Maksutov or Schmidt corrector could still be used, placed
in the light path before the diagonal mirror and out of the way
of the light reflected from the parabolic mirror. If one is
willing to place the corrector even farther from the flat mirror,
I think a less curved corrector could be used; mightn't that make
them cheaper (at the expense of ending up with an even bulkier,
L-shaped telescope)?

Those correctors are there to correct for a spherical, rather than
paraboloidal, primary. I don't see how your design makes manufacture
and alignment of the correctors any easier or less expensive.

I don't mean to be discouraging. You're thinking about these issues,
and ways to solve them, and that's the right thing to do. I hope you'll
continue to try ideas out, and I for one am happy to act as a sounding
board.

Brian Tung
The Astronomy Corner at http://astro.isi.edu/
Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.txt

Robert Maxwell Robinson wrote:
pppp......................S
ppp ... S
pp ... S
pp ... S
p ...
p.....................|= Eyepiece
p ...
pp ... S
pp ... S
ppp ... S
pppp...........S

Ooh. ASCII art. I like it. (Honestly.)

The obvious fact of this design that makes it look worse than Isaac
Newton's design is the large, heavy flat mirror instead of a small,
light one. But here are the advantages I see that make me ask:

It's not the large that concerns me; it's the flat. Much less expensive
to create a flat secondary a couple of inches across than one that's
40 percent again as large as the primary (and that's assuming you only
want 100 percent illumination at the center of the field of view).
Trouble creating the flat is why most Newtonian manufacturers would rather
use spiders than optical windows--especially in large sizes.

And it's not true that the secondary doesn't steal light. It most
certainly does--but instead of being an obstruction that steals the
light, it has a hole in it that steals the light. The light that would
have bounced off where the hole for the eyepiece goes is missing from
what gets sent on to the primary. That creates essentially the same
light loss and (more importantly) diffraction effects. You *could*, as
you suggest, put a finderscope (my recommendation) or a detector beneath
the hole, so to speak, but the diffraction effects are why you do not
want to just "add the second image" back in to the principal image.

Incidentally, using the finderscope reintroduces the precise alignment
requirement that was taken out of the collimation process. If you
change the orientation of the steering mirror, you will also remove any
alignment you had between where the main telescope and the finderscope
are pointing. (Unless you have some fancy mechanism for halving the
angle of steering.)

As far as the balance issues are concerned, if you were to mount it on
a German equatorial mount, say, the issue is not the weight distribution
along the length of the scope. That is corrected by putting the rings
(or whatever the support mechanism is) further up or down the telescope.
The counterweight is to balance out the scope around the fulcrum that
turns in right ascension--which will still be a problem with this design.

This can be solved for either design by putting the telescope in a
Dobsonian mount and using an equatorial platform.

5. The prime focus would be somewhere after the light had passed
through the hole in the steering mirror. This presents a golden
opportunity to use an iris to eliminate the farthest off-axis
light and enhance contrast when viewing the moon or other bright
objects, does it not?

How is that different from a set of baffles?

5. A Maksutov or Schmidt corrector could still be used, placed
in the light path before the diagonal mirror and out of the way
of the light reflected from the parabolic mirror. If one is
willing to place the corrector even farther from the flat mirror,
I think a less curved corrector could be used; mightn't that make
them cheaper (at the expense of ending up with an even bulkier,
L-shaped telescope)?

Those correctors are there to correct for a spherical, rather than
paraboloidal, primary. I don't see how your design makes manufacture
and alignment of the correctors any easier or less expensive.

I don't mean to be discouraging. You're thinking about these issues,
and ways to solve them, and that's the right thing to do. I hope you'll
continue to try ideas out, and I for one am happy to act as a sounding
board.

Brian Tung
The Astronomy Corner at http://astro.isi.edu/
Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.txt

Robert Maxwell Robinson wrote:

Hi, I'm new to the group. I have been learning about telescope
designs for a month or so, and have a question that I haven't been
able to find the answer to; I thought one of you might like to answer
it.

Would it be a Bad Idea to reverse the order of the two mirrors? The
flat elliptical mirror would have to grow to have the same diameter
(along it's _shorter_ axis) as the parabolic mirror, and would be
similar in position to what I think is called a Steering Mirror.
Light would hit the steering mirror, then the parabolic mirror, then
pass through a hole in the steering mirror and go directly into the
eyepiece, like this (only longer):

pppp......................S
ppp ... S
pp ... S
pp ... S
p ...
p.....................|= Eyepiece
p ...
pp ... S
pp ... S
ppp ... S
pppp...........S

I can't believe noone has considered this simple variant on a
Newtonian before; so does anyone know the name of this design? Also
I've never heard of one being constructed, so there must be some
significant problem with it. Can anyone tell me what it is?

You have described (very well) the telescope Charles Fundingsland
invented, built and patented in the 90's, called the "Fundyscope". Mr.
Fundingsland published his 6" aperture design in S&T, but I don't recall
the year and month - maybe someone can look that up. The George B. Wren
II Supernova Search Telescope (SNST) at McDonald Observatory is the
largest Fundyscope in the world, with a Galaxy Optics 18" f/4.5 primary
mirror and 24.25" diameter steering flat made by Mike Marcario at High
Lonesome Optics. I derived the tracking equations and algorithms for
SNST, and Wayne Rosing (also a VP at Google) implemented the tracking
equations and made the thing work, and it worked very well. Bill Wren
used it to discover several supernova. See
http://hej3.as.utexas.edu/~www/SN/.

The Fundyscope steering flat has to be VERY flat to prevent image
astigmatism, on the order of 1/20 wave peak-to-valley, and must be
supported by an edge/back flotation system that can maintain that flatness
over the full angular pointing range. Achieving 1/20 wave P-V precision
on a 24" flat right up to the edge requires a truly skilled optician such
as Mike Marcario. The hole in the flat must also be a tapered 45º cone to
prevent vignetting the field at maximum mirror tilt. Making the steering
flat is the main drawback to Fundyscopes.
Mike