28 replies to this topic

[nexstar11]

There are several great papers detailing the advantages of RC and CDK optical systems.  Here is a great write up that captures differences and contrasts between the 2 systems.

https://www.astrosys...against-cdk.pdf

 

It is clear, that when discussing RC vs CDK or CDK vs RC, it is important to compare 2 corrected systems. It is clear, that a corrected RC outperforms a corrected DK (CDK).

Edited by nexstar11, 05 August 2024 - 03:16 PM.

[bobzeq25]

Depends on the particular RC and CDK. Depends on the usage.

The CDK is vastly easier to collimate.

[Midnight Dan]

Keep in mind that the paper was authored by ASA, who makes RCs and not CDKs.  There might be a teensy bit of bias there. :-)

 

-Dan

Edited by Midnight Dan, 05 August 2024 - 03:24 PM.

[Pickwick]

It makes sense to compare a corrected DK vs a RItchey-Chretien under a cost point of view: the DK is far cheaper than the RC due to spherical secondary (which is very easy to make in contrast to the hyperbolic secondary of an RC) and an elliptical primary (which is also cheaper than an hyperbolic primary of an RC).

 

The reason is that there is a machine able to work a mirror giving it a perfectly spherical shape and spherical shapes are also much easier to test and guarantee proper crafting.

Adding crafting on the spherical mirror will make it elliptical, then parabolic and lastly hyperbolic. Those are obtained starting from a spherical mirror and consuming the mirror to increase its eccentricity.

 

However at the end of the job it is much ore difficult to test the optic plus the added work.

 

Of course you can add a two lenses corrector to correct the coma of a Dall-Kirkham. The Ritchey Chretien is already coma free so the corrector will focus only on the field curvature and the astigmatism. Therefore, using a corrector with the same number of lenses, the RC will have a better correction.

 

You may also increase the number of lenses for the DK corrector and then the DK will be better. But increase the number of the lenses will increase the cost, again...

 

You may check on www.telescope-optics.net

[nexstar11]

Keep in mind that the paper was authored by ASA, who makes RCs and not CDKs.  There might be a teensy bit of bias there. :-)

 

-Dan

Dan, I certainly realize that but based on the spot diagrams and other information they referenced, I don't believe this is biased.  I'm eluding to the same point that ASA mentioned in their paper, that is if we are going to compare the optical systems, we need to compare apples to apples.... i.e.... CDK vs CRC , or DK vs RC, not CDK vs RC.

[freestar8n]

There are several great papers detailing the advantages of RC and CDK optical systems.  Here is a great write up that captures differences and contrasts between the 2 systems.

https://www.astrosys...against-cdk.pdf

 

It is clear, that when discussing RC vs CDK or CDK vs RC, it is important to compare 2 corrected systems. It is clear, that a corrected RC outperforms a corrected DK (CDK).

That write up ignores the thinking at the time (and it still persists) when the CDK first appeared - and that is that the RC is a well corrected and flat field system just with its two mirrors - and that is why they were (but not so much anymore) the design of choice for professional observatories.  People thought the RC *was* already corrected just because of its fancy and expensive curves - so there is no unfairness at all in comparing a *cheaper and better corrected* system with an RC.

 

At that time the RCX/ACF came out saying how it was well corrected "just like an RC is" and has a flat field.  But without the extra corrective lenses near the focal plane - they both suffer from field curvature and astigmatism.

 

As for not being able to use the CDK without those lenses - I don't know anyone who cares about that.  The only real downside of the lenses is the reduction of available spectrum usable for spectroscopy - and *that* is the big motivation for the pure, two-mirror design of the RC professionally.  *Not* how well corrected the field is.

 

In addition, you can similarly get RC's with integrated correctors that are even better corrected because the adjusted the curves on the mirror will only work with the corrective lenses.  So - the same sacrifice made for the CDK can be made for the RC - with an additional gain in performance.

 

In terms of overall field performance you have:

 

DK < RC << CDK < RC with add-on corrector < RC with integrated corrector.

 

But in terms of having a large field that is diffraction limited and flat - the CDK is able to deliver with much simpler curves at a lower cost - and the improvement beyond the CDK is small compared to the big jump from RC to CDK.

 

Frank

Edited by freestar8n, 05 August 2024 - 08:28 PM.

[nexstar11]

Depends on the particular RC and CDK. Depends on the usage.

The CDK is vastly easier to collimate.

I partially agree. I certainly am not referencing something like GSO vs PlaneWave. I think there would be an obvious winner there.  Let's assume this is something like RCOS/Optical Guidance Systems and PlaneWave quality.

[nexstar11]

It makes sense to compare a corrected DK vs a RItchey-Chretien under a cost point of view: the DK is far cheaper than the RC due to spherical secondary (which is very easy to make in contrast to the hyperbolic secondary of an RC) and an elliptical primary (which is also cheaper than an hyperbolic primary of an RC).

 

The reason is that there is a machine able to work a mirror giving it a perfectly spherical shape and spherical shapes are also much easier to test and guarantee proper crafting.

Adding crafting on the spherical mirror will make it elliptical, then parabolic and lastly hyperbolic. Those are obtained starting from a spherical mirror and consuming the mirror to increase its eccentricity.

 

However at the end of the job it is much ore difficult to test the optic plus the added work.

 

Of course you can add a two lenses corrector to correct the coma of a Dall-Kirkham. The Ritchey Chretien is already coma free so the corrector will focus only on the field curvature and the astigmatism. Therefore, using a corrector with the same number of lenses, the RC will have a better correction.

 

You may also increase the number of lenses for the DK corrector and then the DK will be better. But increase the number of the lenses will increase the cost, again...

 

You may check on www.telescope-optics.net

Can you elaborate on what additional advantages would adding more lenses to CDK would produce?

[GlennLeDrew]

I've heard from users who said they were driven to near insanity when collimating an RC, to the point of getting rid of it in exchange for something like an SCT or CDK. The RC's aspheric secondary has a unique axis of symmetry, requiring that the two mirrors' axes be also coincident to within a fairly tight limit. A spherical secondary has no unique axis of symmetry and so can be wildly de-centered, the only downside being non-symmetrical field illumination.

[quilty]

Doesn't this apply to any CC type as well? both mirrors have their axis which should match perfectly at direction and no offset

[Midnight Dan]

I've heard from users who said they were driven to near insanity when collimating an RC, to the point of getting rid of it in exchange for something like an SCT or CDK. The RC's aspheric secondary has a unique axis of symmetry, requiring that the two mirrors' axes be also coincident to within a fairly tight limit. A spherical secondary has no unique axis of symmetry and so can be wildly de-centered, the only downside being non-symmetrical field illumination.

In most cases, that's due to a design flaw, not an inherent RC problem.  Most of the smaller, solid-tube RCs have the focuser mount attached directly to the primary mirror, with a single set of collimation screws which moves both simultaneously.  This makes collimation an iterative, frustrating process.

 

The larger RCs are designed with the primary mirror and focuser separated, each having their own set of collimation screws.  This makes collimation, with the right tools, a straightforward, single-pass process that's easy to do.

 

-Dan

[bobzeq25]

Doesn't this apply to any CC type as well? both mirrors have their axis which should match perfectly at direction and no offset

 
 

In most cases, that's due to a design flaw, not an inherent RC problem.  Most of the smaller, solid-tube RCs have the focuser mount attached directly to the primary mirror, with a single set of collimation screws which moves both simultaneously.  This makes collimation an iterative, frustrating process.
 
The larger RCs are designed with the primary mirror and focuser separated, each having their own set of collimation screws.  This makes collimation, with the right tools, a straightforward, single-pass process that's easy to do.
 
-Dan

This is just not true.

The problem is that, when collimating, the hyperbolic mirrors interact. This ONLY applies to scopes with TWO hyberbolic mirrors.

You CANNOT collimate an RC in a single pass when both mirrors are out of collimation. You have to go back and forth, sneaking up on collimation.

It drives people crazy. It's a major reason why, as stated in post #9 "I've heard from users who said they were driven to near insanity when collimating an RC, to the point of getting rid of it in exchange for something like an SCT or CDK."

It is true that coupling the focuser to the primary makes things worse.

BTDTGTTS. When I got my 6RC collimated, after many hours, I discovered that focuser sag threw the scope out of collimation with changes of altitude. Some people fix that with a Moonlight, which costs as much as a 6RC.

I bought a nice 3.7 R&P focuser. It came attached to a TS 130mm F7. Problem solved. <smile>

Edited by bobzeq25, 08 August 2024 - 06:14 PM.

[quilty]

This is just not true.

The problem is that, when collimating, the hyperbolic mirrors interact. This ONLY applies to scopes with TWO hyberbolic mirrors.

Y

I think in a CC both mirrors interact as well. What's the difference?

[Rasfahan]

I think in a CC both mirrors interact as well. What's the difference?

Tolerances in an RC are much tighter for collimation.

[quilty]

nevertheless the mirrors interact (what else should they do) and both mirrors have their axis which must be aligned
And surely you can't collimate a CC in one step when both mirrors are off.
You can do neither with an SC when the sec. mirror is offset.

[freestar8n]

I think the main thing at play here that makes collimation easier is if the primary is fixed and has no adjustments.  I think the cdk is collimated just using the 3 screws on the secondary - and that means fewer degrees of freedom to get lost in.

 

If the primary is also adjustable it becomes much more complicated.  And when people add sensor tilt adjustment on top of that - combined with software tools that don't measure "tilt" properly - it's no wonder people end up in the weeds.  But it's not so much to do with the curves on the mirrors...

 

Frank

[Matthew Paul]

 
 

The problem is that, when collimating, the hyperbolic mirrors interact. This ONLY applies to scopes with TWO hyberbolic mirrors.

This is simply not true. 

 

In ANY compound optical system the two surfaces [or more] interact. This is the very trait that makes them a compound optic.  

Edited by Matthew Paul, 08 August 2024 - 08:15 PM.

[Midnight Dan]

 
 This is just not true.

The problem is that, when collimating, the hyperbolic mirrors interact. This ONLY applies to scopes with TWO hyberbolic mirrors.

You CANNOT collimate an RC in a single pass when both mirrors are out of collimation. You have to go back and forth, sneaking up on collimation.
 

That hasn't been my experience.

 

I use the Glatter laser collimator.  I start by putting it into the focuser and using the focuser collimation screws to aim it at the center spot on the secondary mirror.  Then I adjust the secondary collimation screws to bounce the laser directly back at the source.  And finally, I use the primary collimation screws to center the projected concentric-circle pattern around the secondary.

 

Once I've done this, I've gone back for a second pass to check it but never had to make an additional adjustment.

 

-Dan

[bobzeq25]

That hasn't been my experience.
 
I use the Glatter laser collimator.  I start by putting it into the focuser and using the focuser collimation screws to aim it at the center spot on the secondary mirror.  Then I adjust the secondary collimation screws to bounce the laser directly back at the source.  And finally, I use the primary collimation screws to center the projected concentric-circle pattern around the secondary.
 
Once I've done this, I've gone back for a second pass to check it but never had to make an additional adjustment.
 
-Dan

OK, our experiences vary. Here are some other people who found it necessary to go back and forth.

https://www.cloudyni...is-easy-or-not/

https://astrophotoni...collimation.pdf

These people, who manufactured RCs (with decooupled focusers), have published a method many find to be the gold standard. It is iterative.

https://www.deepskyi...ure_Ver_1.0.pdf

I could list more examples forever. Forever.

A reason for different experiences. Here are some factors that could make a one step process acceptable.

The scope started out not far from collimation. Mine was delivered in what I considered to be far from collimation.

The imagers standards are lower.

My guess is that your scope was delivered not far from collimation. Many are not so lucky.

They post threads like "borderline ready to sell this piece of XXXX" which recounted a long struggle. They sold the scope.

Others are more lucky (or have low standards). Some post that they did not have to touch collimation with a new RC.

It's a crapshoot. I find it hard to believe that your method eliminates the issue, it's BASIC to telescopes with two hyperbolic mirrors. Professional observatories are well equipped with optical experts who can manage the process. Their very expensive RCs need collimation infrequently. Most imagers are not optical experts. The relatively inexpensive quality does not compare.

The CDK has a spherical primary. The issue does not arise. Just as a SCT does not have the issue. Spherical primary.

Whatever you want to say about RCs, their reputation as being difficult to collimate is not because people just don't know the magic method.

It's because two hyperbolic mirrors are hard to collimate.

Other point. Illustrative but not relevant to almost everyone here purchasing a scope.

It is possible to make a good quality RC that holds collimation well, because the mechanical stability is excellent (the RC is peculiarly sensitive to mechanical quality) and truly decouples the focuser. Here's one from CFF. The gold thing at the front is the focuser, it shoves the secondary up at back.

They sell very few. Few want to pay for a quality RC. It's so much easier (and cheaper?) to just get a CDK instead.

Edited by bobzeq25, 08 August 2024 - 11:48 PM.

[freestar8n]

The CDK has a spherical primary. The issue does not arise. Just as a SCT does not have the issue. Spherical primary.

Whatever you want to say about RCs, their reputation as being difficult to collimate is not because people just don't know the magic method.

It's because two hyperbolic mirrors are hard to collimate.

The CDK is a modified Dall-Kirkham, so its primary is a pretty strong ellipsoid and just as much an asphere as the primary of an RC.

 

But the CDK, like current SCT's, has a non-adjustable primary - so there is no back and forth at all...  And that makes for much easier collimation - even though the primary is an asphere.

 

The ACF has an ellipsoidal secondary.  Is it much harder to collimate than an SCT with a spherical secondary?

 

What makes collimation much easier is only having one adjustable mirror.  Whether that mirror is asphere or not.

 

Frank

[bobzeq25]

The CDK is a modified Dall-Kirkham, so its primary is a pretty strong ellipsoid and just as much an asphere as the primary of an RC.
 
But the CDK, like current SCT's, has a non-adjustable primary - so there is no back and forth at all...  And that makes for much easier collimation - even though the primary is an asphere.
 
The ACF has an ellipsoidal secondary.  Is it much harder to collimate than an SCT with a spherical secondary?
 
What makes collimation much easier is only having one adjustable mirror.  Whether that mirror is asphere or not.
 
Frank

OK, I misspoke. Doesn't change the main point. The REASON Planewave CDKs don't have a collimation adjustment for the primary is that the optical design means you can collimate the secondary independently. Like an SCT.

This problem is a function of the dual hyperbolic mirrors. I know of no other telescope with them.

Edited by bobzeq25, 08 August 2024 - 11:55 PM.

[quilty]

I think the main thing at play here that makes collimation easier is if the primary is fixed and has no adjustments.  I think the cdk is collimated just using the 3 screws on the secondary - and that means fewer degrees of freedom to get lost in.
 
If the primary is also adjustable it becomes much more complicated.  And when people add sensor tilt adjustment on top of that - combined with software tools that don't measure "tilt" properly - it's no wonder people end up in the weeds.  But it's not so much to do with the curves on the mirrors...
 
Frank

This is exactly what I think, too. the more elements to adjust the more complicated.
Adjusting both mirrors is like (not quite) an equation with thwo unknown variables which can be solved by iteration.
Which makes it much more complicated than a normal SC (when there's no sec. mirror offset). Any additional element as focusser and senor tilt would exceed my capabilities.
Fortunately spidered sec. mirrors use to be centered and the primary mirror can be adjusted indepently along the spider vanes.
(in order to adjust a Mak's primary mirror a centered crosshair is most helpful)

To make it short, an RC wouldn't scare me off as long as it's just about both mirrors, neither should it do to anyone, not much of a difference to my CC.

When you had to guess, how many CCs are out there compared to RC numbers?

Edited by quilty, 09 August 2024 - 02:15 AM.

[Rasfahan]

The problem is not only the two adjustable mirrors but also the tighter tolerances. Even a slight decentering of the mirror axes badly compromises field performance. Compared to an SCT or Mak the necessary adjustments are tiny and the effort involved is huge. I speak from experience with the kind of CFF RC bobzeq showed above. The hugely better mechanics than my previous GSO enable me to actually collimate it correctly. If it‘s badly miscollimated, I still need about an hour to get it good. Visual collimation aids and a laser are only the first step, they‘re really too coarse for good results.

I think that most CCs are used visually. In that case miscollimation first affects off-axis performance while on-axis is still seeing-dominated (unless seeing is truly excellent). As they‘re mainly used for high magnifications, this decreased performance isn‘t visible to the observer. RCs are mostly used photographically, so compromised field correction is easily visible.

[Midnight Dan]

OK, our experiences vary. Here are some other people who found it necessary to go back and forth.

https://www.cloudyni...is-easy-or-not/

https://astrophotoni...collimation.pdf

These people, who manufactured RCs (with decooupled focusers), have published a method many find to be the gold standard. It is iterative.

https://www.deepskyi...ure_Ver_1.0.pdf

I could list more examples forever. Forever.
 

 

Your first link is from a person who has a 10" solid tube, which does not have separate collimation adjustments for the primary and focuser.

 

The second and third links are methods involving visual inspection of images, not using collimation tools.  My statement about single pass collimation included the caveat, "with the right tools".  I have no experience with the methods in these links, but they are significantly different from what I use, and may be why they need iteration.

 

 

 

 My guess is that your scope was delivered not far from collimation. Many are not so lucky.

 

Not so.  It needed some significant adjustment when I got it.  Also, after a year one of the fans went bad so I had to disassemble the rear of the scope to replace them.  This, of course, threw the collimation out again.  And again it was a single pass to achieve collimation.

 

I understand the issue that the mechanical center and optical center of a mirror may not be coincident.  But I suspect in most RC's that difference is small and inconsequential.  I also own the SkyWave software that gives you accurate collimation without relying on a mechanical center spot.  After using my laser collimation tools, I found that SkyWave could not improve my collimation, at least within the limits that my seeing imposed.

 

-Dan

Edited by Midnight Dan, 09 August 2024 - 09:12 AM.

[jhwilmes]

The CDK is a modified Dall-Kirkham, so its primary is a pretty strong ellipsoid and just as much an asphere as the primary of an RC.

 

But the CDK, like current SCT's, has a non-adjustable primary - so there is no back and forth at all...  And that makes for much easier collimation - even though the primary is an asphere.

 

The ACF has an ellipsoidal secondary.  Is it much harder to collimate than an SCT with a spherical secondary?

 

What makes collimation much easier is only having one adjustable mirror.  Whether that mirror is asphere or not.

 

Frank

I wonder if the ellipsoidal secondary would be harder to collimate... I do not have one myself, but I have seen threads comparing the Meade ACF to the standard Celestron SCTs, and the commentary seems to be the ACF is harder to collimate.

 

OK, I misspoke. Doesn't change the main point. The REASON Planewave CDKs don't have a collimation adjustment for the primary is that the optical design means you can collimate the secondary independently. Like an SCT.

This problem is a function of the dual hyperbolic mirrors. I know of no other telescope with them.

Agreed.  Spherical secondaries are significantly easier to collimate - only the position of the secondary needs to be changed, all angles of rotation are valid.  Primaries for telescopes with spherical secondaries should have similar difficulty to collimate as for a Newtonian, if they are collimatable.  I suspect classical cassegrains (CCs) would be about as difficult to collimate as RCs, but in practice they operate at much higher focal ratios, and the mirror curves are less extreme, which makes collimation slightly easier.