316 replies to this topic

[MKV]

It was touched on in post 76 and a matching test plate and interference might be the easiest method if one was available. But is anyone actually going to make one just to test an already existing secondary? However, if the 2ndry needs to be made then the making of the test plate at the same time is not too much more trouble and can be used, as Tim53 is doing.

There are many professional optical engineers and opticians here who could have made these test plates (Mike I. Jones, Glenn Drew and Ed Jones, Mark Harry, etc. -- using work facilities or their own tools at home), not to mention excellent ATM opticians like Dick Parker, DavidG, and others, but  none has volunteered so far as I know, even though there is enough interest in this subject (just judging from the length of this thread alone) to settle this conundrum once and for all. 

 

In fact, a group contract could have ordered a hundred or so such test plates for a nominal price per person and had them made by a professional lens maker company (Newport, Thorlabs, Melles Griot, or a few (one or two) by local labs, such as U. of Arizona for example). They could be distributed to various ATM clubs for that purpose since there are always C8s around.

 

Short of any empirical results, everything is a speculation.

[davidc135]

Yes, it's not a big deal to make the test plate. It doesn't even have to be fully polished. But no one has because every ATM has had more interesting and worthwhile ways tp spend their time. And we don't need to acquire hundreds, just one, assuming the 2ndry RoC is sufficiently constant. An approx f2 2.5'' diameter plate would be good for C8s. Anyone fancy a go?

 

Maybe not essential but the interference test would work better if the mirror coating is stripped, which is an extra hassle.

 

Believing that everything is a speculation is a very disproportionate view.

 

David

[MKV]

Believing that everything is a speculation is a very disproportionate view.

Interesting terminology in this issue, but I agree-there's a disproportionate amount theorizing compared to empirical results; an abundance of roundabout approaches - my Norwegian ex used to call it "going around a bend a bit", or just beating around the bush in good ole' American parlance - rather than the well tried and proven empirical method routinely used by professionals, and also within reach of ATMs.

 

Speculation is a good initial, step, but empirical proof is what nails it. If a spherical test plate shows an optical null, the secondary is a sphere. End of story. 

[GlennLeDrew]

My optics fabrication was restricted to the shop at work, until a decade ago. If I had the stuff required to do so at home I very well might have made up a test plate...

 

I wonder if this would provide the requisite sensitivity for a test. Utilize the fact that an aspheric surface delivers an image upon reflection that suffers astigmatism. The reflection of a point source from a spherical surface, as long as the reflection is perpendicular, will be absent astigmatism. Replacing the reflector with one having a non spherical surface at the point of reflection will suffer astigmatism. The question is, what optical means will be sufficient to discriminate the aberration?

 

If a magnifying instrument is to examine the reflected image, a means to deliver the image on the optical axis of the test system is required, such as via a beam splitter. Or, if the test apparatus is at a suitable distance, the point source could be located just outside the optical path, obviating the need for a beam splitter.

 

Just an idea for a potential approach that might be cooked up with available components.

[MKV]

I wonder if this would provide the requisite sensitivity for a test. Utilize the fact that an aspheric surface delivers an image upon reflection that suffers astigmatism.

Not if you use a collimating lens. That should be easy and cheap for small optics in this case. A bulls-eye irregularity should be able to resolve aspheric departure.

 

I am sure there must be a surplus C8 secondary somewhere to be stripped and tested. A test plate can be made then to its exact ROC and tested. Of course, it would have to be a genuine C8 secondary, not some mystery convex mirror. 

 

Is it not possible to buy a C8 secondary from Celestron?

[davidc135]

My optics fabrication was restricted to the shop at work, until a decade ago. If I had the stuff required to do so at home I very well might have made up a test plate...

 

I wonder if this would provide the requisite sensitivity for a test. Utilize the fact that an aspheric surface delivers an image upon reflection that suffers astigmatism. The reflection of a point source from a spherical surface, as long as the reflection is perpendicular, will be absent astigmatism. Replacing the reflector with one having a non spherical surface at the point of reflection will suffer astigmatism. The question is, what optical means will be sufficient to discriminate the aberration?

 

If a magnifying instrument is to examine the reflected image, a means to deliver the image on the optical axis of the test system is required, such as via a beam splitter. Or, if the test apparatus is at a suitable distance, the point source could be located just outside the optical path, obviating the need for a beam splitter.

 

Just an idea for a potential approach that might be cooked up with available components.

Would the below diagram work? If the high quality DX finder objective was stopped down by half, its under-correction would be cut to one sixteenth whilst any asphericity on the 2ndry stays the same, or nearly.

 

 

At (correction) f34 the reflected image of the 50micron pinhole should show astig if it's there. Care needs to be taken to make sure that the 2ndry is aligned.

 

David

Edited by davidc135, 21 September 2023 - 04:49 PM.

[Matthew Paul]

Not if you use a collimating lens. That should be easy and cheap for small optics in this case. A bulls-eye irregularity should be able to resolve aspheric departure.

 

I am sure there must be a surplus C8 secondary somewhere to be stripped and tested. A test plate can be made then to its exact ROC and tested. Of course, it would have to be a genuine C8 secondary, not some mystery convex mirror. 

 

Is it not possible to buy a C8 secondary from Celestron?

What era Celestron C8 secondary are you looking for?

[davidc135]

I added a crude green filter to the pinhole above in post 131 and took Ronchis inside and outside of focus. The objective was masked down to 1 inch and the Celestron secondary moved horizontally with repect to the objective by 1/2 inch. 

 

The grating was orientated at 45^o to the axis of astigmatism in case that was needed as well as vertically. Vertical Ronchigrams were also taken with the secondary and objective centered, for comparison.

 

The green film used as a filter is distorting the optics. I can redo them without the filter if anyone wants.

 

 

 

This simple test was a great idea, Glen.

 

Astigmatism should show up as clocking between the inside and outside images. There's nothing much significant that I can see, whether the components were centered or off centre.

Once more, tests are very strongly saying that the Celestron secondary is essentially spherical.

 

David

 

The outside focus looks a touch cw compared to the inside, inline with a tiny asphericity?

Edited by davidc135, 21 September 2023 - 06:20 PM.

[davidc135]

Before taking the Ronchigrams I looked through a 7.5mm ep at 115x. Because of the false colour I tried the filter and got a circular image of the pinhole surrounded by scattered light due, I think, to the green film. It seemed the same on either side of focus but I wasn't able to film it and so resorted to a grating.

 

David

[GlennLeDrew]

David,

I was just spouting off a notion that had just occurred to me, and had not applied any deep thought to the matter.

 

A Ronchi might prove to be insufficiently sensitive. An examination of the reflected image through a close focus telescope (due to the nearby virtual image located not very far beyond the mirror's surface) might prove to be more definitive. An exit pupil small enough to present diffraction might enable to directly see the ellipticity of the astigmatic image. A known aspheric surface, such as an SCT corrector, could furnish something of a baseline. Different regions on the corrector having differing rates of slope could provide an indication of a threshold sensitivity. Provided this idea is actually viable!  😉

 

I don't have the bits on hand to try this out myself.  😞

Edited by GlennLeDrew, 21 September 2023 - 08:43 PM.

[starspangled]

FWIW  in  Robert  Peikal's  'Making Schmidt Cassegrain Optics'  , in the design chapter he talks about the  `compact variant'   "which is the basic system used in commercial  SCT' s"  where the secondary surface is more or less aligned with the corrector plate , and using two spherical mirrors . ( P39 ) .

 

That would lend weight to the idea that the very short figuring runs on the secondaries at Celestron were more about system compensating for minor aspheric or zonal errors on the mirrors .

Edited by starspangled, 21 September 2023 - 11:36 PM.

[MKV]

Would the below diagram work? If the high quality DX finder objective was stopped down by half, its under-correction would be cut to one sixteenth whilst any asphericity on the 2ndry stays the same, or nearly.

 

At (correction) f34 the reflected image of the 50micron pinhole should show astig if it's there. Care needs to be taken to make sure that the 2ndry is aligned.

If it would work in theory can be determined only by raytracing. This would give you wavefront residuals as well tolerances, includes alignment tolerances. 

 

In practice, the whole test setup would have to be calibrated to account for its errors first.

 

The light source would have to be monochromatic.

 

A 50 micron pinhole will not produce a flawless spherical wave signal. You'll need either a single mode optic fiber with or without a GRIN lens, or a good spatial filter.

 

Components should be held and mounted in precision optomechanical mounts.

[davidc135]

David,

I was just spouting off a notion that had just occurred to me, and had not applied any deep thought to the matter.

 

A Ronchi might prove to be insufficiently sensitive. An examination of the reflected image through a close focus telescope (due to the nearby virtual image located not very far beyond the mirror's surface) might prove to be more definitive. An exit pupil small enough to present diffraction might enable to directly see the ellipticity of the astigmatic image. A known aspheric surface, such as an SCT corrector, could furnish something of a baseline. Different regions on the corrector having differing rates of slope could provide an indication of a threshold sensitivity. Provided this idea is actually viable! 

 

I don't have the bits on hand to try this out myself. 

I think your idea is very good and at f34 the Ronchi test should be well able to detect any significant astigmatism. Telescope-Optics.net illustrates the increasing sensitivity to astigmatism and visibility of clocking with f ratio at bottom page 4.8.3.

The axis of the objective was centered on the 0.5r (correction) zone of the secondary in my test. It would be good to compare with a known figure such as a sct corrector neutral zone area tested at a comparable f ratio. Maybe I have some stuff.

 

David

 

PS Correction above, .75r zone of secondary changed to 0.5r.

Edited by davidc135, 22 September 2023 - 04:34 AM.

[MKV]

What era Celestron C8 secondary are you looking for?

Aren't all C8 secondary mirrors the same?

[MKV]

Would the below diagram work? If the high quality DX finder objective was stopped down by half, its under-correction would be cut to one sixteenth whilst any asphericity on the 2ndry stays the same, or nearly.

 

At (correction) f34 the reflected image of the 50micron pinhole should show astig if it's there. Care needs to be taken to make sure that the 2ndry is aligned.

What is the focal length of your DX finder, and the ROC of the convex secondary mirror?

How far from the mirror was DX lens? 

What was the skewed angle of the secondary?

What was diameter of the precision pinhole?

What was light source? 

What was the frequency of Ronchi screen?

[davidc135]

What is the focal length of your DX finder, and the ROC of the convex secondary mirror?

How far from the mirror was DX lens? 

What was the skewed angle of the secondary?

What was diameter of the precision pinhole?

What was light source? 

What was the frequency of Ronchi screen?

I think most of the answers are in the thread but:

DX finder fl is 8.75''

3mm gap

ROC is 245mm

The 2ndry wasn't skewed

Pinhole is 50microns

80 lppi

 

Your emphasis on forensic precision, ray tracing, grin lenses etc seems to be divorced from commonsense and what's needed to answer this thread's question.

 

Which, as I've pointed out a number of times, is whether the secondary is sufficiently aspheric to be part of a coma free compact f2/10 sct.

 

The answer is of the order of 2.75 fringes on the surface.  I get this from the following.

 

The spherical 8'' f2 primary suffers 22.55 waves SA. Do you need ray tracing to tell you that?

 

Telescope Optics (R&vV)in their design give the 2ndry's conic as -0.88 and the corrector g as 0.834 IIRC. If you accept that and the figure of 0.712 as the rquired g in an all spherical design as given by Vla Sacek, then you get adifference in corrective power between the two designs as:

 

(0.834-0.712)x22.55 = 2.75 waves

 

Similarly, working from the 2ndry cc of -0.88 of 2'' f2.5 its difference from a spherical optic is:

 

(2X22.55)/2.5³ = 2.88 waves, out by a seventh or eighth.

 

Why do you need ray tracing? All this talk just camouflages very simple things.

 

Were my Ronchigrams, measurements of knife edge travel and the astigmatism test all unable to pick up at least 2.5 waves of asphericity?

As I've also said the tests could be polished up in a number of ways but there wouldn't be enough difference to affect the outcome.

 

Isn't it obvious that this secondary hasn't a snowball's chance in hell of being part of a coma free telescope.

 

David

Edited by davidc135, 22 September 2023 - 04:20 AM.

[MKV]

Actually a 1971 link to Celestron catalog confirms that the secondary is definitely NOT spherical but figured aspheric. Thus, unless someone can prove/show that current C8's have simple spherical secondary mirrors, this debate should be finished. 

 

 

[MKV]

 

 

The spherical 8'' f2 primary suffers 22.55 waves SA. Do you need ray tracing to tell you that?

Of course not. 3rd order min. rms OPD in waves = D⁴/(256λR³), where D is the aperture diameter, R is the ROC and λ is the wavelength the same units as D and R. That comes to about 232 waves. 

 

Telescope Optics (R&vV)in their design give the 2ndry's conic as -0.88 and the corrector g as 0.834 IIRC. If you accept that and the figure of 0.712 as the rquired g in an all spherical design as given by Vla Sacek, then you get a difference in corrective power between the two designs as:

 

(0.834-0.712)x22.55 = 2.75 waves

If I remove the secondary mirror from a C8, the residual wavefront error at the primary focus is actually 5 waves, i.e. 10 fringes, not 2.75.

 

[KBHornblower]

Actually a 1971 link to Celestron catalog confirms that the secondary is definitely NOT spherical but figured aspheric. Thus, unless someone can prove/show that current C8's have simple spherical secondary mirrors, this debate should be finished. 

 

C8 Testing and Guarantee_1.jpg

Those words from Celestron could apply to either of two extremes, or something in between.

 

1. Secondary spherical by design, minor aspherizing to correct for errors in the other two elements.

 

2.  Secondary hyperboloidal by design along with different Schmidt plate, for correcting coma. In that case "aspherizing" would not be an appropriate word for the touchup figuring, as the surface was already aspherical by design.

 

If Celestron was already designing coma out of the system in 1971, why was it such a big deal when Meade started the ACF some decades later?

 

This thread remains inconclusive.  It is so long with so many twists and turns that I cannot remember who said what about what, or when they said it.

[KBHornblower]

Let me add that the description of the star test looks like nothing more than checking for spherical aberration or zones on axis, with nothing about coma.

[davidc135]

 

 

If I remove the secondary mirror from a C8, the residual wavefront error at the primary focus is actually 5 waves, i.e. 10 fringes, not 2.75.

 

 

I didn't say if the secondary was removed then the residual error at focus would be 2.75 waves.

 

I said the difference in correcting power between a corrector designed for an aplanat and one designed for an all spherical sct was 2.75 waves. (ie 0.122g).

Therefore it is also the difference in correcting power between the corresponding secondaries and so is what the test is searching for but is failing to find.

 

David

[MKV]

Those words from Celestron could apply to either of two extremes, or something in between.

Sounds like reading too much into it. One way or another, retouching the seconday means it's aspherized. Celestron never mentions erradication of coma. I don't think the C8 was ever intended to be an aplanat.

 

If Celestron was already designing coma out of the system in 1971, why was it such a big deal when Meade started the ACF some decades later?

I think that's an excellent point -- and only reinforces the impression I stated above, namely that full coma correction was never a stated goal by the Celelstron people until just a few years ago.

 

This thread remains inconclusive.  It is so long with so many twists and turns that I cannot remember who said what about what, or when they said it.

Agree 100%. 

[Matthew Paul]

Aren't all C8 secondary mirrors the same?

I don't know the answer to that question.

 

There were wishes of having a C8 secondary and talk of different eras of telescopes. I might be able to supply a stripped C8 secondary mirror if someone was serious about making a test plate and learning the shape first hand. This is why I asked if there is a specific era of secondary that was desired. 

[DAVIDG]

Sounds like reading too much into it. One way or another, retouching the seconday means it's aspherized. Celestron never mentions erradication of coma. I don't think the C8 was ever intended to be an aplanat.

 

I think that's an excellent point -- and only reinforces the impression I stated above, namely that full coma correction was never a stated goal by the Celelstron people until just a few years ago.

 

Agree 100%. 

 If  you read  Piekels books  on what was Celestron was all about he mentioned that what Celestron was doing was making the power of the corrector great then 1 This is from interviewing Bob Johnson . What this does is make the wave form from the combination of the spherical primary and the corrector be over corrected  like that  of a hyperbola   So when you aspherize the secondary you have a system that is acting like a Ritchey Chretien and as better coma correction for astrphotography which was the main marketing reason  for selling these scopes just like it is today.  Also one  needs to understand that from business stand point your not going to tell the world exactly what your doing so other can take advantage of it. 

   From  a business stand point the skilled labor to aspherized the secondaries was one of the most expensive part of the process. So  in recent times to reduce costs they redesigned the system to use spherical surfaces which now includes a lens group as coma corrector. If you understand the optical industry the "holy grail" is spherical surfaces because they are cheaper to make. So it is actually cheaper to make a  system with more surfaces  but having them spherical then  fewer with aspheric system. This is why both Celestron and Meade were coming out with these newer system.  They had better coma correction then the first version but were actually cheaper to manufacture. If you look at the Meade coma free system which they claimed was an RC type and got sued for and lost, it has no aspheric surface but a number of spherical ones to get around the issues of the manufacturing costs. 

   So back to the original design,  Bob Johnson was a smart guy and his goal was to sell an excellent astrophotography system that had good correction both in field flatness and coma correction that fit on standard 35mm film frame. So by over correcting the corrector and aspherizing the secondary he was able to achieve that. If you don't have a aspheric surface somewhere in a classic SCT system it has poor coma correction and that is just an optical fact.

    To complete with Celestron, Meade had to figure out away to make their scope have the same or better performance but at a  cheap cost. So they looked at the process and saw that hiring skilled people to hand figure the secondaries was the most costly part of the process. To over come this they had the secondaries made out the US and had them aspherized. Knowing that the process is variable in the aspherizing process they understood that each secondary would be slightly different in it's figure. So what they did when assembling the scopes was to swap the secondaries in an out to the test jig to try to find one that meet their spec.  It didn't require a skilled optician, you just had to train someone to look at  an image the scope produced and learn when it meet their spec. If they went with a spherical secondary Celestron would soon have pointed out the poor off axis correction.  The opposite is also true, if Celestron had used a system with poor coma Meade would have pointed that out right from the start and their system was better  They both changed their design when labor costs in the US started to rise since they were cheap to make and even cheaper when you make the optics were the labor is cheaper. 

 

  The bottom line is you  have to look at the design from both an optical stand point and also a business stand point to understand the why's and what's of why they were being done.  People just are making assumption based on what they believe the optical design is but missing the main reason, MONEY.

 

                   - Dave 

[davidc135]

In post 133 Ronchigrams were taken of the wavefront reflected off the off centre half of the C8 secondary to see if they could pick up the astigmatism expected from an aspheric surface. There was only a tiny amount clocking between the inside and outside of focus images.

 

To check whether this could be due to insensitivity I repeated the experiment using small parts of the C8 corrector plate but this time using transmission rather than reflection.

 

As I maintain that a secondary figured to correct for coma should be under-corrected by approx 2.75 waves wf I tried to find a part of the plate that would be equivalent.

As the original test was carried out at f34 I chose my 75mm f15 achromat stopped down to 33mm (so also at f34)  to deliver the focused rays to the grating and camera.

The precision pinhole was situated at the focus of a Meade 8'' sct the collimated rays from which were received by the achromat with its mask.

Between the Meade and the achromat the rays passed through a section of the C8 corrector which could be moved sideways to present a section either near the edge or near the centre.

 

As spherical correction is proportional to the 4th power of the aperture radius I imagined that that each annulus of corrector could deliver correction equal to that of its OD minus that of its ID. And that a 33mm circular aperture would be a small part of that annulus with the same asphericity.

It happens that the innermost annulus just outside the baffle with an ID  (that should be IR) of 33 and an OR of 66mm has an ashericity of 2.8 waves.

 

I take the full sized corrector (with r = 1) to correct for 16 waves, assuming a spherical 2ndry:

 

(0.66)⁴-(0.33)⁴x16 = 2.84 waves

 

whilst the outermost annulus has around 13 waves.

 

The first two images were taken near the baffle with the other pair closer to the edge.

 

 

 

Inside of focus is cw.

 

Whilst the optics here are more complicated and I could well have misunderstood it, this test seems equivalent to the one using reflections off the C8 secondary and which showed so little.

Here the clocking between inside and outside is obvious and lends yet further support to the 2ndry being a touched up sphere.

 

If these Ronchigrams clearly show astigmatism due to off centre aspheric surfaces, then the C8 secondary, whose off centre surface showed virtually no astigmatism, must surely be essentially spherical.

 

That's me about done with tests. All of them have firmly said sphere to me but if anyone can take on the test plate job that certainly would be the clincher.

 

David

Edited by davidc135, 23 September 2023 - 03:07 PM.