971 replies to this topic

[DAVIDG]

 It's a  rainy Sunday here in Delaware so I got  a couple things done inside the house and have the Eagles game on the radio. So I have some time to play. 

    I pulled out my 8" f/2 spherical mirror to test the Dx-8 corrector  I got at Stellafane this years. The mirror by the way is from a Coulter Lensless Schmidt Camera kit that I also got at Stellafane many years ago.

     So here is a picture of  just the spherical mirror being Ronchi tested. The lines are nice and  straight showing it is nice sphere. 

 

 

   Next  I placed the corrector in front the spherical mirror. The corrector is not going to correct the wavefront so I'm not looking for straight line but I'm looking for a  smooth symmetrical pattern. this would indicate a smooth wavefront that can be corrected by secondary if it's figure and the primary are also smooth. 

 In a Schmidt Newtonian the corrector should fully correct the wavefront and when placed at the correct distance from the primary show straight lines. In a Schmidt Cass.  the power of the corrector determines the wavefront of the combination. So if the corrector under corrects the wavefront the complete scope has the optical characteristic of the  Dall Kirkham (DK) . In a DK it uses a elliptical primary and spherical secondary. It has good on axis performance but a lot of off axis coma. If power the corrector is increased it can fully correct the wavefront the spherical primary so now it acts like a parabola and used in a SCT the secondary is aspheric and coma is reduce to that of what a Classic Cass has with a parabolic primary and hyperbolic secondary. The power of the corrector can being increased more so now the wavefront of it and the spherical primary is overcorrected and it act like a hyperbola . In  a  SCT the secondary is aspherized even more and now the system acts like a Ritchey Chretien (RC) which uses a hyperbolic primary and and hyperbolic secondary. RC are corrected for coma and a SCT can be made to have similar performance. 

   So  here are the results. This looks promising. The pattern is symmetrical and mostly smooth. So there is hope that this corrector  could work. I need to assemble  the telescope with  this corrector and test the complete system with double pass autocollimation to see what the total wavefront looks like.

 

       

 

               - Dave      

Edited by DAVIDG, 12 September 2022 - 08:31 AM.

[mfalls]

. 

   So  here are the results. This looks promising. The pattern is symmetrical and mostly smooth. So there is hope that this corrector  could work. I need to assemble  the telescope with  this corrector and test the complete system with double pass autocollimation to see what the total wavefront looks like.

 

 dx correct stellafane 8 2022.jpg      

 

               - Dave      

So this is the corrector from the 8001. I just purchased an 8001 pro OTA. Perhaps I could make it available to you if that was useful at some point. Rod Mollise indicated in his SCT buying guide there might be a difference between the 8001 and 8001 pro. Anyway, thanks for all you do.

[Gil V]

Your findings are of great interest to me, David. Looking forward to your further evaluation. I have always wondered what the root cause was behind the Dynamax optical deficiency.

I’m leaning towards improperly sourced raw material for the corrector - combined with a just-barely-workable process that failed to remove those issues. We only polished one side, so any inherent issues in one surface would remain in the finished piece. If a poor R/M piece was used as the production blank on the master block, there’s no telling what that would mean for the production run. (Well, there is, and it’s not good).

Our inspection processes were not satisfactory - clearly they were not as good as what we are getting through this forum. We clearly did not detect the defects during assembly. When I built my scope, I hand-selected the best glass I could find - and I assembled hundreds of them, so I looked through A LOT of glass. No wonder I liked mine. I found the outlier!!

I look at it like this. If when I was 19 years old I was as talented as David is today, I could have pointed out the flaws and stopped production. And promptly been fired for refusing to pass finished product that my managers classified as acceptable.

Just my .02. Keep up the good work!

[DAVIDG]

 Gil,

   As I have said many times the purpose of making commercial telescopes is  to make a profit. People seem to miss that in these threads so if the telescope that is produced, sells, that is the quality that you get.  Criterion was trying to compete with Celestron and to do that, they needed to make the Dynamax line cheaper then what Celestron was selling their scopes for . So they had to figure out how to cut production costs. This why they were buying many of the parts from Japan including the secondaries and most likely why they were purchasing the glass for the correctors from a local supplier of window glass.

 

    Also if you haven't made optical surfaces  you  do not appreciate the variability in the process especially when your trying to make 1/4 wave or better optics. Each surface comes out slightly different and for systems that have multiple surfaces, those errors add. So the only way to have the complete system be diffraction limited is to test, figure, test  and figure.   That costs time and that is money. So the idea that you can take optical elements from one telescope and swap them into another, and have it be diffraction limited, is flawed, the odds are low that it  will work.  You will get an image but like I said the odds are low that it will be 1/4 wave.  This why companies like Cumberland  which makes the optics for Questar and Astrophysics have the  reputation they do because they test, correct, test again and correct until they get the true wave front to 1/8 wave or better and that testing and correcting takes time and that is money and you have to pay for that. 

   

It was the mid 1970's to early 1980's so there was no ' Net to easily get information  so a lot of testing techniques required one to dig to find them. They were published in books and Journals, many of which I have but again you had to dig for them and also understand them. 

 

     I get the impression that  the management at Criterion really did not have a  full understanding optical testing.  I believe you tested the system by looking  at an artificial star created by HeNe laser and at  focus of 8" Newtonian and you looked at the in focus image. If so,  that  is part of  the problems, the in focus image does not have the sensitivity and information that the out of focus images do on each side of focus. That is what you do when you correctly do a star test. you look at the defocus images  So if management  understood this the errors would have clearly  been seen. 

 

So even if you did understand how to test  each component correctly which would not have been difficult as my posting show I  wonder if your management would have accepted the findings and try to improve things or not ? Criterion seem to be just barely making it and if any improvement cost money and reduced production numbers they might have had the attitude that what they are making is good enough because the number of telescopes returned for a refund is so low enough not to change anything. 

 

   When I get some time I'll reassembly my scope with this corrector and test the complete system by double pass. Like I just said just because the corrector looks to test better, that doesn't mean in the complete system will be better.  You have to test and correct the errors in the complete system to have the result be a  diffraction limited or better. 

 

                    - Dave 

Edited by DAVIDG, 12 September 2022 - 03:27 PM.

[tim53]

Boy, wouldn't it be nifty if Criterion's (or B&L's) master block still existed somewhere?  ...such that someone could try making a new corrector to old specs with optical glass instead of window glass?

[DAVIDG]

Boy, wouldn't it be nifty if Criterion's (or B&L's) master block still existed somewhere?  ...such that someone could try making a new corrector to old specs with optical glass instead of window glass?

 It does. The gentleman I purchased the corrector from at Stellafane told me he has it.

 

                      - Dave 

[tim53]

That's interesting.  So apparently he doesn't want to part with it.

 

How hard would it be to reverse engineer the figure of the corrector, knowing the f ratio of the primary and measuring the secondary with a spherometer?

 

I'm hoping my complete DX-8 is at least good enough to maybe only need the flat side polished optically flat, though I could try living with it as a DSO instrument only.

 

-Tim.

[davidc135]

 

 

I'm hoping my complete DX-8 is at least good enough to maybe only need the flat side polished optically flat, though I could try living with it as a DSO instrument only.

 

-Tim.

The figure on the 'flat' surface is integral to the performance of the plate and can't be removed

 

.  David

Edited by davidc135, 13 September 2022 - 10:16 AM.

[highfnum]

The dx8 the scope that almost could

 

for commercial scopes  which  is the easiest to produce

and has highest probability to be "good" ?

[DAVIDG]

That's interesting.  So apparently he doesn't want to part with it.

 

How hard would it be to reverse engineer the figure of the corrector, knowing the f ratio of the primary and measuring the secondary with a spherometer?

 

I'm hoping my complete DX-8 is at least good enough to maybe only need the flat side polished optically flat, though I could try living with it as a DSO instrument only.

 

-Tim.

 The math is all in the back of  "Telescope Optics" so any ones  that wants to can  calculate a design starting with  the primary focal length and total focal length of the system.  The formula are for an all spherical system which has coma and systems with no coma with a spherical primary and aspheric secondary, or aspherical primary and spherical secondary or both primary and secondary aspheric. The formulas give the aspheric conic constants and the 4th order terms for the corrector.  I codes these all up in  an Excel spread sheet and then takes these values and put them in OSLO to raytrace the results.  

 

                       - Dave 

[DAVIDG]

The dx8 the scope that almost could

 

for commercial scopes  which  is the easiest to produce

and has highest probability to be "good" ?

 Yes, that is why you see people giving positive comments over many years for 60mm f/15 refractors, 4" f/10 and 6" /8 Newtonians. The Laws of Physics are  work with them and tolerance range is large for producing a 1/4 wave or better images.  

   Criterion understood this with their beloved RV-6. The mirrors in them have a  50" not 48" focal length. Why ?  Because if  you look  in Texereau's book " How to make a Telescope" there is a table of when a mirror can be spherical and just at 1/4 wave and for 6" mirror that is 50". So this was a reference that Criterion had available at the time they were making Newtonians and understood if they got a 6" mirror with a 50" focal length  close to a sphere with just a little correction the odds were very high that it would be at least 1/4 if not better. So they didn't have to do any critical testing to be in that envelope. That saves time and that saves money and they could sell an RV-6 for $199 and make money.  

  You have to keep in mind the business aspects along with  the technical aspect of different telescope models since that explains why they were made the way they are  and why they sell for the prices they do.   The same is true for the Dynamax series and  why Criterion was doing what they did to design and build them.

 

                  - Dave 

[Joe Bergeron]

That Celestron EdgeHD white paper is a masterpiece of marketing, very well written. If I didn’t already have one, I’d want to run out and buy one right now. 
 

It consistently describes the secondary mirror as spherical, and talks about why they decided not to go with a hyperbola or anything other than a sphere. If those mirrors were aspheric, it’s hard to see why Celestron wouldn’t brag about that. 

[DAVIDG]

That Celestron EdgeHD white paper is a masterpiece of marketing, very well written. If I didn’t already have one, I’d want to run out and buy one right now. 
 

It consistently describes the secondary mirror as spherical, and talks about why they decided not to go with a hyperbola or anything other than a sphere. If those mirrors were aspheric, it’s hard to see why Celestron wouldn’t brag about that. 

 The HD Edge most likely does have a spherical secondary because this design is different then the classic C8 The reason is that if you have designed commercial optics you understand that  spherical surfaces are easier to manufacture then  aspheric ones and  so they are cheaper. That is why SCT were be able to be manufactured in the first place because they can be made cheaper  with a spherical primary and Schmidt corrector then the two aspheric surfaces in other cassegrain designs 

   So it is cheaper to use more elements with spherical surfaces then an  aspheric one. Camera lenses are a good example of this.  Celestron is trying to stay in business and to  do so they need to reduce production costs and increase profits.  So they needed to  figure out how to do that. One thing they did was move production out of the US to reduce labor costs. Next they needed to reduce production costs and one of the most labor intensive is hand figuring the secondary .

    So to correct the coma in a Schmidt Cass one can aspherize the secondary or you can add a couple of spherical surfaces lens to do the same  to make coma corrector/field flattener with a spherical secondary. Since the production cost is cheaper you can increase your profit and your marketing people can then hype the fact that you have this  wonderful complex design that has wonderful optical correction. By the way  same concept is used in a Corrected Dall Kirkham Cass which also has a spherical secondary and without the lens corrector has off axis coma

    Like I keep saying you have to look at both the technical and business reasons way  telescopes are made the way they are. 

 

                      - Dave  

Edited by DAVIDG, 15 September 2022 - 09:05 AM.

[Senex Bibax]

Just wondering if any manufacturers have successfully developed and implemented an automated or computerized CNC process for figuring optical components. Maybe the characteristics of the glass and the precise, complex curves involved are not conducive to such an approach.

[CharLakeAstro]

The short answer is yes. The machining of precise, complex curves are but one reason to employ multi-axis CNC machining. Automating the repeat machining on multiple identical parts to the same tolerances is another. The tolerances required are well within the abilities of higher-end CNC machines. 

 

Just wondering if any manufacturers have successfully developed and implemented an automated or computerized CNC process for figuring optical components. Maybe the characteristics of the glass and the precise, complex curves involved are not conducive to such an approach.

[DAVIDG]

The short answer is yes. The machining of precise, complex curves are but one reason to employ multi-axis CNC machining. Automating the repeat machining on multiple identical parts to the same tolerances is another. The tolerances required are well within the abilities of higher-end CNC machines. 

   I agree that high end CNC can make surfaces to very high mechanical  tolerances but they can't  figure  surfaces  to fractions of wave that can be required for high end optics.  You still need to polish them with pitch  and pitch and glass don't behave the same way each and every time. Ask any optician who makes precision optics especially the guys who make optical flats. Once they get those planetary polishing  machines dialed in they never turn them off or it takes days to recondition the laps to get them to make optically flat surfaces again.  

     The problem still is that when you have multiple element system  each surfaces has residue error and these errors add. So if you want a better then diffraction limited ie 1/4 wave you need to test and correct and test and correct. This is one of the reason why designs use as many spherical surfaces as  possible because they can be made to higher tolerances fairly easy. So it is cheaper to use more elements with spherical surfaces then fewer with aspheric surfaces. 

     Another example is the  not so loved Bird Jones reflectors  They use a spherical primary ( cheap to make) and a  two element corrector/barlow (cheap to make) over a  standard Newtonian design with  one parabolic (aspheric ) mirror. More element yes, but spherical so cheaper to make . This is why Meade and Celestron  with these coma free designs are using a group of corrective lenses that have spherical surfaces.  

 

                   - Dave 

[davidc135]

Right- CNC it to a machanical spec, then hand finish the final figure to optical spec.

 

 

Most of the difficulty lies in achieving the final smooth figure so there's little advantage to using cnc.

 

Criterion adopted the same reasonable (on the face of it) manufacturing approach as Celestron. It's well covered earlier in this thread but glass discs would be pulled by vacuum against a plate whose profile is the opposite desired on the finished corrector plate. Any irregularities on this first surface should be projected through the corrector blank and subtracted from the eventual Schmidt profile; which is why it's  impossible to improve a faulty DX corrector.

 

The worked surface would be first ground and then polished flat to give, in theory, a perfect corrector plate. But what were the reasons behind so much going wrong?

 

They could have been a bit more selective in the glass especially in the figure of the contact surface. But the one Dave shows above is not bad.

 

Dust free working conditions. A micron speck of dust between surfaces would be a local 1 wave error.

 

But I think that they always rushed the polishing and the last sensitive part of manufacture which is the main reason why the plates seem uniformly poor.

 

It'd be interesting to see DPAC results of any scts that were much better than 1 wave off. GilV mentions one and it is said that the latest B&L examples such as Mfalls' 8001 pro were better.

Hopefully DaveG's one above will buck the trend but I don't hold out much hope. The irregularities in the Ronchigram stand out against huge amounts of SA in this test. ie the test is insensitive.

 

David

[highfnum]

Bird Jones reflectors

eek!

has anyone ever seen a good example of this type?

 

i have one

its much worse than dx8

[DAVIDG]

" The worked surface would be first ground and then polished flat to give, in theory, a perfect corrector plate. But what were the reasons behind so much going wrong?"

 

    There are two reasons I believe why it didn't work One was the quality of the glass they were using. It seems to be just regular window glass  to start so not very uniform.  They could have possibly  got around that problem if  they simply scanned  both sides of the sheets of the glass they received by interference against an optical flat and used the areas that were optically smooth as the back surface of the corrector  Second their  method differed from Celestrons to try get around Celestron's patent. Celestron used a Master Block with an  opposite Schmidt curve ground into the face. The block was one solid piece of thick glass and the corrector blank  was pulled down against this Master surface by vacuum.

  Criterion's Master surface was a thin plate  with flat back that was  pulled down by vacuum against a flat surface  to hold it in place and then the corrector blank was placed on top of the Master surface and it was also pulled down against the surface with vacuum. So you had  four surfaces being pulled against each other vs two in the Celestron method. So you have many more sources of error in achieving  a smooth Schmidt profile. 

   When I spoke to the  gentleman who purchased what was left of Criterion and was working there when Bausch and Lomb took it over he said he pointed out the problem with their method.  He told me they did not test their method and he showed them the problem by showing them the interference fringe patterns. 

     They were making Newtonians up to this point and like I said the RV-6 used 50" focal length mirror since the tolerance were wide in producing a 1/4 wave image. I have tested a handful of RV-8 mirrors which are 8" f/8 and they were spheres so not at 1/4 wave and there are reports here of their 12" f/7 Newtonian when the primaries were tested  and were very much undercorrected and not at 1/4 wave or better.  Then there is 4" SCT  which are also known to have major optical issues This keeps pointing back to a lack of a critical understanding off  how to make and test optics. Nothing against them but this just seems to be the facts. 

 

                        - Dave 

Edited by DAVIDG, 16 September 2022 - 11:01 AM.

[markb]

As the victim of the cheap glass used to make very rough B&L ex-Criteriion 6000 correctors, it's sad to think how a lovely 6" SCT design, with apparently good primaries, was turned into trash by poor planning and ill-advised sourcing of badly chosen materials, plate glass in this case.

 

While the products had to be able to be manufactured for a specific price point, there was no reason for Criterion and, initially B&L,  not to shop around for a plate glass supplier that could deliver glass with a better surface, and test each batch to find the side to grind (or toss a substandard batch of cheap 1/8 plate).

 

Making products to a price does not mean making expensive but useless products. The auto industry learned a similar lesson in the 80s.

 

I'd love to source a 'tested good' quality NOS corrector for my still new-in-box (plastic carry carry at this juncture) B&L 6000. As far as I can tell, the primary and secondary are well figured and, presumably, well polished.

Edited by markb, 17 September 2022 - 04:07 PM.

[Geo31]

Sad, yes, but for me, more surprising that the issues were either not known or not understood. History is filled with simple mistakes with devastating consequences, especially when trying to survive.  Just as true today and will be true in the future.  We’d like to think not, but it’s true.

 

I’m more surprised (MUCH more) that Bausch & Lomb bought Criterion.  It made NO sense.  Other than a handful of cheap private label scopes sold as Bushnell, B&L was LONG out of the amateur optics business.  I know this very well as I’m from Rochester and got into this hobby in the early 70s.  And now, as someone who spent a long career in managerial accounting, the end of Criterion was a forgone conclusion the minute B&L bought Criterion.  There is no way a mega corporation like B&L could be profitable with a small, lean outfit like Criterion.  The corporate overhead alone would keep Criterion from being profitable.  Most likely, someone at B&L just had to have Criterion, forget the numbers.

 

Lastly, I wonder if Ralph Darin was still at B&L when this went down or if he had retired already?  I know when I got to know him in the mid 70s he was close to retirement.  Of course, I don’t know what area of the company Ralph worked in, but if he was there you’d think they would have tapped into his knowledge.

[Hyokin]

dead topic? any updates dave?