Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Homemade 0,65m f 1.5

  • Please log in to reply
68 replies to this topic

#51 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 01 October 2019 - 10:08 PM

Vla, But would not the all spherical surfaces be Much Easier for an Amateur to do???

 

Keep the wedge under control, and then hit the radi to the specs...maybe a few test plates to make

for those surfaces that cannot be tested otherwise.

Nothing's easy about building f/1.5 system like this, even only half as big. Tolerances are nightmarish. In addition, the speed sets short limit to exposure (before sky background becomes to bright), the brighter sky, the more so. Borisov also mentions that problem. Makes more sense to relax the speed somewhat, because there's more to gain than to lose. Here's an example, 400mm f/2.34, for comparison.

Attached Thumbnails

  • 400.png

  • PrestonE, clivemilne and 555aaa like this

#52 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 01 October 2019 - 10:33 PM

There are several reasons to compare primarily in RMS spot sizes:

1. It's correlation with encircled or ensquared energy is pretty good for geometrically-limited systems, like discussed ones on wide field.

2. It's usually the primary design criterion for entire development cycle from global search and glass selection to tolerancing and it's very quick and useful for estimation of performance gains and harms at every design step, just by current MF value.

3. FWHM is really useful only for nearly pure gaussian PSF, like seeing-dominated or perfectly Gauss-apodized extreme cases. In the case of non-gaussian shape like non-apodized diffraction PSF, especially aberrated, or complex geometrically-limited spot it's an approximation on the brink of vulgarization.

 

I personally like Zemax universal plots like DENC vs. field, that strictly meet design targets like "80% energy in 2x2px square", but more and more often use rough estimates at design search state, like "system with less than 1px RMS spot radius over the field is good for surveillance work", especially with clear trend to extreme speeds and great practical error margins.

 

Usually it's quite difficult to make frank assessment of real quality of fast astrograph just because of coarse sampling with operational sensor, and in most cases there is no collimation/attestation fine sensor/scanner in the observatory. So overall design cycle and particulary speed/FOV improvements are made in careful steps based on prototypes with conservative parameters, because collimation tolerances dominate over nominal residual aberrations.

We should differentiate practical from accurate. RMS spot size is better practical criterion than Airy disc, but lacks accuracy for the final assessment. For the same RMS error, the RMS spot size for primary astigmatism is three times smaller than for primary spherical (more than that for balanced secondary), and more than two times for primary coma, while its 80% energy square is only two times smaller for the spherical, and 50% smaller for coma. Those are not small discrepancies. That's why we have that the system with a singlet corrector has nearly twice worse RMS spot size, but nearly as good contrast transfer as the 3-element corrector system.

 

As for the actual FWHM, that is the actual energy distribution, and no geometrical substitute can be as good. Both, long and short-exposure PSF do fit under Gaussian envelope. The aberrated FWHM, before seeing - which are Gaussian here - indicates the after-seeing FWHM.

 

The tolerances is a subject that hasn't been addressed, except very generally. It's a very important, but missing piece of information in this case.


  • PrestonE likes this

#53 An18

An18

    Lift Off

  • -----
  • Posts: 20
  • Joined: 25 Sep 2019

Posted 02 October 2019 - 03:54 AM

Keep the wedge under control, and then hit the radi to the specs...maybe a few test plates to make

for those surfaces that cannot be tested otherwise.

 

The ASC 330 that Jones designed for me is also All Spherical, but less lenses I think

and no Margin Primary...

 

Richter-Slevogt may be more practical for amateur without test plates - with front lens bound to second and primary mirror, and field lens can be made quite rough with simple spherometer. Here is example of 250/400 system for FLI ML16803.

Attached Thumbnails

  • 250-400.jpg
  • 250-400-16803.JPG

  • 555aaa likes this

#54 Mike I. Jones

Mike I. Jones

    Skylab

  • *****
  • Posts: 4,442
  • Joined: 02 Jul 2006
  • Loc: Fort Worth TX

Posted 02 October 2019 - 07:58 AM

Y'all know if all the optical designs ever posted on Cloudynights, complete with prescriptions, were compiled into a book, it would rival any other similar book in existence, right?


  • PrestonE, LarsMalmgren, Oregon-raybender and 4 others like this

#55 MKV

MKV

    Cosmos

  • *****
  • Posts: 8,246
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 02 October 2019 - 10:57 AM

And these are the spots, with 18µm pixels overlaid, and the MTF out to 30 cy/mm.  The Nyquist cutoff for 18µm pixels is 1/(2*0.018)=27.8 cy/mm.

 

attachicon.gif 300mm f1 J-band astrograph spots.jpg

 

attachicon.gif 300mm f1 J-band astrograph MTF.jpg

Defocus tolerance (inside the tube!) is ≤ 0.02 mm (0.0008 in) ! How realistic is this?



#56 An18

An18

    Lift Off

  • -----
  • Posts: 20
  • Joined: 25 Sep 2019

Posted 03 October 2019 - 08:06 AM

Defocus tolerance (inside the tube!) is ≤ 0.02 mm (0.0008 in) ! How realistic is this?

It's very easy for focusers like FLI Atlas. In practice, I usually take final focusing iterations with F/1.5 systems on 9 um pixels arrays with 100 steps of Atlas, e.g. 10 um shift. And there is a lot of precision for finer focusing for systems even faster than F/1.

 

Now Santel produces 400/500 astrograph for KAF 16803 and 50100 chips, next version will be probably 500 F/1 for 61x61 10um px. Extreme version with D400 F300 also covering GSENSE 6060 is developed.

The first Santel 400 F/1.25 is on the picture, focusing and collimation stability weren't problems at all on initial star tests.

Attached Thumbnails

  • image004.jpg

Edited by An18, 03 October 2019 - 08:07 AM.

  • PrestonE and Dave O like this

#57 MKV

MKV

    Cosmos

  • *****
  • Posts: 8,246
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 03 October 2019 - 11:49 AM

Thanks, but how is this accomplished under actual nighttime observing conditions? Temperature and air density variations alone can produce larger shifts in focus,  making it impossible for the image to stay in sharply focused without adaptive optics. This doesn't even take into account guiding errors, vibrations, light scatter, and so on, all of which contribute in one way or another to blur enlargement above and beyond the Airy disc size (for an f/1.5 a 0.25 λ defocus is only 2.5 μm). On the other hand, if these scopes are used only for imaging then a need for such fine focusing iterations may be meaningless.

 

Mladen



#58 An18

An18

    Lift Off

  • -----
  • Posts: 20
  • Joined: 25 Sep 2019

Posted 04 October 2019 - 05:00 AM

For systems with same samplings of Airy disc wavefront tolerances are also same, regardless of F-ratio. For pixel-limited like discussed ones, Airy disc if by far undersampled. Actually, the much larger geometric PSF is frequently undersampled too. So, in practice even complex catadioptrics with Cassegrain focus, like Shenker 200/300 "Genon", manufactured by Gennady in past years, don't need refocusing in the night, BK7/steel pair works great.

 

Sonnefield system shown in post #56 uses titanium, for even greater focus stability.


  • PrestonE and Gleb1964 like this

#59 An18

An18

    Lift Off

  • -----
  • Posts: 20
  • Joined: 25 Sep 2019

Posted 05 October 2019 - 02:08 PM

That's why we have that the system with a singlet corrector has nearly twice worse RMS spot size, but nearly as good contrast transfer as the 3-element corrector system.

That's simply not true and it has nothing in common with RMS/PTV ratios of different abberations/orders.

We have 103/52/39 MTFA50 frequencies for 65-triplet system and only 53/34/28 for singlet for the same wavelength set and fields. This is in a good accordance with RMS spot size differences - both near twice at the center, 1.53/1.77 on 1.8 degrees and 1.39/1.59 at 2.5 degrees from frame center. RMS Spot MF ratio is 1.63x and it indicates that better system can be used with pixels about 1.6x smaller, which is near to MTF-based values.

 

The real advantage of the triplet version is in much better correction of lateral color on large field, triplet with glass selection gives us degrees of freedom to correct spherochromatizm simultaneously with lateral color. On narrow field singlet can achieve near the same quality.

Attached Thumbnails

  • 650-lc.JPG


#60 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 05 October 2019 - 06:27 PM

Well, if it's "not true" there's something wrong with Zemax and OSLO, since they are showing it is true. And if you are saying it's not, don't you feel you should support it with specific design data and MTF graphs?

 

Btw. I wasn't talking about RMS-to-PV ratio, but about RMS spot size ratios for any given RMS wavefront error of different aberrations. Maybe you should pay more attention to what is actually written down?



#61 An18

An18

    Lift Off

  • -----
  • Posts: 20
  • Joined: 25 Sep 2019

Posted 06 October 2019 - 07:36 AM

MTF graphs are below, numbers above are calculated from MTFA MF operands.

 

You wern't clear enough, saying that "RMS spot size is better practical criterion than Airy disc", as Airy disc size can only be used a PTV spot criterion and not any WF one, while not useful when near.

 

Systems of discussed parameters just aren't designed with RMS WF because they're far from DL performance - we deal with near-zero Strehls for the most of the practical FOV (for 650 F1.5 this is at least 61x61mm, using 650 mm barrel of glass for ridiculously small chips like 37x37 may occur only due to poverty, like in Russian case).

Attached Thumbnails

  • 650-sing-mtf.JPG
  • 650-tripl-mtf.JPG

Edited by An18, 06 October 2019 - 07:57 AM.


#62 Vla

Vla

    Viking 1

  • *****
  • Posts: 611
  • Joined: 08 Mar 2014

Posted 06 October 2019 - 01:25 PM

Those are obviously not the systems in question, i.e. Mike Jones' (triplet corrector) and mine (singlet corrector). It is only clear that it is not clear which plot refers to which system. If we look at the average (T+S)/2 transfer, the bottom systems is better in field performance, and certainly on axis, but its wide S-T spread indicates significant asymmetry in energy distribution (not good). Where do we go from here? Btw. I was never talking neither in terms of P-V nor RMS WFE, my focus was - as it should be - energy distribution, i.e. ensquared energy.

 

Below is polychromatic MTF for the singlet 650mm f/1.5 system I posted last, for the same field and spectra range.

Attached Thumbnails

  • smtf.png

Edited by Vla, 06 October 2019 - 02:26 PM.


#63 Oregon-raybender

Oregon-raybender

    Optical Research Engineer

  • *****
  • Vendors
  • Posts: 633
  • Joined: 13 May 2010
  • Loc: Oregon, South Western Coast

Posted 06 October 2019 - 02:04 PM

I like Mike Jones suggestion on optical designs, Perhaps a listing on CN is in order?

I am game to list mine, if you have Zemax, it's under telescope reference designs.

 

Starry Nightswaytogo.gif


  • PrestonE and LarsMalmgren like this

#64 Lucullus

Lucullus

    Viking 1

  • -----
  • Posts: 645
  • Joined: 14 Feb 2012

Posted 16 October 2019 - 05:18 AM

Short meta-topic question: are you all professional opticians? What does it take for a non-optician amateur astronomer to contribute meaningfully to such a discussion?

#65 coinboy1

coinboy1

    Apollo

  • *****
  • Posts: 1,303
  • Joined: 03 May 2011
  • Loc: Tulsa, OK

Posted 31 October 2019 - 11:40 AM

I like the idea of having a resource and compilation of all optical prescriptions of various designs posted on cloudy nights. It would be of immense value to many people! 


  • PrestonE likes this

#66 PrestonE

PrestonE

    Vanguard

  • *****
  • Posts: 2,195
  • Joined: 29 Apr 2005
  • Loc: San Miguel de Allende,Mexico

Posted 31 October 2019 - 02:32 PM

Ok, Who is going to set up this Wonderful Resource???   I think it needs to be an Optics Guy, thus not me...

 

Perhaps the Moderator will set up the link Pinned to the Top of the ATM page and we can copy and paste

the links in???

 

Best Regards,

 

Preston


Edited by PrestonE, 31 October 2019 - 06:10 PM.


#67 Benach

Benach

    Vanguard

  • *****
  • Posts: 2,125
  • Joined: 24 Jan 2008

Posted 31 October 2019 - 04:35 PM

I drop Mike's 250mm Super-Houghton in if requested.


  • clivemilne likes this

#68 Oregon-raybender

Oregon-raybender

    Optical Research Engineer

  • *****
  • Vendors
  • Posts: 633
  • Joined: 13 May 2010
  • Loc: Oregon, South Western Coast

Posted 31 October 2019 - 06:31 PM

I'll send mine.

 

Starry Nightswaytogo.gif



#69 Benach

Benach

    Vanguard

  • *****
  • Posts: 2,125
  • Joined: 24 Jan 2008

Posted 01 November 2019 - 01:52 AM

Short meta-topic question: are you all professional opticians? What does it take for a non-optician amateur astronomer to contribute meaningfully to such a discussion?

A few of us have experience in professional experience but certainly not everybody. To name a few from whom I know they have it or strongly suspect they have it: Mike I. Jones, Ed Jones, Oregon-raybender, Vla, Mark Cowan, An18, me. Someone else like MKV has read quite a lot in the field of optics and optical testing but I happen to know that he has no professional experience.

 

How you can contribute, simple: read some advanced books about optics, do not hesitate to ask questions and certainly do not hesitate to make more advanced designs if you feel you like them more. The problem is that a lot of ATMs build "only" Dobson/Newtonian telescopes. They make nice looking ones and since so many examples exist online, it is easy to find examples, spreadsheets, books, CAD-files etc. of these telescopes. There is a myriad of examples you can take your inspiration from. If you have difficulty judging and evaluating the designs above, read books like Telescope Optics, evaluation and design as a starter. For me that book was a great stepup to professional optics and it was also written with that intention.

 

Last of all: Do not hesitate to ask questions. Let me share a personal anecdote: When I was ATMing as a high school student about 20 years ago, I learned how to grind and polish optics. I learned it from a guy but I noticed that he circumvented the more difficult questions. At the same time, I was in a e-mail newsgroup with amateur astronomers and Harrie Rutten (the author of Telescope Optics) was in there as well. He regularly contributed to the newsgroup. I met him, give or take a few days, exactly 21 years ago we got along very well. I never hesitated to ask him questions up to the point that some members in the newsgroup became annoyed at me that I asked so many questions. :) But boy, I learned a lot from him.

If you feel ashamed of asking questions, you shouldn't in the first place, but you can also send a PM to the writer of a post. If that doesn't work: feel free to PM me. Am often willing to answer questions.


  • PrestonE and clivemilne like this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics