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

Ronchi Rulings

  • Please log in to reply
39 replies to this topic

#1 macleod

macleod

    Ranger 4

  • -----
  • topic starter
  • Posts: 325
  • Joined: 21 Sep 2014
  • Loc: Napier, New Zealand, 4110

Posted 18 July 2019 - 04:53 PM

In a topic discussing  "turned down edge ? " there is a lot of excellent discussion on the Ronchi test. I,m attempting a very fast mirror, 8in f2.6  and am using a 133LPI grating . However I am aware of seeing diffraction effects at the SIDES of the image . I read the edge off the center of the image principally , for tde. So should I be using a 50LPI grating instead ? Will a 50LPI grating  still be ok for figuring at f2.6 ? Thank you , Macleod.



#2 BGRE

BGRE

    Vanguard

  • *****
  • Posts: 2239
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 18 July 2019 - 05:41 PM

A sinusoidal phase grating eliminates all diffraction orders other than 0,+1, -1.



#3 Steve Dodds

Steve Dodds

    Owner - Nova Optical

  • *****
  • Vendors
  • Posts: 233
  • Joined: 02 Apr 2007
  • Loc: Utah

Posted 18 July 2019 - 06:09 PM

At f/2.6 diffraction is going to be minimal, but once you get closer to a parabola there is going to be too much spherical aberration to use a 133, there will just be too many lines.  


  • sopticals likes this

#4 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22712
  • Joined: 22 Feb 2010

Posted 18 July 2019 - 06:31 PM

The lower the frequency the lower the sensitivity to correction.

 

you could get a 8-10” flat from an old comparator and test it in the scope as dpac.

 

get the lines straight in dpac and you got a great scope.

 

i did (2) of my 8” f 3.5 with a Ross lens for nulling and it turned out great

 

i got a nice 100x400 pcx, it’s the easiest null by far. Great for  figuring a mirror. 


Edited by Pinbout, 18 July 2019 - 06:32 PM.

  • Venetia2004 likes this

#5 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 18 July 2019 - 07:00 PM

mcleod, an f/2.6 sphere will not suffer too much from diffraction effects but it will require to be 98.5% or better parabola and that may be too didifuclt to see even with 133 lpi Ronchi screen.

 

8in_f2_6_133LPI.jpg

 

Autocollimation is the way to go since it doubles the error and makes it easier to see. A Ross null test is easier to interpret than standard ROC test but it's not twice as sensitive. 

 

 



#6 macleod

macleod

    Ranger 4

  • -----
  • topic starter
  • Posts: 325
  • Joined: 21 Sep 2014
  • Loc: Napier, New Zealand, 4110

Posted 18 July 2019 - 07:22 PM

Thank you all . I doubt I,ll ever get to 98.5% correction ! That is perfection , but for wide field DSO viewing could less be acceptable ? Macleod



#7 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22712
  • Joined: 22 Feb 2010

Posted 18 July 2019 - 07:30 PM

Don’t sell yourself short.



#8 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22712
  • Joined: 22 Feb 2010

Posted 18 July 2019 - 07:33 PM

https://rover.ebay.c...tm/254189392639


  • PrestonE likes this

#9 macleod

macleod

    Ranger 4

  • -----
  • topic starter
  • Posts: 325
  • Joined: 21 Sep 2014
  • Loc: Napier, New Zealand, 4110

Posted 18 July 2019 - 08:09 PM

Danny thank you so much for all your help!  Unfortunately being retired I cannot justify nearly NZ$ 300 for that mirror if I was still working I wouldn,t hesitate. I,m doing this mirror making to keep " my brain active " , besides loving doing this type of precision work - a bit like the Dentistry I did for 60yrs !  best wishes, Macleod.



#10 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 19 July 2019 - 10:13 AM

Thank you all . I doubt I,ll ever get to 98.5% correction ! That is perfection , but for wide field DSO viewing could less be acceptable ? Macleod

You can still try. What do you have to lose? A mirror like yours, f/2.6, that'S 98.5% paraboloid is a good, but not perfect mirror that should give you 1/7 wave pvw on the bench. You can go as low as low 97.5% for a 1/4 wave residual for the mirror itself (not counting wavefront-degrading effects of misalignment, atmosphere, eyepieces, central obstruction, air currents, etc.) if that makes a difference. All I was saying is that it's difficult to discern the difference using the Ronchi test at that focal ratio, iow that the Ronchi test will not let you easily achieve even the minimum requirements for the mirror. 

 

8in_f2_6_a.jpg

 

But a Foucault test should. It will provide deep, high contrast shadows, and plenty of knife travel distance. If it's a smooth curve, test the 90% and the 70.71% zones, and estimate the surface using the good old r2/2R rule (for a moving source and k-e), where r is the zonal radius multiplied by the conic constant. 



#11 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8889
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 19 July 2019 - 10:56 AM

."All I was saying is that it's difficult to discern the difference using the Ronchi test at that focal ratio, iow that the Ronchi test will not let you easily achieve even the minimum requirements for the mirror."

    This is why I would recommend that the OTA be made and then the mirror can be tested via double pass autocollimation using an oil flat and an ultra bright LED. Now you have 2x the sensitivity as the Ronchi by itself and a null test were dead straight Ronchi bands indicate an excellent figure. It is easier to judge the straightness or  should I say  lack of  straightness in the bands in DPAC test vs trying to match a Ronchi pattern when testing at the ROC.

 

                   - Dave 


  • PrestonE and coinboy1 like this

#12 gr5org

gr5org

    Vostok 1

  • -----
  • Posts: 105
  • Joined: 06 Sep 2018

Posted 19 July 2019 - 11:52 AM

A Bath Interferometer can test an 8 inch down to about F/2.2.  An F/2.6 is much easier.  You should be able to build a Bath for less than the cost of a Ronchi screen - how much did you pay for the Ronchi screen?  In general though F/2.6 is getting tricky no matter what method you use.

 

If you don't want to pay $300NZ for a flat then maybe you should stick to foucault and do lots of zones?  Do you have a foucault setup with a micrometer to move along the optical axis?  I'm not sure what the best answer is.  I'm very biased to Bath Interferometers.  Cheap, simple, and amazingly accurate.  If you have a friend with a 3d printer the remaining optical parts are < $60 USD (not sure about shipping to NZ).



#13 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3826
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 19 July 2019 - 12:10 PM

The reflectivity of an uncoated Pyrex mirror autocollimated on an oil flat is 0.0049% and indeed needs a very bright source.  Another easier and inexpensive test is the forgotten immersion null where the mirror is tested in a pan of water.  Testing vertical is practical for fast mirrors like this.  A monochromatic light source must be separated from the KE and a single mode fiber is perfect.  The KE is placed  near the ROC which is 42 inches here.  The reflectivity is 51 times brighter at 0.25%.  The spacings are ray traced easily and are not very sensitive.


Edited by Ed Jones, 19 July 2019 - 12:11 PM.

  • coinboy1 likes this

#14 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 19 July 2019 - 03:06 PM

."All I was saying is that it's difficult to discern the difference using the Ronchi test at that focal ratio, iow that the Ronchi test will not let you easily achieve even the minimum requirements for the mirror."

    This is why I would recommend that the OTA be made and then the mirror can be tested via double pass autocollimation using an oil flat and an ultra bright LED. Now you have 2x the sensitivity as the Ronchi by itself and a null test were dead straight Ronchi bands indicate an excellent figure. It is easier to judge the straightness or  should I say  lack of  straightness in the bands in DPAC test vs trying to match a Ronchi pattern when testing at the ROC.

Yep, DPAC with a knife-edge will go a long way in correcting this fast mirror. Don't forget, the wavefront values are doubled!.

 

This clearly shows that an 8-inch f/2.6 even at 98.5% correction is only going to be about 1/6 wave pvw, even under ideal conditions  This may help some understand the great demand such surfaces pose for a top-notch mirror.  It also shows the incredible efficiency of the DPAC test.

 

Oil flats make testing larger mirrors possible, and are economical. However, despite discussions on CN and other portals, no definitive solution has been found (as far as I know) how to prevent thin and large mirrors from sagging if they're tested in a vertical position.

 

If the Ronchi test were used, it would fare worse. At 98.5% correction a 133 lpi screen would show barely noticeable curvature with 2 bands/pupil.  So, again, a Ronchi test will not get you much better than 1/5 wave pvw.

 

8in_f2_6_c.jpg


Edited by MKV, 19 July 2019 - 05:59 PM.

  • coinboy1 likes this

#15 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 19 July 2019 - 03:15 PM

The reflectivity of an uncoated Pyrex mirror autocollimated on an oil flat is 0.0049% and indeed needs a very bright source.  Another easier and inexpensive test is the forgotten immersion null where the mirror is tested in a pan of water.  Testing vertical is practical for fast mirrors like this.  A monochromatic light source must be separated from the KE and a single mode fiber is perfect.  The KE is placed  near the ROC which is 42 inches here.  The reflectivity is 51 times brighter at 0.25%.  The spacings are ray traced easily and are not very sensitive.

Yes, the immersion null is another method. It's pretty labor-intensive, however, as everything has to be on the floor, and you have to spoon small amounts of liquid in and out. I tried it and had problems with vibrations, too much near-by traffic makes it almost impossible even at night. 


Edited by MKV, 19 July 2019 - 03:15 PM.


#16 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 19 July 2019 - 03:30 PM

A Bath Interferometer can test an 8 inch down to about F/2.2.  An F/2.6 is much easier.  You should be able to build a Bath for less than the cost of a Ronchi screen - how much did you pay for the Ronchi screen?  In general though F/2.6 is getting tricky no matter what method you use.

 

If you don't want to pay $300NZ for a flat then maybe you should stick to foucault and do lots of zones?  Do you have a foucault setup with a micrometer to move along the optical axis?  I'm not sure what the best answer is.  I'm very biased to Bath Interferometers.  Cheap, simple, and amazingly accurate.  If you have a friend with a 3d printer the remaining optical parts are < $60 USD (not sure about shipping to NZ).

Well, interferometry will always beat other tests hands down because of the wealth of information no other test provides. I originally included simulated interferograms but decided not to use them, since it's unlikely the OP will go for anything like a Bath, given his earlier description of what he wants from the mirror (basically a low power wide field sky sweeper). But since you mentioned, I am including inerferograms. One thing is clear: DPAC with k-e is every bit as sensitive as an IF.

 

However, given the separate beams architecture of the Bath I always found it difficult to get satisfactory results with very fast mirrors or lenses. If you have your f/2.2 Bath igrams, it would be nice to see them. 

 

Cheers!

 

8in_f2_6_d.jpg


Edited by MKV, 19 July 2019 - 05:56 PM.


#17 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3826
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 19 July 2019 - 04:11 PM

 

It's pretty labor-intensive

It's simpler than autocollimation to align (only one mirror to align), with no secondary it has almost no obstruction using a fiber, the mirror is face up and under less stress, and it's brighter.  Water thickness is not sensitive, a 1/8 inch error is insignificant.  Yes it can be sensitive to vibration but no more so than in my water flat video .  I parabolized my first mirror a 12 inch F/4 (living right on a highway) and I highly recommend it when you have the vertical space.  Very low cost, all you need is a fiber optic and a pan of water.


  • coinboy1 likes this

#18 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 19 July 2019 - 04:23 PM

It's simpler than autocollimation to align (only one mirror to align), with no secondary it has almost no obstruction using a fiber, the mirror is face up and under less stress, and it's brighter.  Water thickness is not sensitive, a 1/8 inch error is insignificant.  Yes it can be sensitive to vibration but no more so than in my water flat video .  I parabolized my first mirror a 12 inch F/4 (living right on a highway) and I highly recommend it when you have the vertical space.  Very low cost, all you need is a fiber optic and a pan of water.

I guess like so many things I'll have to try it one of these days again. But an almost finished spherical wave IF (see avatar) is likely to divert me in the other direction for a while.


Edited by MKV, 19 July 2019 - 05:54 PM.


#19 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22712
  • Joined: 22 Feb 2010

Posted 19 July 2019 - 04:54 PM

Danny thank you so much for all your help!  Unfortunately being retired I cannot justify nearly NZ$ 300 for that mirror if I was still working I wouldn,t hesitate. I,m doing this mirror making to keep " my brain active " , besides loving doing this type of precision work - a bit like the Dentistry I did for 60yrs !  best wishes, Macleod.

What DaveG said - an oil flat, get it close as possible then build your ota, so you know it comes to focus - then test it dpac with oil in a pan - used motor oil from changing the oil in your car- and adjust correction accordingly.  You don’t need a thick layer of oil either. 


Edited by Pinbout, 19 July 2019 - 05:00 PM.


#20 BGRE

BGRE

    Vanguard

  • *****
  • Posts: 2239
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 19 July 2019 - 05:32 PM

To avoid the significant curvature near the edge of a liquid flat, it should be somewhat greater in diameter than the test surface (about 75mm/3inches is recommended with a water flat for example). Averaging of the  results obtained with many short exposures taken with an interferometer can sometimes be useful in minimising the effects of vibration provided that the frame rate isn't synchronous with the vibration.

 

With Ronchi test using a long exposure (up to several minutes or so) can also be very effective.

If the source is too bright and cannot be dimmed sufficiently just average a large number of exposures (do not attempt to align the fringes in the exposures before averaging).



#21 ckh

ckh

    Surveyor 1

  • -----
  • Posts: 1916
  • Joined: 21 Feb 2015
  • Loc: Arizona

Posted 19 July 2019 - 05:56 PM

Ed,

 

I never heard of the immersion null test. Sounds intriguing. The papers I've found are behind a paywall. Any source of info?

 

Thanks,

Carl



#22 MKV

MKV

    Cosmos

  • *****
  • Posts: 7960
  • Joined: 20 Jan 2011
  • Loc: Florida, USA

Posted 19 July 2019 - 06:01 PM

Ed,

I never heard of the immersion null test. Sounds intriguing. The papers I've found are behind a paywall. Any source of info?

Carl, watch Ed's Youtube video on water test. I believe he may have a video on his site (Optical Ed) as well.

 

Mladen


Edited by MKV, 19 July 2019 - 06:02 PM.


#23 macleod

macleod

    Ranger 4

  • -----
  • topic starter
  • Posts: 325
  • Joined: 21 Sep 2014
  • Loc: Napier, New Zealand, 4110

Posted 19 July 2019 - 06:07 PM

WOW !! Thank you everyone for your contributions . What have I got into ? An 84yr old retired dentist even reading this stuff is  nightmare!  Seriously I love it, but only marginally understand some of it - certainly not DPAC etc. Although I,m keen to learn, that is why I,m doing this project in the first place.

So I,ll go as far with Ronchi as I can and then move to one of the better double pass tests for final accuracy . So which one to go for, remembering I,m going to learn from GO ! I can build the tester in my off time. Ross Null - a beam splitter and lens quite affordable and doable for me. Bath/oil tests ?? Quite like the idea of Ed Jones test , but not sure of the fibre-optic bit, although I have a heap of single strand very fine FO here - left over from our FO/Broadband install.

So I,ll take the advice, and have a go at MKV,s 98.5% - why not ? Macleod.



#24 macleod

macleod

    Ranger 4

  • -----
  • topic starter
  • Posts: 325
  • Joined: 21 Sep 2014
  • Loc: Napier, New Zealand, 4110

Posted 19 July 2019 - 06:09 PM

Danny -any reference to oil-flat test ? Macleod



#25 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22712
  • Joined: 22 Feb 2010

Posted 19 July 2019 - 06:10 PM

Hey Macleod 

 

another thing about matching ronchi to computer computer generated images 

 

I don’t know a single person who does it that doesn’t also star test the mirror 

 

star testing helps confirm what side of correction your on and if your in the ball park

 

check out this thread

 

i posted some examples and how to. 

 

https://www.cloudyni...et-8-newtonian/




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