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

Who wants to make a Ross Null Lens?

  • Please log in to reply
174 replies to this topic

#26 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8,798
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 15 August 2020 - 06:21 PM

"In the Ross null test the light goes through the lens twice (double pass),...."
Yes, and it also doubles the "half-error" each surface of the reference iintroduces into the result.
********
 One of the other defects any lensmaker is aware of, is typically with methods used, there is a visible error that occurrs at about the area of 1/3rd the diameter of the lens, in the central area.
As I said before there is a REASON why 1/20th wave errors of a certified reference cost $15,000.
If you don't believe this, contact ZYGO in CT and ask them what a 4" reference is gonna cost.
Personally, i wouldn't entertain the thought of a 4" reference element cheaper than mid-4 figures would be guaranteed to perform adequately in an arbitrary test.


  • MKV likes this

#27 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 16 August 2020 - 10:07 AM

"In the Ross null test the light goes through the lens twice (double pass),...."
Yes, and it also doubles the "half-error" each surface of the reference iintroduces into the result.
********

✓ waytogo.gif

 

 

One of the other defects any lensmaker is aware of, is typically with methods used, there is a visible error that occurrs at about the area of 1/3rd the diameter of the lens, in the central area.
As I said before there is a REASON why 1/20th wave errors of a certified reference cost $15,000.
If you don't believe this, contact ZYGO in CT and ask them what a 4" reference is gonna cost.
Personally, i wouldn't entertain the thought of a 4" reference element cheaper than mid-4 figures would be guaranteed to perform adequately in an arbitrary test.

Reality ✓

 

Ceravolo's Ross null lens is an 80 mm f/4 precision PCX lens (1/10 wave on the surface, each side) which originally sold for $350 more than a decade ago. Today it's priced at $650! Of course it comes with a Zygo certificate. Again, Ceravolo never suggested the Ross null lens should be done with commercial grade (i.e. 120/80) optics. He simply couldn;t do that as a professional optician. Such lenses are specifically cassified as unsuitbale for testing.

 

Ceravolo also never intended to use his Ross null lens with an interferometer because, as you say, interferometer-grade reference elements 100 mm in diameter cost $15,000. He couldn't sell such a lens for $650 and make it profittable.

 

Take, just for example, the tiny reference element for his spherical wave interferometer: it's a 25.4 mm lens three times smaller than his Ross null lens), which sells for $600! As you say, there's a reason for it.

 

But if someone can crank out such lenses at home, by hand and without sophisticated equipment, more power to them! They should qualify for a Nobel Prize in optics. If they're doing it just for the experience" -- have fun! :o) 


Edited by MKV, 16 August 2020 - 10:07 AM.

  • PrestonE likes this

#28 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 24,885
  • Joined: 22 Feb 2010

Posted 16 August 2020 - 10:28 AM

 

Again, Ceravolo never suggested the Ross null lens should be done with commercial grade (i.e. 120/80) optics.

In his write he sure uses it to make a 1/60 ~ mirror. Certified with his IF.

 

start making mirror instead playing with software and you’ll see how it is really helpful.

 

maybe China should use it instead whatever they use cause a lot of commercial telescopes are not that great.



#29 GLS

GLS

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 01 Aug 2020

Posted 16 August 2020 - 11:33 AM

More power to  Peter to get paid big bucks for his time and effort.  That's why he is in business.  It is surely worth it.  He seems like a nice enough fella as well; not one to lord his superior talents over us poor little glass pusher wannabees.  Gil


  • Pinbout likes this

#30 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 16 August 2020 - 12:15 PM

Again, Ceravolo never suggested the Ross null lens should be done with commercial grade (i.e. 120/80) optics. 

 

In his write he sure uses it to make a 1/60 ~ mirror. Certified with his IF.

start making mirror instead playing with software and you’ll see how it is really helpful.

Maybe you need to re-read (if you've ever read) what Ceravolo actually wrote. Here, with my emphasis:

 

The precision Ceravolo Ross null lens is 80mm in diameter, with a focal length of 390mm. When the test set up is calculated for a 16” f/5 mirror, only 54mm of the diameter is used, thus producing a null with only 1/60 th wave residual—perfect, for all intents and purposes.When used with a 30” f/4.5 mirror, the COS null lens will have 70mm of its aperture utilized, and the residual wavefront aberration is 1/13th wave peak to valley. With a perfect set up and perfect figuring on stable substrate one can produce a figure better than 1/20 th wave peak to valley

He explicitly writes the 1/60 wave residual is in connection with his precision lens (only). Furthermore, where do you see that he actually achieved this result? Nowhere! It was a software calculation for an ideal system.

 

And for your information, I have been making mirrors for over 30 years, so thanks but no thanks. 

 

BTW, why don't you use your Bath? It's much better than the Ross null.

 

 

More power to  Peter to get paid big bucks for his time and effort.  That's why he is in business.  It is surely worth it.  He seems like a nice enough fella as well; not one to lord his superior talents over us poor little glass pusher wannabees. 

Who are you referring to? People who offer cautious encouragement, but  also caveats with good intentions? That's rich.


Edited by MKV, 16 August 2020 - 12:16 PM.

  • Mark Harry likes this

#31 GLS

GLS

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 01 Aug 2020

Posted 16 August 2020 - 12:51 PM

I'm referring to Peter.  Any problem?  

I'm sitting by the phone waiting to hear from the Nobel Committee. But thanks for asking.  Gil


Edited by GLS, 16 August 2020 - 12:57 PM.


#32 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 24,885
  • Joined: 22 Feb 2010

Posted 16 August 2020 - 08:17 PM

 

BTW, why don't you use your Bath? It's much better than the Ross null.

Not to make mirrors, only to qualify 

 

even my 16” perf flat is quicker than a bath

 

but the Ross is the easiest 


Edited by Pinbout, 16 August 2020 - 08:20 PM.


#33 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8,798
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 17 August 2020 - 06:25 AM

"But if someone can crank out such lenses at home, by hand and without sophisticated equipment, more power to them!"
******
Wishful thinking, most certainly.

If anyone has one of these so-called 80mm Ceravolo lenses, I will volunteer to take it to work, and test it for any error it may have--- on both sides.
That way, it will be certified.
But don't be disappointed if the 1/60th wave throughput error is just wild speculation.
Testing will put to rest any conjecture in this thread.
*****
There are 3 of these IF's at work. You enter the lens diameter, the aperture desired, and hit "enter". It not only QUALIFIES, but QUANTIFIES---in about 3 seconds to an exact focal length of sub-micron radius accuracy. Cost of these IF's is $150k each. (not including the reference spheres) Certainly better than something built in a garage.



#34 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 17 August 2020 - 06:56 AM

Not to make mirrors, only to qualify 

even my 16” perf flat is quicker than a bath

but the Ross is the easiest 

From someone who avidly advises others how to make mirrors, I'm surprised to read this. In the final stages of finishing astronomical mirrors and optical flats to a fraction of a wavelength accuracy, regardless what test you use, quicker and easier should never be deciding factors. To the contrary, this is when you really want to take your time.  

 

Delicate optical surfaces take time to settle down after even a quick "round the barrel" touch-up, or when handled with bare hands. It takes seconds to reacquire your optics on a precision optomechanical kinematic mount.

 

You can most certainly use the Bath to qualify and to make mirrors with ease. 



#35 GLS

GLS

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 01 Aug 2020

Posted 17 August 2020 - 07:53 AM

Pinbout, in your classes how many  mirrors were made with Ross and star tests?  Were the attendees pleased?  Just curious.  Gil



#36 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 17 August 2020 - 08:23 AM

"But if someone can crank out such lenses at home, by hand and without sophisticated equipment, more power to them!"

******
Wishful thinking, most certainly.

I think so too, but I wish to give everyone the benefit of a doubt. It certainly would be helpful if everyone simply spelled out, step-by-step, just how they intend to achieve that level of precision. If they don't want to share their "trade secrets," maybe they should just make their precision optics, have them certified for a few hundred $$$, and then write about it -- for bragging right, if nothing else. But, then, why not just spend those $$$ to buy a finished product? :o)

 

If anyone has one of these so-called 80mm Ceravolo lenses, I will volunteer to take it to work, and test it for any error it may have--- on both sides.

That way, it will be certified.
But don't be disappointed if the 1/60th wave throughput error is just wild speculation.
Testing will put to rest any conjecture in this thread.
*****
There are 3 of these IF's at work. You enter the lens diameter, the aperture desired, and hit "enter". It not only QUALIFIES, but QUANTIFIES---in about 3 seconds to an exact focal length of sub-micron radius accuracy. Cost of these IF's is $150k each. (not including the reference spheres) Certainly better than something built in a garage.

Mark, this is a fantastic offer. Hope there's one soul who will take you up on it.

 

Years ago, I toyed with the idea earlier of getting one as an Offner nulling compensator, but decided against it because of it's longish focal length and a bulky setup, as well as its a price tag. It was selling for $650+S&H. Even if they keep that price (they promised they'd lock it in for me), I need to have the front surface facing the mirror AR coated, and then there are possible customs tariff charges from Canada, all in all I'm thinking at least $900-$1,000 in all. Not worth it. 

 

OTOH, I can't imagine there are too many folks killing themselves to get one...so maybe they will be flexible. The lens itself used to cost half that about 15 or so years ago (I'm guessing, since there's no date on the copy-right online PDF file), so yeah we call that nowadays "progressive". :o)

 

I also found this CN classified (posted Jan 2020) never used (serial number 4) selling for $700! What a deal...sarc.  https://www.cloudyni...94860-ceravolo-ross-null-lens-sn-4/.  Says bought but never used -- yeah, right. And this is far from the only person trying to get rid of it -- but not for a bargain. They all try to recoup the money they spent and maybe even make a few bucks more. Some deal.

 

Actually, all of this is a little dated by now. DFTFRinge IF analysis program by Dale Eason (who has indebted the ATM community with free sophisticated software, and deserves recognition) has a digital null, so even an AC flat is no longer a necessity. 

 

But I think Gil should take you up on your generous offer since the OP is eagerly waiting for that phone call from the Nobel Committee (just kidding with you, Gil). :o) 



#37 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 17 August 2020 - 08:34 AM

Pinbout, in your classes how many  mirrors were made with Ross and star tests?  Were the attendees pleased?  Just curious.  Gil

That's a fair question of course, but satisfaction is one thing and precision is another. If all you care about is satisfaction then different standards apply. Celestron and other companies guarantee satisfaction. You can always return their product, knowing someone else will be happy with it no matter what the other person thought was wrong wiht it. Satisfaction is not an objective criterion; precision is

 

Some Delmarva mirrors have won Stellafane awards. I'm not sure for what classification, but the Delmarva classes went on for about 15 years, if not longer and I'd say many if not most were satisfied, based on what I have read from DavidG, and maybe he can help with the number of actual Stellafane awards received for Delmarva mirrors (many of them made with the help of a very good autocollimation test, and a lot of expert assistance). And maybe he can also tell us who received what award for and what classification.



#38 GLS

GLS

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 01 Aug 2020

Posted 17 August 2020 - 08:35 AM

Mark, are you saying Ceravolo's 80 mm lenses were made with garage equipment or just mine?  Just trying to determine who your target of ridicule is.  

For the record I didn't build mine in the garage, but probably would have but for a boat project.

As for garage built myths I was upset to learn that Jobs and Wozniak didn't build the first Apple in the garage but elsewhere in the house.

I will be happy to send my lens to you for testing although I've never said it was 1/60 wave or 80mm.  PM me your address and I'll send it to you.  Test at 3", 2.75, 2.5 and 2".  It is 4" OD but in  use masked off at 3".  Maybe Grade A Schott isn't better than top of the line Ohara or Schott but it is hoped to be better than the bottom of an old Coke bottle. Thanks for the opportunity.  Gil

P.S. work address is preferred as a signature will be required.


Edited by GLS, 17 August 2020 - 09:22 AM.


#39 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8,798
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 17 August 2020 - 09:36 AM

I am not trying to RIDICULE anyone, or anything! I want to present some factual information, and you guys can make up your own mind.
As to what I use for IF, I have related it already- in having excellent examples at work to use, but with size constraints I do not test mirrors there. The rails are only ~2 meters long that can afford the accuracy they require there.
A Ross lens should be fairly simple to test, and take just a couple minutes to test, and make a print-out.
If you read Mladen's post about residual error claiming 1/60th, that's where I am commenting on. That claim is awfully "lofty" and sounds far too good to be true. It's my 2¢, but I did a carload of lenses and tested them in the last couple years to make such a comment.
I have a picture of a PCV lens, which doing a a batch of about 150, may give you an idea just what i was doing---and was my first job to complete there. From generating-grinding-polishing- and with 100% verification of each piece.
Also, one of the IF's.

Attached Thumbnails

  • interferometer at PPO resized.jpg

  • PrestonE and MKV like this

#40 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8,798
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 17 August 2020 - 09:43 AM

The lens blank. This was a setup blank, that was fine ground, but wound up slightly undersize for CT.
Took me a week to get 150 finished to tolerance.

Attached Thumbnails

  • A    lens prodge at work 5.9 resize CN.jpg

  • MKV likes this

#41 GLS

GLS

    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 01 Aug 2020

Posted 17 August 2020 - 09:44 AM

Looks good.  Awaiting where to send it. PM, please.  Gil



#42 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8,798
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 17 August 2020 - 10:08 AM

Sent.


  • PrestonE and davidc135 like this

#43 tommm

tommm

    Viking 1

  • *****
  • Posts: 853
  • Joined: 16 Dec 2015

Posted 17 August 2020 - 10:25 AM

Sent.

This will be very interesting! I'm not interested in "who wins the argument", just in seeing how good a lens a person might be able to make. I've always been intimidated to try making lenses, so it will be nice to see test results - especially if it turns out to be fairly good. Thanks for doing this testing, and to Gil for volunteering his lens. To me that's what it is about - getting data to determine how just what surface accuracy we have.


Edited by tommm, 17 August 2020 - 10:28 AM.

  • PrestonE and davidc135 like this

#44 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 24,885
  • Joined: 22 Feb 2010

Posted 17 August 2020 - 12:39 PM

 

If you read Mladen's post about residual error claiming 1/60th, that's where I am commenting on. That claim is awfully "lofty" and sounds far too good to be true. It's my 2¢, but I did a carload of lenses and tested them in the last couple years to make such a comment.

that's not MVK's claim, that's Ceravolo's claim on a 16" f5 ... with his 80mm lens that only uses partial use of the lens.... written 2003.

 

and on a 6inf4 and ridiculous residual error of 1/132~

 

at Delmarva we were always taught that the lens would produce 1/8~  on wavefront ... or better. nothing lofty. But Ceravolo says 1/20~ surface

 

Dick Parker's AC would produce 1/10~ on the wavefront.

 

Ceravolo's lens is made and then quantified on a Zygo from BMV Optical


Edited by Pinbout, 17 August 2020 - 12:39 PM.


#45 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 17 August 2020 - 03:47 PM

...at Delmarva we were always taught that the lens would produce 1/8~  on wavefront ... or better. nothing lofty. But Ceravolo says 1/20~ surface

Dick Parker's AC would produce 1/10~ on the wavefront.

For an AC with a good flat, and a precision Ross null test, using a knife-edge, not Ronchi grating, that's doable with skilled workers. Remember, Ceravolo's calculated wavefront errors presumed a perfect lens of his design, and the following "With a perfect set up and perfect figuring on stable substrate one can produce a figure better than 1/20th wave peak to valley." (page 5 of his PDF report; my emphasis added).

 

Now, we all know that no perfect setup and no perfect figuring is real (although professional labs come mighty close to perfection for all practical purposes, but not the way I have seen ATMs do it).

 

That statement of claim is a purely mathematical construct for a specific mirror configuration and should be taken with a grain of salt just as the a prabolic on-axis raytrace is a dimensionless dot, and not an Airy disc.

 

So, again I think people inadvertently (not intentionally) read into it something that Ceravolo probably never intended to say.


  • Mark Harry likes this

#46 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 24,885
  • Joined: 22 Feb 2010

Posted 17 August 2020 - 04:18 PM

His lens 1/10 wave 

 

1/20 surface 

 

you use perfect way too much

 

even in his 2003 article - the Ross test will reveal errors that you will never see in a star test, at some point getting rid of them is just ego.


Edited by Pinbout, 17 August 2020 - 04:21 PM.


#47 duck

duck

    Sputnik

  • -----
  • Posts: 43
  • Joined: 11 Jun 2020
  • Loc: madera ca

Posted 17 August 2020 - 04:52 PM

Who wants to make a Ross null lens?  I do.  In fact, the BK7 element of a f/15 Fraunhofer refractor objective looks like a promising candidate.  Dreaming of a 24" f/4&f/20 Newt/Cass.  Needs a 6" guide scope, so there's symbiosis.

 

I couldn't tell you what the exact precision of the mirrors I've made with a combination of testing.  I can tell you that the last one, a 12 1/2" f/4.0&f/20 Newt/Cass with a fused quartz primary is by far the best.  I'm a fortunate guy who has the $ to invest in test optics.  Better be fortunate, because the quoted price for a 24" fused quartz blank 3" thick, generated and cored, was > $20,000.  OUCH!

 

Have made my last 4 mirrors with using un-nulled knife edge to get to 1/2 correction.  Then Ceravolo's Ross null lens to get a null with the knife.  Then nulled IF with Ceravolo's ref element.  Takes lots of experience to get rid of lens generated astigmatism.  Always go back to the nulled knife to look for roughness, especially when using small laps.  I had to use a 1" pitch covered with felt on the last mirror, thanks to a sequence of blunders.  Smoothed and final correction using a lap made from the plug.

 

I digress, but I like to digress.  


  • Pinbout and GLS like this

#48 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 17 August 2020 - 05:56 PM

His lens 1/10 wave 

1/20 surface 

No. Caravolo's precision lens is 1/10 wave surface on each side. That equals 1/20 wave wavefront because the refraction factor is about 0.5 (1 - crown glass refractive index), hence 0.1*0.5 = 0.05 or 1/20. In mirrors, the factor is 2, hence 0.1*2 = 0.2 or 1/5. Mirrors that form an image are spec'd on the wavefront. Flats are spec'd on the surface, like the lenses, but the peaks and throught are doubled on reflection at 90 degrees, as opposed to halved in lenses..

 

 

 

you use perfect way too much

Not I. Ceravolo mentoned perfect setup and perfect figuring in his calucations. Stay with the program :o)

 

even in his 2003 article - the Ross test will reveal errors that you will never see in a star test, at some point getting rid of them is just ego.

That doesn't surprise me. I think the star test is way overrated. Ceravolo wrote an article years ago in which he says that most people couldn't tell a difference between a 1/4 wave telescope and a 1/10 wave one. He knows very well that the atmosphere overtakes the performance of excellent mirrors and lenses, except on a very rare occasion.

 

BTW, that article was copyrighted in 2003, but Peter Ceravolo published articles on the Ross null test in 10 years earlier in the ATM Journal (which was edited by Richard Berry). So, I am sure if the price of about $300 for his Ross null lens is the 1993 price or the 2003 price. My guess is it's the latter. 



#49 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 24,885
  • Joined: 22 Feb 2010

Posted 17 August 2020 - 06:03 PM

 

 

But in practice, a lens size that
produces a residual wavefront error of 1/10th wave or better will work very well. This set
up will produce a 1/20th wave surface if the mirror is figured perfectly   

 Peter came and taught one year at Delmarva if i'm not mistaken, before I was into that stuff. guest star...

 

 

his conclusion - for those who haven't read his article

 

 

 

I have tried to be thorough and practical in describing the implementation of the Ross null test. At first reading, with all my caveats and warnings, the whole process may sound intimidating. But with persistence and practice—qualities in abundance in the average telescope maker—the Ross null test can be mastered. 
The Ross null test was introduced to the telescope making community over ten years ago. Since then computers have changed the hobby. Doug George’s software has made it easier to implement the test. Affordable interferometry [link to the booklet page] and computerized fringe analysis [link to COS Quick-Fringe page], places professional capabilities in the hands of the hobbyist. Telescope makers are no longer limited to the tools and techniques of the last century.

Edited by Pinbout, 17 August 2020 - 06:09 PM.

  • RichA likes this

#50 MKV

MKV

    Cosmos

  • *****
  • Posts: 9,065
  • Joined: 20 Jan 2011
  • Loc: Southeast, USA

Posted 17 August 2020 - 06:27 PM

 

 

Peter came and taught one year at Delmarva if i'm not mistaken, before I was into that stuff. guest star...

But in practice, a lens size that produces a residual wavefront error of 1/10th wave or better will work very well. This set up will produce a 1/20th wave surface if the mirror is figured perfectly

And you wrote

 

 

 

His lens 1/10 wave

 

1/20 surface

Please write in full sentences. You either misunderstood what he said or misquoted him again. All it means is that a lens that will produce 1/10 wave on the wavefront (the lens  surface is 1/5 wave) will require a mirror 1/20 wave surface to create the same wavefront error. In other words, the mirror has to be four times better surface-wise than an equivalent lens.

 

In the Ross test, however, the two surfaces and the lens gets traversed twice hence the 4:1 advantage is lost. A much simpler and easier to understand rule is that in the Ross null test the lens has to have at least twice as small a wavefront error as the mirror. In reality even smaller than that because of other concomitant factors. The same is true of similar (Burch, Offner, etc.) lens test configurations, which also require high precision lenses. 




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