Jump to content


Photo

9.25 inch refractor project

  • Please log in to reply
407 replies to this topic

#126 Mark Harry

Mark Harry

    Vendor

  • *****
  • Vendors
  • Posts: 6188
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 13 March 2013 - 06:20 PM

Here's what I use for polishers. It's a type we used to use on flats. I discovered they can aspherize surfaces just as well as those square types above, but far more predictably. Pitch on these was real old, perhaps 10 years or more, #73.
I never, never let mine get glazed- generally they're just as black and shiney as when first made. These 2 were for a 6" F/10 primary, the far one didn't work, the near one smoothed and corrected the mirror in 20 minutes. Minor differences, but those differences either made it work, or it didn't. This type of polisher can do down to around F/4.5 paraboloid without a problem.
M.

Attached Files



#127 Mark Harry

Mark Harry

    Vendor

  • *****
  • Vendors
  • Posts: 6188
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 13 March 2013 - 06:30 PM

I'm fairly CERTAIN the flint is not biconcave; but rather a negative meniscus.
M.

#128 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 13 March 2013 - 11:29 PM

Gord, I think the best and most practical way to test convex lens surfaces is by the use of test plate interference method. The last surface to be figured can be tested by DPAC test agains a flat mirror, with a fully assembled objective, or using a spearate test plate.

Jim also mentioned an interesting oil test , which requires vertical testing. The oil would have to have a refractive index very close to that of the glass used. The test would require raytrace analysis for a proper test setup. However, cleaning and refilling the oil in the exact volume would make this method highly impracticsl for figuring purposes.

And yes, R4 in Jim's case is convex rather than concave, so he has three unknowns.
Mladen

#129 jimegger

jimegger

    Apollo

  • *****
  • Posts: 1200
  • Joined: 22 Dec 2005
  • Loc: Palmer,Alaska

Posted 14 March 2013 - 07:28 AM

Mladen, Yes the objective design has 3 convex surfaces, R1 , R2 and R4. As to the oil or any liquid for that matter , all you need is for one surface of the lens to be covered so as to prevent a flat surface to the knife edge. It would not matter if it was just covering or a couple of inches deep although just covering it would be best and all you would need. All you want to do is see the backside curve with the front side negated. At the very least it would be an easy way to tell which surface is the bad or worst one so you know which one to work on. The interference method is good if you are doing production work but for a hobbyist doing many different lens designs you would need a plethora of tooling. With a single pass collimation system such as I have using the 12.5 inch mirror You can accurately test many different designs with one setup. Same for the DPAC.

#130 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 14 March 2013 - 09:17 AM

There is some misunderstanding on how the oil test is done. The oil test is done by having the lens enclosed in a tube that is as long as the radius of curvature of convex surface you want to test and the tube is filled with oil. The oil has a refractive index that matches that of the glass so what your doing is making a test tunnel that has replaced the air and filled it with oil to cancel out the glass of the lens. Your testing thru the lens and the convex surface now acts like a concave surface. If done this way then when the convex surface is spherical it will test just like a spherical concave mirror but your going to need enough volume of oil to fill the tube not just cover the lens.
If you just cover the one surface of the lens with enough oil to form a flat surface and by doing so make a plano convex lens and test thru the oil surface, you have to take into account the flat surface. If you use a fluid that has different refractive index then the glass you have to take that into account as well. So if you tested this way and figured the lens to show a null and assume that the convex surface is spherical you'll be way off. The convex surface won't be a sphere but a hyberbola.
Here are the results from OSLO showing what happens if only an enough oil is used to make a flat surface on one side of the lens and having a perfect sphere on the convex side. As one can see the spot diagram is very bad and the convex surface would look like it is strongly aspheric. So this is not a very good test for a convex spherical surface and a test plate using interference testing is much better. To do interference testing all you need to do is flash polish the concave tool used to grind the convex one and polish it spherical. When the interference fringes are straight then the convex surface is also spherical.

- Dave

Attached Files



#131 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 14 March 2013 - 09:27 AM

Here is what happens if you assume wrongly that if you just use enough oil to make a flat surface on the one side of lens and test thru this and think that if you figure the convex surface to show a null it will be sphere. In reality what you have done is put a very strong hyberbola on that surface and when you assembled both elements of the achromat and test it you would be wondering why the image is not very good.
Here is the spot diagram showing the results. The spots are very good and the lens would test like a perfect spherical mirror but again what has happen is that the convex surface is now a hyberbolic with conic of about -1.2 and not spherical.

- Dave

Attached Files



#132 Mark Harry

Mark Harry

    Vendor

  • *****
  • Vendors
  • Posts: 6188
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 14 March 2013 - 03:14 PM

Had the same results when I tested a couple lenses. Though the convex is spherical, the test reveals an aspheric surface.
If you test the whole achro, (with a flat)you should get a null; which is what I did.
M.

#133 Mark Harry

Mark Harry

    Vendor

  • *****
  • Vendors
  • Posts: 6188
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 14 March 2013 - 03:16 PM

I'm almost thinking that there might be an error between R2 and R3- which can give an indication of over/undercorrection. With polishers like that, I would think the R1 surface likely has a huge hill in the center.
M.

#134 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 14 March 2013 - 07:00 PM

To do interference testing all you need to do is flash polish the concave tool used to grind the convex one and polish it spherical. When the interference fringes are straight then the convex surface is also spherical.

Dave, the problem today is that most people don't use solid glass tools but rather ceramic tile tools which, obviously, cannot be used for contact interference. Purchasing extra glass disks just to turn them into matching test plates makes a one-time project really costly and wasteful, so it is no surprise ATMs shun away from such techniques and seek alternative ways to get the job done.

That's why picking a well-thought-through design from the start, one that matches your skills and tooling, is so critical. For ATMs, refracting objectives or compound correctors with as many matching surfaces as possible is the most economical approach. One such example is the Houghton corrector, where the rear element is biconcave and the front element biconvex where R1 = -R3, and R2 = -R4.

Almost symmetrical telescope objectives are also possible, especially with modern glass melts. If one can devise an objective where three out of four surfaces are known to be spherical, either by Foucault or by contact interference of matching surfaces, then the system can be nulled by DPAC by working on the front surface of an otherwise finished objective. Also, configurations with very long, almost flat R4 can be advantageous.

Unfortunately, Jim here is working in the dark, except for R2 (which can be contact tested against R3). Short of making two additional test plates his engima will not be easily solved. Assuming he determines R2 is spherical, then it would be a matter of judgment whether R1 or R4 or both are the culprits.

I would guess it's probably R1 because its relative power (shorter radius of curvature) is more likely to affect the overall wavefront error, than the weaker R4 surface, and is morel like to distort the wavefront even with smaller figure errors than R4.

I don't have time to do this right now, but this can be illustrated easily using a simple raytrace analysis of wavefront variation over different conic constant values for for R1 and R2. Maybe later on tonight I can find some time to do this analysis.

Mladen

#135 jimegger

jimegger

    Apollo

  • *****
  • Posts: 1200
  • Joined: 22 Dec 2005
  • Loc: Palmer,Alaska

Posted 15 March 2013 - 12:41 AM

I do know the problem is in the crown because while working ONLY on the crown the objective went from under-corrected to over-corrected while using the convexing tools. I over shot the mark because I was trusting the flat to be giving me an honest reading which it isn't. Even though it is over shot I am now working it slowly back to where it should be and checking it with star testing. I have put 1 hour of polishing on the crown on both R1 and R2 while watching the star tests get better each time. When I get back from a trip in a week I'll be setting up the 12.5 inch mirror for the single pass collimation test and doing both Foucault and star testing in the shop so I won't have to wait for stars to come out. It should speed the whole process up considerably. R3 is right on a sphere and R4 is a long radius curve so it won't have as much effect on the image. It really doesn't matter as long as the objective nulls out in the Foucault test.

#136 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 15 March 2013 - 08:30 AM

Jim,
Are you only testing the crown using double pass with the optical flat ?
- Dave

#137 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 15 March 2013 - 10:38 AM

Jim, in similar configurations to yours, the R2 will contribute a lot more spherical aberration then R1 if it is not spherical. This is just another reason why it would pay to determine if R2 is spherical or not, and the only practical method I can think of with your tooling is the contact interference test against R3.

In MIJ's 4-inch f/15 configuration, with R2 = -R3, the LA will change 2.2 times more if R2 is given cc = 0.3 (oblate), rather then if R1 is given the same conic. The R2 will give LA = 0.0585, whereas R1 in LA = 0.0262.

Mladen

#138 jimegger

jimegger

    Apollo

  • *****
  • Posts: 1200
  • Joined: 22 Dec 2005
  • Loc: Palmer,Alaska

Posted 25 March 2013 - 02:31 PM

I have tested the crown only with the flat but also the whole objective assembly.

#139 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 25 March 2013 - 03:51 PM

I have tested the crown only with the flat but also the whole objective assembly.


Jim,
What is purpose of testing the crown in combination with autocollimation flat ?

- Dave

#140 jimegger

jimegger

    Apollo

  • *****
  • Posts: 1200
  • Joined: 22 Dec 2005
  • Loc: Palmer,Alaska

Posted 25 March 2013 - 08:46 PM

To see if it is or has spherical surfaces. Look for turned edges and spherical aberration.

#141 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 25 March 2013 - 10:29 PM

Jim,
The crown element by itself will show a huge amount of spherical aberration when tested by double or single pass autocollimation even if the surfaces were perfectly spherical. Only when the crown is combined with the flint will of the combination of the two shows a null when perfectly figured and you'll be able to determine if the surfaces are spherical. Just like the refractive index of the flint offsets the chromatic abberation of crown, the curves on the flint also offset the spherical aberation of the crown. It's the combination of both elements that correct for color and spherical aberration. Each by themselves has a large amount of both.
If you misuderstood this then this maybe the reason why you believe there is an issue with the flat. Since the crown element when tested by itself using double pass autocollimation would always show undercorrected spherical aberration but since it is a lens, all the errors are opposite of that of a mirror. So the picture you posted of the crown that looks like a parabolic mirror is not over correction but actually under correction. So if you believed that the test of crown was showing overcorrection and when you tested the assembled lens by a star test and it showed undercorrection, then that is were the confusion is coming from and misuderstanding that the quality of the flat maybe the cause.

- Dave

#142 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 25 March 2013 - 11:19 PM

To see if it is or has spherical surfaces. Look for turned edges and spherical aberration.

Jim, I am a little confused. Could you explain the rationale for testing a single biconvex lens by autocollimation?
Mladen

#143 jimegger

jimegger

    Apollo

  • *****
  • Posts: 1200
  • Joined: 22 Dec 2005
  • Loc: Palmer,Alaska

Posted 25 March 2013 - 11:27 PM

I did not realize that the crown could not be looked at alone to determine its aberrations. I have never read anywhere stating that the single lens elements could not be tested this way. One learns something new all the time ! Apart from that, I did also use the flat for both elements combined along with doing star testing. Originally the star testing showed under correction. After using the convexing laps it went to over correction according to the star tests. Now I am working on the crown to get it back gradually towards the proper amount of correction. I do understand the difference between mirror and lens Foucault appearance being essentially opposite and what needs to be done to correct it. Right now the lens is over corrected meaning the center has a shorter focal length than the edge so I am working to remove glass more from the center on the crown. It is gradually going towards the objective of equivalent star images on both sides of focus.

#144 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 26 March 2013 - 08:18 AM

Hi Jim,
Here is an OSLO plot for just the crown element of your lens being tested via Double Pass with perfectly spherical surfaces. The wave front error is close to 250 waves !

- Dave

Attached Files



#145 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 26 March 2013 - 08:27 AM

Jim,
Here is a drawing that might make it easier to understand why the zones on a lens are opposite of that of mirror. The drawing shows both a mirror and a lens with a hill on the surface. In the case of the mirror, the light is reflected off the convex surface of the hill and focuses long. In the case of the lens the light is refracted thru the hill and focuses short. You would see what looks like a hole when you tested the lens but it is actually a hill on the surface.

- Dave

Attached Files



#146 MKV

MKV

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 20 Jan 2011

Posted 26 March 2013 - 11:33 AM

Dave, that is a very good pictorial example. The problem is that even a simple lens has two surfaces and both may have hills or holes. The illustration below shows a hypothetical situation where a hill on one surface and a whole on another results in complete correction!

But assume that the correction is not perfect. Trying to fix one surface will result in overcorreciton or undercorrection, so it's essential to know that all surfaces are spherical in order to know where exactly to work on. However that would require at least two tmeplates.

Attached Files



#147 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 26 March 2013 - 12:28 PM

Mladen,
I agree that when both surfaces of the lens are unknown it is difficult tell which side or sides has the errors. As has been sugguested a good approach is to test the surfaces using interference against a test plate.

- Dave

#148 saemark30

saemark30

    Apollo

  • -----
  • Posts: 1187
  • Joined: 21 Feb 2012

Posted 26 March 2013 - 02:55 PM

Great project. Can you tell me how you make the pitch laps?
Do you use curved ceramic tool or curved glass tool as base?
I am wondering if the base can be flat.

#149 jimegger

jimegger

    Apollo

  • *****
  • Posts: 1200
  • Joined: 22 Dec 2005
  • Loc: Palmer,Alaska

Posted 26 March 2013 - 06:35 PM

So..... a simple lens will not be able to be tested via Foucault test even using a single monochromatic wavelength of light .... because..... why is that ?

Mladen, the test plates are of course one way to see which surfaces are spherical ... but there must be others as well. It can not be that there is only one way.

From what I have read in ATM books, an objective will work just fine if a defect on one surface is compensated by a correcting defect on another surface such as the hill hole scenario Mladen did above.

I have the glass tools to make test plates from if it becomes necessary. For now I am going to keep working towards a better star test and Foucault test by more work on the crown with blending strokes to try and get the center down for a longer focal length there. It is so close now I can taste it !

#150 DAVIDG

DAVIDG

    Aurora

  • *****
  • Posts: 4881
  • Joined: 02 Dec 2004
  • Loc: Hockessin, De

Posted 26 March 2013 - 07:21 PM

So..... a simple lens will not be able to be tested via Foucault test even using a single monochromatic wavelength of light .... because..... why is that ?


Jim,
A singlet, using spherical curves can be designed to have no spherical aberration at only one wavelength but in your case the curves on the crown element were choosen to have spherical aberration so the crown with not null when tested by itself. In an achromat it is the choice of the curves and the index of refractive of the elements working together to both correct for color and spherical aberration at the same time. In reality your objective is only fully corrected for spherical aberration at only one wavelength as well, that being in the green at 546 nm. All the other wavelengths are either slightly over or under corrected. This is called spherochromatism ie spherical aberration as a function of wavelength.
The new book by Smith, Ceragiolli and Berry "Telescope, Eyepieces and Astrographs" has two excellent chapters on achromats and apochromats and it explains how they work and what each element is doing.
I agree that a lens will work well when a defect on one surface is corrected by another, but right now you have 3 convex surface that are unknown. You need to get them close to, if not perfectly spherical to have a chance of correcting the assembled lens. So I agree with Mladen, your best bet is to use test plates to test and figure the convex surfaces. Once they are well figured to spherical surface you can test the assembled lens by double pass autocollimation and tweek a surface (usually R4) until the lens nulls perfectly Now you'll have a very well figured lens that most likely show no spherical aberration when star tested and will give jaw dropping images.
If you believe your close then test the assembled lens by double pass autocollimation. There is no guessing and it is a far more sensitive and much less error prone test then all your others test methods. If your close then the lens should show only faint zones and look nearly spherical.

- Dave






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics