22 replies to this topic

[peleuba]

Over the last 12 months or so there has been an increasing amount of interest in testing one's own optics.  The most often sited test here in the Refractor forum is the Double Pass Auto Collimation test as it gives you an excellent picture of surface quality and spherical correction of the lens.  It can also give insight into how a lens behaves at different wavelengths - ie. "color correction" as well as spherochromatic correction and focus point as a function of wavelength.  Its not a very sensitive test for astigmatism or coma.  For these, the star test is preferable.  And, actually, when testing a lens you should seek to have at least two corroborating test methods before determining the "goodness" of an optic.  In other words DPAC and star test that agree with each other is far stronger then any single qualitative test method. 

 

Below are tests of a 76mm ED doublet and a 9.25" SCT.  The SCT (has the central obstruction)  These telescopes both suffer from some minor spherical aberration of opposite sign - the 76mm telescope is undercorrected and the SCT is over corrected.  You can see the correction is "opposite" as the curvature of the bands are opposite when comparing the resulting images of the different telescopes on the same side of focus.  You can also see that 76mm scope is has a very smooth lens.  The SCT has a somewhat rough corrector but not too bad - I've seen worse.

 

Inside of focus images are on the left; At focus images in the center; Outside of focus on the right.

Attached Thumbnails

• 
• 
• 

Edited by peleuba, 12 March 2020 - 01:17 PM.

[DAVIDG]

 Having done double  pass for many many years, a couple of corrections. First  double pass is VERY sensitive to astigmatism. Line up the grating one on side of focus so the lines are vertical, then moved to the other side of focus. If the lines rotate by any amount you have some level of astigmatism.  I do agree with your statement that you should use as many test methods as possible and they all should agree. If not don't take the best results as the truth but at the same time double pass has very few source of errors. So the confidence in the results are  very high compared to other test methods were many times the errors in the result are unknown. I  have never seen optics that test well in double pass testing badly with other test method when they are correctly done. I have seen thou many time optics tested by other test them that  show good result test poorly with double pass because of unknown errors in the other test methods. 

  Also I judge both of the your optics as having a fair amount of spherical as in around 1/4 wave from doing this test and fabricating optics with for over 35 years. 

 

                - Dave 

[peleuba]

 Having done double  pass for many many years, a couple of corrections. First  double pass is VERY sensitive to astigmatism.

 

Hi Dave - Thanks for feedback I think you only made an attempt at one correction.  

 

And, yes - I have seen astigmatism in DPAC myself.  I did not say it was impossible to see.

 

However, I do contend that its difficult to do better then the star test when detecting and quantifying astigmatism.  I am not speaking about "weekend warrior" type of casual star testing.  I am specifically speaking of indoor testing as I perform it - on a bench with a collimator (parallel light), a precision 5µ pinhole and some aspherics that presents an artificial star at infinity to the telescope under test. 

 

There is no alignment of a Ronchi grid; no looking for a subtle circular movement of the bands as you carefully rack in/out of focus.  The star test is unequivocal and unmerciful when looking for astigmatism.  Its plain as day to see.  And, with only a little effort one is able to reasonably quantify it.

 

As an aside, back in February of 2011 you and I exchanged PM's on CloudyNights as well as some private email about testing a 160mm Fluorite APO I had at the time.  You invited me down to Mount Cuba and the DELMARVA Mirror Making Workshop happening later that Spring so that I could get a look at the double pass jigs and you could evaluate the 160mm lens.  At that time I explained how I was performing indoor star testing and you encouraged me to set up a DPAC rig.  I am sorry to have never made it to the DELMARVA event and its my loss to have never met you in person.  

 

In any event, DPAC is a terrific test for the very reason you specify - very little chance of error and gives an overall picture of the optic.  But its not the best tool in the box to  detect, assess/quantify astigmatism.  To this end, DPAC in conjunction with the star test is tough to beat as each has its own strength.  And, when the results agree one can be reasonably certain of the quality of the optic.

 

Agree on the 1/4 wave assessment.  Maybe slightly better.  Refractor is very smooth.  

Edited by peleuba, 12 March 2020 - 01:34 PM.

[DAVIDG]

"However, I do contend that its difficult to do better then the star test when detecting and quantifying astigmatism.  I am not speaking about "weekend warrior" type of casual star testing.  I am specifically speaking of indoor testing as I perform it - on a bench with a collimator (parallel light), a precision 5µ pinhole and some aspherics that presents an artificial star at infinity to the telescope under test."

 

   I'm sorry but I have to disagree. The reason is double pass as the name states  double errors including astigmatism. It is very easy to see even a very small amount of rotation of the bands especially if you quickly move the grating from one side of focus to other. You don't have to align the grating in any orientation, it is just easier to see if the bands rotate when you start with having them vertical on ones side.  It also has the advantage that you see if the total figure has astigmatism or  is it come from an off axis asymmetrical zones or in  a refractor a wedged element. A wedged element will show up as bands being curved or distort on one side and symmetrical on the other.   Another  advantage over a star testing is it much easier to see if you have made an improvement or not in both the over all wavefront and the astigmatism if you changed something.  One has to be careful when trying to fix one error and introducing another. 

   You might want to try something with your refractor. Flip the complete lens around on the test stand so "rear" of the lens is now skyward ie facing the flat and see how it tests. It is not uncommon to have an element or the whole lens backwards in the cell and that would cause the test results your seeing.  Many times a flipped lens will not produce an image that is total fuzzy mess but one that doesn't look to bad. I  have seen it a number of times. It will only take a few minutes and it might uncover  an error that is easy to fix and a much better image.

 

            - Dave 

[peleuba]

 

   I'm sorry but I have to disagree.

 

Thanks for the response.  No need to be sorry.  Yes, understood - yes we disagree.  Hopefully some others will chime in.  

 

 

 

The reason is double pass as the name states  double errors including astigmatism. 

 

Respectfully, this is not my first rodeo vis-a-vis optical testing so am not sure why you are treating it as such. For the last 20+ years I have a had a very good teacher.  And, I've  learned a lot from your posts, too.  Nevertheless, I understand and have mastered the fundamental principles of testing in Double Pass and why its sensitive.  Moreover, your defense of DPAC is a bit odd to me as no one is implying that DPAC is not a good test, on the contrary!  Its my favorite one to use - just not for astigmatism or coma (or wedge which we I'll discuss further, below).  I prefer the star test for these and perhaps confirmation with DPAC.   

 

But I need to ask, have you ever tested indoors on an optical bench with a collimator generating parallel light and a precision pinhole to use as an artificial star?  When we last corresponded (2011), I came away with the impression you had not done this.  So, my point is this:  I have done both and for astig, I like the star test.  DPAC is incredibly useful, fun and addicting.

 

 

You might want to try something with your refractor. Flip the complete lens around on the test stand so "rear" of the lens is now skyward ie facing the flat and see how it tests. It is not uncommon to have an element or the whole lens backwards in the cell and that would cause the test results your seeing.

 

The refractor above is a TeleVue 76.  Images are decent.  I will test as you say and see but doubt that the lens is in backwards.  I have tested enough TeleVue OTA's to know that the optics can range from barely diffraction limited to being very good.  

 

Now - Lets talk about wedge and astigmatism.  Below is a small aperture APO that has nearly perfect spherical correction but suffered from a very slight amount of trefoil astigmatism and very slight wedge.  None of this was evident in DPAC.  Perhaps it was there, but its easier for me to see in the star test.  And that is what this point/counterpoint discussion you and I are having is all about.  See the images below.  

 

 

====

First image is star at very high power on the bench.  Notice the color in upper right quadrant - that is tilt/wedge.  Notice the triangular shape - that is Trefoil Astigmatism.

 

Second image is typical high power star test intra and extra focal of same tekescope with points of interest in the outer ring labeled. 

 

Third Image,same telescope, is one done in polarized light to see if strain exists in the lens.  The reason this is important as the manner in which the wedge was adjusted out was to turn plastic screws on the side of the lens cell.  You can see the affect of the screws pushing on the glass in polarized light.  Its a very sensitive test and no detrimental affects were seen once finished.

 

The fourth image is a DPAC test of the same scope.  Jail bar straight.   

 

=====

Attached Thumbnails

• 
• 
• 
• 

Edited by peleuba, 13 March 2020 - 12:19 PM.

[Kent10]

Paul, thanks for the discussion and pictures.  I am enjoying it. 

 

What do you look for to tell if the lens is smooth or rough?  Why is the in focus image of the 76mm dark on one side and bright on the other?  What does that tell you?

 

Thanks, Kent

[peleuba]

Paul, thanks for the discussion and pictures.  I am enjoying it. 

 

What do you look for to tell if the lens is smooth or rough?  Why is the in focus image of the 76mm dark on one side and bright on the other?  What does that tell you?

 

Thanks, Kent

 

Hi Kent,

 

The image of the 76mm refractor is dark on one side and light on the other because its very near best focus where I used a singe line of the Ronchi grid to cast a shadow.  This is very similar to what you would do if the lens was being tested using a Foucault tester.  The shadow helps with contrast in viewing the surface features.

 

A rough surface looks that of a golf ball and/or it will have circular full circumference polishing artifacts.  The image I posted above - of the refractor has none of this.  The SCT has some.

 

Mike Lockwood has some good photos of optics with rough surfaces in his "hall of shame".  See here:

 

http://www.loptics.c...l/foc_hall.html

[Kent10]

Thanks Paul.  Mike Lockwood's pictures really help to show the roughness.  That is probably about as bad as it gets, so it is easy to see.

[peleuba]

Thanks Paul.  Mike Lockwood's pictures really help to show the roughness.  That is probably about as bad as it gets, so it is easy to see.

 

Mike photos show Newtonian mirrors.  There are other examples of SCT's that are (a lot) worse.  The SCT has three optical elements that either refract or reflect light - the corrector, the primary and the secondary.  Most of the roughness on SCT's is in the corrector plate.  The Criterion SCT's are horrible.

 

IIRC, Dave G. - who is participating in this thread - has undertaken a project to remake the corrector of a Criterion SCT.  I am sure he can relate some horror stories as he has seen many more then I ever will.

[Jeff B]

Hi Paul and thanks for posting this thread.

 

Yup, my experience with DPAC (with mostly refractors) is really similar to yours.  While good for seeing spherical errors, zones and gross astigmatism (clocking and/or non-symmetrical line bowing about focus), I much prefer star testing to get a good handle on astigmatism and, especially, coma (I have no idea what coma looks like in DPAC or if it can be seen at all).   Trouble is, I don't have an indoor star test such as yours.

 

But when looking for astigmatism in DPAC, I've indeed seen clocking of the lines through focus, but I am unclear as to how much clocking relates to how much astigmatism there is.  Also, I've experienced what I call "false positives" regarding astigmatism.  I see some clocking but I do not see it outside in star testing even up to ~35-40X per inch of aperture.  Maybe I'm not going high enough in power to catch it out, in which case it is a very small amount, but this also points out me not having any type of "calibration" for astigmatism seen in DPAC versus visually under star test (which I'm much more calibrated on...if it's simple astigmatism).  One thing I've always wandered about was, can there be differences between laying a scope on its side for DPAC and then pointed up in outdoor star testing.  Of course, indoor star testing is mostly done, I imagine, with the scope "lying down".

 

Now for lower order SA, how about some computer generated images showing 0, 1/8, 1/4, 1/2 and 1 wave of SA? 

 

Thanks again Paul!

 

Jeff

[Steve Allison]

Jeff, Paul and David, thank you all very much for your interesting, and highly authoritative and informative posts!

 

I have learned a lot from all of you and as my understanding of the optical side of our wonderful hobby has grown, so has my enjoyment! Please continue to share your expertise with the rest of us on this great site.

Edited by Steve Allison, 13 March 2020 - 02:02 PM.

[m9x18]

Ditto Steve. I could not have said it better.

[DAVIDG]

 If you were to take take two image of the lens in DPAC mode with one taken  inside of focus showing three lines and one outside you would see that the bands will have rotated. Also if you go to one line and use the grating as a knife edge or switch to a knife edge  you would see the lens doesn't darken uniformly  from  the astigmatism. Remember the closer you get to the focal plane to more sensitive the test. If you defocus using an artificial or real star by too much you won't detect astigmatism either. So to compare apples to apples between a star test and DPAC for astigmatism you need to be at the same sensitivity level which is at the same distance from the focal plane. 

  Coma is an off axis aberration and DPAC  like a Foucault test or many other optical test is on axis test. Coma shows up in DPAC test as the Ronchi lines all bowing in one direction like this )))). What that means thou is that you don't have the optics being tested and the optical flat aligned correctly.  

  If you want an  idea of what different levels of spherical aberration would look like, use a program that simulates Ronchi tests of a spherical of same aperture and F-ratio of the  optics your tests. Tell  the program to show the Ronchi  pattern for what the mirror would look with 2x the wavefront that it would really be. For example for  a 6" f/8 mirror with 1/4 wave of under correction, it  would look like a one with 1/2 wave of under correction in DPAC. since the errors are doubled. 

  The bottom line is that if you believe you have 1/10 wave optic  and you test them with DPAC and position the grating so you see line lines they should be very straight and as you move it from one side of focus to other you need to fight to see any tiny wiggle in the lines. If not your don't have true 1/10 wave optics. 

  As I have said many times use as many different tests as you can but don't take the best results as the truth. All the results should agree, if not your doing something wrong and you should find out what.

 

                   - Dave 

[Jeff B]

Thanks Dave.  

 

With some refractors (the older Meade ED doublets for example), I can get coma "on axis", if the elements are not properly centered on each other.  I've a couple of so-so achromats and will look for the ))), which I believe I've seen before.  

 

The thing is, if the elements are decentered, is that also a form of wedge and if so, would I be lumping a bit of astigmatism in with the coma?

 

Jeff 

[Eddgie]

I am a major fan of DPAC testing for showing errors and I would love to see more tests of more scopes.

 

While I am a fan, the reason I have never invested in it and continue to use star testing is that while the DPAC is great at showing errors, I do not really know how to quantify errors when I see them. For example, if there is astigmatism (as discussed earlier) I would not really be able to judge how much astigmatism was present and because of that, it makes it less attractive to me than star testing.

 

Same with spherical aberration. It is pretty easy to get a very good approximation of the amount of SA using the star test, but I am not competent in making that assessment using just the DPAC images.

 

I do love seeing the tests though but it would be even better if we had a tool or an example library that we could refer to for quantification. There are many of these fir star testing.

 

What would be even more exciting would be a piece of modeling software for DPAC like Aberrator 3.0 for the star test. That would be awesome!

 

But that is why I still like star testing. There are tools and examples that make it possible to get a reasonable approximation of the errors and Aberrator 3.0 even does an MTF plot.

 

Love seeing these tests though and hope to see many more!

Edited by Eddgie, 13 March 2020 - 11:26 PM.

[peleuba]

What would be even more exciting would be a piece of modeling software for DPAC like Aberrator 3.0 for the star test. That would be awesome!

 

 

 

If you want an  idea of what different levels of spherical aberration would look like, use a program that simulates Ronchi tests of a spherical of same aperture and F-ratio of the  optics your tests. Tell  the program to show the Ronchi  pattern for what the mirror would look with 2x the wavefront that it would really be. For example for  a 6" f/8 mirror with 1/4 wave of under correction, it  would look like a one with 1/2 wave of under correction in DPAC. since the errors are doubled. 

 

The two programs I use are Diffract and Ronchi for Windows.  Both are free.

[peleuba]

As I have said many times use as many different tests as you can but don't take the best results as the truth. All the results should agree, if not your doing something wrong and you should find out what.

 

 

Dave - other then autocollimation what other tests are you using?  I know on the Criterion corrector re-work, you are using a Newton interferometer of sorts  where you are doing some contact testing between the corrector and a flat.  Do you perform a lot of star testing?  Is there any test you prefer to corroborate the results in DPAC?

 

Also - any comments on my post #5 above as it pertains to wedge.  In my images, wedge is easy to see in the "at focus" star image; not easy so easy to detect in DPAC.  This is why I use both tests.  I also have a couple of interferometers - Bath and a Michleson (variant of Twyman-Green).  But have found Interferometry to be a time suck.  

 

Thanks for participating in this thread.

[Jeff B]

John (precaud here on CNs) taught me how to use a single white light LED (really red, blue and green LEDs in one unit) and then separate the the individual red, green, and blue colors using software, in my case an anchirent version of Paint Shop Pro.  Thank you sensi John    I later discovered that PSP allowed me to also separate the individual colors in monochrome as well.   I find this useful as it eliminates any color funny business from my eyes and the computer screen.  

 

As the color images are all sourced from the same white image, taken at the same discrete position in the light cone (in/out/at focus), this can give me some insight as to where the focus of the individual colors fall relative to each other.   But that's all.  I gave up on trying to figure out if one color's focus fell in or outside that of another color.

 

For example, here are the red, green and blue monochrome images derived from single white light images for inside and outside of focus for an 8" F9 Istar achromat.  Initially, I concluded that the red and green focus positions were really very similar to each other and that the blue focused a bit closer in.  I further concluded that this was different from a classic achromat where the red and blue focus positions are close to each other and are a bit "long" relative green.  Well, directly measuring the focus positions of red, green and blue with individual, color LEDS showed I was wrong on both counts.  In fact, looking at the individual color focus positions, showed it is indeed a classic achromat, with the red and blue close to each other (within ~1/2mm) and both a bit long relative to green (~2mm or so).

 

It still is useful information to me to see focus position differences in DPAC using this technique but I really need to be careful about drawing any conclusion other than they seem a bit different from each other.

 

Jeff

Attached Thumbnails

• 
• 
• 
• 

[peleuba]

I much prefer star testing to get a good handle on astigmatism and, especially, coma (I have no idea what coma looks like in DPAC or if it can be seen at all).   Trouble is, I don't have an indoor star test such as yours.

 

A single pass indoor star test apparatus is somewhat easy to build.  You could use one of your large, high quality, refractors as a "collimator" to test smaller lenses.  You can PM me if you want plans to set it up.

 

Below is a device a friend constructed so I could star test in double pass.  The device utilizes a white LED, a precision pinhole, and beam splitter plus some additional plumbing.  You place it in the focuser of the telescope to test, the light traverses the telescope, hits the flat, returns through the telescope and the output is viewable via the beam splitter in an eyepiece or a camera.  The auxiliary optics in the device need to be of very high quality as any aberration is added directly to the the result.  Its an expensive device, but fun to use.

Attached Thumbnails

• 

Edited by peleuba, 16 March 2020 - 02:10 PM.

[Jeff B]

A single pass indoor star test apparatus is somewhat easy to build.  You could use one of your large, high quality, refractors as a "collimator" to test smaller lenses.  You can PM me if you want plans to set it up.

 

Below is a device a friend constructed so I could star test in double pass.  The device utilizes a white LED, a precision pinhole, and beam splitter plus some additional plumbing.  You place it in the focuser of the telescope to test, the light traverses the telescope, hits the flat, returns through the telescope and the output is viewable via the beam splitter in an eyepiece or a camera.  The auxiliary optics in the device need to be of very high quality as any aberration is added directly to the the result.  Its an expensive device, but fun to use.

Really cool and thanks Paul.  

 

I'm very interested in the double pass device.

 

Jeff

[TG]

I too have thought of using a cheap surplus cube shed beam splitter to view completely on-axis because the usual DPAC mode results in off-axis viewing and vignetting of the pupil in all scopes except refractors. Here are the problems I don't know the answers to:

• In my DPAC tester, the LED is jammed up against the grating. Using a cube beam splitter, to view fully on-axis, there are two possible configurations:
  • Place the grating immediately after the beam splitter and the LED before it. Does the LED need a lens to focus it on the grating or is this not going to matter? Will the beam splitter affect the results?
  • Place the grating+LED before beam splitter as usual. In my simulations with OSLO, the beam splitter does introduce spherical aberration at faster f/ratios so this would be a concern. I'm not sure whether this would also be the case with (1)
• Wouldn't a thin plate glass filter be better than a cube beam splitter? I though using a interferometric filter at a 45° angle would work. Good idea or bad?

[starman876]

A single pass indoor star test apparatus is somewhat easy to build.  You could use one of your large, high quality, refractors as a "collimator" to test smaller lenses.  You can PM me if you want plans to set it up.

 

Below is a device a friend constructed so I could star test in double pass.  The device utilizes a white LED, a precision pinhole, and beam splitter plus some additional plumbing.  You place it in the focuser of the telescope to test, the light traverses the telescope, hits the flat, returns through the telescope and the output is viewable via the beam splitter in an eyepiece or a camera.  The auxiliary optics in the device need to be of very high quality as any aberration is added directly to the the result.  Its an expensive device, but fun to use.

I want one  

[Jeff B]

How about cannibalizing a bino-viewer for its beam splitter?

 

Jeff