25 replies to this topic

[WxObserver]

I tried to make an artificial star by poking a hole in aluminum foil with a needle. I'll skip the details, but I know the hole is less than 75 microns, and maybe less than 50 microns. It is illuminated from behind by a HeNe laser (632nm). 

 

Below is the defocused image it produces from a distance of 40 feet through an Askar SQA70 (336mm focal length) on an ASI294mm mono camera. The image is enlarged 2x for clarity, so at this scale, one pixel is about 1.15um.

 

What can be deduced from this? Is there something wrong with the artificial star? Was it a mistake to use a laser for illumination? Does this reveal anything good or bad about the telescope?

 

Kind of new to all this and not sure what to make of this. Wasn't expecting all the rings there.

 

[maniack]

What were you expecting? This looks pretty picture perfect to me.

[WxObserver]

Well, I wasn't expecting all the rings, just one ring for each optical surface and maybe not concentric, but then I don't think I fully understand how the test works either. Can you point me to some learning on this?

[RichA]

Well, I wasn't expecting all the rings, just one ring for each optical surface and maybe not concentric, but then I don't think I fully understand how the test works either. Can you point me to some learning on this?

When you are inside or outside focus, you get multiple rings which increase in number the further from focus you get.  When you are in focus, you get a diffraction disk and a ring or two, or more, depending on the illumination level.  

Edited by RichA, 21 June 2025 - 11:31 PM.

[maurycy]

Getting rings is normal. If anything, it shows that your pinhole is small enough to be used as an artificial star. As for the telescope itself, that looks just fine, although you should try examining the pattern closer to focus, which will make subtle problems more apparent. 

[WxObserver]

Thanks maurycy -- I'll try closer to focus in the next day or two and see what happens.

[davidc135]

The quality of the pinole wouldn't matter so long as it is 75 micron or smaller. It needs to be the equivalent of half the Airy disc diameter or less at the distance of 40'.

 

An f4.6 optic has a 6 micron Airy disc roughly, so a minimum of 3 microns. I divide 40' by the f.l of the scope to get a factor of 36 and divide that into 75 to get approx 2 microns.

 

It's usual to take images of the artificial star at exactly equal distances either side of focus and compare them. At focus and say 5 and 10 waves defocus where 1 wave defocus in an f4.8 optic is 0.10mm.

 

40feet is a bit too close and the scope should show significant under-correction, maybe even a 1/4 wave, as a result. If you can, increase it to 150feet. 

 

A 1/3 diameter paper disc obstruction can be suspended in front of the scope and the sizes of the shadows compared at +/- ten waves defocus. If there is under-correction, the shadow should be noticeably larger inside focus with the artificial star close at 40'.

 

With some visual scopes the outer rays of the focused beam might clip the baffles when the source is close, but not likely in an OTA made for imaging.

 

David

Edited by davidc135, 22 June 2025 - 02:56 AM.

[Eddgie]

 

 

Kind of new to all this and not sure what to make of this. Wasn't expecting all the rings there.

 

Defocused-Star-Crop-2x.jpg

As others have mentioned, the rings are to be expected.

 

As for making anything of the test, there really isn't anything that can be determined from seeing this pattern because even a poor instrument can pass when there is too much defocus. The more you  defocus, the more sensitivity the test loses, except for the case of zonal errors, which become easier to see with more than optimal defocus.

 

Optimal defocus is 10 waves of light, but you need to make the comparison the exact same amount of focus both inside and outside of the point of best focus. 

 

At f/4.8, the amount of defocus required is 1.01mm in side, and 1.01mm outside.  (Note. If one does not have a 1.01mm feeler gauge, 1mm is fine. The most important thing is that you are exactly the same amount inside and outside. 1mm is more than close enough. 

 

With refractors, color mixing can be an issue, so for testing, it is best to use a green filter, or using a 33% obstruction.

 

I have attached three simulations. 

 

Perfect f/4.8 aperture with no obstruction, 1.01mm of defocus in side and outside of best focus

 

 

F/4.8 with .25 wave of spherical aberration, 1.01mm out of focus inside and outside. Notice that the pattern is larger on one side than the other. This is a signature of spherical aberration but it is not easy to determine the exact amount, as the patterns are both fairly large, and the eye would have trouble telling them apart, but if you can image them, it is easier to see the difference in size.

 

 

f/4.8, .25 wave with 335 obstruction and .25 wave of spherical aberration. Note that not only is the pattern different in size, but but it is much easier to see that the shadow of the secondary obstruction is much larger on one side. If you could accurately measure them, and this is not easy to do, you would find that the obstruction was 25% larger on one side than the other. This would be how you measure the spherical aberration.  You are looking at the difference in the shadow size. With very small amounts of SA, it can be difficult to tell the difference, but with modeling, you can get close.

 

 

You can model using the free software Aberrator 3.0, which is a free download. It is easy to use. 

 

I would repeat this. Color mixing in refractors can make the outer ring look fatter on one side than the other, and this can make it hard to get a good result. Again, the solution is to use a green filter or red filter is the scope is designed for imaging. The secondary obstruction can also be used and is better than no obstruction, but it can be slightly affected by color mixing as well.  

 

Star testing is very powerful, but one should read up on how to do it. 

 

Hope this helps.  

Edited by Eddgie, 22 June 2025 - 08:16 AM.

[WxObserver]

Thanks for all the input 

 

I will try the experiments Eddgie suggested soon. 40 feet is about the best I can do at this time, unfortunately.

 

Below is another image that was taken with 383um of defocus. This is fairly precise because I've calibrated the EAF movements using a dial caliper. This was taken before I read the posts by davidc and Eddgie. This one is enlarged 4x so about 0.58um/pixel. Is the non-uniform illumination from bottom-left to top-right a concern?

 

Because of the EAF calibration, it's easy for me to take exposures at specific amounts of de-focus, with an accuracy of 25um or better.

 

How do you calculate the de-focus distance -- was that based on a specific wavelength? (remember, I'm using a 632nm HeNe laser as a light source so this is a very narrowband measurement)

 

    "Star testing is very powerful, but one should read up on how to do it."

 

That's partly what I was asking -- where can I read up on this? Do you have any links to web pages, PDFs, videos, etc?

 

 

 

[Eddgie]

 

 

 

How do you calculate the de-focus distance -- was that based on a specific wavelength? (remember, I'm using a 632nm HeNe laser as a light source so this is a very narrowband measurement)

 

    "Star testing is very powerful, but one should read up on how to do it."

 

That's partly what I was asking -- where can I read up on this? Do you have any links to web pages, PDFs, videos, etc?

 

 

Artificial-Star.jpg

The defocus is a function of focal ratio. 

 

I use the program Aberrator 3.0 to calculate it. This is a free program and easy to use. You put in an aperture, obstruction (if any) and the focal ratio, and "Focus (Wave)" to 10.  There is a formula, but this is just much easier for me. The program won't do scopes less than 100mm of aperture, but again, defocus is a function of focal ratio so you can put in any aperture you wish. For this measurement, you just need the focal ratio

 

Richard Suiter has written a very good book on Star Testing. It is far more than a book just about star testing, and I have repeatedly recommended that anyone that is involved in amateur astronomy read it. 

 

You may be able to get it through your library system. 

 

https://shopatsky.co...pes-2nd-edition

Edited by Eddgie, 22 June 2025 - 01:52 PM.

[WxObserver]

"You may be able to get it through your library system."

 

It's only 35 bucks -- peanuts compared to all the $$ I've spent on this hobby. I'll just buy it.  

[davidc135]

Defocus distance varies in proportion to the square of the f ratio and, I'm sure, directly with wavelength. More important, as mentioned, is having both intra and extra focus distances the same. Figures given are for 550nm.

 

The image at 383 microns or 3 waves looks v. good. Is that inside? The slight brightness bias may be temperature effects or perhaps a touch of coma. Can you get an image at focus?

 

If you have an ep that provides a high enough magnification, just racking +/- a few waves through focus visually will give a useful idea of the correction. Depending on whether it's under or over-corrected, one side of focus will appear more diffuse with the rings less clearly defined than on the other. But maybe it's not recommendable using a laser source.

 

At fast f ratios, eps and Barlow lenses may contribute to spherical aberration, apart from any resulting from the test setup.

 

David

 

PS The correction also may naturally be slightly under in red light compared to green, but it's not much.

Edited by davidc135, 22 June 2025 - 03:15 PM.

[WxObserver]

Okay -- ran the test with 1.01mm out and in focus. The outermost dark ring is 17% smaller for the image at 1.01mm out focus (image sensor 1.01mm farther away from the focus point) -- compared to the image 1.01mm inwards from focus.

 

Is this an indication that the distance from the artificial star is too close, or a problem with the telescope...or is that unknowable at this point?

[WxObserver]

Defocus distance varies in proportion to the square of the f ratio and, I'm sure, directly with wavelength. More important, as mentioned, is having both intra and extra focus distances the same. Figures given are for 550nm.

 

The image at 383 microns or 3 waves looks v. good. Is that inside? The slight brightness bias may be temperature effects or perhaps a touch of coma. Can you get an image at focus?

 

If you have an ep that provides a high enough magnification, just racking +/- a few waves through focus visually will give a useful idea of the correction. Depending on whether it's under or over-corrected, one side of focus will appear more diffuse with the rings less clearly defined than on the other. But maybe it's not recommendable using a laser source.

 

At fast f ratios, eps and Barlow lenses may contribute to spherical aberration, apart from any resulting from the test setup.

 

David

Yes -- that 383um image is inside the focus point. This is imaging only with ASI294MM mono camera -- I don't have any way to do visual imaging. Will get an focused image too...

[davidc135]

Okay -- ran the test with 1.01mm out and in focus. The outermost dark ring is 17% smaller for the image at 1.01mm out focus (image sensor 1.01mm farther away from the focus point) -- compared to the image 1.01mm inwards from focus.

 

Is this an indication that the distance from the artificial star is too close, or a problem with the telescope...or is that unknowable at this point?

Can the inside and out images be shown side by side? I would have expected the outer defocused image, judged by the outer bright ring, to be larger.

 

David

Edited by davidc135, 22 June 2025 - 03:16 PM.

[WxObserver]

Sure.

 

Also below is the focused image, enlarged 4x using nearest-neighbor to keep the pixelation obvious. It's also stretched a bit so the intensity drops off faster than this in the linear image. It's fairly under-sampled I think, so not sure how much you can deduce from it...?

 

 

 

[davidc135]

It's hard to understand the at focus image but the in/out pair are text book illustrations of under-correction other than the out is smaller than seen in post 8. Are you sure of the exact position of best focus? If the two star images were the same, I'd be worried.

 

David

[WxObserver]

Defocus distance varies in proportion to the square of the f ratio and, I'm sure, directly with wavelength. More important, as mentioned, is having both intra and extra focus distances the same. Figures given are for 550nm.

 

The image at 383 microns or 3 waves looks v. good. Is that inside? The slight brightness bias may be temperature effects or perhaps a touch of coma. Can you get an image at focus?

 

Yes -- that 383 microns is inside. I notice that the brightness bias switches direction on in-vs-out focus images. Is that a useful clue?

 

Trying not to get sidetracked here, but I've noticed that there's a small amount of stellar eccentricity when seeing is extremely good -- with an aspect ratio on the elliptical stars of maybe only 1.10:1 or 1.15:1. When seeing is worse, that seems to swamp out the elliptical nature.

 

I tried running DynamicPSF in PI on the focused stellar image and it came up with an elliptical aspect ratio of 1.10:1. This is indoors, so there obviously cannot be any issues with seeing or guiding.

 

Could this brightness variation indeed be coma that could explain that? That would answer one big question I've been pondering.

[WxObserver]

I'll double check the focus position and re-take the two in/out images.

 

P.S. and just FYI.

 

1) When in focus, the exposure has to be dropped to 100usec to avoid over-exposing the central pixels. That's a 1mW laser IIRC.

 

2) When making the pinhole, the needle puncture was compared to a piece of AWG 40 magnet wire (89um dia) under a stereo microscope. The hole was close to that size, and the foil wasn't flat at the hole -- the needle had deformed the foil vertically. So, I put the foil between two small pieces of polished marble and rubbed a bit. After that, the pinhole was no longer discernable under the microscope but light was getting through. This might be a generally useful technique (or not if I just got lucky).

[WxObserver]

Wow, what a difference 100um makes! I re-checked the focus position and took two more images in and out by 1.01um. Now the two outer rings are almost exactly the same diameter.

 

So in the end, other than some possible coma, does this indicate this telescope is very well collimated? 

 

[davidc135]

IIRC, assymetry due to coma doesn't change sides re inside v outside focus but there's another level of complexity in understanding Petzval designs and the subtleties of their performance, compared to doublets or triplets.

 

It would be worth a forensic, visual examination in white light with a high power eyepiece to see what's happening in that suspicious, at focus image. It does suggest coma.

 

David

Edited by davidc135, 22 June 2025 - 04:51 PM.

[davidc135]

It may be that a slight decentration in the lens assembly could cause the brightness bias above and the 10% irregularity in the star images but I'm guessing. Elliptical stars nearer the edge could be caused by the focal surface not being completely flat.

 

David

Edited by davidc135, 22 June 2025 - 05:04 PM.

[WxObserver]

The elliptical pattern usually pretty constant across the image. 

Eccentricity 0.4 is a 1.09:1 aspect ratio and 0.45 is 1.12:1 aspect

 

Edited by WxObserver, 22 June 2025 - 05:25 PM.

[WxObserver]

I went back and capture several de-focused images -- 590um intra-focal -- at center and four corners of the image. Note this is a smaller sensor -- 13x19mm so this doesn't show what's going on at the edges of a full-frame sensor.

 

Result below -- this is a 2x enlargement, so 1.15um per pixel. This is pretty obvious coma, isn't it? Although it doesn't appear quite as bad in the upper-left corner.

 

Still not sure what to make of the non-uniform illumination of the patterns.

 

 

 

[Eddgie]

This is a fast instrument, so any tilt in the system is going to have some difficult to predict affects on the off axis performance. 

 

The first thing I always recommend when trying to diagnose any kind of aberration in a refractor is to ensure that the focuser is not tilted and the sensor is perfectly square to the focal plane. The faster the system, the more impact even tiny amounts of tilt can have. 

To be honest, I am not sure exactly  what the issue is, but the best place to start is ensuring perfect alignment of all components.  Again, the faster the instrument, the more pressure on the mechanics of the system.