Colimation with artificial star

[Davey-T]

Made an artificial star and just wondering how folks do this indoors on a bench when I've just measured my scope and nearest it can focus to is about 15 meters.

Have they got really long workshops or am I missing something here ?

Dave

PS: Too windy outside

[vlaiv]

You can do collimation on defocused star pattern so no need to bring it to focus

On a side note, if you want to do collimation or anything else indoors - you can use another scope to do it. Place artificial star at focal plane of a second telescope and turn that telescope towards the telescope you want to collimate - sort of aperture towards aperture setup.

Second telescope will act as collimation lens for artificial star and it will provide collimated beam - like real star at infinity. You should be able to reach focus with primary scope quite easily in this case.

Mind you, if you want to measure quality of optics of the first scope - you need seriously good second scope that won't produce very aberrated wavefront, or alternatively you need to know wavefront of second / collimation scope and subtract that from result.

[Davey-T]

20 minutes ago, vlaiv said:

You can do collimation on defocused star pattern so no need to bring it to focus

Is it better to be able to go both sides of focus or not necessary ?

Dave

[Ikonnikov]

14 minutes ago, vlaiv said:

You can do collimation on defocused star pattern so no need to bring it to focus

On a side note, if you want to do collimation or anything else indoors - you can use another scope to do it. Place artificial star at focal plane of a second telescope and turn that telescope towards the telescope you want to collimate - sort of aperture towards aperture setup.

Second telescope will act as collimation lens for artificial star and it will provide collimated beam - like real star at infinity. You should be able to reach focus with primary scope quite easily in this case.

Mind you, if you want to measure quality of optics of the first scope - you need seriously good second scope that won't produce very aberrated wavefront, or alternatively you need to know wavefront of second / collimation scope and subtract that from result.

If the reference scope is much larger in aperture than the test scope wouldn't this help to reduce the apparent error (assuming the lager scope was reasonably well collimated)? I seem to remember people on CN using big dobs for this purpose.

[vlaiv]

3 minutes ago, Davey-T said:

Is it better to be able to go both sides of focus or not necessary ?

Dave

Not really necessary to do that as collimation issues will be equally seen both sides of focus - only mirror reversed.

 

3 minutes ago, Ikonnikov said:

If the reference scope is much larger in aperture than the test scope wouldn't this help to reduce the apparent error (assuming the lager scope was reasonably well collimated)? I seem to remember people on CN using big dobs for this purpose.

I don't think so. You certainly need larger or at least the same aperture of "reference" telescope to be able to cover whole aperture of measured telescope for optical quality tests, but wavefront error won't be reduced if one is using only part of the aperture of reference scope - at least not random part.

If you know that you have very good section of mirror and you make sure you use that section of the reference mirror - then yes, it will help. Another way large mirror helps is that same mechanical surface quality translates in less relative wavefront error on large mirror (at least I think so), but large mirrors are harder to make and on average I guess you end up with same level of wavefront error (basically you get what you pay for and if you want very good figure, it's going to cost in large aperture).

[Davey-T]

1 minute ago, vlaiv said:

Not really necessary to do that as collimation issues will be equally seen both sides of focus - only mirror reversed.

Thanks Vlaiv, will give it a try later.

Dave

[Stuart1971]

1 hour ago, Davey-T said:

Made an artificial star and just wondering how folks do this indoors on a bench when I've just measured my scope and nearest it can focus to is about 15 meters.

Have they got really long workshops or am I missing something here ?

Dave

PS: Too windy outside

I thought that using this method you needed to be around 50-100 metres away... to get an accurate result, when they do it in workshops I would not think they use an artificial star method, but maybe a laser set up, that works very closely to the scope, but also very expensive.... 🤔

[johninderby]

Have you seen this article on CN?

https://www.cloudynights.com/articles/cat/user-reviews/the-easy-way-to-align-your-telescopes-optics-indoor-artificial-star-collimation-r2798

Edited June 29, 2020 by johninderby

[Peter Drew]

No real need for anything sophisticated as long as you can reach focus.  There is a big difference between checking collimation and optical quality.  Using a shorter than ideal distance just affects the apparent spherical aberration.  Not suitable for binoculars though due to parallax.   🙂   

[Davey-T]

8 minutes ago, johninderby said:

Have you seen this article on CN?

https://www.cloudynights.com/articles/cat/user-reviews/the-easy-way-to-align-your-telescopes-optics-indoor-artificial-star-collimation-r2798

Haven't got a spare mirror, lol, same problem though wants scope 15 meters away.

Dave

[Davey-T]

1 minute ago, Peter Drew said:

No real need for anything sophisticated as long as you can reach focus. 

House isn't long enough 😁

Dave

[Peter Drew]

You can use extension adaptors as long as you avoid "sag".  I wonder whether having the artificial star next to the telescope and a mirror at the furthest distance you can place it would give you double the distance, the optical quality of the mirror shouldn't be an issue as you are only checking collimation. (Haven't thought this through yet!)  🙂

[inFINNity Deck]

Have you seen my article? It is in Dutch, but should translate when opened in Chrome.

I use a 12" Newton in combination with a 9 micron glass-fibre artificial star. The artificial star sits in the focal point of the Newton. Mind you that when using a second scope, that the secondary obstruction of the reference scope is smaller than that of the scope to be collimated. If you scope has moving mirrors it is important that the scope it collimated as close as possible to infinity focus, this to minimise influence of mirror-shift/mirror-flop. For the rest it is quite easy, mutual distance in my set-up is about 10cm. I also recommend turning the scope over (180° rotation along its optical axis), as some scopes have terrible mirror-flop, see my second article on this (you will be surprised by the manufacturer's response).

Nicolàs

 

 

[andrew s]

24 minutes ago, Peter Drew said:

You can use extension adaptors as long as you avoid "sag".  I wonder whether having the artificial star next to the telescope and a mirror at the furthest distance you can place it would give you double the distance, the optical quality of the mirror shouldn't be an issue as you are only checking collimation. (Haven't thought this through yet!)  🙂

I built an auto collimator using an old 50/50 beam splitter some years ago. By injecting an artificial star in the side beam you could have a plane mirror as near as the scope as you liked and then collimate it. The unit just plugged into the focuser. Edmund optics made something similar for ££££££

Regards Andrew 

[vlaiv]

31 minutes ago, Stuart1971 said:

I thought that using this method you needed to be around 50-100 metres away... to get an accurate result, when they do it in workshops I would not think they use an artificial star method, but maybe a laser set up, that works very closely to the scope, but also very expensive.... 🤔

There are two "points" that you need to worry about when using artificial star.

First point is point where artificial star is resolved. You want angular size of your artificial star to be at least half that of airy disk in order not to be resolved. This distance is not all that big with very small artificial star.

For example - using 150mm aperture telescope and medium sized artificial star - about 20um (most artificial stars are in range of 50 down to 10um), calculation goes like this:

150mm aperture creates 1.7" airy disk, so we want our artificial star to be at a distance so that it presents about 0.8" of angle with its diameter.

http://www.1728.org/angsize.htm

handy calculator for those things, we input values - solve for dist a nce, angle in arc seconds and we will put artificial star diameter in millimeters rather than microns - so 0.02

Solution is 5156mm, or 5.15 meters.

That is fairly close and in general you need to worry about that only if you have larger artificial star and larger aperture (like 50um and 200mm+) then you need 10m+ for resolving criteria.

There is second criteria - and that is amount of spherical aberration we want to introduce or rather not introduce. That is important for optics testing.

Here we can do following: Let's say we want to have less than 1/10 wave spherical due to distance and we work with 150mm aperture. What is the distance we want to place our artificial star?

1/10th of 0.5um (500nm green light) is 0.05um. What angle will give difference in hypotenuse and side of 0.05 and what would be sides length?

Here is a nice image to explain it. Artificial star is at A. B is center of our aperture and C is edge of our aperture. AC edge (hypotenuse) is longer than AB edge (Adjacent side). We know CB - as being half of aperture, or 75mm.

We need AC - AB to be 0.05um long. So first equation is:

AC - AB = 0.05um => AC = 0.05um + AB

Second equation is Pythagoras

AC^2 - AB^2 = CB^2 => AC^2 -AB^2 = 5625mm

(0.05um + AB)^2 - AB^2 = 5625mm

0.0025um + 0.1um * AB + AB^2 -AB^2 = 5625mm => 0.1um*AB = 5625000um - 0.0025um

AB = (5625000 - 0.0025) / 0.1 = 56249999.975um = ~562450mm = ~56.2m

We need at least 56.2m of distance for 150mm telescope to have 1/10 spherical from artificial star.

 

[Davey-T]

38 minutes ago, vlaiv said:

We need at least 56.2m of distance for 150mm telescope to have 1/10 spherical from artificial star.

Will have to take it to the end of the garden 😁

My artificial star is a 5mm ball bearing because I had some lying around, difficult to calculate the diameter of the reflected torch light.

Dave

[Peter Drew]

42 minutes ago, andrew s said:

I built an auto collimator using an old 50/50 beam splitter some years ago. By injecting an artificial star in the side beam you could have a plane mirror as near as the scope as you liked and then collimate it. The unit just plugged into the focuser. Edmund optics made something similar for ££££££

Regards Andrew 

I also used to have a bench collimator when I was in business, that's why I raised the query.  Mine was a small ball bearing illuminated by a collimated light source and the resultant artificial star viewed at its image in a 12" F6 paraboloid.   🙂   

[vlaiv]

4 minutes ago, Davey-T said:

Will have to take it to the end of the garden 😁

My artificial star is a 5mm ball bearing because I had some lying around, difficult to calculate the diameter of the reflected torch light.

Dave

Not really, or rather - it depends on measurement method used.

For example Roddier test (WinRoddier implementation) allows you to select if you are using artificial or real star and in case of artificial star - distance to telescope. It will automatically remove spherical term produced with close target so you can have even 1/4 or 1/2 wave of spherical and it will be accounted for.

On the other hand - that much spherical with other tests will mask other aberrations so you won't be able to tell how good optics is.

Any test that produces Zernike polynomials of your optics allows for removal of appropriate spherical term - you just calculate spherical magnitude like above given aperture size, wavelength and distance and you remove it from appropriate Zernike term and then synthesize wavefront again.

This is common practice - and you will often see some terms remove in optics reports - like coma from newtonian parabolic mirrors since it is inherited in design and will be 0 at optical axis (but can have small value if test was done at an angle and is thus removed).