Jump to content


Photo

Star Test: Definitive check for collimation?

  • Please log in to reply
80 replies to this topic

#1 GregBuchholz

GregBuchholz

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 13 Feb 2012
  • Loc: Kennewick, WA

Posted 22 March 2012 - 09:38 PM

Am I correct in thinking that the star test is the definitive check for collimation? That is, you crank up the power to 40-50x the aperture in inches, point to a star, defocus a little, and make sure you get 5 or so perfectly concentric diffraction rings. (Also described in chapter 6 of Suiter's Star Testing Astronomical Telescopes). Not to be confused with the defocus-on-a-star-until-you-see-the-secondary-shadow-blob-approximatly-centered test.

If the star test shows concentric circles, then there would be no further geometric adjustments to primary, secondary, focuser, etc. that could improve the image. Likewise, if you get the perfectly concentric diffraction rings, then subsequently using a cheshire, autocollimator, barlowed laser, etc. will all show perfect optical collimation (if not physical centering of the components).

Let's ignore issues of atmospheric seeing which might preclude you from performing the test on an actual star.

Does that sound correct?

#2 Mirzam

Mirzam

    Skylab

  • *****
  • Posts: 4447
  • Joined: 01 Apr 2008
  • Loc: Lovettsville, VA

Posted 22 March 2012 - 09:43 PM

The in-focus star showing a single symmetrical ring of uniform intensity all around is the best test, seeing permitting.

JimC

#3 GlennLeDrew

GlennLeDrew

    Voyager 1

  • *****
  • Posts: 10671
  • Joined: 17 Jun 2008
  • Loc: Ottawa, Ontario, Canada

Posted 22 March 2012 - 10:04 PM

For visual observing I'd go no further than the star test. But for imaging I would tweak for radial symmetry across the full field.

#4 john gabriel

john gabriel

    Ranger 4

  • *****
  • Posts: 343
  • Joined: 13 Dec 2011
  • Loc: Southern California

Posted 22 March 2012 - 10:09 PM

.....point to a star, defocus a little, and make sure you get 5 or so perfectly concentric diffraction rings. ...

Hi guys. If it has good optics, can a 6" F/8 newt be collimated to attain such a concentric ring figure? When I defocus, I get the secondary silhouette in center, but can see no concentric rings like Greg's link shows. Thank you.
John

#5 GregBuchholz

GregBuchholz

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 13 Feb 2012
  • Loc: Kennewick, WA

Posted 22 March 2012 - 11:10 PM

For visual observing I'd go no further than the star test. But for imaging I would tweak for radial symmetry across the full field.


Hmm. What, or how does one tweak for radial symmetry? Anyone have a link to share?

#6 GregBuchholz

GregBuchholz

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 13 Feb 2012
  • Loc: Kennewick, WA

Posted 22 March 2012 - 11:32 PM

.....point to a star, defocus a little, and make sure you get 5 or so perfectly concentric diffraction rings. ...

Hi guys. If it has good optics, can a 6" F/8 newt be collimated to attain such a concentric ring figure? When I defocus, I get the secondary silhouette in center, but can see no concentric rings like Greg's link shows. Thank you.
John

Yes you should be able to see the diffraction rings. You defocus ever so slightly to see them. Seeing the secondary's shadow means you've gone too far out of focus. Also, you need fairly high magnification. For checking and adjusting collimation, it is easier to see errors on the outside-of-focus, compared to inside-of-focus. And crummy seeing makes the rings jiggle about. Lately, I've been playing around with the Hubble artificial star, and generally like it. With an artificial star you can test things regardless of atmospheric seeing or clouds.

#7 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6652
  • Joined: 21 Oct 2006
  • Loc: California

Posted 22 March 2012 - 11:57 PM

For visual observing I'd go no further than the star test. But for imaging I would tweak for radial symmetry across the full field.


Hmm. What, or how does one tweak for radial symmetry? Anyone have a link to share?


Radial symmetry requires the adjustment of the secondary mirror -- not an easy task with a star test. Use the proper collimation tools.

Radial symmetry means the focal planes of the eyepiece and the primary mirror are parallel.

#8 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6652
  • Joined: 21 Oct 2006
  • Loc: California

Posted 22 March 2012 - 11:59 PM

For visual observing I'd go no further than the star test. But for imaging I would tweak for radial symmetry across the full field.


With a Paracorr or/and with a wide FOV quality EP, radial symmetry is needed for visual observation.

#9 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6652
  • Joined: 21 Oct 2006
  • Loc: California

Posted 23 March 2012 - 12:03 AM

An example of a star test. Defocus by a little amount at high magnification. It is best if you do not see the diffraction pattern of the spider vanes.

Attached Files



#10 cheapersleeper

cheapersleeper

    Gemini

  • *****
  • Posts: 3273
  • Joined: 22 Jan 2010
  • Loc: Sachse TX

Posted 23 March 2012 - 12:30 AM

In general, how "quiet" do the little wiggly spikey things have to be to allow a decent test? With my newer bigger, thickish mirror, there's a whole lot of shaking going on, with regard to my out of focus star, not my mount, but the star.

#11 FirstSight

FirstSight

    Duke of Deneb

  • *****
  • Moderators
  • Posts: 9850
  • Joined: 26 Dec 2005
  • Loc: Raleigh, NC

Posted 23 March 2012 - 01:03 AM

IF you have an accurately placed center spot
AND you have high-quality collimation tools, such as the Catseye sight tube, cheshire, and autocollimator
THEN star testing doesn't really improve on the results, but merely at best confirms them. Also, how well and easily star testing can be done is atmospheric-condition dependent, whereas collimation with quality tools is not so at all.

What I find a quick star test in the middle of an observing session to be valuable for is as a quick and easy indicator for whether my collimation is starting to drift off as the night goes on. But if so, IMHO it's still easier to fix it accurately by bringing out the quality collimation tools for a tune-up. YMMV depending on what collimation tools you have available.

#12 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1598
  • Joined: 26 Jul 2004

Posted 23 March 2012 - 04:31 AM

Am I correct in thinking that the star test is the definitive check for collimation? That is, you crank up the power to 40-50x the aperture in inches, point to a star, defocus a little, and make sure you get 5 or so perfectly concentric diffraction rings.



I would say it is the definitive check to see that the marker on the primary actually corresponds to the optical center (the center of the paraboloid). I have not heard of an instance where anyone has shown a significant offset on an accurately round, machine-polished mirror, though (but many stories of a spot placed significantly off the mechanical center of the mirror - usually by the manufacturer!). But I believe such a test, if done at all, should be done carefully by star collimation, followed by a check of any decentering (as seen with any tool showing the collimation of the primary only), and repeated to see if any deviation is consistent - sacrificing a night of unusually good seeing.
But once you know that the marker is true, simple tool collimation to it will no doubt give a quicker and more accurate routine collimation.

And keep in mind that this is just a check of the primary and its axis, not of the focuser axis.

Nils Olof

#13 Sarkikos

Sarkikos

    Hubble

  • *****
  • Posts: 16741
  • Joined: 18 Dec 2007
  • Loc: Per sylvam ad astra

Posted 23 March 2012 - 07:19 AM

Nils,

Is is possible that a star test would not be able to distinquish between a merely good collimation and a very close collimation? Would the star test be able to tell, for instance, that the telescope has been collimated to the extra precision produced by a 2-Pupil Infinity XLK Autocollimator in the hands of a skilled user?

By the way, I never thought of star testing as an efficient or particularly good way to collimate a telescope, or to test the collimation. Star testing is a good way to test the telescope's optics. Maybe that's why it's called star testing? :thinking:

:grin:
Mike

#14 ahlberto

ahlberto

    Apollo

  • *****
  • Posts: 1348
  • Joined: 06 Sep 2008
  • Loc: lisbon-portugal

Posted 23 March 2012 - 07:47 AM

I dont know if it is a definitive test but i usualy check the star test with a 4mm UO ortho.I have a SW 8" F5 and the center spot on mi mirror was of for about 1 mm so wen i colimated with the laser and check with the cap things dont matched... :p I recentered and now i can achieve a perpect collimation.I have to say this is the second blue 8" tube and the optics are so amazing good...You put a rochi tester and...perfect..no aberrations,no turned edge..perfect...you pull a star test with a 4mm and the in/out rings are text book perfect.The coatings are perfect without any pin holes...Sorry for the of topic guys :jump:

#15 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6652
  • Joined: 21 Oct 2006
  • Loc: California

Posted 23 March 2012 - 08:00 AM

Mike, I did the following experiment -- anyone can try it:

1) I collimated to the best of my ability using my Catseye tools
2) I decollimated the primary mirror by twisting only one of the primary knobs clockwise. I gave it a generous twist.
3) I recollimated using a star by slowly twisting the same knob counter-clockwise until the star test indicated good collimation. I wanted to get the edge of what would be considered good star collimation. Again, do it slowly.
4) I compared the result with the cheshire (Blackcat). It was little off.
5) I repeated the above but by twisting the knob in the opposite direction. Again, it was off in the opposite direction.

Conclusions:
1) A quality collimation tools will get you to the center of the “good” collimation range but star collimation is a hit&miss somewhere in that range.
2) If a collimation tool disagrees with the star collimation, it does not necessarily means the collimation tool is at fault.

Saying it differently, star testing should agree with the results of quality collimation tools but quality collimation tools might not 100% agree with star testing.

Personally, I believe star collimation is practical to confirm good collimation but is impractical to use to collimate. Then again, that is my personal opinion. Not to mention it can’t be used to adjust the focuser axis.

Jason

#16 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6652
  • Joined: 21 Oct 2006
  • Loc: California

Posted 23 March 2012 - 08:09 AM

The first page of this thread has good insight
http://www.cloudynig...7/o/all/fpart/1

#17 Sarkikos

Sarkikos

    Hubble

  • *****
  • Posts: 16741
  • Joined: 18 Dec 2007
  • Loc: Per sylvam ad astra

Posted 23 March 2012 - 08:14 AM

Jason,

Saying it differently, star testing should agree with the results of quality collimation tools but quality collimation tools might not 100% agree with star testing.

Personally, I believe star collimation is practical to confirm good collimation but is impractical to use to collimate. Then again, that is my personal opinion. Not to mention it can’t be used to adjust the focuser axis.


I'm not surprised by the results of your experiments or your conclusions.

Mike

#18 howard929

howard929

    Member

  • *****
  • Posts: 3515
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 23 March 2012 - 08:46 AM

2) I decollimated the primary mirror by twisting only one of the primary knobs clockwise. I gave it a generous twist.
3) I recollimated using a star by slowly twisting the same knob counter-clockwise until the star test indicated good collimation. I wanted to get the edge of what would be considered good star collimation. Again, do it slowly.


At that point in the procedure, did the star appear to be as sharply defined at focus as it was after you completed the rest of the process with the cheshire?

Howard

#19 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1598
  • Joined: 26 Jul 2004

Posted 23 March 2012 - 09:13 AM

Mike,

Is is possible that a star test would not be able to distinquish between a merely good collimation and a very close collimation?



Depending on the size of the mirror, and the seeing (also related to the size of the mirror - good seeing for a small mirror may not be goood for a large one), I'd say. Below is a (simplistic) simulation of coma, in and out of focus with faultless seeing (from my collimation page). The difference between merely good and very close is in the asymmetry of rings only visible (seeing permitting) quite near focus.
Star testing can be used to test many aspects of a telescope's optics, including the centering of the primary's optical axis (= collimation of the primary).

By the way, I never thought of star testing as an efficient or particularly good way to collimate a telescope, or to test the collimation.

I fully agree with you here. But looking through the archives, you can find many who recommend it as the ultimate, or at least final, step of routine collimation.

Would the star test be able to tell, for instance, that the telescope has been collimated to the extra precision produced by a 2-Pupil Infinity XLK Autocollimator in the hands of a skilled user?



The precision of the primary's collimation is essentially determined by the BlackCat, not the XLK. Despite wishful thinking - the extra precision is with the focuser axis which is less demanding anyway. But the Blackcat + hotspot allows very precise collimation - Jason has posted some good pics showing this. After collimation with quality tools of whatever type and make, and a correctly placed center marker, it should be difficult or impossible to show an imperfection on the star test, I trust.

The first image pair is perfect, the third with a 1/14 wave RMS error ("quarter wave"), the 2nd midway, the 4th: 2.5 times the error of #3.

Nils Olof

Attached Files



#20 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6652
  • Joined: 21 Oct 2006
  • Loc: California

Posted 23 March 2012 - 09:16 AM

At that point in the procedure, did the star appear to be as sharply defined at focus as it was after you completed the rest of the process with the cheshire?

Howard


yes

#21 GregBuchholz

GregBuchholz

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 13 Feb 2012
  • Loc: Kennewick, WA

Posted 23 March 2012 - 11:09 PM

Radial symmetry means the focal planes of the eyepiece and the primary mirror are parallel.


I'm assuming that non-parallel focal planes means that for instance if you focus on an object on the left side of the view, the right side will be out of focus? Why wouldn't this be important for visual observing? Or is it the case that long exposure imaging is more sensitive to focus issues?

#22 GregBuchholz

GregBuchholz

    Explorer 1

  • -----
  • Posts: 58
  • Joined: 13 Feb 2012
  • Loc: Kennewick, WA

Posted 23 March 2012 - 11:16 PM

Mike, I did the following experiment -- anyone can try it:

1) I collimated to the best of my ability using my Catseye tools
2) I decollimated the primary mirror by twisting only one of the primary knobs clockwise. I gave it a generous twist.
3) I recollimated using a star by slowly twisting the same knob counter-clockwise until the star test indicated good collimation. I wanted to get the edge of what would be considered good star collimation. Again, do it slowly.
4) I compared the result with the cheshire (Blackcat). It was little off.
5) I repeated the above but by twisting the knob in the opposite direction. Again, it was off in the opposite direction.


It sounds like there is a range of primary adjustment knob movement for which the star test appears to give the same result. Could you quantify about how much knob movement that is for your particular setup? Is that an eighth of a turn for example?

Thanks.

#23 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1598
  • Joined: 26 Jul 2004

Posted 24 March 2012 - 04:55 AM

It sounds like there is a range of primary adjustment knob movement for which the star test appears to give the same result. Could you quantify about how much knob movement that is for your particular setup? Is that an eighth of a turn for example?



It would say more to have an idea of how much "a little off" could be in terms of distance, as this in turn would give an idea of how much coma would be left after star collimation.

Nils Olof

#24 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1598
  • Joined: 26 Jul 2004

Posted 24 March 2012 - 05:08 AM

I'm assuming that non-parallel focal planes means that for instance if you focus on an object on the left side of the view, the right side will be out of focus? Why wouldn't this be important for visual observing? Or is it the case that long exposure imaging is more sensitive to focus issues?


Comparing defocus on one side of the field to the opposite is not easy - moving your gaze all the way, you might re-focus unconsciously (at least if you are much younger than I). And with moderate tilt (1/30 of diameter), defocus will be efficiently hidden in much larger coma. With a coma corrector, as you will use for wide field imaging, much of the coma will be eliminated, and any defocus error is captured in the image.

Anyway, I still wait to hear anybody actually reporting a visual problem of this nature...

Nils Olof

#25 howard929

howard929

    Member

  • *****
  • Posts: 3515
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 24 March 2012 - 08:54 AM

4) I compared the result with the cheshire (Blackcat). It was little off.




That's a curious result considering that collimation with a star test is independent of the location of the primary mirrors center spot (as Nils pointed out earlier) while it's the focus of the cheshire. Might it be possible that with repeated results of that test procedure that it indicates a misplaced mirror donut/marking instead of a more exacting collimation??

Howard






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics