Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Concise thread about autocollimators+improvements

This topic has been archived. This means that you cannot reply to this topic.
434 replies to this topic

#276 Gianluca67

Gianluca67

    Viking 1

  • -----
  • Posts: 629
  • Joined: 26 Nov 2008

Posted 22 January 2011 - 04:53 PM

I also find the Catseye acetate template a bit difficult to use to EXACTLY center the mark on the mirror for the same reasons you pointed out. I have an f/4 mirror (8") and I was able to place the triangle around 0.35 mm from the true center regardless the help of another person. Now I have ordered the hotspot to replace the triangle and I would be interested in learning a better technique to be able to place the hotspot within 0.1 mm from the true center of the mirror.

#277 Nils Olof Carlin

Nils Olof Carlin

    Vanguard

  • *****
  • In Memoriam
  • Posts: 2,227
  • Joined: 26 Jul 2004

Posted 23 January 2011 - 07:06 AM

Don,

I like to use two crossed plastic rulers (any strip of thin plastic should do) to put a center spot. Measure the mirror diameter and divide by 2 - attach the spot at the center and mark one radius from the center of the spot along all 4 arms. Maybe you could try to glue some small blocks just outside the radius marks - this should totally prevent sliding when you press down the center (I hope).

BTW if you have a Cheshire that does not quite match your center spot, you could put a "field stop" inside the Cheshire tube, with a diameter just a little larger than the spot (that needn't be illuminated). That way, you will see a thin bright circle, and you can easily tell if it is not uniformly illuminated. That's one way I like, but there are other good ways.

Nils Olof

#278 CatseyeMan

CatseyeMan

    Vendor (Cats Eye Collimation)

  • *****
  • Vendors
  • Posts: 568
  • Joined: 16 Dec 2004

Posted 23 January 2011 - 09:08 AM

Yes, the HotSpot is available in reflective white. In fact, the other colour is yellow, not red.


The reflective "white" HotSpots (you and a few others received on PM inquiry) are from a limited protoytype run and are not currently offered on the web site; however, I expect a "production" batch of these to be ready within the next 2 weeks and will be adding these to the spot selection drop-down option list for ordering on the site; I'll make an announcement in the vendor's forum when they become available.

#279 CatseyeMan

CatseyeMan

    Vendor (Cats Eye Collimation)

  • *****
  • Vendors
  • Posts: 568
  • Joined: 16 Dec 2004

Posted 23 January 2011 - 09:30 AM

I'm ready to buy the new CAM. Any idea when it comes out?

Keep an eye on the Vendor's forum for availability of the XLK-C soon.

#280 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 47,254
  • Joined: 23 Jun 2003

Posted 23 January 2011 - 11:47 AM

Don,

I like to use two crossed plastic rulers (any strip of thin plastic should do) to put a center spot. Measure the mirror diameter and divide by 2 - attach the spot at the center and mark one radius from the center of the spot along all 4 arms. Maybe you could try to glue some small blocks just outside the radius marks - this should totally prevent sliding when you press down the center (I hope).

BTW if you have a Cheshire that does not quite match your center spot, you could put a "field stop" inside the Cheshire tube, with a diameter just a little larger than the spot (that needn't be illuminated). That way, you will see a thin bright circle, and you can easily tell if it is not uniformly illuminated. That's one way I like, but there are other good ways.

Nils Olof

The acetate template has to slide a bit to allow the center to be depressed into the mirror. Even when I've been successful at installing the centermark, the rings on the outside of the template do move in toward the mirror when the center of the template is depressed to the mirror. The rings on the template I have move from slightly outside the reflective area of the primary inward to just inside the reflective rim when I depress the center of the template.
And it is almost never an equal inward movement on all sides. Most of the time, one side of the template moves inward more than the other, leading to the positioning of the centermark away from the dead center of the mirror by a millimeter or so.
That this may be in acceptable tolerances for an f/5 mirror does not ease my qualms about the inaccuracy of placing the centermark on the mirror.
I dream of an f/3 scope with larger aperture, and tolerances are considerably tighter on that short a scope.

What I would like to see is for someone to come up with a sure-fire, repeatable, method for installing a centermark to an accuracy of 0.1mm or better. So far, I know of no technique that is more accurate than about 1mm or maybe a little better by luck. The Hotspot and the XLK autocollimator really demand it, and my fuss-budgety quest for perfection will keep me from resting until I find the method.

I discussed this with a friend who is a very experienced mirror maker, and he took the pragmatic approach of "why go more accurate than is necessary to produce images commensurate with the quality of the optics?". I understand that point of view, but it's just like collimation: collimating to the center of the acceptable envelope means mechanical movements during the night will not take collimation (assuming the scope is stiff enough) outside the envelope of "diffraction-limited Strehl ratio".
If the centermark is off-center, even the most perfect collimation, using known tools, will not attain the goal of putting collimation dead center in the "envelope of tolerances".

I am pragmatic about mechanical sag and flexures and movements. I've discovered the hard way that these are always going to be there--even if they are only related to differential cooling of sub assemblies. So the best we can do is to at least eliminate the sources of error we can control, and the perfect positioning of a mirror's centermark seems to be a reasonable place to start.

#281 lightyear44

lightyear44

    Apollo

  • *****
  • Posts: 1,127
  • Joined: 29 Dec 2009

Posted 23 January 2011 - 11:57 AM

Agreed, Don. If we can control something then we should. It at least eliminates one source. We are constantly moving towards perfection. Is it attainable? That depends on how nit-picky you want to get. Are we close? Likely, at least with the things we can control. -David.

#282 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 47,254
  • Joined: 23 Jun 2003

Posted 23 January 2011 - 12:03 PM

David,
I've come up with several methods to exactly mark the center of the mirror to the required accuracy, but placing the collimation marker on the mirror to that precision currently escapes me.
I'm working on it, though.

#283 Nils Olof Carlin

Nils Olof Carlin

    Vanguard

  • *****
  • In Memoriam
  • Posts: 2,227
  • Joined: 26 Jul 2004

Posted 23 January 2011 - 12:28 PM

...the rings on the outside of the template do move in toward the mirror when the center of the template is depressed to the mirror.

So will the plastic rulers. Let's calculate an example: take a 500 mm f/4 mirror. The radius is 250 mm, but the path along the mirror is
2000*arcsin(250/2000)=250.66 mm (don't forget to switch the calculator to radians, if you try this :grin: )
- if the plastic ruler were pressed down to the surface all the way, you'd have to put those blocks 0.65 mm outside the radius marks.
But more likely, the plastic would form a rather straight line along a chord from edge to center, and be
2*sin(0.5*arcsin(250/2000))*2000 = 250.49 mm. So put those blocks 0.5 mm outside the radius marks, and you should be right. You know whether you are, if those blocks are tight against the mirror once you press down (of course you check without the actual center mark, or at least with its protective coat in place :grin: ).

#284 Gianluca67

Gianluca67

    Viking 1

  • -----
  • Posts: 629
  • Joined: 26 Nov 2008

Posted 23 January 2011 - 05:34 PM

I strongly agree with Don. A repeatable method to position the centermark as close to the dead center of the mirror as possible is still to be addressed. Eliminating collimation errors means we can stay within the "envelope" during an observing session without having to recollimate. Taday we have tools (Blackcat + Hotspot, XLK, XLK-C) capable of delivering high results in collimation but not a repeatable method to reach sub 0.1 mm in placing the center spot. I am interested in this discussion as I would like to replace my Catseye triangle soon.

#285 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 23 January 2011 - 06:06 PM

Here is a little mod to the method I described earlier

1- Jam the mirror between wood blocks with about the same height as the mirror
2- Secure only one side of the template with heavy weight or place another wood block on the top and drive few screws through the top/bottom wood blocks with the template in between
3- Now any template slippage is controlled and will be negligible. For 12.5” mirror with 4mm sagitta, that would be 0.05mm – OK, maybe 0.1mm in real life. You could shift the template by 0.1mm away from the secured side to compensate.
You can try the above with some mockup object to test the idea and consistency before actually doing it on the mirror.
Jason

Attached Thumbnails

  • 4335058-center.png


#286 UmaDog

UmaDog

    Vanguard

  • *****
  • Posts: 2,054
  • Joined: 15 Sep 2010

Posted 23 January 2011 - 06:46 PM

Completely agree, Don. The placing of the centre mark is the part of the process which bothers me most.

When I replaced my ring-binder with a triangle I did something very similar to what you describe. I used books instead of wood blocks and chose to sandwich the template between books on both sides, not just one. The weight of the books is such that the template can slip through am as you push downwards. Obviously you want the resistance to be the same on both sides. I understand why motion would be negligible with your method, but is there a reason why you rejected weights on both sides?

#287 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 47,254
  • Joined: 23 Jun 2003

Posted 23 January 2011 - 07:13 PM

Jason put the weight on only one side so the other side COULD slip. Calculations show the error in descending to the mirror would be in the neighborhood of 0.1mm.
That accuracy would be obtained if the acetate template was fairly rigid, descending as a chord of the straight line from edge to center.
Nils Olof's technique of attaching blocks to the template that, when the template has been pushed into contact with the mirror, would contact the outside of the mirror when the template had been pushed down to the mirror surface would seem to be fairly fool-proof if the distances were correctly calculated.
I'm working on a technique that involves a plexiglas disc and a center plunger with the centermark attached. It wouldn't be as cheap as an acetate template, but it would be extremely accurate.

#288 UmaDog

UmaDog

    Vanguard

  • *****
  • Posts: 2,054
  • Joined: 15 Sep 2010

Posted 23 January 2011 - 07:26 PM

I suppose I was asking: wouldn't it be more accurate if the resistance were the same on both sides? You know you're going to get an error, albeit small, if it's fixed on only one side; but the error may be smaller, but perhaps more uncertain, if it's fixed on both.

#289 michael_m

michael_m

    Ranger 4

  • *****
  • Posts: 351
  • Joined: 16 Nov 2010

Posted 23 January 2011 - 07:34 PM

Hi guys, this is my first post! :jump: Been a member for a couple of months. CN is definitely the place to be for learning and sharing.

Well, it seems to me that the acetate template is meant to be a flat surface that is trying to form to a parabolic surface. There is no perfect way to match the two without some slippage, warpage or deformation, making it difficult to center mark with super high accuracy, although it is easier with large mirrors and at f/5 versus f/4 as Don points out.

Why not hassle out the details in the manufacturing process to create a template that is already a parabola, one that very easily fits the mirrors figure. One that has virtually zero slippage. Jim Fly could make templates at the most common focal ratios like f/3.6 (Steve Kennedy mirrors), f/4, f/4.5, f/5 etc. Near misses in focal ratio would probably still work nicely.

This might possibly lead to repeatable 0.1mm accuracies that Tom is talking about. It's worth looking into at least. It obvious that the accuracy of the center marking is not up to the accuracy of the collimating tools. They need to match better to get the most out of the fine CATSEYE collimation tools.-michael

#290 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 23 January 2011 - 07:46 PM

How about adding a spacer between the template and the spot -- just enough to have the spot around 0.25mm above the mirror surface.
In this case, I would revert back to the idea of applying weights on both sides -- not just one.

Attached Thumbnails

  • 4335320-center2.png


#291 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 23 January 2011 - 07:49 PM

Hello Micahel,
Welcome to "active" membership at CN ;)

Why not hassle out the details in the manufacturing process to create a template that is already a parabola, one that very easily fits the mirrors figure

I prefer the idea I just mentioned since most of the mirror surface will remain untouched.

Jason

#292 lightyear44

lightyear44

    Apollo

  • *****
  • Posts: 1,127
  • Joined: 29 Dec 2009

Posted 23 January 2011 - 07:54 PM

Both seem like ideas with potential. And, welcome to Cloudy Nights, Michael. I'm sure you'll have plenty to contribute around here. Me? I tend to get help, more than I can give it...but it's a great community. (And NICE choice of scope builder/ratio, I might add ;) ;) ;) :rofl2:) -David.

#293 lightyear44

lightyear44

    Apollo

  • *****
  • Posts: 1,127
  • Joined: 29 Dec 2009

Posted 23 January 2011 - 07:55 PM

Hmmmm. Good point, Jason. -David.

#294 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 23 January 2011 - 07:58 PM

Back to this animation
I measured the difference accurately and came up with 0.3mm (0.012"). But I do believe the HotSpot placement was more accurate because I spent more time working on placing it.
Jason

Attached Thumbnails

  • 4335349-template.jpg


#295 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 47,254
  • Joined: 23 Jun 2003

Posted 23 January 2011 - 08:15 PM

How about adding a spacer between the template and the spot -- just enough to have the spot around 0.25mm above the mirror surface.
In this case, I would revert back to the idea of applying weights on both sides -- not just one.

Jason,
Your drawing is similar to the idea I had except the horizontal surface is a piece of rigid transparent plexiglas with markers for various diameter mirrors. In the center is a spring loaded plunger that has a lip of about 0.1" under the plexiglas. The plunger in the center of the plexiglas is spring loaded up, away from the mirror. The center marker is attached to the bottom of the plunger, the plexiglas piece is placed on the mirror and centered according to the template markings (the plexiglas is transparent and the markings are on the clear plexiglas).
The plunger is then pushed in toward the mirror, and the center marker is affixed to the mirror by pressure on the plunger. The plunger then springs upward away from the mirror, and the accuracy of the installation is only determined by the accuracy of putting the center marker on the plunger.
If this seems like a viable product, I know someone who could make a sample.

#296 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 24 January 2011 - 12:06 AM

spring loaded plunger

But wouldn't the plunger mechanical tolerence eats away from some of the additional accuracy we are after? That plunger needs to be accurately centered and needs to be sturdy enough not to sway when it is pressed. We are talking sub 0.1mm accuracy.
Jason

#297 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 24 January 2011 - 12:29 AM

Another data point.
The difference between my center spot placement versus where the original spot was placed by the mirror maker is 0.45mm.

Attached Thumbnails

  • 4335837-template2.jpg


#298 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 24 January 2011 - 01:56 AM

To keep things within perspective, here is a reminder to how small a 0.5mm distance -- see attachment.
A 0.5mm is about the width of that ridge around an American penny.
Don, getting down to 0.1mm accuracy consistently using affordable mechanics by average/first time users is tough.

Attached Thumbnails

  • 4335915-penny.jpg


#299 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 7,413
  • Joined: 21 Oct 2006

Posted 24 January 2011 - 04:16 AM

Pull out an American penny.
Can you see the 'I' in LIBERTY?
The width of the 'I' is around 0.1mm -- actually 0.1mm is still smaller than the width of the 'I'

Attached Thumbnails

  • 4335991-penny2.jpg


#300 sixela

sixela

    Hubble

  • *****
  • Posts: 15,550
  • Joined: 23 Dec 2004

Posted 24 January 2011 - 04:32 AM

To keep things within perspective, here is a reminder to how small a 0.5mm distance -- see attachment.


But the great thing about a template is that a 0.1mm offset in the middle generates a 0.2mm difference between the two opposite edges, and that's something I can plainly see.


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics