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

Center spotting: Polished mirror or O.D.?

This topic has been archived.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

#1 Intihuatana

Intihuatana

    Mariner 2

  • *****
  • topic starter
  • Posts: 270
  • Joined: 15 Jan 2014

Posted 05 April 2014 - 05:00 PM

What to use to center spot your primary?, the outside diameter of the mirror blank or the outside diameter of the mirrored surface?

Enclosed picture of the 200mm primary mirror of my f/4 AT8IN newt show about 1.5-2 mm difference between them when using the Catseye template.

...is the error relevant for an f/4 if chosen the wrong center?, if so what to chose?. Found a couple threads about this but no definitive response.

Thanks in advance for any advice.

Attached Thumbnails

  • 6449479-photo 4-2 - 2014-04-05 at 17-47-10.jpg


#2 tazer

tazer

    Surveyor 1

  • -----
  • Posts: 1882
  • Joined: 22 Dec 2011

Posted 05 April 2014 - 05:30 PM

I'm eagerly watching this thread because I had the same question myself when spotting my primary. I ended up aligning the template with the glass rather than the mirrored surface, which is how I interpreted the Catseye instructions:

Align the appropriate mirror diameter template outline ring (or arc segments) with the perimeter or bevel edge of the mirror.



#3 Bill Weir

Bill Weir

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 01 Jun 2004

Posted 05 April 2014 - 05:48 PM

Is that uncoated area a bevel? If so then you want to use the mirrored portion as the guide. That' s what the instructions mean when they distinguish between the perimeter or the beveled edge. http://www.catseyeco...ing_hotspot.pdf Sometimes the polish goes right to the very outside edge and sometimes the blank is beveled before polishing. In the beveled cases the parabolic centre of the mirror isn't always the geometric centre. Use the shiney as the guide.

Bill

#4 CatseyeMan

CatseyeMan

    Vendor (Cats Eye Collimation)

  • *****
  • Vendors
  • Posts: 566
  • Joined: 16 Dec 2004

Posted 05 April 2014 - 07:06 PM

Actually, the prevailing advice from renown parabolic opticians is to use the "glass perimeter" as the reference rather than the inside edge of the bevel because of the manner in which the parabolic figure is generated on the spinning blank centered on the axis of rotation. The bevel is ground after figuring (and coating I think) and is sometimes uneven. The objective in applying the spot IS to place it at the geometric center of the blank which is assumed to be at or very close to the parabola vertex. I can see that I need to augment my Spotting Procedure to make that a bit clearer to the reader.

#5 Pinbout

Pinbout

    ISS

  • *****
  • Posts: 22545
  • Joined: 22 Feb 2010

Posted 05 April 2014 - 09:17 PM

I wouldnt grind the edge after figuring. But if it is hand ground while grinding it could be as you say ... uneven.

#6 Intihuatana

Intihuatana

    Mariner 2

  • *****
  • topic starter
  • Posts: 270
  • Joined: 15 Jan 2014

Posted 05 April 2014 - 11:02 PM

...the prevailing advice from renown parabolic opticians is to use the "glass perimeter" as the reference.. .


Thanks...makes sense. In any case worst case scenario I'd be entitled to a new set of center spotting triangles!.

#7 tazer

tazer

    Surveyor 1

  • -----
  • Posts: 1882
  • Joined: 22 Dec 2011

Posted 06 April 2014 - 09:30 AM

I can see that I need to augment my Spotting Procedure to make that a bit clearer to the reader.


As a novice to mirror spotting (my first was last year with the Catseye Hotspot) here's my take. After removing the primary and looking at it, I noticed the bevel for the first time. I had no idea what its purpose was as I wasn't familiar with the mirror grinding/figuring/coating process. What I did notice was that it wasn't perfectly even. So, I used the perimeter of the mirror to align the template as it seemed more natural.

Of course I know now that the bevel is there to protect the edges from chipping but at the time I scratched my head for a bit. Stating the purpose of the bevel and that most mirror manufacturers grind the mirror with respect to the glass perimeter would certainly clarify things.

By the way, since the factory installed center spot on my mak-newt was off by 3-4mm, based on your template, my first collimation after respotting (using the Blackcat and XLKP) produced amazing results. Views are sharp and crisp. It's almost like having a new scope. Really a great set of tools and one of the best investments I've made since entering the hobby.

#8 howard929

howard929

    Member

  • *****
  • Posts: 4603
  • Joined: 02 Jan 2011

Posted 06 April 2014 - 10:20 AM

I've yet to try this but removing the blank and placing it on a level surface, a small round bearing should settle at the center of the only thing that counts, the parabola.

Placing a center spot with a pair of tweezers so the hole in it is centered around the bearing is doable.

edit to add: I'm quite sure if Einstein was still alive that he'd Approve Of This Message.

#9 Howie Glatter

Howie Glatter

    In Memoriam

  • *****
  • In Memoriam
  • Posts: 1104
  • Joined: 04 Jul 2006

Posted 06 April 2014 - 10:53 AM

". . removing the blank and placing it on a level surface, a small round bearing should settle at the center . ."

I remember that this was discussed here in the past. Your procedure assumes there is no wedge to the blank. You would have to level your table with a very sensitive level.

#10 howard929

howard929

    Member

  • *****
  • Posts: 4603
  • Joined: 02 Jan 2011

Posted 06 April 2014 - 11:13 AM

In no way did I think I was the first person here to propose that. It's obvious to anyone familiar with general relativity.

Polling for wedge on a blank would be easy with a caliper. Adjusting for any should also be simple with pieces of paper under the short edge. Leveling is of course another matter, doable though I believe. The only problem here is that the only ones making money from this are those selling calipers, levels and small round bearings.

#11 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 1545
  • Joined: 01 Oct 2007

Posted 06 April 2014 - 11:16 AM

If the mirror's edge (outside diameter) has been machined, then generally one should use that for centering.

I indicate off the outer edge when I center up mirrors on the turntable so that I can scribe a permanent series of concentric circles to mark the center.

A bevel is just a safety measure to help prevent chipping, and some shops apply it hastily and non-uniformly. Do not assume that it is uniform or accurate.

If the mirror is not round, as happens with cheap glass or molded blanks, then all you or I can do is make a good estimate of the center and put the mark there.

#12 Howie Glatter

Howie Glatter

    In Memoriam

  • *****
  • In Memoriam
  • Posts: 1104
  • Joined: 04 Jul 2006

Posted 06 April 2014 - 11:29 AM

". . the only ones making money from this are those selling calipers, levels and small round bearings."

I bought one of those levels about 30 years ago, and it cost $150 then. From Poland, hand-scraped base, and if you put a cigarette paper under one end, the bubble would travel an inch down the internally ground and polished vial. However,the high precision ball bearings are cheap.

#13 howard929

howard929

    Member

  • *****
  • Posts: 4603
  • Joined: 02 Jan 2011

Posted 06 April 2014 - 11:39 AM

Howie, of course YOU have one of those levels. I'm not surprised in the least. :bow:

OTOH, I'd use my $60 4 foot level or the $35 digital I own and put enough trust in that and Einstein.

#14 Bill Weir

Bill Weir

    Mercury-Atlas

  • *****
  • Posts: 2608
  • Joined: 01 Jun 2004

Posted 06 April 2014 - 09:43 PM

Actually, the prevailing advice from renown parabolic opticians is to use the "glass perimeter" as the reference rather than the inside edge of the bevel because of the manner in which the parabolic figure is generated on the spinning blank centered on the axis of rotation. The bevel is ground after figuring (and coating I think) and is sometimes uneven. The objective in applying the spot IS to place it at the geometric center of the blank which is assumed to be at or very close to the parabola vertex. I can see that I need to augment my Spotting Procedure to make that a bit clearer to the reader.



So I see I mistook the meaning of your instructions. I made the incorrect assumption that distiguishing between the perimiter of the glass and the beveled edge meant they were two different things. So they aren't? Cool. I think just leave the bevel part out and say line things up with the outside of the glass. For me this was all a moot point with centre spotting my 20" as it has one of those nice little "Bulls Eye" that Mike etches in.

Bill

#15 Intihuatana

Intihuatana

    Mariner 2

  • *****
  • topic starter
  • Posts: 270
  • Joined: 15 Jan 2014

Posted 07 April 2014 - 12:02 AM

Took me a great while to figure out a method to center the template in the mirror with some confidence....and I'm certain I still may have at least 1 or 2 mm of error...(you just can't use your bare eyes to discern how far the template's circumference is relative to the edges of the mirror blank...as the position changes depending on the angle of vision)...would this kind of error degrade the collimation quality/images in my 8" f/4 scope?.

Wondering why mirrors can't be center spotted at the factory during the grinding process ( drill a physical hole)?...we spend hundreds in collimation tools that may not be worth much if the mirror can't be center spotted accurately to start with?.

#16 orlyandico

orlyandico

    Cosmos

  • *****
  • Moderators
  • Posts: 9365
  • Joined: 10 Aug 2009

Posted 07 April 2014 - 01:17 AM

I also recently re-spotted my AT8IN with Jim's template.

The mirror is indeed beveled, which makes centering the template problematic. I am afraid that I just followed the bevel line (it is not immediately evident how to center the template on the mirror's perimeter, since as the OP noted parallax can throw off the centering).

I'm not sure how good the centering has to be, but I've gotten the roundest stars ever (still not perfectly round though) since doing the process and using the XLKP autocollimator.

#17 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 7383
  • Joined: 21 Oct 2006

Posted 07 April 2014 - 01:30 AM

Consider using this method

Posted Image

Place risers next to the mirror at four points. Use the risers edge to center the template. The risers will also help keeping the template flat. Place weights on top of the template over the risers to keep the template steady.
This is the method I used to center spot my mirror.

Jason

#18 howard929

howard929

    Member

  • *****
  • Posts: 4603
  • Joined: 02 Jan 2011

Posted 07 April 2014 - 08:15 AM

I also recently re-spotted my AT8IN with Jim's template.

The mirror is indeed beveled, which makes centering the template problematic. I am afraid that I just followed the bevel line (it is not immediately evident how to center the template on the mirror's perimeter, since as the OP noted parallax can throw off the centering).

I'm not sure how good the centering has to be, but I've gotten the roundest stars ever (still not perfectly round though) since doing the process and using the XLKP autocollimator.


A star test after careful collimation will reveal an error in center spot placement. If the spot is placed so that it's within the sweet spot it's obviously close enough. As you rightly noticed, with a mask it's more prone towards error then it is to exactness, hit or miss.

There's also a way to determine the center of the blank using a digital camera and some software, also nothing new, it's been around for many years. FWIW I'm liking the roller bearing idea quite a lot. It's the only method that actually forces exactness.

#19 Intihuatana

Intihuatana

    Mariner 2

  • *****
  • topic starter
  • Posts: 270
  • Joined: 15 Jan 2014

Posted 07 April 2014 - 09:59 AM

Orly good to know your collimation went well, I've also seen your posts about reinforcing the focuser, changing the collimation springs, etc...

Jason thanks, as a matter of fact I had already followed your advise from other posts about using weights when applying the center spot. Most of my concern although was with centering the template... this is what I came up with:

Pictures here
https://flic.kr/s/aHsjWtkRyt

1. Insert three push pins visually in the template around the centered mirror blank diameter along the three radial axis
2. Use a tablet +level as a "leveled magnifying glass" to take pictures of the three axis in the template (I used my iPad and iphone + level app)
3. Center the push pin's head with its own base to visually align the camera
4. Compare pictures and do as many iterations as possible until the O.D. of mirror blank has the same distance in the three axis (including changing push pin positions)

My first "visual" positioning was terribly off... by at least 3mm in one of the axis ( and I have 20/20 vision...)

#20 CatseyeMan

CatseyeMan

    Vendor (Cats Eye Collimation)

  • *****
  • Vendors
  • Posts: 566
  • Joined: 16 Dec 2004

Posted 07 April 2014 - 06:43 PM

I also recently re-spotted my AT8IN with Jim's template.

The mirror is indeed beveled, which makes centering the template problematic. I am afraid that I just followed the bevel line (it is not immediately evident how to center the template on the mirror's perimeter, since as the OP noted parallax can throw off the centering).

I'm not sure how good the centering has to be, but I've gotten the roundest stars ever (still not perfectly round though) since doing the process and using the XLKP autocollimator.


Unless the bevel width varies noticeably by more than a 1/2 mm, using the inside bevel edge (mirrored surface perimeter) should yield sufficient centering accuracy as exhibited by your performance-improvement experience .

#21 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 41739
  • Joined: 23 Jun 2003

Posted 08 April 2014 - 03:49 PM

Ahem:
The bearing technique will not work.
First, because the center of the mirror is nearly flat and there is very little gravitational assist to the centering of the ball.
Second, because there is friction between the bearing and the mirror surface, it is unlikely to stop at the dead center.
Third, small steel balls have surface irregularities (unless of very high accuracy, unlikely in anything outside of an expensive, sealed, precision bearing), so they will not necessarily roll over the irregularity.

No, I'm afraid this technique is unlikely to find the center. Jim Fly's acetate template is FAR more accurate, and easily reveals errors smaller than 1mm.
Want to prove it for yourself? Put a steel ball down 5mm off center on your mirror and see if it rolls. It didn't on my mirror.

#22 Howie Glatter

Howie Glatter

    In Memoriam

  • *****
  • In Memoriam
  • Posts: 1104
  • Joined: 04 Jul 2006

Posted 08 April 2014 - 07:18 PM

". . small steel balls have surface irregularities (unless of very high accuracy, unlikely in anything outside of an expensive, sealed, precision bearing). . '

Don, I agree regarding the impracticality of this center-finding method, but the hardened steel balls used in what are technically called rolling element anti-friction bearings are very cheap, and commonly accurate in sphericity to 0.00004". True, when they are assembled with hardened steel precision races, they cost a lot more.
Most people do not realize that the development of these bearings and the invention of the technology for their mass-production was one of the things that made the modern world possible (which would make it a mixed blessing).

#23 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 7383
  • Joined: 21 Oct 2006

Posted 08 April 2014 - 08:46 PM

Let us consider my mirror with 1200mm FL for reference.
I found using Catseye template accuracy to be around 0.2mm or better. If I were to use the ball bearing technique I should expect a consistent 0.1mm accuracy or better -- otherwise there will be no incentive to use it over the template.
At 0.1mm off center, the tangential plane has a slope of 0.0000417 for my mirror. That is around 8.6 arcseconds.
Question: Assuming perfect ball bearing, can anyone honestly level the mirror with an accuracy better than 8.6 arcseconds? That is equivalent to raising one end of a 24m bar by 1mm.
It would be impossible for an average Joe with average means to achieve this leveling accuracy.
I am referring to leveling the tangential plane at the optical axis of the mirror -- not the bottom.
Jason

#24 Intihuatana

Intihuatana

    Mariner 2

  • *****
  • topic starter
  • Posts: 270
  • Joined: 15 Jan 2014

Posted 09 April 2014 - 06:27 AM

Jason thanks, as a matter of fact I had already followed your advise from other posts about using weights when applying the center spot. Most of my concern although was with centering the template <on my unevenly beveled mirror>... this is what I came up with:

pictures here
https://flic.kr/s/aHsjWtkRyt

1. Insert three push pins visually in the template around the centered mirror blank diameter along the three radial axis
2. Use a tablet +level as a "leveled magnifying glass" to take pictures of the three axis in the template (I used my iPad and iphone + level app)
3. Center the push pin's head with its own base to visually align the camera
4. Compare pictures and do as many iterations as possible until the O.D. of mirror blank has the same distance in the three axis (including changing push pin positions)

My first "visual" positioning was terribly off... by at least 3mm in one of the axis ( and I have 20/20 vision...)


Any comments appreciated...pictures in the provided link.

#25 azure1961p

azure1961p

    Voyager 1

  • *****
  • Posts: 13595
  • Joined: 17 Jan 2009

Posted 09 April 2014 - 08:31 AM

I also recently re-spotted my AT8IN with Jim's template.

The mirror is indeed beveled, which makes centering the template problematic. I am afraid that I just followed the bevel line (it is not immediately evident how to center the template on the mirror's perimeter, since as the OP noted parallax can throw off the centering).

I'm not sure how good the centering has to be, but I've gotten the roundest stars ever (still not perfectly round though) since doing the process and using the XLKP autocollimator.


The bevel line would seem to make the most sense . Look at it this way - what if the mirror was egg shaped but the reflective coating was the perfect circle? The mirror could be shaped like a tarantula - all that matters (to my mind and how I did it ) is the relevance if centering that reflective surface and placing the dot smack in the middle. The mirror could be shapers like a brick - so long as that aluminized circle has its center found - your good.

I could be wrong but Im seriously doubting I am.

Pete


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