Jump to content


Photo

center mark and collimation questions

  • Please log in to reply
42 replies to this topic

#26 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6655
  • Joined: 21 Oct 2006
  • Loc: California

Posted 11 December 2012 - 09:21 PM

We had a similar lengthy discussion a while back starting with this page
http://www.cloudynig...3532750/page...

#27 howard929

howard929

    Member

  • *****
  • Posts: 3524
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 12 December 2012 - 10:15 AM

a few years ago I posted about an idea I had:
A rigid plexiglass disc the size of the mirror is rested on the mirror's edge, then centered.


The rest of that idea seems good, though a suggestion. Rather then a plexi disc placed by hand and eye, a piece of square plexi that references both edges of a right angle fixture that the mirror also references. ie: a rigid system that forces accuracy.

#28 howard929

howard929

    Member

  • *****
  • Posts: 3524
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 12 December 2012 - 10:54 AM

I have a question that's haunting me here. The primary mirror I have is figured and polished only in what I hope is the center with a small "ring" of rough glass. I can easily understand why it's done that way.

Does the 200mm stated size reference the polished section or the overall size of the blank? Further, who could guarantee that a 2 year old 200mm mirror that I have would be the same size as a brand new one or any other 200mm mirror to within "what?" tolerance?

When I verified the placement of the new spot I referenced the polished section but I never thought to measure its size since that didn't matter.

For the prospect of producing a mask, size does matter if forced accuracy is sought. Thoughts or insights?

#29 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 22897
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 12 December 2012 - 11:29 AM

When determining the center of a mirror, use the outer edge of the mirror as a reference. The beveled edge is done after the mirror is polished, and it can vary in width.

#30 howard929

howard929

    Member

  • *****
  • Posts: 3524
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 12 December 2012 - 11:46 AM

When determining the center of a mirror, use the outer edge of the mirror as a reference. The beveled edge is done after the mirror is polished, and it can vary in width.


More to the point. Is the outer edge of any given sized mirror always the stated size of the mirror? And if so, to within what accuracy? I have a feeling that the only one(s) who know these answers are the ones who actually make these mirrors. Does GSO and Synta make for example, primary mirrors that are always a given size to very tight tolerances and would they be he same? I sort of doubt that but it doesn't hurt to ask.

#31 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 22897
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 12 December 2012 - 02:44 PM

I have seen some companies that have changed their sizes to conform with American sizes.
Metric mirror sizes would be logical at 200mm (7.87"), 250mm (9.84") and 300mm (11.81"), and I've seen these sizes.
American sizes are logical at 8" (203mm), 10" (254mm), and 12.5" (317.5mm). There aren't any 12" American mirrors that I know of except in catadioptric scopes.
If a manufacturer were using metric-sized equipment but wanted to conform to American size standards, we might get mirrors of 205mm, 255mm, and 305mm to correspond to 8", 10", and 12".
And, indeed, I've seen that, from more than one company, and varying over the years.
There is usually great consistency in a particular size of particular scope over a couple years. But in the long run? I wouldn't count on it.

Yes, the size of the mirror is measured on the outside edge of the glass itself. The reflective surface is always a little smaller than that, but the width of the bevel varies greatly. I've seen zero bevel (on a refigured mirror) and I've seen around 1/8" (!). The mirror with the large bevel had a bevel of varying width, to boot. The center of the mirror wasn't the center of the reflective surface!. The mirror performed fine in the field, though. I wondered if the mirror maker increased the bevel size to help take care of a Turned Down Edge or some other issue.

#32 howard929

howard929

    Member

  • *****
  • Posts: 3524
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 13 December 2012 - 12:32 PM

That's bad news for pre-made masks or right angle fixtures. Makes me wonder how many mirrors aren't polished in the exact center of the blank and it's no wonder that so many find star testing reveals the need for tweaks.

#33 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 22897
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 13 December 2012 - 02:42 PM

That's bad news for pre-made masks or right angle fixtures. Makes me wonder how many mirrors aren't polished in the exact center of the blank and it's no wonder that so many find star testing reveals the need for tweaks.


You misinterpreted what I said.
When the mirror is made, the center of the mirror is the glass blank's center. The anti-chip bevel is put on the mirror AFTER the mirror is finished.
If the bevel is exactly equal all the way around, the center of the mirror blank and the center of the reflective surface coincide. The two circles are concentric.

I have seen one 16" mirror with a bevel that varied in width all the way around, and was quite large (looked like it had been done with a file). The center of that mirror was the center of the glass blank, as you would expect. But it wasn't exactly in the center of the aluminized reflective surface because the outside edge of the reflective surface and the glass blank were not concentric!

I did not measure to see if the center marker was positioned in the center of the aluminized surface or the center of the glass blank, but given the precision by which center markers are applied, I don't think it would have mattered. The maximum error caused by the exaggerated bevel would have been less than the likely error in application anyway.

Your other point, about the center of the mirror's figure of revolution not being the center of the blank? Unlikely. Machine-made mirrors would have to be eccentric (i.e. not round) for this to occur.

The point to take from this is: When positioning a center marker on your mirror, use the outside edges of the glass as the place you determine the center from. IF the bevel is perfectly even, you can use the edge of the reflective surface.

I do it a very easy way: place the acetate template down on the mirror, and place pins through the acetate at 4 places to hold the acetate template in place over the mirror (the pins press against the side of the glass mirror). Then, I use a pencil with an eraser tip to press down the center of the template until the center marker is stuck on the mirror. The pins prevent a lateral movement of the acetate when I'm pushing down, and even if there is differential stretch in the acetate as I press down, the lateral movement of the center marker is a small fraction of a millimeter, and I can be off a LOT more than that before the error makes collimation iffy.

#34 tag1260

tag1260

    Apollo

  • *****
  • Posts: 1102
  • Joined: 07 Oct 2012
  • Loc: Ohio, USA

Posted 13 December 2012 - 03:11 PM

Boy!! Ask a couple questions around here and you get a heck of an education!!! :cool:

I guess , not being around ATM'ers, I still can't actually see how your mirror CAN'T be made off center.

I know I see .5 mm out of place for your center mark but in reality, how far off can you be before it REALLY affects your collimation in that we can see it?

Thanks again for all the discussion. (Even those of you who are talking WAY over my head!!! ;) )

Tag

#35 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 22897
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 13 December 2012 - 04:00 PM

Usual rule of thumb is that the center marker should be within 0.005x the diameter of the mirror (on a 10", that's 1.27mm).
Since the error, IIRC, in a cheshire, due to the double bounce, is doubled, that means within 0.635mm from exact center (someone correct me if I don't remember correctly).
0.635mm is 0.025", or 1/40 of an inch.

#36 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 473
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 13 December 2012 - 04:03 PM

When the mirror is made, the center of the mirror is the glass blank's center.

Not necessarily. More below.

The anti-chip bevel is put on the mirror AFTER the mirror is finished.

That is completely wrong.

The bevel is put on BEFORE generating or grinding is done to prevent chipping of the edge of the blank during those aggressive operations. (The bevel may need to be enlarged again during grinding if a lot of glass is removed.)

After the mirror is done, the bevel provides some protection from chipping should the mirror be banged into something hard. (I often ease the sharp edges of the bevel, front and back, to further reduce the chance of future damage from handling mishaps.)

If the bevel is exactly equal all the way around, the center of the mirror blank and the center of the reflective surface coincide. The two circles are concentric.....

No, that is not enough to guarantee that the optical center and mechanical center are coincident.

If the blank has WEDGE, then it is thicker on one side than another. In that case, all bets are off - even if the bevel is uniform the center of the optical figure may not be the mechanical center of the blank. Finding the true optical center can then be very difficult.

Your other point, about the center of the mirror's figure of revolution not being the center of the blank? Unlikely. Machine-made mirrors would have to be eccentric (i.e. not round) for this to occur.

Sorry, but that is also wrong. If they have wedge this could happen too.

Thankfully most blanks are machined with good equipment, and therefore have uniform thickness.

I have seen mirrors that are non-round, have non-uniform bevels, and have significant amounts of wedge (up to 3/16"). However, none of those were from major glass suppliers, who will generally provide nicely round blanks of uniform thickness.

All that I can say is that if a mirror has a uniform edge thickness (which can be measured carefully with calipers), it is VERY LIKELY that the optical center coincides with the mechanical center. It could even have a non-uniform bevel and still be fine, though that would throw off edge thickness measurements.

However, for a mirror that is non-round or has wedge, all bets are off, and center spot at your own risk.

#37 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4782
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 13 December 2012 - 04:03 PM

...I see .5 mm out of place for your center mark...how far off can you be before it REALLY affects your collimation in that we can see it?

It depends...

For your f/5 Dob, the high magnification (300X to 600X for your 12-inch aperture) primary mirror axial tolerance is a bit less than 1mm. If your primary mirror center spot is offset 1mm, you'll need to add 1/2 of that offset to your final primary mirror alignment result, which means you'll need to keep your Cheshire alignment within 1mm of "perfect" (the Cheshire magnifies the primary mirror axial error read 2X). This is the "worst case scenario", and is the one I use since I rely on tool alignment to get the job done (I like to know that my alignment is up to the task when my seeing gets really good).

Of course, both the center spot placement error and the Cheshire error have direction and should be considered as vector quantities. This basically means, in the "best case scenario", a residual Cheshire error could cancel out a center spot placement error. Unfortunately, the only way to be sure is to verify the alignment on a star (when I would rather be observing Jupiter).

Tolerances aside, considering the effects of gravitational flexure on truss OTAs and torque forces on the UTA, the closer you can keep the primary mirror axis to "perfect" alignment, the more likely you'll stay within tolerance throughout an observing session--which means you won't miss out when those moments of extraordinary seeing unexpectedly arrive.

#38 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 22897
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 13 December 2012 - 04:22 PM

Mike's post is illuminating.
I guess I was assuming a round blank with no wedge, but home-made mirrors can sometimes have wedge generated by the grinder, if I recall my mirror-making class back in the '70s.
Mike, thanks for correcting me on the bevel. In our class, most chose to ignore a bevel, and those that did one did it after the mirror was made, and solely for edge protection. I can see how having the bevel done first would help avoid chips.

But, though you have seen some horror stories in mirror blanks, how many commercial mirrors are likely to be delivered to customers that aren't round and have significant wedge or which have radically non-uniform bevels? I would think a vanishingly low number. I would certainly counsel an amateur to assume the center is the center unless placing a center marker there results in visible aberrations in the star images.

#39 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4782
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 13 December 2012 - 04:27 PM

...I have seen mirrors that are non-round, have non-uniform bevels, and have significant amounts of wedge (up to 3/16"). However, none of those were from major glass suppliers, who will generally provide nicely round blanks of uniform thickness.

I tested a 13-inch (early) Coulter primary that had a "wrinkle" on the back of the mirror from about 10 o'clock to 2 o'clock. It looked ugly, but the mirror gave a pretty good image. I haven't seen one with 3/16-inch wedge though! I'm guessing if you had to refigure such a piece of glass, you would start by removing the wedge (assuming there's enough glass), yes?

All that I can say is that if a mirror has a uniform edge thickness (which can be measured carefully with calipers), it is VERY LIKELY that the optical center coincides with the mechanical center. It could even have a non-uniform bevel and still be fine, though that would throw off edge thickness measurements.

I recall chamfering my first ATM project, a 10-inch mirror, after it was parabolized. Back then, Edmund was selling blanks with 1/4-inch chamfers, but University Optics blanks had chamfers that were barely 1/16-inch. I wanted all the aperture I could get, but I got careless and ended up with a small chip on the edge, removable, with about an 1/8-inch chamfer... :shrug:

#40 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 473
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 13 December 2012 - 10:59 PM

Mike's post is illuminating.
I guess I was assuming a round blank with no wedge, but home-made mirrors can sometimes have wedge generated by the grinder, if I recall my mirror-making class back in the '70s.

Yes, that's a good recollection - non-machined, hand ground mirrors are at greatest risk of having wedge since they have never been trued up by a machining operation.

Mike, thanks for correcting me on the bevel. In our class, most chose to ignore a bevel, and those that did one did it after the mirror was made, and solely for edge protection. I can see how having the bevel done first would help avoid chips.

If a bevel gets to less than 1/16" wide, during grinding, then I widen it. If the bevel goes away and one so much as touches it against something hard, there is a risk of a chip coming off.

But, though you have seen some horror stories in mirror blanks, how many commercial mirrors are likely to be delivered to customers that aren't round and have significant wedge or which have radically non-uniform bevels?

Not very many, thankfully, but any mirror that is not machined front and back could have an issue.

Molded blanks are at risk if they are not machined on the back, as are those that are simply cut from sheet glass that is not flat like old sheet Pyrex, which had an uneven surface.

Also, don't forget about non-monolithic blanks (fused, etc.) that the maker may not wish to subject to certain stressful machining operations. If these are not machined front and back, they should be checked.

So, if you see the back of a mirror and it is not machined, just be aware of the possible issue.

The same is true for the SIDE of the blank - a molded blank is often not perfectly round, and it is good to be aware of that too, especially if you are setting the cell to have little clearance around the mirror. The mirror may eventually rotate due to vibration in transport, etc. and may become wedged in the cell. This condition will show up as triangular or astigmatic star images.

I would certainly counsel an amateur to assume the center is the center unless placing a center marker there results in visible aberrations in the star images.

That's good advice. Just remember that a non-machined back or side of a mirror means it should be looked at carefully.

I haven't seen one with 3/16-inch wedge though! I'm guessing if you had to refigure such a piece of glass, you would start by removing the wedge (assuming there's enough glass), yes?

Yes, that was a large BVC blank and it was machined away.

The most maddening thing I have to deal with are smaller non-round blanks, often cut from sheet glass. These pop up from time to time. As polishing is done, the mirror may rotate slowly on what it is placed on (carpet, etc.). This is normal. However, if the retaining "bumpers" on the turntable are not set with sufficient clearance, the mirror will eventually rotate far enough that the wider parts of the mirror become wedged between the bumpers and this inevitably will cause nasty astigmatism or other distortions.

The moral of the story is - a nicely machined blank is a wonderful thing, and it is something that I insist upon starting from. (Refigures must be treated carefully since dimensional uniformity and anneal are not guaranteed to be ideal.)

By the way, it is standard practice for me to center a finished (but uncoated) mirror on a turntable and scribe a series of concentric, precisely centered circles into the glass to mark the exact center of the mirror. This makes center spotting much easier, and my clients love it.

#41 Chucky

Chucky

    Messenger

  • *****
  • Posts: 420
  • Joined: 16 Apr 2010
  • Loc: Dublin, Ohio

Posted 14 December 2012 - 06:47 AM

<< By the way, it is standard practice for me to center a finished (but uncoated) mirror on a turntable and scribe a series of concentric, precisely centered circles into the glass to mark the exact center of the mirror. This makes center spotting much easier, and my clients love it. >>

Now this is nice! I've never heard of such a practice.

#42 okieav8r

okieav8r

    I'd rather be flying!

  • *****
  • Administrators
  • Posts: 4373
  • Joined: 01 Mar 2009
  • Loc: Oklahoma!

Posted 14 December 2012 - 06:51 AM

Mike, what can ATM'ers do to avoid getting improperly made or machined blanks? Who to buy from? Who to avoid? I've been wanting to try making my own 6" or 8" mirror and build a telescope around it for quite a while.

#43 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 473
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 14 December 2012 - 02:00 PM

Now this is nice! I've never heard of such a practice.

A commercial client requested it a while back, and after I figured out how to do it I decided to offer it as a standard feature. Since I mostly make fast mirrors, accurate center spotting is even more important.

Mike, what can ATM'ers do to avoid getting improperly made or machined blanks? Who to buy from? Who to avoid? I've been wanting to try making my own 6" or 8" mirror and build a telescope around it for quite a while.

I can't recommend or warn against specific vendors because that would violate the TOS.

Machined blanks will probably be more expensive, and whether you hog out by hand or use machined blanks to save time is up to you. In either event, check the anneal of the blank before devoting too much time to it.

There is nothing wrong with molded blanks or those cut from sheet. Just make sure they're reasonably round and check the edge thickness after you start grinding. If one side is significantly thicker, grind on that side until it is not. (Wedge in refractor lenses effectively makes them a prism, so lens makers check for this and grind it away.)

With hand grinding, wedge is less likely to cause problems than with machine work.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics