Jump to content


Photo

center mark and collimation questions

  • Please log in to reply
42 replies to this topic

#1 tag1260

tag1260

    Apollo

  • *****
  • Posts: 1175
  • Joined: 07 Oct 2012
  • Loc: Ohio, USA

Posted 09 December 2012 - 10:36 AM

All the discussion here about collimation and I have a hypothetical question on center marks.

Can your center mark be in perfect center but still not be in the proper location for perfect collimation? If so, how would you achieve collimation? Trial and error?

Also, when using a Chelshire eyepiece you set your secondary then finish collimation with the primary. Is there any harm in fine tuning your collimation with your secondary? Seems as though this is what you do when using an autocollimator, don't you?

Thanks

#2 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1614
  • Joined: 26 Jul 2004

Posted 09 December 2012 - 10:53 AM

Can your center mark be in perfect center but still not be in the proper location for perfect collimation?


In principle, yes - in practice, it must be rare at least with machine-figured and accurately round blanks. But what you can do is try a precise star collimation when the seeing is good enough, adjusting to make the diffraction rings accurately symmetric near focus (as seen in the exact center of the FOV).
This done, check with a cheshire or equivalent that the spot appears centered - if not, try again and see if the error is consistent - if so, move the spot accordingly. Is there anybody out in the wild who have actually done this?
Anyway, this particular test is the only reason I can think of for doing a star collimation.

Is there any harm in fine tuning your collimation with your secondary?

If you fine tune the aim of the primary by adjusting the secondary, you will move the focuser axis (from secondary to primary) by much more than the primary axis (from secondary to focus). The aim of the focuser axis is less critical than that of the primary, and you might have a hard time seeing any ill effect, but it would be prudent to check the focuser axis after doing such a fine tuning.

Nils Olof

#3 GaryS

GaryS

    Ranger 4

  • *****
  • Posts: 381
  • Joined: 30 Oct 2006

Posted 09 December 2012 - 12:35 PM

Can your center mark be in perfect center but still not be in the proper location for perfect collimation?


I've never seen it and I suspect it's highly unlikey, if not impossible.

Also, when using a Chelshire eyepiece you set your secondary then finish collimation with the primary. Is there any harm in fine tuning your collimation with your secondary?


That's how I do it. (You can read all about my collimation technieques in the article, "A Beginner's Guide to Collimation" on my web site.) The key is to reserve that adjustment for the final tweak.

Regards,
Gary

#4 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6665
  • Joined: 21 Oct 2006
  • Loc: California

Posted 09 December 2012 - 02:03 PM

Also, when using a Chelshire eyepiece you set your secondary then finish collimation with the primary. Is there any harm in fine tuning your collimation with your secondary?


Fine tune with respect to which reflections?
If you are fine tuning by centering the cheshire eyepiece cross-hair with the primary center spot then this might be unacceptable.
If you are fine tuning by aligning the primary center spot with the cheshire eyepiece pupil hole reflection then it is more acceptable.
But if you are interested in an overall accurate collimation then do both alignments.

Seems as though this is what you do when using an autocollimator, don't you?


Well, that is not entirely true. It depends on the complete steps you follow.
A) If you use the single hole autocollimator without a cheshire then the recommended method is to start off with CDP by adjusting the secondary mirror then end it with stacking the center spot reflections by adjusting the primary mirror (last step is primary adjustment)
B) If you use the dual hole autocollimator without a cheshire then you stack P+2 by adjusting the secondary via offset pupil then stack reflections by adjusting the reflections via central pupil – re-iterate until both show classic alignments. Of course, you can always start with the above paragraph then fine tune with the steps outlined in this paragraph.
I know one vendor who includes instructions to align with the cheshire then end it with only the autocollimator (secondary adjustment) – no iterations. These are wrong instructions. You need to iterate.
Jason

#5 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6665
  • Joined: 21 Oct 2006
  • Loc: California

Posted 09 December 2012 - 02:07 PM

Also, when using a Chelshire eyepiece you set your secondary then finish collimation with the primary. Is there any harm in fine tuning your collimation with your secondary?


That's how I do it. (You can read all about my collimation technieques in the article, "A Beginner's Guide to Collimation" on my web site.) The key is to reserve that adjustment for the final tweak.

Regards,
Gary


Hello Gary,
Why not do both?
You never clarified which reflections you align when you fine tune collimation by only adjusting the secondary mirror using the cheshire eyepiece.
Jason

#6 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 09 December 2012 - 02:32 PM

...Can your center mark be in perfect center but still not be in the proper location for perfect collimation? If so, how would you achieve collimation? Trial and error?

I'm in agreement with Nils Olof's assessment.

Also, when using a Chelshire eyepiece you set your secondary then finish collimation with the primary.

By Cheshire eyepiece I assume you mean a combination Cheshire/sight tube collimation tool. If so, then yes--bottom edge of the sight tube to place the secondary mirror, sight tube cross hairs to align the focuser axis, and the final step, bright Cheshire ring to align the primary mirror axis.

Is there any harm in fine tuning your collimation with your secondary?

"Fine tuning" is usually reserved for the axial alignments. Most users adjust the secondary mirror tilt to fine tune the focuser axial alignment, and then finish with the primary mirror collimating screws to fine tune the primary mirror axial alignment. As Nils Olof noted, there's a reason for the alignment sequence. While small adjustments to the secondary mirror tilt impacts both axial alignments, adjustments to the primary mirror tilt have virtually no effect on the focuser axial alignment.

Seems as though this is what you do when using an autocollimator, don't you?

No. Assuming the primary mirror alignment has already been fine tuned with a Cheshire eyepiece, any obvious residual error visible in the autocollimator is likely to be focuser axial error. When the error is reduced at the secondary mirror, the earlier Cheshire primary mirror alignment will be changed, which forces the user to repeat the process until both axes are corrected to the resolution of both tools.

A more direct approach with an autocollimator is to carefully decollimate the primary mirror first, then correct the focuser axial alignment by adjusting the secondary mirror tilt, and finally, adjust the primary mirror tilt to correct the primary mirror axis.

Whether you're using a combination Cheshire/sight tube collimation tool, a laser collimator, or an autocollimator, I suggest that you always verify/correct both axes for best image performance.

#7 tag1260

tag1260

    Apollo

  • *****
  • Posts: 1175
  • Joined: 07 Oct 2012
  • Loc: Ohio, USA

Posted 09 December 2012 - 03:44 PM

By Cheshire eyepiece I assume you mean a combination Cheshire/sight tube collimation tool. If so, then yes--bottom edge of the sight tube to place the secondary mirror, sight tube cross hairs to align the focuser axis, and the final step, bright Cheshire ring to align the primary mirror axis.




Blackcat Chelshire, but a combination tool for preliminary secondary alignment.

If all of your rings appear to be concentric when finished, are you not still achieving the same end results?

Thanks for all the replies.

#8 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 09 December 2012 - 06:27 PM

The BlackCat Cheshire can easily read a primary mirror axial error as small as two hundredths of an inch (close enough for most scopes and most applications).

For establishing true centering of the secondary mirror under the focuser, you really need a sight tube. Without the sight tube, the focuser axis could easily be as much as a quarter inch or more out of alignment. For an 8-inch aperture, that's pushing the tolerance, even without coma correction. You might be able to do better than the other readers of this thread by reading the concentricities, but you'll certainly do much better with a sight tube.

Regarding concentricity--the only circles that must be concentric are the bottom edge of the sight tube (or the bottom edge of the focuser drawtube), the reflected edge of the primary mirror, and the reflection of the underside of the focuser (with the primary mirror center spot in the center). The real (viewed directly--not a reflection) edge of the secondary mirror and the edge of the silhouette reflection of the secondary mirror do not have to be concentric, although one can be depending on whether you choose an offset or centered alignment.

#9 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6665
  • Joined: 21 Oct 2006
  • Loc: California

Posted 09 December 2012 - 06:42 PM

If all of your rings appear to be concentric when finished, are you not still achieving the same end results?


Not all will look concentic. Just look at my avatar.
Refer to the very recent thread:
http://www.cloudynig...5554348/page...
Interestingly, I noticed you have several posts in the above thread.
Jason

Jason

#10 tag1260

tag1260

    Apollo

  • *****
  • Posts: 1175
  • Joined: 07 Oct 2012
  • Loc: Ohio, USA

Posted 09 December 2012 - 06:53 PM

Yes. That thread is what has me thinking about my collimation techniques. I seem to be achieving the same sort of result, as pictured in your picture of your XT10, but I have been tweaking my secondary to fine tune it at the end.

#11 tag1260

tag1260

    Apollo

  • *****
  • Posts: 1175
  • Joined: 07 Oct 2012
  • Loc: Ohio, USA

Posted 09 December 2012 - 07:10 PM

As for the center mark question, it's raining and foggy and I'm just sort of thinking about things.......You know how it goes!!!!!!

#12 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 09 December 2012 - 09:19 PM

As for the center mark question, it's raining and foggy and I'm just sort of thinking about things.......You know how it goes!!!!!!

I have to admit I've spent some time pondering the "center mark question".

It's remarkable that to date, it seems that tool (Cheshire) alignment always delivers the optimal star pattern when verified with a star test. It's remarkable that so many Newtonian primaries, whether amateur or professionally figured, seem to align properly with a well placed center spot. I mean, even the best available primary mirror isn't perfect!

I think, perhaps, the center spot works so well not because it so precisely represents the center of the primary mirror, but because it precisely represents the edge of the primary mirror. After all, what's the reference for placing the center spot? The edge!

Look at an 8-inch f/6 mirror. The central 4-inches occupies an area about 12 square inches while the outer 2-inches occupies an area of about 36 square inches. And while the outer ring operates at f/6, the central 4-inch diameter circle is operating at f/12 (it's spherical for all practical purposes)! And we haven't even considered that the shadow of the secondary mirror robs even more area from the central circle.

I think that most of the image information is coming from the outer edge/annulus of the primary mirror, so the critical alignment is the edge alignment, and the reference for that, is the center spot.

OK, let the flames begin!

#13 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1614
  • Joined: 26 Jul 2004

Posted 10 December 2012 - 06:09 AM

hi Vic,

It's remarkable that to date, it seems that tool (Cheshire) alignment always delivers the optimal star pattern when verified with a star test. It's remarkable that so many Newtonian primaries, whether amateur or professionally figured, seem to align properly with a well placed center spot. I mean, even the best available primary mirror isn't perfect!


it is indeed possible to make a paraboloid mirror with its optical center offset (for a Herschelian or as part of a segmented telescope), but it is very much more difficult. With symmetric figuring, the optical center will apparently automatically fall in the center of the blank (even if the bevel is asymmetric?).

I would say most of the information comes from the area between the rim and the center :grin: Even if a paraboloid mirror has all sorts of figuring errors (I won't go into Zernikes here), there is still a point in the focal plane where (low-order) astigmatism is zero, and a center mark is a way to make sure this point is centered in the focuser.

There are many posters here (and elsewhere) who claim they finish tool collimation by tweaking with a star collimation. Why? Do they actually see a star image that calls for a tweak?? If so, the conclusion would be that the marker is offset from the optical center - and judging from your experience, also from the geometric center. If it turns out that star tweaking is necessary, the thing to do would be checking the centering of the spot and likely move it.

Nils Olof

#14 howard929

howard929

    Member

  • *****
  • Posts: 3644
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 10 December 2012 - 10:14 AM

There are many posters here (and elsewhere) who claim they finish tool collimation by tweaking with a star collimation. Why? Do they actually see a star image that calls for a tweak?? If so, the conclusion would be that the marker is offset from the optical center - and judging from your experience, also from the geometric center. If it turns out that star tweaking is necessary, the thing to do would be checking the centering of the spot and likely move it.

Nils Olof


If there are many that do perform a star test and find a tweak is necessary it's because that many have used the most recommended method of placing the center spot that just doesn't work too well. At best, one has to get real lucky to achieve anything close enough to the center with one of those. I bought one, checked it for accuracy and threw it in the garbage.

#15 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 10 December 2012 - 01:31 PM

...If there are many that do perform a star test and find a tweak is necessary it's because that many have used the most recommended method of placing the center spot that just doesn't work too well. At best, one has to get real lucky to achieve anything close enough to the center with one of those. I bought one, checked it for accuracy and threw it in the garbage.

I'm not sure which method you're describing. I find that as long as the primary mirror center spot is placed within about 1/2mm of center (relative to the edge), I get excellent axial alignment with the tools I use. I've used just about every method available, from a full diameter paper disk folded in quarters and then nipping off the vertex, to a marked spot measured with a precision rule, to the acetate overlay provided by CATSEYE. They all give good results if the person using them is diligent. And many economy mirrors that come pre-center spotted from the manufacturer are often corrected to significantly better precision with these methods.

#16 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 10 December 2012 - 02:19 PM

...Even if a paraboloid mirror has all sorts of figuring errors (I won't go into Zernikes here), there is still a point in the focal plane where (low-order) astigmatism is zero, and a center mark is a way to make sure this point is centered in the focuser.


I guess my point was that the "point in the focal plane" is defined more by the outer zones than the central zones, the center mark is simply a reference. When I used a zonal mask to figure the mirrors I polished, the mask was referenced from the edge. I'm unfamiliar with interferometry (before my time as an ATM), but I'm unaware of any mirror maker using this testing method to define the center of the paraboloid. I understand what you're saying, but I still feel that the central 4-inch diameter has only 1/2 the resolving power of a full aperture 8-inch primary mirror (not considering CO), so the outer annulus is responsible for "the rest of the information". Until the two zones are working together in concert, the central 4-inch diameter's tolerance to decentering is much more relaxed than the full aperture tolerance.

There are many posters here (and elsewhere) who claim they finish tool collimation by tweaking with a star collimation. Why?


I think in many cases, it's because they are too quick to settle for "good enough". Then there's sloppy focusers, gravitational flexure, and other inadequate mechanical designs. I also see some posters chasing the wrong alignment signature, often at the expense of the more critical axial alignments.

Do they actually see a star image that calls for a tweak?? If so, the conclusion would be that the marker is offset from the optical center - and judging from your experience, also from the geometric center. If it turns out that star tweaking is necessary, the thing to do would be checking the centering of the spot and likely move it.

Until recently I thought most collimation cap pupils were reasonably well centered--but I've been shown that even with a simple collimation cap, you should verify first.

FWIW, when I observe, every star in every field of view is a litmus test for seeing, equilibration, optical performance, and collimation. If something's amiss, the collimation is the easiest to verify (with the proper tools) and correct if necessary.

#17 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1614
  • Joined: 26 Jul 2004

Posted 10 December 2012 - 02:56 PM

I guess my point was that the "point in the focal plane" is defined more by the outer zones than the central zones,


Of course I agree. Take a f/4.5 mirror as example - the <=1/14 wave (front) RMS zone at the focal plane has a radius of about 1 mm. Say you can tell to within half of this (you may do better, but for the sake of argument)- this means an error of at most 1 mm at the primary.
Stop it down to 0.707 of full diameter - ít is now f/6.36, and the error margin is 2.83 times as large (proportional to the 3rd power of f/).
Thus, if you want to verify collimation with any precision, you need the mirror out to the edge. But I can't say offhand what to expect from the obstructed zone - likely, very little.

Even if it takes a bit of care to place the spot accurately, measuring its centering with a ruler ought to be trivial, I think.

Nils Olof

#18 howard929

howard929

    Member

  • *****
  • Posts: 3644
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 11 December 2012 - 10:52 AM

I'm not a fan of referencing the edges of a primary mirror with a circular mask by eye or with placing a new spot on the mask the same way. If memory serves and yes, I did throw out the one I bought, the line thickness on the CatsEye mask was .279mm, half of the .5mm tolerance I wanted to be under. Sure, it needs lines that can be seen but therein lies part of the problem I found with it. IME, using that mask as instructed is too open to placement errors of the new center spot and rotational errors while placing it over the mirror. Then pressing it down by hand even with the block method doesn't ensure accuracy.

It's flat when it needs to be shaped like the mirror is. It was loose and slid around too easily. It wasn't a ridged system that has any aspect built into it that causes or forces accuracy. Not to within .5 mm. IMO.

#19 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 11 December 2012 - 11:18 AM

I'm not a fan of referencing the edges of a primary mirror with a circular mask by eye or with placing a new spot on the mask the same way. If memory serves and yes, I did throw out the one I bought, the line thickness on the CatsEye mask was .279mm, half of the .5mm tolerance I wanted to be under. Sure, it needs lines that can be seen but therein lies part of the problem I found with it. IME, using that mask as instructed is too open to placement errors of the new center spot and rotational errors while placing it over the mirror. Then pressing it down by hand even with the block method doesn't ensure accuracy.

It's flat when it needs to be shaped like the mirror is. It was loose and slid around too easily. It wasn't a ridged system that has any aspect built into it that causes or forces accuracy. Not to within .5 mm. IMO.

Interesting.

I've used the CATSEYE template with good results on numerous occasions and on a variety of large aperture, fast focal ratio mirrors. I didn't have the same problem you had with the line thickness as the various diameters rarely lined up exactly with the mirror blank edge. As long as the line falls inside or outside the edge, it's pretty easy to judge concentricity to less than 1/2mm (probably closer to 1/4mm). When the sagittal depth exceeds more than about 3mm, I usually try to get a handle on the centering first. The CATSEYE template helps here too as the center is accurately perforated. This allows the user to mark the center with a sharpie marker, which provides another opportunity to assess (and correct if needed) the center mark. Once the center has been accurately marked, the template can be used to transfer a perforated CATSEYE triangle or HotSpot with very good accuracy. Even using this procedure, I always verify centering with a precision rule, and when I'm satisfied with the placement of the center spot, then I firmly press it onto the mirror surface.

I can also place the spot without the CATSEYE template, but it's more tedious and requires additional handling of the center spot close to the mirror surface. While an 8-inch aperture isn't too hard, a 16-incher can try your patience, and a 24-incher can try your nerves!

As I noted in my earlier post, regardless of the method, precise centering requires diligence.

#20 howard929

howard929

    Member

  • *****
  • Posts: 3644
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 11 December 2012 - 12:10 PM

I too placed a sharpie mark where I thought the center was using the thinnest marker I could buy, .5mm. It took 4 tries to get it within 1mm using a photographic method for verification that's exceedingly accurate to within .007mm.

I'm going to go out on a limb here and suggest that one cannot accurately verify the results you mention with any consistency using a rule or standard caliper to within .5mm. The rule relies on hand and eye placement since it's flat and the mirror isn't. A standard caliper slides off the end of the mirror when the other end of it dips down towards the center of the mirror. IMO this entire topic is a disaster. WE shouldn't have to become engineers because THEY don't etch out a spot in the coatings exactly at the center of these damn mirrors during the coating process.

Accuracy to .5mm? Gotta be kidding. I'm used to anything like 1/4" is fine and dandy and I hated the fact that I had to come up with some way to get it done.

#21 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 23135
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 11 December 2012 - 12:27 PM

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.
In the center of the disc is a spring-loaded plunger with the center marker on the bottom end, below the disc. The plunger is pushed down toward the mirror, sticking the center marker on the mirror.
To keep the plunger from wiggling as it descends toward the mirror, the center of the disc is built up to a thickness of about 1/2" where the plunger passes through it, preventing any lateral wiggle when the plunger is pushed down.
The procedure would take only a second, and noting would sag into the mirror, and the positioning of the marker on the bottom of the plunger would be the only source of inaccuracy.
I could see this resulting in accuracies of small fractions of a millimeter in the positioning of the center marker.

If I have time next year, I may make one of these. If the disc were clear, with diameter markers, it could be made for many sizes of mirror.

#22 howard929

howard929

    Member

  • *****
  • Posts: 3644
  • Joined: 02 Jan 2011
  • Loc: Low End of High Ground

Posted 11 December 2012 - 12:46 PM

Yes, that's an excellent idea. A ridged system that forces accuracy is always going to best a hand and eye method.

#23 tag1260

tag1260

    Apollo

  • *****
  • Posts: 1175
  • Joined: 07 Oct 2012
  • Loc: Ohio, USA

Posted 11 December 2012 - 01:14 PM

Sounds like an excellent idea. I think there's a quote somewhere-"If you build it they will come".
You are going to market this when you do, aren't you?

#24 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4792
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 11 December 2012 - 01:28 PM

...WE shouldn't have to become engineers...Accuracy to .5mm? Gotta be kidding...

Well, I'm not kidding, and I'm not an engineer. :shrug:

Final measurement using a millimeter rule from the edge of the mirror to the vertex of the triangle gives a read that's an easy 1/2mm tolerance. I use a block at the edge of the mirror to keep the rule "stop" consistent and I measure all three vertices. But, ymmv...

#25 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1614
  • Joined: 26 Jul 2004

Posted 11 December 2012 - 04:35 PM

I'm going to go out on a limb here and suggest that one cannot accurately verify the results you mention with any consistency using a rule or standard caliper to within .5mm. The rule relies on hand and eye placement since it's flat and the mirror isn't.



You shouldn't have to be an engineer to use a flexible ruler of some kind - I take two plastic rulers taped to a cross, and with the marker taped in the center (the free quadrant). Two strips of cardboard would do as well, no doubt. For verifing, a metal tape measure would do (you could put a few pieces of tape not to risk scratching). web pic






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics