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

RC collimation with no laser or stars

  • Please log in to reply
263 replies to this topic

#1 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 29 July 2021 - 06:32 PM

The Holy Grail of collimation for me is a simple and cheap procedure which can be done during the day, inside, and produces optimum results without a star test.  See how close you think I've gotten here.

 

My previous effort gave me good results, but I tried it again recently with a different laser and it didn't work as well.  It didn't converge on a final result well and it resulted in a mediocre collimation, I think because it was hard to interpret the laser position.

 

So I went back to the internet and found the "hall of mirrors" test here, attributed to Jared Wilson.  The test itself is not a full collimation, but it is a very sensitive and precise way to verify that your two mirrors are coaxial.  I put it together with a Cheshire and tilting focuser to produce a procedure which is quicker, easier, cheaper, more precise, and didn't require a final star test.  I'll describe it below, but first some philosophy.

 

Leery of lasers I have become leery of lasers in collimation, either the spot or the hologram varieties.  "What's not to like about a laser?", I hear you ask.  "They're so straight!"  Very true.  But that does you no good unless they are exactly centered and aligned when installed in your scope.  The best ones are good at this, but there are still problems.  A semiconductor laser is a bar, not a point.  In order to approximate a point light source you need to stop it down with a circular aperture, but that's not perfect.  Especially with hologram patterns, any residual non-point shape is magnified and results in a pattern which is not rotationally symmetrical.  The lines or circles it projects vary in thickness, which makes it impossible to judge position accurately.  The hologram itself also needs to be exactly collimated with the laser.  If it isn't, it's as bad as if your laser were off center.  And both holograms and points suffer from shimmer, dazzle, and flare, especially when reflected off a convex secondary.  It's difficult to decide where the exact center is, especially when the flare is asymmetrical.  You can sometimes reduce your brightness to tame this, but then the image becomes faint and difficult to judge.  A plain old Cheshire is much easier and cheaper to machine to at least as good tolerance as a well-adjusted laser, and the visual feedback is clearer.

 

Star tweak?  A final defocused star test is usually the gold standard for collimation.  It's sensitive and tests the system with point light sources.  I've always found it to be nearly impossible though.  Finding and keeping a rich enough star field near the zenith not long after sunset, waiting 30 s or so for each defocused image, and trying to interpret the fuzzy results and what to tweak is difficult, time-consuming, and frustrating.  If you had a daytime, indoor procedure where the smallest adjustments have a visible effect and the metrics were optically correct and easy to discern, how could a star test improve on that?

 

(next slide please)


Edited by TinySpeck, 30 July 2021 - 11:56 AM.

  • Sky King, MikeECha, sunrag and 3 others like this

#2 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 29 July 2021 - 06:32 PM

So, on to the procedure itself.  You'll need only a Cheshire, hex wrenches, maybe a flashlight, and your eyeballs.  You'll also need adjustable focuser tilt.  This procedure works well in broad daylight, but may also be done indoors while pointing the telescope at a light wall.  The telescope can remain on its mount or on the bench.

 

My scope is a GSO-built 8" Ritchey-Chrétien.  My Cheshire is a 1.25" model in a self-centering 2" adapter from AstroMania.  It looks like this:
cheshire.jpg

The margin around the reflective plate appears as a dark circle when you peer through the pinhole, and that is a metric in this procedure.

 

First, your focal length must be correct (1600 mm for the GSO 8" RC).  See step 1 (only) in this post if you need to adjust.

 

Next, mount the OTA back spacer(s) and focuser as you will have them for imaging, and set the focuser for its nominal infinity focus. 

 

Now the procedure:

 

figure.png

 

1. Reset the focuser tilt to zero.  For a Moonlite you should see no looseness in the focuser and it should rotate with a slight drag.

 

2. Adjust the secondary: with moderate light into the Cheshire cutout and looking through the pinhole, adjust the three secondary tilt setscrews until the center dot (which is the reflection of the Cheshire pinhole) is exactly centered in the secondary donut.  Tighten a secondary setscrew (CW) to move the dot in the direction of the setscrew.  Do not touch the center Philips screw.  Loosen the opposite setscrew(s) slightly before tightening a given setscrew, and keep all setscrews snug during the process.

 

3. Adjust the primary (coarse): The coarse adjustment only needs to be done once, typically.  With a strong light into the Cheshire cutout so it illuminates the body of the secondary assembly from the inside, center the secondary body inside the inner edge of the primary.  (At least I think this is the secondary body.  It's the gray ring as shown in the figure anyway, and you can watch it move as you adjust the primary.)

 

Adjust the primary (fine): use the Hall of Mirrors test below.

 

Only adjust two of the silver primary setscrews so your focal length isn't altered in the process.  Put a piece of tape over one of them to keep your mitts off.  Loosen the other two black setscrews slightly, adjust the silver setscrews, and retighten the black setscrews evenly.

 

4. Adjust focuser tilt to center the Cheshire margin ring in its ring of light.  This also centers the optical axis in the tube and eliminates the asymmetrical outer light margin shown in the figure.  You tighten a Moonlite focuser setscrew (CW) to push the margin ring away from it.  Keep all setscrews snug, loosening and tightening as required.

 

5. Iterate steps 2 - 4 until converged (which happens quickly).  Be as picky and perfectionist as you possibly can be.  Every circular thing you see in the Cheshire pinhole should be exactly concentric.  Force yourself to look all around each circle.  Even the tiniest imperfections will show up in your image.  My first attempt at this wasn't quite as perfect as I thought it should be, and when I looked again I found and fixed some minuscule imperfections which improved the results visibly, to what you see in the next post below.  My wife swears this is the first time I have EVER not been picky enough.

 

Hall of Mirrors test ----------------------------------------------------------

hall of mirrors.png

David Cortner credits Jared Wilson for this, and here is my version.  Stand at arm's length from the end of the OTA, hold onto the OTA to keep yourself as still as possible, and look grazing down the sides of the secondary past each vane.  Reflections should appear as a straight tube with "scallops", as shown in David's picture (the "tube" is not quite visible).  Each vane, its reflection, and the middle of the tube & scallops at infinity should all line up, and each vane should look like the others.  Tighten a primary silver screw (CW) to pull the reflected "tube" toward it.

 

(next slide please)


Edited by TinySpeck, 30 July 2021 - 01:53 PM.


#3 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 29 July 2021 - 06:32 PM

Results

 

I generally image using a field flattener / focal reducer (Apex ED-L), which requires a different optical train and spacing than the scope at its native focal length.  The FF/FR doesn't defocus the same as the plain scope either – all defocused stars "point" toward or away from the image center and look themselves off-center -- so a defocused image is not helpful.  The primary tools I have to assess image quality are therefore PixInsight's AberrationInspector and FWHMEccentricity scripts.  The danger with this approach is I'm adding a FF/FR to a scope which has been collimated without it, so if problems appear it may not be clear who is at fault.

 

I took about 200 guided 60 s subs of M29 as it crossed near the zenith and used PixInsight's SubframeSelector analysis to pick an optimal sub with low FWHM and eccentricity.  This eliminates atmospheric and guiding effects as much as possible and should provide a clearer picture of just the optics.

 

Here is the optimal sub with the collimated scope and the ED-L, no cropping, no calibration or processing other than PixInsight CosmeticCorrection and Debayering, straight out of the ZWO ASI294MC camera.  The camera sensor is 19.1 x 13.0 mm, slightly less than APS-C size.  The image has been stretched with PixInsight's default ScreenTransferFunction.

 

stars.jpg

 

(next slide please)


Edited by TinySpeck, 29 July 2021 - 06:52 PM.


#4 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 29 July 2021 - 06:52 PM

And a mosaic of the image periphery:

 

mosaic.jpg

 

And here are the PixInsight contours:

 

contours.png

 

The FWHM isn't perfectly flat, but it is symmetrical and only increases about 0.6 pixel from the center to the corners.  This may be due to imperfect field flattening.  The threshold for visual detection of eccentricity is about 0.42 according to PixInsight, and the entire image falls under that.  The stars do look round in all tiles of the mosaic, too.

 

I'm pretty happy with this.  It only requires a Cheshire and no star testing.  You do have to be very careful to get everything as perfectly centered as you can, but that can be done in about 20 min, and during the daytime.  I'm experimenting with a pinhole video camera now to blow the Cheshire image up on the computer screen and overlay circular graticules.  That should take the guesswork out of some of the procedure.  Any ideas on how to capture the Hall of Mirrors images on the computer screen?


Edited by TinySpeck, 29 July 2021 - 06:56 PM.

  • starhunter50, MikeECha and astrobananeck like this

#5 PaulE54

PaulE54

    Vostok 1

  • -----
  • Posts: 150
  • Joined: 17 Nov 2019
  • Loc: NW England

Posted 30 July 2021 - 02:05 AM

Just a suggested refinement. On my RC6, I found that if the primary mirror lock ring was too loose, the primary mirror could flop. If too tight, I'd get "pinched optics" effects. Similar problems can happen if the rubber O ring under the primary mirror lock ring has seen better days. 

So it would be worth checking the hall of mirrors effect with the scope rotated to different orientations, to check that nothing in the scope optical chain is loose. Otherwise results are likely to be perfect with the scope lying horizontally, and then "off" to some degree when the scope is pointing up. 

 

Clear skies

 

Paul


  • TinySpeck and nebulachadnezzer like this

#6 dg401

dg401

    Vostok 1

  • -----
  • Posts: 101
  • Joined: 11 Jul 2021

Posted 30 July 2021 - 03:49 AM

I got quick convergence with your procedure.  It's too hazy out to test under the stars, so I'll have to wait to evaluate performance.

 

To satisfy my curiosity, I inserted my Farpoint laser after completing your procedure.  The laser hit the marked center of the secondary perfectly and also reflected back perfectly.  This is notable because both your procedure and the laser give the same visual result, and the laser results are awful under the stars.  I'm hoping there's some secret sauce to your procedure that the laser fails to address. 


  • TinySpeck likes this

#7 Terry White

Terry White

    Viking 1

  • *****
  • Posts: 561
  • Joined: 13 Sep 2017
  • Loc: Knoxville, Tennessee.

Posted 30 July 2021 - 09:27 AM

The daylight collimation procedure for RCs, or any other scope for that matter, has already been developed. It uses the gold-standard nighttime star testing method but replaces the star with the sun's bright reflection off of a polished sphere placed in the telescopes "far field". It's described on page 88 and Appendix F in Suiter's book "Star Testing Astronomical Telescopes." The are many refences to it here on Cloudy Nights.


Edited by Terry White, 30 July 2021 - 10:06 AM.


#8 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 305
  • Joined: 04 Jan 2021

Posted 30 July 2021 - 09:43 AM

The Holy Grail of collimation for me is a simple and cheap procedure which can be done in daylight and produces optimum results without a star test.  See how close you think I've gotten here.

[Bringing my discussion over from another thread at your request.]

Is it worth noting that the diagram showing the Cheshire reflection isn't exactly what you'll see through the Cheshire? At least if I recall correctly you won't see the *outer* edge of the primary as depicted without removing the baffle, which is a procedure I don't recommend anyway. Your notes and call-outs are correct, but that one aspect of the illustration may confuse someone.
 

There's always the risk that the Cheshire is off-center. I can't decide if the rubber gasket compression style self-centering adapters are a good or bad thing. I have the Hotech one that came with a laser but I've put a Cheshire into it before. I tend to prefer a close-tolerance milled aluminum 2" to 1.25" adapter that came with one of my Moonlite focusers, but still nothing is perfect.
 

I'd like to see a Cheshire with a threaded front so I could screw it into the focuser (with appropriate adapters). That's one somewhat nice thing about the Russian-made adapters you can buy for the Takahashi scope, which is sort of a magnified concept of the same thing (but unfortunately also prone to internal tilt in my experience).

I haven't seen the Hall of Mirrors effect for RCs descried before. That's very interesting. I checked my RC10 and results look good. However, I would like to suggest an improvement to your text around that section to clarify what should be seen and what adjustments to make to get it there.

You wrote, "Tighten a primary silver screw (CW) to pull the reflected 'tube' toward it." Could you explain in more detail what is being pulled which direction, and what screw would correspond to that? I *think* you are saying that if you see asymmetry in the Hall of Mirrors that an adjustment is required to make it symmetrical but I'd appreciate it if you could elaborate on that step.
 

Thanks for putting this together!


  • TinySpeck likes this

#9 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 305
  • Joined: 04 Jan 2021

Posted 30 July 2021 - 09:51 AM

The daylight collimation procedure for RCs and any other scope, for that matter, has already been developed. It uses the gold standard nighttime star testing method but replaces the star with the sun's bright reflection off of a polished sphere placed in the telescopes "far field". It's described on page 88 and Appendix F in Suiter's book "Star Testing Astronomical Telescopes." The are many refences to it here on Cloudy Nights.

While it is true that Suiter's process is a daylight procedure, it's most certainly not a bench procedure unless your bench has a view out a window to a distant target. Suiter's process suffers from a similar issue as other artificial star processes in that they are extremely limited by the distance to and visibility of the artificial point source. If these are not sufficiently far away, and do not present sufficient contrast vs the background, they're not accurate enough.

As the author noted, and as he and I discussed in a prior thread, the Holy Grail for RC collimation is a procedure that could be completed entirely on the bench, i.e. in a normal indoor setting (not, say, a warehouse-sized building).


  • TinySpeck and xthestreams like this

#10 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 10:04 AM

The daylight collimation procedure for RCs and any other scope, for that matter, has already been developed. It uses the gold standard nighttime star testing method but replaces the star with the sun's bright reflection off of a polished sphere placed in the telescopes "far field". It's described on page 88 and Appendix F in Suiter's book "Star Testing Astronomical Telescopes." The are many refences to it here on Cloudy Nights.

Thanks, Terry.  I don't have enough room for a good artificial star test.  I tried putting one across the street one time and the heat waves rising from the street made a mess of the image, too.


  • MikeECha likes this

#11 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 10:06 AM

I got quick convergence with your procedure.  It's too hazy out to test under the stars, so I'll have to wait to evaluate performance.

 

To satisfy my curiosity, I inserted my Farpoint laser after completing your procedure.  The laser hit the marked center of the secondary perfectly and also reflected back perfectly.  This is notable because both your procedure and the laser give the same visual result, and the laser results are awful under the stars.  I'm hoping there's some secret sauce to your procedure that the laser fails to address. 

Since the laser is a little muzzy, maybe the fact that it looks the same hides a greater precision without it.  Also, I hope you were picky enough!  grin.gif  Please let us know how it works for you.


Edited by TinySpeck, 30 July 2021 - 10:18 AM.


#12 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 10:19 AM

Is it worth noting that the diagram showing the Cheshire reflection isn't exactly what you'll see through the Cheshire? At least if I recall correctly you won't see the *outer* edge of the primary as depicted without removing the baffle, which is a procedure I don't recommend anyway. Your notes and call-outs are correct, but that one aspect of the illustration may confuse someone.

Good point.  The outer bright ring may or may not be visible as shown.  As I recall, I could see some brightness around the outer periphery, but it wasn't the crisp outline shown in the figure.  At any rate, make everything concentric!

 

You wrote, "Tighten a primary silver screw (CW) to pull the reflected 'tube' toward it." Could you explain in more detail what is being pulled which direction, and what screw would correspond to that? I *think* you are saying that if you see asymmetry in the Hall of Mirrors that an adjustment is required to make it symmetrical but I'd appreciate it if you could elaborate on that step.

It's easier to see than describe.  Yes, this step is intended to make the Hall of Mirrors perfectly symmetrical.  The "tube" is the repeated reflection of the side of the secondary, I believe; at any rate it looks like a tube extending into infinity.  If that "tube" appears to be bending in one direction rather than receding straight back, tighten the primary screw on the side it's bending away from (e.g. bending "up"?  Tighten the "down" screw(s).)  You may need to loosen/tighten both available primary screws to achieve your desired effect.  I found it easier to do them one at a time, though, so I could be clear on their separate effects.


Edited by TinySpeck, 30 July 2021 - 01:57 PM.

  • nebulachadnezzer likes this

#13 Terry White

Terry White

    Viking 1

  • *****
  • Posts: 561
  • Joined: 13 Sep 2017
  • Loc: Knoxville, Tennessee.

Posted 30 July 2021 - 10:25 AM

While it is true that Suiter's process is a daylight procedure, it's most certainly not a bench procedure unless your bench has a view out a window to a distant target.

Well, the title of this thread is "A daylight RC collimation procedure" not an indoor bench RC collimation procedure. Daylight implies outdoor.

Suiter's process suffers from a similar issue as other artificial star processes in that they are extremely limited by the distance to and visibility of the artificial point source. If these are not sufficiently far away, and do not present sufficient contrast vs the background, they're not accurate enough.

The Sun's reflection off of a polished sphere will be brighter than any incoherent illuminated pinhole source, so visibility is not an issue. Also, black fabric drop cloth can be used to get the desired contrast. Yes, for a large scope this means a sphere has to be placed a few hundred feet away, yet many seem to be able to do the sphere test quite well, despite the distances required. YMMV.


Edited by Terry White, 30 July 2021 - 10:53 AM.


#14 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 305
  • Joined: 04 Jan 2021

Posted 30 July 2021 - 10:27 AM

It's easier to see than describe.  Yes, this step is intended to make the Hall of Mirrors perfectly symmetrical.  The "tube" is the repeated reflection of the side of the secondary, I believe; at any rate it looks like a tube extending into infinity.  If that "tube" appears to be bending in one direction rather than receding straight back, tighten the primary screw on the side it's bending away from (e.g. bending "up"?  Tighten the "down" screw(s).)  You may need to loosen/tighten both available primary screws to achieve your desired effect.  I found it easier to do them one at a time, though, so I could be clear on their separate effects.

I get it, but I do think this could do with a clearer explanation. It may in fact require an illustration.

Thanks for introducing me to the effect either way. I think you're onto something with this process. I haven't had a chance to test it, and don't expect to so long as my RCs remain in good collimation, but in principle everything you describe here seems to be on track.

I have no way to judge how close this gets to perfect collimation, but it certainly seems like it will get you very close. Even if this gets you 95% there, that's a great head start on a star collimation.



#15 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 305
  • Joined: 04 Jan 2021

Posted 30 July 2021 - 10:44 AM

YMMV.

Yes, it may. Quite literally.

Not everyone has a setting allowing placement of a target hundreds of feed away. And as the author has noted, placement of a distant, reflective target is affected by conditions such as heat shimmer. It also requires clear skies.

The point of the procedure described here is that it's a bench process that can be done entirely within about one meter of the scope. It can be done on cloudy/rainy days over overnight if necessary. Perhaps the author may wish to correct the title and description to address the misunderstanding caused by the use of the word "daylight."


  • Terry White and TinySpeck like this

#16 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 11:14 AM

Well, the title of this thread is "A daylight RC collimation procedure" not an indoor bench RC collimation procedure. Daylight implies outdoor.

 

The point of the procedure described here is that it's a bench process that can be done entirely within about one meter of the scope. It can be done on cloudy/rainy days over overnight if necessary. Perhaps the author may wish to correct the title and description to address the misunderstanding caused by the use of the word "daylight."

Good feedback.  I've requested a moderator to change the title to "RC collimation with no laser or stars".  I hope that's more to the point, and sorry for the confusion.


Edited by TinySpeck, 30 July 2021 - 11:15 AM.

  • MikeECha and nebulachadnezzer like this

#17 Terry White

Terry White

    Viking 1

  • *****
  • Posts: 561
  • Joined: 13 Sep 2017
  • Loc: Knoxville, Tennessee.

Posted 30 July 2021 - 12:36 PM

Gerrit, I would love for there to be a cheap "holy grail" gold-standard for bench collimating my RC. So help me understand something. RCs are well-known to have strong optical axes due to the strong optical corrections that require using hyperboloids on both the primary and secondary mirror. It's my understanding that the mirrors of all the GSO RC store brands are NC machined, figured and polished. The optical axes of these mirrors are determined in the final stages of mirror figuring and are quite hard to measure. However, secondary mirror center spots, made after the mirrors have been polished and coated, are easily made by marking the mirrors while they're rotating it in a fixture. This secondary center spot is a geometric center and not an optical center.

 

The primary and secondary mirror's location on the geometric axis of the OTA are determined by the tight NC machine tolerances specified on the fabricated support hardware. No provision for making any offset adjustment, other than to perhaps rotate the secondary in it's holder, is provided. Also, both the primary and secondary baffles are similarly located on the geometric axis by the same tolerances on the machined support hardware.

 

Optical RC collimation is achieved when the optical axes of both the primary and secondary mirrors are co-linear, combined with an optimal mirror spacing that nulls out spherical aberration. So is your modified hall-of-mirrors technique an optical collimation technique, or a geometric alignment technique? Does it rely on centering mirror baffles, centering edges of the primary or secondary mirrors, or secondary center spots? If so, it's seems to be a variation on the collimation telescope method, which is just a geometric alignment technique. The guide for the DSI gold-standard technique goes into some detail about the efficacy of geometric-based collimation tools and how they may make your collimation worse. To quote (traditional method = laser collimators and collimating telescopes):

 

"The traditional method of collimating references several physical points within the system and relies on certain assumptions that may or may not be true. First, the method requires a center spot (or circle) on the secondary mirror. This spot is usually located in the physical center of the secondary mirror. It assumes the mechanical center of the secondary mirror is also the optical center of the secondary mirror. Further, the method references the edges of both the primary and secondary baffles. It assumes both baffles are coaxial with the OTA and that the system optical axis will eventually be coaxial with these as well. Neither of these will be the case to at least some degree. That is why it is often reported that when a scope is collimated with the traditional method that it still does not perform well. Conversely, a well collimated scope may not look collimated when viewed with a collimation telescope."

 

and,

 

"The mechanical centers of both the primary and secondary mirror are almost never the optical center. They can vary by a few thousands of an inch to as much as a tenth or more. This is often not taken into account when the instrument is assembled. The procedure presented here is very tolerant of this. Optical analysis as well as practical experience shows that this procedure produces good results in the presence of these types of issues."

 

I, like you, find the star test difficult to do, and would love to be able to use a better method. I'm just having trouble seeing how your modified hall-of-mirrors technique will replace star testing, which is your original stated goal of this thread.


Edited by Terry White, 30 July 2021 - 12:53 PM.

  • xthestreams and nebulachadnezzer like this

#18 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 01:40 PM

Gerrit, I would love for there to be a cheap "holy grail" gold-standard for bench collimating my RC. So help me understand something. RCs are well-known to have strong optical axes due to the strong optical corrections that require using hyperboloids on both the primary and secondary mirror. It's my understanding that the mirrors of all the GSO RC store brands are NC machined, figured and polished. The optical axes of these mirrors are determined in the final stages of mirror figuring and are quite hard to measure. However, secondary mirror center spots, made after the mirrors have been polished and coated, are easily made by marking the mirrors while they're rotating it in a fixture. This secondary center spot is a geometric center and not an optical center.

 

The primary and secondary mirror's location on the geometric axis of the OTA are determined by the tight NC machine tolerances specified on the fabricated support hardware. No provision for making any offset adjustment, other than to perhaps rotate the secondary in it's holder, is provided. Also, both the primary and secondary baffles are similarly located on the geometric axis by the same tolerances on the machined support hardware.

 

Optical RC collimation is achieved when the optical axes of both the primary and secondary mirrors are co-linear, combined with an optimal mirror spacing that nulls out spherical aberration. So is your modified hall-of-mirrors technique an optical collimation technique, or a geometric alignment technique? Does it rely on centering mirror baffles, centering edges of the primary or secondary mirrors, or secondary center spots? If so, it's seems to be a variation on the collimation telescope method, which is just a geometric alignment technique. The guide for the DSI gold-standard technique goes into some detail about the efficacy of geometric-based collimation tools and how they may make your collimation worse. To quote (traditional method = laser collimators and collimating telescopes):

 

"The traditional method of collimating references several physical points within the system and relies on certain assumptions that may or may not be true. First, the method requires a center spot (or circle) on the secondary mirror. This spot is usually located in the physical center of the secondary mirror. It assumes the mechanical center of the secondary mirror is also the optical center of the secondary mirror. Further, the method references the edges of both the primary and secondary baffles. It assumes both baffles are coaxial with the OTA and that the system optical axis will eventually be coaxial with these as well. Neither of these will be the case to at least some degree. That is why it is often reported that when a scope is collimated with the traditional method that it still does not perform well. Conversely, a well collimated scope may not look collimated when viewed with a collimation telescope."

 

and,

 

"The mechanical centers of both the primary and secondary mirror are almost never the optical center. They can vary by a few thousands of an inch to as much as a tenth or more. This is often not taken into account when the instrument is assembled. The procedure presented here is very tolerant of this. Optical analysis as well as practical experience shows that this procedure produces good results in the presence of these types of issues."

 

I, like you, find the star test difficult to do, and would love to be able to use a better method. I'm just having trouble seeing how your modified hall-of-mirrors technique will replace star testing, which is your original stated goal of this thread.

I was concerned about this too, and thought that my first less-than-perfect results were due to the irreducible difference between the mechanical vs optical axes you describe.  When I was able to then discern the tiniest imperfections using the technique here, correct them, and see my results improve significantly, though, I became a believer.

 

It's possible that my scope is just a perfect specimen, with optical and mechanical axes perfectly aligned, but I doubt that!  Given the good quality I was able to achieve, then, I have to question some of DSI's assumptions.  Maybe nowadays (or by some GSO magic) the secondary donut really is well aligned with the optical axis, and maybe the alignment between the optics and the mechanicals doesn't matter to this method.

 

The "tube" and scallops you see in the Hall of Mirrors test look to me like reflections of the secondary cell (the mirror holder).  If so, it might be that all you need is good alignment of the secondary mirror to its cell for HOM to succeed -- you aren't relying on alignment to the tube.  Or maybe it's even better: it may be that misalignment of the secondary and its cell would only result in a slightly different look to the HOM effect but wouldn't affect the results.  My hands are getting a little wavy here.

 

Perfect concentricity in the Cheshire view is important too.  The HOM and Cheshire both converged well for me (and @dg401 reports this above, too).  I will ponder the interaction of the three adjustments and see if I can convince myself that they only depend on optical axes.

 

I'm hoping others will try this technique, if only to prove that I have a perfect scope.  That would increase its resale value.  grin.gif


Edited by TinySpeck, 30 July 2021 - 01:43 PM.

  • nebulachadnezzer likes this

#19 Juicy6

Juicy6

    Explorer 1

  • -----
  • Posts: 67
  • Joined: 21 Mar 2014

Posted 30 July 2021 - 02:00 PM

I was concerned about this too, and thought that my first less-than-perfect results were due to the irreducible difference between the mechanical vs optical axes you describe.  When I was able to then discern the tiniest imperfections using the technique here, correct them, and see my results improve significantly, though, I became a believer.

...or you experience the 'squinted collimation' David07 is describing in his post http://www.cloudynig...read/?p=9224176

 

His method with a cheap 58 mm cardboard disc works like a charm:

 

"In my experience, the collimation of the GSO RCs, that is the design where the focuser extension tube is attached to the primary mirror support, is extraordinarily difficult. Any method based on using a laser or Cheshire inserted into the focuser is immediately compromised because of the lack of a fixed reference in the telescope design. There are unknown alignment errors between the primary mirror and it’s holder and the mirror holder and the focuser. Moreover, moving the primary mirror, when collimating the scope, also changes the pointing direction of the focuser. I spent 18 months fiddling about with my RC8 until I read the Austria Teleskop blog and that gave me the idea for the collimation method I now use.

I began by taking the scope to an optician friend with an optical bench and artificial star.  The first discovery was that after I had spent weeks of adjusting the secondary to primary separation to get the specified focal length of 1624mm, a Ronchi grating test quickly revealed that the scope was overcorrected: the mirrors were too far apart. We adjusted the mirror separation until we got a properly corrected image. The focal length was 1660mm, as it still is. So problem number one, the actual focal length might not be as advertised.

 

To collimate the scope:

 

With the scope on a bench, I removed the focuser and extension tubes. I then measured, with a calliper, the inside diameter of the central hole in the primary mirror holder. I cut a piece a styrene about 2mm thick to exactly fit the hole and drilled a 1mm hole in its centre. Now gently fit the styrene disc into the central hole of the primary mirror holder.

 

Next, mark the position of the central screw holding the secondary mirror. A little dab of white acrylic paint, for example. Also, mark the side of the secondary boss and mirror holder. This will enable you to replace the secondary exactly in the correct place and not alter the scope focal length.

 

Carefully unscrew the secondary mirror central screw, holding the secondary mirror. Count the number of turns to release the screw and note the number. Don’t touch the three collimation screws.

 

Place a light behind the telescope and looking from the front of the scope, sight the hole in the centre of the styrene disc though the centre of the hole that the secondary securing screw came from. Now adjust the primary mirror so that the reflections of the secondary support vanes coincide with the support vanes themselves. You should see a tiny spot of light in the centre of the secondary screw hole and the reflections of the support vanes will be hidden behind the vanes themselves.

 

The primary mirror is now aligned with the secondary holder.

 

Replace the secondary mirror. Count the number of turns and align the paint marks.

 

Next, reach into the scope tube and unscrew the baffle tube. Let it rest on the inside of the tube.

 

Now bring the lamp to the front of the tube and set it to shine onto the styrene disc.

 

Go go to the back of the scope and look through the small hole on the centre of the styrene disc. You should see the central alignment ring on the secondary. Adjust the secondary until the reflection of the hole in the centre of the styrene disc is centred in the secondary alignment ring.

 

The secondary is now pointing at the centre of the primary mirror.

 

Replace the baffle tube. Replace the focuser etc. Sell your laser collimator.

 

Check the collimation on a star in the centre of the field of view. It should be really close.

 

On a night of good seeing I get HFD readings of 1.6 to 1.7 arc seconds at focus. Previously I couldn’t get below 2.5 arc seconds. I don’t use a tilt plate and I’m pleased with the images I get.

 

Hope this helps,

 

David"

 

Christer, Sweden


  • lambermo and silentrunning like this

#20 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 305
  • Joined: 04 Jan 2021

Posted 30 July 2021 - 02:09 PM

Thinking about the Hall of Mirrors effect, it does at least seem that achieving it at multiple spider vanes depends upon alignment of the optical rather than mechanical axes of the primary and secondary in addition to their relative alignment to the secondary holder and the spider.

 

The HOM is a series of reflections between the two mirrors, so if you can see the same "scallop" effect at all four spider vanes and none of them curve away, above or below the spider, then that seems to say the mirrors are optically coaxial. Doesn't it?

Please correct me if I'm wrong about this.
 

One could argue that perhaps the edges of the primary and secondary might not be representative of the figuring of the entire mirror, but short of a pinched mirror we have to start with some kind of assumption that GSO did a decent job of this. Otherwise the RC system wouldn't have its characteristically flat(er than many other reflectors) field.



#21 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 03:27 PM

...or you experience the 'squinted collimation' David07 is describing in his post http://www.cloudynig...read/?p=9224176

There's a lot to unpack there, but I don't think squinted collimation would result in even and low FWHM and eccentricity. 

 

...

His method with a cheap 58 mm cardboard disc works like a charm:

...

A link to that other thread would have resulted in a more compact post and less duplication.  I'm glad that worked for him, but "drilling a hole in the center" of the disk is not very precise.  The method also requires quite a bit of disassembly and reassembly, and I would like to see more carefully demonstrated results.  Checking that a star in the center of the field is "really close" and tossing off a couple HFD numbers is not convincing to me.

 

I like the recommendation to "sell your laser collimator" though.  grin.gif



#22 Juicy6

Juicy6

    Explorer 1

  • -----
  • Posts: 67
  • Joined: 21 Mar 2014

Posted 30 July 2021 - 03:45 PM

...but "drilling a hole in the center" of the disk is not very precise. The method also requires quite a bit of disassembly and reassembly...

I use this tool for that: https://www.amazon.c...677725&sr=8-102

 

Christer, Sweden



#23 dg401

dg401

    Vostok 1

  • -----
  • Posts: 101
  • Joined: 11 Jul 2021

Posted 30 July 2021 - 03:49 PM


Next, reach into the scope tube and unscrew the baffle tube. Let it rest on the inside of the tube.

Fine if you're an RC8 owner.  RC6 owners would need to enlist a child to perform this function as there is not enough clearance around the spider vanes for an adult hand to reach into the tube.


  • astrobananeck likes this

#24 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 474
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 30 July 2021 - 05:59 PM

I use this tool for that: https://www.amazon.c...677725&sr=8-102

 

Christer, Sweden

I figured you were using a compass-like thing like this.  That's what I mean by "not very precise".  I made a disk for exactly this application once by drilling a hole in a plastic sheet, roughing out the plastic circle around it, mounting the plastic on the drill bit I used for the center hole (now embedded in a block of wood), clamping the block to a disk sander table, and rotating the plastic against the disk sander around the drill bit.  I started with the hole and rounded the plastic around it, in other words.  It made a nice precise central hole and a disk which slipped into the telescope back like butter, but it resulted in a poor collimation.  I wasn't using the same method you describe though.


Edited by TinySpeck, 30 July 2021 - 06:01 PM.


#25 dg401

dg401

    Vostok 1

  • -----
  • Posts: 101
  • Joined: 11 Jul 2021

Posted 30 July 2021 - 06:00 PM

I've already found that this procedure produces a result with an identical visual outcome to using my Farpoint laser.  Whether the optical outcome using this procedure is superior remains to be seen under the stars.

 

I did it once.  Maybe I got lucky.  Can I do it twice?  Maybe I should add a degree of difficulty?

 

So I dis-assembled my RC6.  I stripped primary hardware to individual components.  I stripped the secondary hardware to individual components (secondary mirror still in cell because, hey, let's not be crazy).  Then I re-assembled everything.  If I can collimate from a pile of parts, that's a fair indicator that this procedure is reproducible.

 

Performed procedure.  Took maybe 15 minutes to converge with much checking and rechecking of cheshire centering and hall of mirrors.  Cheshire dead center.  Hall of mirrors identical viewed from any angle around circumference of OTA.

 

Cross-checked against my 2 inch Farpoint laser.  Beam is hitting secondary precisely in physical center and beam reflection is hitting laser source precisely in center (or at least as close to both centers as my eye can detect). 

 

Conclusion:  Using this procedure, I can collimate my RC6 to a visual endpoint at least as precise as a laser... and I can reproduce the result.

 

But does it work under the stars?  Sunset in 1.75 hours.  We shall see.


Edited by dg401, 30 July 2021 - 06:11 PM.

  • TinySpeck and nebulachadnezzer like this


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