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

Refining Collimation

  • Please log in to reply
33 replies to this topic

#1 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 09 October 2019 - 08:23 AM

I may attempt to better place the primary center mark to the parabolic center of my primary. It is almost dead on the geometric center as gauged by a paper template with the center clipped out. The hole at the center lays nicely over the hole in the paper reinforcement ring. Believe it or not, it came that way from the factory.

 

However, I still need to make some minor tweaks on a focused star for high power collimation. So, I suspect the optical center is offset by about half the diameter of the center mark. When collimated on a star and checking the Cheshire, the edge of the center marker crosses the center of the pupil reflection. I could mark it, and may need to, or just remember the orientation. But, if I clean the mirror and rotate it, I will lose the orientation of what I think is the optical center. 

 

So, couple of questions. First, is there any reason not to refine the center mark location to the optical center of the primary mirror? I'd think doing so would be necessary for the ability to repeat collimation using tools. Second, it looks as though it'll be best, or even required, to redirect the focuser axis closer to the offset optical center, too. So, it's be nice if there was actually a center marker there to aim the cross hairs. What's the best way to confirm optical center? I'd imagine the star test, but an artificial star is easier to work with because it does not drift across the field. Get the best in focus pattern possible at the center FOV, then note the location of the center mark relative to the Cheshire, right? 

 

A couple of anomalies I've been chasing lately and wondering if others have experienced the same or similar. When I collimate using a slightly defocused star, the Poisson spot looks to be off center. I take this to be residual coma and chase the Poisson spot to the center. Without much luck, I might add. I am starting to think I am cashing some slight focuser axial misalignment. It's easier, actually, to star collimate in focus because the Airy disc is well centered in focus, it's just a matter of chasing more obvious coma in some parts of the FOV. In my case, the western half of the field shows increasing coma as the star drifts through it. It's darn near perfect at the eastern edge (and slightly north to south, too).

 

Another interesting observation is kind of scary, the appearance of astigmatism also in the western edge of the FOV. Again, it's perfectly round at the eastern edge and increases toward the western edge as I am testing for it. I think I am beginning to see some off axis astigmatism as a result from slight miscollimation. I do not see astigmatism where I suspect the primary axis is located (eastern FOV at very high magnification 500 to 600x). Here, the stars are round just outside focus on both sides. I can see it at lower magnification, too, but I am testing at high magnification because I am doing high magnification observing. I also understand astigmatism does not increase in the same rate as coma with field angle, but it certainly looks much better corrected after star collimation. 

 

So, all of this makes me think to relocate the primary center mark to the optical center so I can repeat best collimation using tools. Yea? 



#2 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 09 October 2019 - 08:55 AM

...So, all of this makes me think to relocate the primary center mark to the optical center so I can repeat best collimation using tools. Yea? 

Is this the 8-inch f/6 Dobsonian in your signature? If so, the axial tolerances are quite forgiving, and the center marker would have to be off quite a bit before it would impact image performance (a 2mm error in center marker placement introduces a 1mm error to an otherwise "perfect" Cheshire alignment--still within the high magnification error tolerance on an f/6 primary mirror).

 

Slight defocus (Poisson spot visible) should not reveal a residual focuser axial error, but you might be interpreting air currents or a host of other possibilities that can impact image performance at very high magnification. If you find yourself chasing the optimal primary mirror alignment on a star, I would consider Mike Lockwood's star alignment method  http://www.loptics.c.../starshape.html

 

That said, there's no harm in relocating the primary mirror center mark. Further star evaluation may reveal that the relocation was successful (this should also pass a primary mirror rotation test), or that there may be something else still contributing to the anomalous image performance issue(s).



#3 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3989
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 09 October 2019 - 08:57 AM

Certainly sounds like a good idea.

 

Ive noticed that using tools still requires adjustments using a star, and have also concluded that my centrespot isnt on the optical axis. I havent measured whether its even in the centre though!



#4 Keith Rivich

Keith Rivich

    Surveyor 1

  • *****
  • Posts: 1699
  • Joined: 17 Jun 2011
  • Loc: Cypress, Tx

Posted 09 October 2019 - 09:23 AM

Many years ago when I first started using a laser collimator I could never quite get the laser to hit the center mark *after* collimating with a cheshire and autocollimator (fine tuned with a red LED in the path). After collimating and star testing to my satisfaction I put the laser back in and put a small sharpie dot on the edge of the center marker were the laser was hitting. Now laser collimating (along with barlowed primary return) is dead on. I occasionally recheck with the cheshire/auto but I have not needed to change anything.



#5 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 09 October 2019 - 10:07 AM

"Is this the 8-inch f/6 Dobsonian in your signature? If so, the axial tolerances are quite forgiving, and the center marker would have to be off quite a bit before it would impact image performance..."

Yes, 8" f/6. The paper reinforcement center mark is 6mm in radius, that happens to be 0.03Dmm tolerance. The edge of the marker, at what looks to be good high power collimation, crosses the center of the Cheshire pupil reflection. This means the center of the marker is 6mm offset (from the Cheshire) after primary mirror adjustment. The cross hairs are between the edge of the marker (Cheshire pupil) and center hole which means the focuser axis error is about 3mm. The cross hairs were initially aimed to within the central hole. So, yes, I think the focuser axis is still in tolerance.

The reason I asked about the offset Poisson spot is because I've seen something similar colllimating my MCT after re-installing the primary mirror housing rotated 90 degrees from factory. Scrolling inside showed /similar/ Poisson spot displacement. The fix was to rotate the mirror housing back 90 degrees to where it was. Scrolling through focus showed much better symmetry, so I concluded the problem might have been focuser axial error (as CATs have it).

It could be something else, such as seeing. I've seen it before using an artificial star with a wavering first bright ring with very small defocus. I chased that for a while, too, before giving up to recollimate with a combo tool.

Yes thanks, I've read Lockwood's article a few times and reference it periodically. I'll look it over, again.

Okay, thank you. Lot to think about. I'm sure Im within the widest tolerance, but not so sure the primary is within the tightest planetary tolerance. Need to think that through.

"...this should also pass a primary mirror rotation test..."

Not sure what you mean. The marker is pretty much dead center. If I relocate it, it will no longer be at the geometric center. Being away from center, it seems it will not pass any rotation tests and with the focuser axis aimed at the new location, the reflection of the primary with respect to the site tube circumference (a collimation signature) will be slightly off center, too. No biggie, we presume the marker is on the optical center, but we're after optical alignment even when it's not. Yes.

"Certainly sounds like a good idea."

It certainly sounds practical if not A-retentive (like me sometime). Just not sure how to reach down the tube and mark the center point. Maybe take a picture of the alignment as seen through the site tube and use the cross hairs on the center mark as a reference relative to the mirror clips. Then put a marker there. It almost sounds like guessing, but I need to be darn sure the star collimation is spot on before I commit.

Thank you for taking the time to comment.

Edited by Asbytec, 09 October 2019 - 10:11 AM.


#6 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 42724
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 09 October 2019 - 11:45 AM

A few comments:

1) It is hard, using conventional rotating table grinding and polishing, to get a center of the paraboloid that is different than the center of the glass disc.

So it is not likely the optical center is displaced from the center of the disc.

2) if the bevel on the edge varies in width, however (and I seen this a LOT), the center of the optical axis might not be exactly at the center of the reflective surface.

The entire reflective surface could be slightly offset.

3) It is far more likely that the reason you are having to tilt the mirror to center the collimation is due to mechanical misalignments in the scope.

If the UTA is offset relative to the mirror box, or, on a tubed scope, the mirror cell is off-center in the tube, then additional mirror tilt might be necessary

to arrive at collimation.

4) What I would do is to use a laser or sight tube for the secondary position and tilt, a Cheshire or barlowed laser for primary tilt (using the current marker),

and then use an autocollimator for elimination of residual errors.  Then, repeat the Cheshire/AC ajustments until neither needs refining.

 

If you STILL see the evidence of miscollimation, then I would suspect there is some wedge in the mirror.  This is a common problem with home-made mirrors, but uncommon in

well-made commercial mirrors made on rotating tables. 


  • 25585 likes this

#7 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 09 October 2019 - 12:47 PM

...The paper reinforcement center mark is 6mm in radius, that happens to be 0.03Dmm tolerance. The edge of the marker, at what looks to be good high power collimation, crosses the center of the Cheshire pupil reflection. This means the center of the marker is 6mm offset (from the Cheshire) after primary mirror adjustment. 

OK--I'm not sure I understand what you're saying here. When you say the reinforcement center mark is 6mm in radius, I assume you mean the hole, or perforation in the paper reinforcement ring is 6mm. If the edge crosses the center of the Cheshire pupil, that would mean the center marker is offset 3mm. And since the Cheshire magnifies any primary mirror error 2X, that means the primary mirror axial error is 1.5mm.

 

...The cross hairs are between the edge of the marker (Cheshire pupil) and center hole which means the focuser axis error is about 3mm. The cross hairs were initially aimed to within the central hole. So, yes, I think the focuser axis is still in tolerance.

 

And I'm not sure what you mean by "...the edge of the marker (Cheshire pupil)..."? 

 

"...this should also pass a primary mirror rotation test..."
Not sure what you mean. The marker is pretty much dead center. If I relocate it, it will no longer be at the geometric center. Being away from center, it seems it will not pass any rotation tests and with the focuser axis aimed at the new location, the reflection of the primary with respect to the site tube circumference (a collimation signature) will be slightly off center, too. No biggie, we presume the marker is on the optical center, but we're after optical alignment even when it's not. Yes.

 

To be clear, what I meant by "pass a primary mirror rotation test", is that regardless of the primary mirror's rotation--if the center marker is offset toward the top of the OTA or the bottom of the OTA--collimation on a star should always track the newly offset marker. If it doesn't--then the collimation issue is somewhere other than on the surface of the primary mirror.

 

...Just not sure how to reach down the tube and mark the center point. Maybe take a picture of the alignment as seen through the site tube and use the cross hairs on the center mark as a reference relative to the mirror clips...

 

I would consider removing the mirror/cell and marking or overlaying the existing center marker with a new marked center marker. You could use several color indexes to better reference where the new center should be located--then take a picture through the Cheshire. You should have no problem achieving 0.5mm accuracy.


Edited by Vic Menard, 09 October 2019 - 01:08 PM.


#8 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 09 October 2019 - 01:04 PM

A few comments:

1) It is hard, using conventional rotating table grinding and polishing, to get a center of the paraboloid that is different than the center of the glass disc.

So it is not likely the optical center is displaced from the center of the disc.

2) if the bevel on the edge varies in width, however (and I seen this a LOT), the center of the optical axis might not be exactly at the center of the reflective surface.

The entire reflective surface could be slightly offset.

3) It is far more likely that the reason you are having to tilt the mirror to center the collimation is due to mechanical misalignments in the scope.

If the UTA is offset relative to the mirror box, or, on a tubed scope, the mirror cell is off-center in the tube, then additional mirror tilt might be necessary

to arrive at collimation.

4) What I would do is to use a laser or sight tube for the secondary position and tilt, a Cheshire or barlowed laser for primary tilt (using the current marker),

and then use an autocollimator for elimination of residual errors.  Then, repeat the Cheshire/AC ajustments until neither needs refining.

 

If you STILL see the evidence of miscollimation, then I would suspect there is some wedge in the mirror.  This is a common problem with home-made mirrors, but uncommon in well-made commercial mirrors made on rotating tables. 

I'm in full agreement with 1), 2), and 4).

I'm not getting what you're saying on 3). If the UTA isn't coaxial with the primary mirror, obviously the mirror axis will not coincide with the tube axis. But that shouldn't have any affect on the Cheshire alignment, or the subsequent star alignment.

 

As far as mirror wedge, assuming the mirror was spindle mounted, I would expect the wedge to be removed in the grinding process or when the curve was generated on the Blanchard table. If it survives grinding... I don't know, I've never encountered that. Hand grinding might be a different story, but I suspect an unlikely one given the OP's scope is an ES 8-inch Dob.



#9 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 09 October 2019 - 05:40 PM

"If the edge crosses the center of the Cheshire pupil, that would mean the center marker is offset 3mm. And since the Cheshire magnifies any primary mirror error 2X, that means the primary mirror axial error is 1.5mm."

 

 

Yes, the edge of the circular marker crosses the Cheshire pupil. The radius of the entire paper reinforcement marker is 6mm. The hole is closer to 5mm in diameter. 

 

-collimation on a star should always track the newly offset marker.

 

 

Got it. 

 

It is hard, using conventional rotating table grinding and polishing, to get a center of the paraboloid that is different than the center of the glass disc.

So it is not likely the optical center is displaced from the center of the disc.

 

The mystery deepens. 

 

<time out to take some pics>

 

I tried to take some good pictures to show it, but it's hard to get good registration with a cell phone. I did get one image that looks pretty well concentric, but it shows some secondary rotation at that angle through the Cheshire pupil. Guess what? The cross hairs are centered on the perforation and the marker is centered in the pupil. I bet rotating the secondary a little, the center mark will move back to the Cheshire pupil (from the visual perspective at the pupil). The marker does appear offset opposite the rotation error. The uneven lighting is due to the ceiling light being just outside the FOV. 

 

For some reason, that's different than the view visually. Visually, I get that same view by taking my eye to the extreme edge of the pupil far enough to obscure half the primary. Maybe I am just not looking at it right. (Please do not mind the silicone blob in the cross hairs. It's there because I dropped my combo and dented the end. I had to hammer it back into round so it would fit in the focuser, again. But, the cross hairs became loose, so I measured them and glued them at center. Plus, with astigmatism, it helps to see the actual cross hair intersection instead of two blurry lines that are difficult to trace.) 

 

Silly me..but here goes one of those embarrassing moments where the fix may have been this simple had I actually caught it during collimation. But, I ran the combo tool out to the apex without any problems. Collimation looked good at the time. But, why would the rotation cause me to adjust the primary mirror (aligned with the spider vane at bottom right)? scratchhead2.gif

 

1 Rotation.png


Edited by Asbytec, 09 October 2019 - 06:04 PM.


#10 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 09 October 2019 - 06:10 PM

Oh, I read on CN the Poisson spot offset is caused by the secondary mirror offset. The secondary does have the bidirectional offset built into the spider. The diagonal does hang a little below the spider hub because the central screw is offset a little below center in the secondary holder. This makes sense because the obstruction causes the Poisson spot diffraction effect. One might still expect it to be centered, but maybe not. 



#11 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 09 October 2019 - 07:36 PM

Here's a view that resembles visual appearance with the focuser all the way down. This is after adjusting for star collimation. The center mark did not expose correctly, so I drew it in. The cross hairs are on the rim of the center mark...just outside the perforation. It's within 0.03Dmm * 203mm = 6mm of the center of the marker. It's radius.


2 Rotation.png

I'm a little confused. Maybe the center mark is okay as is. I had to pull one spider vane a little to get the secondary under the focuser. It shows as slightly off center in the image above. There may be some tilt error involved. I wonder if reworking collimation (secondary rotation) might help bring everything in line. But when I apply a little pressure to the spider to correct the error seen in the post above, the center mark moves further from center. Stumped.

Edited by Asbytec, 10 October 2019 - 01:16 AM.


#12 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 10 October 2019 - 09:42 AM

Oh, I read on CN the Poisson spot offset is caused by the secondary mirror offset. The secondary does have the bidirectional offset built into the spider. The diagonal does hang a little below the spider hub because the central screw is offset a little below center in the secondary holder. This makes sense because the obstruction causes the Poisson spot diffraction effect. One might still expect it to be centered, but maybe not. 

Your (50mm) secondary mirror offset is about 2mm. That's 1-percent of the primary mirror diameter. This means the Poisson spot should be offset about the same amount, about 1-percent of the star pattern.

 

Although I wrote it down in my response, I missed the 6mm radius offset in your earlier post. The location of the center marker in your annotated image in post #11 is difficult to accept, and if that's what you're seeing after star alignment, I'm not sure I can accept that either. Of course, it's possible--I just feel like it's improbable. I've reread your opening post and other comments, and here's my concerns:

 

From your original post, you seem to be pushing your scope to its theoretical limits--50X per inch of aperture and more. This kind of magnification pushes everything--the optics (including the eyepiece and your eye), optical supports (to minimize and/or eliminate stress aberrations), temperature equilibration (tube currents), upper atmospherics, axial alignment reliability and consistency (for example, mechanical gravitational flexure), and of course, expertise when making the alignment read.

 

On top of this, your axial alignment error tolerances are quite forgiving--a little sloppy should be good enough to deliver a good image.

 

I have to ask, has this scope's image performance ever met your expectations? I just feel like a 3mm error in the primary mirror centering could indicate sloppy parabolizing or something else wrong with the glass--but if that's the case, I don't believe the mirror would have ever delivered a good image. On the other hand, if the mirror has performed well in the past (up to, let's say 300X), then something else is going on... 


Edited by Vic Menard, 10 October 2019 - 10:41 AM.


#13 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 10 October 2019 - 01:43 PM

I've tried to annotate your image in post #9. The green circle is the donut (centered relative to the primary mirror edge, also annotated in green), the blue circle (with cross hairs) is the underside of your focuser/combo tool. Is this the alignment you see through your combo tool after aligning on a star?

 

 

 

 

Attached Thumbnails

  • post-29900-0-80352700-1570660765-1.png


#14 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 10 October 2019 - 02:00 PM

Looking at the annotated image above--I think I see part of the problem. The glue blob at the intersection of the sight tube cross hairs is getting in the way of the Cheshire alignment read. I can almost visualize how you have to move your eye around to see the primary mirror center marker. Unfortunately, you need to keep your eye centered to minimize parallax--if you're not centered, the read precision suffers. 

 

I'm also guessing that your Cheshire is 1.25-inch. If that's the case, how precise is the fit between the focuser, adapter, and Cheshire? Do you use the same adapter with your Cheshire and your high magnification eyepiece?

 

Finally, I recall you mentioning positioning the pupil near the apex. You should only use the apex when you're trying to optimize the secondary mirror placement/offset. Don't use the apex when you're assessing/correcting the axial alignments. You should always assess/correct the axial alignments with the pupil near the focal plane. Also, never rack the focuser drawtube fully in or fully out, as this can mechanically torque the focuser drawtube into a misaligned position. And remember, when you're aligning on a star at very high magnification, the star needs to be carefully centered in the field of view and any defocusing should be at most one or two (more) diffraction rings. And at high magnification, you will see everything, including zonal errors, air current lensing, glare effects (don't use a very bright star!)...ideally, you would like to see a focused Airy disk with a complete diffraction ring, anything else and it's up to you to sort out what you're seeing and if collimation has anything to do with it--which is why I suggested Mike Lockwood's procedure, although thermal and atmospheric instability can still play havoc, even with his simplified read. 

 

That's all for now...  waytogo.gif


Edited by Vic Menard, 10 October 2019 - 02:31 PM.

  • cloudypatio likes this

#15 nirvanix

nirvanix

    Vanguard

  • -----
  • Posts: 2062
  • Joined: 07 Jun 2007
  • Loc: Medicine Hat, AB

Posted 10 October 2019 - 03:23 PM

Sorry Norme, when I read what you write I become confused grin.gif

 

FYI:

I have center spotted my two dobs to what I think is within +/- 0.5mm. Didn't feel it was possible to do better.

 

I don't offset my secondaries. Even so, I'm able to get the reflection of the primary centered in the secondary as seen through the focuser, align the crosshairs of my AstroSystems Lightpipe with the primary center spot, and achieve a closed optical system with my AstroSystems autocollimator. This has produced excellent lunar/planetary results with mags north of 400x (seeing permitting).

 

Best wishes.



#16 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 10 October 2019 - 07:46 PM

Thank you all for replying. After reading Don's comment above about being almost impossible to have the optical center very far from the geometric center, I began to doubt my previous collimation. So, reset everything and started from scratch. After positioning the secondary, I presume it's normal to begin focuser axial alignment by initially not having the center of the primary centered. And the cross hairs are not on the center spot. Achieving that in step 1 would be a miracle. However, using tilt "only" I simply cannot keep the reflection of the primary centered because the secondary does rotate when the collimation screws are adjusted. I do not want to "cheat" and tilt the secondary by hand to center the cross hairs. I may be adding too much additional tilt error instead of collimating the focuser axis. I went through 4 hours if iterations between the two and gave up. 

 

The location of the center marker in your annotated image in post #11 is difficult to accept...

 

Me, too. 

 

you seem to be pushing your scope to its theoretical limits...and of course, expertise when making the alignment read.

 

Yes, this is why I am trying to get the scope within the high magnification tolerance. The entire FOV should be diffraction limited for coma, but it does not appear to be so. Up this high magnification is where I do see some residual coma in part of the field. And I believe some off axis astigmatism. (I went through Lockwood's article to remove as many potential sources of astigmatism as possible). The star images are otherwise "perfect" (coma coma astigmatism free) in the eastern part of the field, but they degrade a bit as the star image drifts toward the west. I am sure I am losing a tiny bit of critical performance in the small FOV. This increasing coma with the westerly star drift is what lead me to believe the colllimation is s tiny bit off. Adjusting the primary tended to improve the image to some degree. 

 

...has this scope's image performance ever met your expectations?

 

I am doing some high magnification study of Saturn's rings. I am still pretty impressed with the image at 0.3mm exit pupil to within the limits of the descent seeing conditions. The mirror is at least descent and so are the images. I conducted a Foucault knife edge, a Ronchi tests at CoC and (probably too insensitive) on a star, as well as a visual star test and shadow break out. It's slightly under corrected and pretty good parabolic figure to the best I can make it out. The Knife edge nulls out very well on the star test and the Foucault knife edge shows the mirror has the expected parabolic figure (but no measurements were made, just looking at the shadows along the caustic at CoC). It snaps pretty well on the moon and Saturn. So, I'm confident the mirror is fine. If it is fine, then the optical center should be well placed, I suppose, but being mass produced I began to wonder about that. And if it is fine, then I have to assume my collimation is somehow off. 

 

Is this the alignment you see through your combo tool after aligning on a star?

 

Yes. But when I look closer, I see some rotational error in that image. You can also see (the reflection of the secondary is offset from the red circles) where I adjusted the secondary laterally from step 1 to center the secondary and square under the focuser. I am wondering if these two adjustments are related somehow and require further refinement of the primary collimation to center the coma free point in the FOV. If so, this is terrible collimation. 

 

The glue blob at the intersection of the sight tube cross hairs is getting in the way of the Cheshire alignment read. I can almost visualize how you have to move your eye around to see the primary mirror center marker.

 

You know what's interesting, and possibly wrong, but interesting? With astigmatism I see blurry cross hairs and try to extrapolate the blur across the center mark. That's difficult. But, I can tell the glue blob is on the center mark because the center mark dims a little as the blob passes over it blocking the light it reflects. This is pretty easy to see. I also tried using the camera to be my "eyes" by taking pictures of each step and each iteration to see how close I am getting. That's pretty frustrating. I can keep my eye pretty much centered, but the camera and my eye do not agree. So, I trust my eye instead of fussing with camera registration. Glasses get in the way, plus my astigmatism is not that bad. I can see everything else. 

 

I'm also guessing that your Cheshire is 1.25-inch. 

 

It's a 2" Light Pipe suitable for f/6. Yes, it registers very well using the 2" adapter. In fact, it's so tight I had to tap it back into round after dropping it just to get it to fit into the focuser. I managed that, but noticed one of the cross hairs was no longer centered. So, I measured, marked and glued the cross hair into position. Upon checking with a ruler across the site tube end with the cross hairs visually centered over the pupil, one cross hair seems dead on, the other measurement from center to the edge is off by about 0.5mm (which is really about 0.25mm error at the cross hair). That should be very good. I am thinking maybe I need a collimation tool that is easier to see well and use visually, but I still have to use the site tube to center the secondary and align the focuser axis without introducing unwanted tilt. With the site tube we can see any induced secondary rotation visually. I am not sure a laser will allow us to that. So, I am undecided. The Cheshire portion seems just fine. 

 

You should only use the apex when you're trying to optimize the secondary mirror placement/offset. Don't use the apex when you're assessing/correcting the axial alignments. You should always assess/correct the axial alignments with the pupil near the focal plane.

 

Yes, after collimation earlier, I racked the focuser (not all the way) out to the Apex to check the reflection of the primary in the secondary. I had it very nice at one point. Clean primary reflection all around, a thin dark line separating the reflection of the primary and the actual secondary edge, and all three clips easily seen and the primary center mark dead center in the Cheshire. The cross hairs were on the center mark within 6mm tolerance from top dead center. That should have been darn near prefect. Still, I needed some primary mirror adjustment at very high magnification. 

 

One difficulty of collimation, first it's pretty dark in there, so its hard to see the secondary holder to ensure it is centered under the focuser. I found if I tilt the OTA  a little and get some better lighting, I can illuminate and see the secondary holder and better estimate longitudinal placement of the secondary along the tubes longitudinal axis. So, I did that and got the great collimation above. Last night I resorted to a yellow background reflection in the secondary (so not to see the primary mirror) to contrast it with the white background behind the secondary on the far side of the OTA. I can get that darn near perfect. It does seem to require a little "lateral" adjustment of the secondary spider vanes to pull the secondary to center of the focuser. No problem, that goes well. 

 

However, and as expected, when removing the yellow background the primary reflection is not centered nor are the cross hairs. That's normal, I believe, so I proceed to collimate the focuser axis without adding unnecessary tilt. I guess we want to correct any tilt induced by the secondary collimation screws, but not *USE* tilt to "cheat" the centering of the cross hairs. I think this is where I am running into trouble mixing tilt with corrections to induced tilt during collimation. So, I added a milk jug washer last night to see if it helps reduce induced tilt on the secondary. The results were not good. Absolutely frustrating going through iterations of steps 1 and 2 and not making any progress. I cannot get collimation anywhere close to being close. 

 

And remember, when you're aligning on a star at very high magnification, the star needs to be carefully centered in the field of view and any defocusing should be at most one or two (more) diffraction rings.

 

Correct. Of course we have to deal with stars drifting or use an artificial star. I do either or both. But, we can watch a star as it crosses the center of the FOV, and we can see it at the edges. In focus, it's not hard to see differing levels of coma when it's present. This is what is driving me nuts. lol.gif

 

Okay, enough for now...coffee is cold. I will hit it again today, starting with step 1. Again. Maybe it is my collimation and not the optical center. The image you annotated above does not look good to me. If that's what the camera sees and if that is my collimation. Yuk! Something is not right. 

 

Edit: OUCH! This was a long reply. I noticed it when I saved it, sorry. No wonder my coffee is cold. 


Edited by Asbytec, 10 October 2019 - 07:57 PM.


#17 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 10 October 2019 - 07:53 PM

Nirv, how are you? You can tell I have a headache? :)

 

I have center spotted my two dobs to what I think is within +/- 0.5mm. Didn't feel it was possible to do better.

 

Yea, my center spot seems to be pretty close, too. Close enough and definitely closer than the image above might imply unless the optical center is off by that much. For completeness, I'll try another method to measure it. I am sure it's not going to be that far off. 

 

...and achieve a closed optical system with my AstroSystems autocollimator.

 

Gosh, all those reflections...I just cannot imagine the learning curve. Scary tool.

 

This has produced excellent lunar/planetary results with mags north of 400x...

 

But I am tempted... :) 



#18 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 42724
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 10 October 2019 - 08:11 PM

 a suggestion:

1) center the secondary under the focuser/sight tube in the up-tube/down tube direction

2) rotate the secondary to center the reflection of the primary on the short axis of the secondary.

3) collimate the secondary

4) repeat step 1 and 2 if necessary

5) repeat step 3.

6) repeat until step 1 and 2 and 3 all work at the same time.  Do NOT move the vanes--the vanes have to be in a direct line in order to avoid double spikes on stars.

7) Collimate the primary.

8) recheck steps 1, 2, 3 and repeat step 7.

All should be just fine at the same time.

Now, if you cannot see the primary center marker easily because of a glue blob, dissolve it with acetone.  There may still be enough glue left to hold the wires in place.


Edited by Starman1, 10 October 2019 - 08:12 PM.

  • Asbytec and airbleeder like this

#19 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 10 October 2019 - 09:14 PM

Don, okay, thank you. Will do, as I understand it, center the secondary as normal then get a rough collimation signature by centering the primary (along the minor axis) using rotation only. It's just an initial orientation to get in the ball park. Still, the entire reflection of the primary is visible after completing secondary placement. It's just not collimated, yet and I go from there trying to use tilt only. 

 

I suspect I am battling induced rotation while tilting the secondary and may be over compensating with too much unintentional tilt to return the secondary to it's initial placement. Probably beyond it's original placement in step 1. I may be overshooting that adjustment for the sake of getting the cross hairs onto the center mark and leaving in some residual tilt error. But, I always check at the apex for a centered primary. I do not always like what I see, but I work with it a little to get the reflection of the primary centered and check the cross hairs, again. At some point, I call it collimated. Then check with a star test...which reveals something is wrong.

 

Yes, double spikes...don't wanna see those. But, to center the secondary requires a little sideways pull (away from OTA center) on the spider. So far, I have not seen double spikes, but I may just shim the focuser a little, anyway. 


Edited by Asbytec, 10 October 2019 - 09:16 PM.


#20 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 11 October 2019 - 06:56 AM

Don, I hope this explanation makes sense. Your suggestion lead to an epiphany. 

 

I centered the secondary in the OTA and shimmed the focused a tiny bit. Got the secondary centered and circular under the focuser. Then I rotated the secondary to center the reflection of the primary along the *major* axis. In other words, if you extend an imaginary line from the focuser through the secondary major axis, the reflection of the primary is centered on that (focuser) axis. All that went well. So, I am leaving that alone. I know that's right, the secondary is centered and concentric and the reflection of the secondary is coarsely centered in the secondary. The three collimation signatures are almost there, just need some tilt. 

 

However, collimating the focuser axis by tilting the secondary led to the same frustrating tilt, de-centered reflection of the primary, and I believe some rotation error. The two collimation screws that move at an angle to the focuser axis are causing the secondary holder to tilt laterally with reference to the draw tube and, I believe, inducing a little rotational error. When I see this, I am tempted to use the spider vanes to pull the secondary laterally back to the center of the site tube to regain even illumination of the white paper background again. First one way, then the other depending on which oddball collimation screw I am adjusting at the time. Then I have to adjust the rotation of the secondary, again, rinse and repeat. I am chasing this rabbit down a bottomless hole, and even adding another shim to the focuser to correct for it (one of them removed later before starting over). 

 

What I realized is, after resetting to the original starting point with the secondary centered and the primary reflection centered on the imaginary focuser axial line (using rotation only), I have to adjust the two screws at an odd angle the focuser axis, first, to reduce the lateral skewing of the secondary holder. One then the other cancels the effect of each. So I make sure the cross hair (set parallel to the focuser axis) is inline with and well below the center spot as I walk the two screws into a position below the center mark along the cross hair (which requires only upward tilt to correct). This helps keep the primary centered on the secondary major axis. Then I correct for any rotation and make sure the reflection of the primary is still centered on the major axis. Rinse and repeat. Then, in the final move toward centering the cross hairs, I walk the secondary tilt *only upward* toward the focuser until the cross hairs are on the center mark. Interestingly, and I dunno why, but this single collimation screw does not induce any rotation, and it certainly does not skew the secondary laterally. Not one iota. The reflection of the primary is well behaved and remains centered. Doing so makes holding the primary reflection centered much easier without any sideways pull on the focuser vanes, rotational correction, or the need for another shim. 

 

Adjusting the secondary in this order prevents the two oddball screws (already set with the primary reflection centered on the secondary major axis) from skewing the secondary mirror laterally and (possibly) inducing some rotation. On final collimation, the center mark slides up the cross hair parallel to the focuser axis until it's centered and adds final tension to the secondary holder. The result is a centered primary reflection not only on the major axis, where I kept it the whole time, but also centered on the minor axis. It's so much easier to set the two oddball screws first and recenter the reflection of the primary on the major axis, then final approach to the center mark by tightening the one collimation screw in line with the focuser axis. It seems more or less randomly adjusting secondary collimation screws causes problems, it's much easier to approach final collimation using one screw parallel to the focuser axis. 

 

A couple of turns on the primary collimation screws (now with motorcycle clutch springs installed - a bit too stiff) and the center mark is centered in the Cheshire. Looks pretty good now. The secondary is centered in the OTA, the focuser has a thin shim (gonna leave it in), the cross hairs are on the center mark, the collimation signatures are pretty darn close, and the center mark is as close to center of the Cheshire as I can possibly get it. Now, just need a clear night...see if the center mark is on the optical axis. I am sure it is, actually.

 

Thanks, everyone. 


Edited by Asbytec, 11 October 2019 - 07:46 AM.


#21 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 11 October 2019 - 09:11 AM

1.) After positioning the secondary, I presume it's normal to begin focuser axial alignment by initially not having the center of the primary centered. And the cross hairs are not on the center spot. Achieving that in step 1 would be a miracle. However, using tilt "only" I simply cannot keep the reflection of the primary centered because the secondary does rotate when the collimation screws are adjusted. I do not want to "cheat" and tilt the secondary by hand to center the cross hairs. I may be adding too much additional tilt error instead of collimating the focuser axis. I went through 4 hours if iterations between the two and gave up. 

 

2.) Up this high magnification is where I do see some residual coma in part of the field. And I believe some off axis astigmatism. ...This increasing coma with the westerly star drift is what lead me to believe the colllimation is s tiny bit off. Adjusting the primary tended to improve the image to some degree. 

 

3.)...when I look closer, I see some rotational error in that image. You can also see (the reflection of the secondary is offset from the red circles) where I adjusted the secondary laterally from step 1 to center the secondary and square under the focuser. I am wondering if these two adjustments are related somehow and require further refinement of the primary collimation to center the coma free point in the FOV. If so, this is terrible collimation. 

 

4.) It's a 2" Light Pipe suitable for f/6. Yes, it registers very well using the 2" adapter. In fact, it's so tight I had to tap it back into round after dropping it just to get it to fit into the focuser. I managed that, but noticed one of the cross hairs was no longer centered. So, I measured, marked and glued the cross hair into position. Upon checking with a ruler across the site tube end with the cross hairs visually centered over the pupil, one cross hair seems dead on, the other measurement from center to the edge is off by about 0.5mm (which is really about 0.25mm error at the cross hair). That should be very good. I am thinking maybe I need a collimation tool that is easier to see well and use visually, but I still have to use the site tube to center the secondary and align the focuser axis without introducing unwanted tilt. With the site tube we can see any induced secondary rotation visually. I am not sure a laser will allow us to that. So, I am undecided. The Cheshire portion seems just fine. 

 

5.) One difficulty of collimation, first it's pretty dark in there, so its hard to see the secondary holder to ensure it is centered under the focuser. I found if I tilt the OTA  a little and get some better lighting, I can illuminate and see the secondary holder and better estimate longitudinal placement of the secondary along the tubes longitudinal axis. So, I did that and got the great collimation above. Last night I resorted to a yellow background reflection in the secondary (so not to see the primary mirror) to contrast it with the white background behind the secondary on the far side of the OTA. I can get that darn near perfect. It does seem to require a little "lateral" adjustment of the secondary spider vanes to pull the secondary to center of the focuser. No problem, that goes well. 

 

6.) ...I guess we want to correct any tilt induced by the secondary collimation screws, but not *USE* tilt to "cheat" the centering of the cross hairs.

OK--there's a lot here to discuss.

 

1.) When you first start aligning the secondary mirror, your first goal is to center the actual edge of the secondary mirror under the focuser. I often start with all 3 of the secondary mirror tilt adjustment screws "equal". This assumes that your spider is perpendicular to the OTA axis, your secondary stalk is 45-degrees to the spider, and your focuser is 90-degrees to the OTA (it's quite likely none of these assumptions are true to any significant precision--but reasonably close is a good starting point). When you've "centered" the secondary under the focuser, it's unlikely that the primary mirror reflection will also be centered in the secondary mirror. That doesn't happen unless the focuser axis is also correct. So, the next goal is to adjust the secondary mirror tilt to center the primary mirror reflection in the focuser. You do this by aligning the sight tube cross hairs with the primary mirror center spot. Of course, when you change the secondary mirror tilt, you also change the secondary mirror placement, which means the secondary mirror is no longer centered under the focuser. What you need to determine is what's causing the placement error. So, now you need to align the primary mirror tilt (make the primary mirror center marker/donut concentric with the bright Cheshire ring). With the primary mirror tilt aligned, if you look at the silhouette reflection of the secondary mirror in the primary mirror reflection, the dark silhouette will probably look ellipsoidal. IF the ellipse points toward the primary mirror or 90-degrees from the primary mirror, the placement error is offset (secondary mirror position closer to or farther from the primary mirror). IF the ellipse points 45-degrees above or below the primary mirror end of the OTA, the placement error is rotation. If you look closely, you can often see one side or the other of the secondary mirror in the silhouette reflection--that shouldn't be visible. The secondary sides define a cylinder, and the cylinder should be aligned to the primary mirror axis. 

 

To summarize:

First--make the actual edge of the secondary mirror concentric with the bottom edge of the focuser/sight tube (try not to use the tilt adjustment if it has been zeroed).

Second--adjust the secondary mirror tilt to align the focuser axis.

Third--adjust the primary mirror tilt to center the secondary mirror reflection in the primary mirror reflection

Fourth--determine how the secondary mirror placement needs to be corrected (rotation or offset) and make the "rough" correction.

Repeat second and fourth steps until the secondary mirror placement and the focuser axial alignment are both correct.

 

2.) Are you sure you're chasing coma? You mention astigmatism (not sure if it's from your eye or the eyepiece). With very small pupils, it's more likely to be the eyepiece. But it can also be tube currents. You mention the west side of the field--is the test star rising, setting, or on the meridian? Do you see convection in the out of focus star image?

 

3.) The adjustments are most definitely related. It's important that the (mechanical) focuser axis intersects the line between the primary mirror and the "center" of the secondary mirror. Perpendicularity is not required.

 

4.) OK--2-inch LightPipe that's suffered an impact. When you make the Cheshire alignment read, is the LightPipe inserted flush to the top of the focuser drawtube? Have you verified that the Cheshire read doesn't change when the LightPipe is inserted flush for inserted part way (for example, when you move the pupil to the apex)?

 

5.) Lighting is, as you say, very important when using a sight tube. You want to minimize glare, but you need plenty of light down that long tube. I also suggest a white piece of paper behind the secondary mirror to make it easier to see the actual edge of the secondary mirror--but don't cover the primary mirror unless the cover is the same size as the primary mirror, placed on the surface of the primary mirror, and preferably, with the center marked. Better yet, just don't cover the primary mirror.

 

6.) See number 1.) above.



#22 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 11 October 2019 - 09:34 AM

...It seems more or less randomly adjusting secondary collimation screws causes problems, it's much easier to approach final collimation using one screw parallel to the focuser axis. 

 

...The secondary is centered in the OTA, the focuser has a thin shim (gonna leave it in), the cross hairs are on the center mark, the collimation signatures are pretty darn close, and the center mark is as close to center of the Cheshire as I can possibly get it. Now, just need a clear night...see if the center mark is on the optical axis. I am sure it is, actually.

It shouldn't be random (unless you have both rotation and offset errors). Rotation errors are probably the most common errors I routinely encounter when helping others sort out their secondary mirror placement. I can usually place my finger on the side of the secondary mirror that's exposed and let the scope's owner see that, if that side is exposed, the secondary mirror is pointing down toward the opposite side. Then all I need to do is to tighten the screw behind the opposite side a turn or so (which moves the sight tube cross hairs "up" an inch or so above the primary mirror center marker, and then rotate the secondary mirror to move the cross hairs back down into realignment with the center marker. Then check and see if it needs more correction, or less...

 

Shimming the focuser is the solution if the (mechanical) focuser axis does not intersect the optical axis (see number 3.) above).

 

Finally, remember what's important:

First--primary mirror tilt alignment

Second--focuser axis (secondary mirror) tilt alignment, and last...

Third--secondary mirror placement.

 

The first ensures that the "sweet spot" will be centered in the field of view. The second ensures that the focal plane will be perpendicular to the optical axis, delivering snap focus when combined with the first. And the last delivers balanced field illumination (most often undetectable visually).


  • Asbytec and airbleeder like this

#23 Vic Menard

Vic Menard

    Fly Me to the Moon

  • *****
  • Posts: 6855
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 11 October 2019 - 09:50 AM

Here's what a rotation error looks like (the secondary mirror is pointing down about 45-degrees--green arrow). If you look closely, the secondary holder has mounting screws. Look very closely and you'll see the two paired mounting screws (large and small violet arrows), and only one unpaired mounting screw (red arrow) on the exposed side (the other screw head/orange arrow isn't visible because of the rotation error).

 

Your secondary mirror isn't mounted in a clamshell with screws, so you don't have the screw heads to help you see this error. But you should still be able to see the side of the secondary mirror--use lots of light.

 

 

Attached Thumbnails

  • Skew1.JPG

  • airbleeder likes this

#24 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 42724
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 11 October 2019 - 10:07 AM

That's why I said to center the reflection of the primary mirror in the minor axis of the elliptical secondary.

You can't do that and have an incorrect rotation of the secondary mirror, as Vic's illustration indicates.

If you then collimate and the primary reflection isn't centered in the major axis of the secondary, then the cure is to move the secondary up tube or down tube,

NOT to adjust its tilt using the collimation screws.

When it is moved up-tube or down tube, you will have to collimate the secondary all over again.


  • Vic Menard, Asbytec and airbleeder like this

#25 Asbytec

Asbytec

    Guy in a furry hat

  • *****
  • topic starter
  • Posts: 15632
  • Joined: 08 Aug 2007
  • Loc: Pampanga, PI

Posted 11 October 2019 - 01:30 PM

...if you look at the silhouette reflection of the secondary mirror in the primary mirror reflection, the dark silhouette will probably look ellipsoidal.

 

Don, yes I've seen this. Now the silhouette looks circular and the offset reflection of the focuser is on an axis pointing toward the primary. Not skewed like before (and I just accepted). Much better.

 

IF the ellipse points 45-degrees above or below the primary mirror end of the OTA, the placement error is rotation.

 

Okay, I have not learned to read this signature in a way I can work with it. Need to put some brain power into understanding it. It makes sense, if the ellipse is pointing somewhere other than the primary mirror, there is some rotation. I've seen it. 

 

First--make the actual edge of the secondary mirror concentric with the bottom edge of the focuser/sight tube (try not to use the tilt adjustment if it has been zeroed).

 

Okay, I've done that. No tilt, only rotation and longitudinal movement toward or away from the primary. 

 

Fourth--determine how the secondary mirror placement needs to be corrected (rotation or offset) and make the "rough" correction.

 

This is the hard part. I've seen it, but have not used this signature to make any corrections thinking the cross hairs are centered, so it much be the way it is. Mechanical construction or something. Yea, it bugs me. 

 

You mention astigmatism (not sure if it's from your eye or the eyepiece). With very small pupils, it's more likely to be the eyepiece.

 

I do have astigmatism, but usually only visible at larger exit pupils. To be clear, I see it only when slightly defocused as alternating ellipses on both sides of focus and only on one side of the FOV. (I have seen slight and highly variable astigmatism and pinched optics in focus and on axis, but Lockwood's "Stars Round" article helped address a lot or all of that). Seeing does not make it easy to see in focus. I thought it might be the eyepiece edge, but the eastern field edge is clear of astigmatism (and coma). The stars are near the meridian. I /think/ it is off axis astigmatism in the scope itself. This is another clue primary axial alignment is off, along with visible coma. I'm sure I'm chasing coma during the star test, I can see the sagittal portion and the tangential portion on field stars bright enough to exhibit diffraction rings (but not too bright).

 

There was more going on, like the offset Poisson spot. I am not sure we can see 1% displacement as Vic said above. So, I thought it might be some tilt error. I'll check again for it. There is definitely seeing involved. 

 

When you make the Cheshire alignment read, is the LightPipe inserted flush to the top of the focuser drawtube? Have you verified that the Cheshire read doesn't change when the LightPipe is inserted flush for inserted part way (for example, when you move the pupil to the apex)?

 

Sometimes it is flush, sometimes not, during collimation. Yes, I have checked the site tube read by rotating the site tube in the focuser. The cross hairs stay put best I can tell. I just checked the Cheshire read, there is a tiny bit of offset on the center mark as the Cheshire is rotated. It stays inside the illuminated hole. Wonder what that means and if there is a way to fix it? (Maybe just spin the Cheshire and average it out through a full rotation)? 

 

I also suggest a white piece of paper behind the secondary mirror to make it easier to see the actual edge of the secondary mirror--but don't cover the primary mirror...

 

I use white paper background. I dunno, I find using a colored paper to cover the primary (from view) to be helpful, at least initially to set the secondary position. Then I remove it. Next, as you suggested, I make a rough correction using secondary *rotation only* to center the image of the primary along the focuser axis. To my mind, that completes Step 1: Secondary Placement. 

 

Lemme read Vic's reply to understand your description of the error and what to do about it. Right now, I do not see it. Thankfully. 




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