434 replies to this topic

[Nils Olof Carlin]

Kudos to Jason for not only identifying the potential, previously unknown (or at least ignored) problems with the AC:

* the vanishing reflections (that can be seen in the offset pupil),
* the defocus problems (solved by the CAM),
* the consequences of having the AC mirror away from the focal plane

but also solving them!

Nils Olof

[hudson_yak]

The pictures and explanations as this topic progresses are truly excellent, Jason. Nice work.

Then we get to the table (what many readers are going to take away from this), and I have to wonder about a few things there:

1) Minor quibble, the cheshire in a combo tool can also be calibrated to a primary mirror mark, so could also earn an Excellent.

2) Was interested to see that there are no Excellent FAE tools(?)

3) Was interested to see that the CAM AC did not get promoted into the PAE row at all, since I thought its intent, one of them, anyway, was to make a cheshire completely redundant.

4) There ought to be some reemphasis that PAE and FAE are really the only two collimation axes. CAE and LAE are interesting to help talk about AC theory only.

Mike

[Jason D]

Hi Nils Olof,
My work is based on the strong foundation laid out by your mathematical analysis of the autocollimator
Jason

[Jason D]

The pictures and explanations as this topic progresses are truly excellent, Jason. Nice work.

Thank you, Mike

Then we get to the table (what many readers are going to take away from this), and I have to wonder about a few things there:

For the record, I did include a disclaimer explaining that the table is subjective and based on my own opinion.

1) Minor quibble, the cheshire in a combo tool can also be calibrated to a primary mirror mark, so could also earn an Excellent.

My grading is relative. A “Good” rating should not be construed as a bad rating just because there is an “Excellent” rating. Using a cheshire like Catseye’s with a triangle and a ring does add more reading accuracy than the sight-tube/combo tool with the distracting cross-hairs. I am speaking from experience since I have used both tools extensively.

2) Was interested to see that there are no Excellent FAE tools(?)

Correct. Getting the “Excellent” rating has high standards. It should involve visual cues that are parallax-free, sharp, and highly-readable. Stacking P+1 does not meet this criteria. A laser collimator does not either.
Let me put it this way, a “Good” rating provides accuracy to within ~0.5mm. An “Excellent” rating provides accuracy to within ~0.2mm.

3) Was interested to see that the CAM AC did not get promoted into the PAE row at all, since I thought its intent, one of them, anyway, was to make a cheshire completely redundant.

The table is meant to rate “direct” and “explicit” alignments – not “indirect” alignments. The CAM AC does not get a “rating” for PAE because it is impractical to “directly” align the pupil reflection against reflection “P”. HOWEVER, as long as a tool gets “direct” excellent rating for two error elimination, it implies “indirectly” the remaining two error elimination get the same rating.
OK, Mike I'll grant you that table could be confusing. I need to think about how to update that post later today.

4) There ought to be some reemphasis that PAE and FAE are really the only two collimation axes. CAE and LAE are interesting to help talk about AC theory only.

I disagree with this line of thinking. PAE, FAE, LAE, and CAE are all means to an end which is “Axial Alignment”. I will follow whichever path that gets me to my axial alignment destination with the greatest accuracy whether it is PAE, FAE, LAE, CAE or even XAE -- whatever that is

Jason

[CatseyeMan]

2) Was interested to see that there are no Excellent FAE tools(?)

Correct. Getting the “Excellent” rating has high standards. It should involve visual cues that are parallax-free, sharp, and highly-readable. Stacking P+1 does not meet this criteria....

Jason

What about stacking P & 3 in the CDP?

[Shawn H]

Jason
Incredible drawings again! You really know how to make a blind man see!

[Nils Olof Carlin]

Since you invite comments, Jason, in #3533575, here are some on account of Mike's comments:

There ought to be some reemphasis that PAE and FAE are really the only two collimation axes. CAE and LAE are interesting to help talk about AC theory only.

Indeed true - IIRC CAE=2*PAE+FAE and LAE=2*PAE+2*FAE (vector sums), and thus, the CAE and LAE should perhaps not be seen as "fundamental" in the sense of being independent of the others (you could, in principle, derive PAE and FAE from CAE and LAE, but this would be pointless).

It could also be made clearer that the PAE and FAE are indeed independent of each other - you may have accurate collimation of one axis and still have some miscollimation of the other (even if that's what you try to avoid). It is only the PAE that causes coma in the center of the FOV, the FAE will cause a tilt of the focal planes but no error in the center of the FOV.

Nils Olof

[Jason D]

What about stacking P & 3 in the CDP?

It depends on the focal length of the scope. For my 1200mm scope, it is hard to see reflection 3 primarily because of its fuzziness.

The left photos are what I see and the right ones is with the camera focused at infinity.



Those with longer focal lengths scopes, reflection 3 will be clearer.

For a reference to others, the following posts explain the relationship between focal length and reflections 1&3 fuzziness:
post1
post2
post3

But the trend seems to be moving towards scopes with shorter focal lengths (< F3.0 scope). Not to mention that imaging scopes tend to have a short focal length – the very scopes that require more accurate focuser axial alignment.

Jason

[Nils Olof Carlin]

One interesting thing, very well illustrated in Jason's post #3532773 (and perhaps easier seen in photos than directly by eye), see the lower set of images:

The reflections occur at different apparent distances from the pupil.
P and 2 are seen one (mirror) focal length away, the AC and CAM reflections at infinity, and reflections 1 and 3 at minus one focal length (that is, behind your head).

When the camera is focused on P and 2, 1 and 3 are obviously fuzzy. On the CAM images, the camera is focused at infinity and both pairs are equally fuzzy (while the CAM rings are sharp), but only half as much as in the previous case. If the camera lens could be moved inwards from the infinity setting, it would be possible to get 1 and 3 sharp - but "normal" cameras do not allow it (a weak negative add-on lens might do the job).

(The thought of seeing images occuring behind your head may be strange - but those images are virtual, and it wasn't before I realized that they are indeed behind the observer's head that I could even begin to make sense of all the reflections.)

Nils Olof

[Jason D]

Just check the post submitted 1 minute before your post. The photo and included links explain the unintuitive concept of seeing an image behind our heads. Of course, you were the first to have explained it to me, Nils Olof. When the camera focuses at infinity then P&2 images are formed behind the retina and 1&3 images are formed in front of the retina by about the same distance – that is why both sets of images would look similar in fuzziness. For scopes with longer lengths, the distance between the four images and the retina is reduced which brings them all closer to focus (retina).
But the eye has the tendency to focus on P&2 on the expense of 1&3 since the eye is designed to see images in front of the eye – not behind.


Jason

[Jason D]

Indeed true - IIRC CAE=2*PAE+FAE and LAE=2*PAE+2*FAE (vector sums), and thus, the CAE and LAE should perhaps not be seen as "fundamental" in the sense of being independent of the others (you could, in principle, derive PAE and FAE from CAE and LAE, but this would be pointless).

It could also be made clearer that the PAE and FAE are indeed independent of each other - you may have accurate collimation of one axis and still have some miscollimation of the other (even if that's what you try to avoid). It is only the PAE that causes coma in the center of the FOV, the FAE will cause a tilt of the focal planes but no error in the center of the FOV.

I covered the PAE/FAE relevant importance in the 3rd post of this thread.

What helped me to think out-of-the-box is thinking of axial alignment as what it takes to coincide two lines – the focuser axial line with the primary axial line. I was not thinking in terms of PAE and FAE – that is how CAE and LAE got more attention from me. Since the objective of this thread is to attain the most possible accurate axial alignment, there is no distinction between the importance of eliminating one residual error over another. The objective is to eliminate all errors: PAE, FAE, LAE, and CAE. Of course, eliminating two is enough to eliminate all four.

Jason

[CatseyeMan]

What about stacking P & 3 in the CDP?

It depends on the focal length of the scope. For my 1200mm scope, it is hard to see reflection 3 primarily because of its fuzziness.

The left photo is what I see and the right one is with the camera focused at infinity.


Those with longer focal lengths scopes, reflection 3 will be clearer.

For a reference to others, the following posts explain the relationship between focal length and reflections 1&3 fuzziness:
post1
post2
post3

But the trend seems to be moving towards scopes with shorter focal lengths (< F3.0 scope). Not to mention that imaging scopes tend to have a short focal length – the very scopes that require more accurate focuser axial alignment.

Jason

... so the question is: What other tool/methodology is out there that is "better" than stacking P & 3 in the AC for insuring no significant FAE?

I'm curious as to how the "diffraction ring" effect (using the Glatter 1 mm aperture option) of the return beam around the Barlow Screen exit pupil stacks up on accuracy? (Vic has mentioned this phenomenon)?

[Jason D]

Jim,
FAE, PAE, LAE, and CAE are not independent. Eliminating two out of the four with great accuracy also implies that remaining two are eliminated with the same level of high accuracy. This is a mathematical fact.
Therefore, eliminating LAE with P+2 stacking and CAE with the CAM with great accuracy implies both PAE and FAE are also eliminated with the same great accuracy.
Having said the above, it is highly recommended to start off the steps with CDP. If reflection 3 is clear enough to eliminate all or most of FAE then the rest of the collimation steps will coverage quicker. If reflection 3 is not clear enough to accurately execute CDP then the same accurate axial alignment can still be achieved but it might require and additional iteration or two.
I can't comment about the diffraction ring affect of the laser collimator because I have never tried it.
Jason

[CatseyeMan]

.... Therefore, eliminating LAE with P+2 stacking and CAE with the CAM with great accuracy implies both PAE and FAE are also eliminated with the same great accuracy.

Sounds like the practical CAM option needs to happen sooner than later stay tuned....

[hudson_yak]

Using a cheshire like Catseye’s with a triangle and a ring does add more reading accuracy than the sight-tube/combo tool with the distracting cross-hairs. I am speaking from experience since I have used both tools extensively.

I agree the xhairs do get in the way, in fact I took mine out for that reason, so mine isn't quite as combo as it used to be. Regardless, it is possible to make a mirror mark that matches up well with the cheshire in a combo tool. For example I made a 8mm diameter circle mark which made an easy eccentricity read within the 9.5mm cheshire dark circle of the combo tool.

Just a general comment on the table, it's a good summary of the foregoing posts in the thread, in the context of your intent to discuss AC theory. I was mostly thinking about a relative novice dropping in here, skipping over all that stuff, seeing the table and thinking ah, here's what I need to buy, and perhaps getting a bit mislead as to how best to start. The disclaimer that the table is just your opinion doesn't really help in that respect.

One way to avoid this would have been to not include the table at all.

Mike

[Jason D]

I see your point, Mike
I edited the post. I removed all ratings and replaced them by "check" marks. OK, I left two checks for what I consider to be highly accurate alignment but I did not explain it.
I did not want to delete the post because understanding which errors each tool checks for is important.
Jason

[demiles]

Jason, I have read over your info and find it quite interesting. I have never used a AC to collimate but do use a site tube to align the secondary and a barlowed laser to align the primary. I double check with a cheshire and all seems well. My question is how easy is the AC to use in the dark? The reason I ask is because once my entire scope reaches ambient temp the collimation does shift a small amount. My Obsession will do it and yes it is repeatable in the same direction all the time. My Discovery also did this as well. I collimate just after setting up then have to recheck the primary after 3-4 hours. There is usually a temp. drop of 20 degrees + during this time.

[Jason D]

My question is how easy is the AC to use in the dark?

I use a clip-on light source as shown in the attachment. To preserve eye adaptation, I added few red translucent layers to the light source. You could also clip the light source to the vanes.
Jason

Attached Thumbnails

• 

[turtle86]

Excellent idea. Do you use something like red brake light tape over the light?

My question is how easy is the AC to use in the dark?

I use a clip-on light source as shown in the attachment. To preserve eye adaptation, I added few red translucent layers to the light source. You could also clip the light source to the vanes.
Jason

[Starman1]

Jason, I have read over your info and find it quite interesting. I have never used a AC to collimate but do use a site tube to align the secondary and a barlowed laser to align the primary. I double check with a cheshire and all seems well. My question is how easy is the AC to use in the dark? The reason I ask is because once my entire scope reaches ambient temp the collimation does shift a small amount. My Obsession will do it and yes it is repeatable in the same direction all the time. My Discovery also did this as well. I collimate just after setting up then have to recheck the primary after 3-4 hours. There is usually a temp. drop of 20 degrees + during this time.

I have found that if I collimate perfectly when the telescope is at ambient temperature in the afternoon, I can watch the collimation changing all over the place as the telescope cools. But once it has cooled, the collimation returns. How fast this happens depends on the scope's materials, pole thicknesses, etc.

[mark Jimenez]

Hello,
I just wanted to show how Jim at Catseye helped me get the AC mirror located at the focal plane on my imaging newt. On mine, and many others that are designed for use with a coma corrector, the focal plane often resides above the top of the focuser.
This is a two pupil (XLK) autocollimator with 1.5" extension built into the barrel. Works great.
Also, I find these tools very easy to use in the dark with a small red LED flashlight.
-Mark

Attached Thumbnails

• 

[Jason D]

Nice!!!!!!!!!
Thanks for posting the photo.
Imager enthusiasts will definitely be interested in this option.
Jason

[Jason D]

Rob, I taped pieces of red translucent wrap. Here is an old photo to show you how it would look like at night. The CAM in the photo is an old makeshift CAM.



With the XLK+CAM I use two source lights: One to illuminate the center spot and another to illuminate the CAM. The production CAM will not require the CAM source light – this is what Jim is trying to resolve.

Photos below are without the red translucent wrap.


Jason

[turtle86]

Thanks! Looks like just the ticket for using an autocollimator in the dark.

[CatseyeMan]

Here's what I use and recommend:

Coast Cutlery #TT75331CP LED clip light retrofitted with a bright 3000 MCD Red LED (Gilway E184 found here) and clipped to the spider; I get the best illumination with the light pulled in close to the Secondary as shown in the pics.

Attached Thumbnails

•