434 replies to this topic

[Vic Menard]

Which makes me think that you consider LAE and CAE to be more important than FAE or PAE.

I do not evaluate PAE... As far as FAE, I do not check for it...

You could have just said "yes".

However, during my initial analysis work I would decollimate then take a photo to evaluate P+3 stack to ensure my flow/tools did in fact yield ~0 FAE.
The point is that if you build confidence by experimentation that PAE and FAE always agree with LAE and CAE then PAE/FAE checking becomes optional...I do not expect others to do the same.

I keep reading this synopsis but I just can't imagine aligning my scope using just the offset pupil in the XLK+CAM, never looking at PAE and FAE! I'm either not sufficiently adept making the required iterative (and possibly "intuitive") adjustments with the XLK+CAM (probable), or there's something at a fundamental level that's causing me to subliminally reach for my laser when I try to consider limiting myself to alignments other than PAE and FAE (almost certainly). I just don't have your level of confidence--and that might be because we have different visual acuity, different scopes, and/or different thinking when it comes to assessing and correcting axial errors.

[Jason D]

I just don't have your level of confidence--and that might be because we have different visual acuity, different scopes, and/or different thinking when it comes to assessing and correcting axial errors.

That is not the reason. After I collimated my scope 100s of times over few weeks and got consistent results, that is when I build my confidence. However, I still reach out to my cheshire just because it is there and it is always gratifying to bag another reaffirmation
Jason

[Vic Menard]

...The...CAM as depicted in the photo has the same outer diameter but smaller inner diameter. The idea here is to balance the interior illumination and eliminate the outer one.

The problem I have with the (current beta) CAM is that it's difficult to illuminate properly. If the illumination comes from the side, it looks like there is a residual crescent from the background reflection when in fact it's just the illuminated edge of the washer. I guess it might still be problematic with two annuluses showing, but at least there would be more brightness in the alignment side of the CAM, and since the offset illumination would affect opposite sides of the inside and outside anuluses, it might improve the read.

[Vic Menard]

I just don't have your level of confidence--and that might be because we have different visual acuity, different scopes, and/or different thinking when it comes to assessing and correcting axial errors.

That is not the reason...

I wasn't talking about your level of confidence--I was talking about mine--and I suspect those are the reasons. I actually have issues reading the alignments to the accuracies you've described, and I know that's at least partially due to my visual acuity (it's also the reason I rely so much on my Glatter). I also know that while my longer focal length helps me to see P-3 more focused, the distance I must observe P-3 (or P-2) from limits the read somewhat (I'm sure visual acuity is at least partly to blame here as well). Finally, I can't get my mind around ~zero FAE, PAE, or any XAE. I have a reasonable understanding of functional or usable read accuracies and how they apply to tolerances and mechanical registrations. And while I understand CAE and LAE, I still think a direct read of PAE and FAE is better than an interpreted read of PAE and FAE, and this is almost certainly because of my experience with the XLK+CAM.

And I'm not a big advocate of reiteration--recollimating over and over again. It's a requirement for good (enough) secondary mirror alignment, but it doesn't have to be for (very) good axial alignment--even considering mechanical registration, etc. With the XLK, I found that if P-2 was suffieciently misaligned to be obvious in the 22 f/4, it was quicker (for me) to just reassess P-3 with the autocollimator and PAE with the Glatter (and be particularly careful) to arrive at a better P-2 alignment in the XLK. I found this worked consistently, and almost always to a "sufficient" precision (P-2 looked like a "perfect" hexagram) on the second try.

[Jason D]

it looks like there is a residual crescent from the background reflection when in fact it's just the illuminated edge of the washer.

Ummm, this is what I get. Left photo is for a slightly misaligned CAM and the right photo for a perfectly aligned CAM.

The barley seen trace of light around the dark ring of the right photo is only seen by the camera. When I align the CAM, it looks pitch black. I do not know why it is different for you.
Jason

[Jason D]

Vic, something sounds amiss. I suspect your XLK+CAM might be slightly misaligned or maybe your scope with different specs than mine is manifesting a new issue that I have not run into with my scope. I certainly explained all the discrepancies I have run intowith my scope and the newly gained knowledge helped me to achieve collimation with consistent visual cues from all reflections. Maybe we need to follow up offline with Jim. Maybe Jim can ship you another XLK+CAM or I can get hold of yours temporarily to analyze it.

As far as your re-iteration comment, I am also not a fan of it. If P+3 stack and a cheshire agree but both disagree with P+2 and the CAM then we can either stop or re-iterate to get all four visual cues to agree. If we do not re-iterate then we end up with the best collimation an XL+Cheshire could provide. If we decide to re-iterate then we improve our collimation.
In other words, I see the required re-iteration required by the XLK+CAM as above and beyond what the XL+Cheshire provides.
Jason

[CatseyeMan]

... You're selling it - a BlackCat placed precisely so that the focal plane is between the pupil and the ring.

...If placed properly, it directly detects PAE (and nothing else).

Touche'

Actually, I think I realize the disconnect after reading the later posts you participated in about Secondary Sag affecting both PAE and FAE.

Here's the point I was trying to make:

Starting from theoretical perfect axial alignment with PAE=0 and FAE=0, if the Secondary is then tilted to cause an FAE of 2mm at the Primary, I don't consider PAE to have been introduced. In other words, it's the Secondary movement that has caused the problem - not a shift of the Primary. If one inserts a Cheshire into this Scenario, they will see a "pseudo" PAE error and unless they realize that it is a Secondary tilt issue, they will be misled into adjusting the tilt of the Primary to recenter the spot in the Cheshire ring.

This is exactly why (in lieu of the CDP) to achieve axial alignment that the iterative process is required alternating the use of the Cheshire (or Barlowed laser) to adjust the Primary tilt (effecting PAE) and the autocollimator to adjust the Secondary tilt/rotation (effecting FAE). Fortunately in this method, the errors diminish to negligibility after 2-4 iterations.

[CatseyeMan]

Vic, something sounds amiss.

I would suspect Vic's illumination method is the likely difference. I don't have another CAM unit built (other than my beta which I will keep here) so perhaps Vic will lend his to you for comparison to yours.

[Nils Olof Carlin]

Starting from theoretical perfect axial alignment with PAE=0 and FAE=0, if the Secondary is then tilted to cause an FAE of 2mm at the Primary, I don't consider PAE to have been introduced.

Jim, I don't quite see how you think here. The primary's axis between the primary and the secondary won't change, of course, but the reflected axis reaching the actual focal plane will - less than the FAE, as much as the distance focal plane-secondary is less than the distance secondary-primary, but still in this case some 0.5 mm or a bit less.
If you had moved the FAE by tilting the focuser, the PAE introduced would be very much smaller.

Nils Olof

[nsldvd]

Concerning CAM illumination, would it not be possible to coat the desired area of the CAM with a phosphorescent coating (http://www.glonation.com/)? After holding it next to a light source it should "glow" on its own for some time afterwards.

[CatseyeMan]

Concerning CAM illumination, would it not be possible to coat the desired area of the CAM with a phosphorescent coating (http://www.glonation.com/)? After holding it next to a light source it should "glow" on its own for some time afterwards.

Great minds think alike . I've actually used a phosphorescent plastic media for the bright CAM element in preliminary beta development work - It works great for the first 5-10 seconds after charging but becomes too dim too quickly for practical use. I plan to try some alternate materials to test for longer half-life - thanks for the link.

[Vic Menard]

it looks like there is a residual crescent from the background reflection when in fact it's just the illuminated edge of the washer.

...this is what I get...I do not know why it is different for you.

I think I know now. Your image clearly shows a directed beam that illuminates only the white washer. The black washer is carefully shielded from any extraneous illumination. I was clearly seeing some off axis illumination on the black washer from my diffuse light source (this even happened with Jim's test setup at PSSG with a "glow in the dark" white washer--must have been getting light from somewhere else). I'll need to try again when I get the 22 out for the next observing session.

[Vic Menard]

Vic, something sounds amiss. I suspect your XLK+CAM might be slightly misaligned or maybe your scope with different specs than mine is manifesting a new issue that I have not run into with my scope. I certainly explained all the discrepancies I have run into with my scope and the newly gained knowledge helped me to achieve collimation with consistent visual cues from all reflections...

I don't think it's a misaligned CAM (I was having the same issues with Jim's beta at PSSG), more than likely it's just a lack of persistence (almost certainly attributable to the round trips back and forth from the 22-inch primary mirror to the focuser, coupled with pushing the limits of the read), although I think I was expecting a quicker resolution on Jim's much shorter test rig with no intervening secondary mirror.

If P+3 stack and a cheshire agree...

...but they always agree since they are direct reads...

...but both disagree with P+2 and the CAM then we can either stop or re-iterate to get all four visual cues to agree.

That's the rub. If after several iterations I still can't get all four (or just the CAE and LAE alignments) aligned to the resolution of the signature reads, I feel compelled to finish with a precision PAE alignment.

...I see the required re-iteration required by the XLK+CAM as above and beyond what the XL+Cheshire provides.

It's certainly above and beyond, but until I achieve your level of confidence regarding the derived PAE precision, and the persistence to push through the necessary reiterations to bring all four signatures into alignment, I'll be keeping a Cheshire (or Barlowed laser) handy for a quick look at (or tweak to) the primary axial correction.

[CatseyeMan]

Agreed Nils.

See my clarification post above to Sixela.

In the Cheshire/AC iterative process of axial alignment, for conceptual simplicity I promote the concept of adjusting the Secondary tilt/rotation to minimize FAE and adjusting the Primary to minimize PAE when in fact as you state, adjusting the Secondary in the process minimizes both FAE and PAE when restacking the images in the central pupil of the AC after Cheshire use. To me it's less confusing to limit the Secondary to only FAE effect for tutorial purposes when conveying the logistics of optical component adjustment sequence with this method.

As for routine adjustment of the focuser tilt to correct FAE, I discourage this practice. Once reasonable focuser squareness has been achieved and initial axial collimation has been accomplished, I attribute any subsequent FAE introduction to Secondary support mechanics shifts which IMHO need to be reversed rather than compensated for via focuser tilting. That said, I also understand that it's sometimes easier to change the tilt of the focuser axis than to re-adjust the Secondary, but at some point if the problem persists with creeping Secondary changes, the original source of the error (the Secondary) may need to be addressed.

Regards,

[Nils Olof Carlin]

Jim, it makes sense indeed to get the focuser axis right before fine-tuning the primary.
One possibility that kind of suggests itself is to put a very thin reflective annulus around the AC pupil - say, 1/4" outer dia (and a mm wide or so?), matching the center hole on the primary marker. This would show the PAE directly, in the same way as the cheshire does, with no penalty to any of the other indicators, and possibly help shortening the procedure some.

Nils Olof

[Jason D]

I think I know now. Your image clearly shows a directed beam that illuminates only the white washer. The black washer is carefully shielded from any extraneous illumination.

I use two clip-on light sources as depicted in the photo of post #3537603. One meant for the CAM and another for the center spot.
I carefully place the first light source clipped on to one of the spider vanes to illuminate the CAM’s reflective ring. The bottom of the focuser shields light from the dark ring. I use the second light source for the center spot. By only keeping the first light source on, I further enhance the CAM contrast (right photo).



By the way, the above photo shows a CAE of 0.1mm

Jason

[Jason D]

Jim, it makes sense indeed to get the focuser axis right before fine-tuning the primary.
One possibility that kind of suggests itself is to put a very thin reflective annulus around the AC pupil - say, 1/4" outer dia (and a mm wide or so?), matching the center hole on the primary marker. This would show the PAE directly, in the same way as the cheshire does, with no penalty to any of the other indicators, and possibly help shortening the procedure some.

Nils Olof

Jim and Vic experimented with this idea years ago. Jim called it the AutoCat. Check out the following post
AutoCat

Here is another post

I have expressed my concerns about the idea in this post

The area around the central pupil is highly sensitive because of the “G4” focal point which is basically the background AC reflection for the central pupil. The “G4” naming comes from your autocollimator analysis page, Nils Olof. If any part of the G4 hits the reflective annulus ring, the whole background will lighted up which would complicate and interfere with the autocollimator operation which is what Vic indicated when he experimented with it.

Jason

[Jason D]

If P+3 stack and a cheshire agree...

...but they always agree since they are direct reads...

With "agree", I meant both "agree" that alignment has been met which is indicative of axial alignment. It is certainly possible to have misaligned P+3 view but aligned cheshire view. As you know, both are independent of each other.

...but both disagree with P+2 and the CAM then we can either stop or re-iterate to get all four visual cues to agree.

That's the rub. If after several iterations I still can't get all four (or just the CAE and LAE alignments) aligned to the resolution of the signature reads, I feel compelled to finish with a precision PAE alignment.

My point is that if you can't get all views (P+2, CAM, P+1, and cheshire) to agree at least once then you need to investigate to root cause the issue. If you know your scope and tools are capable to getting all views aligned but you decide at the moment not to take the time to align all and rely on the cheshire then that is a different story. I am trying to differentiate between: “Yes I can do it if I take the time” versus “No it can’t be done”

...I see the required re-iteration required by the XLK+CAM as above and beyond what the XL+Cheshire provides.

It's certainly above and beyond, but until I achieve your level of confidence regarding the derived PAE precision, and the persistence to push through the necessary reiterations to bring all four signatures into alignment, I'll be keeping a Cheshire (or Barlowed laser) handy for a quick look at (or tweak to) the primary axial correction.

It comes down to the trust level of your tools. I aligned with the cheshire and taken photos. I aligned with the XLK+CAM and taken photos. I’ve taken 100s of photos. I found out that the XLK+CAM has consistently at least matched the cheshire results. That is part of confidence building and trusting the tools. That is why I feel comfortable with the XLK+CAM. Having said that, I still reach out for my cheshire because it is there and to keep my XLK+CAM confidence high.

Jason

[Vic Menard]

...I am trying to differentiate between: “Yes I can do it if I take the time” versus “No it can’t be done”...It comes down to the trust level of your tools.

I'm sure it can be done, but for me it comes down to my ability to accurately read the signature alignments, and then make the best available correction. I'm pretty good at reading concentricities and hexagrams, but trying to balance the two against each other (CAM and LAE) became frustrating (for me) several iterations before I reached the full read precision. That said, I feel I can read P-3 to almost the same resolution as P-2 (given my 88-inch focal length), but I know I can't guarantee zero FAE with my best P-3 read (P-2 proves that). In that respect, I can't fully trust my read of P-2 to guarantee "perfect" parallelism. And since the CAM puts the alignment reference point (what P-2 is parallel to and the only actual "point" on the alignment axis when CAM and LAE are used in tandem) a focal length away from the focal plane, I just feel more comfortable with another point that's actually on the alignment axis.

[Starman1]

Vic's comments about the readability of tools brings to mind my first uses of the XLK. The two stacks are P-2 and 1-3. P-2 is clear and sharp, while 1-3 is a little "fuzzy".
So differentiating which group is which should be easy, right?
Wrong.
With my glasses on, P-2 is fuzzy and 1-3 is sharp. With my glasses off, P-2 is sharp and 1-3 is fuzzy. At first, I couldn't tell which stack was which because I wasn't certain whether glasses on or glasses off gave the correct read. I use glasses for the sight tube (to see the wires), and no glasses for the cheshire (because I can focus that far away).
Then Jason mentioned the P-2 stack didn't suffer from parallax as much as 1-3. By moving my head back and forth it became obvious which stack was P-2, because 1-3 would stack and unstack while P-2 remained nearly constant. Now I know that I always put the XLK into the focuser one way (I have a mark on the body) and NOT use glasses.

So when it comes to readability of the tools, you have to experiment a bit to adapt to the tool use for your circumstances (focal length, vision, light levels).

I'd still like to see the commercial development of the Krupa collimator. It would be a lot easier to read than the bottom of a barlowed laser or a BLUG, and it could be parallax-free.

[Jason D]

Vic, Don, points well-taken...

Don, I found it interesting how your eye-glasses make a difference. I do not wear glasses (not yet) so I never had a similar experience. But your experience made me think and I came up with an explanation. You must be long sighted (OK I know this is a personal matter; therefore, you do not have to answer). When the eye is long sighted then the lens focuses images behind the retina. Which means that reflections P and 2 will form behind the retina; therefore, they will look fuzzy. Now, reflections 1 and 3 by the virtue of their location behind our heads will always form in front of reflections P and 2. For what is considered health eyes, P&2 will end up on the retina (sharp images) and 1&3 will end up in front of the retina (fuzzy images). But for long sighted eyes, the whole thing will move away from the eye lens; therefore, P&2 will end up behind the retina (fuzzy image) and the 1&3 will end up ion the retina (sharp images). If this is true, then short sighted eyes are out of luck for all four reflections since none of the reflections will form on the retina.

Talking about readability, I am still interested to hear your opinion about the following center spot shape.



Left: The suggested center spot
Middle: View when it is used with the cheshire. The outer edge of the newly suggested center spot will be easier to align with the cheshire.
Right: View with P+2 stacked. The P+2 circle will make it easier to stack

And the newly suggested center spot will give the impression your mirror is radioactive

Jason

[sixela]

Jason, I think you should start the "now really concise, I promise" autocollimator thread .

[Jason D]

Jason, I think you should start the "now really concise, I promise" autocollimator thread .

Ummm, may be I shall call it the "mother of all concisions" thread

[Galaxyhunter]

Ummm, may be I shall call it the "mother of all concisions" thread

That would be fine. Post the images that we need to see through the tools, then LOCK the thread.

[erick]

Before this thread is locked a question:- It seems to me that the suggestion is that a 2" AC would be more useful than a 1.25" AC. In order to clearly identify the reflections, the wider AC mirror is better? True?

I have a 1.25" AC, but have never had a mechanically stable scope to use it seriously on - but I have a quality scope coming shortly.