Jump to content


Photo

Autocollimator acuracy

  • Please log in to reply
34 replies to this topic

#1 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 15 February 2013 - 01:21 PM

In another thread, I raised the question about the accuracy of the AstroSystems 1.25" autocollimator. (This unit is the only viable choice for those of us wanting a 1.25" autocollimator.) I bought one of them, and after stacking the reflections, as per DOn's suggestion, I rotated the tool in the drawtube in 90* increments, and examined the pattern of the reflections. In both scopes, 4 reflections were visible. (Contrary to the manufacturer's instructions, I secured the tool with the setscrew after each rotation, the results were much better.)

On my F/8 scope, there was very little change. The maximum deviation was about 1/2 the diameter of the center donut (which is a 5/16" circle).

On my F/5 scope, the maximum deviation took one of the reflections to the edge of the mirror, about .46" off-center. (Interstingly, the axis of maximum error is exactly in line with one of the adjustment screws on the top... suggesting that the unit could be adjusted to better accuracy.)

When I asked the manufacturer about this, they replied:
"The mirrors in our autocollimators are aligned optically on a 100 foot path, giving 2-3 arc minute accuracy. I don't believe that the accuracy of the ID of a drawtube is accurate enough to see any deviation."

So the questions are:
Is that degree of accuracy good enough?
At what focal ration does that level of uncertainty translate into image deterioration?
If it were better, would the results translate into a visibly better image?

Given the small deviation, I think we can assume that, for my F/8 scope, this level of accuracy definitely is good enough.

But is it good enough for an F/4.9 scope?

I think it would be very beneficial to have some quantifiable answers to this. Autocollimators provide a very real benefit, and there's no reason why every end user who faces this dilemna needs to redo the same research that every previous person who faced it want through.

#2 spencerj

spencerj

    Apollo

  • *****
  • Posts: 1149
  • Joined: 17 Nov 2004
  • Loc: Londonderry, NH

Posted 15 February 2013 - 01:45 PM

Given the small deviation, I think we can assume that, for my F/8 scope, this level of accuracy definitely is good enough.

But is it good enough for an F/4.9 scope?


Definitely good enough in the F8 scope. At F4.9 it would still be ok. Not perfect, but better than just a collimation cap.

If the F4.9 has a 2" focuser, you are introducing other factors (like the 2" to 1.25" adapter). That is not ideal, but you can always confirm the result with a quick star test.

As for there being definitive answers to the accuracy of autocollimator reflections, they exist. A quick search of the archives will quickly lead you to some very long and complex threads.

#3 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 15 February 2013 - 01:58 PM

Thanks Jason. I'd rather not confuse this discussion with adapter issues, etc. Let's assume those details are in order and just deal with the AC uncertainty.

The thrust of my questions isn't about the accuracy of the reflections, but to see if we can define, in laymans terms, at what point this autocollimator (the ONLY choice we have for 1.25" systems) cannot be relied on, not just for technical reasons, but because of the image integrity that could result.

#4 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4766
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 15 February 2013 - 01:59 PM

...Given the small deviation, I think we can assume that, for my F/8 scope, this level of accuracy definitely is good enough.

If your f/8 scope is a 6-inch aperture, the high magnification focuser axial tolerance is about 0.2-inch (about 3-percent of the aperture diameter). The error you describe is about 0.16-inch, so you're probably OK. I would prefer a P-3 read after carefully decollimating the primary mirror to specifically target the tilt error.

But is it good enough for an F/4.9 scope?

It depends on the aperture and whether or not you intend to use a coma corrector.

I think it would be very beneficial to have some quantifiable answers to this. Autocollimators provide a very real benefit, and there's no reason why every end user who faces this dilemma needs to redo the same research that every previous person who faced it want through.

There are potentially significant registration issues when using a 1.25-inch autocollimator in a 2-inch focuser. To achieve the "very real benefit" you're looking for (assuming your high magnification eyepieces are 1.25-inch and highly corrected for off-axis aberrations), you may need to invest in a precision 2- to 1.25-inch adapter.

#5 csrlice12

csrlice12

    Voyager 1

  • *****
  • Posts: 10443
  • Joined: 22 May 2012
  • Loc: Denver, CO

Posted 15 February 2013 - 02:11 PM

Yup, trying to use a 1.25" collimater on a 2" scope could be an exercise in frustration if it's not a precision machined one (and most are NOT, they may fit fine for visual/photo use, but not collimation). It's why I bought the Glatter/Tublug...no muss, no fuss, drop the laser on the floor? No problem.....any other system????????????

#6 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4766
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 15 February 2013 - 02:24 PM

I'd rather not confuse this discussion with adapter issues, etc. Let's assume those details are in order and just deal with the AC uncertainty.

Since the adapter defines the focuser axis for your 1.25-inch autocollimator, if the axial errors we're assessing are smaller than the manufacturer's spec, it may be difficult to leave the adapter out of the discussion.

The thrust of my questions isn't about the accuracy of the reflections, but to see if we can define, in laymans terms, at what point this autocollimator (the ONLY choice we have for 1.25" systems) cannot be relied on, not just for technical reasons, but because of the image integrity that could result.

Axial alignment tolerances (high performance/high magnification):
Focuser axis w/o coma correction, about 3-percent of the aperture diameter.
Focuser axis with coma correction, about 0.5-percent of the aperture diameter.
Primary mirror axis, about 0.005mm times the focal ratio cubed.

#7 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 15 February 2013 - 04:37 PM

If your f/8 scope is a 6-inch aperture, the high magnification focuser axial tolerance is about 0.2-inch (about 3-percent of the aperture diameter). The error you describe is about 0.16-inch, so you're probably OK. I would prefer a P-3 read after carefully decollimating the primary mirror to specifically target the tilt error.


I haven't ID'ed the individual reflections yet, it's tougher with round center spots... I'm working on getting triangular ones.

But is it good enough for an F/4.9 scope?
It depends on the aperture and whether or not you intend to use a coma corrector.


The F/4.9 has a 9.8" dia. mirror, no corrector.

There are potentially significant registration issues when using a 1.25-inch autocollimator in a 2-inch focuser...
Since the adapter defines the focuser axis for your 1.25-inch autocollimator, if the axial errors we're assessing are smaller than the manufacturer's spec, it may be difficult to leave the adapter out of the discussion.

Understood, and I've tackled that problem independently and have done things to make insertions very repeatable, so one issue at a time, please... let's just talk about the contribution of the autocollimator's worst-case inaccuracy... that is the matter at hand.

Axial alignment tolerances (high performance/magnification):
Focuser axis w/o coma correction, about 3-percent of the aperture diameter.
Focuser axis with coma correction, about 0.5-percent of the aperture diameter.
Primary mirror axis, about 0.005mm times the focal ratio cubed.

Great, thank you for those guidelines, Vic.

trying to use a 1.25" collimater on a 2" scope could be an exercise in frustration

Doesn't have to be... I know how to address one problem at a time... it's doable... I don't want another laser... I don't want 2" tools... I want an 1.25" autocollimator that is reliably accurate and useful.

#8 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4766
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 15 February 2013 - 06:41 PM

I haven't ID'ed the individual reflections yet, it's tougher with round center spots... I'm working on getting triangular ones.

If you can get a near perfect stack, when you decollimate the primary mirror reflections 1 and 2 will move away from P (the primary mirror center spot), leaving reflection 3 hiding (mostly or completely) behind P. Then you can tweak the focuser axial alignment by adjusting the secondary mirror tilt to carefully stack P-3. Once P-3 is stacked, you can rotate the autocollimator and get a good read of any inherent autocollimator mirror tilt.

(You may also be able to tweak the P-3 alignment by rotating the autocollimator in the adapter.)

But is it good enough for an F/4.9 scope?
It depends on the aperture and whether or not you intend to use a coma corrector.

The F/4.9 has a 9.8" dia. mirror, no corrector.

Three percent of your 9.8-inch primary mirror diameter is about 0.3-inch. Once you can reliably isolate P-3 and you've determined any inherent tilt error in the autocollimator, you should be able to get a better understanding of the autocollimator's usability. Remember, P-3 magnifies any residual focuser axial error 2X.

I am a little confused though. If your f/8 scope is indeed a 6-inch, the focal lengths of your two scopes are nearly identical, which means the reflection deviation caused by the autocollimator mirror tilt error should be pretty much the same in either scope...

...Regardless, once you've assessed and corrected the focuser axial alignment, I would suggest using a Cheshire eyepiece (or collimation cap or other derivative) to assess and correct the primary mirror axial alignment.

#9 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 15 February 2013 - 07:58 PM

If you can get a near perfect stack, when you decollimate the primary mirror reflections 1 and 2 will move away from P (the primary mirror center spot), leaving reflection 3 hiding (mostly or completely) behind P. Then you can tweak the focuser axial alignment by adjusting the secondary mirror tilt to carefully stack P-3. Once P-3 is stacked, you can rotate the autocollimator and get a good read of any inherent autocollimator mirror tilt.


That's great, I'll try it in the morning.

(You may also be able to tweak the P-3 alignment by rotating the autocollimator in the adapter.)


Now that's called making create use of the problem! :)

Three percent of your 9.8-inch primary mirror diameter is about 0.3-inch. Once you can reliably isolate P-3 and you've determined any inherent tilt error in the autocollimator, you should be able to get a better understanding of the autocollimator's usability. Remember, P-3 magnifies any residual focuser axial error 2X.

OK, thank you. So, despite the different F ratios, the allowable FAE is the same for both scopes... interesting.

I am a little confused though. If your f/8 scope is indeed a 6-inch, the focal lengths of your two scopes are nearly identical, which means the reflection deviation caused by the autocollimator mirror tilt error should be pretty much the same in either scope...

If that's the case, then I definitely have to look into it further. One possible cause is that the current center spots on both scopes are not accurately placed and destined to come off as soon as the new ones arrive. The 9.8" spot is further off than the 6", and its primary is probably aligned to the accurate center point (a known offset). Everything I'm doing now is a) tool assessment/familiarization, and B) rehearsal to get comfortable with the procedure, for when the new spots arrive and I can do it for real...

Regardless, once you've assessed and corrected the focuser axial alignment, I would suggest using a Cheshire eyepiece (or collimation cap or other derivative) to assess and correct the primary mirror axial alignment.

That's a definite will-do. I plan on doing this unto all the tools agree and it can be repeated with the same results!

Thanks again for your help.

Edit - I just learned, Randy at AstroSystems had some circumstances to deal with, he's sending a new AC out (after checking it) along with the triangular spots.


#10 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6637
  • Joined: 21 Oct 2006
  • Loc: California

Posted 16 February 2013 - 02:43 AM

When I asked the manufacturer about this, they replied:
"The mirrors in our autocollimators are aligned optically on a 100 foot path, giving 2-3 arc minute accuracy. I don't believe that the accuracy of the ID of a drawtube is accurate enough to see any deviation."


Let us think about the above logically.

If you buy a laser collimator and followed the proper steps to collimate then rotated the laser collimator by 180 degrees to notice the beam has shifted by 2mm off-center, will you be satisfied? I just described a laser collimator with 3 arc-minutes error used with a 1200mm FL scope just like my XT10.

An autocollimator with a 3 arc-minutes error is the same as the above described laser collimator.

I believe autocollimators have to meet higher bar than other collimation tools in terms of precession. Any other autocollimators will either not add value or will make collimation worse.

Can you imagine some laser collimator vendor adding a statement on their website claiming their laser collimators trace a 2mm diameter circle at 1200mm!!!!

Jason


#11 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1580
  • Joined: 26 Jul 2004

Posted 16 February 2013 - 08:27 AM

An autocollimator with a 3 arc-minutes error is the same as the above described laser collimator.

I believe autocollimators have to meet higher bar than other collimation tools in terms of precession. Any other autocollimators will either not add value or will make collimation worse.



If you do an apparently exact collimation with this AC of 3 arcmin error, the focuser axis will be 1 mm off the exact center, but the primary's collimation is not affected, regardless of focal ratio (at least as long as the AC is close to the focal plane).

With a paracorr or similar coma corrector, the primary axis should ideally be centered at the level of the lens system, not the focal plane - not easily done in practice, I think. But the error in this case would be less than 1/10 mm, and the coma induced would be approximately what it is so far off center without the corrector.

Nils Olof.

#12 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4766
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 16 February 2013 - 08:53 AM

If you do an apparently exact collimation with this AC of 3 arcmin error, the focuser axis will be 1 mm off the exact center, but the primary's collimation is not affected, regardless of focal ratio (at least as long as the AC is close to the focal plane).

Since we're aligning P to reflection 1 (not the pupil), I would expect a residual primary mirror alignment error of 1/2mm (with a truly "perfect" stack). That's the reason I suggested a Cheshire or equivalent.

#13 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6637
  • Joined: 21 Oct 2006
  • Loc: California

Posted 16 February 2013 - 10:46 AM

Hello Nils Olof,
My points are based on logic and common sense:
1- Given a perfect casing and a laser with its aperture perfectly aligned with the casing opening but the laser collimator is tilted by 3 arc minutes inside the casing, will such a laser collimator be acceptable by its owner knowing it will trace a circle with 2mm diameter for 1200mm F4.7 reflector?
2- Assuming the above hypothetical described laser collimator is produced in quantity and offered for sale by some vendor, will it be appealing to mention that the laser collimators from that vendor are guaranteed to trace a circle not larger than 2mm in diameter at 5 feet!!
3- A perfect AC but with a 3 arc-minutes tilted mirror matches is equivalent to the above laser.
4- Specifically to Astrosystems, they claim their AC testing is done at 100ft to ensure a maximum error of 3 arch-minutes. An equivalent laser collimator will trace a 2 inch diameter circle at that distance. Is that an acceptable test or even a proper claim? If I claim to have aligned a laser collimator at a distance of 100ft then it is an expectation that the laser beam will be confined to an area of 1mm or 2mm upon rotation. It is meaningless to claim that I have aligned a laser collimator at 100ft with the laser beam confined to a circle with 2 inch diameter upon rotation.
If any of the above math is incorrect, please correct me.
If my logic does not make sense, please correct me.

But there is more to evaluating an autocollimator aside from a mirror tilt.
http://www.firstligh...tocollimator...

Jason

#14 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6637
  • Joined: 21 Oct 2006
  • Loc: California

Posted 16 February 2013 - 10:51 AM

Since we're aligning P to reflection 1 (not the pupil), I would expect a residual primary mirror alignment error of 1/2mm (with a truly "perfect" stack). That's the reason I suggested a Cheshire or equivalent.

Hello Vic, the "PAE=0.5*FAE" formula can't be applied in this case since the center spot is assumed to be centered. Your statement would make sense only of the center spot is off. In this case, PAE will be 1/2 of the center spot placement error.
Jason

#15 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1580
  • Joined: 26 Jul 2004

Posted 16 February 2013 - 11:07 AM

Since we're aligning P to reflection 1 (not the pupil), I would expect a residual primary mirror alignment error of 1/2mm (with a truly "perfect" stack). That's the reason I suggested a Cheshire or equivalent.



Here's how I think: with perfect stacking, the AC axis (if you call it that - I do) points at the primary's spot, and the primary's axis points to the center of the AC - they are coincident. But if the AC mirror is tilted 3 arcmin to one side relative to the barrel, tilting the barrel 3 arcmin to the other side will place the AC in the position where it again shows perfect stacking. Both the AC and the primary axes are coincident. But the AC barrel, and thus the focuser, is now tilted 3 arcmin.

If you let the AC be tilted but the focuser straight, and stack P and 1, you have indeed introduced a 1/2 mm miscollimation of the primary axis, as you say. But in this situation, reflections 2 and 3 are no longer stacked with P - they are off by 2 mm (in opposite directions).

Nils Olof

#16 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 16 February 2013 - 11:08 AM

Thanks for your input, guys. It looks like 1mm is the consensus for the maximum FAE of this unit.

I would like to clarify, the purpose of this thread is not to trash AstroSystems or their products. Quite the opposite. I'm using this info to encourage them to raise their standards and offer a 1.25" unit that is more precise and meets the needs that exist, because NOONE ELSE IS OFFERING ONE. They are the only game in town in 1.25" autocollimators. It needs to be a good one. If that results in it costing a little more, so be it. This is a calibration tool. It MUST BE precise.

When I sent the info to AS, he did not challenge it, but said he would send a replacement and would personally check it first. That's commendable service. But it would be better if it were "right" in the first place. As it is, the big winner is the USPS who gets paid to carry it three times.

And I also want to encourage Jim Fly to come out with a 1.25" version. There's nothing like competitive pressure to raise quality standards, and Jim's standards are pretty high. From what I have learned, unless you are using 2" eyepieces exclusively, there are many good reasons to standardize on the 1.25" format for collimation. Most importantly, since collimation accuracy becomes more critical at high magnifications, I want a collimation system that duplicates the setup of my high-magnification eyepieces.

And in keeping with that POV, I'd also like to see an AC that, just like an eyepiece, is designed to register off of the inside of the drawtube and not off of the drawtube lip. Both makers of autocollimators tell the user to leave the tool loose on top of the drawtube, using the lip as the reference. This introduces the requirement that the drawtube or adapter be perfectly square to that portion of its cylinder that an eyepiece registers onto. There's no good reason to require that. And it is not stated outfront. The user is left to discover that on his or her own.

The reference vector is defined by how an eyepiece sits secured in the drawtube. The tool should calibrate that vector from the same surface that an eyepiece would.

#17 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4766
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 16 February 2013 - 11:25 AM

Thanks for your input, guys. It looks like 1mm is the consensus for the maximum FAE of this unit.

At 1200mm focal length.

The reference vector is defined by how an eyepiece sits secured in the drawtube. The tool should calibrate that vector from the same surface that an eyepiece would.

To the prescribed tolerance--agreed.

#18 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6637
  • Joined: 21 Oct 2006
  • Loc: California

Posted 16 February 2013 - 11:31 AM

Thanks for your input, guys. It looks like 1mm is the consensus for the maximum FAE of this unit.

As I mentioned, there is more to AC evaluation then a tilted AC. Refer to the link I provided.

I would like to clarify, the purpose of this thread is not to trash AstroSystems

What would you even say that?
I believe their claim is misleading. Readers see the large (100ft) number followed by the small (3 arc minutes) and assume a precise testing setup but it is not. I have already clarified their testing setup is equivalent to aligning a laser with 1 inch radial error at 100ft.
Aside from the 3 arc-minutes claim, their documentation and procedure include errors. I have clarified these errors several times but AS do not seem interested in making any corrections. For example, you should never end an autocollimator session with only adjusting the secondary mirror without checking the primary.
Jason

#19 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 16 February 2013 - 11:46 AM

Give it a rest, Jason. If you want to trash them, please start a separate thread for it. They are not the only company that uses obscure specmanship, has unstated requirements, or has quality control issues. This is a "trust but verify" world.

In the past few months I've bought a new dob that came with a disfigured mirror and a cockeyed focuser; two combo tools that were way out of spec; two eyepieces that were clearly defective. They put me on an "accelerated learning process" to assess their defects, and either correct or replace them.

The AstroSystems 1.25" autocollimator is a unique product. The company has been responsive. I'm working for a positive outcome, and to nudge them to raise the bar in the future.

#20 Howie Glatter

Howie Glatter

    Vendor

  • -----
  • Vendors
  • Posts: 858
  • Joined: 04 Jul 2006

Posted 16 February 2013 - 12:02 PM

Precaud said:
" in keeping with that POV, I'd also like to see an AC that, just like an eyepiece, is designed to register off of the inside of the drawtube and not off of the drawtube lip. Both makers of autocollimators tell the user to leave the tool loose on top of the drawtube, using the lip as the reference. This introduces the requirement that the drawtube or adapter be perfectly square to that portion of its cylinder that an eyepiece registers onto. There's no good reason to require that. And it is not stated outfront. The user is left to discover that on his or her own.
The reference vector is defined by how an eyepiece sits secured in the drawtube. The tool should calibrate that vector from the same surface that an eyepiece would."

Precaud, it's very good that you have raised this point. Many people advocate pressing the accessory face to face against the drawtube when tightening the drawtube clamp screw, and this can prevent accessory tipping if the drawtube inside diameter surface is accurately square with its circular front face, as it usually is. However, it is possible for a drawtube bore to be out of square with its face, and then tightening the clamp screw will force the accessory out of full contact with the drawtube face.
For this and other reasons I have advocated the use of the drawtube bore rather than the front flange face as the reference surface for focuser axis direction.
But then we still have the problem of inconsistent registration of a smaller cylindrical accessory within a larger cylindrical drawtube due to the ease of slippage between the two clamped surfaces which are almost flat(in the sense of infinitesimals) at their line of contact. This is the problem that my Parallizer™ principle is intended to fix. Although I'm using it with my 2" to 1.25" adapter, I think the prime places to apply it are in the drawtube itself, and also on the cylindrical reference surface of collimation tools. Interested parties are invited to contact me, as it could improve alignment tolerances, as well as help fund my retirement.
(I have a patent on it :-)

#21 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1580
  • Joined: 26 Jul 2004

Posted 16 February 2013 - 12:03 PM

Jason,

I fully agree with your views on the First light optics page.

But a question such as "will such a laser be acceptable to its owner..." is a question about the owner rather than the laser or AC owned. And it is basically of emotions, not logic or even common sense.

I can answer for myself: I like to think in terms of tolerances - in engineering, you (or rather your professional customer) decide the acceptable tolerances, then you see to it that you meet them, but you don't expect your customer to pay extra for much tighter tolerances than needed. I might expect a layman customer, not accustomed to think in terms of engineering, to react differently - how many of us don't go for "perfect collimation"?.

If I had reason to believe that those 3 arcmin (laser or AC) were the maximum errors, and the workmanship in general acceptable, I would accept it if the price were right, knowing that collimation would not introduce significant optical problems. I would mark the laser, or AC, to make sure I always insert it rotated the same way, though.

But I would consider stability as well, perhaps see it as the more important aspect - a knock must nor send the error far outside tolerances.

We know that some people buy inexpensive laser collimators that are known to be easily knocked way out of tolerances, others pay premium price for better stability and workmanship.

But if you suggest that any manufacturer could easily do better than that, I am sure it is so - at a cost depending of time needed. Possibly only marginal, and highly likely a good investment in professional reputation.

If the testing is done at 100 ft or other distance is irrelevant, if the tolerance is given in arcmin (or arcsec).

Nils Olof

#22 Jason D

Jason D

    Fly Me to the Moon

  • *****
  • Posts: 6637
  • Joined: 21 Oct 2006
  • Loc: California

Posted 16 February 2013 - 12:36 PM

Nils Olof,
I believe there are two different views being discussed. If the collimation goal is to stay within tolerance then reasonable collimation tool errors will be acceptable. But if the collimation goal is to stay within tolerance then why even bother getting an autocollimator?
But there is another view which is what I have been trying to convey. Many, including me, strive for “perfect” collimation and let scope mechanical imperfections use up collimation tolerances. A quality autocollimator with proper use and experience can get me there. That is why I do not evaluate autocollimator errors against known collimation tolerances but I evaluate them against an almost-perfect criteria.
Jason

#23 precaud

precaud

    Apollo

  • *****
  • Posts: 1397
  • Joined: 05 Dec 2012
  • Loc: north central New Mexico

Posted 16 February 2013 - 01:13 PM

Thanks for weighing in, Howie.

However, it is possible for a drawtube bore to be out of square with its face, and then tightening the clamp screw will force the accessory out of full contact with the drawtube face.

For most of the scopes out there, I think we can say "likely" instead of "possible"...

For this and other reasons I have advocated the use of the drawtube bore rather than the front flange face as the reference surface for focuser axis direction.

Makes complete sense to me.

But then we still have the problem of inconsistent registration of a smaller cylindrical accessory within a larger cylindrical drawtube due to the ease of slippage between the two clamped surfaces which are almost flat(in the sense of infinitesimals) at their line of contact.

Agreed again. And in addressing this problem, choosing a different surface to collimate to should not be part of the solution.

This is the problem that my Parallizer™ principle is intended to fix.
<marketing info snipped>
...as it could improve alignment tolerances, as well as help fund my retirement

Gotta respect a man with the courage to put his future security on the line behind his creativity! :bow:
If I'm unable to resolve this problem passively, I may end up having to make a contribution to your retirement... :)

(I have a patent on it :- )

Congrats and good luck with the patent. How long did it take them to process? I filed one two years ago this Friday, and still haven't heard a peep...

#24 Nils Olof Carlin

Nils Olof Carlin

    Surveyor 1

  • *****
  • Posts: 1580
  • Joined: 26 Jul 2004

Posted 16 February 2013 - 01:25 PM

Jason,

But if the collimation goal is to stay within tolerance then why even bother getting an autocollimator?



I don't have an answer to that one. :grin:

The reason I took an interest in autocollimators was curiosity - at the time, apparently no one had much of an idea what kinds of reflections (or even how many) there were, and what could be derived in terms of tolerances. Unless you knew, there was no way of knowing even whether "perfection" would meet reasonable tolerances!
It took me a long time to figure out the "anomalous" reflections (1 and 3), but this done, it was straightforward to see how reflections depend on respective axial errors (even to find the unexpected and AFAIK unknown fact that stacking of 3 and P only depends on the focuser axis!)

1 mm at 1250 mm f.l. corresponds to one inch at 100 ft. And to a tilt of
0.04 mm across a 2 in. focuser - much less than a human hair. What does common sense say?

But this is amateur astonomy, done for pleasure, and there is plenty of room for debate and different opinions.

Nils Olof

#25 Vic Menard

Vic Menard

    Aurora

  • *****
  • Posts: 4766
  • Joined: 21 Jul 2004
  • Loc: Bradenton, FL

Posted 16 February 2013 - 01:59 PM

...why even bother getting an autocollimator?


I don't have an answer to that one. :grin:

For most users, I think the autocollimator provides a means of improving the axial alignments achieved with their other tools. With practice, I've found that I can routinely achieve a "good" stack simply by being a little more careful with my laser alignment. There is a level of expertise involved, especially when the mechanicals are less cooperative (a condition common to many inexpensive Dobsonians).

With regards to the precision of collimation tools, I have to agree with Jason that properly executed tools should be transparent when making the more "critical" axial alignments. This transparency should extend such that the user's tools provide "agreement" instead of confusion. Finally, concerning tolerances, I believe they should be applied cumulatively to all of the mechanicals (throughout their respective motions), not just the collimating tools.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics