Grzybowski collimator

[Stefan Buda]

About 20 years ago I made a small alignment telescope with cross-hairs that can focus from a few inches to couple of metres. A few days ago I decided to turn it into a Grzybowski collimator, by adding the light sources and a 2" barrel to the front end.

Hopefully I will be able to use it to fix Andy01's TOA-130 problem.

[Atmos]

Looks a bit like one of those Hubble Star things.

[Stefan Buda]

No, this is an auto-collimator. It plugs into the eyepiece holder and looks at retro reflections from lens surfaces.

[Atmos]

Ah! So it’s kinda like a Tak collimating scope on steroids ;)

[Sunfish]

Very interesting device. I tried to make a collimating scope with an old lense but I could not sort out a light source. Is the placement of the led critical in relation to each other and the centre ?

[Stefan Buda]

The Tak one is not very useful for refactors. With this one you can see the retro reflection from each air/glass interface.



Yes, the placement of the light sources need to be accurate and also the optical/mechanical axis of the whole device.

[sharpiel]

Hello Stefan. I see you’re still at the forefront of optical-engineering excellence. Is there a thread link to Andy’s TOA problems? I have a TOA 130 also so want to be abreast of inherent issues...I haven’t been on here much since December 2019.

[Stefan Buda]

Hi Les,

Nothing is seriously wrong with Andy's TOA. The objective appears to have a bit of a tilt and that prevents the flattener/corrector to work properly.

The rest of the problems were not from Tak hardware. The field rotator could not be locked in a way that would guarantee squareness and the focuser had a bit of a tilt too. Focuser tilt can be adjusted but only in one plane, which is not enough to get it perfect.

Other than that, he inherited a bunch of adaptors/extenders with wedge errors.

[sharpiel]

Ok. It's always been a fear of mine that one day misadventure would require re collimation of my TOA. Would your "Grzybowski collimator" in the hands of a reasonably handy intelligent person enable self re collimation in situ?

[Stefan Buda]

This collimator is good for adjusting only the tilt of the lens groups, not the spacing. Andy's scope needs the whole objective to be adjusted, not the individual lens groups. At least that is what I hope, as I have not looked at it yet. My assumption is based on test images that Andy sent me.

If you need a full collimation, you must use the collimator in conjunction with Roddier testing that will tell you how much spherical aberration you have due to incorrect spacing. Not difficult but very time consuming unless you have a suitable optical bench where you can set up and artificial star.

[Andy01]

TOA now delivered to the master for what we all hope will be the final step in a long saga of rectifying tilt issues.
Thanks to Stefan for all of his care and attention done on this :thanx:

[Stefan Buda]

First look at Andy's TOA-130.

My estimate, based on defocused star images taken through a 3D printed mask, for the objective tilt, seems spot on. I was expecting about 3mm at the sensor.
However there is a bit of a problem because about one quarter of the error is coming from the focuser. As I rotate the focuser, the error varies from 50% to 100% at the best and worst orientations.
If I eliminate the objective tilt, the remaining focuser tilt might become insignificant though.
All four sets of reflections seem consistent so I'm happy that I won't need to touch the inter lens adjustments.
The attached image shows only three sets of reflections.

[Andy01]

Interesting! Seems that the collimator is showing the error up well - the spacing between reflections appears to be inconsistent.
Not sure what options there are regarding the focuser - wait & see I guess. :D

[Stonius]

You say it does the lens cell tilt and not individual element spacings, but I wonder if, given that you are seeing reflections from all lens surfaces, this would work for colimating the individual lens elements within a cell (thought not their spacings) so all lens elements are centered with respect to one another?


Cheers
Markus

[Sunfish]

There are probably a lot of Tak owners anxious to see the result. I know my focuser is off the centre of the lense cell by a millimetre. Could collimation help? Very interesting to find out. I see Grzybowski collimaters can be purchased for not too many hundred euros so I am surprised they are not more common considering their effectiveness.

[Stefan Buda]

You are misinterpreting the reflections - they are consistent but it is not possible to tell from the photo.



Yes, you are correct. The problem is that it is not possible to change the tilt of individual elements without affecting the spacing as well.


I'm not sure, but maybe those collimators come without the required cross-hair reticule eyepiece. If you need to supply your own, then it will add to the cost.

I will post the result of the collimation I did last night once I have the images prepared.

[Stefan Buda]

Ok, here's the result of about an hour of fiddling with the cell adjustment screws last night.

There is a mismatch of about 1mm between the mechanical and the optical axis. This, I think, is due to a number of errors such as focuser tilt, probably a bit of lens cell and rotator eccentricity, as well as cross-hair eccentricity.
The cross-hair reticule came from my junk collection and I never tested it.

Anyway I adjusted the the lens cell so that the position of the dots don't change relative to the cross-hair when I rotate the focuser. That should produce consistent images regardless of camera orientation. Otherwise if I cancel the offset in one particular position, I get twice as much error when change orientation by 180 degrees.

The two attached images were taken at 180 degree rotation of the focuser, with fixed, tripod mounted camera looking through the eyepiece of the collimator.

And yes, Andy, there is a slight misalignment of the two lens groups as well, but I would not worry about that, as it is very minimal

[Sunfish]

Good to know that the focuser offset does not get in the way of better collimation.

Many seem to think it is too hard to do without centre of the focuser-on the objective . So it seems the Tak designers have thought of that and minor focuser misalignments over time are inevitable and are not critical to good collimation. Good tolerances in the system.?

Would changing the focuser make a difference?

[Stefan Buda]

I'm not blaming the focuser for the offset. I think it has a contribution too, as perfection only exists in mathematics, not in engineering.

For visual use, without a field corrector or reducer, such an offset would not matter, but when used as an astrograph the optical axis must go through the center of the corrector group.

Anyway, I played a bit more with the collimator and rotator and came to the conclusion that most of the offset is coming from my collimator's telescope.
This little scope was put together from junk optics and the objective doublet must have a bit of wedge error. One day I'll make a good sighting scope for this, if I find a need for it.

[Stonius]

I'm sure you guys must have thought of this, but you couldnt use one of those focuser tilt rings you get for RC scopes, could you?


Markus

[Stefan Buda]

Yes it would work but it's not worth the trouble considering that we are talking about a tilt, coming from the focuser, of something like 15 microns across the sensor. I bet the sensor alignment in the camera is not much better than that.
Not a worry considering that the theoretical depth of focus is well over 40 microns at f/7.7

[Sunfish]

Thanks to your posts Stefan I managed to find an led illumination method for for my collimation telescope project . Managed to collimate my Tak. Not perfect but much better and i can do another round after a star test.

Better do you think or is there something I am missing?

Not too hard with only one lense cell i suppose. The LED does change the world.

After and before images along with my T2 lense collimator frankenscope and reticle eyepiece projection.

[Stefan Buda]

Great work. It definitely looks better. These collimators should be more common.

[Joanna]

thanks，i Managed to collimate my Tak. Not perfect but much better and i can do another round after a star test. the optical axis must go through the center of the corrector group.

[garymck]

Any chance you give details of how to make one of these collimators?
Cheers
Gary

[Stefan Buda]

Do you have access to a lathe and a milling machine?
If you do, I can give you some guidelines.

[tempestwizz]

I too would be interested in fabricating one of these devices. I have access to a lathe and crude milling equipment. I would like to test/confirm my efforts at collimation of my Tak TOA130 were effective. My scope is located in remote Laos, and the concept of shipping is daunting.
Unfortunately, at the moment I’m in Australia, waiting for the opportunity to return to wife and family in Laos, but on my return I would like to try this out.

[garymck]

Hi Stefan,
I have a Mill and a Lathe, as well as a cnc router.

cheers
Gary

[Stefan Buda]

Perfect, but I'm going through a bit of a hectic time. I'll get back with some guidelines in a few days.

[Tinderboxsky]

I have all the equipment too and am following this with interest.

[Sunfish]

If you have a 75mm focuser and a simple doublet or perhaps a triplet and can make a plate which screws into the focuser with a small hole in the centre with three equally spaced leds on the back as far apart as practical, that can work. That is what I used.

Will only work with a big 75 mm focuser other wise the reflected separation is too small to see without adding the collimating scope lenses as Tak and Grzybowski do. Stefan could make them easily and you could all buy one.

We could call it the Buda collimator.

I only added the t2 lense and projection reticle spacer in a 2inch eyepiece clamp to confirm the result by taking the photos.

[Sunfish]

Like this worked for me for an obvious correction on a doublet. Three LED on the back of a cheap 72mm to t2 adapter. Maybe 50mm would work. I used an adapted camping light but you can buy a three led disk online.

The spacing of the led needs to be at least 20mm centre to centre to give enough separation for naked eye adjustment viewing through a 5mm hole in the centre.

You can also use an eyepiece projection tube and two old stacked Plossl to get focus for a closer view.

[Stefan Buda]

Sorry for the delayed response. Among other things I had to deal with a HDD crash on my main PC. Slowly restoring all my apps now.

I'll let the photos do the explaining. The most difficult thing was getting the power cable to the LEDs go through the housing in a neat way.

[tempestwizz]

Hi Stefan, as you say pictures are worth lots of words. Combined with the pic from your first post on the topic, it makes sense to me now, and quite do-able.

Would I be correct in assuming that a wider spacing of the lens , ie larger diameter of the ring of leds, would improve resolution? At first glance,(and specifically for my TOA 130) I think I could fit the leds into a 2.7" extension tube Ive made already. That would allow me to space them a bit wider than the 2" at the focuser.
Also, I'm assuming the finer the LEDs, the better the resolution? Earlier, Ive used a fibre cable to work as an artificial star. This gives a star diameter of about 50 microns. Maybe i could insert glass fibres instead of LEDs and feed with a laser? I think TOSLink plastic fibre would not be much of an improvement over mini LEDs. Another also is that maybe the fibre idea would not be bright enough to reflect back of all the surfaces of a triplet?

Thanks for the information in any case. Brian