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

My take on GSO Ritchey-Chrétien collimation

  • Please log in to reply
106 replies to this topic

#76 MikeECha

MikeECha

    Messenger

  • *****
  • Posts: 405
  • Joined: 15 Sep 2018
  • Loc: Charlotte, NC

Posted 19 July 2021 - 08:54 PM

A lot to unpack there.  Hoping there's something there that clicks!

Yes. I know, I have seen these discussions many times but I have refrained from posting my experience because is based on a long problem solving process. I spent last year's fall and part of winter experimenting, checking and learning about this.



#77 MikeECha

MikeECha

    Messenger

  • *****
  • Posts: 405
  • Joined: 15 Sep 2018
  • Loc: Charlotte, NC

Posted 19 July 2021 - 09:47 PM

Yes, OP approves.  grin.gif   

 

@MikeECha -- @nebulachadnezzer is right, star quality over the whole field can't be determined from your AberrationInspector mosaic.  We need to see a full field of stars, and while you're at it do FWHMEccentricity on a focused field to get contours of FWHM and eccentricity.  That's an even more rigorous test, and will show imperfections that are not apparent by looking at the image.

Here is eccentricity. I am not sure how to read this or even the settings.

 

Eccentricity.JPG



#78 Rasfahan

Rasfahan

    Viking 1

  • -----
  • Posts: 878
  • Joined: 12 May 2020
  • Loc: Hessen, Germany

Posted 20 July 2021 - 01:05 AM

I don‘t know which camera you used, but even if it was the ASI294MC in „unbinned“ mode (best case) you have an FWHM of 4“ and only 2 stars are measured. There are too few stars visible to judge the field and the two stars that PI can measure are very bloated. The single corner star I see in the top left shows significant off-axis aberration (probably astigmatism). The method you describe means a full disassembly, a huge investment of time. My experiments with using the DSI method were far quicker with acceptable results.



#79 MikeECha

MikeECha

    Messenger

  • *****
  • Posts: 405
  • Joined: 15 Sep 2018
  • Loc: Charlotte, NC

Posted 20 July 2021 - 06:38 AM

I don‘t know which camera you used, but even if it was the ASI294MC in „unbinned“ mode (best case) you have an FWHM of 4“ and only 2 stars are measured. There are too few stars visible to judge the field and the two stars that PI can measure are very bloated. The single corner star I see in the top left shows significant off-axis aberration (probably astigmatism). The method you describe means a full disassembly, a huge investment of time. My experiments with using the DSI method were far quicker with acceptable results.

 

Rasfahan

 

There has been terrible seeing for a long time and It is the ASI183MC pro at 1365mm .37"/pix bin 1. The disassemble is only because of my conditions and was not different than when you clean the mirror which in fact I did as it was a second hand scope. If I have to collimate again I do not have to disassemble. Everything is now in place and I know they work.

 

The problem I see on many of these discussions is that everyone wants perfection with the easy button. It is not going to happen. As a design engineer I know how many hoops I jump thru to make things look easy to the end user.

 

I did the disassemble mainly to confirm my thoughts and to take the mystery out of the way. Now I know how the scopes are actually built and design features I did not understand before. Case in point: the use of the center screw, the "bad GSO focuser design that wobbles" that does not wobble and is actually not  that bad and,  the focuser attached to the back plate that is not a bad thing. In fact, I find it is a good thing if you find the way to use it to your advantage which I did.

 

I use this RC as learning prototype in case I want to buy a larger/expensive one.

 

Yes I see the elongation by zooming 1:1 but the rest of the field looks pretty good to me. I think I am at the point of diminishing return and trying to finetune this scope further with a coarse thread screw is a waste of time.

 

I also try to find an adapter from coarse to fine thread but, they do not exist with the thread pitch on the scope. That would have been sweet. 


Edited by MikeECha, 20 July 2021 - 06:38 AM.


#80 Rasfahan

Rasfahan

    Viking 1

  • -----
  • Posts: 878
  • Joined: 12 May 2020
  • Loc: Hessen, Germany

Posted 20 July 2021 - 09:37 AM

With the 183MC you have worse than 4“ FWHM. There is no way to correctly judge optical performance, because the seeing will blur all stars to circles. That you still have the amount of elongation shown in the corner of the small 183MC chip points to a serious misalignment in your scope or very bad sensor tilt - my 183MC has the latter problem. You should get far better corner performance from the scope.p I would love to believe a laser collimation of RC scopes was feasible but no manufacturer of premium RCs actually recommends to use one: The method by DSI is well described online, and CFF recommends preliminary alignment with mirror distance by backfocus measurement, then using the Tak scope and final refinement under the stars (the latter only for on-axis aberrations with the primary, so they are very sure of their secondary assembly). I have not found the ASA reference document so far, would be an interesting read, I think.

 

Edited to add that there is no comparison at all between the mechanical workmanship of my GSO RC8 and my CFF RC10. Because the mechanical parts of the latter are much better aligned, collimation is much easier. I need a CTU for my ASI2600MC, though, for optimal performance.


Edited by Rasfahan, 20 July 2021 - 09:42 AM.


#81 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 467
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 20 July 2021 - 10:11 AM

Here is eccentricity. I am not sure how to read this or even the settings.

The "star support" is only 2, meaning the script only found two useful stars.  The sub you're giving it only seems to have a dozen or so, but you need hundreds to get a meaningful result from FWHMEccentricity.  Start with a richer star field, and experiment with the "star detection sensitivity" and "upper limit" values till you see "star support" in the hundreds.  Then clicking "Support" will produce contours of FWHM and eccentricity.  The resulting "_stars" image will show you which stars the script found.  It's good to examine that a little to make sure it's evaluating good stars.



#82 MikeECha

MikeECha

    Messenger

  • *****
  • Posts: 405
  • Joined: 15 Sep 2018
  • Loc: Charlotte, NC

Posted 20 July 2021 - 10:17 AM

This thread is about methods that people use to collimate RCs and I was asked to share my experience.

I think the OP has an 8" version of GSO compatible with mine

Methods do not make miracles happen.

#83 MikeECha

MikeECha

    Messenger

  • *****
  • Posts: 405
  • Joined: 15 Sep 2018
  • Loc: Charlotte, NC

Posted 20 July 2021 - 12:55 PM

The "star support" is only 2, meaning the script only found two useful stars.  The sub you're giving it only seems to have a dozen or so, but you need hundreds to get a meaningful result from FWHMEccentricity.  Start with a richer star field, and experiment with the "star detection sensitivity" and "upper limit" values till you see "star support" in the hundreds.  Then clicking "Support" will produce contours of FWHM and eccentricity.  The resulting "_stars" image will show you which stars the script found.  It's good to examine that a little to make sure it's evaluating good stars.

TinySpeck thank you for the explanation I will check If I can find an image with more stars. The problem is I finished collimation late last year and since then I have been trying to image the Iris Nebula at 1365fl. With that small field of view there are not too many stars. But as soon as weather permits I will get some glob images with good star field. I can tell you I do have better out of focus stars on the corners than those in your image on the lower left side



#84 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 467
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 20 July 2021 - 01:06 PM

TinySpeck thank you for the explanation I will check If I can find an image with more stars. The problem is I finished collimation late last year and since then I have been trying to image the Iris Nebula at 1365fl. With that small field of view there are not too many stars. But as soon as weather permits I will get some glob images with good star field. I can tell you I do have better out of focus stars on the corners than those in your image on the lower left side

Yeah, it's tough to find a dense star field sometimes.  It should be above about 70 deg altitude, too, to avoid atmospheric dispersion (which can make a surprisingly nasty mess of your stars).  There's a cheat you can do for testing, if you want -- see https://www.cloudyni...ch-star-fields/ for a trick I just stumbled on recently.

 

Also beware that dense globular clusters like M13 won't give you good star coverage in the cluster.  The stars are not separable by FWHMEccentricity.  If you can find a non-globular cluster (or almost any image into the Milky Way) you might have better success.


Edited by TinySpeck, 20 July 2021 - 01:07 PM.


#85 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 301
  • Joined: 04 Jan 2021

Posted 20 July 2021 - 01:54 PM

The thread is about best methods to collimate RCs, specifically GSO-built RCs. To that end I'd say any discussion about collimation methods is welcome.

The Holy Grail would be a bench method that achieves (at least near-) perfect collimation of an RC. In my experience & evaluation, including all methods described in this thread, such a method does not yet exist. My experience includes collimating my own CF tube RC8 and CF truss RC10, as well as bench-collimating SCTs, Newts, and refractors, all of which are easy by comparison. The difficulty (or rather, need for accuracy) seems to increase with aperture, likely due to a combination of focal length and the larger size of the flat field. What may be good enough for an RC6 may leave visible astigmatism with an RC10.

OP's method is a hybrid, bench up to the point of primary alignment which takes place under the stars. If it works, that seems decent. I'm not 100% convinced on the laser/binocular method of secondary alignment (and I personally don't like combining binoculars and lasers). OP's method potentially eliminates half of the DSI method (which involves iterating between tweaking primary and secondary) so if it's accurate enough that does reduce the time required tweaking under the stars.

In my experience, secondary tilt on the GSO scopes is so extremely touchy that stars at (essentially) infinity focus are the only way to get there. The Tak scope isn't good enough. I have one. No laser is good enough. I have many. However, if OP's method gets you near to perfect secondary alignment then that can save a lot of time under the stars.

The vaunted DSI method is all stars (after setting focal length, primary position, and focus tilt if necessary -- all of which done at the bench the same way as OP's method). DSI is typically performed *beyond* infinity focus. Using DSI beginning from fairly bad collimation doesn't take a huge amount of time under the stars. I think it takes longer to get your mind around how the DSI method works than to use it.

If IFI ever ships SkyWave that will be an exciting (if somewhat expensive) alternative to squinting at diffraction patterns.


  • TinySpeck and MikeECha like this

#86 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 467
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 20 July 2021 - 07:24 PM

The Holy Grail would be a bench method that achieves (at least near-) perfect collimation of an RC. In my experience & evaluation, including all methods described in this thread, such a method does not yet exist. My experience includes collimating my own CF tube RC8 and CF truss RC10, as well as bench-collimating SCTs, Newts, and refractors, all of which are easy by comparison. The difficulty (or rather, need for accuracy) seems to increase with aperture, likely due to a combination of focal length and the larger size of the flat field. What may be good enough for an RC6 may leave visible astigmatism with an RC10.

OP's method is a hybrid, bench up to the point of primary alignment which takes place under the stars. If it works, that seems decent. I'm not 100% convinced on the laser/binocular method of secondary alignment (and I personally don't like combining binoculars and lasers). OP's method potentially eliminates half of the DSI method (which involves iterating between tweaking primary and secondary) so if it's accurate enough that does reduce the time required tweaking under the stars.
...

I totally agree -- that's my Holy Grail too.  I almost got there with the procedure outlined a-way at the start of this thread.  The single tweak with one centered star to fine-tune the primary is a lot easier than most star tests.  Kudos to Vixen for that.

 

I think I have something Grail-ish with my latest iteration, but I want to make sure it came out well enough.  I'm waiting for a clear night to snap some subs and do some careful analysis.


  • nebulachadnezzer likes this

#87 dg401

dg401

    Explorer 1

  • -----
  • Posts: 94
  • Joined: 11 Jul 2021

Posted 20 July 2021 - 11:41 PM

I hinted at it previously, but I'm a huge fan of a good tri-bahtinov mask for evaluating on axis coma and off axis astigmatism.

 

I will propose the following... if I'm wrong, that's ok, it's happened before, wasn't the first time, won't be the last...  if a tri-batinov mask shows focus in all three orientations both on and off axis, you have "good" or at least decent collimation.  No out of focus images needed, the mask will show us any coma/astigmatism via one or more out of focus mask axes.

 

Here's an example of the best on-axis coma performance I've managed to achieve recently on my RC6.  The mask is nearly perfectly focused in all three orientations:

 

vega Tri 2

 

Of course, as soon as I move the bright star (and who doesn't pick on Vega this time of year?) to the corners, the astigmatism is painfully obvious, and once I try to tweak out the astigmatism, I screw up the on-axis coma, well because... that's how it seems to go with these things...


Edited by dg401, 20 July 2021 - 11:42 PM.


#88 dg401

dg401

    Explorer 1

  • -----
  • Posts: 94
  • Joined: 11 Jul 2021

Posted 21 July 2021 - 12:14 AM

I'd like to get the OP's opinion on the primary/focuser alignment.

 

Assume we have a means to tilt the focuser in any directly and lock it there, maybe a tilt-plate, or a Moonlite CS, or even an off the shelf GSO focuser that just happens to line up perfectly with the optical center of the primary...

 

Why wouldn't our first... our very first step be to line up a laser dot from the focuser projected onto a white card with it's dimmer reflection off the primary mirror back onto the card?  This is perhaps the only collimation step we can do with almost absolute accuracy.  If the laser dot and it's reflection coincide, the primary and the focuser are aligned.  Period.  And since the primary and the focuser move in unison, they will remain aligned.  Anywhere the laser dot in the focuser points, so also points the primary mirror.  Anything the focuser is aligned with, so is the primary.

 

This seems like such an absolute no-brainer first step... or am I continuing my fine tradition of missing some salient point?


Edited by dg401, 21 July 2021 - 12:15 AM.


#89 MikeECha

MikeECha

    Messenger

  • *****
  • Posts: 405
  • Joined: 15 Sep 2018
  • Loc: Charlotte, NC

Posted 21 July 2021 - 08:36 AM

Yes, aligning focuser and primary is the base for a reference to collimate the secondary. That is a no brainer and it is the first step of my method. However, with reset I use I can repeatably get back to focuser aligned with primary without disassembly.

#90 nebulachadnezzer

nebulachadnezzer

    Ranger 4

  • -----
  • Posts: 301
  • Joined: 04 Jan 2021

Posted 21 July 2021 - 10:27 AM

It's worth noting that the truss models do not have the focuser attached to the primary mirror. On those models the rear flange attaches to the back of the cell and includes a focuser collimation capability at that point. I zeroed mine and confirmed with a laser that the beam hits the center spot on the secondary. For what that's worth...

Checking for focuser tilt using a laser is a terrible method but it's about all we have until we get out under the stars. Diode lasers aren't round, and they aren't necessarily mounted straight in the chassis of the tool. Even the HG diffraction grating can't be trusted. Don't believe me? You can buy my HG (I need to sell it -- I don't use it). At *least* try rotating your laser in your focuser before you conclude that it's perfect.

The old adage says, "It's a poor craftsman who blames their tools." However, I would also say, "It's a poor craftsman who fails to test their tools."

My point is that laser can get you "close" and to that extent it's useful if you are concerned that the tilt is *way* off. If you manage to get the secondary off-center in the OTA, good luck to you, but the laser should help you fix that too. I would argue, however, that neither the laser nor your eyes (even with binoculars) are good enough to judge the accuracy to perfection. Arguably it's also unlikely the spot on the secondary mirror is perfect on GSOs.

Both primary and secondary alignment eventually involve small tweaks like 1/16-1/20th of a turn to achieve near-perfect collimation. With a laser you won't see much change. Under the stars the change is evident in how far the stars move across your sensor. The difference in on-axis coma and off-axis astigmatism is apparent as well.

My PoV remains that all of these things eventually lead to final alignment under the stars using the DSI method. Bench methods may reduce the amount of time required for star alignment, and that's a good thing, but I have to caution against methods requiring disassembly of the scope unless there's strong evidence that particular scope is badly assembled.

Again, I have bench-collimated everything but an RC with great success (except a CDK which I've never owned). The mirror alignment of RCs is the most critical I've encountered.



#91 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 467
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 21 July 2021 - 10:37 AM

I'd like to get the OP's opinion on the primary/focuser alignment.

 

Assume we have a means to tilt the focuser in any directly and lock it there, maybe a tilt-plate, or a Moonlite CS, or even an off the shelf GSO focuser that just happens to line up perfectly with the optical center of the primary...

 

Why wouldn't our first... our very first step be to line up a laser dot from the focuser projected onto a white card with it's dimmer reflection off the primary mirror back onto the card?  This is perhaps the only collimation step we can do with almost absolute accuracy.  If the laser dot and it's reflection coincide, the primary and the focuser are aligned.  Period.  And since the primary and the focuser move in unison, they will remain aligned.  Anywhere the laser dot in the focuser points, so also points the primary mirror.  Anything the focuser is aligned with, so is the primary.

 

This seems like such an absolute no-brainer first step... or am I continuing my fine tradition of missing some salient point?

Would you place the card inside the secondary and view it using an inspection mirror?  The angle of the card would affect the shape of the pattern, but maybe you could use that to your advantage to "elongate" it in different axes.  I'm with @nebulachadnezzer on the fuzziness of lasers in general though.  It would be difficult to align the laser source blob with the reflected blob (probably of a different shape).

 

Also, I'm not sure that you want to use focuser tilt to align the focuser and primary.  I think you need to adjust primary and secondary tilt to get their axes lined up and then focuser tilt to align that axis with the OTA tube.  But I'm not positive about that...



#92 dg401

dg401

    Explorer 1

  • -----
  • Posts: 94
  • Joined: 11 Jul 2021

Posted 22 July 2021 - 01:55 AM

Would you place the card inside the secondary and view it using an inspection mirror?  The angle of the card would affect the shape of the pattern, but maybe you could use that to your advantage to "elongate" it in different axes.  I'm with @nebulachadnezzer on the fuzziness of lasers in general though.  It would be difficult to align the laser source blob with the reflected blob (probably of a different shape).

 

Also, I'm not sure that you want to use focuser tilt to align the focuser and primary.  I think you need to adjust primary and secondary tilt to get their axes lined up and then focuser tilt to align that axis with the OTA tube.  But I'm not positive about that...

This alignment would be done with the primary cell/focuser removed from the OTA.  No secondary to get in the way.



#93 Stonius

Stonius

    Vostok 1

  • *****
  • Posts: 118
  • Joined: 28 May 2016

Posted 24 July 2021 - 08:29 AM

Is there any reason to assume that the centre spot on the secondary actually represents the optical centre of the mirror and not the physical centre? I'm thinking there's no guarantee they're the same at all.

 

Markus


  • TinySpeck likes this

#94 jmiller1001

jmiller1001

    Messenger

  • *****
  • Posts: 421
  • Joined: 21 Mar 2016
  • Loc: Santa Cruz and Carmel, CA

Posted 24 July 2021 - 10:10 AM

@Stonius - I would not assume that the mechanical and optical are the same.  Case in point: My (former) Orion/Synta RC10 Truss scope: the two were not co-incident.  I aligned everything mechanically, and under the stars, the scope showed slight mis-collimation - I had to dial everything in with a star test.

 

My (current) CFF250 RC Scope - the two are the same.  And, this is not even a guarantee with high-end scopes - as I've heard a number of stories where the two centers are not aligned for some very high end scopes.  YMMV.



#95 Paul Garais

Paul Garais

    Viking 1

  • -----
  • Posts: 584
  • Joined: 07 Oct 2018
  • Loc: Sendenhorst, Germany

Posted 24 July 2021 - 10:26 AM

@Stonius - I would not assume that the mechanical and optical are the same. Case in point: My (former) Orion/Synta RC10 Truss scope: the two were not co-incident. I aligned everything mechanically, and under the stars, the scope showed slight mis-collimation - I had to dial everything in with a star test.

My (current) CFF250 RC Scope - the two are the same. And, this is not even a guarantee with high-end scopes - as I've heard a number of stories where the two centers are not aligned for some very high end scopes. YMMV.

Are you happy with your CFF? Is the collimation stable and possible to make fine adjustments without fiddling around again because it changes after fixing the bolts?
And most important: Is it stable throughout the sky or is the mechanic not stiff enough for heavy imaging trains?

#96 jmiller1001

jmiller1001

    Messenger

  • *****
  • Posts: 421
  • Joined: 21 Mar 2016
  • Loc: Santa Cruz and Carmel, CA

Posted 24 July 2021 - 11:39 AM

@Paul Garius - I love my CFF250.  Ironically, the optics in my Orion RC were very good, and mechanically, it was sound.  I had an opportunity to pick up one of the last CFF250s and I jumped at the chance - my CFF92 and CFF105 refractors are great, and Catalin Fus is great to work with.

 

The CFF250 shipped from Poland to California collimated and arrived collimated.  So far, I have not had to touch the collimation, and it holds regardless of orientation.  The mechanics of the Orion/Synta are very good for the price; the CFF is a few steps higher in quality (and is lighter).

 

Optically, the CFF is first class and mechanically, it's built like a tank - it can easily support a full imaging set-up.


Edited by jmiller1001, 24 July 2021 - 11:41 AM.

  • Paul Garais likes this

#97 Paul Garais

Paul Garais

    Viking 1

  • -----
  • Posts: 584
  • Joined: 07 Oct 2018
  • Loc: Sendenhorst, Germany

Posted 24 July 2021 - 12:14 PM

@Paul Garius - I love my CFF250. Ironically, the optics in my Orion RC were very good, and mechanically, it was sound. I had an opportunity to pick up one of the last CFF250s and I jumped at the chance - my CFF92 and CFF105 refractors are great, and Catalin Fus is great to work with.

The CFF250 shipped from Poland to California collimated and arrived collimated. So far, I have not had to touch the collimation, and it holds regardless of orientation. The mechanics of the Orion/Synta are very good for the price; the CFF is a few steps higher in quality (and is lighter).

Optically, the CFF is first class and mechanically, it's built like a tank - it can easily support a full imaging set-up.

Thanks a lot. They seem still to have some in stock with GSO optics. Sounds like a good deal.

#98 TinySpeck

TinySpeck

    Messenger

  • ****-
  • topic starter
  • Posts: 467
  • Joined: 08 Oct 2017
  • Loc: Seattle area

Posted 24 July 2021 - 12:46 PM

Is there any reason to assume that the centre spot on the secondary actually represents the optical centre of the mirror and not the physical centre? I'm thinking there's no guarantee they're the same at all.

 

Markus

If they don't coincide, it throws the idea of "perfect" bench collimation using the secondary donut into a cocked hat.  (Does anyone use that expression any more?)  In that case you would have to rely on optical collimation, like with a laser, which has other problems.  Maybe a star test is unavoidable if you really want the best collimation.  Some people swear by the DSI technique, but I can never find a rich enough star field at a high enough altitude for it.  As I recall there isn't an adjustment for focuser tilt either, which seems like you're leaving out a useful control point.


Edited by TinySpeck, 24 July 2021 - 12:50 PM.

  • MikeECha likes this

#99 jmiller1001

jmiller1001

    Messenger

  • *****
  • Posts: 421
  • Joined: 21 Mar 2016
  • Loc: Santa Cruz and Carmel, CA

Posted 24 July 2021 - 02:08 PM

@TinySpeck - I think that a star test is inevitable and necessary to assure "perfect" collimation.  Fortunately, I've never had to touch the primary adjustment of an RC - just the focuser tilt and the secondary.



#100 jmiller1001

jmiller1001

    Messenger

  • *****
  • Posts: 421
  • Joined: 21 Mar 2016
  • Loc: Santa Cruz and Carmel, CA

Posted 24 July 2021 - 02:09 PM

@Paul Garius - I got mine with GSO Optics - they are fine - on both my CFF and Orion RC's, they were very good.


  • Paul Garais likes this


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.







Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics