64 replies to this topic

[Cedric_r]

Hi all,

 

I have recently purchased an RC 14" (Omegon; same model as the GSOs, TPOs, etc around). Coming from a Newton 12", the learning curve for collimation is brutal. I have tried to follow the DSI method, which is easy to understand and apply, but I'm hitting a problem that I can't trace. I'm hoping that someone will know what that is.

 

I can defocus a star and align primary, then move to a star field and align secondary as the method describes. I can have pretty good results:

http://raguenaud.spa...s/day2-star.jpg

 

The problem is that there is a clear distortion in the field that makes all that work useless because it's not on good bases. Here is a sequence going from outside focus to focus to inside focus. See how the stars are distorted one way before focus, then another after ocus.

http://raguenaud.spa...tion-591671.jpg

 

http://raguenaud.spa...tion-486697.jpg

 

http://raguenaud.spa...tion-429438.jpg

 

http://raguenaud.spa...tion-378542.jpg

 

http://raguenaud.spa...tion-251301.jpg

 

 

Before I attempted DSI, I did a laser collimation (mostly following a method I used to apply to my Newtons):

First I adjust the focuser onto the secondary by using a laser (collimated best I could on a wall 6m away) and rotating it to make sure its dot in the secondary wobbled a bit, that's inevitable when you rotate by hand, but never enough to leave the center of the secondary (what is marked as center, obviously), or enough to describe a circle, by adjusting the focuser screws. I considered that this told be the focuser was in line with the optical line, assuming the secondary is centered. Then I center the reflection of the concentric rings onto the primary from the secondary (takes time; possibly readjust the focuser centering if you made big changes and start again a few times) by adjusting the secondary's screws. Then I center the concentric rings reflected onto the wall by the primary so that the secondary reflection (very faint light only visible in the dark) is centered with the rings from the laser and the rings are concentric and equally spaced.

http://raguenaud.spa...-on-primary.jpg

 

 

I have done the whole thing every night from scratch (laser collimation to star test) for the last 4 nights and I'm not getting closer because I can't figure out what causes the star distortions.

 

Does anyone have an idea?

 

Thanks.

Edited by Cedric_r, 10 March 2021 - 03:26 AM.

[Supernova74]

I’m no expert in this field however I will take a pun on it.and I’m sure other members will correct me if I’m wrong,however we all have to learn even if it’s by our mistakes.have you tested the scopes optics via just useing a eyepiece firstly!? How is the star images in and out of focus!? I’m also assuming the telescope has been cooled down appropriately.anyway from the best of my knowledge it looks like some form of feild curvature,have you introduced a feild flatner when useing camera

[petercoxphoto]

Hi Cedric -

I also recently bought a GSO RC12 and while it arrived with reasonable collimation, it does have some off-axis problems. Poor weather have prevented me from trying to collimate it yet, but I plan to use the DSI method also - so I'll be following this thread with interest.

 

Just a few points - you don't tell us what part of the frame the problem images come from, and this is critical to figuring out the error. 

 

I would suggest producing an image showing a matrix of 9 panels showing the edges of the field and the center as shown in the DSI PDF. This can be done very easily in NINA, using their aberration inspector in the imaging tab. I'm sure there are other ways to do it, but it's pretty painless with NINA, even if you don't normally use it for imaging.

 

I also note that the DSI method doesn't talk about using lasers at all - you purely collimate based on the star test. It would seem wise to follow that specifically rather than mixing methods.

 

As I said, I have no experience collimating RCs (like you, I also have an imaging Newtonian that I regularly collimate), but I've studied the DSI guide pretty closely and while I'm approaching the collimation question with some trepidation, I think patient, methodical and precise following of the guide will yield good results.

 

I suppose one other thing to bear in mind is your local seeing - if your seeing is very bad on a given night, you might not be able to get any usable results as the star shapes may be too distorted. Working closer to the zenith would help there, as well as using an IR pass filter.

 

Cheers,
Peter

[Cedric_r]

Hi SN74,

 

I'll try to answer your questions:

 

I haven't tested with an eyepiece, I don't have any. I've never done any visual astronomy.

 

The telescope was colled down over 1 hour before starting that the effect is consistent through the night.

 

I have no flattener in the optic line. I thought the RC should cover the camera sensor without distortion. But there indeed seems to be a large amount of curvature that shouldn't be there.

 

Thanks.

[petercoxphoto]

Cedric -

Regarding curvature, that RC should cover an APS-C sensor without any visible curvature. You should only need a flattener for a full-frame sensor or larger. Any apparent curvature you're seeing would be collimation-related.

 

Cheers,
Peter

[Supernova74]

Ok wouldn’t it make more sense firstly to possibly test with an optical eyepiece would that rule out any collimation issues and able to attempt an in out of focus star test.then you could rule out any problems when a camera is introduced in the optical train and then troubleshoot from there if it’s necessary for a feild flatner! Go from there

[Cedric_r]

Hi Peter,

 

There are fairly few stars in the field, so creating a panel of the corners doesn't show much. The star distortion affects the whole field, with minor variations in the corners. The images are uncropped.

http://raguenaud.spa..._266_mosaic.jpg

 

Note: I can't use NINA, it doesn't know SBIG cameras and SBIG has never produced ASCOM drivers for its cameras.

 

Yes, the DSI method doesn't mention lasers, but I wanted to start with a semi collimated state. It shouldn't matter anyway, just more adjustments if you start from a really bad point.

 

What really bothers me is these stars pointing to the right outside focus, then xwitching to pointing left inside focus. I can't figure out what configuration would do that across the whole field. If the center was nice and round, I'd think curvature, but it's more like a plane crossing the sensor line, if you see what I mean. I've been thinking about it for 4 days and I can't think of a good explanation.

 

Thanks.

Edited by Cedric_r, 10 March 2021 - 05:49 AM.

[Cedric_r]

Cedric -

Regarding curvature, that RC should cover an APS-C sensor without any visible curvature. You should only need a flattener for a full-frame sensor or larger. Any apparent curvature you're seeing would be collimation-related.

 

Cheers,
Peter

Agreed, it was my impression.

[petercoxphoto]

Hi Peter,

 

There are fairly few stars in the field, so creating a panel of the corners doesn't show much. The star distortion affects the whole field, with minor variations in the corners. The images are uncropped.

http://raguenaud.spa..._266_mosaic.jpg

 

Note: I can't use NINA, it doesn't know SBIG cameras and SBIG has never produced ASCOM drivers for its cameras.

 

Yes, the DSI method doesn't mention lasers, but I wanted to start with a semi collimated state. It shouldn't matter anyway, just more adjustments if you start from a really bad point.

 

What really bothers me is these stars pointing to the right outside focus, then xwitching to pointing left inside focus. I can't figure out what configuration would do that across the whole field. If the center was nice and round, I'd think curvature, but it's more like a plane crossing the sensor line, if you see what I mean. I've been thinking about it for 4 days and I can't think of a good explanation.

 

Thanks.

 

Hi Cedric -

You did the mosaic with PI, I see - even better. You can get more stars in the field by increasing the size of each mosaic panel, say to 512 pixels. Or downsampling your data. Can you post a full frame?

 

One thing I know about the RC design is that they are very sensitive to primary->secondary distance and the focal length as a result. If your efforts have changed that distance, that can have an effect as well. What is the published focal length of your scope, and what does PI's ImageSolver script say?

 

Cheers,
Peter

[Epox75]

I'm no expert but could pinched optics possibly be a reason for such shaped stars?

[Cedric_r]

I'm no expert but could pinched optics possibly be a reason for such shaped stars?

 

Hi there,

 

I thought of that, but I was careful not to over-tighten the primary. Also it doesn't explain the changing stars.

 

Thanks.

[Cedric_r]

Hi Cedric -

You did the mosaic with PI, I see - even better. You can get more stars in the field by increasing the size of each mosaic panel, say to 512 pixels. Or downsampling your data. Can you post a full frame?

 

One thing I know about the RC design is that they are very sensitive to primary->secondary distance and the focal length as a result. If your efforts have changed that distance, that can have an effect as well. What is the published focal length of your scope, and what does PI's ImageSolver script say?

 

Cheers,
Peter

Yes. My first task was to get to (close) the nominal FL. I bought the telescope second hand and the previous owner had fiddled with it a lot. The FL was way off. I've brought it back to <0.5% away from nominal before I started collimating: 2833mm (varying by a mm or 2 depending on secondary adjustments) instead of 2845mm.

[Cedric_r]

Peter, here is a different series of images (full frame) from the previous night that shows a lot more stars (M67). It also goes from outside focus to focus (4th image) to inside focus.

http://raguenaud.space/images/foc1.jpg

http://raguenaud.space/images/foc2.jpg

http://raguenaud.space/images/foc3.jpg

http://raguenaud.space/images/foc4.jpg

http://raguenaud.space/images/foc5.jpg

http://raguenaud.space/images/foc6.jpg

http://raguenaud.space/images/foc7.jpg

[petercoxphoto]

Hi Cedric -

That does make the problem much clearer. I suppose the bright side is that the in-focus frame is not nearly as bad as the out of focus images would suggest. I'm at a loss as to the scale of the problem. Can you get the on-axis donuts perfectly round by adjusting the primary via the DSI method at all?

 

Did you take care to orient the camera using an aperture mask as they suggest, so you know you're adjusting the right screws?

 

Cheers,
Peter 

[Cedric_r]

Hi Cedric -

That does make the problem much clearer. I suppose the bright side is that the in-focus frame is not nearly as bad as the out of focus images would suggest. I'm at a loss as to the scale of the problem. Can you get the on-axis donuts perfectly round by adjusting the primary via the DSI method at all?

 

Did you take care to orient the camera using an aperture mask as they suggest, so you know you're adjusting the right screws?

 

Cheers,
Peter 

I didn't manage to get a very round star when inside focus like DSI says last night. But I did the night before.

http://raguenaud.spa...s/day2-star.jpg

 

The problem is that I can make the central star round, but that's in the context of the pointy stars when inside focus. So when I go back to focus it's no longer round.

 

The changing stars makes the autofocus really hard because the star size isn't linear. I use SGP and it struggles, having to restart several times to get focus (sometimes not very well).

[petercoxphoto]

Hi Cedric -

So one point is that DSI says to do collimation outside focus, not inside.   I wonder if that might be part of the problem?

 

Adjusting the Primary Mirror to Remove On‐Axis Coma

There are several techniques for collimating out on‐axis coma with the primary mirror.  Any method you
are comfortable with is fine.  The method we use is to look at relative image intensities of on‐axis,
outside‐of‐focus star images.  Coma is present when star illumination is not uniform.

Focuser Positions and Directions

Collimation adjustments are based on out‐of‐focus images taken outside‐of‐focus.  This is illustrated in
the last image shown below.  To take outside‐of‐focus images, you need to move the focal plane IN from
a focused position.  This means the secondary mirror should be moved IN towards the secondary
focuser and away from the primary mirror.  The IN button on the focuser controller does this.  Outside‐
of‐focus images are typically taken 50 counts from a focused position for an RC14C.  Scopes using an
external focuser would move the CCD image sensor away from the scope.

 

As this procedure is written in the context of secondary focusers, it's a little confusing, but the last sentence in the above quote (my bold text) clarifies it for our GSO scopes.

 

Cheers,
Peter

Edited by petercoxphoto, 10 March 2021 - 06:32 AM.

[Cedric_r]

My first thought was that the focuser wasn't aligned with the 2 mirrors and was creating an oblique plane that explains the stars. But when I put a laser in the focuser and rotate, it doesn't move, and when I move the focuser in/out, the laser dot doesn't really move either. So I assumed that proved alignment with the secondary.

[Cedric_r]

Hi Cedric -

So one point is that DSI says to do collimation outside focus, not inside.   I wonder if that might be part of the problem?

 

Cheers,
Peter

I tried that too but it made no difference

[petercoxphoto]

Hmm. The only other thing I can think of is that the secondary is just grossly misaligned. But if it was that bad, you'd expect the in-focus images to be much worse, I'd have thought.

 

At this point I think we need a real expert on this stuff to weigh in as I'm firmly out of my depth now. 

 

Cheers,
Peter

Edited by petercoxphoto, 10 March 2021 - 06:37 AM.

[Cedric_r]

My first thought was that the focuser wasn't aligned with the 2 mirrors and was creating an oblique plane that explains the stars. But when I put a laser in the focuser and rotate, it doesn't move, and when I move the focuser in/out, the laser dot doesn't really move either. So I assumed that proved alignment with the secondary.

Of course there is a good chance that the center of the secondary isn't the optical center. How do you align the focuser with the optic train when you can't use the secondary for that?

[petercoxphoto]

This is something I'm wondering as well. I'm not sure how the focuser axis is handled as part of the DSI method. I'd be very interested to hear how that should be handled. 

[petercoxphoto]

Ah, I thought I'd seen a thread with a similar problem. Take a look at this - the OP here had exactly the same problem you have: https://www.cloudyni...tchey-chretien/

[Supernova74]

Just a matter of interest really is this common practice for a brand new scope!?still a shame I think you haven’t got an eyepiece to use as just feel this could rule out some possibilities!? Especially the collimation,however seeing in the photos you supplied does not look to far out.so this only leaves 2 other possibilities feild curvature,pinched optics highly unlikely as I cannot see any spike patterns.

[petercoxphoto]

Also these threads:

 

https://www.cloudyni...chretien-scope/

https://www.cloudyni...3#entry10293755

[Jared]

What you are seeing is some combination of on-axis coma and astigmatism. The astigmatism is what makes the elliptical shape flip depending upon which side of focus you are on.

1) If the focuser were not flat it would not affect stars near the center of the field. It would cause stars to have different amounts of defocus in different ent corners. This isn’t your issue.

2) Because both mirrors are hyperbolic you can’t just adjust one mirror to fix on-axis coma, then the other to fix corners. It’s an iterative process, unfortunately.

3) Lasers are fine as a starting point, but the results with the camera are ultimately all that matters.

4) Here is what I did with mine in the past... Correct coma in the center of the field using the primary. Put the Poissant spot in the center of the star, even if the star isn’t quite round, using primary mirror adjustments. Then use the secondary mirror adjustments to make sure the shape (and direction) of elongation is the same in all four corners.

5) Making changes to the secondary to fix the corners will throw off the primary—on-axis coma. You will likely need to repeat step for a couple times, but it should converge on a solution.

That should take care of collimation, ignoring focuser tilt. It won’t fix astigmatism, though. It looks like you have some. Hopefully not much, and that lack of collimation is making the astigmatism appear worse than it is.

Astigmatism is specifically the problem where stars across the entire field are concentric but elliptical with the orientation of the ellipse flipping by 90 degrees on either side of focus. It’s not a collimation issue. If you do have much astigmatism, you won’t be able to collimate it out. The cause lies elsewhere. It could be an optical defect. It could be cooling related—a 14” mirror holds a lot of heat and it could take well over an hour to stabilize, especially with falling temps. It could be in the camera rather than the scope, also. You’ll want to confirm by rotating the camera to see whether the direction appears to rotate. If it appears to stay the same—horizontal or vertical or diagonal—when you rotate the camera, then it’s in the camera not the scope. If it changes orientation when you rotate the camera 90 degrees, it’s in the telescope.

After you are sure collimation is correct—I don’t think you are there—start to eliminate any other possible causes for astigmatism. Check the camera rotation as I described. Let the scope cool fully. This could take a while, and on some nights it might not ever happen fully. Big mirrors are finicky, even if they are made of substrates that should be thermally stable. If you still have an issue, it may be an optical defect.