Robin
I've just gone through the process of learning how to collimate a 200mm f/4 and now 300mm f/4 Newt for AP.
You have had plenty of good advice here already. There is also lots of info on the internets but the majority applies to collimation for visual, which is quite a different process, and has different objective.
Here are two of my learnings:
1. Analyse the collimation by looking at the shape of the stars. It sounds obvious, but it's the name of the game. You might get lucky and the collimation with the laser or cheshire is still ok when you put your camera in, but this would be good luck. Tools will get you in the ballpark after you rip the scope apart and put it back together. I find that I can only get it dead nuts by pixel peeping. When you do nail it, it will 'snap' like magic and then you'll be afraid to touch it!
2. Make sure nothing moves. That includes the focuser and primary. If you're in the process of collimating, and something moves without you realising, you'll be going in circles unknowingly.
If you have a robofocus, make sure it holds focus all over the sky and through the night. Holding focus is critical - good focus makes good collimation look better. If you're slightly out of focus, it will really accentuate all collimation issues. Initially I was loosing focus **during** a 10 min exposure. This weekend I took the same drastic action I did with my 200mm. There could be an earthquake directly underneath my observatory and the primary mirror is not going to move one micron. Tonight was the first proper test run - so far it looks perfect.
James, could I ask for more detail about how you're ensuring focus, what drastic action etc? To do with primary, focuser, or both?
I need to store my scope horizontally so I can close the observatory roof. I was finding that the mirror was tilting forward during the day and then settling back down into the cell as the night went on (and also jumping around at times). I knew it wasn't thermal contraction because it was going in the wrong direction. The cell had small pads on the back that can expand and contract. Also only one of the 3 points on each triangle was stuck to the mirror, the other 2 were free to separate from the mirror, allowing the mirror to fall forwards. To make it worse, the mirror clips on the side are angled forwards too, so when the scope is horizontal, the mirror really tilts forward a lot. The solution was to glue the mirror to each triangles with 3 blobs of silicone. Simple and easy. I also "tightened" the mirror clips, so they just start to grab when I slide a piece of paper between the clips and the mirror. Since this a standard GSO newt, I suspect the all have the problem.
To my amazement, I found the tension screw on the bottom of the focuser works loose every night. I couldn't believe it was true (or possible) so I drew a little arrow on it, and sure enough, in the morning I could see that it had unwound itself about 1/4 of a turn. I still can't figure out the mechanics of how that is possible, but it happens. So I used some sticky tape to hold it in place.
These are both simple things really, but it takes time to sort them out one by one.
I wrote a little program to extract data from my focusmax logs so I could see what is happening using excel. The attached graph shows the before and after of how the scope holds focus. Yes there is still room for improvement, but at least I have a chance now.
I'm pretty sure everything is on axis now Rob, just needs a little more time but looking at the forecast it won't be this week.
That's the result I'm looking for James. No robofocus on this one, Moonlight focuser on a prostar tube and the tube has been strengthened around the focuser. You have a keller corrector on that tube don't you?
That's the result I'm looking for James. No robofocus on this one, Moonlight focuser on a prostar tube and the tube has been strengthened around the focuser. You have a keller corrector on that tube don't you?
Troy, have a look here for some info.
Yes, it's the 3 inch corrector from ASA - I can't fault it.
Do you have a motor on the Moonlight? If so, it would be interesting to see how well the scope holds focus. If you can confirm nothing is moving, then it's just a matter of turning some screw one way or the other to get it collimated.
I just realised there is a picture of the offending tension screw on my site (before I drew the arrow) pic
There is also lots of info on the internets but the majority applies to collimation for visual, which is quite a different process, and has different objective.
Hi James,
Why do things become different for imaging and visual purposes?
Collimation of a newtonian reflector focuses solely on aligning the optical axis of the telescope. If you wish to take collimation to the ultimate level, it also involves making the mechanical axis of the telescope coincidental with the optical axis. Making the optical and mechanical axes of the telescope coincidental does not affect optical performance in any way. It affects the pointing accuracy of an DSC system employed and the ongoing tracking accuracy of the telescope only.
Quote:
Originally Posted by Moon
You might get lucky and the collimation with the laser or cheshire is still ok when you put your camera in, but this would be good luck.
A laser in itself is not adequate for collimating a telescope properly. A laser is only suitable for adjusting the "tilt" of the primary mirror, after the secondary mirror has been correctly positioned in the telescope and its axial rotation properly adjusted.
Quote:
Originally Posted by Moon
Tools will get you in the ballpark after you rip the scope apart and put it back together.
This is not correct. Someone who knows how to collimate a telescope properly can collimate a telescope with a very high degree of accuracy using high quality collimation tools like the Catseye or Tectron Tools. Both of those collimation sets include an "Autocollimator" which is very accurate indeed. The only time these tools will not be 100% accurate is when the mechanical centre of the primary mirror (where the dot is) is not in fact coincidental with the optical centre of the primary mirror. Due to modern mirror production techniques this is a pretty rare occurence these days.
This is not correct. Someone who knows how to collimate a telescope properly can collimate a telescope with a very high degree of accuracy using high quality collimation tools like the Catseye or Tectron Tools. Both of those collimation sets include an "Autocollimator" which is very accurate indeed. The only time these tools will not be 100% accurate is when the mechanical centre of the primary mirror (where the dot is) is not in fact coincidental with the optical centre of the primary mirror. Due to modern mirror production techniques this is a pretty rare occurence these days.
It depends on your definition of collimated. My definition is the stars on a full frame chip are symmetrical - and hopefully round. The view I see with my Catseye is not the endgame. It's the second last step. Final tweaking is still required after the Catseye is removed and the camera is inserted.
James
Im with John on this one James, the right tools in the right hands will get it just about perfect, you cannot not be right when your looking at 3rd or 4th reflections eg correct me if im wrong the image has gone from your eye piece down the tube and back again 3 times now on my 1200FL F4.7 newt thats nearly 7 meters do some quick trig on that and you will find you are in the micron region of accuracy. I believe that auto collimators are readily able to get within 10 micron of optical alignment (assuming your reference spot and the like are right).
Pixel peeping can be hit and miss unless your seeing is BRILLIANT eg your stars are virtually still and even so can still hide optical collimation as if your seeing is better than 1 arc second (which it generally is not) then depending on your camera/FL set up your pixel is actually under the seeing limit! My F4.7 + qhy9 mono produces a 0.93 arc second per pixel so you see its kinda close ish. Then you have to contend with your mounts guiding/tracking and hopefully you are 100% guiding on a pin point if it wobbles your star has a whoopsy in it are you able to detect the difference in a guide error vs optical error from looking at a pixel?
I know its getting nit picky and all theoretical but at the end of the day if you take a image especially with newts you will see the ones who have nailed their collimation look at the diffraction spikes they generally have clearly seperated rainbows in them. I know that when i get my collimation right they are very distinct when i am out.... they are just whiteish.
Don't get me wrong, I really like my autocollimator thingie and I'm not saying you don't need one - you do. But we are not replacing it with an eyepiece here - it's being replaced with a camera. Lucky for me, my Catseye has two holes, so I can also view the alignment off centre and this helps a lot. But remember the sensor might not be square, the weight alone is probably enough to push the scope out of collimation - as per the video on my website.
It's also a practical matter - for the same reason lots of people use the CCD to do the polar alignment rather than swap the camera out with a reticle and then swap back to the camera, it also sometimes easier to use the CCD to do the the final collimation. Removing the camera to insert the autocollimator and getting it back in the perfect position is not always easy.
I'm not talking about the endless theory of collimation here which we all know back-to-front and no-one can despute that. It's the practical side that we all have to come to grips with and find a way that works for our own setup. For me, its easiest to use the CCD Sensor as the reference plane, and work back from that.
Lets say you collimate with a catseye and get it all perfectly stacked. Then you take an image and notice that the stars are wonky on the right hand side only. What do you do? Put the catseye back in and start over? Maybe. The alternative it to center a star and push it left or right using the collimation adjustments and see if it fixes it. That's what I would do.
James
I guess you are right with the focuser thats why I countered that with utilizing screw down fittings all the way down to the focuser and a FT 3" focuser because i knew that the standard Skywatcher 2" sucked big ones and with the addition of a 3" OAG the QHY9 mono plus associated filters in the filter wheel there is not a hope in hell the Skywatcher can hold onto that with any reasonable accuracy.
Secondly if your moving your collimation that much from when your autocollimator out to when you put your camera in start fixing! When i do my collimation i have a reference screw which i always screw down first and then in a clockwise fasion tighten down the remaining two this makes sure that the camera is sitting virtually in the same spot as the auto collimator and hence my collimation routine is still valid.
The other thing that you have to remember that its well and good to collimate to your focuser if you have a stationary focuser but if you want to frame your object to take full advantage of your sensor you are screwed big time because you will have to not only have to do a focus routine on every object you will have to collimate on every object which is not really feasable and in the case of my FT it has a 360 degree rotatable focuser which holds focus. So at the end of the day you want to get your collimation perfect so to speak that way you have the best flexibility.
You can see from my images that i have round stars across the FOV admittedly this isn't a 11k like yours but then again if im getting pin point stars to the edge of FOV now there will be no difference if i increase the sensor apart from if there is any residual error it will be more evident but still under the seeing celing.
In relation to the tilt of the ccd sensor, if it has tilt then utilizing the cats eye gear will give you a big piece of mind and a base line that the imaging plane isn't tilted. this will then show up inadiquacies of other components and allow you to fix them as you go.
I guess at the end of the day it all comes down to your equipment, if your trying to get ultra fine and precise alignment using the bottom of the line gear you will always struggle to push 5hi7 up the hill. on the other hand if you get the top of the line gear... it just works seemlessly. Most of us spend our lives somewhere inbetween and as you say you do what works for you but for me I am a tinkerer and i wont stop till i have figured out how to solve my issues not band aid them for the long term.
It depends on your definition of collimated. My definition is the stars on a full frame chip are symmetrical - and hopefully round. The view I see with my Catseye is not the endgame. It's the second last step. Final tweaking is still required after the Catseye is removed and the camera is inserted.
James
Hi James,
IMO the Catseye autocollimator is capable of getting you much closer than "in the ballpark". I always check my collimation on a defocused star at high power (200X plus) after using the autocollimator and invariably my Tectron autocollimator has me almost spot on and I don't need to tweak anything. If I do need to tweak it, it is about 1 time in 5 and about 1/10th of a turn on 1 screw, which 95% of people wouldn't detect or wouldn't worry about anyway.
I think there are 4 possible problems causing the autocollimator to not be giving you the desired results. These are in order of probability IMO.
1) Your triangle or spot has not been correctly positioned on the primary
2) You have flexure in your focuser drawtube and housing, or play in the drawtube rollers, which manifests under the load of the camera.
3) Your focuser is not squared to the optical and mechanical axes of the telescope and the camera focuses in a significantly different focuser position to that which you are using the autocollimator.
4) The optical and mechanical centres of the primary mirror are not coincidental.
That all having been said you need to use a method that works for you with the equipment you have. However, that doesn't mean it is what the "majority" needs to do.
John
For visual you are 100% correct, but I thought we were talking about imaging here.
James
James,
I asked you in an earlier post "what is the difference between the two". I don't see any, other than a need to compensate for flexure or slop in the focuser, due to the weight of the camera and that IMO is not a collimation issue, it's putting a bandaid on a focuser issue.
IMO a defocused star at 200X is a lot more telling on miscollimation and any other optical issues (to someone who knows what they are looking for) than what you see on your CCD chip.
James tonight I am going to take a few photos of my telescope to show you that even with the mirror not perfectly centered, and the focuser not perfectly aligned to the mechanical center, the cats eye equipment will yield perfectly round stars at the focuser as last night I finally got the scope under the stars for the first time in nearly 6 months (grr weather gods and uni and work and life). I didn't use any form of collimation with the scope nor did I do a defocused star test it is purely auto collimator and stacking the required astrisims (when they all dissapear and your left with P1 amen).
I just want to make sure that when people who are starting out for the first time start reading these posts which they enevitably will that the correct information gets out there. From there people can tweak the methods to suite themself.
Looks like we might get a couple of hours with stars in the sky tonight, it's been crappy weather all week, so I have it back on the bench.
Last run I again got concentric stars inside focus but non concentric outside focus. The secondary offset is 3mm but it should be 5mm. That offset is built into the spider however the 2 short legs, the ones away from the focuser, are 2mm different. I think one has been stretched somehow and I've had to shim the other one by 2mm to get the bolt centered in the horizontal plane.
Is this discrepancy in the offset going to cause me grief? I seem to be having a lot of trouble keeping the secondary centered and round in the sight tube and pointing at the primary center at the same time.
Also these red triangles are rubbish, even with a halogen lamp stuffed down the tube, refection 3 is almost impossible to see. I notice they sell white and yellow center marks so maybe one of those would be better? I should mention that to focus an eyepiece in this scope you need about 80mm worth of extension tubes in the focuser so the autocollminator is not being used at the eyepiece focus position.
Robin are you talking about the red catseye triangles? I never had any problesm seeing 4 with mine - but thean again I had the you-beaut, super duper, catseye clip light - have you tried illuminating it with something red - a super bright LED maybe?
for what it is its hideously overpriced (solidly built though), but it is nice and bright, and it just clips on the end of the tube - hard to beat. I also use mine as a headlamp (clips onto the brim of a cap nicely, and with the little ball joint can point wherever you need it) and keyboard illuminator - clips onto the laptop screen.
