Collimation of Newtonian Telescopes using SharpCap

[admin]

Collimation... a subject on which there are an awful lot of opinions

In theory there are so many adjustments that you could make...

* Tilt of the primary mirror in 2 directions
* position of the primary mirror in 2 directions across the tube
* Tilt of the secondary mirror in 2 directions
* position of the secondary mirror in 3 directions (2 accross the tube, 1 up/down away from the primary)
* twist of the secondary mirror about the axis of the primary
* position of the focuser in 2 directions (along the length of the tube and around the tube)
* tilt of the focuser in 2 directions

Phew... that's 14 separate adjustments if I've counted right.

Fortunately you don't really need to worry about all of these... Some we almost never change (position of the primary mirror relative to the center of the tube for instance). Some don't need adjustment because they can be covered by one of the others - for instance moving the secondary position up/down/across the tube allows you to correct for the primary position and/or the focuser position. Some we tend to adjust infrequently - the position of the secondary left/right/up/down is typical for this - get it right once and don't mess with it again (I may write another post later on how to get this all set up correctly).

The ones we adjust frequently are the tilt of the primary and secondary mirrors, and that's what I'm intending to discuss in more detail here.

So... let's think what we mean by getting good collimation when adjusting the tilt of the primary/secondary mirros. Good collimation means

1) There is zero coma at the center of the image frame. All Newtonians show coma off-axis - we want light that is coming into the telescope along the optical axis of the primary mirror to be hitting the camera sensor at or near to the middle of the sensor. That way the effects of coma will be evenly spread around the edge of the frame with the sharpest stars in the middle

2) No tilt of the camera sensor with respect to the focal plane of the telescope. If there was a tilt then one side of the sensor would be out-of-focus (too close in) and the other side would be out-of-focus (too far out). Another way to look at this is that the light that comes in along the optical axis of the telescope needs to hit the sensor perpendicular to its surface.

That's it... nothing else (as far as I can work out). Note that there is no mention of right angles, 45 degrees or anything similar in these descriptions. In theory it's possible to satisfy these conditions with the secondary mirror at 30 degrees or any other angle to the axis of the telescope (the focuser would have to be at a pretty odd angle, but it could be made to work). In fact there's no need for a secondary at all - you can have a prime focus reflecting telescope where the camera is inside the tube at the prime focus of the primary.

So...

How can we use SharpCap to arrange good collimation according to the two conditions above? The key is to realise that both conditions are satisfied if

A ray of light leaving the primary mirror from its center spot along the optical axis of the mirror reflects off of the secondary and then strikes the imaging sensor at right angles to the sensor surface at (or close to) the center of the imaging sensor

Actually, this hypothetical ray of light is interesting, because to leave the primary from the center spot along the optical axis, it must have *hit* the primary at the center spot along the optical axis, which means that it must double back on itself - any light on this path must start at the center of the camera sensor (or very close to), hit the secondary, hit the center of the primary along the optical axis and then follow the reverse path back to the sensor. Realising this gives us the ability to collimate

Contd in part 2

[admin]

Part 2...

The next step is trying to put some of the theory above into practice. The first thing is to understand the hardware that you need:

* A camera that can fit into the telescope drawtube *with* a lens on it

Guide cameras that fit into a 1.25" drawtube (either partly or entirely) are good for this, since they typically have a C or CS thread, and you can buy C/CS thread lenses of various focal lengths relatively cheaply. Here's the setup I've been using

PXL_20220707_134258084.jpg (389.08 KiB) Viewed 10012 times

This is an Altair GPCAM guide camera (the model doesn't really matter - it's the 1.25" body format that is important) with a 6mm CS thread lens on it. I've also put one of those sticky rings that you use to reinforce the holes in pieces of paper on the lens with a couple of pen marks on it. More on why later...

This can be placed into the drawtube of the telescope - like this

PXL_20220707_134342080.jpg (790.55 KiB) Viewed 10012 times

And here's a view of what you see through the camera when it is in the drawtube (after adjusting the exposure/gain/illumination a bit)

Capture_00001 14_50_40Z_.jpg (127.65 KiB) Viewed 10012 times

Going from the outside in, you can see:

The end of the drawtube
The secondary and secondary holder
The reflection of the primary (and primary holder) in the secondary
The reflection of the secondary mirror in the primary
The reflection of the drawtube all the above mirrors
The camera inside the drawtube, with the paper disc and pen marks around the center of the lens
Right in the middle, the center spot of the primary mirror

You may need to adjust the focus of your lens to be able to see most of these things in decent focus (they will not all be in crisp focus at once - the distances are too different).


So, now we are going to make an assumption - that the sensor that we are going to image with is installed square in the camera, meaning that it will be sitting at right angles to the axis of the drawtube. That means that we can find the line perpendicular to the center of the sensor that we need to find by finding the center axis line of the drawtube.

Now, looking back at the image we see through the camera, the light comes from different angles to illuminate the different pixels in the image. Somewhere in the image there must be a pixel where the light comes almost exactly straight down the axis of the drawtube to get to that pixel. We want to find that pixel as it will allow us to get on with collimation. Note that it won't necessarily be the pixel right in the middle of the sensor, since the sensor might not be perfectly centered in the camera body.

How do we find that special pixel? Well, if that pixel is looking straight along the axis of the drawtube, and we rotate the camera inside the drawtube, everything should appear to rotate in our image *except* that pixel - everything should appear to rotate around that special pixel in the image. We can use this to find the on-axis pixel....

continued in part 3...

[admin]

SharpCap has just added (11th July 2022) a beta 'Multiple Reticules' tool that allows you to place multiple reticules on the image at once - up to 6 independent circles and 2 lines. There are various ways to use this tool to find the 'on axis' pixel, but one is as follows:

Select the 'Multiple Reticules (Beta)' tool from the reticules list, and show both the line reticules.

[Note - this is being treated as a beta feature for now - it will work in a newly released version of SharpCap for up to 60 days after the creation date of that SharpCap version, then it will stop - to continue using it you will need to update SharpCap to a newer version. This may be developed further to provide guidance for the process below. This may become a SharpCap Pro feature]

Place one end of each line onto an easily identifiable feature in the image - in this case I have chosen screws in the primary mirror mount. You can move each end of the line by dragging it with the mouse.

2screws1.JPG (77.09 KiB) Viewed 10011 times

Now rotate the camera in the drawtube through about 180 degrees and move the other end of each line to the new position of the corresponding feature - this means that each of the two lines will run from the old position of the feature to the new position.

2screws2.JPG (58.08 KiB) Viewed 10011 times

Now look at the two dashed lines - the dashed lines are at right angles to the solid lines at their midpoints. Where the dashed lines cross is our first estimate of the position of the 'on axis' pixel. It won't be entirely accurate, but we can refine it later.

Enable one of the circle reticules and place its center point where the two dashed lines cross - you can place the center point with the left mouse button and also use the zoom functionality in SharpCap to zoom in to get the position just right

crossing.JPG (59.86 KiB) Viewed 10010 times

[admin]

Ok, so now we have a fairly good idea of which pixel in the camera is looking straight along the axis of the drawtube, so the next step is to adjust the secondard mirror to place the center mark of the primary so that it is centered on our 'on axis' pixel. Use the normal adjustment screws on the secondary mirror for this - it should be relatively easy as you can see the result of every adjustment you make on screen. Don't worry about perfect accuracy just yet as we haven't refined the position of the 'on axis' pixel yet.

primary1.JPG (45.47 KiB) Viewed 10009 times

Once the center spot is in place, adjust the primary collimation to place the reflection of the sticky ring on the camera lens concentric with the primary center spot in the image, like this

primary2.JPG (31.31 KiB) Viewed 10009 times

Now we need to refine our 'on axis' pixel position - re-show one of the line reticules and place one end right in the center of the primary mirror center spot...

refine1.JPG (33.14 KiB) Viewed 10009 times

Rotate the camera in the drawtube through 180 degrees and place the other end of the line reticule on the new center position of the primary mirror center spot, like this

refine2.JPG (29.76 KiB) Viewed 10009 times

The half way point of the line (where the dashed line crosses it) is the new, improved, position for the 'on axis' pixel. You can now go back and re-adjust the secondary and then primary mirrors based on the new 'on axis' pixel.

If it moved a long way then you probably ought to repeat the refinement procedure at least once more (particularly if your turning through 180 degrees was a bit sloppy). Remember to re-adjust the secondary and primary mirrors to get everything concentric around the 'on axis' pixel each time. Remember to make sure that the camera is pushed firmly into the drawtube and tightened in for each measurement - a self-centring adapter is ideal as any slop in the drawtube can be an issue. You may reach a point where you find that you cannot refine the 'on axis' point any further - every adjustment leads to another small adjustment - at that point you have to make a best estimate, which will be within a few pixels of the correct point.

Once you have finished refining, it's worth noting down the co-ordinates of the 'on axis' point - as long as you do not change the setup (lens/guide cam) then this should be pretty much constant and can be reused next time.

At this point, perform final collimation based on the best 'on axis' pixel position you could achieve, adjusting, as usual, the secondard first and then the primary. This time, lock up the primary adjustment, watching the screen for any shifts caused by tightening the locking bolts.

final.JPG (27.36 KiB) Viewed 10009 times

Let's review what we've achieved so far...

* We have identified a pixel that corresponds to light that is coming down the axis of the drawtube by rotating the camera and finding the point the image rotates about. This light will hit our imaging sensor centrally and perpendicular to the sensor.

* We have arranged that light travelling along that path reflects from the secondary and hits the primary in the middle of the center spot *and* reflects back up the same path, off the secondary a second time and back to the center line of the drawtube (because the image of the paper ring on the camera lens is also centered).

This should imply that the telescope is collimated according to our two criteria given in the first post above.

Why do I say should? Well, there are a lot of things in the optical system that can shift around - a prime example is the focuser which could flex under the (probably) heavier load of the imaging camera or as the telescope orientation is changed when slewing to a target. It's probably best to do a star test and final collimation tweak once all is set up for imaging.

cheers,

Robin

[admin]

This feature is now available for testing in SharpCap 4.0.9152.

Feedback is very welcome - I do not usually use reflecting telescopes (I find they are rather 'high maintenance')

cheers,

Robin

[Ddaniel84]

Hi Robin,
It seems to be like OCAL system.
Of course OCAL is a full system with camera and software. I haven't read all your large explanation, but the fundamentals seem to be the same. Is it thrue ? Where are differences ?
Best, Dominique

[admin]

Hi Dominique,

yes, the camera assisted system to get things concentric goes back a lot further than the OCal device. What the OCal added is the calibration of each camera so that the central pixel on the camera in relation to the 1.25" nosepiece is pre-measured for you. With this system, you have to rotate the camera in the eyepiece holder to calibrate the central pixel for yourself (the one that looks along the axis of the focuser is the one that the rotation appears to be around).

cheers,

Robin

[Ddaniel84]

Thanks Robin
Good idea to include that within Sharpcap!
Is Sharpcap able now to manage meridian flip?
I use NINA for this reason, and the high capabilities of its Advanced Sequencer. It don't manage .SER format. It would be a good idea to create a plugin to allow NINA to use Sharpcap.
Cheers
Dominique

[admin]

Hi Dominique,

there are all the building blocks in the sequence editor to handle meridian flips, but nothing automatic that just deals with it for you. It's on my todo list (probably not until autumn/winter - too much else to do and also it needs a lot of testing for this sort of feature).

cheers,

Robin

[Ddaniel84]

Thanks Robin
I'm really impressed by all the enhancements you have done. Bravo!

Dominique