Dumb question about newtonian optics geometry

[markse68]

My gut feeling is it shouldn’t matter apart from even field illumination and risk of vignetting, but does it matter if the focuser optical axis isn’t exactly perpendicular to the primary mirrors optical axis? Eg if the secondary isn’t at the perfect height (is there a way of actually ascertaining that?!?) and you end up compensating with the optical axis angles to get good collimation, does this cause any aberrations like astigmatism? Surely not?

Mark

[Pixies]

Hi Mark,

In my non-expert opinion....

The issue is not the angle you show in your diagram, its the fact that in such a case - where the parabolic mirror would be tilted slightly off-axis - the incoming parallel light rays will not be on-axis to the mirror. This will cause coma (I think)

See p3 and p4 of http://scipp.ucsc.edu/~haber/ph5B/parabolic09.pdf

[markse68]

thanks Pixies- that’s an interesting thought. I had just assumed that the primary would “see” an incoming beam with optical axis non-parallel to the ota and risk vignetting by the opening of the ota. But if this is the case, how on earth do you establish the accurate correct height of the secondary mirror? I don’t think you can rely on centring the secondary with concentre/sight tube/ w/a camera as that relies on having the secondary offset perfect too. You would almost have to remove the secondary and then collimate the primary to a laser perfectly central and aligned to the ota centre axis first, then adjust the height of secondary until you could reach collimation-without touching the primary. That would be very difficult and time consuming I think!

Mark

[Pixies]

2 minutes ago, markse68 said:

thanks Pixies- that’s an interesting thought. I had just assumed that the primary would “see” an incoming beam with optical axis non-parallel to the ota and risk vignetting by the opening of the ota. But if this is the case, how on earth do you establish the accurate correct height of the secondary mirror? I don’t think you can rely on centring the secondary with concentre/sight tube/ w/a camera as that relies on having the secondary offset perfect too. You would almost have to remove the secondary and then collimate the primary to a laser perfectly central and aligned to the ota centre axis first, then adjust the height of secondary until you could reach collimation-without touching the primary. That would be very difficult and time consuming I think!

Mark

Hi Mark,

In modern scopes, the shape and offset of the secondary are such that when they present a perfect circle concentrically under the correctly aligned and perpendicular focuser when collimated correctly.

As for off-axis targets. That's a fact of Newtonians - the more off-axis it is, the more coma you see. That's why when you are collimating a scope with a star-test, the star has to be dead-centre. As you move the star away from the central axis, you'll see the diffraction rings start to 'bunch-up' to one side as the coma increases. Our scopes are usually slow enough that coma doesn't really bother us visually if viewing objects off-axis in a well collimated scope.

Well - that's how I understand it to be.

 

[markse68]

But what if the offset isn’t accurate? I have no idea about mine as it came in a frankenscope package  

Ive had an idea- there’s a recent thread on here with a very clever method to adjust sensor tilt as used by Starlight Express using a laser bouncing off the sensor to a target, then adjusting the tilt until the spot remains static when the camera body is rotated in a v-block. I could do similar by taking my ota sans secondary/spider and resting it between 4 fixed castor wheels so it can be rotated, and adjusting primary tilt likewise. Then adjust secondary only for perfect rectilinear collimation 🤔

Mark

[Pixies]

I've seen articles on how to calculate the correct size and offset for a secondary mirror. Scope builders will be very familiar with this, I'm sure. However, I also recall seeing an article about offset that shows an older method where the secondary is offset differently with the primary slightly tilted. Perhaps that's more applicable to a spherical primary mirror, where off-axis objects are less affected by coma,  rather than a parabolic one.

Anyway - I'm not an expert and I'm sure a few will be along shortly. But I reckon some scope-builders will be a good source of advice.

As for your cunning plan to align the primary first, then (I assume) align the secondary from  that... The trick will be making sure you can accurately align the scope with the laser. I would have thought it easier to make sure your focuser is straight/perpendicular to the tube, then you can use a correctly configured/offset secondary. The issue will be finding the correct secondary configuration, but that must surely just be a case of finding the correct advice/experience?

[johninderby]

Assuming the focuser is square to the tube a concentre will allow you to perfectly centre the secondary. A concentre is not like a simple sight tube as the way it is made allows for accuracy that is simply impossible with a sight tube or cheshire.

As for the primary adjust it physically level by using a steel rule and the adjustment bolts. Then when the secondary is adjusted with the concentre check collimation and you will find collimation is very, very close and only minor adjustment of the primary is needed.

[Pixies]

OK. I have been thinking a little more. Forgetting about the correct offset/size of the secondary for now.

1) make sure the focuser is set up correctly. I.E., its central axis intercepts and is perpendicular to the central axis of the scope.

2) the centre of the primary is marked.

Now - it doesn't matter about the offset or size of the secondary, as this only affects field illumination. Worry about that later! Just adjust the secondary angle so that the primary centre spot is directly central to the focuser view. You will have to make sure the secondary is facing the focuser by rotating it and looking for the widest angle as seen through the focuser. As long as 1 and 2 above are correct, you now know for certain that the secondary is at 45degrees and the central axis of the focuser is aligned to the tube axis (via the centre spot). All you now need to do is the usual primary collimation so that the primary is aligned back up the same axis - either laser or Cheshire/collimation cap. 

When the primary centre spot is dead-centre to the focuser and the primary reflects the collimation device back to the same central point - it's all aligned and straight. Just assuming 1 and 2 above.

Then you can worry about secondary size and offset. But this is all to do with field illumination.

 

Any comments? I think that's correct.

[markse68]

But if the secondary height isn’t perfect won’t you just end up with the green drawing in my first post?

I don’t think i’d need to align the laser perfectly for my other proposal- it’d rely on the primary being perfectly optically centralised to the ota- i’d have to assume the optical centre of the primary was concentric to the physical centre. I’d just be looking to make the reflected laser dot static as the ota rotated. Hmm but the primary would decollimate the laser beam so that might not work at all actually- i’ll have to give it a try- edit the mirror should focus the laser beam to a point at the focal length distance from the mirror which would be outside the tube so should be ok if target is placed close enough.

@johninderby the concentre must rely on perfect secondary offset i think to bring optical centre of secondary to intersect with optical axis of ota? That’s what I can’t guarantee and am trying to figure a way around. Plus i bought wrong concentre and can’t focus the rings 🤦‍♂️ Plus i use guide cam with fish eye lens and overlay concentric circles to do same thing

Mark

Edited September 8, 2021 by markse68

[johninderby]

This article is useful. Although it is about replacing a secondary mirror gives some useful info on the offset.

http://www.deepskywatch.com/Articles/replace-offset-collimate-secondary.html

[markse68]

I'm having a go at getting more accurate orthogonal collimation to see if that improves my star shapes- still not convinced it will but worth a try.

I used a small trolley we had at work and jammed some bar stock in the castors to keep the swivelling wheels aligned. It's just about a perfect size for my OTA

I had previously glued a bolt to the centre of the back of the primary so I could jiggle the mirror to make sure it was seated properly and not hanging up on any of the cell contact points- this is handy also for this to pull the mirror back against the cell supports as it's going to rest horizontally on the wheels of the upturned trolley.

Before starting I measured and adjusted the primary in its cell to be equidistant from the cell mounting points where it attaches to the OTA so hopefully the optical axis should be close to the centre axis of the OTA- the Orion cell is nice for this.

A laser pointer reflects off the mirror onto a target in front of the OTA- I had removed the entire secondary assembly.

Rotating the OTA on the wheels the reflected laser described a circle- the optical axis wasn't coincident with the OTA axis

After adjustment the reflected dot is much closer to static- non circularity of the OTA tube and possibly the wheels made it impossible to get it absolutely perfect but it should be close enough.

Just an elaborate way of doing this stage- maybe John's suggestion of just measuring the mirror edge to bottom of OTA tube would have been accurate enough but nice to know its really there.

Now to recollimate the scope using only the secondary. This could take some time...

Mark

 

Edited September 9, 2021 by markse68

[alacant]

On 08/09/2021 at 11:02, markse68 said:

it shouldn’t matter

Hi

If you use a Cheshire sight tube correctly -preferably one with cross hairs- it doesn't matter where the focuser aims.

You've already had a look at Telia's collimation myths and Seronik's no nonsense guide so you're most of the way there!

Cheers

[markse68]

This was the result of my playing last night.

Not sure what it means but looks like my secondary offset is wrong. And that my field illumination will be offset as a result.

This makes sense as I now believe my secondary assembly to come from a Vixen R200SS- an f4 scope. Certainly looks very like it. My scope is f4.5 and should have a different offset.

Mark

[markse68]

3 hours ago, alacant said:

Hi

If you use a Cheshire sight tube correctly -preferably one with cross hairs- it doesn't matter where the focuser aims.

You've already had a look at Telia's collimation myths and Seronik's no nonsense guide so you're most of the way there!

Cheers

Hi,

are you saying the answer to my initial post question is that there is no effect of having the primary optical axis meet the focuser optical axis at other than 90deg? It was suggested that it could lead to coma and i’m curious if it is leading to the elongation of star shapes i’m seeing infra/extra focus. 

Thanks,

Mark

[alacant]

2 hours ago, markse68 said:

elongation of star shapes i’m seeing infra/extra focus

May be better to concentrate upon stars which are in focus. Coma correctors act in mysterious ways.

Not sure to which telescope you refer. One is an 8.75" f7, the other either a vixen f4 or vixen f4.5.

If the shot along your focuser was produced by collimating using a cross hair Cheshire, the offset looks just like our f3.9 and so with a good cc, you should have decent stars edge to edge. If it's around f4-ish.

from Telia:

myth: You have to square the focuser very accurately

I'm not quite sure of even what "square" is supposed to mean - likely it means set perpendicular to the tube, or possibly to the optical axis - or both, always assuming you have made them coincide. There is nothing wrong with doing it, of course, but the secondary is optically flat, and the angle of reflection is not critical. Most secondaries are made to look circular when tilted 45 degrees (to reflect 90 degrees), but if the angle deviates from this by a few degrees, the only consequence is that the secondary will appear slightly elliptic - it won't affect the image.

 

Edited September 10, 2021 by alacant