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

Mirror edge support - over thought?

  • Please log in to reply
19 replies to this topic

#1 astrokeith

astrokeith

    Apollo

  • -----
  • topic starter
  • Posts: 1,094
  • Joined: 14 Mar 2010
  • Loc: Surrey, UK

Posted 28 July 2021 - 04:01 AM

I read and see lots on Dobsonian mirror edge supports. Slings, wires, whiffletrees & glue. Some can get very passionate about both the need for, and the type of solution.

My own 18" has a good 18pt support, but a very simple 4 pt whiffletree edge support. Before that just two posts at 90 degrees. The mirror is 40mm thick.

I've seen all sorts of thermal problems, but never an edge induced issue. Hmmm.

 

While thinking about another thread (observing Jupiter & Saturn at low altitude with a 16") I posted a comment that I rarely look below 20 degrees elevation.

 

I've used the on-line edge support analysis tool and it thinks that my set up is good, but it also seems to only worry about mirrors angles looking at the horizon. This is extreme and probably almost no one does it?

 

My mirror sits on eighteen 20mm diameter thick felt pads. The friction between these and the mirror back is considerable. My mirror wont slide on them until its angled at more than 45 degrees. So most of the time my edge support is doing almost nothing. At 20 degrees elevation a quick measurements suggests the edge support is exerting just 2kg of sideways force on my 12kg mirror.

 

No wonder i havent seen any issues. Are we all over thinking this?

Attached Thumbnails

  • Screenshot 2021-07-28 at 09.44.15.png

  • Kunama and retroman2 like this

#2 Kunama

Kunama

    Aussie at large

  • *****
  • Posts: 6,137
  • Joined: 22 Oct 2012
  • Loc: Canberra, Australia

Posted 28 July 2021 - 04:16 AM

You may well be right Keith,  I used felt pads on my previous scope build but the new one has 18 hard nylon discs 18mm diameter and this transfers weight to the stainless wire slings well before 45 degrees is reached.  My wire sling is connected to linear bearings that run on 10mm shafts.  

 

I previously had it set up with a pair of 90degree slings that crossed under the mirror edge at the 6 o'clock position. This system was great in that it didn't allow the mirror to shift laterally.  I switched to a single 180 degree sling when I was trying to track down some astigmatism..... only to find the cause was excess padding behind the secondary mirror.  I will return to the dual sling system in due course. seems the important detail is to support the mirror at the CoG....in my case 19mm from the back of the primary.

 

My mirror is an 18" F3.54 and 42mm thick at the edge. 

 

Like you I very seldom view targets below 20degrees and in fact have a stopper at 9 degrees.


Edited by Kunama, 28 July 2021 - 04:24 AM.

  • retroman2 likes this

#3 astrokeith

astrokeith

    Apollo

  • -----
  • topic starter
  • Posts: 1,094
  • Joined: 14 Mar 2010
  • Loc: Surrey, UK

Posted 28 July 2021 - 04:57 AM

Which raises the question as to any difference in performance between felt pads and hard nylon supports?

 

My gut feeling is that felt pads should be better, but it does mean that my support metalwork is closer to the mirror aback and hence restrict cooling air flow.

 

Nylon will tend to ensure the mirror is sitting against its edge support and hence kept central. I swing mine in altitude to make sure the mirror is settled against the whiffletree and then assume it stays there, (cant see why it shouldn't)



#4 sixela

sixela

    Hubble

  • *****
  • Posts: 16,056
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 July 2021 - 05:18 AM

My mirror sits on eighteen 20mm diameter thick felt pads. The friction between these and the mirror back is considerable. My mirror wont slide on them until its angled at more than 45 degrees. 

That's usually bad design. In effect, you need to use the "mirror glued to back support points" option in PLOP or the cruxis.com simulator to see what a stuck mirror yields. Everything else assumes a frictionless back support.

 

I used to have felt pads and in my case the result was very bad when it didn't slide until it was supported by the edge support (400mm mirror, 33mm thick).

 

Apart from that, the fact you don't see problems is probably because you have a good edge support in the first place.

 

I spent two evenings tweaking a scope where the mirror was supported by a sling where the sling was attached to the mirror box and the attachment points were not moving when collimating the mirror. It turned out to be very important to get the sling attachments at the correct height with respect to the mirror on both left and right sides.

 

 

This is extreme and probably almost no one does it?

 

Uhm -- there aren't many alternatives when you want to observe objects with a declination that make them culminate at 7° above the horizon. Perhaps you don't want to, but I once observed M8 set into the sea at one site...and there was surprisingly little atmospheric extinction. Ditto for seeing Omega Centauri set under some faraway hills in Namibia...

We've had years now for which observing Jupiter and Saturn meant aiming quite low. And obviously, any astigmatism would be seen obviously (if you had an atmospheric dispersion corrector as well).


Edited by sixela, 28 July 2021 - 05:44 AM.

  • Jon Isaacs likes this

#5 astrokeith

astrokeith

    Apollo

  • -----
  • topic starter
  • Posts: 1,094
  • Joined: 14 Mar 2010
  • Loc: Surrey, UK

Posted 28 July 2021 - 05:43 AM

That's usually bad design. In effect, you need to use the "mirror glued to back support points" option in PLOP or the cruxis.com simulator to see what a stuck mirror yields. Everything else assumes a frictionless back support.

 

I used to have felt pads and in my case the result was very bad when it didn't slide until it was supported by the edge support (400mm mirror, 33mm thick).

 

Apart from that, the fact you don't see problems is probably because you have a good edge support in the first place.

 

I spent two evenings tweaking a scope where the mirror was supported by a sling where the sling was attached to the mirror box and the attachment points were not moving when collimating the mirror. It turned out to be very important to get the sling attachments at the correct height with respect to the mirror on both left and right sides.

I can see that felt pads are part way to glue. But surely no way as near 'stuck'! So I dont think the glued back analogy is right. Somewhere in-between - but where?!

 

Based on my 45 degrees before the mirror slides, I did a quick fag packet calculation of the 'tension' across the mirror back surface as the steel support changes with temperature, and find it two orders of magnitude smaller than that necessary to distort the mirror. If I assume the elasticity of silicone rubber glue, then i indeed get a problem. So I think my fag packet calc is good enough.



#6 sixela

sixela

    Hubble

  • *****
  • Posts: 16,056
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 July 2021 - 05:54 AM

I can see that felt pads are part way to glue. But surely no way as near 'stuck'! 

I either slides or it doesn't. If it doesn't -- and you're claiming that with angles up to 45° it doesn't slide-- then the force of gravity in the plane of the mirror is entirely counteracted by the static friction, and even though it's in the plane of the mirror these forces are not in a plane that contains the COG so they will introduce bending forces the glass will need to counteract though deformation. It might just as well be glued to the pads.

 

For an f/4.5:
 

 

Reduction of Strehl Ratio at horizon: 0.21 ; at 45° altitude: 0.11

 

It's not a catastrophy, it might not be that easy to see at lower powers, in 'average' seeing you might not notice it at the powers you can use effectively, but it's still there. And I know some 18" mirrors far thinner where it would be a real nightmare (I've used a couple with mirrors made out of Borofloat 33, which only goes up to 25mm).

 

You can simulate it in PLOP3D (only tricky thing is you usually have to edit the angles and add 90°, since PLOP will angle the mirror to the right when it simulated edge support), wil yield similar results (and a pretty picture of the wavefront deviations).

 

Of course in practice you are luckier if the mirror is indeed close enough to the edge support. The felt will 'give' a bit until part of the weight is supported by friction and part is supported by the edge support. Still, not good design and performance will not be repeatable (been there, done that, designed by someone else until I rebuilt the scope). In such a setup it's not uncommon to have the mirror move a bit when you've been observing close to the zenith (or sometimes over it) and moved the scope in Az, and then when you go back down to another object suddenly the views are worse until you give it all a good whack.

 

If you get frictionless support points you no longer have to worry about it, especially with good (also low friction!) edge support points. You already have a whiffle tree, so that's a no-brainer. That setup is also quite a bit more tolerant about not touching the mirror exactly in the COG than a sling, whose ends can pull on the mirror quite easily unless they're on linear bearings.

 

I've debugged enough scopes (including my own before I rebuilt it) to know that we're not just imagining things and worrying about nothing. I saw Obsessions with obvious edge support problems despite the humongously thick mirrors that means they get away with more (but are vastly less practical than scopes with thinner mirrors provided those are well supported too).


Edited by sixela, 28 July 2021 - 06:08 AM.

  • SandyHouTex likes this

#7 astrokeith

astrokeith

    Apollo

  • -----
  • topic starter
  • Posts: 1,094
  • Joined: 14 Mar 2010
  • Loc: Surrey, UK

Posted 28 July 2021 - 07:06 AM

I either slides or it doesn't. If it doesn't -- and you're claiming that with angles up to 45° it doesn't slide-- then the force of gravity in the plane of the mirror is entirely counteracted by the static friction, and even though it's in the plane of the mirror these forces are not in a plane that contains the COG so they will introduce bending forces the glass will need to counteract though deformation. It might just as well be glued to the pads.

 

For an f/4.5:
 

 

 

It's not a catastrophy, it might not be that easy to see at lower powers, in 'average' seeing you might not notice it at the powers you can use effectively, but it's still there. And I know some 18" mirrors far thinner where it would be a real nightmare (I've used a couple with mirrors made out of Borofloat 33, which only goes up to 25mm).

 

You can simulate it in PLOP3D (only tricky thing is you usually have to edit the angles and add 90°, since PLOP will angle the mirror to the right when it simulated edge support), wil yield similar results (and a pretty picture of the wavefront deviations).

 

Of course in practice you are luckier if the mirror is indeed close enough to the edge support. The felt will 'give' a bit until part of the weight is supported by friction and part is supported by the edge support. Still, not good design and performance will not be repeatable (been there, done that, designed by someone else until I rebuilt the scope). In such a setup it's not uncommon to have the mirror move a bit when you've been observing close to the zenith (or sometimes over it) and moved the scope in Az, and then when you go back down to another object suddenly the views are worse until you give it all a good whack.

 

If you get frictionless support points you no longer have to worry about it, especially with good (also low friction!) edge support points. You already have a whiffle tree, so that's a no-brainer. That setup is also quite a bit more tolerant about not touching the mirror exactly in the COG than a sling, whose ends can pull on the mirror quite easily unless they're on linear bearings.

 

I've debugged enough scopes (including my own before I rebuilt it) to know that we're not just imagining things and worrying about nothing. I saw Obsessions with obvious edge support problems despite the humongously thick mirrors that means they get away with more (but are vastly less practical than scopes with thinner mirrors provided those are well supported too).

I'm still not convinced!

 

So my mirror does slide, when under a sideways force of about 6kg at the pad/glass interface. So at worst case the back of the mirror is seeing a force across its back surface. The force is spread evenly across the 18points, but the worst effect would be caused by the outer pads, cant remember but these are on about a 14" diameter. on average each pad can exert a maximum side force of 6/18 = 330gF. For the life of me I cannot get any significant bending of the mirror. Quick calc with a thin 24" mirror does show a risk. So perhaps I'm lucky.

 

The moment the scope is moved, I suspect the stiction in the pad/mirror interface is likely to be released in practice. So again I'm getting away with it.



#8 sixela

sixela

    Hubble

  • *****
  • Posts: 16,056
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 July 2021 - 07:14 AM

For the life of me I cannot get any significant bending of the mirror

In what way do you suggest that using e.g. PLOP3D or the Cruxis calculator with the support points glued to the back are any different than what you describe?

 

In what way do you suggest the FEA analysis done by these tools is incorrect? The only thing PLOP introduces when it models that 'glued' edge support is static frictional forces along the back plane of the mirror that prevent the mirror from moving.

 

I remember discussing the actual failure modes with Nils Olof a long time ago. I haven't heard anything from you to suggest any changes to what I learned there.

 

But of course whatever works for you is good enough for you. If it works, it works, but you were asking about sound design principles.

 

I'll see if I can kick PLOP3D's tires once more if I have the time -- I tend to trust it more than any back of the envelope calculations that might not fully characterise what happens.

 

Here it is with a 450mm mirror, 40mm thick f/4.5 mirror glued to 18 points. Right is "down".

At 45° of the horizon first, assuming your mirror is indeed stuck on the felt pads when slanted this much, but the effects of a "stuck" mirror that doesn't use its edge support at all at 80° altitude is also not negligible:

450_40_glued_45deg.jpg 450_40_glued_80deg.jpg

 

RMS errors on the wavefront are 44nm at 45° of zenith and 7.4nm at 10° of zenith. Corresponding Strehl ratios are 0.78 and 0.99 (but frankly, even that's already more than a good 6 point cell with good edge support would yield as degradation).

 

So yeah, at 45° you'd better jog your mirror to transfer all the weight that's possible onto the edge support (which is what low friction back support points do).

 

At 20° of the horizon of course your felt pads will not be able to let the mirror stick too much -- it'll overcome static friction and slide onto the edge support. Fortunately, because a glued mirror gives you a colossal error (58nm on the wavefront, Strehl ratio of 0.65. Not unusable, but not what I'd call 'good'.)


Edited by sixela, 28 July 2021 - 10:24 AM.


#9 Mike Lockwood

Mike Lockwood

    Vendor, Lockwood Custom Optics

  • *****
  • Vendors
  • Posts: 2,001
  • Joined: 01 Oct 2007
  • Loc: Usually in my optical shop

Posted 28 July 2021 - 10:24 AM

Eventually felt sticks.  Moisture and dirt build up over time.  I've seen it happen.

 

Felt also compresses.  When the felt compresses, the mirror moves slightly.  If the edge support has friction (i.e. it doesn't use rollers), then that friction will distort the mirror as the felt is loaded and unloaded with the weight of the mirror, and that will cause distortion of the optic.  (Even the metal of the cell itself still flexes slightly as the cell loads and unloads, but felt adds to it.)

 

The ideal support case is a low friction rear support and edge support that can only apply force on the plane of the center of gravity of the mirror, and not in the direction of the optical axis.

 

You might just be fortunate that it works, or more likely there is some distortion that is concealed by other factors.


  • GUS.K likes this

#10 Redbetter

Redbetter

    Voyager 1

  • *****
  • Posts: 11,161
  • Joined: 16 Feb 2016
  • Loc: Central Valley, CA

Posted 28 July 2021 - 10:29 AM

With respect to observing very low near the horizon, I mostly avoid it for several reasons:  very poor seeing that low (e.g. as comparing the view through a refractor for confirmation), high levels of extinction and sky brightness, and high levels of atmospheric chromatic dispersion.  My recent observation of diffuse globular NGC 6352 at 5 degrees in Ara demonstrated how tough this can be, with ~4 magnitude of extinction observing over the hazy illuminated valley from 4,000 feet. 

 

Somewhere below 15 degrees about all I am likely targeting with the 20" is globulars and going that low for them is unusual.  From an 8,000+ foot site that has a clearer and darker southern horizon I do observe some galaxies at ~20 degrees (Sculptor and Fornax dwarf galaxies) to as low as about 11 (Grus quartet), but it is rare.  It is tough to get clear enough nights with decent seeing where it is worth going so low on targets with typical galaxy surface brightness.  And when I do get such nights, things higher in the sky will be much better as well...so it becomes a question of "what is the best use of great conditions, looking at something low where it will still be marginal, or looking at other objects higher where they will be at their best?"

 

The main problem I have seen with a sling pointed this low is that if it is not well positioned it seems to alter the tilt of the mirror slightly--so that it isn't fully flush along all of the felt pads.  That was killing me with the old thin (and worn/torn on on one edge) manila sling until I replaced it.  Collimation was shifting, which masqueraded as astigmatism or was added on top of it...hard to say which was the primary culprit.

 

But if there is no tilt being added to the mirror plane, I do wonder if there is much value in worrying about small stress deformation with the mirror approaching on edge.  There is just so much else going on down low that if the mirror remains in collimation, I doubt it matters much.  Or perhaps I should restate that to:  if the edge support holds the primary in collimation when pointed near horizontal, deformation of the edge is likely a secondary or even tertiary concern, at least in the 20" range.  That might not be true as the mirror size increases. 


  • Jon Isaacs likes this

#11 astrokeith

astrokeith

    Apollo

  • -----
  • topic starter
  • Posts: 1,094
  • Joined: 14 Mar 2010
  • Loc: Surrey, UK

Posted 28 July 2021 - 10:34 AM

In what way do you suggest that using e.g. PLOP3D or the Cruxis calculator with the support points glued to the back are any different than what you describe?

 

In what way do you suggest the FEA analysis done by these tools is incorrect?

 

I remember discussing the actual failure modes with Nils Olof a long time ago. I haven't heard anything from you to suggest any changes to what I learned there.

 

But of course whatever works for you is good enough for you. If it works, it works, but you were asking about sound design principles.

 

I'll see if I can kick PLOP3D's tires once more if I have the time -- I tend to trust it more than any back of the envelope calculations that might not fully characterise what happens.

 

Here it is with a 450mm mirror, 40mm thick f/4.5 mirror glued to 18 points. Right is "down".

At 45° of the horizon first, assuming your mirror is indeed stuck on the felt pads when slanted this much, but the effects of a "stuck" mirror that doesn't use its edge support at all at 80° altitude is also not negligible:

attachicon.gif450_40_glued_45deg.jpgattachicon.gif450_40_glued_80deg.jpg

 

RMS errors on the wavefront are 44nm at 45° of zenith and 7.4nm at 10° of zenith. Corresponding Strehl ratios are 0.78 and 0.99 (but frankly, even that's already more than a good 6 point cell with good edge support would yield as degradation).

 

So yeah, at 45° you'd better jog your mirror to transfer all the weight that's possible onto the edge support (which is what low friction back support points do).

 

At 20° of the horizon of course your felt pads will not be able to let the mirror stick too much -- it'll overcome static friction and slide onto the edge support. Fortunately, because a glued mirror gives you a colossal error (58nm on the wavefront, Strehl ratio of 0.65. Not unusable, but not what I'd call 'good'.)

Don't get me wrong, I've used PLOP for years and trust it completely. But all models have their boundaries.

 

The question is around 'stuck'. I am wondering what is going on at the felt pad. There will be 'creep', 'give', etc much of which is different to a glued support point.

 

Take for instance thermal expansion of the support metal work. This will apply an enormous lateral force at the contact points. A glued point would deform a little and transmit the force to the bottom of the mirror. But under such forces my felt pads will slip. So under some circumstances at least, glue and felt aren't equivalent.

 

I strongly suspect my felt pads (about 5mm thick) are compliant enough to not transmit any significant forces, but I cant think of any way of verifying this. I know I dont have a problem, as I have been using this mount for 15 years and I nor any of my friends have seen astigmatism or other mirror distortion issues. The felt get renewed every 3-5 years.

 

Perhaps my felt is thick and robust enough to have worked OK.



#12 sixela

sixela

    Hubble

  • *****
  • Posts: 16,056
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 July 2021 - 11:16 AM

I've seen all sorts of thermal problems, but never an edge induced issue. Hmmm.

TL;DR: absence of evidence is not evidence of absence. Especially since I have seen edge support induced issues (and it's not like there is a dearth of information about them, including Mike Lockwood's little page with experimental results).
 

There will be 'creep', 'give', etc much of which is different to a glued support point.

Doesn't really matter. The end result is a force that exactly counteracts the mirror's tendency to slide down due to gravity, since in equilibrium you're telling us the mirror isn't moving anymore. There is F, there is m, so unless something cancels F through F=m.a a is non-zero. Since you told us that eventually a is 0 in equiilibrium...
 
One might even posit that the PLOP3D simulator fits felt pads that apply static friction better than it models glue points, in general (as you quite rightly point out).
 
 

A glued point would deform a little and transmit the force to the bottom of the mirror. But under such forces my felt pads will slip. So under some circumstances at least, glue and felt aren't equivalent.

Correct. Which is why I said exactly what the PLOP edge support simulator does in this case.
 
 

I strongly suspect my felt pads (about 5mm thick) are compliant enough to not transmit any significant forces

If there isn't a "significant force" due to static friction, then the mirror slides down under the effects of gravity when the mirror is slanted 45°.

Yet you are telling us it doesn't. I still fail to see how felt pads would keep the mirror from sliding down without applying forces, compliant though they be. They'd have to be magical or telekinetic at the least, not just compliant.
 
 

Perhaps my felt is thick and robust enough to have worked OK.

No, you're moving the scope while inclined low enough until the mirror wins against the felt pad's static friciton and rests on the edge support despite the friction of the pads, and then you're lucky they're compliant enough not to make the mirror stick somewhere it should not later (but I've seen scopes where you'd not want to bump the scope while pointed closer to zenith -- including the former build of my current scope). Tat's not good design, even though it may 'work'.
 
If by now you aren't convinced that you need to design a good edge support and that people are just 'making a fuss', you're just making an argument based on personal incredulity. There are sound reasons to worry about this, for the simple reason that relying on sheer luck or 'I just shake the scope a bit when I start observing and mentally sacrifice a goat' (which seems to work with your edge support and your felt pads but which I would not really advise if you had a sling!) doesn't always work reliably for all scopes. I've seen enough scopes in my life with edge support issues to make an argument with a single counterexample where it does work rather unconvincing.
 
To me, the simple fact that you can slant your mirror 30° and it will still stick to me means there is far too much static friction for the design to be what I call 'good'.

Edited by sixela, 28 July 2021 - 11:32 AM.


#13 Bill Weir

Bill Weir

    Mercury-Atlas

  • *****
  • Posts: 2,962
  • Joined: 01 Jun 2004
  • Loc: Metchosin (Victoria), Canada

Posted 28 July 2021 - 11:31 AM

http://www.loptics.c...rorsupport.html

 

Mike’s excellent article working through all aspects of support.

 

Bill


  • Asbytec and GUS.K like this

#14 astrokeith

astrokeith

    Apollo

  • -----
  • topic starter
  • Posts: 1,094
  • Joined: 14 Mar 2010
  • Loc: Surrey, UK

Posted 28 July 2021 - 12:06 PM

Yes my mirror doesn't slide off it pads at 45deg, but that doesn't mean it isn't moving a little due to the compliance of the pads. It doesn't have to move much to remove or reduce stress.

 

I start my observing session with the mirror in contact with the edge support - I always check collimation is stable by swinging the telescope fully in Altitude a few times, hence the mirror will be in good contact with the whiffletree.

 

So there is no opportunity for the mirror to 'slide', as it is already on the edge supports.

 

During the session as I move around the sky, no doubt there is some inertia being applied to the mass of the mirror and I know it continues to touch the edge supports (I have checked many a time) probably though the felt pads allowing the minute movements necessary. This is the 'creep' I refer to.

 

There may be some lateral force still being produced by the felt pads, but as the mirror is in contact with the edge supports it will be much less than if it was glued.

 

What I am desperately trying to explain is the complexity of what might be happening in practice, and how it may compare with a model. I have been a scientist for 40 years and know that reality and theory often don't agree. The reasons are sometimes obscure and also interesting. The models are exact in describing the characteristics of what they include - but they don't include deformation of felt pads and the friction/stiction of the felt to glass surface. They don't include the practical impact of the dynamic forces of moving a telescope around. These and other factors are outside the scope of the models.

 

As Mike Lockwood pointed out, felt will get contaminated and compressed. Perhaps my regular renewals are reducing that potential problem. Also my mirror only sits on its pads for a few days of the year. In the UK we usually get small and gentle temperature changes during the session. Perhaps that helps. My mirror is on the thick side, compared with the 'trend'. So many variables.

 

I'm not saying felt pads are better, and probably next time I work on my scope I will try some hard nylon ones. I'm just saying that in my case they are working very well.



#15 sixela

sixela

    Hubble

  • *****
  • Posts: 16,056
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 July 2021 - 01:52 PM

It doesn't have to move much to remove or reduce stress.


If it is at 45° and not down on the edge support, then no, there are stresses as long as it ends up with zero acceleration towards the bottom into an equilibrium position.
 

I'm just saying that in my case they are working very well.

I had a scope that worked exactly like yours and it was usable, but "very well" is not how I'd describe how it worked. It was possible to jolt the mirror out of contact of the edge support and it'd stick there and show astigatism until I gave it a whack or turned it near the horizon to "reset" things.

Edited by sixela, 28 July 2021 - 01:55 PM.


#16 Starman1

Starman1

    Vendor (EyepiecesEtc.com)

  • *****
  • Vendors
  • Posts: 52,068
  • Joined: 23 Jun 2003
  • Loc: Los Angeles

Posted 28 July 2021 - 03:05 PM

I read and see lots on Dobsonian mirror edge supports. Slings, wires, whiffletrees & glue. Some can get very passionate about both the need for, and the type of solution.

My own 18" has a good 18pt support, but a very simple 4 pt whiffletree edge support. Before that just two posts at 90 degrees. The mirror is 40mm thick.

I've seen all sorts of thermal problems, but never an edge induced issue. Hmmm.

 

While thinking about another thread (observing Jupiter & Saturn at low altitude with a 16") I posted a comment that I rarely look below 20 degrees elevation.

 

I've used the on-line edge support analysis tool and it thinks that my set up is good, but it also seems to only worry about mirrors angles looking at the horizon. This is extreme and probably almost no one does it?

 

My mirror sits on eighteen 20mm diameter thick felt pads. The friction between these and the mirror back is considerable. My mirror wont slide on them until its angled at more than 45 degrees. So most of the time my edge support is doing almost nothing. At 20 degrees elevation a quick measurements suggests the edge support is exerting just 2kg of sideways force on my 12kg mirror.

 

No wonder i havent seen any issues. Are we all over thinking this?

Thicker mirrors tend to exhibit less in the way of support-induced aberrations.

But, there should not be enough friction to prevent the mirror from sliding.

You have a good number of support points, but they should be more frictionless.

However, as the mirror rests on the 4 points in the whiffle tree, you need really only make certain the contact points are in the front-back center of gravity for the edge support.

If you look at the wavefront error in PLOP with the mirror at a 45° angle, you can evaluate whether you're OK with your current setup.


  • GUS.K likes this

#17 Piero DP

Piero DP

    Viking 1

  • -----
  • Posts: 702
  • Joined: 28 Jan 2015
  • Loc: Cambridge, UK

Posted 28 July 2021 - 04:26 PM

I only have experience with two Newtonians.
Based on this, it seems to me that it is far easier that mirror cell-dependent aberrations are caused by incorrect lateral support rather than the back supports. I've used felt and nylon pads without issues. The felt pads I use don't seem to stick or cause friction and the mirrors slide very easy on them.

Both mirrors are laterally supported by sling cable: the 16" via an original Glatter's sling, the 12" via a sling installed on the same plane where the mirror triangles are installed. In both, it is critical that the sling is at the COG of the mirror (all the time) otherwise astigmatism can appear. Velcro pads help make sure this alignment is maintained.

I found that fix lateral supports (e.g. 3 strong stoppers with some pads touching the mirror) can cause very severe astigmatism but also spherical aberration. Whiffletrees with roller bearings solve this issue as the mirror can slide rather than being constrained.

In the end of the day, to give their best, mirrors should not be constrained by external forces like friction.

I don't think people are too fussy when they worry about the lateral support of their primary mirror. It's very critical to get it right, and it seems to me that the consequences of an incorrect installation are not well understood by most.
Personally, I am glad that this topic continues to be discussed here on CN.

Yeah.. a lot of people blame the seeing as the major cause for not being able to use high magnifications. Although this is certainly possible, I do wonder how many of those telescopes have poor lateral support.. in a similar way I wonder how many of those telescopes have air boundary layer issues above the mirror..

Edited by Piero DP, 28 July 2021 - 04:36 PM.


#18 a__l

a__l

    Gemini

  • -----
  • Posts: 3,193
  • Joined: 24 Nov 2007

Posted 28 July 2021 - 07:26 PM

astrokeith, for 40 mm thickness, half the number of felt points is sufficient. Perhaps then sticking will be less. Check with the laser if you have any displacement. I have felt pads on all my telescopes, including Zambuto 14.5 (37 mm thick). I see no problem with beam offset.

If your felt is dirty, replace it. This is 10 minutes of time. And observe planets for your pleasure.

 

If I had something stiffer than felt, I would probably have fewer telescopes now.
For many years of travel, there have been quite hard jumps by car (at night).


Edited by a__l, 28 July 2021 - 07:36 PM.


#19 MitchAlsup

MitchAlsup

    Fly Me to the Moon

  • -----
  • Posts: 5,795
  • Joined: 31 Aug 2009

Posted 29 July 2021 - 05:06 PM

The mirror cell and edge support of my 13" F/3 are constructed from ball transfers and the friction is so low that the primary can swivel around like a lazy Susan--at any angle--yet the primary is never more than 0.004" from center of the mirror cell due to the adjustments I performed when the mirror was installed.

 

End of page 4 and beginning of page 5::

 

https://www.hobby-ma...ld.65805/page-4


  • Mike Lockwood and Kunama like this

#20 a__l

a__l

    Gemini

  • -----
  • Posts: 3,193
  • Joined: 24 Nov 2007

Posted 29 July 2021 - 08:17 PM

You just have to give evidence in the form of an interferogram that it does not have astigmatism. I have demonstrated several times for my 14.5 "ZOC.
For ATM, it is easier to make support on a steel cable (I also demonstrated it) and forget about the problem. Anyway, for the thickness of the mirrors that Zambuto made.

 

(I can't see the link in full due to AdBlock Detected).


Edited by a__l, 29 July 2021 - 09:17 PM.



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.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics