Support Point Numbers for Mirror Cell?

[glend]

This is sure to open the proverbial 'can of worms' but I need some assistance with what is 'reasonable' on this cell design area.

The mirror is a 250mm f/5 GSO (BK7) with an edge thickness of 34mm. This is for an imaging newt build.

I have PLOP, and have run the numbers but I am unclear about what it is telling me re support points. I gather the idea is to make the PV error and RMS error as small as feasible but what is overkill in the real world (particularly for imaging) .

In reading through this thread below, about the 10" Serrurier build, it is clear that only three bottom support points were used, is this enough?:

http://www.pbase.com/rolfolsen/10_inch_serrurier_truss_tube_newton ian

Obviously three is easiest, and certainly there are no shortage of web links that say its ok to use three up to 10", or even 12", includng the Stellafane tutorials on cell construction (which says use 3 at 71% of radius).

In David Kreige's book 'The Dobsonian Telescope' it specifically notes to 'support mirrors of 10" and under on three points. But then my mirror is not quite a full-thickness mirror in the common 1/6 ratio definition.

I checked this calculator are well, Johann's link is here:

http://www.digilife.be/club/johan.vanbeselaere/atm/optics/mirrorcell/mirrorcell.htm

It says it should have nine. And it gives this informative narative:

supported by 3 points on the edge = lambda/6

supported by 3 points on 70% of the radius = lambda/18

supported by 9 points flotation system = lambda/171

supported by 18 points flotation system = lambda/407

if we tolerate 1/4 of the Rayleigh tolerance (lambda/32 surface peak-to-valley), which seems reasonable, allowing room for other defects such as figure errors in both mirrors, bad seeing,... than this mirror should be supported by a 9 point flotation system !

But when they say we tolerate (lambda/32), I am assuming this is for visual use.

My reading of R F Royce's site found that: "The Rayleigh limit basically asserts that if the wavefront reaching the eye is distorted or deformed by spherical aberration more than 1/4 wavelength of yellow-green light, the image will be perceived as degraded. Any wavefront reaching the eye having a de-formation of 1/4 wave or less will be perceived as essentially perfect."

The key words in all of that maybe 'the eye'.

So what's the real world feedback on this? I'd like to keep it simple and that means three to me, but if image quality is going to be compromised in a noticable way then I am willing to go with the recommendation for quality.

After hours of research I could not find any definitive answer so over to you guys. Thanks in advance.

[Shiraz]

three points should do the job. I have a Skywatcher 10 with 3 point support and it works exceptionally well - the mirror can move around by a few thou on three cork supports and is located by 6 felt edge support stops.

For an imaging system (not for visual though), with a 10 inch mirror you really don't need to be too concerned with slight distortion of the optics - you will rarely get better than 2 arcsec FWHM seeing and even slightly imperfect optics can do much better than that (eg 1/3 wave optics is going to be indistinguishable from perfect optics most of the time).

The biggest problem may be the secondary mounting - astigmatism is the issue. Chinese secondaries can be pretty thin and bendy - a thicker (premium) secondary with high reflectivity coating is probably worth the money

[ariefm71]

For visual use, will a 3 point support works for a 10" f/6 full thickness mirror or a 6 point will be better? Mirror cell will be fitted into a 12" aluminium tube. Also what is the consensus of gluing the mirror into the cell with silicone RTV to avoid edge support?

thanks,
Arief

[glend]

Download PLOP and run your numbers. My 10" f5 has been in use for months now and produces excellent results photographically - which was the subject of thus thread originally. Six is a strange number as usually it is arrived at by doubling three point using rockers at the same spacing distance in radii. I suspect that nine would be better based on my use of PLOP.

[mental4astro]

The support of a mirror to resist flex in it, much of this has to do with the thickness of the mirror itself. Your 10" mirror is very thick. 3 point support for such a thick 10" mirror would be fine. Six points is better. Nine points helps too as Glen mentions. Only thing is you don't gain more support for THIS mirror by putting more support points in. Such a massive mirror does not require so many points. It won't hurt it. Only thing you are creating is more work for yourself.

So how many points? In your case, work out how you will build the cell, and how complicated it will be to make each support system. You are fortunate to have such a thick mirror here as all three support systems will work equally well. My suggestion for you is to make things simple here :thumbsup:. You won't gain anything with complicating matters here. But, but, you will gain invaluble experience by making a 6 or 9 point system in understanding the guts of Newtonian design.


Don't glue the mirror in place unless the whole of the cell is waterproof! Every now and then you will need to clean the mirror, and you don't want to compromise the cell by soaking it.

With your concerns about edge support, two things: 1, don't make it point contact, spread the load over an area. You have a massive thick lump of glass to spread the load over - take advantage of this. 2, the silicone will flex when the mirror is tilted during use. This will see the mirror move ever so slightly, especially it being so heavy, and enough to alter collimation. You need to support its weight. And if you go for 6 or 9 point supports, these will have more movement in them than a 3 point. You are asking for trouble with mirror shift if you don't support the mirror from the sides. Gluing the mirror will work, but it is sloppy workmanship and makes for a compromised mirror.

[ZeroID]

My 10" uses a 6 point but the mirror is not 34mm thick. 6 point was easy to build but also introduced some design elements that made it interesting to build.
Careful how you read the PLOP diagrams, it uses the same range of colours for the differences across the surface. It's only when you start to interpret the numbers that you realise they are are referring to smaller and smaller differences. With more support points the picture may look more complex but the flatness of the surface is far more.

I used industrial Velcro strips under my mirror. Totally flexible, no need for side retainers and it can support the mirror upside down.

[clive milne]

What Ray is saying is essentially correct. In the interests of completeness, however, it should also be pointed out that in a typical imaging rig the ccd under samples the focal plane to a large extent. You would need to be operating at something like f/24 and have perfect seeing before >1/4 wave optics became a major concern whilst imaging.

best,
c

[ariefm71]

Thanks guys.

Arief