Mirror cell

[dark star]

I am trying to figure out a mirror cell design for my 350mm diameter 25mm thick plate glass mirror. I am tryiing to use PLOP which is a free mirror cell design software. I have two problems. One is that when I run a mirror cell design with 18 supports it gives me a worse PV and RMS error than with less supports! Am I making a silly mistake and the higher the figure the better, but surely if it is a figure for error less should be better?

The other problem is I cannot make any sense of the output of PLOP for the actual dimensions of the mirror cell and placing of support points. Could anyone who has used PLOP give me some advice, preferably in very simple terms!

Ideally I would like to use a 6 point support. Is this possible/desirable for this size and thickness of mirror? I am planing a triangular cell made from aluminium square tube and 3 teeter bars.

Also, should I glue the mirror to the supports with silicone glue? Some people say that you must not do this as it will cause astigmatism, others say it is fine.

What is an acceptable PV and RMS error?

David

[Peter Drew]

I hardly dare say this but my 30" F4.1 13/8" thick mirror is "plopped" on a disc of industrial grade plastic bubbles attached to a substantial backplate, that's it!. Seems to work OK. 

[Moonshane]

A pv of 1/10 is better than 1/6

The figures are fractions of a wave as I understand it so more is less in this case.

As long as the mirror is evenly supported it should be fine. I dont like sticking them in bit others do.

I would say 18 point would be ideal for a 14" but my 12" has only three

[dark star]

If I use the automatic cell design in PLOP for a 6 point cell I get PV error of 1.27 and an RMS error of 2.86. For one part of the cell, which seems to be made up of 3 bars with 2 support points on each, I get figures of X of CG = 77.677, Y of CG = 44.847, radius of CG = 89.69 and angle of CG = 30.00. I am sure for someone any good at maths it will be obvious how this relates to the position of the parts of the cell in relation to my mirror. Unfortunately I am very bad at maths!

Are the above errors acceptable?

[ronin]

PLOP site used to have more information and they said that a 6 point mount was often better then a 9 or 12 point mount. However on a 350mm mirror and at 25mm thick I would have thought that a 12 point would perfrom better.Simply the diameter and thickness, if it had been 40mm thick then maybe a 6 point.

The 6 point is like a set of 3 bars each with 2 mount points as I recall, (?) try the same but swap the bar to a triangle with a mount point at each apex, thus a 9 point mounting.

[dark star]

I tried a 9 point cell in PLOP which gave me a PV error of 1.808 and an RMS errror of 4.18, which is worse than for a 6 point cell. An 18 point cell gives a PV error of 3.01 and RMS error of 5.43, which is even worse! Obviously this makes no sense so I must be doing something wrong.

David

[dark star]

Actually, the PV error you have to times by the power of minus 5 and the RMS times to the power of -6. Plus I mixed up my radius of curvature which is 3380mm with my focal length which is 1690mm ! The size of my secondary mirror is 83mm. 

David

[skybadger]

Have you let plop have free range with the variables -not constrained it by ticking a check box somewhere. I have seen designs where increasing the number of supports is worse but not often and only by constraing the radius of the supports.

Plop provides a graphical output of the support locations and a separate view of the bar/triangle - have you seen both of these ? With this info I typically draw them on a cad drawing. of of my own.

in my own recent builds the thing I forgot was to allow access to the mirror centering screw and mirror clamps. They got obscured by the trusses. Worth checking

Mike

[dark star]

I have managed to figure PLOP out by doing a lot of reading on Cloudy Nights-an American website.  I think I have managed to figure out the dimensions. In PLOP in parts dimensions the figures next to the parts are the actual size of the parts and the figures above are the position, for which I had to draw an xy graph and do a lot of reading! The only way to figure out the position of the parts is to use a CAD drawing program or draw them on graph paper like I did. I used a compass and a protractor.

I am going to go with a 6 point cell. There is a web site called Astro Bananas English a Photoscope which shows 3 teeter bars attached to a circlular piece of wood, I think this may be simple enough for me to make! it also shows nylon screws as side supports and mirror clips. I can always improve on the cell once I have made it. I need to finish the cell and upper cage so I can calculate the size and height of the mirror box. On the above website the mirror seems to be floating on the cell and not glued, PLOP only gives options for glue or a sling. If I findthe nylon screws give problems as edge supports I will try roller bearings for roller skates.

David

[ronin]

Just downloaded and played with PLOP with your dimensions.

From what I get the 6point is better then the 9 point (probably centre sag as only 3 around the centre) but the 18 point is better then either, the 27 point came out a bit better still but not by much.

For a 6 point I got:

P-V Error = 1.45e-5

RMS Error = 3.13e-6

For the 18 point I got:

P-V Error = 3.40e-6

RMS Error = 5.79e-7

Reasonable improvement on the 18 over the 6, but somewhat more difficult to construct.

For the record the 27 point gave:

P-V Error = 2.26e-6

RMS Error = 4.02e-7

Tiny bit better then the 18 point but not by much.

Also found this site, really more a quick check then anything:

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