Primary mirror support cells

[memoryman]

Been checking on various primary mirror cell support ideas and didn't realise that the stock three bracket/clamp method is not adequate, or even particularly effective at preventing flexing and stress! :shock:

Apparently you can use 6, 9, 12 up to 52 point cell designs for really big mirrors, but that for most large relatively thin mirrors a 6 point design works at least as well as, if not even better than, a 9 point design! :notaclue:

http://www.google.co...bih=653#imgrc=_

One method even suggests using Astatic levers:

http://www.astrosurf...taz/astatic.htm

Has anyone any experience of modifying their scope, or of a friend doing it and was it successful?

[umadog]

Been checking on various primary mirror cell support ideas and didn't realise that the stock three bracket/clamp method is not adequate, or even particularly effective at preventing flexing and stress! :

It depends on the size and thickness of the mirror. For a lot of mirrors a simple cell should be quite adequate. I've not heard that the cells in mass-produced Newtonians are not adequate for the job. It's generally large mirrors that are problematic. The second link you list discusses mirrors of 400 mm diameter and up.

[astronymonkey]

Ditto what umadog says.

I have homemade 6inch dob with a mirror thats nearly an inch thick so the thickness means at this size flex wont be an issue. With big mirrors though the thickness relative to the diameter drops in order to have a mirror of a manageable weight and that wont take for ever to cool down. Many commercial 16inch scopes have 18 point cells to spread the weight and provide support but on most amateur scopes, if reasonably well made, the viewing limits will be set by the quality of the optics themselves along with the seeing at the time, rather than a poorly supported mirror.

When supports do cause problems they tend to be due to the mirror not being adequately fixed in the cell so it can move slightly when the scope is tilted between the horizon and the zenith , or the cell is tightened so much around the mirror in that it can pinch the optics and cause some distortion.

Cheers

[Peter Drew]

My 17" 1.5" thick plate glass mirror sits on a disc of heavy duty bubble plastic material, I've never noticed any support problems in over 20 years use.

[umadog]

My 17" 1.5" thick plate glass mirror sits on a disc of heavy duty bubble plastic material, I've never noticed any support problems in over 20 years use.

Do you not have cooling issues, though?

[cs1cjc]

I understand that many have used bubble plastic with success, but it is not really possible to analyse the error induced by such a support. If you want to see how good (or bad) a support is then PLOP ( http://www.davidlewistoronto.com/plop/ ) is an excellent tool. It is simplest to use the automated design feature.

In passing, according to PLOP, a 17" plate glass mirror, 1.5" thick, focal length 2000mm (guess) and secondary 75mm (guess) on three points only at the optimum radius of 0.43 of mirror would have an induced surface error on the high side at RMS 2.4e-5mm. A simple 6-point support (three circumferential bars supported at centre with mirror supports at the end, equally spaced at 0.59 of the mirror radius) would reduced the induced error to a very acceptable RMS 3.35e-6mm.

[starman345]

But why would there be a need to analyse the error if the owner was happy with how it worked and could see no problems at the eyepiece? I've built a cell with PLOP and it works ok, but it certainly is not the only way. There are a lot of nice telescopes out there giving great views that were built long before PLOP was around.

I'm curious what this "heavy duty" bubble plastic is that Peter mentions? I'm guessing this isn't the regular bubble wrap used in packing?

[Peter Drew]

Regular bubble wrap does work but it gradually deflates, the material I used was intended for covering swimming pools, it's heavy duty and the 10mm bubbles cannot be popped by finger pressure. The circular pad works as an air bed with hundreds of support points and air can flow freely round each, amateurs in the USA have also used foam backed carpet as an alternative. I stress this is for visual use, no guarantee for imaging. My 30" F4.1 1.5" thick plate glass mirror also sits happily on the same material, it does however have a 3" central aperture to locate it on a rear cell boss and a sling to carry the weight at low altitudes. This system was in use long before "PLOP" and works for me.

[umadog]

But why would there be a need to analyse the error if the owner was happy with how it worked and could see no problems at the eyepiece? I've built a cell with PLOP and it works ok, but it certainly is not the only way. There are a lot of nice telescopes out there giving great views that were built long before PLOP was around.

Well, PLOP just calculates the optimum support doesn't it? I imagine you can get near-optimum without it just through rules of thumb. Perhaps the need for PLOP has increased recently since large, thin, mirrors are now more common.

[starman345]

Well, PLOP just calculates the optimum support doesn't it? I imagine you can get near-optimum without it just through rules of thumb. Perhaps the need for PLOP has increased recently since large, thin, mirrors are now more common.

PLOP calculates optimum support for a particular mirror but only for a certain set number of support points, 3,6,9,12,18 etc. and tells you which would be more suitable, a great design tool for sure. But, who is to say 100, 200, or 500 support points would not be just as good or better, I think a big problem as Chris C says above is the fact you can't analyse it..... so you don't really know for sure until you look through the eyepiece.

I think for a first time cell builder PLOP is the way to go because it will get him in the ballpark.

[cs1cjc]

Well, PLOP just calculates the optimum support doesn't it? I imagine you can get near-optimum without it just through rules of thumb. Perhaps the need for PLOP has increased recently since large, thin, mirrors are now more common.

Except that the results are different from the old rules of thumb and in some cases significantly different!

Three points at about 0.4 of the radius for my 200mm, 20mm thick mirror gives an induced error of 40% that of the old rule of thumb, at 0.7 of the radius. Moving the supports was a trivial upgrade, much cheaper than a new complex cell.

Larger thinner mirrors can be supported on three points than was thought before finite element analysis.

One still sees the equal force 9-point design which is actually worse than the simpler 6-point support. The best 9-point support as used by Orion UK only a little better than a 6-point and is not so easy for an amateur to make.

However it is also true that traditional, thick mirrors are often supported adequately by supports designed by rule of thumb.

[Peter Drew]

I've made Newtonians from 10"-17" aperture with perforated primarys that had been profiled similar to those in SCT's and mounted in the same manner, no radial support, lighter weight and faster cool down.

[memoryman]

The mirror on my scope is made of Pyrex and has a diameter of 254mm and is 25mm thick. Has anyone made a cell support structure that would be relevant and would be willing to let me have the measurements and recommended materials, so I can have a go at constructing one?

[cs1cjc]

*Deleted*

[cs1cjc]

I have guessed F/6 and a secondary of 60mm. If you confirm the correct sizes I can run the calculation again, but it will probably make little difference to the result. Plop says supporting the mirror on just three points, at 0.41 of the mirror radius will give an induced RMS error of 5.8e-6mm. This is marginal, with the author of Plop suggesting an RMS maximum of 4.3e-6mm.

The next simplest support is a 6-point support with three bars pivotted in the middle with a supports at each. The six mirror supports are arranged to be equally spaced around a circumference at a distance of 0.59 of the mirror radius from the centre, so for your mirror around 150mm from the centre (the exact distance is not really critical). This arrangement gives an induced RMS error of only 1.01e-6mm so anything more complex is overkill! I attach an image of the output from Plop for this.

[Peter Drew]

I would try a disc of bubble pack or foam backed carpet first, it will cost next to nothing in time or materials. If it doesn't work for you then there's always the more sophisticated approach.

[1parsec]

As mentioned already, PLOP is a great tool.

I used it for the design of the 9 point support cell for my Newtonian.

Without using PLOP I wouldn't have had a clue where to put the supports points !

I plugged the dimensions from the PLOP output in to my drawings and went with that.The end result seem to be spot on.

Dave.

[cs1cjc]

Dave,

That is a lovely piece of work!

[cs1cjc]

...The next simplest support is a 6-point support...

No! A 4-point support is also possible. The mirror supports are arranged at the corners of a square frame, in this case at 0.54 of the mirror radius from the centre. This frame in turn is supported on three points, two of which are on one side, in line with the mirror supports and the third is is the middle of the opposite side. The frame should have enough flex so that the force on each of the four mirror support points is equal. The frame supports can designed to be used for collimation. This support design is adequate for this mirror, with an induced RMS error of 2.7e-6mm. Plop output attached.

[memoryman]

Well I finally plucked up courage and took the advice of a more experienced astronomer than I, namely Michael Herman (California USA) advice and designed and constructed a 6pt PLOP optimised floating support cell for the mirror. I fitted and tested it three weeks ago and gave it a good work out at an observing session at Rosliston Forestry Centre on Friday night 12th July. It passed with flying colours! In fact the other three observers were so impressed with the scopes performance that they judged it have the two top views of the evening, one of which was M13 through a Williams Optics SPL 3mm eyepiece. Totally against expectations in less than perfect viewing conditions, was the detail it revealed, which was amazing!

It's so fantastic to see pinpoint sharp, bright, stars instead of flaring etc!

The project took a total of 5 weeks overall and included flocking the OTA, cleaning the corrector lens/plate and secondary, fitting a new Revelation Astro Crayford 10:1 focuser with Takahashi 75mm extender, replacing the adjuster spring on the secondary with a compression spring OUTSIDE the secondary holder as recommended again by Michael Herman), cleaning and respotting the primary mirror with a Farpoint red triangle and collimating the entire optical path. I also blacked the bevel edges of both the secondary and primary mirrors with a Sharpie pen (the primary was relatively easy and only took 2-3minutes, but the secondary was very fiddly and took ages!

There were plenty of sweaty, nervous and frustrating moments when I couldn't quite figure something out! I really enjoyed the project, although it was quite frustrating at times as I tried to be very precise with measurements and angles etc! There were also many sweaty moments of worry when something wasn't working out! Not to mention waking up at 4am with an idea and having to go downstairs to write it up or try it out!!

But I really enjoyed the challenge and innovating using materials in our garage and having to change tack according to how things were working out.

The pivot parts method came to me as I was looking through various nuts & bolts. I saw a *screw chrome ball combo and immediately thought, "That looks like a perfect ball joint! Now how can I make this work?".

In the end I had to rebate the wood to allow the ball to sit low, so the gap between the aluminium plate and the wooden base was just sufficient to allow pivoting, but without pushing the mirror any more forward than necessary (it sits 9mm higher than I presume the mirror sat when Meade assembled it originally and 6mm higher than its clipectomy days! I worried that the PLOP data would be compromised by this fact as I'd used the standard focal length figure of 1016mm. So I ran the program again using 1007mm focal length figure and it was different by 1/100mm, so no real need to worry as even I'm not sad enough to be that precise!!

The screw connection on the ball allowed for precise upward & downward adjustment of the plate, so I could get the mirror sitting level on all 6 pads with equal pressure. I used two 1.5mm drill bits to fine tune the gap and 2 x 2mm nail heads (cut off and filed smooth) to restrict lateral movement of the plate without compromising pivot efficiency. It works beautifully and I'm really proud of myself for persevering and making it work.

I now much better understand how the scope works and have gained a massive amount of confidence overall.

I have thought about building another floating support cell using a different method and materials in the future, which will sit lower and be closer to its designed focal length.

P.S. Whilst attempting to remove the mirror from its clipectomied prison, I forgot to mark the primary mirrors orientation on the mirror assembly and before I realised, I’d moved everything around so couldn’t be sure where it had fitted. DOH !

So when I put the completed mounting back on the scope I decided to check the performance by rotating the mounting by 90 degrees clockwise, collimate then as far as possible replicate an identical viewing **cycle for each rotation to see if there were significant differences in image quality.

I’ve completed the testing in all 4 positions and haven’t noticed any real difference in views and certainly no bad effects, so I'm not so sure that mirror, secondary, corrector lens/plate alignment is as critical as people suggest (at least with this Meade scopes optics).

I'm not sure if you can post images on this site? If I can, I'll post relevant images of construction for you all to see.

*It was a mirror screw with the chrome ball as an aesthetic touch. I drilled through the aluminium bar with a 5.5mm metal drill bit and countersunk the hole. Then, to avoid reaction between the chrome and alumnium I made three bearings for each hole out of milk bottle plastic (superglued in place), which was both strong and slippery.