If I'm going to see farther than others, it will be because I'll be standing on the shoulders of y'all

June 13, 2010 7:00 PM

Help me build a giant parabolic reflector planet-o-scope of SCIENCE!

I'm working on a plan to build a Newtonian reflector telescope, mainly to get high-resolution pictures of the planets. Since my plan involves a lot of different steps that I have little experience with, I thought I'd tap AskMeFi for advice/evaluation.

Here's the basic plan:

1. Make parabolic mirror (1.5 m focal length)
a) Get 24 inch cake pan
b) Borrow an electric pottery wheel
c) Obtain ~2 gallons of resin (not sure what type yet)
d) Pour resin in pan, then allow to dry as it spins at the proper speed on top of the pottery wheel.
e) Using a silver deposition procedure involving silver nitrate, sodium hydroxide, ammonia, and sugar, silver the parabolic surface (I'll probably allow it to deposit while spinning so the silver solution coats the whole parabolic surface)

2. Make aperture (barrel or huge PVC pipe, or something)
a) Mount cake pan with parabolic mirror inside aperture

3. Build secondary mirror assembly to deflect beam off the parabolic mirror's optical axis and to a lens assembly outside the main barrel of the telescope.

4. Obtain a series of convex and concave lenses to make a short-focal-length eyepiece lens. (combined focal length: 25 mm)

5. Mount SLR camera at eyepiece

6. Construct rotatable mount for telescope.

7. Install telescope mount on our roof deck in Cambridge, MA.


Does this seem plausible? Any suggestions for modifications or additions to the plan? I've never made a telescope before, so I feel like there are probably some common pitfalls that I'm unaware of. Thanks!

posted by Salvor Hardin to science & nature (18 answers total)

Fascinating. This rotational parabola, where have I heard of that? Wasn't it originally a pan of mercury? Won't the shape depend on viscosity, and in turn temperature, humidity, etc? Will you be able to monitor the shape while it's forming?

You are planning to let that form set, before applying the reflective coating, right?
posted by StickyCarpet at 7:40 PM on June 13, 2010

It would be extremely difficult to get the focal length correct, which would require controlling the rotation to a very precise speed. You'd likely end up with a parabola, but not necessarily the one you want.

Why are you trying to come up with a new approach for this when existing approaches work fine?
posted by Chocolate Pickle at 7:40 PM on June 13, 2010

Also, the seeing in Cambridge is going to dreadful. Between city lights and pollution, you're not going to get a very good view of anything no matter what kind of scope you have.
posted by Chocolate Pickle at 7:41 PM on June 13, 2010

You might be better off using some kind of plaster of paris on the potter's wheel and just brute forcing the proper parabolic shape. Then coat that. Or make it out of clay and fire it...?

But unless you are skilled or lucky, you are probably going to get a pretty lumpy lens. You'd probably get better luck taping an aluminized mylar sheet on top of a tube and hooking a vacuum behind it.
posted by gjc at 7:47 PM on June 13, 2010

Are you sure that the method you describe of making a parabolic mirror actually makes a parabola? I only ask because I tried to make a large parabolic mirror once and later on it turned out that the curve I was using as a model, a hanging chain, actually makes a catenary curve instead of a parabola. But Galileo made the same mistake so I don't feel so bad.
posted by XMLicious at 8:01 PM on June 13, 2010

This is an interesting question, and it's going to take me a little while to find the relevant literature, but for the moment here is a laundry list of my concerns, most serious first:

1. The surface of the resin is not going to be smooth to the level that you need. The mirror needs to be accurate to roughly the wavelength of light, and preferably to a tenth of that. That is, really, really smooth. I would be very surprised if any resin would set up suitably.

2. You're not going to be able to fix the surface of the resin much after it has set up. If it was a glass mirror, the final stages of polishing need to alter the shape by a handful of nanometers. I don't think you will be able to remove such fine amounts of material from your resin mirror.

3. Your mirror will probably deform in use as you tilt it away from the zenith. How much it deforms depends on the specific material.

4. You don't need an eyepiece if you just want to mount a camera. You don't even need a lens on the camera. The telescope produces a converging beam of whatever focal length you picked, and it'll form an image right on the camera's detector. If taking pictures is your primary goal, you need to pick the right focal length for your telescope such that it matches the pixel size and areal extent of your camera's detector. For example, say I have 4.5 micron pixels (aveage for a crop sensor), and I want that to be roughly 1 arcsecond on the sky because I'm in a city and don't expect good seeing. That gives me a 36.5 inch focal length, or a focal ratio of 1.5. That's pretty steep curvature. You say you want a 1.5m focal length, which is probably a little more reasonable, but f/2.5 might still be problematically fast.

5. I have no idea about your silvering compound.

Ok, that's the main stuff right now. I don't follow amateur astronomy, so they might have some ways to do this that I've never heard of. There are zero mirrors made of resin in the professional community. The only non-glass options that I've seen explored are carbon fiber, and there seems to be some hangups with that (the fibers make it difficult to get smooth). Fiberglass might be something to think about.
posted by kiltedtaco at 8:04 PM on June 13, 2010 [1 favorite]

You'd be better of getting a book like Build Your Own Telescope. It's a fantastic, down to earth book. Also, I have a video about building a large 16" Dobsonian telescope where Dobson himself is teaching. A 16" telescope is really big. Telescope optics are most forgiving at long focal lengths (like 8 times the diameter of the mirror or more) the shorter the ratio, the more exacting the mirror has to be made (like inaccuracies less than a wavelength of light. A coated resin mirror will slump to much as the telescope angle changes, it just wouldn't be rigid enough to maintain the required shape.


I've read about telescopes using spinning trays of mercury, but I'm sure the used something more precise than a potter's wheel. It could only view stars directly overhead.
posted by bonobothegreat at 8:16 PM on June 13, 2010

Dittoing kiltedtaco's first two points about getting the final shaping of the lens right. I ground an 8" parabolic mirror, intended for a Dobsonian, but was not able to figure it (remove the few nanometers needed to get the wavefronts aligned) because I'd used heat-treated glass and removing the most miniscule amount of glass changed the stresses in the mirror such that its shape changed. There's no way resin will work (in my opinion.)

Grinding a lens from a glass blank is quite satisfying, in a mindlessly repetitive way, but a 24" mirror would take a long time.
posted by anadem at 8:37 PM on June 13, 2010

Ok, quick look through the optics literature turns up nothing on pure-resin mirrors, but quite a few people working on carbon fiber mirrors with resin on top to prevent fiber print-through. These mirrors are all also formed by creating a convex mirror with the same shape, usually out of glass. Forming this convex piece is actually more challenging for amateurs than the concave.

The two room-temperature-cure resins I'm seeing used are Vantico RenGel and RenLam. Conveniently they give surface roughness values, 33 angstroms and 95 angstroms respectively. That's quite good, so maybe the roughness wouldn't be as much of a problem as I thought. There still remains the challenge of getting the material to the right shape though.
posted by kiltedtaco at 8:49 PM on June 13, 2010

Would a 24" light bucket really be helpful in getting good pictures of planets other than Uranus or Neptune (or Pluto if you count that)?

I'd think the big things for getting high resolution pictures of planets would be (1) Very high quality magnifying optics and (2) Good seeing conditions.
posted by ROU_Xenophobe at 8:55 PM on June 13, 2010

I think your plan won't work for the first few iterations, but the reasons why will be interesting enough that you should go ahead with it anyway.

I don't know what kind of velocity feedback there is on a typical potter's wheel, but you can help to stabilize the rotation rate just by adding mass to the wheel.

I wouldn't worry about not being able to control the focal length -- estimate what you expect and be prepared to adjust your design to be longer or shorter.

I would expect that any resin you can cobble together, unless you're a much more careful chemist than I am, will change its density as it cures. I'd expect that even if the liquid resin makes a nice parabola, the solidified resin will end up lumpy.
posted by fantabulous timewaster at 9:18 PM on June 13, 2010 [1 favorite]

Obligatory link to Sidewalk Astronomers and plans for building your own Dobsonian telescope. Been there done that!
posted by intermod at 9:43 PM on June 13, 2010

I actually tried this out of plaster and sorry to say that it's not very plausible/ feasible. Doesn't work when the material hardens at a differing rate. I think it would ALL have to slam to hardness, like Ice-9. Having worked with resin for years, I don't think it would fare much better. Also: resin is likely to expand and contract with heat/cold more than glass (which despite it's stability, still needs to be properly managed in terms of it's thermal expansion i.e. you need to let one cool down when you take it from your car out for a gazing party.

The bigger the mirror the more stable it needs to be. Such that even 1" thick glass mirrors have problems deforming under their own weight for the larger telescopes, giving astigmatism to the telescope unless it's been planned for.
posted by asavage at 10:48 PM on June 13, 2010

If you want to try a little experiment, get the cake pan and the resin and the silvering materials and do the whole thing without any rotation -- that should result in a flat mirror which you can use to optically judge the best-case scenario for what you could expect to get out of anything curved. I'm just guessing here but I'd imagine that it would be inferior to even the cheapest of bathroom glass mirrors even by just inspection with the naked eye.
posted by Rhomboid at 3:07 AM on June 14, 2010 [1 favorite]

Thanks for the input folks! A lot to think about.

It sounds like the consensus is that the biggest problem is that the resin won't form a smooth/uniform enough surface for my purposes. Any suggestions for how one might grind down a resin mirror after the resin sets? I don't have a clear idea on how you'd grind a mirror smooth while maintaining the parabolic shape.
posted by Salvor Hardin at 3:47 AM on June 14, 2010

Nthing all the mirror steps.

6. Construct rotatable mount for telescope.

If by "high-resolution" you mean "high-magnification" (i.e. lots of detail), then you left a lot out of this step. High-mag is going to mean than the object quickly rotates out of view. And I mean quickly. Even at relatively low magnification, you will be able to see the object moving through the field of view.

You need an equatorial mount. You can buy mount kits, but if you are making your own mirror you are probably planning to make your own mount and making a good one won't be trivial.

Plenty of people have built very large telescopes. You should check out the local star parties and see what others have done.
posted by DU at 5:09 AM on June 14, 2010

Read up a bit on making glass mirrors. In a nutshell, you rub two flat, circular glass "blanks" together with grinding powder in between them and eventually one becomes spherically concave (the other convex). The challenge is turning this perfect sphere into a parabola. The shorter the telescope, the more "parabolic" the shape has to be (the more it needs to depart from being a perfect sphere). Then you need to build a knife edge tester in order to see where the different zones of your mirror are reflecting light to and polish to correct errors. Using resin isn't going to help you bypass this callenge.

Also, you probably wouldn't want to use an SLR camera. Amateurs have been getting better shots of the planets since using digital cameras and Photoshop, which you can use to stack images and average out the rippling effects of the atmosphere.
posted by bonobothegreat at 8:56 AM on June 14, 2010

Amateurs have been getting better shots of the planets since using digital cameras and Photoshop, which you can use to stack images and average out the rippling effects of the atmosphere.

That's actually a really great point. With lucky imaging you could probably use a more reasonable-to-make-yourself 8-10" mirror and get the same results. Of course, this requires that you take many exposures, which in turn requires a better mount....
posted by DU at 9:07 AM on June 14, 2010