Thats rather awesome. Remembering stuff like manually polishing telescope lenses was what made me wonder if we couldn't just 3D print the basic shape of these lenses in future, then put them through a final 'polish' like that, to achieve lenses of the correct shape in a much quicker time than making them from scratch from blanks...

Looking forward to seeing how the mirrors turn out!

Nah, making them from blanks will almost certainly be easier. 3D printing is actually really, really slow, especially for finer objects.

Probably a stupid idea, but what about cutting the basic lens shape from a 3D CNC mill and polishing it down that way?

I imagine casting them is still WAY easier, but using a CNC, you make design modifications quickly and without needing to fabricate new molds.

The issue with 3 axis mills is you always end up with the tool profile in the final piece, the usual way you get "smooth" shapes on a 3 Axis mill is to rough it out then use a small ball nosed cutter with a very small over lap to produce the final profile. That would not be accurate enough for the outside of a lens.
I guess you could likely do a final polish, to get the effect but I have to wonder if the mill would actually save you any time.

You probably be better off using some sort of single point cutter to rough the shape i.e. a lathe, but again your still going to need the final polish pass.

Yeah, that's a good point. A lathe isn't a bad idea, though. How are prototype lenses typically fabricated?

Thanks for your comments,

Cnc is ideal but in 2003 when I was a teenager, there was no money for cnc and no 3D printer.
I could only use some hobby materials and my brains.

Material list
the casting resin for lenses: 12 euro
diy spinning lens holder (fibreboard): 15 euro
diy vacuum forming pump and ring hole (1mm aluminum): 15 euro
polishing and sandpaper (p40 to p1000): 25 euro
silicone mold (standard bathroom silicone) : 5 euro
+ drill from my parents.

Fabrication of the resin casted lens (March 2003):
1. The lens material index is calculated, then the curvature of the lens surface is drawn on paper (diameter is very important).
2. Vacuumforming by blowing surface into a hemisphere with custom diameter ring hole.
2 (note) The diameter of the lens itsself and the diameter of the vacuum formed hemisphere are custom.
How smaller the hemisphere, how stronger the lens.
3. make a cylindrical shape and glue it WITHOUT GAPS on the inner side (for convex) or outher side(for concave) of the hemisphere.
4. Take the glued hemisphere and cylinder apart from the rest of the hemisphere around the cylinder by cutting (with siccors).
5. Put it into a mold like a ship (not over the open surface of the cylinder). you can put weights for maintain it right.
6. If the mold is dry, you can remove the 'ship'.
7. Poor the casting resin.
8. Let it dry.
9. remove it from the mold and grind the upper surface.
10. Fix it on a drillshaft (in the center) and let it spin.
11. Grind it with a sandpaper fixed on a metal shape with the same curvature as the lens.
12 polish in the same way.
13 remove from shaft and grind/polish the flat surface on a horizontal grinding/polishing disc.

I'll send an illustration because it is complicated that way .

The Focal length is the most difficult part

Yes, I know. I was responding to Okta's idea of projecting the microdisplay's image onto a flat white curved surface instead of a mirror, and that it would not work unless you added another lens anyway.

I have previously mentioned the idea of using the flexible OLED displays for a Head Mounted Display. This would hopefully solve the issue of light on one portion of a projected image from somewhat washing out other portions of the screen. However, this technology has not been coming as fast as hoped. The writer in the article below says Samsung's late 2012 launch has been put off to perhaps early 2013. And even then, the mass market being perhaps delayed to 2014 or later.

Of course we have CES 2013 coming up in Jan. Perhaps there will be some surprises.

http://www.oled-info.com/thoughts-about ... -plausible

Joe Dunfee

I just checked the pictures again I see how it works now, i would be looking into a mirror with the image set at the correct focal length.

3dpmaster: when you say focal length is the biggest problem, do you mean the interfaces on areas of all the reflectors matching up or over all?

The Issue is the focal plane. Some beams from the observer's eye (set to infinity) have a larger focal length than other beams also set to infinity. With the technique of Mr. Hajime N., most of the rays are inverted to +/- infinity before they reach the collimated lens.

The idea to set two micro projectors on the top of the head to project the image to a curved screen is not bad. If we use a pair of reading eye glasses, we can bring the curved screen closer to the observer's eye. The best screen to use, is a toroidal curved screen. The projector lens has to be FISHEYE to get a wide projection at short distance. I have to try this technique, I hope It will works also.

Correction, the focal POINTS are not on a planar surface. Image below.