Thread: Columnating light from a non point-light source??

I want to be able to take light from a light pipe, maybe about 2 cm diameter, and columnate it into a beam that can be sent over distance.
ie, non imaging optics.

How can this be done?

Will it work to use a long cone to 'pointize' the light from the pipe, so that it can then be put through standard lenses, and how long would the cone need to be?

You will not be able to do this very efficiently. You can re-image the exit of your light pipe anywhere you want but you will be losing a lot of light.

At the exit of your light pipe imaging a grid. In each grid element the light rays are all going in random directions. The result of this is that to your eye the exit of the light pipe looks just as bright when you look at it from the side at you do near straight on. This is oblivious not true for a laser source, from the side you can only see a little scatter but if you look straight at it (Do not do this) it will look blindingly bright.

Also keep in mind there really is not such thing as collimated light. A beam of light can only be consider to be collimated for a finite distance. For a Gaussian laser beam, which is the type of light that can be collimated for the longest distance one has the following relationship:

Where is half the possible collimation distance and is the minimum width of the beam and is the wavelength of light.

If I known more about your optical system and what you wanted to do I may be able to help more.

Could I not just place that at the focus of a parabolic reflector and ceate a beam?

It doesn't need to be perfectly columnated, but the more the better. It's for a signalling device fed off a light pipe. A laser would obviously be easier and better, but for this application it unfortunately needs to be white light.

Yes if this is just for and electric eye type set up then yes a parabolic mirror would work but those are hard to come by. What is the distance between the light source and the collector?

I'm going to be taking a crack at making parabolic reflectors with mylar (reflective plastic foil). Hopefully I can get them precise enough.
Or else there must be some little thing I can pull to pieces and get a pretty decent parabola out of?

The legnth of the light pipe, or the distance from the initial liht source to it's start?
The pipe will be about 1.5 meters (5 feet), the light to the pipe will vary, but I don´t think I'll have any problems there.

I'm hoping to get a mile or so usable range on the beam, tho I'm probably dreaming.

OK I'm beginning to picture what you want to do. I would suggest that you make large in diameter and as sallow reflector as you can. A parabolic shape is hard to make so I would suggest making it spherical. Maybe you can mold your mylar over a large ball. Then place your light pipe facing the mirror at about half the radius of the ball. You will need to slightly adjust the location of the light pipe to maximize it effect for different distances.

Be sure to post your results.

I think I'd still try to use parabolic. I think you can make a parabolic surface by letting gravity work for you. I'm not 100% sure of that though, and you'd have to use a secondary flat mirror to aim it.

Gravity will make a Catenary which is a hyperbolic cos for you. This will produce more aberrations than a sallow spherical shape. Using a sallow spherical surface is the standard and easy thing to do.

Ah. I remembered wrong then. I thought there was a fairly easy way to get a parabola, but I can't remember what it was.

Anyone have any idea how to model a Dall-Kirkham type reflection setup. ie Ellipsoid and sphere. Parabola - parabola and parabola - hyperbola are easy enough, but I can't get the Dall-Kirkham to work.

I'm modeling reflections in rhino3d.

I'm also simming some conical setups, but it doesn't look like anything more than a 5:1 reduction is likely. Not sure that will be enough.

Ah. That's what it was. A rotating pan of liquid forms a parabola. Oh well. That's probably not practical. (I guess you could always try it with wax to get the shape down though.)

Edit: Actually, wax wouldn't be good. It changes volume too much as it heats/cools. I don't know what material would be good for such a project. Most actual liquid mirror telescopes leave the mirror liquid.

What about a rotating stage with a bucket of water in the freezer?

I'm not real sure. Due to surface tension and the expansion of water as it freezes, I don't think you'd get a perfect parabola, but it may be close enough.

If you could find a material that changed volume very little as is melted/froze (or if you could control the temperature precisely enough to minimize the effect), you could (probably) make a parabolic dish out of it, then use that as a mold to make a more permanent mirror.