If you had, say, light from a torch or whatever, directed it through a metal pipe or fiber optic to get it kind of directional, and then into some arrangement of lenses; what kind of maximum range could you expect before it completely lost focus?
And would it still obey the inverse square law, or is that just for expanding spheres of light?
I have a little 1W LED flashlight and a surplus copier lens (about 4" diameter). Pairing them gave me a beam I could see reflected off of rock unaided over 2 miles away against a mountainside. Using a small night vision monocular I could see the beam was (just guessing) about 50' wide at that point. That's when I discovered the light-emitting portion of the LED was square in shape.
To maximise your ability to produce a focussed beam, you need your light source to be as small as possible. Incandescent lamps are poor for this, because the filament is spread out over a relatively large area.
Using a light pipe [such as fiber optic] won't help to collimate your beam, total internal reflection means that as the angle of incidence == the angle of reflection, the light entering the pipe at random angles will exit just as randomly.
Lasers can produce a close to point source effect because they selectively amplify light along the axis of their resonant cavity, ensuring that the light that does exit is highly parallel - semiconductor lasers are not as good at this as gas lasers because their resonant cavity is very short.
If I had the light coming out of a tube, maybe a couple centimeters diameter, and spreading out, is there a series of lenses I could use to get it back to a point or column?
Like Stokes said - a Fresnel may be best, even a letter-sized sheet will focus pretty well. I do like the compact nature of the copier lens though. With that I just held it up against the flashlight about 2-4 inches away to see what it could do. Maybe I'll try a Fresnel later tonight - I have a letter sized one somewhere.
I had planned to use the copier lenses for a point-point optical data link but never got around to it. Come to think of it, that also why I bought the letter-sized Fresnel lens..
I was thinking of some kind of signaling device, really it's just curiosity to see what can be done. A laser pointer would be the obvious solution but the battery life can be pretty limited. And it's less fun. I'd prefer a solid lens solution as the whole thing would be more compact.
registrationsucks2 sez: "If I had the light coming out of a tube, maybe a couple centimeters diameter, and spreading out, is there a series of lenses I could use to get it back to a point or column?"
You can focus it to a point, but there's nothing you can do that will fully collimate the beam, i.e. it will always diverge.
BTW a laser pointer can easily be driven with an external power supply.
A bit of divergence is ok, as long as it's manageable. 50' over 2 miles I can live with. But I need to get focused to a point first. I'll do some research on optics, see what lenses will do it.
You need a "point source: light emitter. Lenses can't change the shape of the emitter, or cause a wide emitter to send out a narrow beam. Too bad, because if they could, then a ten watt flashlight bulb could be made just as dangerous as a ten watt laser. Or use big lenses to convert a square meter of sunlight into a thin kilowatt beam.
The usual way of producing narrow intense beams of white light is to use lenses on a xenon arc lamp with a very tiny arc. Those "DJ dance floor lasers" which aren't lasers? Those use a xenon arc point source.
On the other hand, search on non-imaging optics. Melt-stretched fiber optic cones, trumpet-shaped mirrors, etc., can shrink a light source by destroying the image information.
you may be able to experiment with the optics in a toy "light gun" (video games, laser tag) The video game light guns use a sensor to pick out a point only a few pixels around (because CRTs scan.) The laser tag guns send an IR code, like a TV remote, in a narrow beam to the receiver on the target.
You could try to replace the IR LED or sensor with a LED.
a more expensive way to go may be to get a professional ellipsoidal spotlight, and use an iris, steel gobo, or glass gobo to give you a fine point. (and no, I don't mean a big follow spot) google "ETC Source Four" and you'll get the idea.
What about using a long cone to 'pointize' the light coming out of the pipe? If the wide end of the cone was 2cm and the narrow 1 or 2 mm, how long would the cone have to be?
The problem wth anything based around an LED, or ellipsoid, is they're assuming you're using a point light source.
The 1W Luxeon LED die is a square, so the beam is square in shape. The copier lens has an adjustable focus which makes it easier to get a tight beam at various distances.
At 40' it produces a square about 14" wide. Extrapolated out, that's 132' at 1 mile, so my guess of 50' at 2 miles may have been a ways off. With the variable focus it might not be as simple as that,, and you might end up wider or narrower. Overall I'm impressed with the very narrow beam. It's easily as bright as any handheld spotlight I've used, but the beam is so narrow it would be useless for anything but a signaling device or data transmission.