Adjustable beam focus for lights...

Is it feasible to have adjustable beam focus for high-end lights like SFs and the like? Has anyone tried and succeeded (or failed?). I think if a light could have adjustable throw, that would be one step closer to the ultimate light.

It's not easy. Most lights use a reflector or collamator to generate a hotspot. Unfortunately they only work when the LED die is at the focal point. When the LED is off by a fraction of milimeter, you end up with a donut beam /ubbthreads/images/graemlins/frown.gif

Ack, I've NEVER seen ANY adjustable focus flashlight worth a damm. I fiddle with it till I get the least objectionable beam pattern on a wall and then might as well glue it in place.

great concept, poor execution

I suspect that when multiple LED's come in and people get used to non circles for a flashlight different methods will come into play.
If you have say three LEDS in a line and with parallel beams, you could flex the material they are on to give flood or moe focused light.

This sort of hits on a question I've had. My understanding is that the new L1 has a tight focus. How well/poorly does a beamshaper work on it?
TIA

it is just not really possible to get a good beam in both settings with a focusable reflector
if you use a faceted reflector and unfocus it (to get a flood) you get a donut hole and the spill shows the facets as ugly artefacts
the only way i see would be to use a flexible reflector, that is able to change its actual focal length /ubbthreads/images/graemlins/grin.gif
and now make that heat resistant ...

Beamshapers work, but adjustable (i.e. defocusing, like Mag, etc,...always seems to stink and have a big donut hole).

Maxabeam has a nice solution, with their partial diffusion filter, you get a decent spot, and then with defocusing, you get a perfect flood (but again, the in-between has a donut hole). I may have to try this with a Mag, since it has a relatively tight focus.

Changing focal lengths will still give you a spot, just a little more or less sidespill. Now, the way to really do it, would be for the reflector to change from smooth to orange peel /ubbthreads/images/graemlins/grin.gif

McGizmo had an interesting concept not too long ago - use a small primary reflector that travels with the light source and fits inside a larger secondary reflector with the same curve.

Here are some sketches I threw together:


Flood: light from primary reflector mostly not in contact with secondary


Spot: primary+secondary essentially behave like single reflector

Disclaimer: Not to scale, reflector curves are not parabolic (far easier to sketch an elipse), original concept by McGizmo, etc...

Wouldn't that "donut" too? Nice idea too, I was thinking the interface of the two reflectors would be "zig-zagged", but I'm sure McGizmo has even better ideas.

I think you would have to do something like use a paraboloid with an annular focus, if that makes any sense.

It depends upon why there is a donut in the first place.

I don't know much about LEDs, but it sounds to me like the emitting patterns are disagreeable for reflector focus in the first place. You would have to graph emitted light intensity as a function of angle from the direction of pointing, i.e., the forward direction. I'll bet you don't have equal amounts of light coming off in every direction, so some parts of the reflector bounce a lot more light than other parts.

If that is the case, it might work to first shape the output into something that can be focused. This might mean placing a reflector around the LED bouncing light backward to a bigger reflector, then out. You would shape the beam with the first reflector so that it was ammenable to focusing with a larger reflector. Even out that emitted pattern before you try to focus it.

That's my $0.02.

Can anybody plot light output versus angle from forward for just the bare LED? I think that would be "illuminating" /ubbthreads/images/graemlins/wink.gif

Scott

[ QUOTE ]
beezaur said:
I don't know much about LEDs, but it sounds to me like the emitting patterns are disagreeable for reflector focus in the first place. You would have to graph emitted light intensity as a function of angle from the direction of pointing, i.e., the forward direction. I'll bet you don't have equal amounts of light coming off in every direction, so some parts of the reflector bounce a lot more light than other parts.

[/ QUOTE ]
LEDs behave much like point-source emitters. Almost all have beam angles of 180 degrees or less.

3mm/5mm LEDs ("commodity" LEDs) are typically focused to some extent by the reflector cup leadframe and package. These do not work so well with reflectors because they're already focused and the emission point of "spill" light varies due to internal reflections in the package.

Luxeon (and most other high-power emitters) feature wide beam angles slightly greater than 180 degrees. They work very well with reflectors designed for point sources.

Almost all LED spec sheets feature emission patterns with relative intensity graphed by angle.

Another way to approach this would be to use multistage optics. Capture the light from the emitter with a reflector, then set the beam angle with a zoom lens. Weight, cost and size would be issues, but the beam quality could be extremely good.

Some very good optics could be made if extremely narrow angle LEDs internal reflectors hit the market. Externally focusing such an anisotropic area source as an LED is a real kludge.

There is a company that does just that, but their lights are rather costly by comparison to most others. This company (name escapes me) makes film equipment like camera cranes and other stuff used in film production. I hear the beam quality is top-notch.