Has anyone ever tried......

Has anyone ever tried to cut down the diameter of the lens on a LS/O (Luxeon Star with optic)?
I want to get a collimated beam but put the LS/O into a small diameter package. I was inspired by Hotfoot's Pencellon (penicillin?).
It would be great to have an Arc AAA-sized package with a LS/O but you have to make the lens diameter smaller.
Any alternative optics (besides the plastic ball)Anybody? Anybody? Bueller? CNC Dan?----Marc

I did that here:

http://www.candlepowerforums.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=14;t=000582

To fit the optics into a Brinkmann 1AAA flashlight. The Brinkmann head is only about 17mm in diameter, whereas the optic is about 20.

-Darin

Evan---Thanks for pointing that out. i had missed your "cram the optics into the head" statement when I read it.
I was looking to go just a wee bit smaller. I was hoping to get a lens that is 8mm or less in diameter and reduce the whole thing to true ArcAAA size in all dimensions.
Two questions:
1.) How did you "cram the optics" into the head? Did you use a lathe to turn down the lens or its black plastic mount?
2.) How tall is the stack of 1/3AAA's? It looks like the height of three is more than the height of one full AAA.
The A23 combined with reduced diameter optics and a tact switch tailswitch should yield an LS that is bright and the size of a Arc AAA that is quite useful.---Marc

Originally posted by MicroE:
Evan---Thanks for pointing that out. i had missed your "cram the optics into the head" statement when I read it.
I was looking to go just a wee bit smaller. I was hoping to get a lens that is 8mm or less in diameter and reduce the whole thing to true ArcAAA size in all dimensions.
Two questions:
1.) How did you "cram the optics" into the head? Did you use a lathe to turn down the lens or its black plastic mount?
2.) How tall is the stack of 1/3AAA's? It looks like the height of three is more than the height of one full AAA.
The A23 combined with reduced diameter optics and a tact switch tailswitch should yield an LS that is bright and the size of a Arc AAA that is quite useful.---Marc

Click to expand...

<font size="2" face="Verdana, Arial">Marc,
Going down to 8mm may be detrimental to the function of the optic. IIRC, the forward light from the emitter is collimated through refraction (and out the center "nub" of the optic), and the off-axis light collimated through internal reflection of the optic (along the outside edge). By sanding that down, you reduce its effectiveness in both a) making a rough surface that doesn't reflect internally (scatters), and b) modifying the angle of the light that hits the new surface. A) can be solved by fine sanding and polishing of the surface, but B) can't really be - it will modify the beam pattern.

To shape the optics, I used a hand file, and some eyeballing (note that I didn't show a picture of the optic ). I ground about 2-3mm off the diameter, just at the largest point, then screwed the head in over the optic. I had to back off and tighten several times to make sure it's centered.

The stack of 3 1/3AAA's is EXACTLY the same length as 1 AAA. Some mfgrs cross-label between 1/3 and 1/4 - the part you're looking for should be 14mm in height.

The A23 only has a capacity of about 40mAh (i think? maybe 80). The 1/3AAA are labeled at 100, but are probably like 120. They can feed the luxeon 400mA for about 2 minutes (direct drive). I resistored it down with 2.7 Ohms, putting it at about 150mA. This way, it lasts at full brightness for about 20-25 minutes.

So there are downsides to the 1/3AAA - charging is a pain, and so is getting ahold of them.

However, after all that, it is SO worth the work to get a luxeon flashlight that small . While it was direct drive, it outshone my roommate's 3D maglight (it's not the size of the light.... )

Good luck!
-Darin

Darin---You are absolutely right. I had not considered the total internal reflection aspect of the optic.
I spent some time examining the lens from a LS/O and realized that it is basically a short focal length plano-convex lens with a large plastic funnel.
I would bet that the designers are using the internal reflections of the funnel in the same way that people on CPF use write-right. It's a beam "smoother".
My understanding of total internal reflection is that the amount of light internally reflected is proportional to the difference of indecies of refraction of the two materials. Since the materials here are plastic and air it seems that there is very little light reflected back into the funnel (only the really oblique rays would get reflected, right?).
If the funnel contained an inner cone of low index material and an outer cone of higher index material it would be much more efficient (but much more expensive). I wonder how much light is lost from the surface of the funnel.
I would expect that you could get a highly collimated beam by replacing the optic with a small diameter convex lens so that the LED would be just outside its focal point. This could probably achieve 90% of the efficiency of the current optic, but would not be quite as collimated. I'm willing to give up a little collimation in order to make a AAA-sized Luxeon. Now all I need is a really small diameter Hastings Triplet. Off to the junk box I go!
Sorry for the rambling. I'm just thinking "out loud".
What do you think?---Marc

Well, I think that the internal reflection of the stock optic is actually pretty good.

I'll show you what I mean (going to try some picture magic here)

Bare emitter

Optics resting on emitter


I locked the exposure on my camera. I don't think we can measure from the pictures, but we can get a pretty good idea about how effective the optics are at reflecting the emitter light SOMEWHERE besides the direction they were going before..

One thing you might try is a small acrylic ball - that's something that many members here have used in leu of the standard optic. However, they often get an exact projection of the emitter die rather than a nicer circular beam. With the white emitters, it's not so bad, since the phosphor coating fuzzes the image of the emitter die and gives a smoother output.

Did you see the other Arc-AAA size mod from last week? He used an acrylic ball as a lens - and ended up with a very similar beam shape to the
original optic.

http://www.candlepowerforums.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=14;t=000622

The other thing is that with a 110 degree output angle from the low-dome emitters, you would need an extremely short focal length lens to gather all of the light - and it would probably have to be touching the emitter dome. For the other mod I linked to, it looked like he drilled a small hole in the ball, in which the emitter fits.

Anyways enough rambling - It would be neat to see different experiments with optic systems for the luxeons - especially a very tightly collimated beam.

-Darin