Since LEDs are more efficient at lower power, we should be able to run two luxeons at, say, 200ma, and have the total light output be greater than a single emitter at 400ma. Multiple emitters make a good floodlight, but getting a tight beam is extremely difficult if not impossible. The 5-watt luxeon was a good attempt at implementing multiple emitters while retaining the possibility of photon management, but I'm wondering if it can be improved on.
Reflectors are the currently most-favored method for creating a long-throw light, so the question is how to get two (or more) emitters working together in a single reflector (mutliple reflectors are too bulky and require alignment, etc). Here's what I came up with:
Two emitters pointed directly at each other, one in the standard position at the base of the reflector, and one in the "recoil" position at the mouth of the reflector. Both emitters would be wired in parallel with a single driver, each getting half of the driver's rated output. Total output should exceed that of a single emitter at the full output.
My first thought is that the emitter at the mouth will block a signifcant amount of light from the base emitter. To check that, I just set a McLux PR up on its tail and placed a loose emitter on the lens. Casual observation revealed no noticable difference unless the beam was aimed at a target less than 6 inches away, at which point I begin seeing a hole in the beam anyway with the single-emitter setup. No doubt a light meter could detect some difference, though. The question is this: will the amount of light from each emitter that is absorbed by the other emitter be more or less than the extra light generated by the greater efficiencies at the lower power levels?
Your reflector will give you a problem. In the above, only one of the luxeons will be at the focal point of the reflector. The other will merely serve to add flood light to your beam (if that's what you're after, then great), but a single reflector won't be able to focus the light from two distinct points.
Evan is absolutely right. Multiple reflectors is the only way to go. Note that the smallest overall diameter (the diameter of a circle that can contain ALL the reflector) is 2.155d (d being the diameter of each reflector) but that for 7 reflectors this increase to a mere 3.00d. For maximum packing (maximum number of round reflector that fits into a circle) the numbers are 3,7,19,37 (anyone want more? [img]/ubbthreads/images/graemlins/grin.gif[/img]). So a 7 Luxeon III light will have a 81mm (3.2") diameter using the 27mm Fraen optics. the beam should still be s single spot at a respectable distance. If the circles are cut slightly as they have done for their 3 optics cluster, then the diameter can be even smaller. Assuming about 1500 Cd from each optics, 7 of them will give about 10,000 Cd. The 2 degree optics may give 40,000Cd (90,000 possible). How to get rid of so much heat in such a small apace is a more difficult problem. [img]/ubbthreads/images/graemlins/icon15.gif[/img]
I'm not sure this is a bad idea at all. Perhaps Don could chime in with his proto Pelican, somehow positioning another (lit) Luxeon at the base shining outwards and seeing if it increases output somehow.
The generic cone-shaped reflectors as we know them are designed to capture light thrown forward and to the side from a light emission point at or close to its small end of the cone. Pelican's pointing-downward-from-above setup may utilize a differently-shaped cone but if it's close, perhaps there is indeed a Secondary ideal focal point at which an outward-shining emitter could be placed, effectively doubling light output within the same beam.
why not use 4 reflectors in a Turbo Head style thingie? there should be enough space and you could have an output comparable to a SF M6 .... I mean similar to Elektrolumens Tristar except it would use reflectors instead of optix ... I have long been wondering why nobody has done this, and this thread finally gives me the opportunity to voice my madness ... [img]/ubbthreads/images/graemlins/grin.gif[/img] [img]/ubbthreads/images/graemlins/grin.gif[/img]
At the risk of repeating myself, why not 7 collimating optics (actually even the Fraen and NX05 optics are reflecting optics. The difference is that they use total internal reflection and not the regular reflection fron highly uniform metallic surfaces)? The overall diameter will be only 3 times that of each optics but we have the power of 7 LEDs. Suspect that cooling may be more of an issue then.
For a collimated beam, whether a parabolic or spherical reflector is used, there is one finite focal point corresponding to the beam. Multiple LED's would be best served with their own individual optic/reflector I believe. I have seen the triple optic lights and hope to make a head with three Pelican reflectors one of these days..........
tylerdurden - true if the beams are perfectly collimated i.e. zero divergence. To get 1 spot you will need to align them for a given distance. BUT we almost never get perfectly collimated beam - at least not with reflector sizes that are manageable - so the hot spot increases in size as the distance increases. Assuming a divergence of 4 degrees, the spot will measure 70 mm at 1m. With a 7 optic cluster, the distance betwen centres of the spots is a constant value equal to the optics' diameter. For 27 mm Fraen, this is 27 mm. Most parts of the spots will be on top of each other. Now 4 degree is a very narrow angle (I am thinking of the 2 degree HALF angle optics from CTP Coil), so most optics will do even better. That is why the Tri Star has what looks like s single spot. This will not work with very narrow beam, large diameter reflectors (i.e. MaxaBeam).
this may be out of topic but i designed a simple reflector using a 3 sided pyramid i know weird pyramid jokes
but it seems to work tho this is for an indoor light
seems to have amplified the light a bit i'm just using 3 12k 5mm's maybe 1/3 watts anyway i've lit up a bathroom nicely the pyramid has near perfect angles and a glass bottom for defuse, not sure if i'm loosing light there but it seems to beat any curved reflectors (from some old 2 million candle power flood
Well, I recall someone mentioning in a thread recently, about the potential of having an emitter in the centre of the lens, rather than at the back of the reflector, and that can work in the right circumstances (Think satellite dish), the problem though, is that the reflector's going to be set up to be focussed with one source or the other, I think that getting it configured for both sources would be near impossible. Like Evan1982 said, one or the other is going to add mainly to floodlighting, rather than the primary hotspot. I'm no optics expert though, so could well be wrong.
Keep in mind that I don't know nearly as much about this as the rest of you.
Could there be a way of putting a side emitter with a high dome emitter on top of it? Then they could share the heatsink. [img]/ubbthreads/images/graemlins/icon3.gif[/img]
Just a thought.
I really like the idea of putting one emitter (a 3 watt) in the middle, and then using shims to aim peripheral emitters towards the middle. This would create a blinding hot spot where they all converge, and would create a ton of spill further out.
For walking, imagine one emitter aimed straight ahead, with emitters on the sides aimed slightly downward at opposite sides of the path. And what would be ultimately cool would be the ability to turn them on individually. In fact, this could be accomplished by yoking 3 flashlights together in a clever (maybe even flexible) way.
On the idea of a side-emitter & a high dome, actually it's possible to focus both of them with just one reflector, for example by pointing the HD towards the reflector
pelican-style, and placing the SE on top of the HD pointing in the other direction. The 2 luxeons would therefore bounce off different (disjoint) parts of the reflector; the reflector could thus be designed to make one single hotspot (albeit the SE hotspot being wider than the HD one) by machining two different parabolic surfaces.
This would be more understandable with a sketch,
but I don't know how to post a picture?
Now THAT is interesting. I think I have a picture in my head of what you're talking about. I'd love to see a sketch, if you can just get it onto a website it should be easy to post. The machining sounds difficult, but I guess it would be pretty easy for a CNC machine to handle. Would the emitters here be stacked slug-to-slug or dome-to-dome? In either case it sounds like they would need to be suspended in the middle of the refelector, as opposed to at the base or at the opening.
This certainly is possible. The problem will be heat dissipation. The diodes will have to be back to back with the emitting sides facing away from each other. It will have to be in the reflector. Getting the heat out but not blocking the light will be difficult. Getting the focus right for both diodes is also going to be a pain.
OK, here is a rudimentary picture of what I described.
The bottom parabola (blue) is to scale, that's to say the HD is really placed at the focal point. For the 2nd parabola (red) I was too lazy, but it's not very hard to calculate; the actual equation of the curve will depend on the height of the heat sink.
Good idea. I've been thinking about the best way to do this for quite a while.
You'd have to arrange your emitter direction and spacing such that the percentage of light gathered by each emitters reflector is MORE than the percentage of total light generated by the emitter.
If in the above configuration, due to the emitter not having a reflector surrounding it completely, each reflector only caught 20% of the emitters light, then the hotspot of the 3 petal configuration would only be marginally brighter than a single LED within one reflector that catches 50% of the single LED's light (the spill light would obviously be brighter). This is because with the single, you've put 50% of 1 LED into the projected beam - let's call that 0.5L. With the 3 and petal, you'd get 3 * (0.2L) = 0.6L in the beam.
With your configuration, the petals would have to be rather large. A better setup would be to angle the emitters to the side, or even slightly towards the back of the reflector, catching more light from each emitter (inverted truncated pyramodal heat sink block).
Someone could construct such a beast by mounting 3 or 4 emitters on a piece of square aluminum stalk, then cutting and merging 3 or 4 large (mag or whatnot) reflectors together. You would have to place the emitters such that they're at the focal point of their respective reflector. The resulting reflector profile might appear light a fat clover of sorts: 3 or 4 circles merging together. Might be an interesting experiment. Might be worth spending $12-$15 on mag reflectors.
evan, your statement about ponting the leds down instead of up might make sense. But, when you say 50% is gathered by a single led reflector and etc, remember that there is more light near the axis of the reflector. In fact, if one assumes a reflector with an aspect ratio 1:1, I think more light will be gathered from a sideward mounted led than from a forward mounted one.
If you point the leds more downward, more light will be reflected, but also the near-led-axis-light (the strongest light) will travel shorter before it hits the reflector.
I figure one could balance it off by pointing the leds further down when using shallower reflectors, and point them more upward (forward) when using deeper reflectors. This because the focal point of a shallower reflector is further away from the reflector bottom.
[ QUOTE ] tylerdurden said:
Since LEDs are more efficient at lower power, we should be able to run two luxeons at, say, 200ma, and have the total light output be greater than a single emitter at 400ma.
[/ QUOTE ]
I have been working on a 2 x Luxeon III flashlight with side-by-side Stars. I originally chose two LEDs to make the neatest package with the rectangular battery I'm using. When I direct-drive a single 3W Star with that battery, I get 1300mA across the circuit. When I direct-drive two in parallel, I get 900mA total. (I assume this is intrinsic to the circuit rather than because of the extra test leads required to hook up 2 in parallel.)
When I look at the luminosity:current curve for the Lux III, it seems like I'll be making at least as much light from 450mA each through two Stars as I would from 1300mA pushed through just one. And the emitters should be much happier that way. And the battery will burn longer and recharge more times.
If I wanted to do a similar trick with a tubular flashlight, I would use multiple optical or reflective elements like Elektro Lumens' Tri-Star, rather than trying to make one reflector serve two sources.