The light intensity decreases at greater angle only because the apparent area of the LED decreases. The luminance would still be the same. So if the reflector catches that high angle light, it should be able to contribute to the throw as expected.
The increase in light collection will not be huge, though; standard reflectors with depth about equal to the diameter collect more than 70% of the light from the LED.
Interested.
About the angle and the recoil concept, i agree the full 180° is not really necessary, but this would probably depend on the LED (dome or no dome, flat or bulbous phosphor) too.
I would probably go for something like 135°, making the reflector some 33% smaller in diameter.
The drawing is ‘for reference only’ 
But a normal reflector cup also uses the wide angled light of the emitter, so i’m not really sure…
Hmmm…
The standard reflector doesn’t reflect the best / brightest bit of the emitted light though…
This is what the ‘problem’ of a normal reflector thrower is, a lot of light is spilled, it’s called “spill”…
A zoomy only uses what would normally be the spill, so with a recoil set up you get both. 
I can’t find proper and affordable parabolic mirrors anywhere though, i have some spherical concave ones, but that just doesn’t work (they’re made for incandescent projector lamps).
In the drawing it is a copper ‘fin’, at least as thick as the thermal slug of the LED’s width.
The weak spot of the heat dissipation is still the LED itself, not the fin.
I think it will be no problem to run any 3535 on 10+ Watts.
3 fins forming a ‘Mercedes star’ would look cool too.
Ah yes, back on topic, maybe i’ll start a topic on the recoil idea, for this light that deeeep reflector 4wheeler proposes looks very interesting. :+1:
If you haven’t already, have a look at this thread. The throw is just a function of the apparent area of the reflector or lens and the luminance of the LED.
The standard reflector misses the light directly above the LED, but even though this light is the highest intensity, it comprises only about 25% of the total light emitted from the LED. This might seem unintuitive, but it is just a consequence of geometry; the light at more of an angle is less intense, but there is more of it. So an aspheric light might collect the highest intensity light from the LED, but it is only about 25% of the total light.
So I agree that a proper recoil setup could be better in terms of light collection, but it won’t be as much of an improvement as you might expect.
I read that some years ago i think, but i’m not sure what the terms used all mean exactly, so it’s kind of wasted on me…
this (i assume) is 70% of the 180°?
Or is it 70% of the actual light emitted?
This depends also on the relative depth of the reflector.
It will have the tighter beam though and the intensity of a reflector and a zoomy combined.
But enough about it here, i will start a topic about a recoil thrower when i come up with a catchy title.
(and i made a new drawing, 145° in stead of 180°)
IM supplied an estimated throw with XPL HI. Response is:
"optical simulation of the xplhi LED with the 100-1332 LED. With 1klm, the intensity collimation is around 200 - 235 cd/lm, and 2.5 deg beam."
Not that I want this project to end up as an aspherical thrower, but does recoil flashlight throws better than an aspherical flashlight? Or is it significant?
Does recoil thrower have spill?
I’ve never own any recoil thrower so I am not sure.
A few things I noticed, and of course I’m interested… for Justin and Valerie, the BLF JV Classic. The classic from Thijsco19 looks almost exactly like an Olight SR-90 Intimidator. Might have to make some changes to the fins to get away from any copyright issues. Other than that, the simplicity of form as a carrier of function just works.
I put an XP-L W2 2B in an Convoy C8 yesterday (day before?) and got 2028 lumens at 6.6A. Something to think about.
What about the intensity in cd?
Will update OP later
Behind the scenes TA and me are pmming with 5ar for a more in depth render of a design that is based on the Q8 design.
It will have its own look
And I overstretched, for the groupbuy we can surely make sure Justin gets a donation for each light sold but after that it is probably much harder to keep that going and by making that a hard demand it could make it harder to find a manufacturer (or make manufacturer decide to make a knock off of the GT to save the $ and hassle). I will ask for it though but for now realistic a donation per groupbuy light.
Lets just hope and pray something comes out of this.
Interested. Please put me on the list.
I’m interested in this light- please add me to the list.
Please add me to the interest list!
Put me in for one please.
Hmm, that would mean a total of ~300kcd? Seems pretty low for such a large reflector? The TN42 gets 2x that much with a much smaller reflector?
This is a domed LED correct? First off where are you finding these? I have not been able to find any W2 in neutral tints (although even 2B is not neutral).
In a pre-made light the LED needs to be a HI LED from the factory. Dedomed versions will be for modders to do later. It is simply not practical to sell a de-domed light from the factory. Too many possible issues plus the tint shift would not mesh with most people, including me.
So that leaves a V3 HI or an XHP35 far as I know. (assuming we could even get a V3 XP-L HI).
So XPL hi or XHP35HI?
Any other LEDs worth considering?
Haven’t seen this level of power in a binned domeless emitter from Cree. It could be in the works, but I haven’t heard anything about it.
Cutter in Australia is selling a W2 2B, it’s warmer than one might think, not sure if that means the tint bin is incorrect or just what is going on there, but it definitely shows up as a warm tint when viewed under a UV source.
Absolute long range output will almost certainly have to rely on the XHP-35. High output HI variants are not readily available and even if they were, the domeless emitter doesn’t produce throw like a de-domed high bin XP-L, proven fact. I’ve gotten 2 dozen of the new XP-L and they came on 10mm aluminum mcpcbs, attached in a sheet. It’d be comparatively easy to de-dome an entire panel at a time in a solvent, so it shouldn’t be completely out of line for consideration here.