Also.
If anyone is in consideration of what has been proposed recently; and more anonymity you seek, for whatever your reasons are…
Then make note of something like “post edited for clarity” within about 12 hours from now as I will do, and delete whatever you wish to clean up, like the egg holders or anything else. Problem solved.
Now we do this as fellow friends and machinists behind closed curtains ( 0:) ), collaborating to bring you the very best BLF egg-holders anyone will ever need.
That’s both an enlightening post, and graph. If the blue reflectivity of metals varies so much then I suspect the specific alloy will matter too. I did some looking and aluminum really is the best material for this application with over 3x the reflectance in the blue region, too bad it oxidizes unless it’s coated.
Ok, i try to find my posts with referenses to this & “clean” it up.
EDIT this thread is done on my posts i think anyway.
If anyone know of any of my posts that could benefit to a bit of scrubbing, please let me know. So we can get some beautiful & shiny jellybean holders done in the future 
And MEM that is a really unfortunate story, i am sorry you had to go through that.
Kloeper Knife Works really has great idea but MEM is original inventor of this.
MEM consider option 4 small jelly bean holder which you can easy send worldwide. Plug and play is what most people want.
That can be used in smaller versions of this light(which will be more in use by hunters and more hunter friendly than 1504 ).
The only thing I’ve done is decide that I like a small collar from the standpoint of it fitting any light. I doubt that was even an original idea. I want to improve on the performance of the ones I’ve made though, and I like the science that is involved, but as far as me coming up with any new original ideas for the collars, not yet.
I think i just got an interesting idea…….
What if we used a XP32 Noctigon with a dedomed XP-G2/XP-L with an big RA covering at least the whole XP32, in the most blue reflecting material + one of these special coating. AND place blue leds on the 3 spaces around the middle one.
Shouldn’t that increase the phosphors excitation, from all the extra blue light coming from the 3 extra leds reflecting back to the middle one?
So MEM & anyone other interested, is it a viable idea?
According to djozz blue XP-E2 test, they top out at 450 lumens & 3A. But if it would work that’s 1350 extra lumens of blue blasting that poor little middle emitter to higher excitation 
My hunch is that only a small percentage of the blue would actually end up focused on the main die. Most would be reflected until lost to absorption. Keep in mind the MCPCB face would still have a low reflectivity.
Maybe the face of the mcpcb could be coated with an reflective coating also.
Cajampa,
My suggestion is that you take that idea, forget you ever had it, and definitely don’t try to solve it. LOL
I’ve been working on remote excitation for some years now and yes it works but is a royal PITA to do right for extremely high surface intensity. Cree finally saw it as a plausible usage scenario and started offering yellow compound wafers for remote excitation experiments and new luminaire designs—though lumen/W is not high the way they intend it to be used.
The way you have explained it Cajampa, means there will be a shower of uncontrolled photons scattering around inside of a hemisphere with no collimation. Take the area of the mirror+area of the circle under the emitter, say that the total light within the cavity is allowed to cover this area, and you can picture quickly how small the primary die is compared to this area total.
Also, as stimulation power is increased, the color falls on the CIE 1931 chart into the blue region. This is OK. Starting with a lower-K (warmer) bin offsets the effect. Sooner or later you end up facing the same barrier, though; heat. The primary problem is removing the heat from the phosphor substrate before the binding agents darken and cause a thermal runaway event (poof). There are some interesting projects I’ve worked with in the form of remote excitation and I continue to. Sooner or later you will have one of them in your own light. The energy increase is not free, of course. What I mean is, with an RA, the input power remains the same. With remote excitation, the “remote” devices are using power as well which must be accounted for in the total wattage draw.
Ok, gotcha i failed to consider that only a small part of the blue light would even reach the middle emitter.
And it sounds like you tried out something similar, but with the remote phosphor angle that didn’t work out, because it couldn’t reach a higher surface intensity than the white emitter plus RA approach.
Would a blue laser (they come in up to 5W now) also heat up the binding agents and cause a thermal runaway event? Hmm maybe that is a given, considering most build that powerful lasers precisely because the want to burn stuff with them 
Yes blue lasers can excite the phosphor, but you have to remember that the semi-conductor below can absorb photons as well as make them. A semi-conductor is also more efficient at lower temperatures. When you have an LED that is driven hard, it is already hot by nature. So by adding an intense concentration of light, you are simply adding an intense concentration of energy, which has to be converted to light, and we know the conversion is not efficient (light obtained vs power input = big difference); end result is an addition of more heat power than light power. What happens is, the laser will create the effect of increased intensity where it strikes, but its heat will quickly soak into the LED, lowering the overall output. If the energy is too concentrated, yes—any amount of heat much over the die’s intended limits will char it black. An RA places light back on the LED in the most proper way. It is the RA which needs to be improved the most, and built to the highest standards, because these gains will be the most advantageous over adding crazy contraptions which no average person will ever be able to practically setup and configure to work without using advanced testing equipment made for lasers and LEDs to tune two independent systems together for a minimal gain vs costs.
I have some interesting lasers I have built. They are incredible when you apply a beam expander, especially on a high-output green. You can see the dot, literally the dot, at 3000+ meters with only a few hundred mW of 532nm. With a beam expander, instead of that parallel look as the beam travels away (even though it’s expanding), like the beam just suddenly ends in space, you see a neat visual effect like the beam is getting smaller and smaller as it goes into the distance. In other words, it really looks like it’s going off into infinity at the same beam diameter. It’s just increasing at a very small divergence angle, is all.
You makes some very good points, thanks for setting me straight
One thing i have been wondering now & again, is it possible to use a beamexpander in reverse as a pre collimator, to make a XM-L2 class emitter appear to the ashperic lens about half the size or the size of the longer throwing XP-G2 but with a higher intensity.
I am sure there is something fundamental i am failing to consider again here :), but it feels to me like it could work, but if it doesn’t please explain to me why so i can stop wondering about it.
this is my review …….
So recently I picked up one of these for the hell of it (XML-U2 variety from GearBest), and while I do find it impressive, I want more power out of it. I’ve chosen the XP-L HI V2 3C LED from iOS, but I want to remain using the two Efest IMR green 4200mAh batteries. I was looking at the LD-2D from iOS for the driver, but I am worried if the pill had enough room in it for that tall of a driver. Would this driver and LED combo make good sense with this lights design? Please let me know what you guys think. Thanks!
This 17mm MTN-MAXlp driver by Richard, is the best driver out there for this light it is practically designed for it nothing come close to its performance in this size.
Thanks for the info cajampa, but the pill in this light is aluminum and it would seem that at high amperages (over 3A) the pill is going to get fully heat saturated and damage the LED for run times over 60 seconds. I want longer run times, so I’m looking to go no more than 3A which I feel would be safer for the LED. Its too bad no one makes copper pills for this light. 
The aluminium pill has excellent heat transfer and capacity and will be fine for any current the led runs at. The need for a copper pill is a BLF-myt, the gain in heatshedding is marginal in a single-led situation. You do want a DTP copper ledboard because that is what really matters, but since you order the led from IOS, it is probably on a Noctigon board so that is accounted for.
Yes, i agree 100% with djozz, you don’t to worry about that as long as you use a DTP mcpcb.
The 1405 is a massive chunk of metal it will take 5A just fine without damaging the led, remember i said this driver is made for this combo
And if you want maximum kcd chose the XP-G2 S4 SB or the XM-L2 U4 1C on a noctigon from IOS & dedome them in some gas If you don’t want to dedome your self you can get it from here
I agree with djozz. I’ve never been concerned about the 5A I send to an XP-G2 in a 1405. I’d get more concerned once you move into MT-G2 territory at 5A, after a couple of minutes being on that is.
how reliable is the new 17mm MTN-MAXlp driver? and what is a theoretical throw with this driver + a dedomed xp-g2 ?