I think it’s no accident that many parabolic reflectors for flashlights are designed such that the diameter of the reflector is roughly equal to the distance from reflector edge to the focal point. Because that would describe exactly a 30° half angle…

Yeah you need to take into account that the LED emits in a hemisphere, so even though angles close to 0 are more intense, the circumference of the circle at that angle is small.
Luminous flux depends not only on intensity but on intensity and area, which you probably found out when reading about Lambertian distribution.
Here’s some more info:

The graph with the lobes shows that at 0 degrees there are 0 lumens emitted, because the area is 0 even though intensity is max.
Also at 90 degrees there are 0 lumens because the intensity is 0 while area is max.

You need to take the area integral of this lobe graph, not just the spatial distribution that is on LED datasheets.

Read my post above, you need to take into account area, not just intensity.

Think about measuring the lux from an LED with the luxmeter at the end of a 1m long stick.
When you move around the hemisphere there is a lot more places the luxmeter can be when you’re near 90 degrees.
When the luxmeter is right in front of the LED, there’s only one spot.

lumens = lux * area

thats a great explanation enderman. I too have been thinking intensity but your analogy perfectly set my imagination straight.

sometimes things are that easy.

Well, the lux meter can be 30deg. to either side of the central axis of the LED and still read 90% luminous intensity of dead center. I found a calculator online and found the surface area of a parabolic reflector for a front-facing LED whose diameter is the approximately the same as the distance from opening to focal point about four times that of the surface area of the opening. But even though the surface area is four times as large, the light intensity I would estimate to be about 2/3 that of the light emitted from a 30deg. to 0. And the efficacy of an electroformed parabolic reflector surface with rhodium coating is around 90%; I’m sure the cast, machined, and polished/coated surface of say a BLF GT to be slightly less. So maybe the ratio of light from the reflector versus light directly from 0-30deg. is more like 2:1 rather than 3:1.

But the reason for using these emitters is to effect a longer throw, not to put the light on an integrating sphere to get a lux reading. Your very own OptoFire uses that 30deg. half angle since that’s the aperture size of the RLT you used,. And the lens is approximately the same size as the reflector on the GT - yet it throws farther than a Black Flat installed in a GT with a reflector - despite the GT’s reflector having over four times the surface area of the RLT collar you used. A collar which you noted is about 33% efficient at collecting light from 30deg - 90deg. So using your theory a larger collar (with the same aperture angle) would be more efficient - but I don’t think so, because any errors on the surface would be greatly magnified dur the distance of the reflector from the LES. But for a larger emitter it may be beneficial.

LouieAtienza, I think one concept you are missing is that of luminance. This is a measure of the surface brightness (of the LED) and it is important in the determination of beam intensity or throw (intensity in cd is equal to luminance times apparent reflector area). The luminance of LEDs does not vary with angle; the reason the intensity goes down as you go to the side of the LED is because the apparent area decreases. See here for some explanation.

The intensity of a flashlight depends on two things, the emitter’s luminance (cd/mm^2) and the front area of the optic.
Just ignore reflectivity for now, to simplify things.
The luminous flux output of a flashlight depends on the angle of light collection, because the angle tells us both the area on the hemisphere as well as the intensity.

Try playing around with these calculators, you might see how things work when you adjust the value sliders:

I do have some understanding of the importance of luminance… which is why I take interest in this thread in the first place! I actually have a few of the Black Flat and White Flat 1mm^2 and 2mm^2; and have already done a couple “learning” builds with them (which I still tinker with). I modded my GT Mini with a White Flat 1mm^2 not too long ago; my Uncle liked it so much he’s likely the only guy in the Philippines who has a GT Mini with a White Flat! I thought Enderman’s canon-style layout was great because it gives me a test bed for swapping emitters and optics and electronics. I’ll have to wait however until it gets warm enough outside to get more accurate long-range lux readings without me shaking from the cold.

The recent comments stemmed how to achieve additional luminance to a standard layout reflector-based flashlight in a retrofit scenario - that is, after de-doming (which I haven’t done yet to my XHP70.2). The question was as to how much benefit there’d be if a spherical reflector with a width coinciding with the 30deg. half angle to take the “spill light” and reflect it back onto the die. Enderman estimates about an 11% gain, and how that’s probably not worthwhile. Coming from someone who chases a 1 percent gain to obtain the maximum out of every component just seemed a bit off to me, when other ideas to gain 11% would be far costlier. So I probably defended my position with some half-true assumptions. Still, I think even half, a 5.5% gain, for $3 in parts from the box-‘o’-parts, and a bit of sweat equity from a man that barely watches TV, and otherwise doesn’t have a clue, is not too horrible. Sure I could just buy a precision aspheric lens, get a MarineBeam Illuminator RLT to harvest the Wavien collar (because that’s probably the only way I could get one now) and spend hours like_ pscal_ to make a 3.508Mcd GT. But since I’m trying to stick to the “budget” part of BLF just trying to come up with ways to get more performance on the cheap…

Thanks for the link - it reinforced some notions I intuitively had as well as gave insight on some that were not-so-intuitive…

I’ll check that out later, thanks for the link…

Mmm, suggested maximum continuous driving current for KW CSLPM1.TG in a C8S?

Gonna pimp an LD-29 for 9 - 10A :-P of output LoL, check the update in HKJ's 2014/02/16 LD-29 review. Wish me luck.

Cheers ^:)

Interesting two lens design (LEDiL Seanna) with an Osram 1mm^2. The first wall shot is with out the exterior Fresnel lens. Not the most practical of lights but it was fun to build. I like to try a center gasket with different thicknesses to see the effect. Maybe removing the first lens and taking a shot. Early stages of this build.

Cool! I have two of these… I fitted one with a 2mm*2, have to take some beam shots. But yes, the interior TIR lens produces a weird beam, like an axicon lens for a laser.

I remember seeing some automotive presentation which recommended conical light pipes before aspherics to improve light collection. So….axicon.

I modded for a german Forum member an Utorch UT02 I measured 440k he 480k with the 2mm2 at 7A with 2S buck, about 1500 lumens

It was the Other Osram package with the die, on white flat the centering ring needs to be sanded down

I upgraded my maxtoch shooter 2x with the CSLNM1.TG. I never cared much for the stock dedomed xml2, very green. I had a bit of trouble at first getting the LED onto an XPL MCPCB, then had a LOT of trouble getting the focus, but I finally got it.



Awesome result! Is that with 2x26350 cells? So is this the white flat 2mm^2 or not? I’m not sure what you mean by “other osram package with the die”.

So I had time to digest this particular tidbit of info and now it’s bugging me. Because if the intensity remained the same, while the apparent area decreased, then that would be the luminous intensity would increase. But I intuitively guess that’s not the case, so I’m missing something here.

Thanks all for your patience and explanations. I do love reading but most of what I’ve learn has been through experimenting and trial-and-error - meaning I am an LED executioner of sorts. But one never knows the limits unless they’re exceeded…

It’s really easy to overthink it. When you think of the intensity being constant over a given area then it makes more sense that the bigger the area, the more light you will see coming from it.

Yes… I understand that. But, as I mentioned, if the intensity of the surface remained the same, and its apparent area became smaller (because we’re looking at it from an angle), then by the definition above the luminance would increase as you moved further away from center. I figure intuitively that is not the case, but I could may well be mistaken. I don’t think it’s overthinking things; We could easily take a lux meter and orbit 180 degrees around an LED and know. But we don’t need to - the manufacturers provide us that data. So something’s amiss in the explanation; either there’s a correction factor based on an angle from the axis of the LED, or something.

I guess we’re thinking of “intensity” differently then. I think of intensity as a constant, like one unit of brightness. More units take up more space. So, if you have less space, you have less brightness. To get more brightness from the same space (or from less space), you have to increase the intensity itself. Like a building has a specific size footprint. You can get more space by adding more buildings, but you need more land to do that. If your land is limited, the only way to get more space is to make the building taller.