I think you’re close to the heart of the matter with that question.
Perhaps it would help to see the LED as an array of point sources, in a flat, square, 3mm*3mm plane…
When I try that, the ability to describe an “ideal” reflector becomes both more accessible and less possible at the same time.
With an incan, that “point” source is also not a “point” by any means. It’s a long, squiggly line, as you can observe for yourself in the “beam artifacts”. Likewise an HID has a “ball” of light, not a “point”.
I don’t think we’re helping matters by trying to realize some definition of “perfection”. It doesn’t help that there seem to be two diverse “camps” — “I only want Throw” vs. “I only want Flood”. I sincerely believe the optimal answer is going to have the word and in it, not any expression of “either-or”…
I really would love to see more people with the Machine Tools to make the parts, doing experiments like this one!!! (And I’m still amazed that y’all can do this level of machine work by hand!) But I do not believe the Paraboloid is “optimal” for an LED. Not what I’ve seen so far, anyway.
Has anyone tried any other shapes? Hyperboloid, anyone? How about getting away from the whole “conic transection”? What happens if you replace the cone with a paraboloid or hyperboloid & cut it in the same fashion?
What if you cut a Paraboloid or a Hyperboloid instead of a Cone, in the same fashion as when making a Parabola? I’m pretty sure the “mouth” end would approach parallel, but I don’t yet own the software (or grayware) to model it. All I can do is ask.
That “bell-shaped curve” of the Typical Spatial Distribution chart does seem to want a different approach, other than what works for a “ball of light” point source.
The SupFire F9 and Solarforce M8 have a similar feature in construction, about 10-12mm out from the emitter is a different angle, then the paraboloid resumes the remainder of the length out to the rim. The M8 has orange peel in this area, the F9 just shows a different magnification of reflection. If I had a spare in each or either I’d cut it in half to see the profile. Each has a nicely defined hot-spot, a bit on the large side which I like, and moderate spill, which I also like. Neither are true throwers in the most literate sense of the word but each does a respectable job in that dept, with the large hot-spot growing to illuminate a broad area out around 200+ yards.
These are both favorites, with the M8 hosting an MT-G2 and the F9 carrying an XM-L U3.
The M8, although considerably smaller than the S2200, has a slightly tighter hot-spot with similar beam profile characteristics to that honking big reflector in the largest of the Solarforce lights. Oh, were they all made so well!
An LED is a bunch of points of light clumped together so no matter what shape reflector you use, it’ll always have a relatively round pattern even though it might be all rings. A filament is a long squiggly line of points with no real center. No reflector of any shape is going produce a good spot or any kind of smooth pattern. The best you can do is focus at a mid point on the filament and hope the rest blends in. As they say, garbage in, garbage out.
Some of this discussion made me go back to the drawing board. Things that I have discovered:
1) A one inch tall reflector will have TONS of side spill unless the inside diameter is quite a bit smaller than 7/8" like I was doing.
2) I need to account for the focus point of the parabola. The equation that I used "y=.875 x^2 has a focus point that is way too high (.28")
3) I need more tooling (no surprise).
I did a test with the 2nd reflector that I made. It does have a hot spot but it has a crap ton of side spill. It is too short for the profile that I made.
I figured out the math for focus point of a parabola then looked at the XM-L2 die. The dome is about 3mm high and wide. Figuring half of that height for my desired focus height gave me 1.5 / 25.4 = .059". I can't attain that in a 7/8" diameter without making it really deep. I could just machine some of the bottom off, but that makes it shorter > more direct spill > not what I want.
So, the current plan:
I'm going to go slightly smaller in diameter to 3/4".
I am also changing the "a" number in the equation to 1.5. That will give me a focus point height of .167" from the bottom of the parabola. I can just machine some of the bottom off to get me in the ballpark.
Damn. I just thought of something else. I'll be back after more Winnie the Pooh type thinking...
Did you see that Scott? You’re supposed to use a Drawing Board! Gosh, what could you come up with if your bar napkins were the size of a drawing board?
Oh, and Buck? Dome? We don’t need no steenking dome!
You got me thinking about why de-doming gives better throw. I have to wonder if it is because the emitter plane is closer to the focal point than the average of the dome. Or is there something else at work.
I have always thought that it is because the source of the light more closely resembles a point source. Small sources of light cast “hard” shadows, larger sources of light cast “softer” shadows. That is why frosted light bulbs are called “soft”. By frosting the glass the source of light becomes larger. Instead of being the filament or emitter, it is now the entire surface of the bulb.
The sun on a clear day casts a sharp shadow. On an overcast day, where the sun is not visible, and the source of light is the entire sky there are no shadows at all!
A light source that is an infinite plane will cast no shadow.
(All of this, I remember from my High School Physics class, I paid attention that day )
I had assumed it was because the dome has an effect on the beam, apparently to get more light “up” (meaning OTF of the reflector) from the LED. Sorta like the wee “optic” on an XR-E, I think. There isn’t much light to work with, and the technology isn’t designed to be a “point source”, and apparently no one (present company EXcluded, I hope) has yet come up with a “good” (as opposed to “workable”) reflector. Still, I’ve yet to stumble across “bare emitter” beamshots with and without domes…
So “the way it’s always been done” is to add the dome. I don’t see it as “protective”, especially since you guys use gasoline (and other, more-scary means) to remove them (w/o killing the emitter)!! Besides, a mounted Optic has got to be more “protective” than any wee plastic dome…
I think it’s just more emissions getting out to the sides of the reflector, which (as we’ve seen over-and-over) gets more light into the focused hot spot…
You, dear Bucket, have the power to prove what’s going on! IIRC, you have a means of holding the LED at/near the Focal Point of your reflectors? Well, perhaps you can adjust it a mm or so each way (with and w/o de dome) to see? (that might even enable you to adjust the beam a tiny bit a-la M*G lights…??)
But NOT if it means putting off the rest of your work, please!! There are plenty of people here who can test & document beam profiles, and many of them can exceed any “ISO9000” levels of QC; and I’m pretty sure all of us would bend over backwards to help you come up with a “really good” reflector. The carving of the tool is the Hard Part.
As lightme pointed out, the tightest hotspot is obtained with a paraboloid reflector, with the lightsource as good as possible in its focus. The size of the reflector (diameter and how deep it is) together with the quality (size and shape) of the lightsource determines the size and quality of the hotspot.
Any divergence from a paraboloid will make hotspot brightness (throw) less. There are no magical speciality 'led' reflector shapes that increase throw compared to a paraboloid. A special reflector for this or that specific led may produce a more pleasing beam by smoothing out imperfect projections of the lightsource (e.g. a square lightsource will give 'flowers' around the hotspot), but the best shape for maximum throw is simply a paraboloid.
Is there a physicist around that knows better than this?
Where to begin? I've been madly crunching numbers in my spreadsheet trying to figure out the best profile. I think theory will only carry me so far and I'm going to have to make several different profiles and experiment with them.
To summarize:
-The drill rod I bought is 7/8 diameter, so I can machine it smaller but 7/8 is the max diameter for now.
-Formula for a parabola is y=Ax2+Bx+C....I am using 0 for B and C so it's just y=Ax2.
-Smaller values of A yield a more "throwy" narrow parabola, but move the focus point higher above the vertex.
-this requires machining more off of the bottom of the mirror to get the LED to the focus point. It also results in a larger hole where the emitter pokes through.
-Smaller values of A also require a longer parabola for the same diameter. Since I'm machining a lot off of the bottom, this isn't a bad thing.
Plan for the day:
Machine new bit using A=0.5, so y= .5x2and leave the total diameter at .875".
I also decided to increment the x value on the lathe by .015 so I have less filing and sanding to do.
First, that’s only True if you can mount a point source exactly at the Focal Point; but PLEASE let’s not start a “Ford vs. Chevy” feud over this! As many have tried to imply earlier, not all of us care to illuminate things hundreds of yards away. And when/if we do, there are MUCH better choices than any LED solution.
“Useful” light is that which illuminates what you are using. For the work I do, that’s seldom far enough away that you’d have to walk for it, nevermind a 100-yard dash. IOW, I can throw the torch much farther than I need for it to throw light. :bigsmile:
“Throw”, on a rifle light, is The Only Thing. A “Fat-Head” (C8-class) on a .30-06 w/a 10x50 scope, and the feral hogs are toast. But the best “thrower” I ever had (a 2C incan), made a pencil-thin beam like a cheap LASER, and I gave it away because it was useless in close where I need light. A stinking-bright spot the size of a quarter does not help at all when you need to see what’s in your hands!
Where I live, there is seldom any opportunity for a “long” rifle shot, and the only place I’ve really NEEDED a long throw was while night fishing. Night fishing really showcases what we’re trying to do here, as you need “spill” aka “flood” for working in & around the boat, and “throw” for navigation. In my world, that means a lot of “flood” and very little “throw”; although when we do need the throw, the Q-Beam keeps the LEDs in pockets.
The idea of meeting both needs in one torch (with a smooth, even, monochrome beam) is The Pipe Dream, I think.
Dim, you might be able to throw the flashlight further than you normally need the flashlight to throw. But let me ask you this…how fast can you run? The reason I’m asking is because I’d like to know how far away you need to see a black Rottweiler coming at you in order to outrun it to safety. Perspective. There are lots of reasons to use a light.
I de-domed and bumped power on a little CR123 Eagletac yesterday. The de-domed XP-G2 now throws a reasonable beam out past 100 yds and gives ample spill to see what’s going between that distance. 270 OTF lumens @ 30 sec from a thumb sized little dynamo. Pretty nice when you need to see in the dark.
My de-domed XM-L2 HD2010 showed me there was nothing closing in for the next 800+ yards when I was walking the trash down the lane on my 270 yd Thursday night walk. Works for me!
I am no “Jessie Owens”, my friend!!! I cannot outrun a fat Rottie in broad daylight. But he cannot outrun my .40S&W hollowpoints either, so we have to find some Truce… Or I’d have to “fling a Cesar move on him” (as my friend likes to put it when I “manage” dogs) and make him part of my pack…
I take your point, of course, well enough; and agree. That’s why I keep preaching “AND, not EITHER-OR” …
You must not be in Hill Country… I remember the big, open vistas when I was in Dallas… 270 yards for a driveway is pretty rare around here, and the trees get in the way anyway.
I shot comp with a .40 S&W, fully understand your point, but you still have to see him to shoot him. If you can hit him, weaving through the trees. If there’s only one. They get sneaky and take the orange and white bullseye off and it’s hard to get em just right.
We live in the country, city folks like to dump a dog that’s gotten too big for their apt or uncontrollable. These “strays” tend to group up, run in a pack that fears nothing. They take down calves, sometimes cows. And whatever else gets in their path. As people like Dobie’s and Rotties, it’s common to see several running together, with a Shep thrown in or a bulldog. Never can tell, and then there’s the coyotes. Some of them are the size of a nice German Shepherd, never thought they got that big. Anyhoo….
I’m between Dallas and Austin, smack dab in the Heart of Texas! Wide open farm country, for the most part.
I found that using a good thrower works great for close work if you turn down the brightness. I usually don’t need a broad beam, just a less harsh one. How often do you really need more than a three inch spread for close work? In any case, a focused parabolic shape is a good starting point to insure the center is brightest. Then you could gradually tulip out the wide end. Personally, I don’t like lights with a well-defined hotspot so if it wasn’t for OP, that’s what I would do.
As far as needing a physicist to figure out what’s happening. Reflectors use simple geometry. Angle in = angle out. That’s it! Lenses are a different story and I prefer to steer clear of them.
I’m on a mountain ridge with woods all around. The next ridge over is the only distant object with no obstructions, otherwise, I only need a tight spot to go between trees without lighting up the closer ones. Black bear is my only concern so making noise is more important than making light. There’s supposed to be rattlers out here but I’ve only seen other kinds of snakes.
I still have a bit of sanding to do to remove the lines but here's one for you to do your angle of incidence analysis if you care to lightme. I took this one from damn close to square.
)
