subscribed this interesting topic…
I used a bunch of Optolife aspheric a couple of years back. I was having trouble getting decent 50mm from Kaiodomain, Dino Direct and the usual suspects. Although not MG quality they were a significant improvement over most China glass at a reasonable price. The AR coating was a good option.
So, an aspheric lens cannot be used to produce a very tight spot at, let's say 10 miles, with a led and a reflector cannot be made that does the same, so in reality leds are no good for real long distance throw (several miles), and lasers are all that remains for that purpose?
oh yes they can throw that far, as long as you make the aspheric lens/reflector big enough. Put a dedomed XM-L in one of those WWII air defense search lights and you have a pencil beam that goes 20 miles.
Where can I get one?
So, like I said, it's not really something we can do in a flashlight, so a laser would be about the only way to go, but a laser beam is too small and not white?
and not legal here, never mind.
Thread derailment warning!
I’ve always wondered about the Bat signal and whether or not it would work as depicted. Wouldn’t a Bat shaped die be the best way to project a Bat shaped image? I guess one is casting a shadow and the other projecting an image. Shadows just seem too easily blurred by distance.
I think Ervin’s combination of reflector and aspheric is the best I’ve seen at generating throw from a modest sized flashlight.
Yes, that must be what he means. Thanks Ruf.
I can not remember the details of Ervin's design, but adding a reflector to an aspheric light does not improve the throw. It will add to the flood around the hotspot, usually in a ringy way.
This has been discussed a lot, so I am saying more than immediately necessary here.
The simple optics approach (expanded):
The ray that goes through the center of the lens from anywhere on the LED goes approximately straight through the lens. So the angle of that ray is not changed. So the angle of the beam is the same as the angle of the LED viewed from the lens. The spot size is the LED size times the ratio of the distance from the lens to the target to the distance of the LED to the lens. For a fixed lens diameter, moving it closer to the LED and making it thicker so it focuses catches more light but makes the spot bigger. Moving it farther from the LED and still in focus makes the spot smaller but loses more light. If the lens is small relative to the angular spread of the LED these two effects compensate and the throw does not change. For typical LEDs and lenses, the outer part of the lens sees a bit less light than the center, so a longer focal length does increase the throw but not as much as it decreases the total output.
More general optics point of view:
Another way to look at it, which I really like, is Dr. Jon’s luminance. Simply put, you can’t make the front of the light look any brighter than the LED. To get a brighter spot, you have to get a wider flashlight.
Relation to other physics problems:
From Liouville’s theorem (Hamiltonian), the phase space of a classical statistical system, such as a beam of light is a constant of the motion. It can get tangled up, such as with a diffuser, so it looks bigger, but it can’t even look smaller.
This is roughly the angular spread times the position spread. So if the beam comes out through an aperture of a certain size, it must have a minimum angular spread. The angular spread of the LED is reduced by the optics, but only by the ratio of the optics size to the LED size. To make the beam narrower than that, you have to remove some of the rays from it.
Do you have the idea that you are helping anyone understand anything with this post??
I know the first point of view is obscure. To me it is interesting that this is a special case of a much more general principle. I know it may not happen, but it would be nice if this helped someone else be aware of the unity of physics.
The other two approaches are simple enough to be of general use to others on the forum. The idea that a ray passing through the center of a lens goes approximately straight is very simple and useful. The idea that the luminance can be understood as the brightness of a surface and especially the idea that this quantity cannot be increased by passive optics have been discussed repeatedly in relation to dedoming (which does increase luminance).
These, and especially the impossibility of increasing luminance, are needed here to give a clear answer and keep the discussion here and elsewhere from bouncing back and forth without a solid conclusion, as it had been.
I am changing the order, so people won’t have to read what they can’t understand to get to what they can.
Well, most of the time, the people who know, can't understand it in a manner that the dummies who know nothing, can understand. My problem is that I walked through high school, in a small town, in the 60's and didn't have to take mandatory classes in 11th and 12th grades, because I had enough "points", by 10th grade, that I could pick for myself. I never went farther than basic math, no algebra, or anything farther than multiply, divide, etc. I never went through any science classes because only 1 year of biology served as the necessary mandatory class. I did literally nothing the last year, taking the mandatory English, speech and athletics classes, (yes mandatory, shows what they were about) and the rest were study halls. I sat through 5 study halls a day. I had absolutely no interest in any of it at all and since they let me do it, I skated through it.
So, when you start quoting stuff like in some of the posts here, it is totally lost on me. What I need is, "you are right, you need the biggest, thinnest aspheric lens you can get and no matter what, you will not produce a small beam at 1 mile or more. It's not possible" That's the kind of thing I need. Hopefully, the smart people here will gain something from the answers, but I won't. Now you know why I will never, ever take an electronics class and try to learn circuits. I would have to take years worth of other classes, to even begin to understand the basics.
The American school system has a poor record. I have been taking care of kids. They are enormously motivated to learn. Somehow the schools beat that out of us. Most successful people in the US either went to school in some other country or learned somewhere else.
Examples: I have a really excellent scientific education and my career ground to a halt early, while my brother who never finished any school, as far as I know, is doing well paying, very interesting work on car engines.
I find aspherics great to diffuse the light evenly with minimal loss bathing a large spot with very even light.
I think he situated the lens within the reflector so that without blocking light that hits the reflector the unfocused light that normally misses the reflector is focused by the aspheric. The aspheric is smaller in diameter than the reflector and sits below the rim.
Yes.
Hello. Thread necroing time!
Well, yesterday, while trying to hot-glue the thing, I dropped over the floor the glass plano-convex lens I had purchased for overhauling my shabby flashlight (30/28mm∅ full/convex surface, 7mm wide), and it chipped… :FACEPALM:
I had to mill a thin, jagged, protruding circular border inside of the head cover which served to hold the just slightly smaller original acrylic lens; with it gone, now I realize I have to get a bigger lens because, as it is now, the head tube is 33mm∅ ID and its cover has a maximum of 36-37mm∅ of lens slot: anything bigger won't go and anything smaller falls into the pit. I'll have to machine the head cover somehow (output hole is just 28mm∅ right now). :FACEPALM:
Well, after an ultra-brief time lapse of depression, I decided I was not gonna throw in the towel…
Honestly, I'm posting this in case there's someone who could give me a (mathematical) hint on these matters, total n00b in optics. I'm taking a look at this: Diameter 37mm Convex Lens Glass for Google Cardboard Virtual Reality VR LA. Looks to have just slightly more focal distance than the one I just broke, which is fine because I am using a pre-collimator, and the result was good testing over an XM-L die with the 28mm∅ convex surface glass lens. Gonna be used over an XHP70.
Duh! Just wrote another chapter here. 
Cheers ^:)
Groovy!!!
With such a focal lenght, my estimations predict ≈82'4% maybe 80+% flux output from a fully focused XHP70 beneath (about 110° of collimated light cone). 
€3'57 to spare anyone? (.357 lol)
Cheers ^:)
I only read through some of the first page, but I don't think anyone really hit this focus thing very perfectly. The point is the non-zero size of the LED.
For geometrical optics, If a single POINT source of light is exactly at the focal point of a perfect lens, the beam will be very tight.. parallel, laser-like. It won't spread out more than the lens size itself, the ideal thrower.
But you don't have a point-like light source. You have an extended light source. The point-like light source situation depends on the fact that every beam of light hitting the lens at a certain lens location strikes that location at exactly one angle (because it came from exactly one place and hit exactly one place), which if the lens is made right, is just the angle needed at that part of the lens for the light to end up coming out straight. But for the extended source, light coming from the edge of source will hit the same lens location at a different angle, and will also thus leave at a different angle, not straight, so you have diverging rays.
The farther away the lens is , the less the angle difference is from two points on the source to any single point on the lens, so all light striking any one point on the lens comes in more nearly at the same correct angle. So the lens needs to be farther away(edit: with correspondingly longer focal length) to focus well, and, to still catch as much light while being farther away, it clearly also needs to be bigger.
This same argument is also precisely why smaller LEDs throw better for the same size lens.
By the way, I learned in American schools, mostly all public schools at that.
Edited for brevity... a little.