Thanks for trying Enderman. :+1:
I was doing some ray traces and calculations on a larger version of a recoil thrower, when I came across some interesting observations…
With a lens, it’s easy to increase the intensity of the LED nearly 2x by using a collar.
However if you use a parabolic reflector to collect nearly 180 degrees of light, you can’t use a collar.
.
In both cases pretty much all lumens are getting used, assuming the lens collects all 60 degrees of light from the collar, and the reflector collects 180 degrees.
The problem comes when choosing diameters.
If you compare a lens of diameter X to a parabolic reflector of diameter X, the entire surface of the reflector is much closer to the LED than the lens is.
This results in a much larger spot, with the same amount of lumens, so less throw.
.
If we distance the reflector from the LED, without changing diameter, the reflector will collect less and less light until it is the same distance away as the lens, collecting 60 degrees.
At this distance, it will be creating the same size spot as the lens, but without a wavien collar it will not be as bright.
The reflector config is also at a disadvantage due to the LED (and wavien collar if you add it to not waste light) blocking the center of the beam.
————————-
Alternatively, if we choose the bottom of the reflector and the lens to be at the same distance of the LED, say for example 3”, the reflector would need to be ~12” diameter to collect 180 degrees of light, which the lens easily does using a wavien collar.
.
In this scenario, the reflector is the better choice because the surface is between 3-6” away from the led, while the surface of the lens is only about 3-4” away.
Both collect the same amount of light, but the reflector makes a smaller spot.
However, since there is a large variation between 3 and 6”, the spot created by the reflector will not be as uniform as the one created by the lens.
————————-
Overall it comes down to how large of a diameter you are willing to use.
Over 150mm it is very hard to find high quality aspheric lenses, and they are not cheap, however 12-24” reflectors are fairly common for under $1k.
.
.
TL:DR
If you use a 5” parabolic reflector that collects 60 degrees, it’s lux will be almost the same as a lens that collects 60 degrees, but worse worse than a lens+collar that collects 60 degrees.
If the parabolic reflector collects 180 degrees it will be about the same number of lumens as a lens+collar system but will make a much larger spot, so less lux.
The only way to get more lux than a lens+collar would be to do reflector+collar collecting 60 degrees with the reflector being larger diameter than the lens, or to have a reflector that collects 180 degrees with much larger diameter than the lens.
.
PS-
I know this is kinda pessimistic, but I’m pretty sure all my calculations and assumptions are correct.
I also assume the lumen loss due to a collar is similar to lumen loss due to a small center circle blocking the beam of a reflector to make things simpler.
It’s late, I probably haven’t explained some stuff too clearly.
If you’re confused just quote what part and I will make more ray traces tomorrow to explain it better.
Running out of time, bound to go to France for a week, maybe longer.
But 180° is a bit pointless, i think 140° would be the best option (without collar).
My point is it would still not be better than an equivalent diameter lens with collar.
It’s possible to get an off-axis parabolic mirror which would eliminate the problem of something blocking the center of the beam, but those are extremely expensive.
It would be nice to know how many lumens are lost by using a collar, unfortunately I don’t have an integrating sphere and few people have collars.
I wonder if someone has both and could find out.
I see your point, but i’m still thinking heat-pipe-beam across the reflector with the LED in the centre.
A collar can’t make up for focussed reflection of the bulk of the emitted light.
But i’m afraid we’ll never know, unless a suitable reflector suddenly pops up…
We’ve been stuck there for quite a while now… 
Hi Dear, I was playing around with this reflector since it is cheap, big and have close focal point. Maybe this can help you. Sorry if you have already know it but I am pretty new at the forum.
This have more diameter but more expensive too:
I mean, I was planning to buy a 11” electroformed reflector that would collect nearly 100% of the light, but even though it would get almost all lumens, the distance the reflector surface is from the LED is between 3” and 6”.
At 45 degrees to the LED, where most light is emitted, is 3.5”.
So even though it would have a lot of lumens the spot size would be larger than even a 150mm lens that collects 60 degrees of light, since the surface of the lens would be at about 4-6” distance (compared to the reflector being at 3-6”)
It would still be a cool light to build since nobody’s really done it before, and lots of lumens is nice, but tbh this will not come close to breaking any throw records even if I did use a 11” reflector rather than 5” like you were planning. (and yes, the reflector I calculated this with had a longer focal length (smaller spot) and collected more lumens compared to a 5” reflector that did 120-140 degrees)
Generally the actual luminous intensity (throw) only depends on the diameter of the reflector and the luminance of the LED. A 11” reflector will easily break all “records”. 10Mcd (based on a luminance of 200cd/mm^2) should be doable without any problems.
That only applies if you compare one reflector to another.
If you compare a reflector to a lens, it’s a whole different story, because the intensity of light projected at a distance is relative to both the distance from the LED to the lens as well as the angle the LED face is at from a point on the reflector/lens.
Assuming the wavien collar recycles almost all light, and the reflector collects almost all light, the 150mm lens on average will have it’s surface farther from the LED than the reflector would.
This means the lens will be making a smaller spot, while the reflector makes a bigger spot, both with a similar amount of lumens.
Therefore, the lens would have higher intensity.
I did play with that reflector and posted the results in this thread. It does not work much better than the miramax reflector unfortunately.
The focal length of a lens has no effect on center beam candela (throw). This has been proven (see Ras posts on CPF).
You can’t compare an LED with Wavien collar and a lens to an LED with a reflector. That’s comparing apples to oranges.
The Wavien collar increases the luminance of the LED.
It collects 75percen of the LEDs light and doubles the luminance (so you have 50 percent of initial light of the LED coming out of the collar).
A lens of the same diameter as a reflector only throws farther because it doesn’t have the hole in the center. So it has a bit more surface area, but it’s a small difference.
The focal length of the lens determines the size of the projected spot.
The longer it is, the smaller the spot is, and assuming constant F ratio, higher intensity.
Same lumens + smaller area = higher intensity.
If you don’t keep constant f-ratio, and the diameter stays the same, having a longer focal distance will STILL increase the lux.
I calculated it here: The "optimal" lens for use in thrower flashlights [mistake fixed]
Maybe you want to link that post you’re talking about?
The entire point of me comparing them is to point out that a smaller diameter lens+wavien collar will still get higher lux than a large recoil reflector.
Where did you get the 50% from? I was looking for tests that showed how many lumens you get when using a collar as compared to stock lumens.
If you can link a source that would be great.
This is based on the false assumption that all that matters is the front area of the optic.
I think you can see clearly that it is false by simply comparing an aspheric flashlight to a reflector flashlight with the same front area.
The throw will depend on the intensity, which is determined by a lot of factors such as LED spatial distribution, distance from the LED to the optic’s surface, angle between the LED and the optical surface (for a single ray) and also the reflection/refraction of the optic.
Jeroen, Just an idea to solve some of your heat related issues with the recoil setup. What if you use blue laser instead? So all you have to do is place tiny dot of whatever white emitting phosphor in the recoil - less heat to manage. Then shoot the phosphor dot with blue phosphor from some points (if you want more power with better efficiency) - easier heat management at laser diode(s). Cooling the phosphor alone is easier than cooling the phosphor PLUS the die. You can use any scrap LED to test as the phosphor point. Ideally the phosphor is deposited on reflective and high thermal conductive material such as polished AlN (found on LED chips) or silver.
- Clemence
I can’t remember which it was, but a car manufacturer made headlights like this years ago (they still may, I don’t keep up).
It was BMW. And now many projector companies…
Laser is far more efficient than LED they say
Well, the “laser” they’re using is still a diode, I think (at least most “laser” products are). But ,the laser diode has a narrower beam than lighting class diodes, so less is wasted.
Typical is 50, some already in the high 64. This is huge compared to 33% with the best LED