This is what I have observed so far without knowing much optics math - hope I’m in the ball park:
Same LED size, larger and deeper reflector (GT-FC40 in Convoy M21B-F-E): larger reflector has smaller hotspot (more “focused” with clearer border and smaller size) and less spill (smaller area and dimmer).
Same reflector size, with different LED sizes (Convoy M21B with B35AM vs GT-FC40): I am not sure about this one but the smaller B35AM LED has smaller hotspot. Not discussing output, just strictly the size of the hotspot. Also my Mateminco PD90S with the gigantic 11x10mm SFH55 LED has a very large hotspot despite of the deep reflector.
Combining small LED with deep reflector: great throw yes, but I believe this will result in small hotspot, limited spill. Perfect if this is a beam profile you’re looking for, but IMHO it’s unbalanced for general use. My favorite light? I prefer large hotspot and bright, large spill. GT-FC40 in M21F/E or SBT90.2 in Wurkkos TS30S. SBT90.2 doesn’t have a large hotspot but make up with 4000 lm output and 1 km throw.
A bigger and/or deeper the reflector just means that it collects more light being cast off by the emitter.
Hotspot is what comes off the sides, hits the reflector, and gets bounced forward, ideally all as parallel rays.
Spill is what comes straight out and misses the reflector entirely.
Throw is purely the result of surface intensity. Dr Jones did the “proof” in an old thread.
That’s pretty much it.
I like the HS-802 because of the nice deep reflector which catches more light, but also pushes the spill farther forward so it’s not lighting up my toes as I’m shining up ahead.
The bigger and deeper the reflector, though, the more critical positioning the emitter to maximise the throw. Not just centering it in the hole for the X- and Y-axes, but also sanding/shimming to raise/lower the chip in the Z-axis as well.
An LED sitting flush on the mcpcb vs “floating” somewhat on excess solder, can affect the focus.
All that twitchiness makes for a light that’d have horrible QC if being mass-produced.
Minor correction: a bigger reflector does not imply more light is collected: in fact, a large and shallow reflector collects less light than a small but deep reflector. What a bigger reflector does well is converting rays from a point source to near-parallel rays, the larger the reflector the closer to being parallel the rays are.
A really nice way to think about intensity is to imagine standing at the receiving end of the beam. What you see is a reflector filled with the image of the LED. Fixing the output constant, either an increase in reflector size or a decrease in emitter size results in the reflector making more "copies" of the LED, resulting in your receiving more light.
As Hoop said, the opening and depth of the reflector determine the angle of the spill. Secondly, the spill luminance is reduced as the emitter gets smaller.
Manufacturers have chosen this setup in the past. It’s very common in ‘tactical’ lights like Armyteks Dobermann (XHP35).
The key to throw is focal distance from LED to optic.
A large optic has a larger distance to the LED.
So with a very deep reflector with a relatively small diameter, the distance between the LED and the reflector is still relatively small.
Yes, with a very deep reflector the distance from the LED to the widest part is relatively large, but extra depth only adds a little bit to how much light is reflected.
This is not worthwhile beyond 1:1 proportions of depth and width.
In fact, most LED flashlight reflectors are approximately 4 wide and 3 deep (like 40mm wide and 30mm deep or 60mm wide and 45mm deep) and even less deep reflectors can be found in throwers.
Extreme reflector depth mostly makes the head taller and more expensive and heavier.
One benefit can be that the spill is as narrow as the corona, eliminating much of the rainbow effect.
And in all more light is cast into a similar direction.
But if you want a tight piercing hotspot you need width rather than depth, and the spill will be more useful because it will be wider.
A 250mm diameter reflector and a 25mm diameter reflector may have the exact same focal length. These two reflectors will have the same parabolic geometry, except in regard to where the parabola terminates, i.e. where it is cut off. The focal length of a parabola is the distance between the vertex and the focus point.
A reflector with a short focal length may have the same linear distance from the emitter surface (focal point) to the clear aperture as a larger diameter reflector with a longer focal length. Example
Not quite! The spill intensity depends almost solely on the total output of the emitter, independent from emitter size and reflector size.
As was previously pointed out, depth does not really matter for throw, what matters is width. However, deeper reflectors do tend to blend the tint nicely, while shallow reflectors tend to produce a fried-egg beam, Thrunite T2 being a particularly egregious example.
I agree to this statement, for this simple non-scientific reason:
I remember those rechargeable flashlights that has even lead-acid type battery power source at the bottom and a separate, very wide reflector yet very shallow with halogen bulb and a grab handle on top, sold usually in hardware stores. Commercial LED flashlights were not even available at that time yet.
I marvel how far they cleanly throw a piercing light to a distance.
If you take two reflectors of the same diameter, and one is taller than the other, (it has a shorter focal length) the taller one will produce higher lux on target, all other components being the same. The taller reflector has more surface area and also collects more light because the spill angle is reduced, which is to say the spill is being collimated into the beam. This also means that the taller reflector will throw as much as a slightly diametrically larger but shorter reflector. For a practical example: the S2 reflector, despite having a ~1.5mm smaller clear aperture than the S21A reflector, produced ~4% more lux on target in my recent tests at 1.3 meters.
Here’s an interesting situation:
The following optical configurations result in the same diameter and also the same amount of parabolic surface area:
That’s an array of seven particularly tall reflectors (this one from KD) vs 1 large reflector of the same overall diameter and with a focal length such that the total surface area of the parabola comes out the same.
The emitters used could be like so:
7 x Osram WF1 (~7mm sq) at 800 lumens each = 5600 emitter lumens
1x SBT90.2 (9mm sq) at 5600 emitter lumens
So these configurations have the same lumens output, the same diameter, and the same optical surface area.
The WF1 setup has higher emitter surface intensity.
Yes, but geometry and focal distance for a reflector are not the same thing.
Same geometry but very shallow reflector doesn’t have the distant part a deeper one does have.
With an aspheric lens you measure the focal distance between the LED and the flat side of the lens, but not with a reflector, of course.
And something that hasn’t been mentioned yet, is that hotwire bulbs throw light to the sides and even behind the emitting source (ie filament), whereas LEDs barely get 150° or so. Not even 180°, and certainly not 270° or more, like bulbs do.
So a wide shallow reflector might still throw a great beam for a bulb, but most of the light coming out of an LED misses it completely.
That’s true.
For an LED you can cut off the bottom part of an incandescent reflector, leaving you with an even shallower reflector.
I’ve done some experimenting with an aspheric to take over where the reflector comes to an end, but the obvious challenge is to mount the lens in the centre with the proper focal distance.
But your point about 150° or so is interesting too.
Because the light at those extreme (for an LED) angles is warmer and greener, giving us that lovely puke color corona we’re all so very fond of.
I have eliminated much of that vomit ring of light by using centring gaskets that protrude so that it shrouds that greenish light.
I also put a layer of white out on the inside of the centring gasket so that it is refected diffusely and it becomes part of the spill, which then will be slightly less cool.
Works pretty well, allthough in real life use a corona is also useful sometimes.
