My Understanding of Emitter Die Size in Relation to Throw

LED Flashlights – General Info
1

The density is about the same (xhp35 hi at 490 lumen per square millimeter compared to xhp70.2 sliced dome at 509 lumen per square millimeter), the problem is the size.

To get long distance you need to make a beam of light as parallel as possible. If the light beam is cone shaped, of course it spreads out and you lose intensity. Lasers are very parallel which is why they go such long distances even though they don’t have to be very bright. So keep that in mind.

Now with a reflector based light, we have the reflector trying to bounce the side light from the emitter at a certain angle to bounce it forward. The key here is certain angle. The reflector shape has to be designed with the light coming from a single point. A point source if you will. If this point source were one atom in size and the reflector had perfect angles we would have a very parallel beam of light that would go a super long distance.

In the real world, the emitter is not a tiny point. It can vary in size quite a bit. The bigger the emitter die size, the more the angles on the reflector are not perfect (dark blue lines) and the less parallel the light beam is. Here is a good picture showing this. The red lines show a perfect angle.

This is why very tiny emitters are such good throwers. Their beams are more parallel (and hot spot sizes smaller)

The old xp-g2 has a die size of 1.29mm2
The xml2 has a die size of 1.97mm2
The xhp35 has a die size of 2.35mm2
The xhp50.2 has a die size of 2.94mm2
The xhp70.2 has a die size of 4mm2

The domes that are put on the emitters tend to magnify the die size by roughly 1.5 times. So an xhp70.2 with dome on is about 6mm in die size. This is why dedomed or sliced domed lights tend to give more throw. It gets rid of its inflated size and shows its true size, but you lose lumens.

Another way to shrink the die size is to make the reflector bigger. Now the die size seems smaller in comparison and you get more parallel beams. This is the key to the GT’s throwing ability.

I could go on concerning intensities, but will stop here.

2

1) “beams are more parallel” is called light collimation.
This only affects the size of the projected spot.
It does not change the intensity.

2) the intensity is linearly proportional to lumens/mm^2 (this is because the measure of lumens is defined as intensity over area)
Something like an XHP70 has far fewer lumens/mm^2 than an XP-G2, which is why it’s intensity is less.

3) shrinking die size and making a reflector bigger will make the beam “tighter” or more collimated, but the reason you get higher lux is not because the beam is more collimated.
If you shrink the die size without reducing lumens you end up with higher lumens/mm^2, aka higher intensity.
If you make a reflector bigger, the front area increases, and this is also proportional to the lux.
This is why the GT performs so well compared to other reflectors, we specifically designed it to have a large front area.

4) not all small LEDs are good throwers.
It is just more common that small LEDs are good for throwers because you can drive them at higher currents before reaching limitations of power and heat.
This means it is (in general) easier to get higher lumens/mm^2 on a small LED.

3

As Enderman stated, die size alone doesn’t account for throw. A large die will project a larger hotspot than a small die, but how far that hotspot throws is based on the intensity of the LED in question. Think of intensity as the lumens produced per square mm of LED die.

Some smaller LEDs, like XPG2 have higher intensity than larger die LEDs like XML2. That is why XPG2 throws better. Many large die LEDs have lower intensity as such an arrangement is usually more efficient. They can produce a lot of lumens with less power. But because the intensity is low they have less throw.

4

:+1:
To put it simply, throw (lux) is completely determined by the intensity of the LED + area of the optic.

Of course, some other things will affect it a little, such as the reflectivity/transmission of the optic and front glass.

.
.
The shape and size of the spot is based on the size/shape of the LED and the distance and position from the reflector, which is what the image you posted is showing.

5

I was trying trying to do a layman’s explanation of why the 70.2 didn’t throw as well as the 35 hi. I didn’t even want to get into intensities or lumens. :wink:

Do you think a newbie would be able to follow what I said or its too complicated?

I tried to see what the lumens per area was.

Xpg2 can do 1,400 lumen. 1.66mm = 843 lumens per square millimeter.

Xhp70.2 (hi) can do 8,000 lumen. 15.7mm = 509 lumens per square millimeter.

Xhp35 can do 2,700 lumen. 5.5mm = 490 lumens per square millimeter.

I think the xhp35 and 70.2 are a pretty similar match. I can see why the xpg2 is a good throwing emitter.

6

Well, what you described is how tight the beam is, not how much throw you get, so it’s pretty much just wrong…

It’s not hard to understand really, if you make something twice as big and it doesn’t have twice the lumens then the intensity is less, therefore less throw.

This is assuming that the LED have a shaved dome or no dome, but yeah it’s pretty close.
The XHP70.2 won’t get anywhere close to 9k lumens in something like the GT though, you need an actively cooled heatsink to get output that high.

If you want to know the true intensity of an LED (measured in cd/mm^2) you can find values here:

7

That’s right, with the dome sliced on the 70.2 the lumens will be less.

Let’s say 8,000 lumen. Then it’s 509 lumen per square millimeter.

Did you understand that I was not talking about lumens or intensities?

8

Dome slicing does not reduce the lumens that much, maybe like 5-10%.
The reason the XHP70.2 is worse than the XHP35 throw in the GT is because you cannot run it at 20A and get that many lumens.
There are heat and power limitations.
.
.

“The density is about the same, the problem is the size.”
“This is why dedomed or sliced domed lights tend to give more throw.”
“This is why very tiny emitters are such good throwers. Their beams are more parallel (and hot spot sizes smaller)”
This information is wrong.

I was correcting you, because throw has nothing to do directly with how large the emitter is, it all depends on the intensity, which is proportional to lumens per area.

So yes, you have to talk about lumens or intensity if you want to compare the throw of different emitters.
What you described in your original post is light collimation, which determines the size and shape of the spot, NOT the throw.

9

PS- if you use the XHP70.2 on a fan cooled heatsink and shave the dome, you can probably get the same intensity as the XHP35

With the dome on it is 2/3 as intense as the XHP35.

As you can see from TA’s tests in a real flashlight, the GT, the 70.2 still gets significantly less throw since it produces a lot more heat and uses a lot more power.

10

I want to remind everyone the title of this thread:
My Understanding of Emitter Die Size in Relation to Throw

So it may be completely wrong, but it’s what my understanding is.

This thread is also not aimed at scholars or professors to pick apart. (Enderman)

This thread is aimed at the average Joe who doesn’t know anything about led to reflector relationships. (Henningap)

So I hope the average Joe can learn a little more without having to spend hours or read an entire book on the subject.

Ps, I’ll try and edit my post to make it a little more accurate, without getting too technical.

11

It’s not a matter of having a “Scholarly” explanation. It’s that you drew an incorrect conclusion from what you described.

It is clear you put a lot of effort into your original post. You also gave in that post quite a good explanation of light collimation. The only problem is you drew the wrong conclusion from it. What you described governs the width of the hotspot, not throw. Having a wider or narrower hotspot doesn’t let you see further…. it just lets you see more at the same distance. But having a more intense (brighter) hotspot does.

If average Joe were to read your initial post only, average Joe would get incorrect info and could come to the wrong conclusion about what affects throw.

12

hey, what if we put 1 convex lens before the led, so that it focus the light right in the center of the reflector(we will have a very bright point)

13

I think I know what your describing, but the focused photons are traveling the wrong direction. We need a tiny point for the light to emit outward, not a tiny point of light being emitted inward.