# how much brighter can a led get?

#### chillinn

##### Flashlight Enthusiast
Yep, unless the law of physics as we know them change, it's not too hard to educate onself on how scientists came up with that number.

Excellent retort. Now can I get some old timers that actually know a thing to explain that in fact the most efficent LEDs on the planet happen to be white LEDs, and how that could possibly be, romteb's personal laws of physics notwithstanding? kthx.

#### Tre_Asay

##### Enlightened
Yes it is true that you can't ever get more light / watt than theoretical green light. This is because light is energy and it doesn't come from nothing. In the case of a flashlight it is taking energy stored in the battery.
Watt is a fixed ammount of energy that can only be converted to a fixed number of photons. Even if 100% of the electrical energy is converted it will not ever make more than 1 watt of light for 1watt of electricity.
Lumens is just a measurement of brightness and green light looks brightest to the human eye (and not by a small amount , have you ever compared 1mw green and red lasers?)

As others have said the way to get brighter LED is to make them handle more watts not use less.

So sorry you will not be getting an AAA light that can put out 5000 lumens for an hour until you have some AAA battery that can put out 10+ amps for an hour.

Even with recycled heat the efficency will not get above 100%.

Oh yeah, wikipedia lists the theoretical maximum efficency of white LED as 260-300 lumens/watt.

#### raggie33

##### *the raggedier*
u all have great info everyone

#### chillinn

##### Flashlight Enthusiast
Yes it is true that you can't ever get more light / watt than theoretical green light.

Ok, was never arguing this point, about theoretical efficiency, but instead, ignoring theory, what is actually true in practice. I'm still waiting for the old timer heavyweights to chime in, but my understanding is no one is spending any R&D of consequence on the efficiency of green LEDs, mostly due to the fact that in practice, green LEDs are far less popular than white LEDs. Most of the advances in LED efficiency in recent decades has been in getting white LED's more efficient, leading to some unexpected (to those only considering theory) anomalies in practice, such that today's white LEDs are more efficient than, say, red LEDs, when theory and physics tells us otherwise. The point being is that far more effort is going into making existing, i.e. real, actual, white LEDs more efficient. We may never have the theoretical 100% efficient green LED, because no one really cares, no one needs them, thus no one is working on that. White LEDs, however, are very popular and there is an enormous interest in getting them as close to their theoretical maximum efficiency as possible. Thus it is very very possible that today's white LEDs are in fact more efficient than today's green LEDs. I don't know this to be the case, but it is possible, and I hope third time around my point is getting across: theoretically, green LEDs are more efficient, but today's advanced white LEDs are getting so efficient, they are surpassing the practical efficiency of actual colored LEDs, not theoretical ones.

#### FlashKat

##### Flashlight Enthusiast
How do you know that we are pretty close to peak luminous efficiency? Did you quote what we have now 10 years ago? You seem to know everything about LEDs. Please let us know your insight.
That's a totally different discussion from luminous efficiency. As LEDs become more efficient they can become much brighter because there is far less heat to deal with.

Heat is a problem. A 30lm/W LED putting out 100lm generates about 3W of heat. A 150lm/W LED putting out 650lm generates about 3W of heat. You get a lot more lumens for the same amount of heat. As luminous efficiency continues to increase you can continue to get more or more lumens without overheating the die. However, luminous efficiency can only continue to advance so far. We're pretty close to peak luminous efficiency.

#### StorminMatt

##### Flashlight Enthusiast
Excellent retort. Now can I get some old timers that actually know a thing to explain that in fact the most efficent LEDs on the planet happen to be white LEDs, and how that could possibly be, romteb's personal laws of physics notwithstanding? kthx.

I'm going to go out on a limb and guess that the most efficient LEDs out there are white LEDs because white LEDs are the most useful. Nobody is interested in a super bright green LED. So nobody is trying to design a super efficient, super bright green LED because there is no money to be made by selling one. White LEDs, on the other hand, are highly salable because they are used for all sorts of lighting.

It should also be mentioned that a green light source is NOT by definition the most efficient possible. From the standpoint of the laws of physics, you can theoretically have 100% conversion of electrical energy to radiant energy in the form of (broad band) white light just as you could to narrow band green light. Or, for that matter, red, orange, blue, or any other color. The only reason why a green light can theoretically produce more LUMENS per watt is that the lumen is not merely a measure of radiant output. It also takes into account the sensitivity of the human eye to the various wavelengths of visible light. So even if a green light cannot theoretically put out more radiant energy for a given input of electrical energy, it can put out more LUMENS because the human eye is more sensitive to green light than other wavelengths (ie colors) of visible light.

#### bykfixer

##### Flashaholic
I miss semi-man about now. lol

SureFire and Toyota are going the direction of green tint.
I don't know the science behind it but do know from the user end it appears brighter than white while being less harsh.

Casual users who like that pink twinged nichia say they loath the hint of green. But 10 years ago never would've tolerated the rosey tint either. Nope, back then they wanted pure white light bulbs, and now fashion wants tan (read incan twinged) LED's.

In 10 more years angry blue may be in vogue. lol

Last edited:

#### romteb

##### Enlightened
and I hope third time around my point is getting across: theoretically, green LEDs are more efficient, but today's advanced white LEDs are getting so efficient, they are surpassing the practical efficiency of actual colored LEDs, not theoretical ones.

No, theoretically a green light source is most efficient, not a green LED, the efficiency of LEDs at a given time is completly determined by the advances in substrates, phosphores and processes, there is no predicting wich led wavelenght or chromaticity will be more efficient, by the way, a white led is actually a blue led with a yellow phosphore on top.

#### KeepingItLight

##### Flashlight Enthusiast
Casual users who like that pink twinged nichia say they loath the hint of green. But 10 years ago never would've tolerated the rosey tint either.

<snip>

I am one who prefers the 5000K neutral-white version. The Nichia emitter in the BLF-348 flashlight is just about perfect. Back in the days of film, I spent a lot of time correcting 4500K color-casts in photos. In most photos, 4500K is bad news. I couldn't shake the bias now, even if I wanted to.

Of course, the attraction of the Nichia 219B is about more than tint. The version used in flashlights typically has a CRI (Ra) of 92. High CRI doesn't matter if all you want to do is change a tire or climb under the sink. If you are out for a pleasure walk, however, the enhanced color is part of the fun.

#### Stereodude

##### Flashlight Enthusiast
How do you know that we are pretty close to peak luminous efficiency? Did you quote what we have now 10 years ago? You seem to know everything about LEDs. Please let us know your insight.
Because the peak is likely around 350lm/W for a white LED with a reasonable CCT and CRI. There are commercial LEDs in the 200lm/W range. Cree says they have 300lm/W LEDs in the lab. There's not much further to go. Hence we're very close to peak LED.

There's a physical limit when converting electrical energy to light energy. Even if you make 0 heat with no other losses you can't get more than 1W of light energy out of a light with 1W of electrical energy in. It's pretty easy to do the calculation for a single wavelength green LED and get the 683lm/W number that's been bandied about in the thread and misunderstood by many (most?). For a white LED the energy is not concentrated at one single wavelength, so there is not a simple calculation. Some people ran a large number of simulations to calculate the maximum theoretical efficiency of white with different CCT, CRI, and emission spectrum and the corresponding lm/W values. Their data is here: http://www.photonics.com/Article.aspx?AID=28677 Please note for the efficiencies they report max and practical lm/W.

#### Stereodude

##### Flashlight Enthusiast
then the answer to your questions is "that there is no limit"
That's totally incorrect. One of the two questions he asked was, "will some day we get a 1000 lumens per watt?" The answer to that is absolutely no.

The answers to your question that site 'luminous efficacy" as a limiting factor are not really correct.
They absolutely are correct. He asked about luminous efficiency directly in the text of his post. The question in the topic, of "how much brighter can a led get?" is impossible to answer, but there probably is no hard limit.

We have to be very careful asking and answering questions like this in plain language. Physicists and engineers use mathematic notation to express these relationships. Plain language is not precise enough and has a bunch of fuzzy trap doors. Efficiency is involved, of course, but not exactly the way you think when discussing Lumens which are a measurement of 'the perceived' amount of 'luminous flux'. If you could convert all energy into your 'light engine' to visible light then the 'luminous efficiency' will go to 100% and you are well on your way to the worlds infinitely brightest LED.
Only if you have a 100% efficiency LED that generates no heat or has any other losses in it, an infinite amount of power to feed it with, and no problems getting an infinite amount of power to the LED's die. But even then it still won't have 1000lm/W luminous efficiency.

Excellent retort. Now can I get some old timers that actually know a thing to explain that in fact the most efficent LEDs on the planet happen to be white LEDs, and how that could possibly be, romteb's personal laws of physics notwithstanding? kthx.
There's nothing quite like announcing to the rest of the thread that the topic has gone *woosh* right over your head.

A lumen is not a measure of radiant energy but one of perceived brightness. Since our eyes are most sensitive to green it is impossible for any LED color or combination of colors to exceed 683lm/W even with 100% efficiency. That's not to say people are in a lab trying to make ultra high efficiency green LEDs (though there might be). It's fairly straightforward to compute the maximum luminous efficiency of a single wavelength light source that converts 100% of the electrical energy to light energy. Since white LEDs are not single wavelength sources it's much more complicated to determine the maximum luminious efficiency of a white LED that converts 100% of the electrical energy to light energy because "white" can be some many different specific things. I've already addressed this in another post in greater detail.

#### racerford

##### Newly Enlightened
I'm going to go out on a limb and guess that the most efficient LEDs out there are white LEDs because white LEDs are the most useful. Nobody is interested in a super bright green LED. So nobody is trying to design a super efficient, super bright green LED because there is no money to be made by selling one. White LEDs, on the other hand, are highly salable because they are used for all sorts of lighting.

.......

Because most "white" LEDs are Blue LEDS with a Yellow Phosphor over it, aren't companies by definition trying to produce super bright blue LEDs in order to produce super bright "white" LEDs? The exceptions would the 3 or 4 LED "white" LEDs that have a combination of Red, Green and Blue leds.

#### Lightmax

##### Newly Enlightened
The days of the Luxeon IIIs an Vs.....

#### chillinn

##### Flashlight Enthusiast
I miss semi-man about now. lol

I do that, too! When there is scientific dissent, semi-man used to refreshingly end the thread with authritay and something like a head kick.

#### chillinn

##### Flashlight Enthusiast
There's nothing quite like announcing to the rest of the thread that the topic has gone *woosh* right over your head.

A lumen is not a measure of ....

Woosh is right, stereodude. Allow me to enlighten us all on the difference between theory and practice. Are theoretical LEDs more efficient than practical LEDs? In fact, they are not, by definition. Theoretical LEDs do not actually exist. They exist in theory, but not in the real and material world. Practical, and thus existing, LEDs will always be more efficient than theoretical LEDs -- but not in theory -- in practice.

#### Bdm82

##### Flashlight Enthusiast
Woosh is right, stereodude. Allow me to enlighten us all on the difference between theory and practice. Are theoretical LEDs more efficient than practical LEDs? In fact, they are not, by definition. Theoretical LEDs do not actually exist. They exist in theory, but not in the real and material world. Practical, and thus existing, LEDs will always be more efficient than theoretical LEDs -- but not in theory -- in practice.
This reminds me of one of my favorite things a political science prof said... "A great theory that is only great in theory is not a great theory."

#### Stereodude

##### Flashlight Enthusiast
Woosh is right, stereodude. Allow me to enlighten us all on the difference between theory and practice. Are theoretical LEDs more efficient than practical LEDs? In fact, they are not, by definition. Theoretical LEDs do not actually exist. They exist in theory, but not in the real and material world. Practical, and thus existing, LEDs will always be more efficient than theoretical LEDs -- but not in theory -- in practice.
Thanks for getting all existential on us and trying to sidetrack things with nonsense mumbo jumbo. For your next act are you going to start pontificating how things traveling slower than the speed of light are actually faster than the speed of light because something travelling the speed of light is only theoretical? Or do you understand that theoretical limits are still limits as they concern the maximum speed of an object?

#### CelticCross74

##### Flashlight Enthusiast
think the thread title should be how much more efficient can LED's get. The answer to that is a LOT and its starting to happen now. I cannot even imagine what Cree will be making in 10 years.

#### easilyled

##### Flashaholic
Thanks for getting all existential on us and trying to sidetrack things with nonsense mumbo jumbo. For your next act are you going to start pontificating how things traveling slower than the speed of light are actually faster than the speed of light because something travelling the speed of light is only theoretical? Or do you understand that theoretical limits are still limits as they concern the maximum speed of an object?

Wasn't the USS Enterprise able to travel faster than the speed of light? :nana:

#### PeterRamish

##### Newly Enlightened
That's totally incorrect. One of the two questions he asked was, "will some day we get a 1000 lumens per watt?" The answer to that is absolutely no.

Please re-read my answer carefully.. You will see that I covered the issue that he asked 2 questions, not one, and I selected and responded to the question posed in the title. I will stand by my answer. He asked "how much brighter can an LED get".. the answer is that there is no limit. His other question is about a calculation of "watts per luminous flux" and is a simple conversion calculation, and I pointed out that is not really much for discussion there. It is just like asking "how many calories are in a gallon of kerosene?" There is one simple numerical answer, and once you have it there is no point in discussing it further.

One of my professors in civil engineering at Cal Poly opened his first day of class lectures with this statement that has stuck with me for better than 50 years, "In engineering I want you all to remember : That if you do not ask the right question you will never get the right answer.."