LEDs luminous limits reached?

Food for thought. Lumileds announced the K2 Led over a year ago now. It appears that the highest flux bin avaialble is U bin. Funnily enough, their U bin for K2 is 87.4 lumens min at 1000mA. Their Lux III U bins are 87.4 lumens min at 700mA. If you use their current factoring curve for K2, you get aprox 68 lumens at 700mA, which is about the same minumum flux as the Lux III T bin! I do not have any figures of production spread (obviously) so the average luminous intensisty on a production run may have increase due to better (tighter?) processes, but it does appear that we are approaching a ceiling in luminous intensity v's current at the moment. I'm sure that the V bin is a real design aim that is proving harder to acheive than imagined, hence the delay.
Obviously, the K2 is better as you can use smaller optics as the active led area is smaller, and also it will not dim with temperature as much as the LUX III which has a higher thermal resistance, so I am not knocking Lumileds acheivements. It just seems that the improvements we crave in this technolgy is not forthcoming and maybe a we have reached a limit without a fundametal change in materials/process. Does anyone have any views or technology updates they can share?

Cheers, MJR

The K2 is an entirely new design, so it's still in its infancy, whereas the Lux I/III are older more well understood designs and therefore more mature. I think you're jumping the gun thinking that the end of LED improvement is near. When the K2 is 4 years old and still not improving, then there will be something to think about.

True. Designs do mature. It's just that Lumileds launched the news to people about 18 months ago and we are still awaiting the high flux bins mentioned - I assume that they are not getting the yeilds expected (in fact the top device luminance seems to have regressed slighly). Any other opinions? Any other news on that high efficacy purple LED that was in the news earlier this year?

I am an LED person and i can't wait for the next step in LED technology but it is a little frustrating how slow it is progressing. Maybe it's the marketing people holding back. If the sales of the current LED's is strong then they will not release the new higher flux LED's just a theory.

leeleefocus said:

I am an LED person and i can't wait for the next step in LED technology but it is a little frustrating how slow it is progressing. Maybe it's the marketing people holding back. If the sales of the current LED's is strong then they will not release the new higher flux LED's just a theory.

Click to expand...

Until they double the output with the same power consumption, I don't see much point in getting excited. I mean adding 30 or 40 lumens to a 60 lumen light is just so ho hum, barely worth talking about to Mr. Eyeball. The current generation put out enough light for most of my uses , if they could match the tint of a surefire incandescent that would be a definite enhancement or perhaps if they started making arrays in a cone shape that would focus better or single die leds smaller. I'm stupid so don't bother telling me how silly these ideas are , I just think that there is alot of room for enhancement in the quality of light and delivering it that can make flashlights much better even before getting 150lm/w flux ratings.

I see occasional articles about major improvements in LED efficiency being achieved in the lab. This one claims the potential to improve LED output by a factor of 256. Don't know if it'll happen, or when, but I think there are things coming.

We shouldn't expect the "new advance every month" culture of computer/mp3/game player marketing to be true for more fundamental physics research like LEDs.

I too look foward to the advancement of led technology, and there will be advancements, but in the meantime look at the bright side, all of the cutting edge lights you now have will remain cutting edge untill the next phase or breakthrough, so its that much longer untill you have to upgrade or replace your lights!!

Hi everybody!

Back in the late 1960's when leds were first introduced we were 'soon' going to have led headlights, stoplights and all that stuff. When I used to buy electronics for a living in the 70's I remember we used thousands of Red Fairchild leds with an output of 0.2mcd, which was thought quite respectable at the time. Later I has an HP 'torch' as a free sample which had a couple of amber 'superbright' leds, and we all thought that was impressive. (It was a pretty usless torch though.) But it really has taken a long time from those dim red lights of 40 years ago to the time when I knocked my eye out looking into an LHS-p in 2003. I still marvel that a led can be made to give such a bright white light, but it was a very very long time coming. Progress is usually a series of bursts followed by quiet periods, and it could well be that we are entering one of those. We have had remarkable progress with the physics of leds lately, but it is unlikely that More's' law applies to them. The trade off between phosphor efficency, die size and current consumption etc may well give us some form of ceiling with the current state-of-the-art. Small tweak type improvements will of course be expected, but nothing stunning. More significant though may be the laws of supply and demand, as I suppose the main thrust for producing very bright leds in vast volumes has been from the mobile phone industry to back light their displays. But once they have enough light for their little screens (like now) there won't be much push for further development. I suppose led area lighting might happen, but the established alternatives seem very cost effective and frankly give a nicer light. So it remains to be seen if current led technology will remain anything other than a niche fashion product. As regards torches (sorry flashlights), sure we all want vast amounts of light, but for most purposes the current crop of 1 and 3 Watt leds are probably quite good enough for most people. And in the mass market where the volume business is, there's a lot of catching up to be done which might keep the likes of Limileds happy for quite a while before they feel the need to make any really significant improvements. So I am not holding my breath for anything dramatic soon, but hope to be proved wrong. Actually I wonder if the next bigh improvements might come from the direction of semiconductor lasers, which do seem to be entering the realms of science fiction.

I would not be looking to Lumileds for the "next big thing" in LEDs. The K2 still looks to be in beta and the marketing hype machine looks to be running out of gas to sustain it; they're simply not meeting their performance numbers.

I think another player will jump out in front, like Cree, Seoul, or Nichia.

Cree has not only been announcing laboratory advances in LED efficiency, but producing the last generation of lab results in actual product. While Lumileds is struggling mightily to reach 60 lm/W, Cree has had the parts on the market for quite some time and is starting to release even better parts. They recently announced a prototype that hit 113 lm/W ... in a fixture, after it had reached thermal equalibrium, verified independently.

LEDs have a long way to go. I'm not sure that any manufacturer has yet experimented with the "quantum holes" concept proposed years ago that should double die efficiency by allowing photons that would otherwise be trapped in the die to escape. LEDs are probably going to encroach on HID efficiency soon - their main limitation will likely be that of their limited power scaling.

idleprocess said:

LEDs have a long way to go. I'm not sure that any manufacturer has yet experimented with the "quantum holes" concept proposed years ago that should double die efficiency by allowing photons that would otherwise be trapped in the die to escape.

Click to expand...

And then you can start stacking dies for even more total output.

The best barometer of progress are the latest lab results since these indicate what is possible. Nichia reached 112 lm/W in April and Cree reached 131 lm/W last month. Eventually (meaning within a year or so for 5 mm LEDs and perhaps 2 years for higher powered ones) these enhancements will make it into production LEDs. This will mean LEDs at least twice as bright as present ones. Making predictions about further efficiency improvements is problematic at best. Maybe LEDs will hit a brick wall at 150 lm/w or so or maybe the progress will continue until we approach the theoretical limits of 100% conversion efficiency. Fact is we still have a tremendous margin for improvement. The best blue chips by Cree output 30 mW with a power input of about 64 mW, for an efficiency of 47%. The aforementioned 131 lm/W white LED is only about 40% efficient. I think we'll probably hit 60% efficiency in the near term (IR laser diodes are already there) with visible light LEDs. After that any further improvements might just be the result of long, slow process improvements. Or perhaps we'll reach 90% efficiency in short order and then the only improvements we'll see will be reduced cost and more consistent tint. No way to tell since predicting anything in the electronics industry is fraught with peril.

As far as Lumileds, they are hardly in the lead anymore nor should their products be used solely as a barometer of state of the art. Cree has had 1 watt LEDs available with efficiencies of 60 lm/W since last year. Seoul Semiconductor had a press release about a 98 lumen, 1 watt LED available later this year. Citizen has power LEDs with efficiencies in the 70 to 80 lm/W range. As efficiency improvements come along it will be possible to increase power input in the same package since less heat will be produced. Eventually we should have 10 watt, 1500 or 2000 lumen LEDs making no more waste heat than today's 3 watters.

Anyway, even if LED efficiency limits are reached soon, which I highly doubt they will be, we can look forward to cost and consistency improvements for a long time. My guess is eventually high-power emitters will be available for less than $1 complete with all the electronics needed to drive them straight off 120 VAC. That will of course be the last nail in the coffin of the incandescent light bulb. If nothing else, that is certainly worth looking forward to as the tungsten filament is one technology which has long outstayed its welcome.

as long as they measure the output in lm/Watt (and having 0.00000001 mA running through the led), they just suck

when output reaches WATT, then let´s speak of powerful led.

It appears Cree is having some troubles as it stock price has dropped almost in half over the past 3 months due to production problems.

http://biz.yahoo.com/ap/060713/cree_stock.html?.v=1

FYI

Do bear in mind that Lumileds has also been undergoing changes during that time. During that period, Philips acquired the rest of Lumileds by buying out the remaining portion that was still controlled by Agilent (which was all part of H-P at one time.)

Mark Richardson said:

It's just that Lumileds launched the news to people about 18 months ago and we are still awaiting the high flux bins mentioned -

Click to expand...

yellow said:

as long as they measure the output in lm/Watt (and having 0.00000001 mA running through the led), they just suck

when output reaches WATT, then let´s speak of powerful led.

Click to expand...

Yeah, I'd like to see lumen/watt ratings along with the actual wattage that the measurements were taken at.

My rapidly-failing memory tells me that when I joined CPF, Q-flux Lux 1s had the status (as in grab ur wallet) that U-flux Lux IIIs have now. They're getting better and the competition will ensure it continues.

Larry

Tomorrows Lights will be as Lasers.........

Thehappyman

These Nichia and Cree Led's mentioned earlier in this thread - do they have a small die area to allow good focusing (such as K2) are are they acheiving good heat dissapation via a large die size?

MJR

I've said this many times before, but I'll say it again.

White LED technology is only about 10 years old. Incandescent lighting (the biggest step since burning flames) has about a 100 year head start. While you probably won't be seeing cars with LED headlamps and police helicopters with LED search lights in the next couple years or so, I say it's still only a matter of time. In the grand scheme of things, LEDs are relatively new, and white LEDs are very new. I'm spitballing here, but take a look at computer technology 30 years ago, and compare it to computers (or even just PDAs or cell phones) today. Apply that same level of advancement to LEDs, and who knows what's possible.

thehappyman said:

Tomorrows Lights will be as Lasers.........

Thehappyman

Click to expand...

So you will have to be a doctor to use them legally