Nichia Develops 60 Lumen Per Watt White LED

[jtr1962]

Do LED's run hotter than incandescents for the same light output? Would that be a problem (like halogen's in some light fixtures)?

LEDs need to have junction temperatures less than 100°C. This means the heat sink they're mounted to needs to be way cooler than that, probably 60°C tops, preferably way less. In short, there's no part of an LED that can get hot enough to start a fire. The LED would be dead long before then.

Halogens get hot precisely because they're inefficient. You need a ridiculous amount of power in a very small space to get enough light. LEDs and the fixtures they're mounted in will get no hotter than today's fluorescents.

 

[jtr1962]

I recently did a "cool" experiment with an LED. I thought the results might be of interest.

 

[ddrueding]

jtr, you are quite crazy. Cooling an LED to -44C in the name of efficiency? Madness. My kind of madness .

Going back to some other stuff: IIRC, a very steady voltage is requred to drive LEDs correctly. If I have a really nice 300W 12V source, whats the best way to use it?

 

[jtr1962]

jtr, you are quite crazy. Cooling an LED to -44C in the name of efficiency? Madness. My kind of madness .

:diablo:

Going back to some other stuff: IIRC, a very steady voltage is requred to drive LEDs correctly. If I have a really nice 300W 12V source, whats the best way to use it?

Actually, they need a constant current source because even a small change in voltage can result in a huge change in current. Not only that, but LEDs differ greatly in their forward voltage at nominal current, even among the same bin.

Anyway, the best way to drive LEDs off your power supply would be to put 3 in series (for white LEDs that would be about 9.5 to 10 volts), and then use a constant current circuit. This can be as simple as the circuit below:

In this case, you would have 3 LEDs in series instead of just one, and you would need to change the value of R1 to reflect the type of LED you're using. As shown the circuit will provide about 20 mA, which is nominal for 5mm indicator LEDs. For Luxeon type LEDs using 350 mA you would use an R1 of about 2 ohms, and an NPN power transistor such as a TIP31. For 12 volt operation the transistor probably wouldn't need a heat sink although it would get a bit warm. The LEDs would certainly need a heat sink since 3 of them would be dissipating about 3.3 watts. With your 12 volt supply you could potentially drive about 70 sets of 3 Luxeon LEDs, or 210 total. At 45 lumens each, that would be ~9500 lumens, about what a fluorescent with 4 32 watt T8 tubes puts out. Don't forget the sunglasses. :mrgrn:

 

[Stereodude]

Not a very efficient drive circuit though.

 

[jtr1962]

Not a very efficient drive circuit though.

It's quick and dirty though for someone who doesn't have much experience working with electronics. Also, since 3 white LEDs take 10 of the 12 volts, it's actually about 83% efficient under the circumstance I mentioned, which isn't horrible. If I were to drive a bunch of power LEDs off 12 volts I would make a few series strings of at least 20 or 30 (more if my pass transistor/MOSFET could withstand the voltage), and use a step-up constant-current switching regulator for each string. Efficiency could be >95%.

 

[Bozo]

OLEDs

http://news.bbc.co.uk/2/hi/science/nature/4906188.stm


Bozo :mrgrn:

 

[LiamC]

http://www.theinquirer.net/?article=32097

RESEARCHERS AT Tokyo University have succeeded in making blue LED components one centimetre across which emanate light uniformly across the surface.

 

[udaman]

I am bad

Alright, it's 1:12AM, day after the what's the # of yrs? of the date of the 9/11/2001 Terrorist attack on the World Trade Centers (unlike jtr, I lost no person I knew when only 3,000 people died that day....while an estimated 700,000+ die every year in China from pollution).

But be that as it may, I thought about all of jtr's LED zealot status over the years at CPF, SR & here . I look at all of his zealot proclimations about how LED's will save the world (Ok, a bit of an exaggeration, but jtr's excessively optimistic outlook on LED advancements are my 'model' of exaggeration ).

Was in CostCo this weekend, picking up my mother's monthy? $300 worth of pills for various senior citzen ailments...it's a very long story that would only make you depressed like Merc . I saw Sanyo Eneloop charger & 8AA 2000ma, 4 800ma AA NiNH batteries for $27. Not bad.

Then I recalled when I got banned from CPF in Feb 2005? (and again, when they goofed on an upgrade that allowed me back for one more post, after jtr sent me a link to a thread he was posting on in the 'Cafe'...hehe, damn it's so funny how they get excited on CPF, and I instantly got banned again!!!

So anyway, lets' look at current LED efficiency to see if jtr's optimistic claims have come true over the last few years, shaw we?

Last time I was on CPF, seemed like Luxeon 3watt LED's overdriven with massivce heat sinks on Mag D cell flashlight could get upwards of 180 lumens for a $$$ price.

Now I looked at http://www.batterystation.com/cr123a.htm

then :

http://www.lighthound.com/

and I see you can get single AA LED flashlights that output more than the formerly very expensive Korean owned/developed Surefire line, for much less that use either lithium ion recharable with protection circuits (recall even just a week ago, Dell laptop on fire with protected circuit Lithium batteries still failing, though there is a newer lower capacity CR123a Lithium Sulfur phosphate tech that's supposed to be fire proof ), or these great Sanyo Eneloop NiMH AA, or AAA rechargeables that hold their charge almost as long as a lithium battery, and don't cost much, can be used in digicams and all sorts of uses where standard AA, or AAA batteries can be used.

200 lumen output in a 2 cell CR123a for just $70 (made in China), not bad when you consider just 2 years ago, the top over driven 2 cell was only around 120 lumens, IIRC.

But I like smaller single cell LED sized flashlights. And they too are up in lumen output, almost as high as xenon bulbs in 2-cell lights that cost $25 or more. Get a single cell LED like those Xenix lines, for just over $50, and you have a long term reliable flashlight. Waiting for those 100+ lumen for 1 watt output. Do these new Luxeon Rebel premium LED's in the flashlights put out more than 100lm/w? Turbo mode, only can operate for 10 min. without damage, did I not read from jtr that cooling wasn't a problem, lol!???

DAMN, unlike Merc on caffiene energy drinks, I'm just about ready to crash into sleep...now!

Somewhere in the back of my brain, there's a post about digicams & macro lens fuction, and flowers and stuff...later

 

[jtr1962]

Do these new Luxeon Rebel premium LED's in the flashlights put out more than 100lm/w?

Look at this thread on CPF. Test results of the Rebel are in post 127. 88 lm/W at 350 mA, and holding above 50 lm/W even at 1.5A. The Q5 bin Cree XR-Es (107-144 lm @ 350 mA), which I have on order, should get into the mid-90s. Release of the R2 bin Crees, which hit 100 lm/W at 350 mA, is imminent. So yes, for all intents and purposes we're at 100 lm/W now, just as I predicted we would be this year. We'll probably hit 150 lm/W by 2010, two years ahead of schedule, and 300 lm/W actually looks possible eventually. Cree also has an LED in the lab producing 1000 lumens at an efficiency of over 70 lm/W. In short, I wasn't an LED zealot after all. LEDs are pretty close to obsoleting incandescent lamps in small lighting applications (<200 lm) right now as you've already seen. The Fenix line produces some great little lights getting well over 100 lumens. By 2010 incandescents for household lighting should be a thing of the past, with mercury-containing CFLs on the way out as well. However, HID and linear fluorescent will be with us for the foreseeable future, at least another decade or more. Hard to compete with linear tubes on price per lumen, at least for the time being. When the T8 fixtures reach the each of their life cycle in maybe 15-20 years they will ultimately be replaced by LED, but that's a long way off.

Anyway, the future looks bright! And thanks for bumping the thread. It's sorely in need of some updating anyway. When I first started it, 60 lm/W was a lab curiosity. Now you can get commodity white 5mm LEDs on eBay for less than a dime each producing 60 to 80 lm/W @ 20 mA. Interestingly, it seems 5mm LED efficiency has stagnated in the mid-80s at most, while power LEDs like the Cree and Rebel have finally overtaken them. Probably makes sense as 5mm epoxy-encapsulated LEDs only last a few thousand hours before dimming appreciably, whereas silicon-encapsulated power LEDs really can last 50,000 to 100,000 hours while maintaining brightness. Oh, and my long-term test Luxeon (currently at 32,000 hours run time) shows no signs of dimming.

 

[jtr1962]

Forgot to mention, as LEDs get more efficient the heat problem you mentioned becomes less and less. See my thoughts on the subject here.

 

[jtr1962]

The Q5 bin Cree XR-Es (107-144 lm @ 350 mA), which I have on order, should get into the mid-90s.

Typo. Should read 107-114 lm @ 350 mA, NOT 107-144. We're not at 144 lumens @ 350 mA quite yet. Probably sometime in 2008.

 

[Bozo]

I've noticed that the majority of tractor-trailers are running LED tail lights. With their long life, it probably cuts down on the number of citations for burned out lights.

Bozo :joker:

 

[ddrueding]

I've noticed that the majority of tractor-trailers are running LED tail lights. With their long life, it probably cuts down on the number of citations for burned out lights.

Bozo :joker:

It also prevents the whole thing from burning out at once, they just go one LED at a time.

 

[udaman]

^^^
Should be in the SR thread, but looks like (after having gotten over my hangover :drinka:, if I'm doing the math correctly ) dd's post above is the milestone 100k'th on SF...w00t! What? No fireworks gif smilies :tgif::beer:

Damn, I'm sorry, I don't have any long rambling thing to say other than that, lol.

 

[Will Rickards]

Saw a link to this LED research over at The Register.
http://www.uc.edu/news/NR.asp?id=7089

 

[ddrueding]

Saw a link to this LED research over at The Register.
http://www.uc.edu/news/NR.asp?id=7089

Salmon Sperm. Neat. I'm concerned about the introduction of biological material having a negative impact on longevity. Anyone have any informed (or not) thoughts on that?

 

[jtr1962]

Saw a link to this LED research over at The Register.
http://www.uc.edu/news/NR.asp?id=7089

Interesting. I noticed the green LED has the greatest improvement by far. This is likely because the best blues are already about 50% efficient, while greens are only around 14%. If we can get the efficiency of greens and reds to the same level as blues are, then a red-green-blue white light source of ~200 lm/W would be possible. White light based on mixing red, green, and blue can theoretically reach close to 400 lm/W. The best we can do using the current blue plus YAG phosphor to make white is around 250 lm/W (assuming a 100% efficient blue emitter).

Negative effects of salmon sperm? I doubt it. Everything is encapsulated in the LED package.

 

[Will Rickards]

Personally I was looking for the april's fool joke thing on it or something. But it looks legit. I also wondered about the longevity. If the DNA would break down from the heat or something over time. I wonder how much post processing they have to do on the sperm? I'm not even sure where on the LED this thing goes. They should have included a diagram or something.

 

[Bozo]

Good read about light polution.

http://www.newyorker.com/reporting/2007/08/20/070820fa_fact_owen

Bozo :joker:

 

[LOST6200]

A typical el Reg sophomoronic article. Why is thr4e no synthencit marcomolecuels being develped?

 

[jtr1962]

Nichia reaches 169 lm/W!

A few of their numbers are very interesting. The first sample was 161 lm/W and 9.89 lm at 20 mA. This implies a forward voltage of 3.071V and an input power of 61.4 mW. They mentioned the output of the blue chip used in this LED was 42.2 mW, hence a WPE (wall-plug efficiency) of 68.7%. This is the highest WPE I've ever heard for a blue LED.

The second sample is the one which achieved 169 lm/W. I found it interesting that they consider ~2.8V as nearly equal to the theoretical limit for the blue emitter used. Assuming similar phosphor conversion efficiencies as the 161 lm/W sample, this blue LED has achieved efficiencies of roughly 72%. In my opinion, most of the 28% which is not converted to light is probably simply not being extracted. In any case these lab results show that 75% to 80% emitter efficiency should be achievable in the near future, at least with blue. If this can be done with green and reds, then 300+ lm/W RGB is possible.

The power LED results were also interesting. At 134 lm/W they break Cree's record. The blue emitter has an output of 651 mW, for an overall efficiency of just over 60% at 350 mA. That's not much worse than their 20 mA results. In any case it looks like the 150 lm/W barrier will soon be broken for power LEDs, and we're closing in on 200 lm/W for indicator-type LEDs.

 

[udaman]

Nichia reaches 169 lm/W!

A few of their numbers are very interesting. The first sample was 161 lm/W and 9.89 lm at 20 mA. This implies a forward voltage of 3.071V and an input power of 61.4 mW. They mentioned the output of the blue chip used in this LED was 42.2 mW, hence a WPE (wall-plug efficiency) of 68.7%. This is the highest WPE I've ever heard for a blue LED.

The second sample is the one which achieved 169 lm/W. I found it interesting that they consider ~2.8V as nearly equal to the theoretical limit for the blue emitter used. Assuming similar phosphor conversion efficiencies as the 161 lm/W sample, this blue LED has achieved efficiencies of roughly 72%.


In my opinion, most of the 28% which is not converted to light is probably simply not being extracted.

Implying what exactly jtr? If they could 'extract' then do you not think it would have been done? Therefore, at present it is not possible to 'extract' more.

In any case these lab results show that 75% to 80% emitter efficiency should be achievable in the near future, at least with blue. If this can be done with green and reds, then 300+ lm/W RGB is possible.

"should" "if"...want to add anymore "maybes" to that list ? We can say the same for halographic storage which is what, a decade after the fact, no where near what was 'should' 'if' possible by the prognosticators like yourself for LED's?

The power LED results were also interesting. At 134 lm/W they break Cree's record. The blue emitter has an output of 651 mW, for an overall efficiency of just over 60% at 350 mA. That's not much worse than their 20 mA results. In any case it looks like the 150 lm/W barrier will soon be broken for power LEDs, and we're closing in on 200 lm/W for indicator-type LEDs.

Power LED, meaning they output enough light to be useful D sarcasm)?
Currently, doesn't Luxeon have a 3 emitter die that outputs 1000lm, or to put into perspective, a weak 60w incandescent bulb equivalent...without any of the nice uniform, soft light sans harsh shadows you get from LED, CFL, or HID?

They break Cree's lab only prototype record, lol. Let's see what each of these produce when they go into mass production, tables could be turned by then and at any rate 129 vs 134lm, no human can detect the difference, you would need a billboard (btw, what type of LED's are then using in the new billboards, if you know? They just started putting them up all over LA, Clear Channel, IIRC) array of thousands to notice that little amount of difference inefficiency/output level. Not sure you'd even be able to tell the difference in illumination levels, if you used those for a backlight on a large screen LDC TV.

300lm/w RGB, how do you make that jump? putting 3 dif, colors together, doesn't somehow magically cause an additive 100% increase in efficiency, lol. Cree's 129lm/w pinhole sized LED, will be nice for some electronics uses (backlit for laptop), and it will probably take 2+ yrs before they get them into mass production, such that the cost will be low enough to use in a laptop computer, sooner the better, though as OLED may overtake them by then.

200lm/w in 'useful' lumen outputs? Not by 2010, not even close. 250lm/w...5+yrs after that, 300lm/w doubt we will live to see that day, but it will be nice if I'm wrong .

 

[jtr1962]

300lm/w RGB, how do you make that jump? putting 3 dif, colors together, doesn't somehow magically cause an additive 100% increase in efficiency, lol.

Very easy-the spectrum is different. The luminous efficacy of the emitted spectrum of blue plus YAG phosphor white LEDs is about 330 lm/W. Furthermore, some energy is inevitably lost in the conversion of blue light to white light. These are termed Stokes losses. The least these losses can theoretically be in this type of LED is roughly 20%, meaning at most 80% of the energy from the blue LED is converted to white light. Even if we have a 100% efficient blue emitter, then we'll at best have a white LED with an efficiency of roughly 265 lm/W.

White made by mixing red, green, and blue emitters is a different animal. By carefully choosing both the center frequency and bandwidth of red, green, and blue emitters, we can end up with a emitted spectrum withan efficiency of up to 400 lm/W. Furthermore, there are no phosphor conversion losses here since no phospohor is involved. This means if we can develop 75% efficient emitters (something which many in the industry consider quite likely), then we can have a 300 lm/W white light source. Even at 50% efficiency, a figure already matched by some production blues, we can reach 200 lm/W once we develop green and red LEDs of similar efficiency. Read about it here.

 

[time]

Fenix flashlights are well-known in places like CandlePower Forums. But there's also a following for cheaper brands such as Ultrafire, and now Romisen.

I just received a Romisen E4 as well as the latest R4 model. The E4 takes either CR123A Lithium or 18650 protected rechargeable cells and has a 'clicky' switch in the end cap. The R4 only takes CR123A dells and has a traditional flashlight pushbutton switch on the barrel.

Both use Cree XR-E LEDs, however the R4 is significantly brighter as well as having a narrower spot and spill, making it more intense again. They both have 'orange peel' reflectors for enhanced spill at the expense of throw.

Both flashlights are regulated, so their light output remains fairly steady for most of the battery runtime.

With a 2500mAh 3.7V 18650 cell, the E4 is drawing less than 500mA, meaning a full-brightness runtime of more than 4 hours. Beyond that, reduced brightness would be available for an unknown number of hours.

With two 3V CR123A cells, both the E4 and R4 are drawing about 670mA, implying a near-full-brightness runtime of somewhere over an hour. That works out at 4W; if regulator efficiency approaches 75%, power to the LED may approach 3W. That's probably 180 lumens and maybe even 200.

In a very crude and dodgy test, my popular but pessimistic light meter says at least 5000 lux at 1m (that would possibly reach 10000 with a smooth reflector), with the spill 100 lux at 1m.

In practise - yes, it's extremely bright. Where my 6V lantern battery flashlight clearly spotlights trees at least 100m away, this thing lights up the whole damn tree - or so it seems in comparison.

My ensuite bathroom has an 8W GE tiny-spiral CFL rated at 470lm mounted in a spherical diffuser. According to the light meter, just aiming the flashlight at the ceiling produced about a third of the fixture's illumination.

Not bad for $16 ...

 

[Howell]

I'd like to find a good shaker flashlight like this.

 

[Will Rickards]

Not bad for $16 ...

Where do I order these $16 wonders? I didn't see the R4 listed on the site you linked.

 

[time]

DealExtreme

There are other HK suppliers, but I picked them as having the best customer feedback. In practise, they exceeded my expectations (I chose the EMS courier option). Note that the simple search on their website appears to be broken - use the Advanced search instead.

BTW, I notice that the R4 is actually $18.80. It's also available in black (which is what I ordered). Please note that it's too narrow to accept 18650 protected cells - you'd have to settle for 17670 or rechargeable CR123A if you don't want to use primary cells.

Also, I suspect most people would prefer a tail cap switch; they're much easier to turn on by feel.

The R4 appears to be the same size as the cheaper F4.

BTW, I didn't bother with AA cell lights because they usually lack either brightness or runtime.

 

[Will Rickards]

I want to keep a flashlight in my trunk. But it may not be in use for long periods. Is there any type of battery that will last in those conditions?

 

[P5-133XL]

Get yourself a hand-crank flashlight.

 

[time]

I realize Mark was being flippant, but a hand-crank flashlight is a really bad idea.

AFAIK, most - if not all - rely on internal rechargeable batteries. If not used for a long time, NiCd will self-discharge so much that they will be unable to hold a charge after cranking. Under such erratic cycling conditions, I'd guess that they wouldn't be much use after 3-5 years anyway.

On the off-chance the device uses a Li-ion cell, it will have lost most of its capacity after three years regardless of usage.

But the main show-stopper is the absolutely pitiful amount of light from the hand-cranked LED flashlights I've seen. You could well be better off with matches.

You're best off with a primary Lithium cell solution, with storage temperatures up to 60C and a claimed shelf life of 10 years (but not at 60C!). For flashlights, that will be either CR123A or AA cell(s). You also need either multiple levels of brightness or a unit designed for long runtimes. The Ultrafire C3 is a popular example.

 

[udaman]

Doubling the highest efficiency prototype LED's...is it commercially possible in the near term???

http://technology.newscientist.com/channel/tech/dn13266-crystal-coat-warms-up-led-light.html

The final LEDs were also better than commercially available LEDs at creating visible light, giving off more than 300 lumens of visible light for every watt of all light emitted.

"The reported values are remarkable," says In-Hwan Lee of Chonbuk National University in Jeonju, South Korea. But problems still remain, he adds. "Making the core-shell nanocrystals is quite difficult."

Yeah, may not ever make it to cost efficient manufacturing economy of scale, sort of like SED monitors are at present stillborn .


DOE study sheds light on commercial LED lamp efficacies
http://www.semiconductor-today.com/news_items/NEWS_2007/JAN_07/DOE_020107.htm

In October [2006] Cree Inc said that it had developed a white-light LED that produces up to 85lm/W. DOE helped fund the development of the LED, which can output as much as 160lm. White LEDs have since been demonstrated with luminous efficacies of over 100lm/W by South Korea’s Seoul Semiconductor (the P4, launched in early December) and 150lm/W by Japan’s Nichia in mid-December. By 2025, DOE’s goal is to achieve
160lm/W in cost-effective, market-ready, solid-state lighting products.

I guess the DOE & I are not quite as optimistic as jtr, 2025 seems about the correct ballpark guesstimate .


LED Lighting Fixtures shatters lamp efficiency record
http://compoundsemiconductor.net/blog/2007/11/led_lighting_fixtures_lmap_sha.html

prototype PAR 38 self-ballasted lamp. LLF's LRP-38 lamp set a new standard for energy efficient lighting by producing 659 lumens at a mere 5.8 watts of wall-plug power, resulting in 113.6 delivered lumens per watt. The LLF lamp would use less than 9% and 30% of the energy consumed by incandescent and fluorescent sources, respectively.

...We used Cree, Inc. XLamp and OSRAM Opto Semiconductors Golden Dragon products in the lamp

Used less than 9%, meaning this lamp is roughly the equivalent output of a 60w incandescent bulb...yawn. Only useful in directional luminaire or directional light areas, for unidirectional incandescent light uses this PAR lamp is mostly useless or inappropriate to clientele's desires. When will they produce one that replaces a screw in 100w or higher lightbulb, for those of us without 'cat like' low-light sensitivity eyes, lol?

Heh, at least they made it in a nice warm incandescent color, just the way jtr likes it... when he wants to romance some little Cantonese cuttie with 'mood lighting', that's not a techno-geek grrl, .

 

[ddrueding]

No pictures of that lamp, either.

 

[P5-133XL]

Here is a link describing LLF's current (not the prototype described above) LR6 product that I found interesting.

 

[ddrueding]

Thanks for the link, Mark. They put all that work into getting an accurate, consistent color balance, and then make it either 2700k or 3500k. Ugh. I've been looking into custom builds and having two dimmers, one for color temp and one for brightness. I like anything above about 4500k, and am right at home with 6500k.

 

[ddrueding]

Something more like what I was planning:

By the end of the year, LLF plans to introduce the LR4, a module designed for 4" (100 mm) recessed cans, and it is developing a 2' x 2' (610-mm square) recessed LED troffer for dropped ceilings. The latter is expected to provide about 4,000 lumens and consume 70 to 75 watts, according to Trott.

 

[jtr1962]

Thanks for the link, Mark. They put all that work into getting an accurate, consistent color balance, and then make it either 2700k or 3500k. Ugh.

Those were my thoughts exactly when I first read about this. 3500K might be borderline tolerable, but 2700K, fuhgetaboutit! That's positively nausea-inducing, irrespective of whether I'm with a Cantonese cuttie or not. There was some talk on CPF about bluish-tinted light being good for setting moods. One person didn't even think it would work well until he lit up a room with a few Nichias. Now he likes it. Trust me, dim bluish LED light is very relaxing, sort of like being under moon and star light. And unlike candlelight, you can still actually see things fairly well.

I've been looking into custom builds and having two dimmers, one for color temp and one for brightness. I like anything above about 4500k, and am right at home with 6500k.

That's kind of my hold grail as well-dimmability along with variable color temp. It seems since the LLF fixtures use a feedback mechanism to control color balance, then variable color temp should be pretty easy to do. I may write them an email suggesting this, or if not, at least suggesting that they put out 4100K, 5000K, and 6500K versions. They're ignoring a huge market here. Not everyone cares for or even tolerates warm light. I'm so used to 5000K or 6500K now that I cringe whenever I see an incandescent.

 

[jtr1962]

Doubling the highest efficiency prototype LED's...is it commercially possible in the near term???

http://technology.newscientist.com/channel/tech/dn13266-crystal-coat-warms-up-led-light.html

This came up on CPF a month ago. Bottom line is they're talking about a luminous efficacy of 300 lm/W, or in other words the efficiency of the emitted spectrum, not the luminous efficiency, or lumens out per watt of power in. A typical white LED actually has a luminous efficacy of about 330 lm/W. If it has a luminous efficiency of 80 lm/W, then the conversion efficiency is 80/330 or 24%, for example. Still, the concept looks promising. If we can make red and green emitters as efficient as the best blues (~50%), then we can make white light at about 200 lm/W. We can reach 300 lm/W with 75% efficiency, something I feel should be possible within a decade.

Yeah, may not ever make it to cost efficient manufacturing economy of scale, sort of like SED monitors are at present stillborn .

True, SED looks dead for now, but OLED looks like it's getting better all the time. Can't wait for a nice OLED screen with super contrast ratios and way less power use than even an LCD.

DOE study sheds light on commercial LED lamp efficacies
http://www.semiconductor-today.com/news_items/NEWS_2007/JAN_07/DOE_020107.htm

I guess the DOE & I are not quite as optimistic as jtr, 2025 seems about the correct ballpark guesstimate .

Umm, they seem to be talking about overall fixture efficiencies in the 160 lm/W area by 2025. Those would make economic fluorescent replacements. I don't expect LED to make many inroads replacing linear tubes or HID until at least 2020. I've often said that. However, long before then, probably in a year or two, LEDs will make lots of sense to replace incandescents, perhaps even CFLs. The 100 lm/W level, which we're at more or less, is the turning point. LLF's fixtures are already viable for certain commercial uses. Once we can get a screw-in bulb replacement to the $25 level or less, LEDs should make significant inroads. As with CFLs, mass production will quickly result in further price drops. At the $5 to $10 level, you're there. Honestly, progress to date has exceeded even my optimistic expectations. I think higher than expected energy prices are one driving force for this.

 

[udaman]

This came up on CPF a month ago. Bottom line is they're talking about a luminous efficacy of 300 lm/W, or in other words the efficiency of the emitted spectrum, not the luminous efficiency, or lumens out per watt of power in. A typical white LED actually has a luminous efficacy of about 330 lm/W. If it has a luminous efficiency of 80 lm/W, then the conversion efficiency is 80/330 or 24%, for example. Still, the concept looks promising. If we can make red and green emitters as efficient as the best blues (~50%), then we can make white light at about 200 lm/W. We can reach 300 lm/W with 75% efficiency, something I feel should be possible within a decade.

Hmm, now I'm getting confused again...maybe tea can explain to this dumb human . If typical white LED (blue driven) has a luminous efficacy of 330lm/w, wtf where they going on in this linked article about...

The final LEDs were also better than commercially available LEDs at creating visible light, giving off more than 300 lumens of visible light for every watt of all light emitted. This figure, known as the "luminous efficacy", is high compared to typical white LEDs.

luminous efficacy vs. efficency, I'm friggin confused:

http://hyperphysics.phy-astr.gsu.edu/Hbase/vision/bright.html#c2

To quote this link mid-page under luminous flux, which I'm sure you've seen recently ...efficiency is not reported as lm/w, efficacy is...so what the hell are the LED manufacturers or industry people, and everyone else reporting anyway, misleading data??? :

http://hyperphysics.phy-astr.gsu.edu/Hbase/vision/lumpow.html#c1

Don't suppose they also mentioned these LED streetlamps over on CPF (you jtr know I don't visit CPF anymore )?

http://www.bbeled.com/products.htm

168W
12,600 lm
High Power LED Streetlight

This is the largest panel, and it would seem it would have to be mounted pretty close the road, low height, to be useful. Couldn't possibly be used on a 'cobrahead' style lamp post that's say 50ft or higher above the road surface which is common, probably the majority even, of streetlamps on highways or main roads throughout the world, yes?

Note 400w HPS is 50k lm.

[URL="http://www.portlandgeneral.com/business/products/outdoor_lighting/lighting_equipment.pdf"]PGE APPROVED STREET LIGHTING EQUIPMENT LIGHTING SERVICES - 2007 R ...

[/url]


Maybe I'll have another brain fart here, just for fun:

Sheets of Cheap Carbon Nanotubes Now a Reality

nanotubes = nanocrystals, lol...non-sequitur? iPod Nano is so geek-hip

 

[Bozo]

CFLs, not all they are cracked up to be????

http://www.msnbc.msn.com/id/23694819/

Bozo :joker: