Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Technology

Cause of LED Efficiency Droop Finally Revealed 308

An anonymous reader writes "Researchers from the University of California, Santa Barbara, in collaboration with colleagues at the École Polytechnique in France, have been able to prove the theory behind LED 'droop.' LED droop is the term for how LEDs emit less light when the amount of current being pushed through them goes above a certain level. 'The cost per lumen of LEDs has held the technology back as a viable replacement for incandescent bulbs for all-purpose commercial and residential lighting.' Now that we understand what causes this, we should start to see research go into technology to circumvent LED Droop. 'LEDs have enormous potential for providing long-lived high quality efficient sources of lighting for residential and commercial applications. The U.S. Department of Energy recently estimated that the widespread replacement of incandescent and fluorescent lights by LEDs in the U.S. could save electricity equal to the total output of fifty 1 GW power plants.'" A pre-print of the team's paper is available at the arXiv.
This discussion has been archived. No new comments can be posted.

Cause of LED Efficiency Droop Finally Revealed

Comments Filter:
  • multiply (Score:5, Insightful)

    by Randle_Revar ( 229304 ) <kelly.clowers@gmail.com> on Tuesday April 23, 2013 @09:22PM (#43532529) Homepage Journal

    >the total output of fifty 1 GW power plants

    Soooo... 50 GW?

    • Re:multiply (Score:5, Funny)

      by Anonymous Coward on Tuesday April 23, 2013 @09:26PM (#43532565)

      Witch!

      Don't you come in here and git all European-like and start a-quotin' your Systeme Internationale at me. I likes my units in New Imperial, and thats how's they're gonna stay if I done have anything to done say abouts it.

    • by Anonymous Coward on Tuesday April 23, 2013 @09:28PM (#43532577)

      Great Scott!!! 50 Jiggawatts!?!

    • Re:multiply (Score:5, Funny)

      by conorpeterson ( 2718139 ) on Tuesday April 23, 2013 @09:30PM (#43532593)

      >the total output of fifty 1 GW power plants

      Soooo... 50 GW?

      Or by my own calculation, 41.32 lightning bolts! Great Scott!

    • Re: (Score:3, Informative)

      by Anonymous Coward

      50 1GW plants do not equal 50GW as the output is not continuously at their maximum. It also doesn't equal one 50GW plant in the fact that one central plant would need a lot of infrastructure to distribute the power around the whole country. Since the economy would be very spread and not localized, it means 50 1GW plants distributed across the network.

      • Re:multiply (Score:5, Funny)

        by NemoinSpace ( 1118137 ) on Tuesday April 23, 2013 @09:58PM (#43532803) Journal
        In much the same way that 50 inane AC responses does equal the coherency of the parent.
        Invariably, like a game of "telephone", the subject switches to cars. -reading at -1 is the only true freedom. Moderation is an oxymoron.
      • Yeah that is true but it is a weird way to say it. Just give us the actual estimated GW savings. Or if you want a comparison, say "equal to 50 large nuclear power plants" (or medium or coal plants or whatever is accurate).

    • Re:multiply (Score:5, Insightful)

      by chrismcb ( 983081 ) on Tuesday April 23, 2013 @09:57PM (#43532795) Homepage
      Yeah, they say that like everyone knows what a "1 GW power plant" is... Is that a typical plant? Is it a small one? Is that the one you might see that powers a university, or one that powers a big city? Or is that about the size of the Manitoba Hydro Limestone hydroelectric generating station?
      Ohh look Mr. Wiki claims that 55 GW is about the peak daily consumption of Great Britain in 2008. Wouldn't that have been more meaningful to say?
    • or,

      50 states? 1 GW each? Stupid summary. Now I'll never know!

    • >the total output of fifty 1 GW power plants

      Soooo... 50 GW?

      "What the hell is a GW?!" Looks around, confused. "What?! I thought "jigawatt" was abbreviated JW? Isn't it easier to say 'almost fifty pre-Fukushima Fukushimas?'"

    • by xelah ( 176252 )
      slashdot....where multiplication is +5 insightful. The point, of course, is to tell readers who may not know that 1GW might be a reasonable size for a power plant. That way, journalists can copy-paste the press release instead of having to do any actual work, so it's more likely to end up in news publications. I think they fail, however, for sticking two photos together on the right, and writing the captions in to the images. Presumably, if it makes the general news, it'll be illustrated with a generic ligh
    • Well, I think more people can relate with the image of 50 power stations than the abstract number of 50 GW. I know this, because I am one of them.
    • by sulimma ( 796805 )

      Almost everywhere in the world this would be 50 GW. But the US use strange units. I don't know. Maybe:
      67 Million Horsepowers?
      47 M BTU / second?
      7E12 kilocalories / week?
      1.6E25 ergs / year?
      2.2 Tera foot pounds per minute?

  • Well, at least after I replace them the first time.

    Brain wandering time: If these are the kinds of lights we'll be putting in long duration spacecraft, it would make sense that they last at least a few decades, since it's cost prohibitive to bring supplies and, if we're talking about Mars, that's potentially a permanent installation (precluding "lunar module" style landers meant to take off later).

    • by ShanghaiBill ( 739463 ) * on Tuesday April 23, 2013 @09:55PM (#43532783)

      This is why LEDs are already used in traffic lights. If you look at the cost of sending out a crew, putting up cones, flagging traffic around the workers, etc., the cost of replacing a bulb can run into kilobucks. Even if the bulb itself is more expensive, it is far more cost effective to use LED traffic lights to avoid the traffic problems, labor costs, and safety problems of burnt out incandescents.

      • by macraig ( 621737 )

        Roundabouts don't consume any power at all or require maintenance. Isn't that greener? Maybe we should be skipping the lights altogether and retrofitting roundabouts? Maybe next we can get rid of all the ubiquitous street lights, which are scarcely a longer blip in human history than automobiles and consume far more power than traffic signals?

      • by Animats ( 122034 )

        If only they lasted. We've had LED traffic lights in areas of Silicon Valley for most of a decade, and many of them are failing, one section of LEDs at a time. There's a huge bin of partially failed traffic light round LED panels at Weird Stuff Warehouse in Sunnyvale. Mostly red and green; the yellows aren't on as much.

        The problem is outright failure, not dimming. That's an indication of a manufacturing problem, like a joint that fails under thermal cycling. [imsasafety.org] Many pre-2004 LEDs are prone to this problem

    • The magical never dimming and never failing LED is a myth or maybe a theoretical concept. Real ones fail and dim.
  • Over here in sweden it's hard to find a good old light bulb that will dim. But I guess I can live with the non dimming led bulbs for now
    • I recently replaced ten halogen GU10 type bulbs in my house with LED ones, and the website I bought them from stocked dimmable ones. They even had replacements for the twelve G9 bulbs that we have in one room, which is great as the non-LED ones blow at a rate of roughly one a month and are very inefficient. We have an energy meter, and our daily consumption of electricity has roughly halved since replacing all the old incandescent and halogen bulbs.
  • by Anonymous Coward on Tuesday April 23, 2013 @09:36PM (#43532649)

    Yo, genius, you misread the resistor markings when you wired up your Arduino circuit.

  • by Anonymous Coward on Tuesday April 23, 2013 @09:45PM (#43532721)

    It's because of Auger recombination. Basically, you stick in too many electrons, and they all mill around talking with each other instead of getting any work done. This is also known as the 'Water Cooler Effect'.

    • So, you're saying that instead of an electron falling into a hole causing a photon to be given off, the electrons are all huddled together elsewhere talking about the last episode of Big Bang Theory?
      • by N Monkey ( 313423 ) on Wednesday April 24, 2013 @12:18AM (#43533627)

        So, you're saying that instead of an electron falling into a hole causing a photon to be given off, the electrons are all huddled together elsewhere talking about the last episode of Big Bang Theory?

        No. They'd be discussing "Quantum Leap"

      • So, you're saying that instead of an electron falling into a hole causing a photon to be given off, the electrons are all huddled together elsewhere talking about the last episode of Big Bang Theory?

        And forming Cooper pairs.

    • It's because of Auger recombination

      You knew that! You should have told the University of California and France'X, they've been searching desperately for years.

  • The ability of marketers to convince people to overpay for absolute garbage, is eclipsed only by people willing to work for peanuts in the middle of the night when they should be sleeping.
    • And "grades don't matter outside school" is what they tell dumb people while they are fitting them for work boots.

  • Leave to our government to have a Department of Energy Energy Frontier Research Center. This just in from the Department of Redundancy Department we can reduce half the E from the DOE EFRC using LED's to become LEED certified says the CEEM.
  • by Anonymous Coward on Tuesday April 23, 2013 @10:13PM (#43532919)

    The little blue LEDs help me when I start to droop. Call your doctor if you don't stop drooping for more than two hours.

  • Pay very close attention to LEDs. Now that we've identified the root cause of one of our biggest problems, in a few years, we'll find ways to work around those problems and extend the lifespan of an LED (and output at higher drive currents) with a minimal loss of light.

    This is EXCITING news, as the uses for this across the entire electronics industry are MASSIVE. Higher-efficiency, longer-lasting LEDs means better optical devices and such, as this same tech can be applied down into solid-state laser diodes.

    I'm literally about to piss myself from this news. The sheer implications of this knowledge are astounding.

    I hope thermal pad and PCB makers are paying attention and prepare, because very soon we'll be pushing a LOT more power through these tiny LEDs, and we'll need the local cooling to compensate.

    I only wonder just how far they can defeat or mitigate this effect, and how. Thicker well walls might be an idea, or perhaps a nano-wire-like growth pattern, like we've seen with the recent development of microwires on graphite sheets, can increase the surface area and reduce the available recombination area, thus forcing electron transport.

    Something to either attract, guide, or force more electrons across the gap seems to be what is needed.

    • by phantomfive ( 622387 ) on Tuesday April 23, 2013 @10:56PM (#43533205) Journal

      I'm literally about to piss myself from this news.

      That's great, this is why I come here. What other site can you visit where people are more excited about a group of electrons than cute cats?

      I'll bet right now you'd rather experiment with electrons than with sex.

    • by Osgeld ( 1900440 )

      I got to play with some on the market now ceramic heatsinks that went though a reflow oven (~260c) and was able to be handled no problem the second it comes out (something you usually want to wear gloves for)

      • by Khyber ( 864651 )

        That 'ceramic' is still aluminum. Aluminum nitride and alumina mix, to be more precise

    • by am 2k ( 217885 )

      very soon we'll be pushing a LOT more power through these tiny LEDs, and we'll need the local cooling to compensate.

      Are you sure? Better efficiency also means that you need less cooling, as more power is emitted via visible light rather than heat.

    • by hyfe ( 641811 )

      I'm literally about to piss myself from this news.

      I would have been more comfortable if that sentence was at the end of your post, rather than at the start.. it now leads to two possible conclusions:
      1. You don't know what literal means, or
      2. You now smell funny.

    • by Twinbee ( 767046 )
      Good post, but I thought the main problem with getting decent LED light bulbs (for example) popular was the cost and that increasing the light produced too much heat. I for one would rather not see a fan-assisted bulb because of noise pollution. Do you think this discovery will help reduce heat to any degree for a given amount of lumens?
  • Basically, past a threshold current, it starts to emit electrons instead of more photons (look up "auger effect").

    Thanks again to "editors", illiterate both in English and science.

    • by c0lo ( 1497653 )

      Basically, past a threshold current, it starts to emit electrons instead of more photons (look up "auger effect").

      Thanks again to "editors", illiterate both in English and science.

      Actually, you need to look for Auger recombination [wikipedia.org] - a bit different than Auger effect.

  • "We were just stepping out to grab a bite to eat," said one LED, who asked that he remain anonymous. "We didn't realize anybody could see a difference. Terribly sorry."

  • Cost Per Lumen? BS! (Score:4, Interesting)

    by guttentag ( 313541 ) on Tuesday April 23, 2013 @11:53PM (#43533493) Journal

    The cost per lumen of LEDs has held the technology back as a viable replacement for incandescent bulbs for all-purpose commercial and residential lighting.

    Really? CREE started distributing LED bulbs a month or two ago through Home Depot for less than $10 each. [theverge.com] I own two of them.

    450 lumens for $9.97 is 0.0222 per lumen. It's rated to last 22.8 years. That's $0.0010 per lumen per year of use.

    Let's compare that to an "equivalent" (the cree is a 40-watt equivalent bulb) incandescent bulb [homedepot.com]. $8.77 for a pack of 6 is $1.46 per bulb.

    300 lumens for $1.46 is $0.0049 per lumen. But it's only rated to last 0.9 years. That's $0.0544 per lumen per year of use. It's more than 54 times more expensive than the CREE. That's before you look at the electricity you'll be saving (6 watts to get more light than you would out of a 40 watt incandescent).

    Home Depot is also selling CREE's 60-watt equivalent:

    800 lumens for $12.97 is 0.0162 per lumen.It's rated to last 22.8 years. That's $0.0007 per lumen per year of use. The incandescent is 77 times more expensive.

    As much as I love CREE LEDs in general, I prefer Philips 10.5-watt bulb [philips.com]. The bulb itself it more aesthetically pleasing (in my opinion) and it diffuses the light better (the CREE focuses all the bulbs in one area and its very apparent from the very bright spot in the middle). I own six of them. Home Depot sells them for $27.97 for a two pack [homedepot.com].

    800 lumens for $13.99 is $0.0175 per lumen. Rated to last 18.3 years. That's $0.0010 per lumen per year of use. If I'm going to spent the next two decades with a bulb, I'll spend the extra three hundredths of a cent per lumen on something I really like. Still less than one fiftieth the cost of an incandescent per lumen.

    The only things I see holding back LED bulbs are misinformation and lack of availability (Home Depot is the only major brick and mortar store I've found that carries them). That, and some freaky designs [philips.com] that don't look like light bulbs... I bought one of these out of curiosity, and its appearance, on or off, just irritates me for some reason... if I was redesigning my living room to look like Quark's, I'd go with these all the way, but since I'm not "that guy" it's in a lamp that I almost never use. Which means it will probably outlive me. It may even survive to the 24th century and end up in Quark's.

    • by FireFury03 ( 653718 ) <slashdot@nexTIGERusuk.org minus cat> on Wednesday April 24, 2013 @02:41AM (#43534131) Homepage

      The only things I see holding back LED bulbs are misinformation and lack of availability (Home Depot is the only major brick and mortar store I've found that carries them).

      I agree, except for replacing "misinformation" with "confusing information", and the manufacturers are responsible for this. Take for example the following photo:
      https://fbcdn-sphotos-a-a.akamaihd.net/hphotos-ak-prn1/11607_10151447644678611_203176319_n.jpg [akamaihd.net]

      We can see here, 2 GU10 bulbs. The one on the left is a 28W halogen, the one on the right is a 4.5W LED bulb (they both have a similar beam angle - 36 degrees for the halogen, 35 degrees for the LED). Both claim to be "equivalent" to a 35W "conventional" (by which I assume they mean tungsten) bulb. However, look at the light output - the halogen claims to output 600 lumen whilst the LED bulb says 200 lumen. So clearly different manufacturers use different criteria for what "equivalent" means - the halogen appears to be saying that its total light output is equivalent to a 35W tungsten, whilst the LED bulb appears to be saying that its brightness is equivalent to a (presumably unshaded) 35W tungsten. By the criteria used for the LED bulb, you could manufacture a tungsten bulb that is labelled as being "more efficient" than a tungsten bulb, simply by narrowing the beam angle with a reflector!

      Some of the bigger brands put even less information on their packaging - on the same shelves were Phillips 5W LED GU10 bulbs that simply gave an "equivalent to" figure - no information about how many lumens or candela they output, no information about beam angle.

      Also, people shopping for bulbs are almost certainly going to be doing like-for-like replacement: if I'm buying a GU10 bulb then the chances are I'm replacing an existing GU10 bulb, which is almost certainly going to be a halogen (since traditionally GU10s are halogen), not an unshaded tungsten bulb with an almost isotropic radiant flux. So telling me what "conventional" bulb it is equivalent to (whether thats done by comparing lumens or candela) is pretty much useless. Instead, I'm most likely to want to know what wattage of halogen its going to replace - if I've got a 50W halogen GU10 already and I'm buying an LED bulb, I want to know which LED bulb will give me the same results as the bulb I'm replacing.

      How is anyone supposed to make a decision when the information provided is either nonexistent or unstandardised and misleading?

      What I needed is for a standardisation of the information provided:
      1. The actual wattage of the bulb - i.e. how much power it is going to draw.
      2. The total light output in lumens.
      3. The brightness in candela. Especially important for bulbs that are traditionally used unshaded, such as GU10s.
      4. The beam angle. Again, important for bulbs that are usually unshaded.
      5. Colour temperature.
      6. What standard bulb this is equivalent to for a like-for-like replacement (i.e. if you're replacing a conventional ~isotropic tungsten bulb then it should be compared against that, if you're replacing a halogen GU10 then that bulb should be the comparison instead). Obviously this becomes problematic where the beam angles are different (e.g. I just bought a LED GU10 with a 120 degree beam - far wider than you'll get from a standard halogen GU10).
      7. The life expectancy of the bulb.

      And this information should be printed on *all* bulbs, even the conventional ones, so that someone in a shop can pick up any 2 bulbs and compare the information between them.

      I was under the impression that the EU had, several years ago, made some of this information (such as the lumen output) mandatory, but there are still a lot of bulbs on the shelves that don't include any of this data.

    • by Inda ( 580031 )
      The price stil holds them back, espeicially for the poorer citizens. You and I can see why buying a bulb for $10 that'll last 20 year is a good thing but when you're down to your last few bucks and you need a bulb, a $0.50 CFL is what you buy... and then you buy another one six months later. Maths doesn't come into it.
  • ...that we won't.

    From the summary: "The U.S. Department of Energy recently estimated that the widespread replacement of incandescent and fluorescent lights by LEDs in the U.S. could save electricity equal to the total output of fifty 1 GW power plants."

    Now, raise your hands, everybody who thinks we'll save energy. Nobody? OK, now raise your hands if you think what we'll actually do is use this as an excuse to leave the lights on 24/7, thereby expending the same or greater amount of energy, AND increasi
  • ... BMW's next year's models will be available with even more bizarre headlighting options.

Kiss your keyboard goodbye!

Working...