The LED technology used in that bulb seems a little dated. I believe it's using Phillips' Luxeon Rebel LEDs.
Looking at the pictures, there's 3 sections and each section has 6 Rebel LEDs. Total of 18 LEDs. All for output of 800 lumens.
A recent LED flashlight can produce 800 lumens with just 1 (CREE XM-L) LED. For example, http://www.fenixlight.com/viewproduct.asp?id=141 If you want a warmer white tint, it's a little less lumens, but you can always use 2 LEDs..
LEDs have come a long way in the last couple of years. It's just unfortunate that light bulb manufacturers do not make and sell new models as quickly as flashlight companies. There seems to be a new model every week with the latest LED for crazy light output.
Philips updates the technology in the Rebel LEDs several times a year. In fact, they have on several occasions demonstrated record-breaking output in this form factor up until earlier this year.
The reason that flashlights are updated so much more frequently than light bulbs is that flashlights don't require any consumer safety ratings. Where a company can release a battery-powered flashlight with no safety testing, a light bulb running on line voltage has to be brought through a rigorous UL, CSA and FCC standards compliance testing.
From the article it appeared that the selling point of the bulb was omnidirectional light at a certain power. In that case a large number of weak leds makes sense.
They aren't blue at all, they are 2700K - and this is a big part of the reason they are 57.8 lm/w. The remote phosphor, while creating a very uniform photometric distribution, reduces the overall light output by filtering the very monochromatic blue light into warm white light.
Does the use of connectors imply that users could replace failed LED boards instead of trashing the entire bulb? Or even upgrading to more efficient LED boards?
It means its possible, it doesn't mean that the owner could do it. It would be good PR on phillip's part to include a postage paid mailing label that would let you recycle dead bulbs. They could recover most of it and save both e-waste and the energy cost of casting a new heatsink.
The energy cost of individually mailing a heavy bulb is greater than the energy cost of casting a heatsink.
Centralized recycling might work, but not a postage paid mailer. Plus who is going to successfully store a mailer for 10 years while waiting for the bulb to burn out.
And in 10 years they won't want this bulb back, there won't be anything reusable in it. Just recycle it for the metal.
Storing the mailer isn't a big problem. When I buy a new bulb, I'll get a new mailer. I toss the first round. I don't disagree with you on the energy cost though.
My biggest reservation about buying LEDs now isn't the price; it's that the lights that will be available next year will be twice as good, and I can't stand the thought of discarding an entire bulb in order to upgrade. It makes no sense to me that the actual LED is inseparable from its quarter-pound aluminium and plastic shroud. A lightbulb should allow you to swap out some small component that includes the LED die and perhaps the driver.
When compact fluorescent bulbs first came out they did something like this.
The ballast lasts much much longer than the bulb, so you were able to remove the bulb while keeping the ballast.
It was a total failure. And in hindsight a terrible idea.
No two companies' bulbs were compatible so most of the time you were not able to find a replacement bulb even if you tried, and when you could find them they cost more than an entire new bulb!
The reason is that the extra cost of manufacturing them in a replaceable fashion costs more than simply tossing them.
I can confidently predict that you will never swap any components from a commercial LED bulb.
And BTW the "LED die and perhaps the driver" is probably 95% of the cost of the bulb. The heatsink and housing is nothing in comparison. There is nothing of value in them to keep. Just recycle them for the metal.
> "No two companies' bulbs were compatible so most of the time you were not able to find a replacement bulb..."
Standardization would be much easier for LEDs than for CFLs simply because of the size of the package. We have plenty of standard bulb bases that could potentially be used for LEDs, e.g. E-10 or PR-6. The biggest issue is ensuring enough contact with the heat sink.
> "The reason is that the extra cost of manufacturing them in a replaceable fashion costs more than simply tossing them."
There's no way to know this without seeing the manufacturing breakdown. The additional cost may be negligible.
> "The LED die and perhaps the driver is probably 95% of the cost of the bulb"
That may be true now, but I personally don't care. Given the choice between replacing an entire bulb and replacing a small piece for slightly less, guess which one I'll choose.
Furthermore, this situation will change quickly as LED bulbs are made in greater quantities. Ultimately the cost of the electronics and the LED will be cents. Think about what's happened to LED flashlights. The cost of the aluminum and plastic fixture is fixed.
Wow, you have a certain mindset and you don't want to change it.
> Standardization would be much easier for LEDs
No, it would be just as hard because LEDs are very new and the perfect format is not known yet. It'll be a decade or more before anyone can think of standardization.
> The additional cost may be negligible.
No, the additional cost will probably double the cost, and require extra material and energy to make it.
> That may be true now, but I personally don't care. Given the choice between replacing an entire bulb and replacing a small piece for slightly less, guess which one I'll choose.
Like I said, a certain mindset and you don't want to change it. Even if this is worse for the environment you would still prefer this? It's greenwashing, it provides no benefit, but it makes you feel better that you didn't throw something away.
You would pay more for the initial bulb, and the new component will cost more than a new bulb - so you pay more twice. And of course you are not paying for nothing, you are paying for extra material and extra manufacturing energy.
> Ultimately the cost of the electronics and the LED will be cents. Think about what's happened to LED flashlights. The cost of the aluminum and plastic fixture is fixed.
The plastic fixture costs cents. The expensive part is making it, not the raw material. This will be true even in some future where high power LEDs are cheap.
I'm not really wedded to a conclusion or "mind set" here. I just don't buy your assertions, which you seem to have pulled out of your ars :)
My proposition is pretty simple: I would like to see LED light bulbs adopt some of the same modularity used in other LED lighting products such as flash lights and bike headlights. I don't have to throw away an entire headlight in order to upgrade, and I don't want to throw away heavy aluminum light bulbs either. I have no idea if there's a market in this, but Nate's money is on the table.
Now, if you you really knew this subject well, you might have pointed out that cooling requirements are pretty specific to the type of LED and driver, and are likely to change rapidly, thus making current shrouds and heat sinks subject to imminent obsolescence. Or you might have pointed out the possibility that remote phosphors will catch on widely, which would seriously complicate the whole arrangement. Or you might have pointed out the fact that the economics of manufacturing mass-market lighting appears to favor arrays of cheap LEDs over single expensive LEDs, which would obviously make modularity difficult or impossible.
If indeed modularity would double the cost of a bulb, as you say, then that would obviously kill the idea as well. If you would be so kind, please provide a citation for this.
It's definitely not user serviceable. Indeed, the SMPS circuitry at the base of the bulb is going to fail before the LEDs do, and that stuff is all encapsulated in epoxy.
I wonder what the actual color temperature of the "blue" LEDs is— I've been hoping for a while that the rise of LED lamps would finally mean shifting to more natural colors.
It's kind of absurd to me that in this day and age we're still trying to make our artificial light look like another kind of artificial light rather than sunlight.
(For reference, daytime sunlight is around 5500k— these bulbs are marketed as 2700k, which is about the color of sunset, sunrise, or a large fire.)
The issue isn't so much the color temperature; it's more about overall color rendering. Black-body radiation, like we get from the sun and incandescents, provides the best color rendering across the whole visible spectrum, assuming a color temp somewhere near the middle.
CFLs aren't too great at rendering, because their phosphors emit very spiky output.
The cool (bluish) white tints of LEDs usually produce the most lumens. As you go to a neutral(5000k) or warm(around 2700k) white tint, the lumens output is lower. I assume they went with a cooler tint to get the most output.
LED production has improved a lot recently and now it's possible to select the exact color temperature of the LED you want to buy. Also, it's now possible to get the same output (800lumens) with only 1-2 LEDs and have a good color temperature without extra steps.
Always have... since the invention of controllable fire by nearly anatomically modern humans? I'm skeptical.
For the entire history of human evolution since the formation of the Earth, that color of light has meant it's almost, or just was, night, and was strictly transient. Just 400,000 years ago it also started meaning that it is night. And just in the last couple of hundred years, with industrialization, it's become the color that most of us see most of the time.
Now, I'm not saying that's necessarily bad— but you gotta wonder why.
My understanding was controllable fire predated anatomically modern humans. I remember one theory that basically said using fire allowed modern humans to evolve with a larger brain because fire aided digestion.
I heard this theory as well, in a history of science class. The professor claimed that you can extract something like 60% more food energy from an egg by cooking it.
Seems to me that humans who liked the color of fire, and by extension liked fire (or vis versa) did better than those humans who did not like fire (or its color).
AFAIK the blue LEDs used in these remote phosphor applications is 550 nm "royal blue" monochromatic chip; you can buy these chips on their own, and they are some of the highest energy output chips available on the market.
There's a big difference between cold light a neutral light. Personally I hate "warm" bulbs. I use the "natural" bulbs which to me appear not cool but neutral. I've read that asian nations also tend to use much cooler bulbs and dislike the warm ones, but I can't find the reference now.
Also, according to your Wikipedia article "Lighting levels that affect the circadian rhythm in humans are higher than the levels usually used in artificial lighting in homes".
Incandescent technology is still superior to any other lighting tech, especially solid state. When one factors in laser sintering of incandescent filaments that increases their efficacy to 3-5 times even LED's, incandescents absolutely rule when considering their performance and cost of production/recycling/logistics/..ect... I love bandgap technology, but it will be hard pressed to beat Edison's brilliance (it hasn't been done yet, despite articles such as this one).
It means nothing. LED lighting is still an infant technology. Any shortcomings are a moving target given how fast the technology is progressing.
The reference to filament sintering is incorrect. First, it's not sintering, it's a process applied to a filament after is has been drawn. Second, the efficiency gain is around double, which still can't even come close to LEDs. Look here for a bit more:
The reference to "production/recycling/logistics" is also wrong if it refers to the EROI of LED lighting, which has already bettered that of incandescent lighting:
This. In particular, I'm concerned by the little switchmode power supply on a non-user-servicable board potted into the depths of the bulb.
If that 5c capacitor dies, a $40 bulb gets scrapped even if everything else is still good. Hopefully Phillips have designed around that, but 20 years is a long time and they have little incentive to do so.
I suppose a proper, universal, standard for low-voltage DC in houses (ala 12v Halogen downlighting only universal, something like PoE) is probably too much to hope for.
I'd be interested in seeing what lightbulbs looked like if we designed them from scratch today -- without, as you note, 150+ years of expectations and needs.
What could they do that the traditional old incandescent can't? What would they look like?
On a more general note, does this "design in a vacuum" have any valid applications? Can this lead to ideas that we can use today, or is it just pie in the sky thinking?
It might look like this: http://www.adafruit.com/products/357 - Buy your lighting by the foot. 12V DC may make more sense for homes these days than 115 AC. The incandescent bulb was the "killer app" for a whole power ecosystem. In older houses in places like San Francisco you sometimes find wiring going through the old gas lamp lines.
> 12V DC may make more sense for homes these days than 115 AC
Have you ever sized any 12v wiring? To carry a load of 22amps 20 feet with a voltage drop of <2% with a 12V system requires 4awg wire. That's crazy big and gets very expensive.
22amps might sound like a lot and lighting may be getting more efficient, but you'll still want to run things like vacuum cleaners.
> To carry a load of 22amps 20 feet with a voltage drop of <2% with a 12V system requires 4awg wire.
The wire size needed to carry a load of a specific number of amps does not vary with the voltage between that wire and some other wire somewhere. Perhaps you mean, "To carry a load equivalent to 22 amps on 120 volts, a 12-volt system needs to carry 220 amps, which requires 4AWG wire to lose less than 2% over 20 feet."
> does not vary with the voltage between that wire and some other wire somewhere.
I'm not following what you're saying here. I was just pointing out that voltage drop can be a significant concern with low voltage systems and controlling that can require thick wires.
From my rough calculations, voltage drop at 12V over 20' of 4awg copper @ 20A is ~1.7% and @ 100A is ~8.6%.
Yes, you're right. I was wrong. Thanks for explaining. My calculations are below. Hopefully you have enlightened me; please let me know if I'm still confused.
4awg wire is ¼Ω per thousand feet (6dB of resistance below 10AWG), or 0.005Ω for 20 feet. 20 amps × 0.005Ω = 0.1 volts. So the voltage drop is 0.1 volts, which is 0.8% of 12V or 0.08% of 120V. If it's 20 feet each way, it's 1.7% or 0.17%.
The low-voltage gets you double: you need more amps to carry the same power (which is what I thought you were saying) and you also need thicker wire to carry the same current.
So, to take a worst-case example, if you have a 900W microwave oven that can tolerate a power supply 10% lower than nominal, and you're running it on 120V 20 feet from the breaker box, you can deal with 6 volts of drop on each side of the romex, at which point you need 900/(120×.9) = 8⅓A. 6V/8⅓A = 0.72Ω, which means your wire can have up to 36Ω per thousand feet, which would be 25AWG, which is about half a millimeter across and commonly used for Ethernet. Not safe, because you're dissipating more than a watt per foot, which might set something on fire, but the microwave will still run. (For safety you'd probably use 12AWG instead, which is 2mm across, if not 10AWG.)
If instead you're running a different 900W microwave off 12V, you need 900W/(12V×0.9), which is 83 amps. But you can't afford more than 0.6V drop on either side of the rails, and 0.6V/83A = 7.2mΩ, so you can't afford more than 0.36Ω per thousand feet, which is 5AWG, which is a copper bar 4.6mm in diameter, thicker than any wire in a normal person's house. Again, that's not to be safe — that's to get the microwave to work at all. You're still dissipating over a watt per foot of cable. To be safe, you need 2AWG, which is 6.5mm in diameter, and a pain in the ass to bend if you don't get stranded wire. 20 feet of copper 2AWG triple-stranded wire costs US$150, which is more than the microwave, and weighs 15 pounds.
Perhaps even more interesting would be "what would they look like if unconstrained by technology"? A lot of science fiction proposes the uniformly glowing ceiling. This would provide a very consistent & uniform light without the eye strain caused when you look directly at a point source.
Another common idiom in science fiction is to replace your windows with giant transparent television sets, functioning simultaneously as window, television, telephone, artwork and light bulb.
Lots of good suggestions here. GE has 'light panels' [1] and one of the things which I am sure we'll see for 'installed' lighting will be 'sky ceilings' that would be a ceiling covered with OLED panels arranged seamlessly which mimic the 'sky' when you are outdoors, providing night views with stars, twilight colors, etc. Probably won't be the 'common' implementation however.
Generally I think one thing that will survive will be 115V ac power distribution. Its got a lot going for it in terms of installed base and understanding its transmission properties. The means LED based lighting will always have some circuitry to rectify and regulate that voltage. Of course they may evolve so it rectifies it to 135v dc and then runs off that.
I enjoyed the psychology comment as well. There were some comments that gas didn't give that 'homey hearth fire' smell when burned back when people were switching from wood fireplaces over to gas.
Even without constraint, I think you will find lighting fixtures don't change too much. You've got area lights and directional lights. The designers will have fun with them, glowing coils, maybe pillows that glow. But a good reading light puts out a uniform beam of light that can be directed easily to your lap. A good area light provides a widely dispersed light without being impossible to look at. Those constraints won't change because the technology has.
"I'd be interested in seeing what lightbulbs looked like if we designed them from scratch today"
The problem is that humans have coevolved with both sunlight and fire. While the shape of the lightbulb may be arbitrary, I think the fact that they provide uneven lighting and cast all sorts of shadows is probably necessary for good mental health. (Because it provides more sensory stimulation, and more closely mimics the natural environment as compared with what you would get from evenly distributing LEDs across the ceiling.)
Well the bulb shape is really a relic from the past. The design has already been disrupted with fluorescent lights, that brought bigger cylindrical shapes that diffuse more.
But even parity was not achieved by CFLs, particularly for dimmable applications. It seems like the latest generation achieving parity (where the CFLs failed to) with the century-old technology is the notable point, more so than by how much they surpassed it.
I've done some searching for CCFLs and it seems very few people are making and selling them, and they only carry 2700K warm white. Can you recommend a good place to buy high-lumen CCFLs that have a more neutral temperature?
Required parity features: bulb form factor, warm white light, dimmable, runs off 115V AC source, screws in to existing sockets. Feature enhancement: long life.
Ummmmmmmmmmmmmm.... I'm pretty sure incandescent bulbs, being a pioneering technology, got to dictate four of those requirements (bulb form factor, triac-based dimmer circuits, 115 V AC, socket form). The final one, "warm white light", is not a feature incandescent bulbs had 150 years ago.
I wonder why people want 'warm white light' so much. It's not natural. Sunlight is in fact much whiter and not so yellowish as what incandescent bulb emits.
I agree with you. Moreover, the eye has a fantastic ability to filter out differences in illumination. If you could design an experiment where someone who had been illuminated by different color temperature lights could be asked what color temperature the light was without remembering what it was from when they were first exposed to it, I bet that people would have difficulty doing so.
It seems pretty natural to me; lower-color-temp lighting has been the standard for nighttime use for a least tens of thousands of years, possibly hundreds of thousands, depending on when exactly fire came into widespread use.
Maybe true, but most of the things you'd look at outdoors are greenish or brownish. Maybe the colors of the things people look at indoors are whiter or bluer?
Meanwhile the >75% reduction in power usage and >800% increase in lifetime are features that have surpassed the incumbent (which are what makes the $40 cost affordable - a feature they are no where near approaching..)
For all the complexity, it's a lot more efficient. Some of the complexity problem comes from adapting LED to design decisions made for incandescent AC lighting in the switches, dimmers, sockets, and power supply. If it were worth it for the user to rewire their house, they could likely get overall simpler and more reliable LED lighting.
There's a parallel to software design, one often incurs complexity when optimizing or adapting one software approach to another. The system might be more reliable overall if you can re-align the entire design, but there's a cost involved in it that might not make it worth it (particularly for physical things with replication costs).
ianaee, but the board from the led bulb looks just as hand soldered as the board from the cfl. I don't think the following argument holds up very well.
"I was quite impressed by the use of connectors in this bulb rather than relying on low-cost-labor for hand-soldering which has been used in the manufacturing of CFLs."
It is a plastic resin that has been blended to a precise dosage with YAG phosphor. You can buy YAG phosphor easily from wholesalers, for example, phosphortech, a Georgia company, sells a variety of colors of phosphor. However, the trick is to precisely and uniformly blend it in with the plastic resin - the reason that it takes so much time and development money to make a light bulb like this.
Looking at the pictures, there's 3 sections and each section has 6 Rebel LEDs. Total of 18 LEDs. All for output of 800 lumens.
A recent LED flashlight can produce 800 lumens with just 1 (CREE XM-L) LED. For example, http://www.fenixlight.com/viewproduct.asp?id=141 If you want a warmer white tint, it's a little less lumens, but you can always use 2 LEDs..
LEDs have come a long way in the last couple of years. It's just unfortunate that light bulb manufacturers do not make and sell new models as quickly as flashlight companies. There seems to be a new model every week with the latest LED for crazy light output.
Another example, http://www.fenixlight.com/viewproduct.asp?id=149 It does over 2,000 lumens with 3 LEDs!! This is just one company. There are many more that also have crazy flashlights.
If you want to learn more about LEDs or flashlights, head over to http://www.candlepowerforums.com/vb/forum.php