LEDs with filaments!

[Illum]

Japanese company Ushio Lighting came out with a 2500K Filament LED lamps for decoration/general lighting purposes...
At 0.6W per lamp they rated the output to be ~18lm...not too shabby but far from bright:duh2:





Chinese link: http://big5.nikkeibp.co.jp/china/news/elec/elec200901070126.html?ref=ML

Japanese Link: http://techon.nikkeibp.co.jp/article/NEWS/20090105/163584/

Up to now the LEDs we have seen only emit light in one direction, the LEDs used in these lamps emit light in all directions, I wonder how they managed with heatsinking. :thinking:

 

[LukeA]

Technologically it's an interesting achievement but it's not too practical. As you hinted, heatsinking will be terrible. That's why the flux is so low.

 

[Sgt. LED]

COOL!

 

[Kiessling]

That's actually very useful for quite a number of "design" lamps that need a pretty bulb. One less argument to avoid LED bulbs, sigh.

 

[jtr1962]

My first thought is WHY? LED completely changes the lighting paradigm. It shouldn't seek to imitate incandescent lamps. Indeed, I'm hoping the whole concept of screw-base lamps gets obsoleted eventually. Anyway, at the low output of 18 lumens these really aren't a practical lighting solution. They're more like an accent light.

 

[LLCoolBeans]

Anyway, at the low output of 18 lumens these really aren't a practical lighting solution. They're more like an accent light.

I think that's exactly what they are intended to be, accent lights. It's at least somewhat practical for that application.

 

[spencer]

The advantage I see to these is that they emit light 360 degrees. To achieve this with regular LED's you would need many pointing in different directions. I like it.

 

[EricB]

Wow! How do they do that? Aren't LED dies like little points? It seems the whole length is lit.

My first thought is WHY? LED completely changes the lighting paradigm. It shouldn't seek to imitate incandescent lamps. Indeed, I'm hoping the whole concept of screw-base lamps gets obsoleted eventually. Anyway, at the low output of 18 lumens these really aren't a practical lighting solution. They're more like an accent light.

That would also be good for chandeliers, and to better emulate faceted, frosted or opaque Christmas lighting (to spread the light around better so it's not concentrated toward the base). I wonder if those can be made rgb. And the higher color temperature whites would look interesting in a filament.

LED's with filaments, LOL! When they first made the so-called "rice lights", I thought they were LED's, until I saw the little filaments. Then, there was a bulb with several small clear rice lights in it, and I thought they were realistic looking 2700K LED's, until I noticed they were filaments.

 

[2xTrinity]

The advantage I see to these is that they emit light 360 degrees. To achieve this with regular LED's you would need many pointing in different directions. I like it.

In general I think the whole idea of emitting light in 360 degrees, then taking a lampshade and blocking out the vast majority of those angles completely stupid for general purpose lighting.

We should be moving AWAY from omnidrectional point sources, to fixtures that are inherently directional, and throw light where it's actually needed -- without wasting the vast majority of it in the process.

To take a design that is inherently both high efficiency and directional, and make it low efficiency and omnidirectional are a couple steps backward.

I Agree it's an interesting curiosity, as it relates to accent lights, but I agree with JTR in this case that I want to see fixtuers that actaully take advantage of LEDs' unique properties to make lighting fixtures that are better and more practical than what we've seen before on the incan and fluorescent front. To try to shoehorn LEDs into a role it was never designed for (similar to the whole concept of CFLs, but worse) IMO would be a shame. As unfortunately I suspect this will be the only option widely available on the market...

 

[Kiessling]

Fortunately, we're not solely composed out of reason and what we "must" and "should" do. There's more. Like omnidirectional light sources. And filaments.

 

[Illum]

heatsinking...refrigerator lamp anyone?

 

[PhotonWrangler]

I don't get it. Looks like a coated filament to me. :thinking:

 

[65535]

I imagine it is an unusual die design, where the die (which is just a semiconducting diode not anything special IIRC it's a pretty uniform material) and is simply formed into a long string shape (likely a long square cross sectioned rect. prism) then coated in a phosphor.

If my logic is right the only reason LED's are the way they are is that the dies are efficiently cut in squares and with the demand for heat sinking they must be mounted down.

 

[purduephotog]

I can think of a really easy way to do this- it's called heat pipes.

Coat a heat pipe with resist, put the LED material down, put the phase change material in... and there ya go. heat goes into the pipe, evaporates, condenses at the top, runs back down. Closed circuit, heat moves away, massive surface area available for lighting.

 

[Oznog]

My god.... the deviation from established process technology required to do this sounds staggering.

 

[rob85635]

heatsinking...refrigerator lamp anyone?

LOL, perfect, hehe.

Rob

 

[PhotonWrangler]

My god.... the deviation from established process technology required to do this sounds staggering.

Yep, that's why I'm skeptical about this product. I'm having a hard time imagining a thin, filament-like strip of semiconductor material. This would be so incredibly fragile that even a light shake would probably shatter it. It also makes no sense to me in terms of the "360 degree" light claim, when it's far easier to achieve the same goal with established methods such as LED arrays, mirrors and frosted globes.

One of those photos doesn't even seem real, the second one showing the "lit" version of the flame tip styled bulb. It doesn't seem to match the "unlit" photo of (purportedly) the same lamp.

:shrug:

 

[jtr1962]

My god.... the deviation from established process technology required to do this sounds staggering.

I think that's one of the biggest reasons I said WHY? You have to jump through all these hoops to try and make an LED which looks like a filament. What are you gaining in return other than having something which looks like an incandescent lamp? It'll likely be expensive as hell, not terribly efficient, fragile, and an omnidirectional light pattern really isn't ideal for room lighting. The 180° emission pattern of regular power LEDs is much more suitable. In fact, it's ideal for mounting on a ceiling. The overall goal of LED lighting has been stated many times to be efficiency. If LED lighting ends up taking the direction of imitating filament-based lamps to a tee, then this goal will be severely comprised.

 

[Kiessling]

Why does a LED need to be formed into a filament? Wouldn't it suffice to put lots of tiny chips ON the filament and then pour phosphor all over, the phosphor thus masking the sinlge chips and glowing uniformly?

 

[jtr1962]

Why does a LED need to be formed into a filament? Wouldn't it suffice to put lots of tiny chips ON the filament and then pour phosphor all over, the phosphor thus masking the sinlge chips and glowing uniformly?

Yes, that would work just fine actually. Even though I question the logic of this whole idea, I'd love to learn more about how they're made. This has to be using some real cutting edge production methods.