[Low_Rider]

San Jose, California - Philips Lumileds, the pioneer and the leader in high-power LED technology, announced today new performance records for high-power white LEDs. Philips Lumileds 1x1 mm2, chip based white LEDs, operating at just 350 mA, delivered 136 lumens for a light source efficiency of 115 lumens per Watt at a correlated-colour temperature (CCT) of 4685K. At 2000 mA, Philips Lumileds white LEDs delivered 502 lumens at a corresponding 61 lumens per Watt. These LEDs are the first high-power LEDs to break through the 100 lumen per Watt mark and demonstrate the real potential of solid state lighting technology.

Philips Lumileds achieved the record results for white LEDs by combining several new and innovative technologies it has developed. The first devices using these technologies will be introduced in a new generation of products during this quarter. These new technologies will continue to proliferate in new, and existing, products throughout the next 12-18 months.

Philips Lumileds breakthroughs in epitaxy, device physics, phosphor, and packaging technologies are critical to delivering the performance required of LEDs as they continue their growth into a preferred light source.

Here we go again, I'll take four please sir!! Official press release here. Discuss… :thumbsup:

Dave.

 

[presslab]

The 500 lumens at 8 watts looks very promising... Now the question is can they get the heat out of it. I have three OSRAM Ostar emitters to play with, which can also put out 500 lumens; at 15 watts though. However, the thermal properties of the Ostar is considerably better than the Luxeon emitters.

 

[achesalot]

I don't thing anyone really thought Lumileds would be down for the full count!
This race will be going on forever.

 

[Low_Rider]

A few thoughts:

- In terms of efficiency there was a lot of hype around the K2 before it's release, and while we did see an increase in light output when it was finally released, unfortunately in reality it was a hot little power sucker that didn't really live up to the hype.

- Lumileds have always shown impressive results with their phosphors in terms of efficiency, low colour shift, and even application on the die.

- While the lumen / watt numbers are incredibly impressive, I notice a rather low colour temperature verging on a "warm white" rather then a brilliant white. However colour temperature and tint are less important on the trails, you can get away with that.

- I really like the lumen / watt numbers at higher drive currents. This is where Lumileds will blow other emitters out of the market.

- What I like most though is the statement "The first devices using these technologies will be introduced in a new generation of products during this quarter". Now we're talking!

- It sounds like they will be integrating their new technology in to existing packages too, so hopefully we may see products that can be used as easy upgrades to existing systems, rather then having to go through the hassle of new mounting techniques, optics, reflectors and all that kind of stuff that we have to contend with when we use other current emitters. :thumbsup:

- Now we just have to play the waiting game to see what will actually happen. I have heard from a number of sources in the past the Lumileds have been working away for a while now, so hopefully we will see a quick release this time around. It's been a long while since they have released a significant product that's a "true" breakthrough.

- I want one now!!!

Dave.

 

[Low_Rider]

Now the question is can they get the heat out of it.

Time will tell, but lumen maintenance of Lumiled products at their rated drive currents has been excellent historically. It will be interesting to see how this new technology will be packaged. Perhaps implementation into existing products will see drive currents lowered a little, but I am sure Lumileds will have a package developed that can cope with the demands of this extra power.

I don't thing anyone really thought Lumileds would be down for the full count! This race will be going on forever.

I think everyone assumed that Lumileds would be working on something while the other manufacturers had their glory for a while. It could even have been a delayed release so that they could make the most of the market with their current range before releasing a new product. Their claimed release time frame certainly looks like that to me.

Hopefully we will see something tangible soon, I don't want to go through the waiting game we had with the K2 again…

And yeah I'd say the race will continue for a fair while yet. Increasing light output at higher drive currents is the current logical focus, as is continued developments in phosphor performance, efficiency, and application on the die. :thumbsup:

 

[[email protected]]

OH! MY!

I tell you what, LEDs are making their mark in a big way.

500 lumens at 8W could make for a l-o-n-g lasting light. I do wonder how big the heat sinks are going to have to be. If they can keep them diminuative it'll be beyond awesome.

 

[Low_Rider]

Forgetting the total wattage of the emitter for a moment, in terms of lumen per watt and the increase in efficiency, the emitter should actually be producing less heat per watt. Obviously these emitters have the capability of running at higher wattages, so adequate thermal management will still be an issue.

At the end of the day I think manufacturers will still prefer multiple emitters from an optical point of view, and in terms of efficiency it's still an advantage to have more emitters, but driven at lower currents. As a result I don't think there's going to be a need for bigger heat sinks, and perhaps we may actually see the opposite.

But if you were to push a single emitter or pair of emitters at their limits then perhaps you would see heat sinks a little larger then normal for a two emitter light, or at least around the same size of current lights around the same wattage.

Minimising resistance between the die, emitter packaging, thermal interface materials and the heat sink is the current focus, and it's this focus on getting the heat away from the die (even if there's generally less heat to deal with) that's the issue, rather then the surface area / mass/ dissipative properties of the heat sink itself.

It's going to be a very exciting year for the lightning industry.

 

[presslab]

The temperature junction-case of a K2 is 9 K/W, and light output is 85% at 90 C. The Cree XR-E is at 8 K/W and about 85% at 90 C, respectively. The Ostar is 5 K/W and 90% at 90 C, respectively.

If Lumileds can get the packaging down to get the heat out, they will really have something!!

The low color temperature is great, I think ~5000 K looks the best to my eyes.

 

[chrism]

Hopefully we will see something tangible soon, I don't want to go through the waiting game we had with the K2 again…

Which is exactly the problem - at the moment this is just vaporware, and given the history I'm not about to get excited until they at least announce production products, or even until they actually appear at distributors.

in terms of lumen per watt and the increase in efficiency, the emitter should actually be producing less heat per watt.

They might be more efficient than previous generations, but given the overall efficiency of turning electrical power into light, the vast majority is still going out in heat - I'd guess that you probably still have 80-90% of the heat output of a Lux 3, so the thermal management issues haven't changed that much. Certainly if you want to get anywhere near the figures you're seeing there, you will still need very good heatsinking (given that I'd imagine they are keeping the die to 25 degrees by supercooling the slug for those tests).

 

[presslab]

(given that I'd imagine they are keeping the die to 25 degrees by supercooling the slug for those tests).

I would assume these lumen spec's are 25 C die as well, as the previous products are spec'd that way. I believe they just quickly pulse the LED on and measure the light output before it has a chance to heat up.

 

[Homebrew]

The low color temperature is great, I think ~5000 K looks the best to my eyes.

I agree, the color sounds really good as long as the tint isn't weird. A less sterile light would work great on the trails.

I think if they can get to the point where a single LED driven with a single 18650 Li-Ion cell for two hours with around 160 lumens (around the same as two T-bin Lux III), I would be very satisfied. A self contained package that was basically a flashlight would be perfect. Anything more should go toward longer run time rather than brighter. Two hours would be a good starting point though. A medium spot on the helmet and a flood on the bars would be more than enough for riding full speed at night.

 

[chrism]

I would assume these lumen spec's are 25 C die as well, as the previous products are spec'd that way. I believe they just quickly pulse the LED on and measure the light output before it has a chance to heat up.

That's exactly the testing regime suggested on the thread over at CPF - I suspect whilst they can rate it at 2A it can't actually handle that much continuous current, so all the excitement over a 500lm emitter is a bit premature.

 

[[email protected]]

That's exactly the testing regime suggested on the thread over at CPF - I suspect whilst they can rate it at 2A it can't actually handle that much continuous current, so all the excitement over a 500lm emitter is a bit premature.

That's a little disappointing, really. It's kind of like weighing a bike without pedals or a saddle and proclaiming the lightest weight. Sure, it's light, but you can't actually ride it that way.

 

[brum]

That's a little disappointing, really. It's kind of like weighing a bike without pedals or a saddle and proclaiming the lightest weight. Sure, it's light, but you can't actually ride it that way.

I don't agree. With the cost of LED emitters, IMHO you would always want to use multiple. This way you can mix optics, have a larger surface area that draws heat away, and multiple emitters at a lower current will be more efficient than less emitters driven harder. However, with 500lm from a single emitters I'd imagine you'd want to build a very small (single emitter), yet extremely powerfull light.
I'm sure Lumileds will come with a product that competes directly with Cree, they are just too big (the biggest). Other than a lot of statements made at CPF thay say that Lumileds are dead if they dont launch a new SLED soon, I dont think they really have to rush (well, for my own projects they do), since the main market is general lighting, where the quality of the emitted light counts. The markets that require quantity of light from as little as possible emitters is really small. Unfortunatly, we and flashlightfreaks are on the latter market. All in all: can't wait! But XR-E's will do for now...

 

[chrism]

I think if they can get to the point where a single LED driven with a single 18650 Li-Ion cell for two hours with around 160 lumens (around the same as two T-bin Lux III), I would be very satisfied.

In that case, prepare to be very satisfied right now. Take a currently available single 18650 with 2400mAh at 3.7V, you have 8.8Wh. Allowing for regulator losses that is enough to drive a currently available U-bin SSC P4 at 1A (~3.6W) for 2 hours or so. At 1A, that U-bin P4 should give out at least 200lm, so even allowing for 10% degradation due to not keeping it cool enough and a 90% efficient optic, that's still 160lm out the front. That's a real world 160lm actually hitting the stuff you want to look at and a real world 2 hour run time, and all with currently available components.

 

[Low_Rider]

I suspect whilst they can rate it at 2A it can't actually handle that much continuous current, so all the excitement over a 500lm emitter is a bit premature.

While I'm not going to believe it until I see it, I don't think we're all that far off running higher currents like 2 amps continuous. My basic understanding from what I have read and played round with is that the current limitations in most of the latest emitters are in the bond wires (current capacity) and packaging (thermal management) rather then the physical die itself.

I definitely agree with brum on most of his points though, particularly from an optical point of view of having more emitters to shape the light. :thumbsup:

A bit of a news update for those who may not be lurking and reading at CPF and other resources, compoundsemiconductor.net has reported that Lumileds could release commercial products incorporating "new technology" before the end of March 2007.

According to Lumileds, the first commercial products to incorporate the enhanced device designs will be released before the end of March 2007, with many others set to follow over the next 18 months.

Solid State Lighting.net has also reported that a license agreement has been organised between Phillips and OSRAM.

OSRAM and Philips have signed cross license agreement covering optoelectronic semiconductors. The agreement involves the mutual licensing of patents for all inorganic and organic LED. "We expect this to put us in an even better position to use LED technology to serve the demands of the market," said Dr. Rudiger Muller, President and CEO of OSRAM Opto Semiconductors GmbH in Regensburg. The agreement relates to patents held by Philips, including the US subsidiary Lumileds, and by OSRAM including its subsidiary OSRAM Opto Semiconductors.

Cheers, Dave.

 

[GMF]

I definitely agree with brum on most of his points though, particularly from an optical point of view of having more emitters to shape the light. :thumbsup:

From a purely optical point of view, i'd have to respectfully disagree. With multiple point sources, you can get weird shadowing effects (although these are greatly diminished far field like we care about). Also multiple emitters won't give you a beam pattern a well designed single optic could (at least in the range of usable bike light). Multiple emitters translates into increased cost, weight, etc.

Admittedly, if you are talking DIY, and you are stuck with off the shelf optics, I definitely agree with you. If you have the luxury of designing your own optic, i'd rather have one super bright LED and one optic. Again, purely optically - the thermal efficiency points are all completely valid and essential to take into consideration.

-Damon

 

[Low_Rider]

I completely agree Damon, I should have distinguished between commercial and home made… Can you tell where my mind is at?

 

[achesalot]

The idea of having a single 500 lumen LED running at 2A is pretty scary! I think that sort of thing is a ways off though. But the idea of a 2x 500lu LED light on my head is sort of exciting. But honestly, it's going to be hard to get real exicted until I have the little devils in my hands... I really prefer a max of about 12 W flowing out my battery pack whichever way you wanna slice the pie. So it could be one 8W scorcher, or two "underdriven" at 6W each giving me around 700-800 lumen... .. Allen... Allen.. wake-up, it's only a dream

Although one light, one optic is probably ideal, I've never had any difficulty getting what appears to be a single beam from a double or Triple LED design. I suppose if you went with some really tight beam optics or deep reflectors you might notice the separate emitters, but in practice, it's not that big of a deal. You really don't want those super-narrow, flashlight-lover kind of beams in a bike light IMHO.

 

[GMF]

Although one light, one optic is probably ideal, I've never had any difficulty getting what appears to be a single beam from a double or Triple LED design. I suppose if you went with some really tight beam optics or deep reflectors you might notice the separate emitters, but in practice, it's not that big of a deal. You really don't want those super-narrow, flashlight-lover kind of beams in a bike light IMHO.

I totally agree with you. In practice, it is not a big deal for bike lights. My experience comes from designing colour changing LED products with reds, greens, and blues. Then, the colour mixing and strange shadowing is a huge problem.
I have, however, played with siamese optics (two LEDs in one optic to reduce size), and then you get weird an undesirable beam patterns. Basically, if you want a really bright LED light, you start getting pretty big because you are stacking up a bunch of optics. In my opinion, HIDs are on the big side as it is, so that is a disadvantage to LEDs

Back on topic, though, I'm not holding my breath for the next generation of lumileds emitters...