Another LED brightness breakthrough announced

[Paul_in_Maryland]

from http://chipzilla.com/?article=27731

White LED doubles efficiency

Japanese boffins make breakthrough


By INQUIRER staff: Wednesday 16 November 2005, 12:05



SCIENTISTS IN JAPAN have made a breakthrough in white LEDs which will mean much brighter units can be made. According to the Nikkei Business Daily, the method converts light using a purple LED and a silicon carbide substrate.

The wire said this converts light with much better efficiency, and will give brightness of 130 lumens per watt rather than the current limits of 70 lumens per watt.

Professor Satoshi Kamiyama of the Meijo University will establish a startup in January to manufacture and sell the LED units.

It will start capitalised at ¥40 million, but other manufacturers are likely to take stakes in the company too.

[jtr1962]

I'll take this with a grain of salt until I can actually buy the LEDs, but 130 lm/W next year would be great, especially given that the timeline to hit 150 lm/W was supposed to be in 2012. Even without this "breakthrough" it looks like we'll probably hit 150 lm/W two years sooner than expected. If this breakthrough pans out, combined with all the other improvements I've read about, we may get well beyond 200 lm/W by the end of the decade.

Now once these can be manufactured really cheaply it's bye-bye gas and glass (the derogatory term used by those in the LED business for incandescents and fluorescents).

[zespectre]

Wow, if this shows up in flashlights then I would think that variable intensity controls are going to be pretty much required or the lights will start to be unusable for close up work.

[Kinnza]

Wow, nice improvement.

Little info in the article, i understand is about no phosphor use at all. Anybody has more info about the tech involved?

[TorchMan]

Nice to see things are moving along, at least in the labs. I wonder about color rendition and tints in some of these.

[LowBat]

Please pardon my novice understanding, but I assume the purple LED will be able to be converted to other colors and white light as well. I also wonder if a 1 watt LED producing 130 lumens will actually do so without needing more power then a current 70 lumen LED.

[ledled]

Purple? hmmm......... any possibility of light reaching the UV spectrue causing eye damage??? As we look at lights a lot

[yellow]

I also wonder if a 1 watt LED producing 130 lumens will actually do so without needing more power then a current 70 lumen LED.

could this be, because the purple runs on 2 V like ordinary LED (instead of the 3.4 V of blue-converted to white LED)?

[LumenHound]

I'm with jtr1962 on this one. Until I can actually purchase one, it's just like vapourware to me.

In other words, lots of hot air.

It sounds great but hearing isn't seeing.

[LumenHound]

More info

High-Efficiency Nitride-Based Light-Emitting Diodes with Moth-Eye Structure

[Robban]

Please pardon my novice understanding, but I assume the purple LED will be able to be converted to other colors and white light as well. I also wonder if a 1 watt LED producing 130 lumens will actually do so without needing more power then a current 70 lumen LED.

I assume that it's working similar to current white LEDs in that the purple LED die is exciting a silicon carbide substrate layer on top of it which in turn emitts white light. Todays white LEDs are actually blue and has a phosphor layer. The phosphor layer is the yellow dot you can see if you look into an LED. If you look into a green one (or any other colour) you will not see this.

Please correct me if I'm wrong here guys. It wouldn't be the first time.

[beezaur]

This appears to be based in sound physics.

A while back there was an artcle in Science that made its rounds in trade journals (though often butchered there). The idea was that you could put holes in the substrate and control the amount of photons that were "born crooked" and got wasted as heat. Essentially the same number of photons (well not really photons, but that will work as a quick approximation) were created, but more were created in a direction that could escape the substrate. More light, less heat, all from a better shape.

This latest appears (at first glance; haven't read the full article yet) to be the same general idea going on here.

Scott

[LEDninja]

The method converts light using a PURPLE LED and a silicon carbide substrate.

I HATE purple tinted LEDs. They reminded me of the cheap Chinese LEDs of a year or 2 ago. Hard on the eyes they are. In the last 6 months most new offerings have a whiter tint. Now we are back to purple. WAAAAA!

-----

While LEDs can be made at 70 lumens/watt, usable light that actually gets out of a flashlight is around 50 lumens (Nichea CS). Assuming the same proportion of usful light can be gotten out of the 130 lumen LED gives 90 lumens/watt. Ummm..... a 90 lumen Proton.... or L1P premiium. I will be blinded!!!

[Kinnza]

Yes, but seems the improvement is related to silicon carbide substrate shaping, allowing more light extraction efficiency, so it might work with other tech.

[InfidelCastro]

This is great news. I wonder if it has anything to do with the breakthrough that makes the Photon Freedom Max possible?

[thesurefire]

I'll take this with a grain of salt until I can actually buy the LEDs, but 130 lm/W next year would be great, especially given that the timeline to hit 150 lm/W was supposed to be in 2012. Even without this "breakthrough" it looks like we'll probably hit 150 lm/W two years sooner than expected. If this breakthrough pans out, combined with all the other improvements I've read about, we may get well beyond 200 lm/W by the end of the decade.

Now once these can be manufactured really cheaply it's bye-bye gas and glass (the derogatory term used by those in the LED business for incandescents and fluorescents).

I concur. I will believe all the specs when I have a product in front of me that forfills them.

[TORCH_BOY]

With LEDS like these Flashlight modding becomes even more fun,

[cmeisenzahl]

http://www.treehugger.com/files/2005..._led_break.php

http://www.treehugger.com/files/whit...through-01.jpg
http://www.treehugger.com/files/2005...ental_inve.php


"It seems like the light emitting diode (LED) world is going from one breakthrough to the next. The last one was the accidental invention of warm white LEDs using quantum dots, and now a Japanese researcher at the Meijo University, professor Satoshi Kamiyama, has found a way to make white LEDs more efficient using a purple LED and a silicon carbide substrate. This new white LED has a brightness of 130 lumens per watt! "Normal incandescent light bulbs produce 15-20 lumens per watt; modern fluorescent bulbs produce between 60-110 lumens per watt; and current LED methods allow for a maximum of 60-70 lumens per watt. In short, if this is real, it's a big breakthrough." Professor Satoshi Kamiyama will establish a startup in January to manufacture and sell the LED units. He already has 40 million yen, but it is expected that other companies will want a stake in this."

[Opto-King]

Hello Paul In Maryland,
Thanks much for your information. Similar theory on better performance white LED is UV dice with RGB phosphor.
As you know, Purple dice & UV dice have bigger energy (shorter wave length) than Blue dice.

Another important factor for white LED shall be its lifetime. Many LED manufacturers tell thier 5mm white Lamp will exist only 9khours at 20mA with 25 celcius.

Please continue to update us later, if you will get more news on this Purple Dice used white LED.

/ Opto-King