World record efficiency more than 100 lm/W !!!!!!!

[ViReN]

Nah .... Not Kidding...

was browsing the site for car's Headlamps... and just bumped on the LED Site...found following....


OSRAM Opto Semiconductors' thin-film chip technology offers an efficiency of
more than 100 lm/W. Read on.. here


Thanks & Regards,
ViReN

 

[raggie33]

holly molly wicked insane cool i bet it has uglycolor though

 

[evan9162]

Hmm.

They didn't mention Vf, but it's easy to calculate. 108lm/W = 2lm / (0.02 * Vf)

Doing some algebra,

Vf = 2 / (108 * 0.02) = 0.93V

And yet, it's emitting at 617nm

Is it possible to emit at 617nm with a Vf of less than 1V?

 

[jtr1962]

Even more interesting is the wall plug efficiency. Assuming 277 lm/W for light emission at 618 nm, I calculate a wall-plug efficiency of 108/277 = 39.0%, which is comparable to that of a low-pressure sodium lamp, the most efficient light source currently known in terms of both lumens per watt and electrons to photons.

The forward voltage drop of 0.93V calculated above seems very low to me, and the press release says it outputs more than 2 lumens at an output power of 20 mA. I think they meant 20 mW, not 20 mA. At 108 lm/W, the LED would output 2.16 lm with an input power of 20 mW. Forward voltage drop is anyone's guess, but if it is around the 3.2V typical of AlGaInP then output should be well in excess of 6 lumens at a forward current of 20 mA. If these can be made in blue with similar quantum efficiencies this bodes well for white LEDs with record efficiencies.

 

[Doug S]

[ QUOTE ]
evan9162 said:
Is it possible to emit at 617nm with a Vf of less than 1V?

[/ QUOTE ]

Yes, using a device called an incandescent lamp /ubbthreads/images/graemlins/grin.gif
With an LED, I don't think so. Without looking it up, I am pretty sure that a 617nm photon has greater than 1 ev of energy so it doesn't look like a quantum effect device like a PN junction will do it.

100lm/W at 617nm? So what? As I recall you can do that in your kitchen in your spare time.

 

[evan9162]

If it is a mistake, and is only 20mW, then that's not too big of a deal. A 300x300 micron emitter is wasted on only 20mW of input power. That would be like running a white luxeon at 50mA, (Vf of 2.9V, 150mW of power) for about 7 lumens out, and claiming 50 lumens/watt.

 

[Doug Owen]

[ QUOTE ]
Doug S said:

Yes, using a device called an incandescent lamp /ubbthreads/images/graemlins/grin.gif
With an LED, I don't think so. Without looking it up, I am pretty sure that a 617nm photon has greater than 1 ev of energy so it doesn't look like a quantum effect device like a PN junction will do it.



[/ QUOTE ]

And again, other Doug, you'd be right. Both times.

617 nm photons are worth 2.01 eV each.....

Photon energy calculator

Doug Owen

 

[ViReN]

is there any posibility that we can use the White Phospor over RED Emmitter .... which could lead us to "Warm White" Light ?

ViReN

 

[jtr1962]

Nope. Phosphers can only reemit at longer (i.e. less energetic) wavelengths, so using red light to excite the phosphor would mean it could only reemit, well, deeper red. Even white LEDs based on blue LEDs are somewhat deficient under about 450 nm for the same reason, but since your eye isn't sensitive to such short wavelengths it really isn't that big of a deal. I tend to think the future of white LEDs will be either UV+RGB phosphor or simply mixing red, green, and blue in the proper ratio to produce white light. If we choose the latter approach then efficient red LEDs like this one will be needed.

 

[ViReN]

I once did try to mix up the 3 colors (with various brightness levels, to match the things up)... but it simply cannot get a *white* light ... it always some thing more or some thing less... so i presume that using RGB to get White .. is kinda difficult... and we might have to play with the *flickering* these LED's in a particular sequence so that our eye persives the light as *White* ....

UV Excited Phospor could have some answer... may be we mix R G B phospors in a proper proportion to emmit various light colors... Possibilities seem endless...

Thanks & Regards,
ViReN

 

[mattheww50]

Phosphors can only output lower energies, but phosphors are not the only game in town. Think about how a DPSS laser works. The input is IR, but the output is in the visible spectrum. There is a hefty efficiency price to be paid for that, but it does work. Molecular frequency doublers and triplers do exist....

 

[NewBie]

[ QUOTE ]
ViReN said:
I once did try to mix up the 3 colors (with various brightness levels, to match the things up)... but it simply cannot get a *white* light ... it always some thing more or some thing less... so i presume that using RGB to get White .. is kinda difficult... and we might have to play with the *flickering* these LED's in a particular sequence so that our eye persives the light as *White* ....

UV Excited Phospor could have some answer... may be we mix R G B phospors in a proper proportion to emmit various light colors... Possibilities seem endless...

Thanks & Regards,
ViReN

[/ QUOTE ]

You can get white from RGB LEDs with no problem. Done it many times. The difficulty is in when you get near white, your eyes seem to lie to you, as to which color to turn up and which one to turn down. I watched one fella fiddle with a setup for an hour, and the result was so so. Grabbed a color measurement unit, and in about 30 seconds had the most beautiful white you'd ever seen. Plus you could put it wherever you wanted on the black body radiator line.

A cheap alternative for this is one of those 50-250 dollar pucks used to set the white point for monitors and projectors.

 

[udaman]

[ QUOTE ]
jtr1962 said:
Even more interesting is the wall plug efficiency. Assuming 277 lm/W for light emission at 618 nm, I calculate a wall-plug efficiency of 108/277 = 39.0%, which is comparable to that of a low-pressure sodium lamp, the most efficient light source currently known in terms of both lumens per watt and electrons to photons.

The forward voltage drop of 0.93V calculated above seems very low to me, and the press release says it outputs more than 2 lumens at an output power of 20 mA. I think they meant 20 mW, not 20 mA. At 108 lm/W, the LED would output 2.16 lm with an input power of 20 mW. Forward voltage drop is anyone's guess, but if it is around the 3.2V typical of AlGaInP then output should be well in excess of 6 lumens at a forward current of 20 mA. If these can be made in blue with similar quantum efficiencies this bodes well for white LEDs with record efficiencies.

[/ QUOTE ]

Hmm, 6 lumens you say? Well that's not too interesting as far as flashlight output levels. So if they come up with world record 1000lm/w efficiency on a microscopic level it's not very interesting to me. When they get these efficiencies into commercially available product in the 1w or greater range, then I'll pay attention. Don't hold your breath though, could be quite a few years!

BTW, in 1998 Japanese researchers have done experimentally 1000w (for commercial greenhouse applications) HID at 180lm/w efficiency. But as is the case here AFAIK, you can't buy it, just a research project at present. If it cannot be cost effectively produced for the current market conditions....I'm going to get pretty fat eating all this popcorn waiting for these things /ubbthreads/images/graemlins/smile.gif /ubbthreads/images/graemlins/popcorn.gif /ubbthreads/images/graemlins/drunk.gif

 

[NewBie]

I wish folks would include ballast efficiencies for the Sodium and HID for their lm/W....

 

[jtr1962]

[ QUOTE ]
udaman said:
Hmm, 6 lumens you say? Well that's not too interesting as far as flashlight output levels. So if they come up with world record 1000lm/w efficiency on a microscopic level it's not very interesting to me. When they get these efficiencies into commercially available product in the 1w or greater range, then I'll pay attention. Don't hold your breath though, could be quite a few years!


[/ QUOTE ]

This part of the press release might interest you:

[ QUOTE ]

Another advantage of surface-emitting LEDs such as OSRAM's thin-film devices is the scalability of output power as a function of chip area, at comparable levels of efficiency. Large-area high-flux devices can be operated at high operating currents while maintaining the efficiency of small LEDs designed for small currents.


[/ QUOTE ]

Of course, I wouldn't hold my breath either waiting for these to get into production, especially with larger die areas and outputs.

[ QUOTE ]

BTW, in 1998 Japanese researchers have done experimentally 1000w (for commercial greenhouse applications) HID at 180lm/w efficiency.

[/ QUOTE ]
That certainly has my interest-current 1000W HIDs max out at around 110 lm/W (I think that figure is for the lamp alone, not lamp + ballast). In fact, if produced these would give LEDs a run for the money for quite some time, including the ones producing white light at 150 lm/W which are supposed to be ready by 2012 (once again, I wouldn't hold my breath waiting for that).


BTW, I think we know each other from another forum. /ubbthreads/images/graemlins/wink.gif Who else but the "other" udaman I know who use Jessica Alba for their avatar? /ubbthreads/images/graemlins/grin.gif