Cree demonstrates 70lm/watt

but is that with WHITE led? or just red?

The article indicates that it's for a white LED.

:thumbsup: Cool! Let's hope we don't have to wait too long for these!!

McGizmo said:

:thumbsup: Cool! Let's hope we don't have to wait too long for these!!

Click to expand...

Note they are doing it within the same XL7090 package, so it is a minimal change to everything.

70 lm/W- thats 70 lumens with 1 Watt of power. This better than it first appears. The thing that is not obvious to everyone, is that when you increase the efficiency of the LED, is that you also get less heat generated by the LED.

Another datapoint, 86 lumens with XLamp 7090 LEDs in development, running at approximately 1.229 watts.

How far out? I'm told under six months.

Quick question:

How do thermal resistance ratings change when efficiency increases so much? Let's say you go from 40lm/W to 80lm/W. You've reduced the amount of power that's wasted as heat - so in theory, the thermal resistance rating should go down.

Let's say the light you're producing has a maximum efficacy of 300lm/W.

Going from 40lm/W to 80lm/W takes you from 87% power going to heat, to 73% power going to heat. So for 1W of input, the less efficient dissipates 0.87W to 0.73W for the more efficient.

If the thermal resistance for the less efficient was rated at 17C/W, then the "real" thermal resistance is something like 20C/W (since not all of the power is going into heat).

So, the more efficient LED should now have a thermal resistance rating of ~15C/W, since more of the power is going to light instead of heat.

Is is possible that the Luxeon III LEDs are really built the exact same way that a Lux I is, but that the Lux III LEDs are sorted to be more efficient, thus having a "lower" thermal resistance?

So 'real' thermal resistance is constant - that is, power dissipated as heat always causes the same amount of junction rise per watt.

But

overall thermal resistance depends on LED efficiency. The more light per watt created is less power dissipated as heat, and thus, less input power is converted to heat, causing lower temperature rise per input watt. So without changing packaging, but by simply improving LED efficiency, you improve thermal resistance.

Make sense?

evan9162 said:

But overall thermal resistance depends on LED efficiency. The more light per watt created is less power dissipated as heat, and thus, less input power is converted to heat, causing lower temperature rise per input watt. So without changing packaging, but by simply improving LED efficiency, you improve thermal resistance.

Make sense?

Click to expand...

Yes and no. In practice you're right that increased LED efficiency lowers the effective thermal resistance. However, the actual thermal resistance is of course unchanged. Also, there won't start to be an appreciable difference between effective and actual thermal resistances for a few more years which is why the power converted to light is usually neglected when doing calculations of Tj. Even at 70 lm/W the LED is maybe only slightly over 20% efficient (assuming a luminous efficacy of ~330 lm/W). This only lowers the effective thermal resistance by about 10% compared to a 30 lm/W LED.

One thing I've been mentioning for a few years now is that higher efficiency means that a given package will be able to handle being driven at a higher power level. While this effect is still practically negligible even for a 70 lm/W LED, it will become important as LEDs hit 150 lm/W and beyond. For example, the LIII package dissipates roughly 2.4W when the LED is driven at it's rated 700 mA (I'm assuming that 0.2W leaves the LED as light energy). Now if we increase LED efficiency to 50% (165 lm/W), you can drive the same package at 4.8W. The power dissipated by the case will be the same, but the same package will be emitting roughly 800 lumens instead of 65. This is over a factor of twelve increase in output for factor of six or so efficiency increase. Naturally, should we ever get close to 100% efficiency, then you can probably even drive 5mm indicator LEDs at a few watts, and get around 1000 lumens out of them. Now if only that day would come sooner...

That sounds great, luxIII performance at 1watt. Have a super long running light, or a tri-light that runs long too!

Only problem, have we figured out how to get a far throwing beam with these like the luxeons?

I'll bet an asphyrical lens would do the trick.

A couple nights ago, I took my 2" asphyrical lens, and held it out in front of a Lux III white, and Red/Orange Lux III. At 1A, the Lux III produced a reading of 25K lux at 1M, and 32K lux at 1.6A. The R/O produced 26K at 1.9A.

The cree part looks to have a nice Lambertian beam pattern, so should produce similar results.

Great news!
I like that they said the holy grail of 150 lm/watt is only two years away! Woohoo! Now I can build ultra-low power lighting for my house... and flashlights... and bicycle lighting.... and ???

evan: true, but where can we get production lenses like that that fit into flashlights at the right focal length, etc..

IsaacHayes said:

evan: true, but where can we get production lenses like that that fit into flashlights at the right focal length, etc..

Click to expand...

They are out there, google around.