Strange question about Luxeon Star Led's...

:help:

(Oops, messed up when posting this thread - sent to wrong category. Hope this works)

Hello. I have always come to the Candlepower Forums for tips and info, but I never registered. Now here I am. Anyways. I was wondering about the shape of the die of a Luxeon Star. I held up an assortment of the Luxeon Stars that have acquired over time, and I compared the sizes of the LED die. The 1W white die is about 1mm wide (approx. 1mm^2), the 3W white die is about 1.5mm wide (approx. 2.25mm^2), and the 5W white die is about 3mm wide (approx. 9mm^2). So, why isn't the brightness proportional to the LED die size? This really bugs me!

Ok, the 3W white is an oversized 1W. It has an area that is 2.25X bigger, so why isn't the brightness at least 2.25X brighter? Throughout this post, I will mostly compare the brightness (in lumens) from what was given on the Lumileds datasheets. The old 1W used to put out 25lm at 350mA, and the cheap 3W (S-binned) put out about 60lm at 700mA. This is pretty close to 2.25X brighter. Now, you can get a decent 45lm 1W Luxeon Star, so you should be able to get a 3W that puts out 100lm just as easily, right? I wish. This is assuming that the 3W uses the same new technology (and efficiency - lm/W) as the latest 1W. Now they do have the U-binned 3W, but they are a bit difficult to get a hold of one. Why is this? Using my theory, the common T-binned 3W (typ. 80lm) still uses the same technology as the 35lm Q-binned 1W. This really isn't too bad; it really doesn't concern me.

What really grinds my gears (ha-ha) is the 5W Luxeon star. This older monster is said to have 4 1W dies together in a single LED. Now this is okay to believe, but when you really think about it, it really makes no sense. Why is the die(s) on the 5W so big? They compare to the size of a 3W die. I understand that a different technology is used in the making of 5W Luxeon Stars, but why is it so large like this? The quad 1W idea can work because 4 older 1W (Q-bin equivalent) at 30lm together makes 120lm. Using 4 1W die with the brightness each from 38lm to 40lm can yield a W-bin 5W (about 150 to 180+ lumens). I have one of these W-bin 5Ws and it is ok (not super bright). Why do they use 4 led die that are the size of the 3W? This gives you a beam, when using a good reflector, that is too big. The 5W are hard to make a good light that has a lot of throw to it. Imagine if they made a 5W with 4 actual 1W die. The surface of the 5W die(s) would measure about only 2mm by 2mm! This would give a nicer beam with better throw. The 3W red/orange has a die that is 2mm by 2mm. The throw is improved because of this, but it is red light that isn't as useful (except for spotting raccoons with).

Wow this is a long post. I have plenty of questions. Now, using my theory on the size of the die, the 5W, using "3W equivalent sized die", should put out at least 65 X 4 lumens (260lm)! This is an average 3W driven at only 700mA. Now I know that it isn't the same, but I am saying all of this assuming that Lumileds is using the same or similar technology, when it comes to lm/W and die size. I know the 5W is a series/parallel set up of LED die (350mA per die). Does this make sense? What if Lumileds did make a 4 X 3W Luxeon Star LED? The package would look the same, it would just run a about 7.8V @ 2A and put out about 320lm.

Also, using my theory with the die size, the 5W has a die that measures 3mm by 3mm (9mm^2). Why isn't the brightness 9X greater than a typical 1W? Why don't they attempt having a LED package with 9 X 1W LED die? This would yield approx 270lm to 405lm at (using the series/parallel setup) 12V at 1050mA.

So why doesn't Lumileds have their 5W with a smaller 2mm by 2mm die? Why don't they just use 4 3W die in the package? How about 9 1W die in a package to replace the 5W? These are crazy and dreamlike concepts, but I was just wondering why the die size was not proportional to the brightness in the whole line of Luxeon Star LEDs. I know that too many die for a higher power LED would mean higher cost, more heat generated, and a greater complexity of manufacturing, but I was just wondering and wishing. No one seemed to have talked about this before, and it has bugged me for sometime. What is with this? Please help me. Maybe others can learn from this, or ideas for a more powerful LED can be created.

Finally, why can't Lumileds use a large, single die that measures from 3mm to 4mm (or more) in size? This would perform well in the super high-end lighting or the automotive industry. I don't know about the chemistry an LED die, or of the process of making the LED die, so this may be impossible. If they cut the die into larger pieces, and mount them on a larger copper slug, would that be awesome? Can you imagine how much light can be produced? I know there would have to be multiple lead wires to connect the massive die to the anode. Also, a large amount of while-light-producing phosphor would have to be slathered on top. Imagine a die that is 6mm by 6mm that emits nearly 1600lm to replace an incandescent headlight bulb. I was just thinking.

This post, again, is way too long, but respond with whatever answers you might have or any comments. I would greatly appreciate any sort of feedback. Thanks!

You could solder a bunch of these 43 lumen luxeon flashes together. 100 would give you a 4300 lumen flashlight


http://www.lumileds.com/products/line.cfm?lineId=17

I think your measurements are off. The dies in the Luxeon I and III are the exact same size. There are no discernable physical differences between the two.

The Luxeon V dies are simply 4 Luxeon I dies in a series/parallel arrangement. The output from the Luxeon III is rated at 700mA (700mA/die), while the output from a Luxeon V is at 700mA (350mA/die), so the current density is twice in the Luxeon III as it is in the Luxeon V - thus, you don't see 4x the output from a Luxeon V as a Luxeon III - however, you do see 4x the output for a Luxeon V vs. a Luxeon I

If you were to run a Luxeon V at 1500mA, you would see more consistent scaling vs. a Luxeon III - however, the thermal path in a Luxeon V isn't sufficient to move the 11W from the junction to the heat sink effectively, thus the Luxeon V would be running at near maximum junction temperature, reducing output and lifespan of the LED.


Making more light from an LED isn't a simple matter if increasing the size of the emitting area. Ensuring even current spreading over the junction area is tricky, and the larger the die (for white/blue/green type LEDs) the more extraction losses you experience. Extraction losses are losses experienced by photons generated at the center of the junction failing to move through the junction material and escape - instead they are reabsorbed into the substrate.

The R/O luxeon is a different beast all together. That one does scale up to a larger die due to the different materials, and the physical structure of the die. The die is an inverted truncated pyrimadal shape (TIP), so it's a 3D shape, rather than the 2D planar shape of the white/green/blue variety. This shape along with the substrate materials, offer high extraction efficiencies even with large die sizes. That's why the R/O Luxeon III is a 2mmx2mm single-die device.

Sorry to hi-jack your thread, I have a question related to this. I have never seen a Luxeon V in action but would like to know if one Luxeon V is better than 4 x Luxeon I for illuminating an area say, 5m radially? I need it for a reverse(parking) light.

Hi. Thanks Carbonium and Evan for replying. Now I am a bit confused. The measurement for the 1W was from a newer batwing Luxeon Star from a Star/O. The die size was significantly smaller. I have a Lambertian 1W that I think had a larger die, but I don't have it with me. I know that comparing the Luxeon III and a Luxeon I or V is like comparing apples and bananas. I was just saying all of this in theory. I had a feeling that the larger die would present problems. So why don't Lumileds create an LED that can handle a far greater junction temperature for the overdriven 5W? Why isn't the gold wire that connects the die to the anode side of the LED made of heavier gauge wire for accidental overpowering situations? I had a 5W that ceased working, possibly due to an excessive voltage and current drawn. As a last attempt to get usable light of some form from it, I took off the lens and carefully removed the phosphor from one small point. I was able to connect my own wire to the die and continue to use it as a flood for some time.

So, if Lumileds can create a LED that can handle greater power and heat (as long as you can properly dissipate the heat), why don't they create a 12W LED that uses the die from a Luxeon III (with the higher current density)? Even though this LED may not be practical for most applications where battery life is critical (like in a small flashlight), it would be useful for automotive use. Plus, this would look rather impressive to some people who look at the amount of light per source. Will they do this at all with the new K2? I know that the K2, so far, only has a single die style LED. I have a feeling that I am a bit impatient. I remember when the first 1W was released. I was so excited to get that Star/O. I wish that Lumileds had far greater selection in the different styles of LED packaging. I am not saying that they do not have enough selection so far. It just takes so much time to create a new product. It would be great if they could widen their range of products, which would be great. Too bad that they don't have an extremely powerful version of the Luxeon Star (like the concept mentioned above). The thing wouldn't be very efficient, but it would be helpful in some applications where it is difficult to use a cluster of Luxeon Stars with several reflectors/optics. Again, I am just dreaming here.

The measurement for the 1W was from a newer batwing Luxeon Star from a Star/O. The die size was significantly smaller

Click to expand...

The lens on a batwing (low-dome) luxeon alters its appearance such that it does appear smaller. However, if you had a lambertian (high-dome) Luxeon I, you would see that the die size is identical to the Luxeon III.

In addition, the phosphor on the white luxeons makes them appear much larger than they really are. Trust us on the measurement (1mmx1mm), that is direct from Lumileds themselves.

:bump: These forums go fast!