Different type of dye in one LED..why not??

[acemelare]

Hello

This may be one very very stupid question..but it's just me thinking outside of the box

But I was thinking:thinking:...there are so many different types of LEDs, each with its own advantages and disadvantages. Why not combine them in one LED.
And why not play with the shape of the dye as well.

For Example: A ring (donut) shaped MC-E dye with a round R2 dye in the middle, installed in a smooth reflector. in a given size
In my mind that would be the perfect combination of throw and spill. But I'm not at all into modding or electronics or LEDs for that matter, but the idea popped into my head today as I was doing the dishes

I don't know if it's is possible to produce or if the result would be as I was thinking=AWESOME. But I think it's kinda weird that none of the big LED producer has tried it (or maybe they have, with poor results)


And if the the Boss of Cree LEDs is reading this, I want 10 % and the possibility to name this new wonder LED.....:laughing::laughing:

 

[gillestugan]

Om not a boss of cree, but I can speculate reason not to make round dies is because of cost and processes in production.
They probably make the dies square as it is easy to cut straight, and you get very little spill.

The cree MC-E, XR-E , XP-E, SSC P4 and The SSC P7 (and probably more) uses the same die, the cree EZ1000.So what you are suggesting is a 5 die package with a round die.

Most of the characteristics of the led comes from the shape of the doom and the size of the emitter area.

 

[rushnrockt]

For Example: A ring (donut) shaped MC-E dye with a round R2 dye in the middle, installed in a smooth reflector.

I'd love to see a die being made in a doughnut shape and then efficiently and at low cost cut out from a wafer. You can make a die into many shapes, but getting it out while still keeping its characteristics somewhat reasonable would make this a non-starter

 

[ifor powell]

plenty of leds with 3 different die in them but it's normaly red green and blue... and they are eash squares

Ifor

 

[Gunner12]

Square or simple shapes are much easier to cut then round dies. Also less loss of material if the dies are cut in a square or simple shape. I also recall someone saying that cutting the dies are more like cutting glass(edit: cutting crystals, even worse), so a circle would be even harder to cut.

The MC-E is essentially 4 XR-Es in one package. They both use the EZ1000 die, the MC-E has 4 and the XR-E has 1. Same with the Seoul P4 and P7. P4 has one EZ1000, P7 has 4 EZ1000.

If they could make that with a reasonable profit(making enough profit means the product will get to the market) then it would be nice to see a round die.

I can see why you had the idea when doing the dishes.

 

[2xTrinity]

Square or simple shapes are much easier to cut then round dies. Also less loss of material if the dies are cut in a square or simple shape. I also recall someone saying that cutting the dies are more like cutting glass, so a circle would be even harder to cut.

Cutting dice is not like cutting glass, which is an amorphoous solid. Semiconductor materials are crystal lattices. The InGaN material used to make LED Dice are specifically cubic crystals. Trying to cut one in a circle would be similar to trying to cut a salt crystal into a circle. If you attempt it, it will most likely shatter into smaller cubic grains. This means even non-rectangular polygonal cuts, eg trying to cut the dice in hexagonal shapes etc. would be very difficult, as well.

IMO I'd like to see more attention paid to designing optics that can actually shape the beam into a beam profile similar to what the original poster is seeking. This would be a more effective approach than trying to make circular, or annular dice. For example, TIR optics are usually a combination of a smooth reflector (the exterior optic/air interface), and an aspheric lens just above the LED itself. The smooth reflector portion might be designed so that it is actually de-focused stlightly, when the aspheric lens is in focus. That would lead to a spot-and-corona pattern similar to what acemelare is seeking.

The only difference is that the spot and spill intensity coudln't be controlled independenly from one another, as a concentric-circle LED could do. The best way to pull this off would be to simply have two LEDs -- one with a narrow-angle optic or reflector, the other in a wide-angle optic or reflector, and control them separately.

 

[rushnrockt]

Cutting dice is not like cutting glass, which is an amorphoous solid. Semiconductor materials are crystal lattices. The InGaN material used to make LED Dice are specifically cubic crystals. Trying to cut one in a circle would be similar to trying to cut a salt crystal into a circle. If you attempt it, it will most likely shatter into smaller cubic grains. This means even non-rectangular polygonal cuts, eg trying to cut the dice in hexagonal shapes etc. would be very difficult, as well.

There is always laser cutting, but that puts it even further into the too expensive to care category. I also think the optics solution makes more sense.

 

[gillestugan]

The best way to pull this off would be to simply have two LEDs -- one with a narrow-angle optic or reflector, the other in a wide-angle optic or reflector, and control them separately.

This has also been discussed in the headlamp forum. Take a look at this headlamp that has four(!) different optics/reflectors. One for each purpose/task.

 

[Th232]

There is always laser cutting, but that puts it even further into the too expensive to care category. I also think the optics solution makes more sense.

Just to tack on a bit about laser cutting, I'm wondering how the heat (admittedly very localised, but still there) would affect the material, it being a semiconductor and all. I use a 7 watt YAG laser at work, and I get the feeling that if I were to use it to cut something like the material used in LEDs (read: vaporise any material in the beam's path), while it might work, the properties would most likely be changed, making it useless for emitting light. The crater thrown up by the blast would also stuff up any kind of uniformity, given the scale of the piece.

 

[rushnrockt]

Just to tack on a bit about laser cutting, I'm wondering how the heat (admittedly very localised, but still there) would affect the material, it being a semiconductor and all. I use a 7 watt YAG laser at work, and I get the feeling that if I were to use it to cut something like the material used in LEDs (read: vaporise any material in the beam's path), while it might work, the properties would most likely be changed, making it useless for emitting light. The crater thrown up by the blast would also stuff up any kind of uniformity, given the scale of the piece.

That would be one of the parts that makes it expensive, even if doable. You'd have to do pulsed cutting and won't be able to do it all at once in one area, just go back and forth on the wafer allowing time for each area to cool down. Or have someone with small hands cut it with x-acto knife

 

[gillestugan]

Or have someone with small hands cut it with x-acto knife

I wouldn't volunteer. The hands are too shaky at my age. (31)

 

[StarHalo]

My new Peak McKinley has seven emitters, three dies each (red, green, blue). It's quite the fireworks show when turned on..