Luminus's new emitter with circular die

Again I don’t need to watch your link as I work in the industry, and again, I have seen with my own 2 eyes the dicing saws.

Lol, so much bad info in that post it's not even funny. Gallium Arsenide is for Red LEDs, blue LEDs are InGaN or GaN. LEDs are generally grown on Sapphire or Silicon Carbide wafers, not a metal.

I work in a class 1000 clean room, but we also have class 10,000 and 100 based on the needs of that particular manufacturing step. Obviously you have a slight knowledge of this stuff, but not very much, and definitely not as much as you think you do.

Uhm, I think Bose works for Cree. Just saying.

Yup, definitely not home made or Fred Flintstone LEDs.

What a suprise! You are obviously a smart guy - maybe. For all I know, you are the janitor there. To smart to watch the video and see how others do it, eh? Or just to smug? I'm aware that different materials are used for different types of product. I saw no reason to write a book to include all the possibilities. Resorting to personal attacks because you got your feelings hurt is not smart at all. I have no idea who you work for and I don't want to know, but I can say with confidence that if you guys are currently using saws to cut out individual LEDs - you either wont be for long or you will be out of business soon. NONE of the big 4 use anything but lasers for their new products. Check it out for yourself. Do some homework and broaden your horizions. The LED industry does not start and end at your place of employment.

Nope, a Process Sustaining Engineer.

I mean, we only LEAD the LED industry in pretty much every appreciable spec.

Big 4? Who would the big 4 be?

Cree, Nichia, Osram and Philips Lumileds, I’d guess.

I don't really care about waste or cutting them in the most efficient method, they could even make individual bases for them to sit on for all I care instead of cutting them out, that would waste nothing. I would just prefer the most efficient flashlight, not the most efficient for the manufacturer to produce.

This is a pretty poor paint picture but the idea is good, traditional square led on the left and round led on the right. I'm sure all that extra area will produce a lot more lumens.

Yea, funny thing is Osram doesn’t make a lot of their own stuff, they buy the chips then all they do is package them.

But if you make the sides of the square equal the diameter of the circle you get more surface area with the square chip and thus more lumens.

According to several sources the top 10 LED companies in 2011 were:

1. Nichia
2. Samsung LED
3. Osram Opto Semiconductors
4. LG Innotek
5. Seoul Semiconductor
6. Cree*
6. Philips Lumileds*
7. Sharp
8. TG
9. Everlight

*Companies have the same ranking when the difference in revenue is within the margin of error. Revenue includes packaged LED sales only.


Altogether holding a market share of 68% while the first 3 had a revenue >1 billion $.

But I these numbers seem weird. Because whenever you read about a big breakthrough or the most efficient LED, its a Cree.

And if you want to maximise throw instead of basic luminous efficiency then a smaller die with the same current (assuming similar efficiency per unit of area) will give much higher surface brightness..

That list is odd, because guess who Seoul Semi gets their chips from, Cree. All Seoul does is package them.

This is true, but to be honest, the flashlight market isn’t even a blip on Cree’s radar, general lighting is. The goal is to replace the incandescent bulb and fluorescent lighting.

In the case of flashlights 99% of the time the leds sit at the bottom of the circular hole in the reflector, to fill that hole would make more light rather than just make it bigger to fit a bigger square led, then you could still fill in around it (round led) and produce more again.

Either way, 2 types, 10 types, 573 types.. I don't really care, the more choice the better.

The more competition the more they will work harder to make better leds.

Yes, true again. I don’t think there are any LED manufactures that are really focusing on the flashlight though. In comparison to the lighting market it is a small segment. I mean, all these LEDs are targeted towards things other than flashlights, they just happen to work in that role as well.

Yeah, I wasn't suggesting that that would be a factor in Cree's design specifications (although with the auto-headlamp market growing you could see the demand for more easily focusable LEDs growing in the future).

I was just trying to point out that the fact that a circular die can supposedly maximise the ratio of space taken up by the die itself in relation to the footprint of the LED package, and therefore alloow for a larger surface area, isn't necessarily a good thing in all respects for flashlights. We can choose XM-Ls over XP-G and XR-E for higher surface area already and this allows slightly better efficiency and higher max drive currents. However, for surface brightness, and therefore for 'throw', a smaller die will still be better. The 'higher surface area in the same footprint' isn't really one of the main benefits of circular dies.

Having said this, I don't actually know what collimation benefits a circular die would actually give? Other than a circular beam in an aspheric lense or other form of direct projection of the die area..

I am more interested by Luminus' new version of the SST-90 without a dome but with a flat layer of optical silicone instead (at least I think it's optical silicone used, I know it's expensive). This will be used in the upcoming Nitecore TM20 I think and as the traditional dome effectively gives the die an apparent increase in surface area/size (and so concurrent apparent surface intensity decrease) this new optical design should allow for much better collimation from the same base die.

Maybe dumb question, but increasing the die size (round or square) will need bigger reflectors to throw well, right?

How's then they stated in the article that it throws well with narrow beam and long distance?

If this is real and not just advertisment, it makes me think the new die is round but somewhat small in diameter, so to make it easier to focus with a reflector.

Or not?

Edit, just read on the flashlight maker page that head diameter is 70mm. Any led throws well with a 70mm reflector...

Will need to wait for more info to seep through...

Take a close look at that picture. The die itself is still rectangular. The active area is a round area in the middle. I don't see any real advantage to that design... it wastes all that material in the corners, which could be put to use making light.

I think diameter shouldn't really be the point of consideration, as a round or circular die can be made at different sizes anyway.

I thought much the same as you though and they are good questions. Increased die size does need a bigger reflector to collimate properly. I could imagine that a round die would allow for a more efficient collimation than a square one in a round reflector, but the differences I am imagining would not be that great, and I could be plain wrong.

I would like to know exactly what the benefits of a circular die of the same emissive area as a square one with the same optics and package etc would be for a conventional reflector.

I would also hazard a guess that these benefits would be much more apparent in dies without the conventional dome over them, which seems to spread the light more evenly anyway..

TSB-70? compare with my TMA-66 TM....

Waiting for real spec.

No dome = wide angle = less throw for aspherics lights.