Quad cree led !

[DM51]

Very interesting find - opens up all sorts of possibilities for new lights. Heat is going to be an issue...

As easilyled suggests, this thread could get buried here in LED Flashlights, so I'll move it to Beyond Flashlights > LED and leave a re-direct to it.

 

[flashburn72]

Is it just me or does the dome already look like its aspherical?

 

[Jarzaa]

With a whole lot of heatsinking, one of these in 2S2P DD off a pair of 14500s could make for an interesting Minimag.

WOW what a sleeper light that would be. You would just have to make a solid copper pill and wire the led direct drive with chips 2s2p and you could have over 500 lumen minimag. I think that would be perfectly doable with one of these leds.

 

[Opto-King]

Hmm, when I first saw this LED I thought that it was nice, but after checking the data of the SSC P7 LED I must say that I don't understand what all the fuz are all about... :shrug:

to me the new Cree LED looks like a P7 copy, please enlighten me about the wonderful thing about the new Cree LED...

 

[tebore]

Hmm, when I first saw this LED I thought that it was nice, but after checking the data of the SSC P7 LED I must say that I don't understand what all the fuz are all about... :shrug:

to me the new Cree LED looks like a P7 copy, please enlighten me about the wonderful thing about the new Cree LED...

Why don't you read the spec sheet.

2 Things pop out to me the ability to power each die separately and warm white(high out put too).

 

[Jarzaa]

Why don't you read the spec sheet.

2 Things pop out to me the ability to power each die separately and warm white(high out put too).

Plus the much smaller size.

 

[lumenlover2]

It can be wired parallel or series, and will probably fit in the existing cree xre reflectors and is about half the size of the p7 !!!
It could sizewise fit properly in an LOD again just for the size... .
It`s wonderful!

 

[saabluster]

That also means that you could optimize your reflector's focus so it is centered about only one of the dice (your "high" beam) and so that turning the other three on would produce a "low" beam pattern that emits light skewed downward and to the sides.

I also anticipate this package hosting a red-amber-gren-blue emitter setup in the near future, where independent control wuld allow variable color.

This is very insightful. Imagine how hard you could run that one die with a thermal resistance of only 3 ºC/W!. There is no doubt now that Cree is king of the led world.

 

[evan9162]

This is very insightful. Imagine how hard you could run that one die with a thermal resistance of only 3 ºC/W!. There is no doubt now that Cree is king of the led world.

A single die won't have a thermal resistance of 3 C/W. The thermal resistance is the rating when all 4 dies are powered. A single die will still have a thermal resistance of probably 8 C/W.

This and the Seoul P7 get a lower thermal resistance because each watt of input power is now spread over 4 dies, and not just one. The per-die thermal resistance is the same as it was before. There have been no improvements in thermal resistance of the individual dies...it's just a multiplicity thing that's changing the spec.

Consider this - An XR-E has a thermal resistance of 8C/W. If you run it at 350mA, Vf = 3.2V, holding the heat sink at 25C, then Tj will be (3.2*0.35*8)+25 = 34C

Now run this guy with all 4 dies at 350mA each. Tj of the dies will be (3.2*0.35*4*3)+25 = 38C.

At 1A, Vf=3.6V, the XR-E Tj will be 54C, but the quad's Tj at 1A/die would be 68C.

Even though the thermal resistance is lower, running at the same per-die current results in a higher Tj. That's probably why these are only rated for 700mA/die - with all 4 dies powered, the per-die thermal resistance is actually HIGHER than with an XR-E.

Why?

It's because the packaging thermal resistance hasn't improved at all. The total junction->heat sink thermal resistance is the sum of the junction->attach thermal resistance, plus the package thermal resistance. By putting 4 dies in the same package, you've reduced the junction->attach thermal resistance by 4 (assuming the same total input power to the device), but the package thermal resistance hasn't changed at all.

The 8C/W of the XR-E is likely 6ish C/W of junction->attach (the die), plus 2C/W for the package. Stick 4 of those same dies into one package, now junction->attach is 1/4 (so 1.5C/W). Improve the package a bit to reduce it, and now you're down to 3C/W, and you haven't done a thing to the thermal resistance of the individual dies. Your only R&D has gone into packaging - after all, why mess around with your dies when you don't have to?

It was the same with the Seoul P7. The thermal resistance is significantly lower than with the P4, but Seoul couldn't do anything to the dies, since they buy them from Cree - it's simply a result of sticking 4 of them in one package, and looking at the total power input to the package being spread out into 4 dies, instead of one.

 

[mr.snakeman]

So, who´se going to be first on the market with this new Cree? I´ve read mention of this type of led on the Surefire website- will they be first? At LIGHTFAIR Int.? If anyone is planning to attend, please let the rest of us know what you find.:twothumbs

 

[IMSabbel]

Reading this, i think thats the reason why we didnt hear anything from Fenix about P7 lights...

Also, i just dont like the parallel nature of the P7. Apart from the "parallel=suboptimal" part, multi-amps currents at low voltages just are not nice. Every contact point, switch, cable sucks away efficiency like cracy.

 

[TigerhawkT3]

With a whole lot of heatsinking, one of these in 2S2P DD off a pair of 14500s could make for an interesting Minimag.

OMG these must be the days when the god of all flashlights finally comes onto the earth !

WOW what a sleeper light that would be. You would just have to make a solid copper pill and wire the led direct drive with chips 2s2p and you could have over 500 lumen minimag. I think that would be perfectly doable with one of these leds.

I guess I have a new project ahead of me.

 

[txgp17]

I can't wait to see what Milkyspit does with one (or three) of those. And image what Gene Malkoff's M60's will be like when he gets hold of a few.

 

[Kiessling]

While retaining the physical size of the XR-E, the die area is still four times larger, meaning our XR-E reflectors most likey won't work.

 

[FredM]

Could this be used by Surefire as a possible replacement int he L4 for the LuxV?

 

[LukeA]

While retaining the physical size of the XR-E, the die area is still four times larger, meaning our XR-E reflectors most likey won't work.

They'll just turn into flood monsters it they fit. The fit concerns me because I can't tell if the die height in this new emitter is the same as the XR-E. If it's not, then reflector backs will need to be shaved down to fit. (the dies look higher to me)

 

[tebore]

Could this be used by Surefire as a possible replacement int he L4 for the LuxV?

Will it be possible? Yes.

Will it be easy? doubt it.

 

[saabluster]

A single die won't have a thermal resistance of 3 C/W. The thermal resistance is the rating when all 4 dies are powered.

Its funny, as I was about to push the "submit reply" button on that post it occurred to me that it was probably incorrect. Then my next thought was "It's ok. Evan will be along shortly to clean up." Thanks. This is one area I've been needing to really buckle down and try to wrap my brain around the formulas involved. I HATE math. So I've been avoiding it.

Maybe you can help clear up something else as well. I was trying to look up the specs for the P7's thermal performance as compared to the new Cree MC-E and found two different results. This one says 7.5 C/W and this(must click on the "W724C0" spec pdf at bottom) one 3 C/W. Why such a different spec for what appears to be the same part?

 

[4sevens]

Reading this, i think thats the reason why we didnt hear anything from Fenix about P7 lights...

Also, i just dont like the parallel nature of the P7. Apart from the "parallel=suboptimal" part, multi-amps currents at low voltages just are not nice. Every contact point, switch, cable sucks away efficiency like cracy.

There are two major advantages of serial over parallel.
1) the same current flows through all the dies so it doesn't risk thermal run away from one of the dies having a slightly lower Vf and pulling more current
2) when you run stuff at high current think about resistive power losses with those tiny bond wires. Resistive power loss is I^2R.

 

[TOOCOOL]

Not sure why you guys are getting so excited over a fancy light bulb

















J/K :nana: