CREE NX : XD16 __ OFFICIAL RELEASE __

Okay i am snooping around and if i am right? It is a XD16A and its a dome less LED?

Cree Extreme Density LEDs
• Lumen Density: First to deliver >200 lm/ mm2
• Optical Crosstalk: Very low
• Thermal Contact: Highest in class
• Soldering: Proven footprint design
• Handling: Very robust package

  • Size
    1,6 x 1,6 mm
  • Product Options
    CRI 70/80/90
  • Maximum Drive Current
    1,5 A
  • Light Output
    540 lm @ 4,5 W
  • Voltage Class
    3 V
  • Viewing Angle
    135°
  • Max Power
    4,5 W

Scroll through here and have a read!

!

!

540lm @ 4.5W =120lm/W at max rating. Still pretty high. I’d say there’s still plenty of room for overdriving with that efficacy. This was a concern since these newer LEDs are getting more aggressive with their drive currents because of increasing rated operating temperature. Assuming there isn’t (which is a silly assumption) any thermal bottlenecking and sudden drop-off at higher currents, we could see 1000 lumens from this LED if it follows the typical curve. I’m trying to keep my expectations low but I’m still very excited for this LED

Clemence has been fiddling with the Nichia equivalent of these leds (the E17 and E21 series) and ran into some new problems with these (CSP) leds (burned phosfor above the gap between the solder pads, photon cross-talk between leds in a matrix) , but he also made progress in driving them at higher currents with a newly developed (expensive!) ledboard. (the Cree NX leds seem to differ from a true CSP led by that there is still a substrate under the die). So indeed keep the expectations limited until the performance in flashlights is proven.

I still don’t know the exact footprint but it won’t differ much compared to E17A or E21A (both have the same solder pads). At the very least it will fit in the VR16S1 as a single or 4p setup. It would be great if it fits in the VR16SP4.
This is not a true CSP, looks like there’s still AlN or SiC beneath the phosphor. If this is true then it would not suffer from premature phosphors overheating like those E Nichias.

- Clemence

IT’S HERE!! :nerd_face:

http://www.cree.com/led-components/products/xlamp-leds-discrete/xlamp-xd16

Thermal Resistance 6 °C/W :person_facepalming:

Interestingly, the XD16’s don’t appear to be cheaper than similar XP-G3’s on Arrow’s site, at least not yet.

What happens to the super low thermal resistance that was floating around earlier?

With that kind of thermal resistance my hope for this LED just about died.

Sad.

Hopefully the thermal resistance is not a game-ender and we’ll still get some good output out of it. The XPG2 thermal resistance is 4C/W.

Yes. This is high compared to the latest 3535 or larger LEDs, but it’s on par with the XQ-E which shares the footprint. Thermal resistance though only matters proportionally to the chips thermal derate curve (think xhp.2 increased performance with equiv °C/W) . According to the datasheets, the improvement in high temperature efficacy (vs XPg2) of the XD16 should offset the higher temps it will see, making it very similar to the XPG2. ALSO, the Vf curve/value will affect the extrapolation of the Flux v Amps as what we should really be comparing in these cases are Flux vs Watts (but circular dependencies arise here so yikes…). Of course this is only eyeballed and napkin type calculations, I’d expect peak output ~6A and 1400 lumens. Sounds too good to be true. But so does the rest of it, at least in my eyes. YMMV.

I still keep imagining 9000 lumens @ ~90lm/W from a 3x3 array (~XML footprint) with amazing throw (XPL HI levels?) and minimal tint shift. Artifacts similar to MTG2, which I find hardly offensive.

@TexasAce

The only thermal values that floated around before were 5°C/W that I remember, so this isn’t a huge departure from that. Let’s also keep in mind the Luxeon V and Nichia E series that claim super low <1 °C/W yet are not crazy impressive (or pathetic for the Nichia) at high currents.

Time will tell.

All good points.

I also would love to see a 3x3 array of these dies done in the same fashion as the xhp35. Or even a 5x5 as a replacement for the xhp70.

Not mentioned yet, the peak efficacy exceeds 170 lm/W. That enables a light that is both high-efficacy (multi-die) flooder and high throw-efficient (1 die) thrower. You can’t do that with Blackie as it’s both larger and way less efficient.

A 3x3 array actually seems to be a practical limit, as it is already RATED at 54W. The XHP70.2 is “rated” at 29W. As you’ve just been realizing that’s a very extreme heat flux to deal with (overdriven) in even the largest air/convection cooled application. We could go 5x5 and run each LED more efficiently, but then we’d lose the throw advantages to gain lumens/W and a larger hotspot (would be nice for a floody, HiCRI, long runtime light). 3x3 seems the best replacement for most aggressively driven existing setups, especially in throwers.

A 3x3 maxed out in Giggles would be amazing. ~XHP35HI throw with greater than 70.2 lumens? Anyone? :beer:

You have convinced me, JaredM. I’m planning on getting some when they are available and putting together a makeshift MCPCB from copper square rod to make a 3x3 array.

Did anyone tried that? :slight_smile:

Are they for sale anywhere?

Where would you get an mcpcb? Would the djozz mcpcb’s work?

Where is Djozz? :slight_smile:

He’s busy.

Here. I have something better than my mcpcb, I have a few of clemence’s two-pad mcpcb, it should perform like a DTP-board, or even better.

But first we need a seller :slight_smile:

I think this emitter will be potent performer, and his higher forward voltage will make it one of the best led emitters for single cell FET driver configurations.