Cree CXB1304 COB

http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXB1304

This one is fresh from the press and has a 6mm round die. No, it is not going to throw, no, it will not be the most efficient led, the surface will probably not be uniform in brightness, the maximum output (1150lm@1A) is pathetic for such a big emitting area, etc. etc. But it does come in a 9V version, making it theoretically possible to fit into a flashlight with 3xli-ion in series. So, to stop for once the discussion of putting COB arrays into a flashlight: who is going to build one :-)

Do you have a link?

..was working on it :-) ..

Subscribed

I think a cob might make a really nice projector. Or I’m wrong again only one way to find out. I have a JaxZ1 that could be made to work, maybe.

FWIW MT-G2 comes in 9V version... just sayin'

Kinda funny a round die with a square MCPCB.

But the image of the die from an MTG2 is unattractive, as is an XML, or XPG. A round die at 3000 Kelvin is much nicer. It’s an entirely different approach, instead of a blast of uncontrolled lighting up everything, it is a precise tuned beam of light. If it all works…

Oh All right, I’ll do it!

3000K at 90 CRI should work nicely. 18º optics, 50mm diameter, should work in a MagLight. Hmmmm……

Mouser has them in 3000 Kelvin, 90 CRI I wonder if I could drive it with an MTG2 driver with a FET…

While you’re at it Scotty you should also look at the Ledil 14236 Winnie-S. :wink:

I have each on the way, should see em by Wednesday…

Anyone have an image of what the array pattern looks like when on? Certainly starting off with a round arrangement, even if it’s like an Mt-g2 in terms of pattern should be interesting in a projector style light. Throw it slightly out of focus and you could have a nice beam.

The Winnie-S is used with the CXA-1304 to light paintings, pretty tight hot spot with very little spill. The CXB-1304 is the same size and shape but more efficient.

I have successfully swapped inerds with 2 D MagLights such that I now have a free 3D light to play with. It was using 3 D size NiMH cells to run a de-domed XM-L2, but now that set-up is in a 2D Mag with a dummy cell on a 32650. :slight_smile:

Wait…that 6mm round die is soldered down to that white board (aka the square star) is it not?

I wonder what that emitter w/o the board looks like and/or how it solders down

If it is a star, could it not be “rounded” a bit

either way…3000K bleh…too “warm” for my tastes…but the CRI is nice

The usefulness will depend on how uniform the brightness is over the die area. If not, a MT-G2 will be better in all aspects. Philips claims a very uniform brightness for its 'crisp white' COBs (but that runs at 34V ).

The Cree datasheet shows a full line, including an 80 CRI 5000K version. (maybe 90 as well, not sure) It’s just new and not readily available at this point. Will keep an eye on that for sure.

In the meantime, I don’t mind a warm tint in the old MagLight, keeping with it’s tradition. :slight_smile:

Damn it, I just got a tray of CXA 1304s!

I tried it on a Jax Z1, very ugly projection. It does make decent output and Vf is close to 11V at 1.5A if I can remember correctly. Heat will limit the output on this one, bad.

Edit: Nevermind, CXA not CXB.

They don't say anything about improved uniformity, so why would the projection of the CXB be better than the CXA? Do you have a picture of the die projection of the CXA?

Heh, its already on my workbench djozz :slight_smile:

Its definately not a flashlight cob chip, not at all
In real life( at least batch 02) performs like 20% better than CXA, and its designed mostly for general lighting and spotlights lighting aplications, probaly retrofits too
Has a GREAT CRI, fairly okay thermal resistance and terrible max load vs temperature perfornace( exacly like CXA) - yet another reason why it cant be used properly in a flashlight, even a flooder

With a proper reflector this one and especialy his bigger brothers 1507-1512 can be used in automobile lights

  • can be achieved a very nice ribbon-like beam pattern
    P.S

They still “hide” real R9 though….even at CRI 95+ binned matrix chips( real one 97) the R9 is like 80 or something( usualy less)