Cree XHP70 up to 4022 Lumens and XHP50 up to 2546 lumens - Multi-die leds.

You do realize the metric system is the tool of the devil, right? 0:) :smiley:

Ok then, it’s about a hair, I ain’t gonna say what kind o hair…

Horse hair?

It was a simpsons reference, Grandpa said: “The metric system is the tool of the devil, my car gets 40 rods to the hoghead and thats the way i likes it”
I’m shocked i can’t find a youtube video, but here is some more info:

https://www.physics.rutgers.edu/analyze/wiki/math1.html

Look what I found…

The solder wire above it is .031” Kester EP256, beside it is a capacitor from C1 of an FET board.

Someone put their XHP70 on the star backwards! :stuck_out_tongue:

Impressive, I think you missed your calling, Cree should be hiring you as an engineer

:stuck_out_tongue:

In a Tesla that should not be a problem :bigsmile:

nice

Nice!
Man, if only they could squeeze those dies just a little closer together! :stuck_out_tongue:

We just need a tiny cuttoff wheel on a dremmel, split the four dies carefully along the seams and sand them down a bit, then superglue them back together without the gap, who’s going to try it first? Dale? :party:

I was thinking JB Weld. I spoke to the guy in Honey-I-shrunk-the-kids and he’s pretty sure if we shrunk it on a horizontal plane it’d come out XM-L sized with XP-G sized dies crammed together snug. Think it’s worth a shot?

Apparently the pack-rat in me wouldn’t let me throw it away. It was in the depression under the handle of my 10 drawer kit box. I thought I’d tossed it so I wouldn’t be tempted to Play with a 12A blacklight…

Edit: Pretty compact little XP-L Quad isn’t it? :bigsmile:

Ok, so let me ask this… since you can plainly see that there are 4 plates to the substrate and since there are 4 connections on the bottom of the emitter to verify they are wired separately, is it not feasible that we could de-dome this array and then airwire it 4P to run on a single 18650 at 3V? A dedicated quad XM class as it were? If not airwire, then modify the star? If I can get my hands on some more of these I think I’m gonna try it…

In case y’all are wondering why it’s so fugly, I de-domed it hot, running at 12.15A. Then, when some phosphor pulled off one of the dies, I scraped the rest off while it was on and played with the UV mega light for a bit. :wink: Most of that was done looking at it through a welders lens while on at full power.

Haha now that would be quite a mod!
Very cool idea, not sure how feasible it is to fiddle with those tiny bond wires though. I remember there was a thread a while ago regarding repairing dead leds by reconnecting the burnt out bond wires. Not sure how successful any of that was though.

I think I can tell why we can’t do that.

The 4 contact points available after dedoming are the same as the 4 connections available on the bottom of the emitter, nothing extra is available by dedoming. That leaves 4 contact points unavailable to us, they are already connected in pairs. Where are they you ask? In DBCstm’s picture they are connecting the upper left and lower left emitter in series and the upper right and lower right emitter in series. I think it’s very reasonable to assume that this connection is happening directly under the dies.

I hadn’t paid much attention to the subject previously, but in the recommended 12v configuration in Cree’s datasheet the thermal pad is not isolated.

Well, it might prove of interest that I realized the comparison shots against my other 2 Eagle Eye X6’s was a bit skewed. The Triple and Quad are running on FET drivers, while the XHP70 had a PZL driver with 7135 regulation chips doling out 5.45A regardless of cells used. A pair of Sony C5’s had the exact same pull as a pair of Red Efest V2 button top 18350’s.

So, I unchained the beast! :bigsmile:

The A17DD-S08 Zener modified driver is now running the show. The pair of Purple Efest IMR18350 allow 7.45A on the top end, with 7 mode levels for easy battery life control.

0.01A for 11.385 lumens
0.10A for 66.93 lumens
0.21A for 158.7 lumens
0.54A for 489.555 lumens
1.86A for 1328.25 lumens
3.81A for 2421.90 lumens
7.45A for 4426.35 lumens
at 30 seconds, after running the other tests, it’s still holding 3650.1 lumens. Considering the small batteries this ain’t too shabby…

I think this solidly surpasses my Triple XP-L now. :slight_smile:

Reckon I’ll have to do the beamshot comparison shot’s all over again…

Edit: For chits and grins I stuck the pair of Sony C5’s (waiting for a charge) in the X6 and tried it on the lightbox, using the 14” long 12Ga Turnigy lead to make ground contact. The light box showed me a brief 5530 lumens. :slight_smile: Crazy stuff man!

Thanks Djozz! Nice to see some easily comparable graphs. Always nice to see some graphs and hear your thoughts.

In the future, could you make the 4,5-9 volt range easier to compare when testing emitters in this voltage range?

Do you plan on testing 6v vs 12v configuration some time just to see if there is any difference?

Thanks also Djozz. It was remiss of me not to comment. Appreciate your hard work as usual.

I'm not sure what you mean, perhaps that the scaling of the voltage axis can be changed so that the voltages show more accurate? I do not usually do that because then the voltage curves get in the way of the light-output curves, that is why I choose an extended voltage scale, and I choose a max so that the scale fits into the grid that is dictated by the light-output axis. But I realise that the voltage scale does not necessarily have to start at zero, that makes a variant possible that shows the voltage better and still does not obscure the light-output curves completely, is this a better graph?: