Luxeon 3W Red LXHL-PD09 Photos

Sorry about the poor quality, but thats all I had at my disposal right now.

Notice the use of the HP TIP die that HP patented years ago to increase light extraction efficiency, much larger die size, and the current distibution "traces" on the top to improve current spreading.

Cool pics... how much brighter are they?

-Mike

At least 2X if you put 1A into them, guess it all depends on the bin you get. This one is G4F.

And you gotta heatsink them really well.

G4F? Did they change the binning structure?

What current are they binned at?

For reference, here's my shots of a Lux I R/O:






To me, it looks like the die is larger on the new Lux III part.

A = Flux bin (G, H, J, K, etc.)
B = Color bin (1, 2, 3, 4, etc.)
C = VF bin (E, F, G, etc.)

G4F

G 155 lumens - 200 lumens
4 620.5 nm - 631.0 nm
F 2.55 Volts to 2.79 Volts

http://www.luxeon.com/pdfs/AB21.PDF

Flux Characteristics at 1400mA, Junction Temperature, TJ = 25ºC
Optical Characteristics at 1400mA, Junction Temperature, TJ = 25ºC
Electrical Characteristics at 1400mA, Junction Temperature, TJ = 25ºC


A really interesting chart is Figure 9. Maximum Forward Current vs. Ambient Temperature
derating based on TJMAX = 135°C for Red, Red-Orange, and Amber.
http://www.luxeon.com/pdfs/DS45.PDF
Take a look at the derating curve, to do 1400mA there is some really important stuff there, as far as cooling the sucker.

Yes, it is considerably bigger.

Interesting, my old red Luxeon I didn't have the TIP die, so they did use the TIP die for the Red-Orange on the Luxeon I... Never had a Red-Orange, learned something today. So for the Luxeon III Red, they have obviously migrated to the TIP die. I wonder if the improved Red Luxeon I actually had the new TIP, but not the old ones?

For example, in the Luxeon I there were the LXHL-BD01 and LXHL-BD03.

What would be involved in making a flashlight with this LED? Would there be significant changes from a light that was designed for the existing white Luxeons (heatsinking, circuitry, optics, etc.)?

I am thinking about having someone make me a ~100 lm red or possibly amber flashlight, but I don't want to get into it if there would be a lot of development involved.

Scott

Evan9162, I merged the two pictures, looks like the height and width is about 2X, which means 4X the area.




Beezaur,
You could use one of dat2zip's buck converters, possibly using a Li-Ion, and do it up in something like the Aleph III.

Maybe someone more knowledgable on the fine points of the Aleph III and the converters could chime in.

Would be one wicked combo if you ask me...

Newbie,

I have a HD Red Lux I, it is also has a TIP die. I would guess that the amber HDs also had a TIP die.

The new batwing Lux I Red/R-O/Amber with higher output also likely have the TIP die as well.


Nice job on the image merge. That would explain a lot, both in the thermal impedance dept and dynamic resistance. The physical bonding method (to the heat sink) doesn't appear to be any different, but there is a substantial difference in the electrical contact on the cathode (top contact). So a physically larger device would naturally lower the thermal impedance.

Lumileds seems to hesitate building larger dies with the InGaN flip-chip family, due to the lower extraction efficiency of larger dies in that configuration. The TIP structure must alleviate extraction efficiency reduction for larger devices.

Beezaur, I have made a custom 2'D' red with a LUX I red, running direct at spec power. O just ordered a 190lm R/O, I will run some tests to find out DD options. If you want to run at 1.5A I dont think there is anything out there that will regulate that power, (like madman and so on)
Jeff

Using R/O is as easy as it can get. You can easily drive these at 700 - 1000ma with a single normal CR123 and a resistor. At these currents, the 123 cell voltage will be very close to the Vf.

Just MHO, but a board is over kill for this arrangement, since the V bat and Vf are so close.

AlInGaP devices like this tend to be very sensitive to heat vs light output, so as Newbie pointed out, good heat sinking is a must.

[ QUOTE ]
WildRice said:
Beezaur, I have made a custom 2'D' red with a LUX I red, running direct at spec power. O just ordered a 190lm R/O, I will run some tests to find out DD options. If you want to run at 1.5A I dont think there is anything out there that will regulate that power, (like madman and so on)
Jeff

[/ QUOTE ]

I could build a 1.4A Linear LDO regulator with about 0.1V of dropout. The Vf for these devices is a little tricky. It's a bit too low to drive with a linear regulator from 3 NiMH cells, because the cells will get overdischarged. 1.4A is asking a lot from 3 alkaline cells - but could be done. With 3D cells, you'll drop the voltage down to about 4.1V at 1.4A. Using a linear regulator, you'll get about 4W into the Lux, and about 1.5W into the regulator, for about 61% efficiency.

Lithium 123 cells would be ideal for my purposes, but it sounds like that kind of a light might be hard to cool with this LED? Is the problem that the LED is more sensitive to heat, and its lifetime would suffer?

I am a complete electronics nincompoop -- If I understand this, the new red-orange (R/O?) Lux III can be run at 1400 mA to produce 190 lm, so long as it is agressively heat-managed. After 20k hrs of operation, it will still produce 50% of its rated output.

So, 700-1000 mA (single 123) should give something like 100 lm? What would happen if an existing Surefire or similar flashlight had its led swapped out for one of these? I don't quite get the whole Aleph build procedure. I'll have to read more about that.

Scott

Beamshots comparing both the 1 watt and the 3 watts at the same drive currents would be greatly appreciated! /ubbthreads/images/graemlins/grin.gif

We're living in interesting times. /ubbthreads/images/graemlins/smile.gif

Newbie, yes, the original part numbers for amber and red did not have a tip die. R/O always did though. Then they came out with improved red and amber that had tip and have different part numbers for them. I remember when this happened, but yet cant give a date. Also around that same time the R/O was only avalible in HD not LD, but then that changed too.

Newbie: As far as drive current, if that "F" Vf bin is close to 2.55, then using 2 NiMH cells should be able to drive it close to specs correct?? Can you try that with some C's or D's or even A's NiMH??

Everyone: The die looks 4x as big. This allows more surface area to conduct heat away, and more area to emit light. Now, what is the beam going to be like?

R/O LuxI: In a mag reflector the beam is bigger than a white lux. Why? Well I guess because even though the die is smaller, the sides of the pyramid reflects light off of the silver "cup" in the luxeon, as not just the front is emitting light.

R/O LuxIII: Will the beam be bigger in a Mag refelctor? Or will it be close to the same? Hmm. Looks like the front surface is bigger but flat like a white luxeon, and the sides of the pyramid are about the same as the 1w R/O. So it could go either way. Who will be the first to post beamshots!!? /ubbthreads/images/graemlins/popcorn.gif

[ QUOTE ]
Nerd said:
Beamshots comparing both the 1 watt and the 3 watts at the same drive currents would be greatly appreciated! /ubbthreads/images/graemlins/grin.gif

We're living in interesting times. /ubbthreads/images/graemlins/smile.gif

[/ QUOTE ]

I second that. Can anyone make some beamshots?

I spoke with Future - they are suppose to be available within 1 - 2 weeks to the rest of us - pretty reasonable also - only $6.50 / each in qty 1 each.

That translates into only about $ 4 of real value since the $ has lost 40% of its international value over the last year.

Hmm if you go to www.lumiledsfuture.com it looks like they are all available as of now (even the red and amber ones) - I know they have the red/orange available now as I ordered a few - ended up snagging some H-flux (200-260 lumens) too, so looks like the getting is good right now!

I talked with someone at Future 2 days ago and I was told they were out of Reds (LXHL-PD09) and it would be "4 to 6 weeks" until they got stock...

Sometimes I get the feeling I'm not always told the truth as the website still claims they are in stock /ubbthreads/images/graemlins/mad.gif/ubbthreads/images/graemlins/mad.gif/ubbthreads/images/graemlins/mad.gif...


I'd also love to see some beamshots. /ubbthreads/images/graemlins/popcorn.gif

Red/Amber probably is sold out due to traffic light makers, etc. R/O should be avalible though...