What the Lux3 is REALLY capable of

Given that a T-bin Lux3 is around 80 lumens when driven at 700mA, that means you're going to get something like an amazing 114 lumens at 1 amp, and pushing that up to 1.2 amps means 137 lumens. I am aware of the thermal dropoff issue (output drops as heat increases) but for calculations let's take it to be linear up to 1.2A.

No wonder my TWOL regularly seems somewhat brighter than my Surefire P60 lamps on fresh batteries even with lens/reflector losses. It's a bit hard to discern with the naked eye but even then I've always gotten a feeling it's just slightly brighter.

I noticed a Lux3 driven at 1.2 amps is very bright, and in fact close to a LuxV. Given that a LuxV sample I'm using is around 150 lumens at spec, I'm overdriving at 0.917A, that would be something like 196 lumens. The LuxV is brighter, but I'm starting to appreciate the tremendous bang for the buck the Lux3 offers especially when you push it.

i like em to they throw so far and i love throw but overall i think a lux 5 puts out more light but a wider panel that i dont realy like as much

As I understand it [now THAT'S funny!], efficiency drops drastically past a certain point and tint can also be affected adversely. What that actually means to the user, I'm not at all clear on. So many variables, so little mental capacity...

The output isn't proportional to current and is even less so once you factor in the temperature rise of the heat sink. In fact, it's a case of greatly diminishing returns. Even with an infinite heat since, an LIII will only give about 40% more light than it will at 700 mA. I think this will occur at roughly 1.4A. For most real world situations you're looking at perhaps 20% more light at around maybe 1A or so. Past that there is very little to be found. IMO, it's probably not worth it to even overdrive the LIII past 1A unless they improve the thermal resistance further.

In another thread it is claimed that Lumileds increased the nominal current to 1.4A and the output to 135 lumens. This pretty much obsoletes the Luxeon V, which is 120 lumens nominal.

Am I right in thinking that LuxIII's have other advantages over the other Luxes, like generally whiter tint for instance, regardless of bins, and no donuthole beam which is apparently common to the 5w?

Yes, there is no donut hole, and an improved phosphor deposition process results in more consistent tints, although I believe the process is now also used on at least 1W Luxeons as well.

The 1.4A LuxIIIs are red/amber/red-orange, not white, blue, or green. Lumileds has just gotten the lower end of the spectrum into the 3-watt package.

i think i complety missunderstand post ya are saying there are a new type of a lux3?

Only new colours in the Lux3 line. Lumileds have not announced any new specs for the white part yet, altho its been hinted at by a forum member.

Also, i think the 80 lumen figure is at 1 amp, not 700ma so dont throw away those LuxVs yet. A good 5 watter is still a lot brighter than any Lux3.

The 80 lumens figure is at 700mA and like I said, it scales well up to 1A 'approved' drive current LINEARLY (according to their graphs) giving up to 114 lumens of light. 5Ws are still brighter, but what I'm saying is that for bang for the buck (or the watt) the Lux3 still works amazingly well. My 5W cost me 4 times what my 3W did and puts out only 40% more light.

I've found with decent heatsinking you can still get a bit of improvement up to 1.2A. I believe some other CPFer has graphs indicating that the ideal point of light vs outpot loss caused by thermal issues is close to that.

Actually, it's a little less than linearly, and Lumileds has a figure of 70 lumens typical at 700 mA now (the previous values were 65 lumens at 700 mA and 80 lumens at 1A). Note that 1A is 1.43 times 700 mA but 80 lumens is only 1.23 times 65 lumens. This is hardly linear. Even assuming 80 lumens at 700 mA, we'll end up with about 98 lumens at 1A, and 112 lumens at 1.4A (both figures assume infinite heatsinking). You can't change the laws of physics-LEDs simply get less efficient as current increases, and more efficient as current decreases (at least down to maybe 20% of nominal).

You won't touch this 80 lumen figure folks are tossing around, you need to derate it significantly, due to the die temperature not being at 25C. Then toss in reflector loss, and window loss, and such. Then add in loss for the slug not being held cold too.

You are going to loose 20% of your light from Tj heating, even if you have a impossible infinite heatsink with perfect heat conduction. Thats at 1A.

You'll loose even more at currents higher than 1A.

Now if you only have a 6D Mag flashlight body with a hotlips or something, you losses due to heating will be even worse than this. Even a 6D Mag is *far* from an infinite heatsink.

LumiLED's specs are with the die at a temperature of 25C, which it jumps right up, 13 degrees C for every 1W of power you put in.

[ QUOTE ]
jtr1962 said:
Actually, it's a little less than linearly, and Lumileds has a figure of 70 lumens typical at 700 mA now (the previous values were 65 lumens at 700 mA and 80 lumens at 1A).

[/ QUOTE ]
Which datasheet are you looking at? I thought that I had the latest ones on file [August 2004 in the case of the DS45, Lux III emitter] and they still show the 65 at 700mA typical.

I agree with Newbie's post. it's almost impossible to maintain spec temp of 25c in normal running conditions.

when you boost milliamps, luxeon generates more heat leading to loss in output. then factor in light tranmission loss. A more realistic loss may be in the 30-40% range.

Don't have a integrated sphere, but do have ARC4X used as a calibrated light to compare others to.


I find that 5watt has a different focal point then a 3watt. If you use a refector optomized for a 3watt. then the 5watt's beam will be defused. leading to the wall of light reputation for 5watts.

My brightest 3watt is nowhere close in output to my brightest 5watt. with both driven at up to 1.5amps.

Sorry to get off topic here, but does anyone have a link to those red lux3's?

[ QUOTE ]
Doug S said:
Which datasheet are you looking at? I thought that I had the latest ones on file [August 2004 in the case of the DS45, Lux III emitter] and they still show the 65 at 700mA typical.

[/ QUOTE ]
Go to Lumileds' web site-it's right on this page. Note the 70 lumen typical figure for the white side-emitting LIII. The Lambertian is still at 65 lumens typical. Apparently most of the LIIIs, at least the side-emitting ones, are making the T bin, albeit barely.

[ QUOTE ]
jtr1962 said:
[ QUOTE ]
Doug S said:
Which datasheet are you looking at? I thought that I had the latest ones on file [August 2004 in the case of the DS45, Lux III emitter] and they still show the 65 at 700mA typical.

[/ QUOTE ]
Go to Lumileds' web site-it's right on this page. Note the 70 lumen typical figure for the white side-emitting LIII. The Lambertian is still at 65 lumens typical. Apparently most of the LIIIs, at least the side-emitting ones, are making the T bin, albeit barely.

[/ QUOTE ]

jtr1962: That's a typo. It appears that they used the 1000mA output for the White SE and also for the Green SE above it. The 700mA value for the white and green SE's is 58 lm per the lastest datasheets.

Thanks, Doug. That makes more sense anyway. Usually the SE puts out fewer lumens than the Lambertian. It didn't make sense to me why the SE was speced at more on Lumileds' site. I'm surprised they let a typo like that get through.

I really really like my UNDER DRIVEN Lux III's in the Pentalux Elektrolumen. They just WORK. Lotsa light, no heat, Mucho runtime on batteries. What can you ask? /ubbthreads/images/graemlins/confused.gif Not for more. /ubbthreads/images/graemlins/tongue.gif I know under driven LED aint gonna fry for at least that 10000 hour or more lifetime. Over drive gains lumens but at lost life time and lower efficiency/more heat. /ubbthreads/images/graemlins/frown.gif
Lux III's sure are some SWEET LED's! /ubbthreads/images/graemlins/grin.gif
Hard to beat, really neat...

NikolaTesla /ubbthreads/images/graemlins/smile.gif

An Arc lamp is the Spark that takes away the Dark--HID Forever!

My Lights

[ QUOTE ]
cy said:
I agree with Newbie's post. it's almost impossible to maintain spec temp of 25c in normal running conditions.


[/ QUOTE ]

Just for testing you could cool them down(as people do with overdriven CPUs. Of course, not practical. The closest thing is using a light underwater.

[ QUOTE ]
My brightest 3watt is nowhere close in output to my brightest 5watt. with both driven at up to 1.5amps.

[/ QUOTE ]

5W has twice the voltage. If you drive them at the same current, the 5W gets twice the power.