What better the Luxeon K2 or Luxeon III?

Jbcourt

Newly Enlightened
Joined
May 31, 2009
Messages
17
Location
Puyallup, WA
What better the Luxeon K2 or Luxeon III? For a flash light app? The Luxeon III is brighter and only has one lead on each side. Does that make it better?
 

DM51

Flashaholic
Joined
Oct 31, 2006
Messages
13,338
Location
Borg cube #51
I'm moving your thread from LED Flashlights to LED, as your question is about the LEDs themselves rather than the lights that use them.
 

Tekno_Cowboy

Flashlight Enthusiast
Joined
Apr 2, 2008
Messages
1,680
Location
Minnesota
Neither would be better. An SSC P4 would be an almost direct replacement, and puts out considerably more light. A Cree LED would also be a much better choice, but the beam pattern is not as nice, in my opinion.
 

HarryN

Flashlight Enthusiast
Joined
Jan 22, 2004
Messages
3,977
Location
Pleasanton (Bay Area), CA, USA
Hi, contrary to popular opinion, it depends.

If you are looking for a really bright, red / orange flashlight, nothing comes close to the Lux III Red Orange driven at 1,000 ma. I know that does not sound like the ideal color, but if you goal is to see at night, it is just amazing.

For a white white application, the K2 TFFC versions are substantially better than the Lux III and put out a heck of a lot more light than any "conventional" flashlight if properly set up.

The advantages of the K2 TFFC are
- Its very small emission area, which makes it very easy to tune for throw, flood, or in between.
- It take abuse like no-bodies business, and if you are making up a flashlight, there is a good chance you will abuse the LED package.
- It has very good thermal characteristics
- It is less sensitive to loosing light output at higher operating temperatures than most LEDs.

In terms of total output, the 4 die packages have more light output, but - it is not so straightforward to make a tight beam with a large emission area.

Unlike Tekno-cowboy, I am not a fan of the S Semi products, but that does not make them bad, just a difference of taste.

The color tint can play a role as well. I personally like the WO white color bin, but others prefer a somewhat more neutral / warm color tint. The warmer color LEDs "CAN" have higher color content (CRI) which aids in seeing colors at night better, but not all warmer tint LEDs have better CRI ratings.

In the end, if I really want to see what is going on outside at night, I will still go for my red / orange Lux III light that Photon Fanatic made up for me. When I turn the light off, I still have a lot of night vision left.
 

Gunner12

Flashaholic
Joined
Dec 18, 2006
Messages
10,063
Location
Bay Area, CA
If you are speaking of white LEDs and the K2 (not TFFC) and Lux III, then I'd go with neither and look at the current generation LEDs like the Cree XR-E, XR-C, Seoul P4, Luxeonr Rebel, TFFC K2, and a few more. The current generation of LEDs have twice or more the efficiency of the previous generation of LEDs, which means for the same power, you'll get at least double the output with the current generation LEDs.

The number of contact points doesn't have much, if anything to do with performance.
 

Tekno_Cowboy

Flashlight Enthusiast
Joined
Apr 2, 2008
Messages
1,680
Location
Minnesota
Unlike Tekno-cowboy, I am not a fan of the S Semi products, but that does not make them bad, just a difference of taste.

The color tint can play a role as well. I personally like the WO white color bin, but others prefer a somewhat more neutral / warm color tint. The warmer color LEDs "CAN" have higher color content (CRI) which aids in seeing colors at night better, but not all warmer tint LEDs have better CRI ratings.

That hit the nail on the head. That is one of the main reasons I prefer SSC emitters. (other than the pleasant beam pattern)
The Neutral and Warm SSC emitters have a CRI of 93, and it definitely shows up in the quality of the light.

That's also one of the reasons I'm not a big fan of luxeon emitters. I have yet to find a tint/CRI combination I like in a Lux emittier.
 

HarryN

Flashlight Enthusiast
Joined
Jan 22, 2004
Messages
3,977
Location
Pleasanton (Bay Area), CA, USA
As you can see, emotions run high sometimes - typical of an internet forum. There are also the usual suspects for promoters of particular product brands.

Let's pick an example to make your life a bit easier. Let's assume you want to build a flashlight that has ALL of the following:
- 100 REAL Lumens coming from the front
- Run the LED within its design spec datasheet
- Will actually output the desired light for a period of time, not for the first 3 seconds
- Is easy to focus for a variety of beams with commonly available relatively small optics or reflectors. (say, 20mm dia)

First, you should assume that your "actual" output from the light will be about 50% of the "rated" LED output - regardless of emitter. There are many losses from many sources that we can go into later.

For those specs, either a Cree XR-E or a Lumileds K2- TFFC will work fine and you will have good output from either from reasonable size battery setups. It is simply not practical to have thorough experience with every kind of LED out there, so I will admit that I have not taken the time to test the SS products more that a quick test. Part of this is that I already had an investment in things that worked with Lumileds and Cree.

Flashlights in this range can be purchased at many hardware stores already, but they are a good place for a beginner. You will notice that most say "xx lumen rated LED", not the real output.

If you are wanting to do some hot rodding, then the decisions are more complex, for the LED(s), optics, battery pack and power management.

The LEDs are all inexpensive enough and easy enough to work with that it is worth building up a small test stand to make your own decisions before actually building the actual light. The light build can take some effort, so it is worth woth the effort.

In my case, I mounted different brand LEDs on Al rods sticking up from a 2 x 4 board. I used aligator clips to connect each one to a buck puck (leddynamics), either one at a time or in series.

You can easily do 1:1 comparisions and just lay on top of the LED your optics or reflectors, and then study the beams on the ceiling. Wear sunglasses and have fun. :D
 

Sober

Newly Enlightened
Joined
Nov 24, 2008
Messages
40
Location
South Africa
Hi There,

I want to build a multi emitter setup using red Luxeon III's and a 4D M@g, what heat sink, reflector, driver and batteries would you guys suggest?

I would prefer to run this using D-sized Nimh which I already have.

Regards
 

blasterman

Flashlight Enthusiast
Joined
Jul 17, 2008
Messages
1,802
That's also one of the reasons I'm not a big fan of luxeon emitters. I have yet to find a tint/CRI combination I like in a Lux emittier.

I don't think it exists.

I built some light bars with neutral white Lux IIIs awhile back, and while brightness isn't an issue their inherently low CRI made anything placed underneath them look terrible.

K2s and Crees much better in this respect.
 

HarryN

Flashlight Enthusiast
Joined
Jan 22, 2004
Messages
3,977
Location
Pleasanton (Bay Area), CA, USA
Hi There,

I want to build a multi emitter setup using red Luxeon III's and a 4D M@g, what heat sink, reflector, driver and batteries would you guys suggest?

I would prefer to run this using D-sized Nimh which I already have.

Regards

Hi - sounds like a nice project. Let's run the numbers.

- Lux III Red or Red / Orange
- Vf = approx 2.5 volts at 1 amp, nearly 3 volts at 1.5 amps max rating
- Red / Orange LEDs are more sensitive to loosing output as things get hot, so it is essential to maximize the thermal path, as well as to drive them closer to 1 amp then 1.5 amps, at least in my experience.

- NiMH x 4 cells = approx 1.2 - 1.4 volts x 4 = 4.8 to 5.6 volts in, dropping to about 4.4 volts at fully discharged, and plenty of current capacity.

There are a couple of approaches to the driver / current regulation that can work:

Buck drivers
- In this case, the V bat must be > than ( LED Vf + overhead voltage for the circuit) Usually, this overhead is 1 - 2 volts, but sometimes is less
- Since you only have about 2 - 2.5 volts of overhead to work with, the absolute safest method would be to run all of the LEDs in parallel
- 4 x 1 amp = 4 amps, which is outside of the capacity of any buck driver I know of on the market, so a single driver is probably out.

- If you want a lot of "features", like dimming, sos, etc, then the drivers from taskled.com are very nice. http://taskled.com/compare.html, but you will need 2 of them.
- I am frankly not sure that in a hunting situation, you want to fool around with a driver that has a lot of software features.
- Another option would be a buck puck, which is a nice driver from these guys http://www.leddynamics.com/LuxDrive/buckpuck.php but - you will need 4 of them. (all run in parallel)

Boost Driver
- In a boost driver, the Vf of the LED string is greater than the V battery pack
- In this case, you put all 4 LEDs in a series string, giving you a Vf of approx 4 x 2.5 = 10 volts.
- A suitable driver would be any boost driver (such as from the sandwich shoppe or again, taskled, such as the maxflex.
- This will work fine, but make sure you have a clear separation of the LED thermal path from the ground and from each other. The thermal pad of the emitter is not neutral (due to ESD protection). I think the stars are electrically isolated, but double check.

Direct Drive or Lightly Resistor managed
- This method is sometimes frowned upon, but if done in a reasonable way can be quite suitable for a project like yours.
- Imagine 2 strings x 2 LEDs in series
- Vf = approx 2 x 2.5 – approx 5 volts.

- You will notice that this is so close to the V battery pack that you could actually get away with just hard wiring the two strings to the battery pack with no drivers at all.
- If I were doing this project, I would put about 1 ohm resistance in series with each LED string. The resistor needs to be rated for the current flow – 1 watt in this case of a 1 ohm resistor.
- Brightness will decrease with battery discharge, but it will also be very easy to see when the batteries are getting low, something that is valuable in the field.
- If you have quality NiMH D cells, they will discharge 2 amps for a very long time with quite flat voltage over 60 – 70 % of the discharge curve. The output will look just like your battery discharge curve.
- You actually will have a hard time seeing a difference that is not at least 25% or more, as you vision for brightness is rather complex and not linear.
- I personally would do it this way, especially if dimming is not a factor, as it is as rugged as your soldering capability, and there is no inductor hum.

Reflectors and heat sinks – need to think about that one, but basically use the largest reflector that will fit in the head.
 

Sober

Newly Enlightened
Joined
Nov 24, 2008
Messages
40
Location
South Africa
Wow, thanks for all the info. I'll have to think about these options.

I would prefer going with the driver router, no SOS is needed.

Regards
 

HarryN

Flashlight Enthusiast
Joined
Jan 22, 2004
Messages
3,977
Location
Pleasanton (Bay Area), CA, USA
The nice thing, is that your light can be mocked up on a bench before building it. It is actually very little effort and most of the parts are pretty inexpensive.

It probably will cost more in shipping than actual parts.

If I were planning a light for hunting, I would definitely mock it up first and make sure I liked how it was going to work. Also, make sure you will be able to see everything that you "need" to see.
 
Top