KL4=1500mA L6=650mA ... why L6 is so much less?

I noticed that my KL4 always get pretty hot when it runs, but that my L6 does not even come close to in terms of heat. Both are supposed to have a 5 watt LED (right? /ubbthreads/images/graemlins/confused.gif ) so I decided to measure how much current its going on each one with different battery combinations.

First I tried the KL4 head on a 1xcell Vital Gear with a 4.2V rechargable cell. I measured 1400-1500mA on the KL4.

I then re-ran the KL4 test but on a 2xcell Vital Gear with two primary 123 cells (two different brands), and I got about 900-950mA. The KL4 gets very warm, but not nearly as bad.

Then tested the L6, and I only get about 650mA (either with 3 brand new SF 123 cells, or with two fully charged Pila 150s batteries). This would at least explain why the L6 barely gets warm.

Why if the L6 so much lower? Shouldn't the L6 be able to dissipate much more heat on the larger head and body combo? Am I missing something here? Is it simply that the KL6 was setup more conservatively by SF? Or is this due to the Luxeon lottery and perhaps my KL4 having a better binned LED than on the KL6?

The other thing, why would the KL4 head draw more on a single 4.2v cell? I can only imagine that the circuit on the KL4 head is terribly ineficcient when trying to work from the single 4.2V cell than working at the "normal" 6V. Any other reason for this? Even if I assume that the "normal" current for the KL4 is about 900-950mA, why is this so much higher than the KL6?

Finally, being that I would like to drive the KL6 much harder given the excellent heat-sinking on the head/body, what options do I have to make this "better"? /ubbthreads/images/graemlins/naughty.gif

Will

Will,

The L4 has a boost circuit in it. This means that the converter is boosting the voltage of the battery to a voltage necessary to run the luxeon at the specified bias. In this case 660mA to the LED. As the voltage of the batteries drop, the converter will draw more current. This is why the L4 draws more current from a single li-ion cell that it doesn from 2X123. The converter has to do a lot more boosting from the single li-ion than it does from the 2X123 causing the L4 to run hotter on the li-ion cell than it does on the 2X123 battery configuration.

If the L6 is a three of more cell light, then it is running a buck converter. It drops the battery voltage to a voltage appropriate to attain the specified bias to the LED, which is probably in the 700mA range. The reason that the current draw is so low with the L6 is because the batteries are supplying more voltage than the LED needs so the current draw is less than the current getting to the luxeon. With boost converters, the light will draw more current than is reaching the LED.

The readings that you took really do not reflect the amount of current getting to the luxeon in each light.

If you want to boost the output of the KL6 and get a meaningful increase in output, then you'll need to double the amount of current getting to the LED to get a roughly 50% increase in output. Just keep in mind that you WILL have heat problems doing this.

Tony,

Thanks so much for taking the time to answer my questions so completely - much appretiated. I do understand what is going on now /ubbthreads/images/graemlins/buttrock.gif

On the KL6, you make it sound as if the current setup is fairly optimal as it is. I had though about replacing the OEM setup with a Badboy (so that I can use 3-4 cells as the OEM setup) set to 1000mA. Is it basically not worth the hasle and expense to do? /ubbthreads/images/graemlins/confused.gif

I did look at the OEM module and because it is using a bare emiter I can't tell what bin it is /ubbthreads/images/graemlins/frown.gif

Will

Re: KL4=1500mA L6=650mA ... why L6 is so much les

Tony nailed it dead on. And the L4 will get hotter because it's a smaller light and has less mass for the heat to spread out into, so it saturates with heat faster and gets hotter.

Re: KL4=1500mA L6=650mA ... why L6 is so much les

The easiest way that I have had to explain regulators is that the power in has to equal the power out. So if the source voltage is less than the Vf (approx 6.84V) you typically have to supply greater current than is required (600ma-700ma typical draw at the lux5 emitter). If you have more voltage than Vf, the current demands decrease.

Also keep in mind the efficiency of the regulator circuit will vary with Vin so that will alter the current draw on the primaries.

Will, I had my L6 modded to run at 1200ma, and I absolutely love it. It's really quite a flame thrower now, and I still get about an hour of full brightness out of it. I recommend getting it done.

Mike

Mike,

PM sent asking for details. Sounds that you have what I would like to have /ubbthreads/images/graemlins/buttrock.gif

Will

Mike,

PM sent too /ubbthreads/images/graemlins/grin.gif
Is this also possible with the L2 and L4?

Re: KL4=1500mA L6=650mA ... why L6 is so much les

That would be cool with an L2...

-tom

Re: KL4=1500mA L6=650mA ... why L6 is so much les

Vin x Iin = Watts

(L4/E1e) 1 X LiIon: 1500ma x 3,7V = 5,55W
(L4/E2e) 2 x 123: 950ma x 5,5V = 5,225
(L6/M3) 3 x 123: 650ma x 8,4V = 5,46W


See that the different current draws and voltage levels add up to pretty much the same Wattage in, I might be off slightly but makes sense overall. The Kl4 seems to work more effieicent on 2 cells and the Kl6 is driven slightly higher, all this makes sense too and matches user reports.

Klaus

Re: KL4=1500mA L6=650mA ... why L6 is so much les

If you are going to run a 5W on 3 or more cells, you'll need a buck converter, like the DownBoy. The BadBoy is a boost converter and with a 5W, 2X123 would be the maximum battery configuration.

Assuming that the L6 runs in the neighborhood of about 700mA, you'll need to double the current to the LED to 1.4A to get less than a 50% increase in light output. The rest of the current just goes toward generating more heat.

Although I can appreciate MikeyTheBull's enthusiasm regarding his KL6 mod, I don't think that you could characterize the modded KL6's output as being like a flame thrower. Well, it will put out lots of light, as any 5W being driven to spec or more will; but the KL6's reflector is still too small for any sort of real throw.

It'd leave it stock and enjoy it.

Re: KL4=1500mA L6=650mA ... why L6 is so much les

Chop,

We would not be considered true flashaholics if we left our lights stock, would we? /ubbthreads/images/graemlins/grin.gif

I went ahead and modified my KL6 to draw 1000mA, and to the naked eye it does look brighter. Maybe only 15-25%? I really can't tell exactly, except it does look brighter.

I did use premium thermal compound (Artic Silver 5) when putting the head together and I can tell that the head now gets warm, where before you could hardly tell. How much is just raw heat vs. the heat better transmitted to the head, I will never know, but I don't care too much as even now the KL6 is not even close to the heat level on my KL4 running from one rechargable CR123 cell. This KL6 head simply has a lot more heat sinking capability /ubbthreads/images/graemlins/tongue.gif

I did follow your advice regarding not going to far and I refrain myself from trying 1.2A or higher. Still, if I kill the emiter, I will simply solder another one - no big deal as access to the LED is trivial on the KL6.

I did order one of the light meters in the GB that closed, so I will report the new KL6 against my stock KL4 head as a reference. I will report the resulting readings later on.

I just could NOT keep this one stock - I could not resist /ubbthreads/images/graemlins/wink.gif

Will

Re: KL4=1500mA L6=650mA ... why L6 is so much les

[ QUOTE ]
Klaus said:
Vin x Iin = Watts

(L4/E1e) 1 X LiIon: 1500ma x 3,7V = 5,55W
(L4/E2e) 2 x 123: 950ma x 5,5V = 5,225
(L6/M3) 3 x 123: 650ma x 8,4V = 5,46W


See that the different current draws and voltage levels add up to pretty much the same Wattage in, I might be off slightly but makes sense overall. The Kl4 seems to work more effieicent on 2 cells and the Kl6 is driven slightly higher, all this makes sense too and matches user reports.

Klaus

[/ QUOTE ]

Yes, Klaus, you got my point. Boost/buck regs typically have an efficency curve that varies with voltage, lower (to a point then things get ugly) more eff and higher less eff.