KL5 on 3x123 --- DO NOT DO IT!

Well, I've been reading some posts speculating on what the KL5 will run on and did a voltage and current draw test. Conclusion is the KL5 uses a boost circuit. I will post graphs later but for now, do not attempt to run the light on 3x123. Even 2 li-ion's may be questionable depending on the forward voltage of the LED. On my particular KL5, the current draw starts to increase very rapidly at a Vin of 7.3V. This means it's going into direct drive at that point. Two Li-Ion's cause the current to go to 910mA which means the LED is seeing about 910mA. I presume 3x123's will drive the LED beyond an amp, probably not a good thing for the circuit as well.

CM

Here's a plot of input voltage vs input current. Data from a KL4 is plotted on the same for comparison. Effiency of the KL5 seems to really suck rocks if this is driven at the same level as a KL4. I was hoping that with the efficiency in the mid 70 percent that the circuit was a buck-boost /ubbthreads/images/graemlins/frown.gif



BTW, I have similar graphs for some Badboys, another KL4 and a KL3. Very interesting data to ponder.

CM

Well, how bright is it on 3x123's? /ubbthreads/images/graemlins/grin.gif

Yeah, but...so? If the Luxeon in there is a first-string unit (don't see why not) it would be no different than a Space Needle II, except perhaps the smaller reflector would have a different beam pattern from the Mag's...hmmm, and mebbe even for the better...and the heatsinking may not be as good (or it may be even better, depending on the KL5's mass and specific internal thermal pathway configuration, to which I have not been privy - yet).

So CM, what does all that analysis mean to use non-techy folks? Are you saying the KL5 is a KL4 with a different reflector?

[ QUOTE ]
Sean said:
Well, how bright is it on 3x123's? /ubbthreads/images/graemlins/grin.gif

[/ QUOTE ]


Really, that's all we really want to know....

LoL

[ QUOTE ]
pahl said:
[ QUOTE ]
Sean said:
Well, how bright is it on 3x123's? /ubbthreads/images/graemlins/grin.gif

[/ QUOTE ]


Really, that's all we really want to know....

LoL

[/ QUOTE ]

Yea guys, don't make me be the first one to do it. I don't have a KL5 yet so I can't do it. /ubbthreads/images/graemlins/tongue.gif

It's your perogative to overdrive the KL5, my post was just to warn anyone who wants to maintain their Surefire warranty on their $160 light. If you want to drive it at 1.5A, that's your choice, not mine.

I did run it at 900+ mA and it was similar in total output to an X3T driven by a BB750 /ubbthreads/images/graemlins/grin.gif Did not run it using 3x123's, I just got the light and the newness has not worn off yet. But when it does, whoo boy, I'm gonna have some fun with this $160 light hehe.

CM

[ QUOTE ]
this_is_nascar said:
So CM, what does all that analysis mean to use non-techy folks? Are you saying the KL5 is a KL4 with a different reflector?


[/ QUOTE ]

I haven't opened it up yet but since it is a boost circuit, I have no reason to believe surefire would redesign an existing boost circuit. The one thing puzzling to me is the relative inefficiency of the converter inside. So to answer your question, the KL5 is really a KL4 with a bigger reflector. It would be interesting to see what they use in the KL6 since the Vf of the 5W spans the buck/boost range for a 3x123 configuration. A buck would go out of regulation pretty early for many LED's but you can't use a boost. You need a buck boost for a 3x123 to properly suck the batteries dry.

CM

My understanding is:
- The circuit in KL5 is very similar to KL4.
- The KL5 circuit pushes more current to LED than KL4 /ubbthreads/images/graemlins/smile.gif, or the efficiency of the KL5 circuit is worse than KL4 /ubbthreads/images/graemlins/frown.gif, or a LED in your KL5 has much higher Vf than a LED in your KL4 (I think it doesn't justify that difference, though).

How do you guys think?

[ QUOTE ]
CM said:

So to answer your question, the KL5 is really a KL4 with a bigger reflector.

[/ QUOTE ]
That would explain why the pamphlet that came with the L5 shows the exact same text and graph (as if copied and pasted) as the one that came with the L4.

It looks to me like the KL5 will be running just fine on one Pila 168s right? Maybe not with the run time of L4 though according to the graph.

Jan

CM, that is very interesting data! The badboy graphs would be also very interesting for me!
On question to your efficiency statement: The VF of the L5 LED seems to be higher. Perhaps the LED is driven harder in the L5, which together results in a higher input current?

JanCPF,

The KL5 will run off a single Li-Ion (that's how I have mine setup).

PeterB,

I'll post BB's graphs later. The BB's have excellent efficiency, in the high 80's and low 90's when Vin is close to Vf of the LED. The Vf difference is related more to the LED's themselves than the drive level I believe. Two KL4's I have show Vf that differ by almost a volt.

CM

[ QUOTE ]
CM said:
JanCPF,

The KL5 will run off a single 123 (that's how I have mine setup).

[/ QUOTE ]

But will it be in regulation with just one 123? On the graph it looks like the 3 volt will be to low, and 3.7 volt from a Li-ion will be okay.

Jan

Correction, I run mine off a single Li-Ion.

Looks like no one made any astute observations from the graph I posted above with the data from a KL4 plotted on the same graph as the KL5 /ubbthreads/images/graemlins/grin.gif I plotted them on the same curve for a purpose. Here's a small clue: What's the significance of the inflection point (the place where the slope changes) above 6V on both curves? Hehe, I love brain ticklers. Any one with the right answer can claim bragging rights to being an electronics expert. I need to entice the nerds to step up to the plate.

CM

[ QUOTE ]
CM said:
What's the significance of the inflection point (the place where the slope changes) above 6V on both curves?

[/ QUOTE ]

That's the point when it drops out of regulation?
And it looks like KL5 is driven harder than KL4.

lak is right - It drops out of regulation and goes into directdrive mode.

Jan

If this circuit works like BB, the drop out will occurs when Vin exceeds LED's Vf.