Results: Testing XM-L, MC-E, SST-50, and XP-G emitters **Updated**

Great setup and great measurments, Match :)

I compared it to the Cree datasheet specs, and it matches quite well up to 2.5A, then your values are a bit below theirs (3% at 3A). I guess that's due to their measurement method: If I remember correctly, they only power the LED for a small fraction of a second to avoid increasing the junction temperature. (Unfortunately I can't recall where I read that.)

Oldbobk,

"Run in series, you have to have a resistor that will run the wattage of the light"

I understand perfectly what you say but just because the above sentence is incorrect what you get is the opposite. In parallel with the emitter the resistor must be much more powerful than one in series.

With an IMR 18650 or an IMR 26650 I get right around 4.5A with the MF UF dropin. I figured they were DD. I keep saying I need to swap the NANJG on my triple XP-G with this UF driver.

I'm perfectly happy with an XM-L @ 2.8A. Actually for a P60 drop-in I think less would even be preferable. I have some A123 LiFePO4 cells that I've been playing with and they turn the MF UF drop-in into a very reasonable light. It would probably be too underdriven for most (1.4-1.5A) but I'm quite comfortable with it.

This is quite possible. For consistancy, all measurements were taken exactly 10 seconds after power on, which also allowed me to get the current dialled into +/- 1ma. At higher power levels, it was quite apparant that output was dropping almost instantly due to, like you stated, junction temperature. Considering the difference between the test heatsink and a standard flashlight, I feel these losses would be considerable at high power in normal practice. One of these days I'll direct mount a bare emitter to the test platform for comparison.

Can someone please decipher arenat for me?

Foy

'Tis in English.

There's two ways you can connect a resistor. Across the LED (in parallel) or between one contact of the LED and the equivalent contact on the battery or driver. (In series)

Hope that is slightly more intelligible.

And since i forgot to say so:

Thanks Match for this!

I reckon that was an enormous amount of work so thanks for doing this.

Since my language skills are inexistent maybe posting what Match stated above some people can get it...

"The lumen difference for any torch or drop-in much above 3a woud be negligable outside a I.S. Any perceived difference would be almost entirely based on the reflector/lens combo. This is going to be my new wish from manufacturers...."Better xm-l reflectors, not more amps!"


Thanks Match!

...wait. wow...

I'm.. I'm.. actually satisfied with my current t6 dropin? no more lusting for higher current drivers, when 3A is more than good enough?

... whatever will I do now?

(seriously, thanks. You've laid that question to rest for good. I gotta ask though, since those 80 stars were for work, what would you have done if you'd blown them? Pushed back the project, or did you over-order 'just in case'?)

Why, go with multi-xml emitter setups of course, silly! :stuck_out_tongue:

[quote=devman]

(... I gotta ask though, since those 80 stars were for work, what would you have done if you'd blown them? Pushed back the project, or did you over-order 'just in case'?)

[/quote]

First rule of business: Why buy one when you can get two at three times the price! (bonus points if you know the movie that's quoted from)

I already had a few extra. After the first run and seeing no evidence of damage, I was content to try out some of the other 80 ordered.

JohnnyMac:

Na. I'd have to go the toilet-plunger route to do that right, and I'm not a big fan. (and unlike match, I don't have access to large blocks of copper and a holesaw to make a svelte version practical)... so it's either go play in the knife forums, or start modding smaller torches. We'll see which based on DX shipping times.

Match:

Contact (google remembered for me). I remember being very disappointed by that movie at the time. The opposite of disappointment is.. mmmm... copper barstock... machine tools that aren't dremel... drool...

First rule in government spending: why build one when you can have two at twice the price? Only, this one can be kept secret. Controlled by Americans, built by the Japanese subcontractors. Who, also, happen to be, recently acquired, wholly-owned subsidiaries...
contact ?? I would have guessed ironman or lord of war



If you have a shot of the driver board, or just the part # off the chips on it, you can tell how it works.

[/quote]




He is correct in that this is in error:

"Run in series, you have to have a resistor that will run the wattage of the light"

It doesn't because the resistor doesn't have to have the same voltage across it as led Vf.

However, your main problem is that putting a resistor in parallel will lower the equivalent "resistence" and therefore not really lower the current much through the emitter. IOW, you'll mostly just heat up that resistor.

Thank's Match! So the general consensus is the XM-L should be driven between 2.4-3.2A for optimal results current/lumen/heat ratio.

I find the KD 2.8A driver the best for the job even if it works at 2.8A for a dozen min or more then drops gradually as Vbatt drops.

We really need some decent buck/boost drivers for cheap!

And for higher voltages 8.4V with few memory modes also!

Good job as usual Match!

Awesome job Match! I think that pretty definitively settles all that nonsense :P

Page 13 for this? Never!