Dual XP-G on a 10mm engine (LSD)

Linger

Flashlight Enthusiast
Joined
Feb 17, 2009
Messages
1,437
Location
Kingston ON
You've seen the xpg mods and the multi-emitters. Nailbender and I took it one further and used an xp-g engine to dual-mount two emitters: the most efficient format for your upgrades, mods, or custom builds.
(Integrated sphere tested: 2xR4's with 515 lumens out the front @ 2.5 amps)
IMG_0963.jpg

A CPF creative collaboration with Nailbender and Linger. Two xp-g R4's, reflow soldered on a single board. These emitters are in parrallel, dividing the current between them and each facing the full voltage.


Background
Never a person prone to exaggeration, David Chow wrote that our jaws would drop when the specs for the new xpg were released. Since the introduction of the xpg a number of new applications have been tried. The rush of announcements in the homemade / modified section reveals widespread acceptance of the xpg and an embrace of the remarkably high efficiency this emitter provides. One of the common themes among these mods is the floody, unfocused image that result from the xpg's larger emissive area, and the use of reflectors and optics designed for the xre / xpe.

Comparison
Here is a head to head comparision of the xpg vs. xre. I control for all other variables with application of standardized components. Both lights use AW 18650, are built on identical DIY copper pill + OP reflector, two 3.0v-8.4v 5mode circuit (the Solarforce circuit). Lights are set on diningroom table ~3ft from beige wall. Hosts SF C2 and Solarforce L2
Low - xpg on Left. Hotspot is much larger and less defined then xre, blends very well towards sides.
IMG_0929.jpg



High (switched hosts) The xpg almost appears as if it were two xre emitters, one with a spot and dim corona just like xre beside it + and additional xre with a boom-m or rippled optic creating that nice bright distribution away from the center. Even in comparisons of the beam edge, the xpg is significantly brighter.

IMG_0927.jpg



Outside the effect is pronounced. The xpg (with xre / xpg reflector) really lights up an area.

Possibility
Here is xpg over-driven to 2.5a, head to head with an mc-e. Both lights use identical drivers for this shoot, though reflectors were much different (xpg smo d26 drop-in, mc-e MG PLI)
100_1693.jpg


Thinking about how best to use this little emitter, I was stuck on something I'd never heard my partner say before, her words when she first saw the xp-g package: "wow, that's tiny.":shakehead The little xpg is small, Cutter compares its demensions to the Rebel.

Implementation
A few weeks ago I put an xpeQ35A, who's small footprint the xpg shares, and placed it on the contact pads beside a Seoul P4.
IMG_0951.jpg

This was one of the 'high cri' binned P4's and while the added depth the higher CRI provided was great I found them a bit cool for wildreness work and decided to add the warm binned Cree. I put in on there with the intention of taking it off. It was a proof, a demo if I could 'reflow' solder w/o paste, to test mixing and blending options between two emitters. (highly successful experiments)

Linger's Line
And this leads to the double, a line of xpg's together on a single pad, sharing the contact points.
IMG_0931.jpg

Two xpg's on a 10mm board, the most efficient LED's at the lower end of the power scale, used together to increase output. This configuration take advantage of all of the xp-g's leading attributes. Small package size for tightly grouped emitters. Large emissive surface and floody beam optimal for grouping multiple emitters together. High efficiency and low vf are ideal for single cell applications, of course multi-cell configurations work well with high-efficient emitters as well. Two xp-g's reflowed in a line on a single engine puts out a brilliant floody beam, slight donut that is not visible outdoors. With ~3.5vf, it is perfect for single cell applications. A higher current driver is indicated as the emitters, in parrallel, can take 2000ma (1000ma each) @ spec. Jtr's White Led testing seems to indicate the 'sweet spot' for the xp-g is at 1400ma. If you used this line configuration, an 8x7135 driver providing roughly 2800ma may be an optimal set-up. Beam quality varies considerably with reflector. Ledil Boom's are fantastic at covering up the entire crispy creme donut chain and if you put one on the xpg line you'll have a perfectly smooth beam. The Fraen may be another excellent option, kicking the range up and providing a defined hot-spot , but this needs additional verification (sun's come up here).:thumbsup:

Nailbender will be offering these, available along with his wide range of d26 drop-ins.
http://www.candlepowerforums.com/vb/showthread.php?t=248771

Best,
Linger
 
Last edited:
Re: Linger's Line, 'multi-die' dual xpg on a 10mm engine

Can you make a dual or quad XP-E/G side firing emitter? The dual side firing would have the LED mounted back to back on a thin-ish vertical heatsink..
 
Re: Dual XP-G on a 10mm engine

wechnivag, I haven't heard anything to suggest that Cree will be going back to side-emitting. Two emitters mounted back-to-back would multiply the waste heat and create a challenging thermal path. I see where you're going with it, but remember Cree builds the xr-e, xp-e, xp-g. This thread is to share the idea of dualmounting xp-g's and introduce it as a means of multiplying the efficency gains: noting that the small package size includes a large emissive surface relative to total footprint, the decrease in waste heat produced, the efficency gains. The parallel assembly extend the efficiency range of the most efficient emitter. Theoretically, the dual xp-g will remain above 130lumens/watt to 1amp.

Here's why it works: the dual xpg pictured next to single and quad emitters (xre, mce, p7, ssr-90)
IMG_0976.jpg
 
Last edited:
Re: Dual XP-G on a 10mm engine

Very cool!! Does having them soo close together create an issue with heat?

How does an XPG compare to a XRE as far as heat output/heat sink needs go with them both running at say 1000mA?
 
Re: Dual XP-G on a 10mm engine

I have been running two side by side on a 14mm board using a 2.5 amps driver and it gets about as warm as a XPG running at 1.4 amps.
While is has a small spot in the middle much like a MCE it is very bright , and outside the spot is little noticed.
I measured at a little over 500lumens.

Dave
 
Re: Dual XP-G on a 10mm engine

I have been running two side by side on a 14mm board using a 2.5 amps driver and it gets about as warm as a XPG running at 1.4 amps.
While is has a small spot in the middle much like a MCE it is very bright , and outside the spot is little noticed.
I measured at a little over 500lumens.

Very cool. Any chance of running it as 2 emitters wired serial on a 14mm board? For bicycle dynamos, its easy to ramp up the voltage, but the dynamos are pretty close to constant current once up to speed.

Steve
 
Re: Dual XP-G on a 10mm engine

Very cool. Any chance of running it ... wired serial on a 14mm board?
Can you find a board wired for serial so close?

The line configuration could be a strong contender for traditional multi-emitter, multi-bin tint blended lights. I see it as a more versatile option than the standard 3 or 4 emitters + optics to improve colour rendiditon and depth. Imagine a Mag D with 3x dual xpgs, each board an xre q35b and xpg r4 or r5. Usable with standard reflectors and eliminating the hollow through use of multiple boards.:)
 
Re: Dual XP-G on a 10mm engine

Can you find a board wired for serial so close?

The line configuration could be a strong contender for traditional multi-emitter, multi-bin tint blended lights. I see it as a more versatile option than the standard 3 or 4 emitters + optics to improve colour rendiditon and depth. Imagine a Mag D with 3x dual xpgs, each board an xre q35b and xpg r4 or r5. Usable with standard reflectors and eliminating the hollow through use of multiple boards.:)

Looking at the spec sheets for the XRE and XPE/G emitters, there's 5.6mm between the contacts on the bottom of the emitter. If you take 2 XPGs, and put them side by side, with the + side of 1 XPG, butting up against the - side of the other one and connected via a solder joint so that they are connected in series, there's 2.3 + .5 + .5 + 2.3 = 5.6mm between the contacts across the series XPG emitters.

So, looking at the numbers, you can mount a pair of XPG's in series across the contacts for an XRE board (in fact, you could make a DIY MCE and get a 2s2p XPG cluster on the pads for an XRE).

You'd have to electrically isolate the contacts between the emitters from the board, but maybe a thin layer of AA could do the trick?

If this works, any sized XRE board would do the trick. I'm pretty sure I've seen XRE boards down to 14mm.
 
Re: Dual XP-G on a 10mm engine

I see what you're saying. These emitters are usually re-flowed onto the boards. What you're talking about isn't just electrically isolating them from the board, but creating a unique electrical path,your own set of leads from one emitter to the next.
syc said:
connected via a solder joint
You'd want this joint to be low resistance as the bottom of the board is a bad place to add more heat, but you'd also want them to be micro thin as you'll between the emitter and it's thermal outlet. You get me? You'd need a wiring harness that was non-existent-ly thin or to etch the path down into the engine so that the remainder of the emitters were still in direct contact with the pads for thermal transfer. You'd then solder this harness directly to the bottom of each emitter.
You're a bold and daring person to undertake that mission syc

Wouldn't you just prefer an xpe Q3 5A + xpg r4 1C? (I'm waiting for solder paste two weeks now, or I'd try to get you some beam shots of this blended configuration)
 
Re: Dual XP-G on a 10mm engine

"wechnivag, I haven't heard anything to suggest that Cree will be going back to side-emitting. Two emitters mounted back-to-back would multiply the waste heat and create a challenging thermal path. I see where you're going with it,"

I recall a web page with multiple experiments using dual side firing, or even triple side firing XR-Es. Can't seem to find it now. He cut the reflector into half and thirds to account for the spacing between the backs of the emitters, and still get the correct focus point. The results were awesome!
 
Re: Dual XP-G on a 10mm engine

I recall a web page with multiple experiments using dual side firing... X # R # - # Es.
...well just as a back-up I tried google for side emitting x # r # - # e :
http://www.google.ca/search?hl=en&safe=off&client=firefox-a&rls=org.mozilla%3Aen-US%3Aofficial&hs=okW&num=50&q=side+emitting+XR-E&btnG=Search&meta=&aq=f&oq=
and did not find the emitter.:shrug:
Lux used to make side emitter's, I's and III's and I have an old V. Cree's website doesn't indicate that they ever used that format for the x-class. What where you thinking of :thinking:
 
Last edited:
Re: Dual XP-G on a 10mm engine

That's pretty cool looking, if you intend to use it as a two eyed mule,
but put it into a reflector, it'll probably be a "dunkin donut" special :crackup:

I had some thoughts similar to a post above a few weeks ago, but never acted upon it,
that's putting 2 XPG back to back to form a "XPG 360". It'll probably work well in Incan reflectors,
as the dual XPG "ball" if stick on to the tip of a thin copper pole deep inside the reflector, may give real nice beam, incan style.
 
Last edited:
Re: Dual XP-G on a 10mm engine

ma_sha1 - look above at the 'floody' more diffused beam of the xpg in xpe reflectors.
I grabbed a few quick pics before going to work, I plan on replacing them afterwards as they are poor quality (the pill was assmebled for a d26 drop-in so swapping in different reflectors for pic was a very diy situation)

xpg line w 34mm Frean - this reflector can put out a decent hotspot, its multiple focal points working to blend light intensity with-in the core
IMG_0979.jpg


xpg line w ledil Boom - Depending on focus this goes from a hotspot with split hemispheres, to generally bright center (no darkness) with two 'eyes' hotspots near the edges of the bright center; the image below is closer to representing the two hotspots; in focus the upper drops closer to the center and the 2nd appears below

IMG_0978.jpg


(Again, appologies for these shots> i needed 4 hands for the set-up)
 
Last edited:
Re: Dual XP-G on a 10mm engine

You'd want this joint to be low resistance as the bottom of the board is a bad place to add more heat, but you'd also want them to be micro thin as you'll between the emitter and it's thermal outlet. You get me? You'd need a wiring harness that was non-existent-ly thin or to etch the path down into the engine so that the remainder of the emitters were still in direct contact with the pads for thermal transfer. You'd then solder this harness directly to the bottom of each emitter.
You're a bold and daring person to undertake that mission syc

Me? My attempts at reflow have been pretty sad.
It might actually be easier just to solder the contacts on the topside of the emitter together.

Wouldn't you just prefer an xpe Q3 5A + xpg r4 1C? (I'm waiting for solder paste two weeks now, or I'd try to get you some beam shots of this blended configuration)
I'd love to have one, just in series :)
Or else, if you can sketch out a voltage divider/current doubler circuit for me that will have 700ma input, I'd pick up one of these parallel LED setups.
 
Interesting...:twothumbs

I wonder whether a triple XP-G light engine would turn a MagLite with stock reflector into a real monster - a XP-G perfectly centered for maximum throw and one on each side for extended flood. What do you think about it? I "only" saw a dual XP-G Mag (quite nice, too^^) yet.
Or to maximize throw, a XP-E could be used...
 
Very interesting! Great work!:thumbsup:

Have you tried increasing the area of the contact points on the MCPCB to increase thermal conductivity? I don't the electrical contacts would be effected much by sharing, but I am concerned about the center contact. Any thoughts?
 
I do not have a temp probe for my DMM so my subjective (touch) impression is that the emitters do a fine job shedding heat to the engine and thus to the pill. In free air the drop-in ramps up heat pretty quick (as expected from ~2.5A), touching the emitters (ceramic substrate and less conductive to my finger?) I felt less heat then touching the copper pill.
 
Just curious how would work with an aspherical lens.
Eny possibility for some beamshots focused and slightly defocused?
Thanks.
 
Background
Never a person prone to exaggeration, David Chow wrote that our jaws would drop when the specs for the new xpg were released.
I vaguely recall the word "guarantee" in there too, but I'm still waiting for my jaw to drop. :popcorn:
 
Top