Build Log, interim name '30k deathray' more XM-Ls than you can shake a stick at.

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
I was feeling the need to upgrade a handheld spotlight in a multi-aspheric sort of manner, like databank 70, or hexaspherical.

Well from a little 5" spotlight stuff got constantly upgraded untill I finally have settled on a general size /power level.



Initial renders for sizing everything up

Some plans details have changed but the general layout is the same.

77861ebc.jpg

not just fun renderings either ;) Got the lenses sorted out, and the mcpcbs, though that could change depending on some other projects by others, and of course the host.

Stuff thats done now:
Lens selection, ahorton aspherics look like the way to go.
Buying Host, lenses and KD MCPCBs
Buying LEDs : XML U2 bin brightness, on cutter site sure looks interesting


Working on now:
Lux boosting R&D. 23.5K cp with 1 XM-L T6 and ahorton aspheric isnt quite enough, U2 bin numbers soon to come, and then other tricks too hopefully.


Stuff left to work on.
Spacers, for standing off the lenses from the LEDs
Mounting plate: Copper, Aluminum? Not sure yet, but either way its gonna be beefy
Heatsinking, lots and lots of heatsinking, probably a fan or two, dont think I wanna carry around enough aluminum to passively cool this.
Powering, National semi has some nice chips which should make perfect drivers, and a few other fun things, and big batteries of course.

Long way to go, but it'll be a monster in the end.

Wont be working on this all the time, busting it out real quick, but it'll get done eventually.
 
Last edited:

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
And lets put this spot on hold in case I need it later
 

srfreddy

Enlightened
Joined
Sep 21, 2010
Messages
921
Location
New England
I've heard that the cutter MCPCB's are pretty bad for heatsinking though.... Thats a Monster! Is that 36?
 

Patriot

Flashaholic
Joined
Feb 13, 2007
Messages
11,254
Location
Arizona
Wow! Really interesting project and it seems that the XM-L is the ideal platform for this kind of array.
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
Those are Kaidomain MCPCBs which I picked up after seeing the cutter ones didnt even have a copper fill on the top, I scraped at one, and its not just a thin layers of FR-4 on top of the aluminum. And like I said, I could be changing to a different mcpcb source if one of the other projects on here works out. My power budget tops out at 400W, so that doesnt leave much room for driving past 3A/led (360W) after driver efficiency is accounted for, so I dont need the absolute best, but the less themal resistance I have from the LED to the heatsink, the higher the thermal resistance that the heatsink can be for the same performance


I'm thinking it will be 4 chains of 9 LEDs making for 4 banks with drivers, it divides up into a nice pattern, the voltage is reasonable, and its not a crazy amount of drivers. I'm definitely not having any LEDs in parallel so balancing them that way wont be a problem. I'm having a PCB made and using 4 Nat Semi led constant current driver controller ICs to make 4, 3A dimmable buck converters, and a microcontroller for managing the 4 of them.

EDIt:
and patriot, if by perfect platform, you mean not $30+/LED like the SST-50's then you are correct :nana: It fits into a nice niche where the performance is high enough, heat dissipation is low enough per LED, and the cost is low enough, this is feasible.
 
Last edited:

Benson

Flashlight Enthusiast
Joined
Feb 15, 2009
Messages
1,145
Well... that'll be a bit of a BAMFlight, won't it?!
:bow::bow:
 

DM51

Flashaholic
Joined
Oct 31, 2006
Messages
13,338
Location
Borg cube #51
Interesting project - very ambitious! As you have said, heat buildup is likely to be your biggest headache, and you may have quite a problem with it in that particular host.
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
cad479a2.jpg

572fd3d2.jpg

Mmmm progress today, just got my cutter shipment in.

~42k potential lumens there :D (extras for R&D and other stuff, 36 slated for the spotlight.
Time for a comparison test between the cutter PCB and the KD MCPCB performance.

Then some more R&D type stuff before I start making lamp assemblies.
 

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
You might need a metal host, that Thor is going to melt unless you don't drive the XMLs to spec.
 

magicmanchris

Newly Enlightened
Joined
Mar 10, 2011
Messages
56
CKOD, your in for a real treat if you see this through to the end. I just fired up my KONG-12 for the first time tonight. This was the 1st KONG that was built using 12 XML's LEDs. I have had such a blast tonight turning this monster on. What satisfaction to go from total darkness to instant 12,000lumens with one push of a button. I can appreciate what your trying to do. I know that something 3 X's the power of my new KONG will be a true killer!!!
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
Thanks for all the positive responses, It'll certainly be a fun toy when its done. And Ma-Sha dont worry about the heat, its nothing more fins and more fans wont take care of. I dont think I'll have to go to water cooling, so it isnt that bad :D The ahorton aspherics only need to be ~8.7 mm from the dome of the LED, and the lens isnt that tall, so that leaves plenty of space in that bell-mouth for heatsinking. I'm hoping I can fit my drivers, DC-DC converters, their heatsinking, and all the other control circuitry in the battery compartment.

EDIT:
new developments in R&D

Did a comparison of the Cutter XM-L boards and KD XM-L boards. Mounted a U2 2S XM-L to one of each PCB (reflowing onto pre-tinned pads with RMA flux)

Screwed the pcb to a heatsink I pulled from an old mobo for the mosfets, using thermal compound(one at a time of course). Ran each LED at 3.73A for 10 minutes to reach a somewhat steady equilibrium. I recorded the Vf at the power supply and light output on a lux-meter 6.5" away, neither the clamp for the heatsink nor the lux-meter moved between tests.

Even soldering the wire leads to the PCBs with them mounted to the heatsink, I noticed the KD board was harder to solder. I also noticed the heatsink became hotter much faster with the KD compared to the cutter PCB (hotter faster is good!)

Results:
Code:
                KD     Cutter
Initial lux    11.65k   9.17k
T+10 Lux       10.31k   7.47k
Lux drop       11.5%    18.5%
Initial Vf*    3.78v    3.73v
T+10 Vf*       3.70v    3.63
Vf drop        .08v     .1v
*The Vf was at the power supply, not at the LED, so there is ~6ft of wire that is probably a touch small for 4A between the LED and the voltage measurement.
I think the results pretty much speak for themselves, the KD boards are the better of the two
 
Last edited:

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
And more goodies for R&D.

c15a917b.jpg


Waterclear epoxy, meant for encapsulating LEDs. like whats usually used for 5mm leds etc... probably cant take the heat of being right next to the die on the XM-L, both are rated at 150C, but I suspect neither likes being at 150. But good thing the epoxy isnt going to be right against the die ;)

1a9b9300.jpg

test-rig, KD MCPCB arctic silver-ed down, used a 2-ton press to hold the pcb down firmly while it cured, hence the ooze around the perimeter. Some low-profile nuts to make a dam, epoxied to the mcpcb, and a threaded ring epoxied to the aluminum to screw on the lens tube assembly. (I find the focal length by dialing in the lens, locking it in with the jamnut, and then gingerly ripping the whole thing off the anodized heatsink, so I can measure the distance from LED to HS face :D )

Some UCL lenses off to the right also, to go on top of the nuts, after I backfill it with epoxy.

Hopefully I can then maintain around the same refractive index from the die, out to the AR-coated surface of the UCL lens. Negating the effects of the dome. Ive done some trials, filling the dam with rubbing alcohol, and got a 5% increase in lux, narrower beam, though, a liquid lens isnt exactly ideal.

Once I get a focal length for the epoxy/UCL lens combo, I can start finalizing dimensions of other components for this.
 
Last edited:

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
Great project.
Karl Dose GmbH makes ship's searchlight with multileds with tir-optics.
It will give you lumen, but won't give you the throw Databank has.

Neat stuff! yeah the TIR stuff isnt usually that tight, I like the combined central lens and imaging quality reflector that the TIR lenses have (I have a 1.25D mag with a SST-90 and TIR lens) gives a nice square projection of the die, and a bit of a hot spot, but its quite floody. I built a SST-90 mag with a DX aspheric for my friend, and the die projection is ~1/4 the size of the projection from the TIR in my mag. Though mine is great for around the house stuff.

I think its more of a market sort of thing that they dont make tight, 1 degree, 2 degree etc FWHM TIR lenses since a 2-4 degree beam is of rather limited use, and arrayed LED spotlights arent that common at the moment. I'd be neat to see a manufacturer make a light with a few SST-90 or -50s and a large-ish (40-60mm) TIR lens have a nice and tight beam.
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
And no luck with boosting lux output via virtual de-doming on its own today. I tried encapsulating 2 different LEDs, the first one, I was using the obvious source of heat to speed the cure of the epoxy (running the LED itself...) It cured up nicely, however the silicone dome pulled back away from the epoxy after the curing. I still tested it of course, and didnt see any increase in lux output, though the beam was narrower. (Which in turn means less lumens out, boo...)

2nd attempt, wasnt perfect, some bubbles in the epoxy, and I cured the whole assembly with a heat gun for a nice even cure, and the virtual de-doming was a success, looking down though the window, I cant see the dome at all, index of refraction is matched well enough that I cant see the dome. Lux was a wash again, slightly less then before encapsulation, same narrowed beam. I'm unsure if removing the bubbles would net a gain in lux that would make this worth doing, though the assembly process I was looking at would be easier to do, and put any bubbles that appear in a more desireable spot. (against the MCPCB rather then against the window)

I'll give it a 3rd shot, with a different method, to avoid bubbles, and hopefully less scattering will make for an appreciable improvement in CP output. The U2 bins are already getting me 30k CP peak, and ~23k CP average(ish).

Not all is lost though, I'll let the 2nd attempt cure completely overnight and then do some tests/thermal cycling on it, and other longevity tests for good measure.
 

netprince

Enlightened
Joined
Sep 7, 2008
Messages
547
Just found your thread. Neat project, I would like to build something like this some day. Subscribed.
 

bshanahan14rulz

Flashlight Enthusiast
Joined
Jan 29, 2009
Messages
2,819
Location
Tennessee
is the dome made of one material, or is it a few layers of differing materials? I was thinking, why not carefully file down the dome and then use the epoxy you have to "wet" the sanded surface? You could do this with #1 or #2 after you have completed tests, to keep from using another new LED.

Cool project, can't wait to watch it mature =)
 
Top