Update:
I broke my lathe machining the stubbornness out of some heatsinks for Old-Lumens. The repair parts just landed, and I should be up and running tomorrow if all goes well.
The driver pictured above is now listed on the cnqualitygoods.com website.
Please stay tuned for more updates and pics to follow.
The 3-up board has been soldered to the driver:
A thin film of Arctic Silver adhesive was placed under the board, then the optic placed on the board, then the whole assembly put under light pressure. The AS should hold the optic in place, obviating the need for CA glue.
Next will be affixing the driver to the light engine.
Since this thing is going to pull 4-5 amps at 12.6v, I wanted to upgrade the wiring I had on-hand. I used some silver plated copper Teflon-coated wire:
Here's the upgrade:
Some heatsink stuff from another light was placed in the driver pocket:
The modified wiring was then soldered:
Since the original pill screwed into the head, I needed a way to provide positive pressure to the bottom of the new light engine. The original pill was used:
I cut a ring from the heatsink, then notched it for wire clearance. I may want to upgrade some day. :)
Here's the battery plate soldered on:
Buttoned up, stock and modded:
Before, without cells:
After, no cells:
Testing.......IT'S ALIVE !!!!!
Next come the night-ending beamshots.
When turning copper alloys try increasing the back rake angle of the tool and if turning slow, speed up. Most copper alloys work harden when turned slow. Sharp HSS tool with a 20 degrees of back and side rake with fast light cuts work well for me. I grind my own bits for copper and aluminum. Most of the carbide ones have angles for steel.
$00.02
Larry
I didn't like this light, solely because there's no change in diameter at the head (iow, looks).
Now you have gone and changed my mind.
I have actually started keeping a prioritized list. Time to figure out where this fits :)
Thanks for the tip. As a novice machinist, I appreciate all info.
Here's a quick iPhone pic. On the right is the stock 700 lumen single XM-L CQG light at 2.8A, on the left the modded 3 x U2 XM-L at ~4A (turbo) estimated at ~3000 OTF.
Even my wife said "wow" at the new torch.
If you want to go with "old school" info check this out.
http://dale.chatham.org/Interests/MetalWorking/HowToGrindLatheTools.pdf
wow is right. that's my kind of beam right there. or even more flood.
Nice work sir! Let me know if you ever start selling these... :) The aesthetics of your light are, well lightyears ahead of the DRY. :bigsmile: To say nothing of the build quality.
Frontpage and Sticky for you.
Wow, my first FP/S !!! Thanks !
I may sell this one, as I ordered 2 more hosts. :) I wish they were a bit more budget...the host is $45 (edit:host only is now available for $35+s/h), the board and optic right at $40, driver $5 (all plus shipping), etc....I would need to get close to twice what the DRY is selling for in the GB just to break even, then pay myself $0.50/hr. 
Thanks again, LarryB. As I said, I appreciate all info. I'm currently poring over "How To Run A Lathe" 1949 edition. It's an awesome resource that hearkens back to a simpler, more honest time.
Well definitely congrats on a job well done. 
By the way, you are on the frontpage, just a little ways down because this post was started a while back, and it sorts by creation date.
Very nice Chicago, it's amazing what you can do with a few "hand tools". Very nice beam out of those optics. Guess the manufacturers of this stuff use those optics for a reason...
Modded lights are usually pricey, if you want to make them look and work good, even with doing your own work. Every time I make a light, I think this one will be cheaper and every time it's more expensive.
Keep up the good work!
Thanks very much for the kind words. I had to take the dogs on an extra walk last night, so we could "test" the light.
Well, testing has revealed some interesting results.
The light gets warm quickly on TURBO, as is to be expected with 3 XM-Ls @ 4-5 amps.
What is surprising is that on HIGH (2.5A) the light gets only mildly warm. It can be run in HIGH comfortably until the cells are depleted, with very little thermal or voltage sag.
With the CW U2-binned XM-Ls, this light is significantly brighter than my WW DRY. HIGH is even slightly brighter than the DRY on TURBO. 
The original torch had one excellent feature - a 3-hour runtime on HIGH. This light does about 3.5 hours on MED, surpassing the original's runtime with the same rough lumen output.
Usually, cranking up the power fourfold leads to compromises in functionality, but this mod seems to have it all - the power, the runtime, the looks.....can you tell I really like it ?
This mod is awesome, thruly awesome.
I always think these threads are the most fun to read since you can see what sets the modders apart from us buyers :p
Great light, I thruly think that this kind of self made powerhouse light is worth much more then any bought xxx dollar flashlight
i really wish i had your skills!
Those are some beautifull dogs btw ;)
I had very little to do with the dogs, but thanks. 
My lathe skills are fairly rudimentary - I am just learning how to operate it, but appreciate the kind words.
There is little difference between a 'buyer' and 'modder,' IMO, apart a willingness to break things for no apparent reason. 
Please get yourself a soldering iron and join the club !!!
Bro, not surprising. The WW DRY in Turbo reads 1600 lumens OTF on my meter. Tail current is to be expected, 3 plus just below 4A range...ard there.
I have a triple U2 coming today. But it's not a DRY so can't compare.
I swapped one of these T6 from a triple XM-L to the Crelant 7G5 which originally had a U2, dropped about 16% in a ceiling bounce. 79 lux -> 66 lux. I don't think on paper the U2 and T6 diff is that great, only conclusion is that the budgetlights have somewhat lower T6 bins or even upper T5 bins.
"It's difficult to say", if you will. :)
I have often questioned the authenticity of binning when purchasing from sources other than those that guarantee it. (e.g. Cutter, LED-tech, LEDsupply, IlluminationSupply, etc...)
Since there are no markings, and few own the proper equipment to discern the difference, my suspicion is that bin fraud is rampant.
They need to do SOMETHING to rid themselves of the tonnes of T5s that are still languishing on the racks. A swap of a sticker, a new UPC, and suddenly they are magically in the T6 business. 
personally, I feel it would be naive to think they're NOT doing that.
even if you had the equipment to test properly you wouldn't have much luck making a claim.
next year we'll get another bin or two higher and be tempted to upgrade.
I've been weighing the cost difference of ~12 xmls - either u2 on copper or t6 (claimed) at DX for about 2/3 the cost. Your comments make me this it is worth going u2 - at the very least I won't be wondering if I got what I paid for...
So how does the beam from those optics compare to the DRY?
I don't suppose you're planning to take a pic in the same spot that you did a whole bunch of beamshots for us awhile back :)