A few months ago I got deep into bin picking for battery packs. One of the packs I picked up was a 6S NIMH sub c pack for some sort of RC vehicle. I brought them home, separated them, and ran them through a refresh cycle on the Opus. They weren't in perfect shape but still plenty of capacity. After refreshing I tried three of them in a 3 up XM-L mag with a nanjg-092 direct drive driver and got 10 amps. I got greedy and tried four. The "poof" of loosing all three emitters in a microsecond convinced me I had to do something interesting and perhaps extreme with these cells.
My recent and hard won success with a 2C mag quad gave me the inspiration. Why not stuff four of these cells into a 3C mag and run four emitters off a direct drive driver? Because four sub Cs won't fit into a 3C mag! But that's why we call it modding. There's plenty of space of the tailcap ID is enlarged to the diameter of a sub C cell. A resistance modified low profile spring glued into the tailcap leaves a perfect amount of space.
Now that I got the power figured out, I wanted to make sure the switch doesn't become a choke point for electrical flow. I was expecting 15 amps or better so I was more keen than ever to route the power around the switch. Soldering an 18 gauge wire from battery positive to led positive and letting the driver do the switching is a good solution.
The heatsink was shaped out of a short chunk of 1.875" aluminum rod and wrapped with a strip of 0.012" copper sheet to make a tight press fit. The heatsink, mag head prep,and reflector prep is the same as with my 2C build.
The chosen emitters are XP-L V6 bin with 3C tint on 16mm noctigon. Because I'm using 4.8v nominal the emitters were wired in parallel. I took some pains to keep the emitter connections as balanced as possible and yet also as low resistance as possible.
The reflectors weren't playing nice inside the head so I glued them together with JB Weld. I'm still not sure how I managed it without ruining at least one reflector.
The driver was a BLF17DD flashed with STAR and off time memory. I chose 5 modes plus moon mode in ascending order. Richard of MTN electronics was kind enough to do the programming. I've only built 10 or so of these FET based drivers but I find it very satisfying for some reason. Programming, here I come!
Top off the reflectors with a UCL and we're done!
I have no ideas how many amps this light actually draws as my DMM quits at 20 amps. The head of the light is too hot to touch at around ninety seconds of run time on high. Hence the turbo timer.
In closing, here's a beam shot in a blizzard.
Thanks for looking.