Thought I'd throw in my two pesos here.
I believe JS's M6-R pack is a very well engineered and built rechargeable power solution for the M6. Its elegant in its design, the genius is the integration of the regulation. Not just stock output with rechargeability, but constant light during the run. My sincere compliments to him for his exceptional achievement. He suggested the idea of boring out an M4 body to run the MN21 and MN20 lamps on a different platform, with 2x 18650s. Great idea. I posted a want ad thread in Machining for having the bore done and haven't gotten a response yet. I wanted to pursue that idea, but it will have to wait.
In the meantime, while others are pondering the issue, I decided to do a little experimenting. My target was to approximate the M6's performance with the MN20 and MN21 with a rechargeable solution that wouldnt require any extensive fabrication or irreversible modification. As many know you cant just jam a bunch of R123s in the stock holder (although that allows the use of the M4 bulbs, havent tried that yet...) because of its tricky three-in-series two-in-parallel double stack of CR123s. Wind up with way too much voltage and the 3.0V rechargeable cells have insufficient capacity to directly substitute the primaries.
I looked up the Kiu thread and brainstormed a bit. I hated the idea of wasted space inside the generous battery chamber but I loved the idea of 18650s, a pair in series. Nice, but... then it hit me. A very simple solution. That worked a LOT better than I ever imagined. I took a Fivemega 3-in-parallel battery holder designed to fit 17670 cells (Pila 168 S) and stacked two 800mah R123s in each of the three spaces, creating a 7.4V nominal 2.4Ah battery pack. I applied electrical tape to the positive end leaving only the central protruding "nub" exposed to mate with the MN21 spring. Almost home. Just a bit too short. Since the Fivemega holders are designed to mate together in series, I used the 1/2" thick top brass component of a second FM holder attached to the negative end of the holder and redundantly used a bit of electrical tape to secure it in place to act as a spacer. This fills the leftover space with a nice low resistance connection and provides a nice mating surface for the stock LOTC switch. Checked the connections, measured the voltage at 8.2V and into the light it went. My theory was that the three-in-parallel of lower individual capacity rechargeable cells (800 vs. 1300mah) would at least match the two-in-parallel setup of stock primary CR123s. Plus the triple parallel would distribute the huge 4.9A load more evenly amongst the cells, reducing waste heat and cell stress. Factor in the lower internal resistance of the R123s and it looked promising on paper....
I had one real concern. Well, two. One was instaflashing the bulb, a VERY expensive mistake ($30/sec). The other was the load on the cells. 4.9A is right on the edge of the 2C upper safe limit for LiIon cells. I had no experience taxing cells that hard before, especially driving a >3A load in a light. I put it together, said a quiet prayer, and hit the switch....
BRIGHTER than stock. Shocking! WHITER too, more intense, greater sear factor. A "Holy Cow" moment! Then I waited to see if it dimmed quickly as I feared it might (within a minute runtime even). Nope. Within 30 seconds it settled in to a constant brightness, like running a Mag85 on quality NiMH cells.
This was certainly a sweet spot. I shut the light down at three minutes to check the cell temps and make sure all was well. It was. There was only a trace of warmth to the cell bodies, whereas the stock primaries got very warm to the touch in the same time frame. A minor Eureka moment. It seemed in real world testing that somehow even though the load was high the new 800mah R123 cells dealt with it rather easily. I then ran the light intermittently for about 15 minutes, observing a dimming similar to stock characteristics but occurring later on during its pack charge cycle.
I was very pleased. While by no means a permanent or care-free solution (that incurs a safety concern - careful use/handling implied), it exceeded my intended goals by surpassing stock M6 performance in both quality and quantity of light delivered over time. I measured 5.10 amps at startup, diminishing to 4.95 within 45 seconds and gradually declining to 4.85 at 5 minutes. I did not test beyond 5 minutes continuous. I figured this solution was best applied to intermittent use of the light, avoiding long (>10 min) runs. The current draws seem to indicate that I hit the target of reproducing the primaries performance of 4.8A @ 6.2V. Exceeded it even. Obviously I'm overdriving the MN21 a tad and dont know how much bulb longevity will be adversely affected, BUT the quality and whiteness of the light makes it worth the cost. It really is a beautiful beam. Only a flashaholic can relate properly.
So there it is. A simple (slightly edgy) no-surgery-required DIY rechargeable M6. Best of all, absolutely NO modification to the light whatsoever, utterly reversible and compatible with the stock holder's use with primary CR123s. Not bad for having some cake and eating it too.
My thanks and credits go to JS and AWR for the inspiration to attempt it and I defer to their superior knowledge on this topic. Project M6: Almost-Guilt-Free-Photons is a success for me!
:naughty: