So, FWIW… I put together a red Convoy S2+ configured with a triple and a FET+1 driver. This was an experiment to find out if it would work, and to find out how the FET and +1 would interact with the emitters. This post is sort of an unassorted collection of notes.
I swapped the pill with an old S2, which made things fit easier. RMM’s aluminum spacer was just right, ish. It’s too long by precisely the thickness of the glass lens, so I wonder if it’s intended for builds with no lens.
The FET+1 is driving a triple XP-G2 S3 3C shining through a Carclo 10508 (medium-spot, frosted). The beam pattern is very nice (smooth, floody). However, the tint is green on the 7135-only modes (looks whiter on the FET modes). I reduced this effect on one mode (~100 lm) by letting the FET contribute about a quarter of the total lumens, which improved the tint compared to letting the 7135 chip do it all. It’s less efficient but looks better.
I haven’t checked maximum output yet… having a difficult time doing spring bypasses correctly. However, pre-bypass I got 1950 lumens on a 3.9V cell so I expect full output will easily be over 2000 lumens. (edit: with a 4.18V cell, it got 2250 lm at start)
On the other end of the scale, it has a nice moon mode which goes from 0.38 lm (at 4.2V) to 0.07 lm (at 3.5V), which is a lot more stable than a FET-based moon. I might try adding PFM to soft-regulate it though.
The levels I gave it (first pass, not fully calibrated yet) work out to: 0.26 lm, 8.8 lm, 98 lm, 382 lm, 913 lm, 1950 lm on a 3.9V cell.
The host is very nice, especially the switch and the color. Half-presses are really easy with this type of button. The bezel is a bit narrower than the optic, so it cuts off a bit of each of the three TIRs. I assume this causes a loss of lumens, but I don’t know how much. The beam pattern seems mostly unaffected.
So, I still have a few things to fix or improve, but it’s pretty nice so far. I figure I’ll adjust the spacer at some point and maybe give it Nichia 219s instead of XP-G2s. Also still have some springs to improve and firmware to tweak. I didn’t even put thermal paste in yet because I expect to be taking it apart soon… but so far so good.
Edit: With a fresh-charged 4.18V cell, it got 2250 lm at start. Modes were: 0.38 lm, 10.3 lm, 121 lm, 425 lm, 1055 lm, 2250 lm. The 7135 channel alone produces 152 lm.