Didn't Dale (DBCustom) build a beast ("Ham'r") with multiple FETs in joint circuitry for way more than 20,000lm using a whole bunch of LH351Ds below some carclos?
Anyway, 4x18A....geeez....that's 72A which equals in way over 250W! What host is able to sustain and dissipate so much heat? Please don't come with some active cooling by ventilation. Rotating, moving parts in a flashlight... :FACEPALM: It will probably take 4x Samsung 30T or 4x Molicel P42A in 4P configuration to get this amount of power without collapsing springs or batteries going nuclear. :-D
I did not mean to joke at all. The light you have linked to is for sure not consuming 4x 18A. It's just a light named "4x18A" but it's one and only XHP70.2 cannot drain more than approx. 18-22A from all 4 batteries. Anything beyond would burn this LED.
I should have read more carefully. Sorry for my misunderstanding. Reading “4x 18A” I was thinking, someone wanted to put 4 (four!) SBT90.2 into one single host.
Convoy M3 at 15-18 amps would be a nice and smaller host. Very heavy light to control some of that wattage. Even the M21C-U has good potential. I’m more into single cell lights.
Second that, PM1 + Z1 should have good throw in zoomed in mode while still having good output in flood mode, and it’s rated at 5A and can easily take 6A from the SST40 linear drivers (or more from the ‘ramping’ one), after all Simon doesn’t like to overdrive emitters, it helps keeping disputes low. I know that zooming flashlights are jack of all trades, master of none, but sometimes the tradeoff is rather useable. Edit: the Boost (U) 2 mm2 version have higher rated limits, 6.6A, 8A pulsed, 10A surge. The Boost 1 mm2 version have rated limits 3.3A/4A/5A but in practical terms it should work with the 5A driver even if out of warranty, but don’t expect many lumens, peak is at about 6A but output is only 5% more than at 5A, it’s useless risk.