You should probably post them up publicly because everyone will want to know!
I was thinking he needs to show the pill and adapter ring, offer the necessary parts for sale! 
I already PM’ed rjorge, asking for details, too :)!!
Pressure’s on, rjorge!
JUST KIDDING :)!!
Have to get myself a BLOODY lathe. Oh and learn how to use it too.
Anyone know what the high and turbo lumens are likely to be given the tailcap readings provided above?
Very stable output , looking nice.
Match’s Mini-Mag was also pretty small, but I’d much rather have rjorge’s Mini DQG.
500 Lumens and 1850 Lumens respectively based on djozz’s work. IIRC those are not OTF lumens though, I figure absolutely no better than 90% efficiency in the TIR. Oops. No. I’ll post the right numbers, but these were not the correct numbers.
Are those numbers correct? Isn't this a boost driver, so emitter current will be less than tailcap?
-Garry
Yeah crap, hold on. Wasn’t paying good attention - thanks.
OK, let’s do Turbo first: 3A input with a fresh KK battery would be 12W input. I figure we get 80% efficiency, I don’t remember if we’ve actually got efficiency figures for this driver. So 9.6W output based on 80. So that’s about 1.6A at 6v, or 1000 Lumens at the LED. OTF would be <900 assuming that the TIR is <90 efficient.
Next High: 700mAh input on a fresh KK should be around 2.8W. Based on the same assumptions this gives us 500mA at 5.4v, or around 350 Lumens. After getting through our assumed 90% TIR we have ~315 Lumens.
I’m using a lot of assumptions and guesswork there of course. When I say “fresh KK battery” I’m assuming 4v. And I’m making a big assumption at 80% efficiency for the driver - I suppose that it could really be anywhere from 70-90% efficient. It can’t be too terrible because people would complain about the heat if it was.
EDIT: To make one of my assumptions more clear: With a boost driver the tailcap current will vary a LOT based on the battery’s state of charge / voltage. Over the course of the discharge the tailcap readings could vary by up to 20-25% (they get higher as the battery becomes more depleted). So it’s important to know what the battery’s state of charge was when the tailcap readings were taken.
Optic is at 88% efficiency with an MT-G emitter, no claimed efficiency with the MT-G2. Other emitters are actually even lower, down to 83% or so.
....and my MT-G2 test was not a bare emitter test (had not build my integrating sphere yet at the time) but an out-of-the-reflector-without glass-lens test so the question here is what is the light loss of an open reflector compared to a TIR? 
Ah, I couldn’t remember for sure and was too lazy to read, sorry! Instead I erred on the side of stacking losses on top of losses ;-).
I think it’s safe to say that the TIR should be more efficient than the reflector by a fair amount. At least 5%, maybe a lot more depending on the reflector. In that case we can adjust my figures upwards to something more like >1050 Lumens and >365 Lumens.
I will post more pictures and dimensions when I get home from work! I will get LED currents too, both with the KK 26650 and a VTC5 18650.
Oh that’s a shame! It’s no better than factory measured by Upz. Image courtesy of Upz
I was hoping it would be better. Aside from the better tint there's no real point for me to do this then.
Somehow I missed all the Vf numbers posted in post #3 of this thread. Again, based on djozz’s previous results those Vf numbers indicate <1100 lumens based on a <6v Vf. Note that rjorge commented in the same post that Turbo seemed brighter than with the 3*XP-G2. I’m not really sure what to make of that based on the numbers you posted in that graph aoeu.
It (“seemed brighter”) may be because the MT-G2 would have more spill than the 3 x XP-G2, causing higher “apparent brightness”)?
I’m interested in the driver. What size is it?
It’s inherently difficult to compare brightness from memory. Especially after the light has been off for a few days for the build.
It’s still a great build but I wouldn’t want to go backwards in brightness especially when stepping up to an MTG2
correction: just reread the post: the MT-G2 test was done with reflector+non-ARcoated lens.
The boost driver puts out a fixed current, at whatever voltage is required to get that current. So the MTG2 with 1.5A @ ~6.1v will be much less power than the original XPG2's 1.5A @ 9.6v. Input power should also fall by around 1/3. But if the total light output only drops by 1/4 or 1/5 that could be an acceptable trade-off.
1st-gen driver is 28.6mm dia, unknown if the gen2 driver is different.