I think the reflector can be made by plastic to save the weight. By doing this not only it can save the overall weight but it also can avoid front-heavy issue.
Now I know most people prefer aluminium reflector, but in fact plastic reflector can be very good too. TK75 uses plastic reflector and it is still one heck of a performer.
Can the tail cap be designed to fit comfortably between the lips for easy hands free operation while working?
Maybe they can offer the XHP70.2 as an alternative option later. With such a huge reflector it can still throw beam over 1km, while putting out 7k lumens.
But of course I would like the XHP-35 HI version come first!
Yeah i would love XPH70 and XPH35 version, and maybe 4x18650 version instead of 8x18650 or sell battery tube separately.
You mean a long-barreled (8*18650) version with an XHP70 and a snub-nosed (4*18650) version with an XHP35?
This one is already a long barrel, 4x2 18650. So, just need a short barrel for those who like weird looking flashlights. Oops, did I say that? :innocent:
One only needs to use 4 cells
Double the cells for more runtimes
The already big TN42 has less lumens but these are throwers so it is not all about lumens it is mostly how the lumens are used.
The Q8 is brighter it the TN42 lights up things much farther away.
And with its size it seems the GT will be a light able to actually run on max as long as the cells allow
The driver is kind-of a ‘2-channel’ driver. The main difference with our conventional 2-ch drivers is that in this case the two channels multiply, instead of add, with the expected impact on the mode-sets and ramping table definitions.
Blue trend on the strobe shot above is pin 5 of the tiny, wired to the EN/PWM pin of the LM3409 (so ‘digital strobing’ there, not analog). Removing/lowering C17 will get it much faster, at the cost of increased ripple current in the LED. The current design is optimized for good constant current control, at the cost of PWM speed, linearity in the very low modes and things like strobing. Simple mod though for anyone with different priorities.
I think you came to a very good compromise with the design. The driver looks great and has great specs. It is also nice to have a working buck driver with which to move buck drivers forward in the future.
Only change I would make to the driver is updating to the 1617 mcu but that is simply because I want to see some firmware developed for it, not because it needs it lol.
Yeah, the multiply-instead-of-add bit is why I was wondering. 
It sounds like the light handles the ramp like this:
Strobe then does one of these two: (?)
Does it make any difference which of those last two methods gets used?
With a cap acting as a lowpass filter, is it simply not possible to get a shorter rise time?
I think it’s a good design; I’m just trying to get a feel for how it works.
(the only time it would even matter is for making a XHP70 party strobe with ~0.3ms pulses, but I don’t think I really need to make one of those… )
Hear you on the 1617 TA, maybe next time! (Have not seen them in retail yet though.)
You have the gist of it TK. I tried both ways to strobe, but the 1st method was by mistake in the strobe levels definition. The 2nd method is the preferred way. Not a huge difference between the two, due to that output capacitance. But if lowering the output capacitance, the PWM/EN pin should react much faster than the analog dimming pin.
The LM3904 evaluation board does not have an output capacitor. The documentation for it shows a 3.5 us rise time when PWMing it. That facilitates neat 30 kHz or more PWM. The compromise is 300 mA ripple on the output.
I know the driver has probably been developed with the XHP35 especially in mind, but after reading Texas_Ace’s data on the XP-L2 V5 and how it’s putting out almost as much light as an XHP35 at 2.5A, any chance at all that a GT XP-L2 V5 version will be in the works in the future? I mean as far as throw goes, the die size of this LED makes it the most apt.
Wrong!
No, if people want different LEDs and drivers then they can mod their own light.
The LED and driver have already been chosen.
Also, not only does the XP-L2 still make a few hundred lumens less than the XHP35 HI, but it also has a dome on it so any advantage of a smaller die is pretty much lost.
It requires very high current to get 2200lm, and has lower efficiency than the XHP35 HI.
Not only that, but dedoming isn’t something that is feasible for a production light like this, so either it would need to be a HI led or it would have to keep the dome on.
oops, forgot the XPL2 wasn’t dedomed… so much for my ‘bright’ observations ha!
Well, if the XP-L HI made more lumens we would have gone with that 
Maybe in the near future we will get better HI LEDs.
XP-L2 HI maybe? 
It’s only a matter of time… fingers crossed! or perhaps one day soon LEDs will be emitting light straight outwards like a laser beam, just without the collimator. Optic fibre technology maybe?
Solution:
Step 1: Find and opaque, sealable container. May I suggest a half gallon milk jug?
Step 2: Find secret hiding spot for container.
Step 3: Add 1 dollar a day into container.
Step 4: October 17, 2017 and a razor blade possibly.
I’m pretty sure shipping and taxes are not included in the $111…?
Also, that’s in USD, in other countries you need to exchange the currency.
And on top of that, other countries often have COD or additional import tax.
Well we aim for the $111 to be with shipping