The Courui has no tail switch and many other e switch lights lack that as well to great nuisance of some
If the tail switch does not make it to the GT we all have to thank Tom for Narsil and ultra low parasitic drain and good lock out/protection hence diminishing the need for a tail switch 
Djozz, well over a million CD for a XHP35HI makes it both impressive and respectable right?
A BLF special “Giga” thrower.
Price under $100 (still/hopefully)
XHP 35 HI with long tube for optical and practical balance
Abel to use either 4 or 8 cells
Approx. 1,5MCD & 2.500 Lumen
8) 8) 8) 8) :sunglasses:
I would be laughing my ass off. Seriousely.
This would be just fantastic!
Let’s get on with it!!!
Grtz
Nicl
Just to clarify things a bit, the candela unit we use to measure throw is simply a way to standardize the measurement. Just because a light has 1,000 kcd doesn't mean you are gonna use it to try to illuminate objects 2,000 meters (1.2 miles) away. 1.2 miles is moonlight level which is a 1/4 lux full moon level, which is mostly useless.
If you want 2 lux of light on a target, that 1,000 kcd light now has a range of 707 meters. And if you want a somewhat effective amount of light, you may need 4 lux for a range of 500 meters, and 4 lux is the minimum amount of light needed to navigate a room (determined by MythBusters).
So as you can see, the 1 Mcd light's range can be a lot more practical in use than you might think, specially if you want 10-20 lux on a target.
Only on BLF, when the gigathrower is produced 
The throw of 1Mcd-plus from a 120mm reflector is a simple extrapolation of what already exists. There is a margin in predictions but I see no reason why the extrapolation would be so radically wrong that 1Mcd is not reached when a XPL-HI is used. The throw with XHP35 will be close, the reflector quality will be decisive if 1 Mcd is met.
The only places I have ever seen an LED flashlight get over 1Mcd is with an overdriven XP-G2 and an aspheric lens. So here is what I think:
I corrected my estimate above for the XHP35, I’m not sure it gets to 1Mcd with a 120mm average reflector. The XP-L Hi should get there though.
The TN42 has a 90mm reflector.
Think about the difference between a 60 mm reflector and a 90 mm reflector - that's 50% increase, 30 mm more. 90 to 120? 33% more. The 60 to 90 change could be in the range of 50% bump in kcd, for roughly a MaxToch 2X to a TN42 or BOSS1.
There's no doubt kcd scales up as reflector width scales up, and if it scales up consistently from 12 mm to 90 mm, proven many times at many widths in that range, why wouldn't it scale up from 90 to 120?
Enderman, that LED emittance chart shows the brightness of the output at different angles. It doesn’t really count up the total lumens collected, such as in a reflector. As you can guess, the deeper the reflector, the more light you’re pushing into the spot. And, of course, as you said, a lens or even a recoil reflector design will certainly have the potential to collect more light. But, if you can get a reflector that goes maybe to the 50 degree emittance angle, you’re still collecting plenty of light. At that angle, you get 70% emittance, plus all the light from all the other emittance angles all the way down. I certainly wouldn’t want to consider a shallow reflector for this light. Spill is nice for a general purpose light, but this is a huge, specialized thrower. It shouldn’t have a lot of spill. I’d really like for someone to do the calculations, as djozz suggests are possible. That way, we’ll know what to aim for. If somebody would post the formula, I’d play with it myself.
Dang, just tested it in my L6 with NarsilTriple on a TA 30MM LDO driver, and it takes 20 seconds to lose power enough for the mode to be memorized. Anything less than a 20 sec wait acts as next mode memory. Hhmm - now have to check if it's LDO specific, or acting like this since brownout detection was enabled, or the form something else.
For tail power switch mode control to work in Narsil, you must be in a mode set (non-ramping) and it always has mode memory, unlike the e-switch which defaults to no mode memory.
That’s true, at a larger angle the radius is bigger, so it’s still a lot of light.
The main thing is that the depth of the reflector needs to be increased to take more of the spill and direct it into throw.
I still don’t think either XHP35 HI or XP-L HI will be able to reach 1M lux. Maybe an XP-G2 would be able to do it?
In a few days I will test XP-L HI against a dedomed XP-G2 side by side, but I doubt the XP-L will be able to reach the same intensity…
Increasing reflector diameter =/= increasing throw.
That’s not how it works at all.
That’s like saying putting larger wheels on a car makes it go faster….
The focal length of the reflector, diameter of reflector, depth of reflector, size of LED die, and precision/smoothness/accuracy of the reflector parabola are all important together.
Ok - that conflicts with pretty much everything we do here, understand, and proved thousands of times. Those other factors are important, but nothing is as important as the reflector I.D. If all that was true, we'd be building pocket throwers and the manufacturers would be making them - they don't exist, so....
If "Increasing reflector diameter =/= increasing throw" was true, that would pretty much put an end to this project.
Personally I like the tailswitch to only turn the light on/off in a dual switch light and control modes with the e-switch. This is the ideal setup IMHO. Although it is nice to have the option to change modes in the firmware I would want my light setup to only turn on/off with the tailswitch.
Although if you can tell us how you are managing to keep the MCU alive for 20 seconds to go to the next mode that would make clicky lights SO much easier. Even with a 100uf C2 cap we are only getting a few seconds.
If you had an LED that was 0.2mm side length and produced 1000lm it would throw the same using a 10mm reflector as an XP-L at 1000lm using a 100mm reflector of the same proportions.
That’s because the size of the LED to the size of the reflector (assuming same diameter and focal length) is the same ratio in both cases.
Unfortunately we do not have LEDs that small with that much output.
Here is a X diameter reflector:
And here is another reflector also of diameter X:
Can you see which one throws farther?
The renderings showing a size comparison to the TN42 make the GT look 50% larger diameter, when in fact it will only be 25%. The TN42 is very impressive, proven, and while yes I’m keenly interested in where the GT can go, I am not so sure the performance level will be as drastically above the TN42 as people are hoping for. Of course, provenance is everything.
The TN42 has a 4” bezel (100mm) It’s appearance is awesome to behold when gazing into the business end. As such, the GT should give that jaw dropping appearance in spades.
I vote for a single tube, for whatever that’s worth. 8 cell, 2 carrier configuration within the single tube.
I would also like to see the fins a little more pronounced, thinner, more of em, but not to an extreme. Thinner should relate to faster dispersion. They don’t need to be on the head where the reflector is, but having fins properly done on the pill section where it matters will be crucial to the light having the long run time necessary to be epic without step downs from Turbo output.
The renderings do indeed look awesome. A massive big brother look to the Convoy L6 which I like a lot.
Enderman, that is a nice idea but for the fact that an LED requires a parabola to have efficient output. Many years of research and development has proven the second design to be more effective with a flat surface LED emitter than the shallow bowl that works with an element style bulb. The difference is in output transmissive angle, the shallow bowl style would give a horrendous beam with an LED source.
I once researched the design for LED’s and one of the main theories was of a square parabolic design, square meaning equal dimensions diameter to height. There have been a lot of variations on that, to some degree, but it’s been the most successful to stay close to square such that a 120mm reflector should have a fairly close to same depth of 120mm. The TN42 “reflects” this, with a very deep wide reflector that shines a proven mile. (The TN42 is actually something like 90-92mm reflector diameter and in the 75-80mm depth area, this is as measured without disassembling)
Thinner fins also means less airflow, there is no fan here so convection has to do all the work.
Maybe someone can look for a peer reviewed research paper on fin density to find the optimal thickness for heat dissipation?
Theories are fine, real world provenance rules. I give you, as evidence, the Olight SR90 Intimidator.
Please note that I DID say “not to an extreme”. Thick bulky fins do not induce induction in my own experience when pushing an LED to it’s limits.
Ok? It’s not a theory that if you leave less space between fins that air has more difficulty moving…this is elementary school knowledge.
I’m just saying that you can find an optimal point where you have more surface area but still good airflow and that is the point with the most heat dissipation.
Cramming as many fins as possible into one spot doesn’t automatically mean it is better.
I think it’s pretty obvious that a solid block of metal (aka infinite fins stacked together) has worse heat dissipation than with fins with good space in between.
btw, the SR90 is only 30W of heat.
I’ve not been too active lately, and we have… a new BLF project!
Nice, put me down for one (with zero spill please 
)
Enderman, this thread discusses the relationship between throw and reflector size. Understanding the concepts in that thread really make it much simpler to think about flashlights and throw.