The L2 has plenty of room to open up the driver bay for a larger driver. Just saying.
The DQG 26650 boost driver works, albeit not at a high power level. If you’re just wanting the creamy white goodness of the MT-G2 then it works fine.
Edit: The DQG is 28mm in diameter, about the max that could fit in the L2 with a modified driver bay. The battery tube would press against the ground ring and hold the driver down. Should work though…
Most of George’s drivers require a minimum of 2 cells including the boost flex and hyperboost. There are boost drivers available but none that I know of that can push an mtg2 beyond 2-2.5A from a single cell. I’ve looked up some boost ic’s that might do the job in a new driver but I’m more of a dumb grunt as driver design goes. I look up parts and stand on my head trying to figure out the specs. There are only a few members that know enough and are willing to use their time to develope new drivers. It seems to take awhile but if you look back over the past 3 years there has been quite a number of new drivers produced by and for the community. Part of the problem is that a boost driver for a 6V led requires about twice the current from the cell as a linear driver for a 3V led which causes 4X the loss of power from internal cell resistance and this loss only gets worse as the cell voltage drops and the driver pulls even more current from it to maintain the same led current. I’m sure it will happen eventually though since so many want it to happen despite the drawbacks. For now it’s low current from one cell or low capacity from two smaller cells that fit in a single cell host. Personally, I like the Small Sun ZYT08 with a series mod as a compromise in size vs capacity(I’m not fond of baton styles).That, or follow Hoop’s thread and get in on a Cu/SS C8 host for which he might be considering a two cell tube. That should be a host built to handle some power. Old lumens has been in contact with Efest and poling interest for a 26350 cell but so far that’s been a non starter.
If your bound and determined to go the single cell boost driver route then look for the DQG boost driver, I believe it’s available again but can’t recall where I saw it. It’s a bigger driver that won’t fit lights that take a 17mm driver.
How much current DQG 26650 driver can provide to 6v MT-G2? 2Amps only?
I think 2-2.5A, somewhere around there. It draws 6A from the cell to do it.
Usually driver calls so cause it regulates output current. Input current may differs with different Vf.
If it so, using 6v mt-g2 with dgq driver senselessly.
There is no such problems at all.
Meteor 4P cells~~3S4P leds or zebralight xhp35 1 cell~~>12V led have more efficincy drivers than most buck ones.
Driver have to pull little more current from cell if you want to have max modes available even with almost empty cells.
2x26350 probably will have 2 times less capacity than one 26650.
The biggest problem of boost drivers is dimentions, there is defenetly no way to make powerfull driver for stock HD-2010 pill for example.
“There is no such problems at all.”
Pure bs, driver efficiency does not include losses in the host or cells due to resistance which rise with the square of the current.
“Meteor 4P cells~~3S4P leds or zebralight xhp35 1 cell~~>12V led have more efficincy drivers than most buck ones.”
Neither of these is available on the open market nor is there anything close to them available, try again.
“Driver have to pull little more current from cell if you want to have max modes available even with almost empty cells.”
Assuming 90% efficiency which is generous, to power one MT-G2 at 8A will take over 15A from a full cell rising above 20A by the time the cell is empty which will declare much sooner given the voltage sag. 5% is a little, more than 30% is a lot.
“2×26350 probably will have 2 times less capacity than one 26650.”
Agreed, which is one of the reasons a single cell host is a poor idea(note that I say poor, not impossible, people do this all the time with 2 X 18350’s and suffer the loss of capacity)
“The biggest problem of boost drivers is dimentions, there is defenetly no way to make powerfull driver for stock HD-2010 pill for example.”
You’re beginning to get the idea. Most 18560 hosts and many 26650 hosts are made for 17mm drivers without much headroom for a large enough inductor or the beefed up input and output capacitors, sense resistor(s), and diodes this would require, not to mention the actual boost ic and other pull down resistors, voltage dividers, OTC, and what have you.
To say the problems don’t exist is absurd. Overcoming them will take more than blind disregard. If you actually want to help instead of just trolling, try looking around the web for a reasonably priced, readily available boost ic that doesn’t require an acre of board space for associated parts and heat sinking. If you’re capable of designing one then quit wasting your time and get to it.
Meteor driver was an open-source project:
http://forum.fonarevka.ru/showthread.php?t=17352
The only closed thing is software, opened one is little different.
There is no problem in making some of them, but I live in city and country where it is hard to get most of components.
MT-G2 at 8Amps: 7V drop, 56W, 3570Lm, less than 64 Lm/W. Is that your target?
Ok, lets imagine that I have trusted that you really need 3600 lumens. Quad 219C can give them:
4x3A, 3.25V drop, 39W, 3600 lumens, 92 Lm/W. (parallel)
3A, 4x3.25V drop, 39W, 3600 lumens, 92 Lm/W. (serial)
This example is good because low voltage drop gives us possibility using this mode with low voltage single cell.
Parallel connection with buck driver (90% eff.) - 13.3Amp from single cell.
Serial connection with boost driver (85% eff.) - 14.1Amp from single cell.
Is this current to big for IMR26650? I don`t think so.
Is each cell in 2x26650 config discharging with 7Amp will have 30% more capacity than one discharging with 14Amp? If you are using ultrafire cells, may be. Basen 26650 4500mAh will have almost no difference.
It is funny to hear about internal cell resistance problems, 5-8 years ago it was normal to have power tool with 20Amp motor and 13R cells inside, and even this config could work for a long time and be in good condition after hundreds of fast charge-discharge cycles.
Now we have much better cells, with much higher capacity, much lower internal resistance, much higher listed discharge rate, and we are talking about bigger 26650 size cells, is it good to dischage it with 15Amp or not. Yes this is not good, you will have about 5% less capacity than rated, but it doesn`t matter if lots of members use linear drivers, improper optics, close its outer edge or even cut them down, leds in 3x rated power and etc. what gives much more lumens lost.
Using boost driver in single 18* cell flashlight is bad idea cause you will have 10W driver for $30+.
Using boost driver in 26650 host is more possible, I think pills can be changed to those which can accept 25mm driver.
Unfortunately, I don’t read Cyrillic and the meteor driver won’t fit into this application. Arguing that a quad is better misses the point. I’ve already conceded that those who want this driver don’t really care about sag losses so save your breath(or fingers). A full translation including schematics and links to whatever software is available would be usefull. Making it fit is a problem to be overcome when we know what to make. Lumens/watt and power handling increase with each generation of LEDs so a discussion of diminishing returns ignores future gains whether from MT-G2 or XHP and is a moot point for many enthusiasts in any case. Build it, make it programmable, make it push as much current as the led can tolerate(you can always use a lower mode), those are the general design parameters around here.
You shouldn`t understand russian (text there is almost useless), but there is enough information to make driver by your own.
Quad isn`t better in all situations, example just shows that there is no difference in cell load with buck and boost drivers (and that sometimes overcurrenting led is possible, but it is better to avoid 8A on MT-G2).
Schematics doesn`t require translating, don`t think I can share software description.
Making it fit is biggest problem that must be solved before any other things cause every mm driver difference means different driver pcb design.
Lumens/watt increase with few %/year, if you are going to use led at 3x rated current, led will eat about 4x rated power and give you about 2x rated output (half lumens/watt).
It is possible to make lots of useless things. I prefer to give an answer to myself for question “why should I go this way, not other one?”. In russia it is called “critical thinking” and is not acceptable by 86% of citizens.
Maybe I like market economy too much, but IMO the only overall factor in lights customisations is the spent money. You can spend lots of them to reduce lost in one factor from 5% to 2, but it looks not good idea if you are loosing 30 because of other factor if this can be fixed much cheaper.
Maybe one day I will move to bigger city and will be able to order factory quality pcbs and buy any components for drivers, but now making boost drivers is out of my possibilities.
Was it efficient to paint the Mona Lisa? But how much impact has it had on the world? Sometimes efficiency is not even part of equation. Why do men climb to the peak of Mt. Everest? Because they can! Is it efficient? Not even a little bit! It costs very much money and some die along the way.
I use the big fat MT-G2 emitter because I LIKE IT! I use it at 12A, even 16A, because I CAN! I use 3 of them in a single light, making 10,000 lumens, because it’s impressive, I could give a rat’s arse for it’s efficiency.
How many people drive vehicles that are not the least bit efficient for their daily driving needs? EVERY DAY, people drive pick-up trucks with nothing to haul, getting 10mpg just because they want a truck. Perhaps if you’re the development team and efficiency is top goal, then you need to think that way. Otherwise, water is cheaper than Vodka and much better for a person, that doesn’t mean Vodka won’t sell…
3xMT-G2 at 12 Amp: 12200 Lm/274W=44.6Lm/W, 30USD+shipping
50xXP-E2 from Mitko (1A each): 15000Lm/165W=90.9Lm/W, 20USD+shipping. Ok, it is not useful to use so many leds, pcbs, optics and etc.
20xXP-G3 that I was offered (2A each): 14900Lm/123W=121.1Lm/W, 30USD+shipping. 7 triples or 5 quads require 60mm space.
I can understand if you are making thrower and run led at 2x-3x current. But if you are making flooder, several small leds are always better than one big (more configurations of voltages and drivers, more efficiency, more different optics).
DB Custom, if efficiency is not main, why blf users are spending so much attention to parts that exert small influence? AR lens, thick wires, copper parts, even internal cell resistance lost. You are buying all best you can find to use them together with 50-60Lm/W LED? I don`t want to say that you are making smth wrong, my post just saying that it is funny to argue about cell resistance or buy 3400 cells in change of 3100 and use them in configuration with 60Lm/W LED.
It’s the challenge to reach the top of the mountain. Every possible gain, regardless of cost, to get the absolute most out of a light. Double AR coated lens, best cells, best wires, every possible way to attain the pinnacle of performance (meaning output)
I was about to say the 12,200 lumens is not feasible, then realized I got 13,500+ from 3 of the 9V MT-G2 in a Trustfire TR-J20 using 3 Trustfire 32650 cells and all the tweaks I could manage.
It’s again, not about the efficiency or even the cost, it’s about getting there at all. And I like getting there with a creamy white tint that I’ve only seen in the MT-G2 series.
It’s not theoretical, it’s not about white papers, it’s about a light in the hand making fantastic beam profiles and output. It’s about USING the light! Illuminating a football field with a hand held flashlight. It’s about schoolgirl giggles from mere achievement. It’s about DOING!
Edit: Are you an electrical engineer? I am NOT! I grew up in a lumber yard, toting building materials like a good mule. I’m a photographer, I got started with flashlights looking for an assist in Macro photography and the search was elusive, so I started “fixing” a bad light, modifying a decent light, and then just flat out making any light produce more output by large margins. So for me, it became about learning how to do it, it became more about the build than anything else. I have physical issues, like many here do. I find the build to be stress relieving, and the results of the build to be gratifying. So I do it for medical reasons, it’s like a stress relieving drug that only constipates my billfold. 
Now might be a good time to explain the difference between efficient and effective. All the little tweaks that are done are to make the LED more effective, not more efficient. In fact, even the best LEDs are terribly inefficient. Most of the power is converted to heat instead of light. But, we call LEDs efficient, because we are comparing them to incandescent bulbs, which are much worse. So, it’s relative in some respects.
The word efficient actually relates to how much of a ‘resource’ is turned directly into ‘product’. By that standard, nuclear power plants are inefficient. Only a small portion of the uranium is turned into electrical power for our homes and businesses. There is a large amount of waste by-product left over which must be disposed. And the proof that it isn’t efficient is that the waste substance is still ‘radioactive’. If it were an efficient system, the waste would be a much smaller amount and wouldn’t have any radioactive energy left.
Now, around here, we’re more concerned with how effective the flashlight can be. That is not about the usage of the resource, but what you get out of it. So, it’s about getting the most out of a smaller resource, rather than how much of that resource is being translated. I don’t think anyone would dispute that nuclear power plants are very effective! Similarly, LEDs are very effective at making light. And some LEDs are more effective than others.
We have learned that AR coatings change the lens so that it lets out more light. We’ve also figured out that spring bypasses reduce the resistance in the circuit, allowing more power to the LED and thusly more light out the front of the flashlight. Adding these performance enhancements isn’t about being more efficient. But these things make the flashlight more effective. If a single MT-G2 can put out as much light as 20 or even 50 other LEDs, then the MT-G2 is more effective, even if the others are more efficient.
I understand that this description can seem confusing for some people. Obviously, if you make the flashlight more efficient, that can also result in it being more effective. ‘Efficient’ and ‘Effective’ are really two sides of the same coin. They affect each other. But, sometimes making something more efficient will make it less effective, and the other way around. LEDs are a perfect example of this. Just look at the Cree PCT. The less power you give them, the more efficient they become (lumens per watt), while becoming less effective at dispelling darkness.
It only has to make sense to one person to be worth doing for that person. And while it’s fine to make alternative suggestions it’s quite rude to insult the choice. There are any number of strange builds that are done that I would never wish to duplicate but I support the efforts wholeheartedly whether they make sense to me or not , are efficient or not, are cosmeticly pleasing or not, whatever. 3-MTG, 30-XPE, 20-XPG, one used AAA Minimag bulb, doesn’t matter, whatever floats your boat is ok around here. Part of enjoying a hobby is learning by doing whether the results are successful or not, and only one person decides what success means.
Right in the point
But more words you are typing, less truth I see.
Mona Lisa is not very useful, but Da Vinci has designed so many useful things, if he was alive and could recieve 1% cost from each machine he had engineered, Mona Lisa cost would be a drop in the ocean.
Americans like trucks, russians like trucks. Now look how many trucks are selling in Japan, South Korea, EU… All resources are limited, and those who live on small areas had understand this better that those who live on big ones.
12-13K lumens from 3xMT-G2…you could get them from 12 XM-L2/XP-L. With less tweaks.
I can`t believe in good tint after you have burned them at 12-16Amp. I don`t own spectrometer, but it would better if smb could make such tests.
Multi-leds usage also has one tweak, you can combine different tints in one light, almost this is used in xph-* when 4a, 4b, 4c, 4d are combined and called simply 4.
Using light is main profit, but how do you use 300W flashlight? It may require portable alternator 
DavidEF, once again.
You can use 3xMT-G2 at 12Amp to get 12K lumens.
Or 12xXP-L at 3Amp to get 12.6K.
First time it will eat 274W, second one - 121W.
After you have made first variant, you are saying about how much springs and wires matter. But if you will go second one, you will understand that they matter not so much 
So, you’d rather pay $86 USD for the same amount of light that you can get for$30 USD? Cells are rechargeable!
I offered xpls for 3usd each, nobody interested.
9v mtg2 cost much more than 10usd.
Modding is a hobby, a form of self expression, as well as a way to accomplish a goal. Don’t be so stuck on the goal that you lose sight of that. The more you try and squeeze someone into a box the more likely you are to break the box. I’m not saying you’re wrong from a purely engineering standpoint but in every other way you’ve missed the boat. Whether it’s from the artistic appeal of a well machined but completely hidden copper sink to a purely personal choice in number/size/type/current of LEDs, there’s a lack of appreciation for individual taste in all of your arguments. That and the presumption that any particular build needs to be justified by some cost/benefit analysis where you and you alone determine the parameters. Loosen up, take a laxative or something, you’re gripping to tight.
Do you have any idea how many people have gone WAY out of their way and paid money to see the famous “Mona Lisa”? I know what you’re saying, I myself think it’s ugly. But nonetheless, it’s one of the most famous paintings of all time!
At any rate, I have seen results from engineers and am not convinced they alone have the answers. Look up “Aggie Jokes”.