I like the International outdoor drivers, but sadly they don’t have one with a 6v output.
Multi-mode is ok but preferably just a low and a high, as PWM makes me motion sick. I was hoping for about a 3 amp high, i know the emitter can take more but I don’t want it to get hot so quick. Low should be anywhere from 50-500 mAh
I was thinking buck driver because I want to use a Nimh battery pack for the drain characteristics and the ease of charging.
Any and all suggestions are appreciated !
P.S. Heatsinking is not an issue when you have an entire machine shop at your disposal.
What input voltage? You'd need to start with a 6D host (if using D cells) to be able to use a buck driver. Using a 2D/3D and a bunch of AAs in series holders wouldn't have squat for runtime.
I’m using a 2D host mostly because a 6 cell (7.2v) 2/3A battery pack will fit perfectly inside the tube, and i just had it laying around.
Using 2/3A cells id only get about 1800 mAh but runtime isn’t really a concern. This is more of a ‘Sprinter’ light and i’m building it for the fun of it.
And by battery pack i mean a premade pack meant for RC cars. I have a handful of ICR 18650’s but i don’t really want to use 8.4v because of the energy wasted by the driver, and I’m not really sure of the discharge rating. They came from a laptop battery so I’m guessing 3-4 amps but they’re unprotected and i absolutely don’t want this thing to turn cruise misslie
All drivers waste energy. A Zener modded linear driver would be simple and relatively efficient and you can get one with the modes you want from RMM. Need a a good heat sink too. Are you going to use the stock reflector?
I suppose you could say that using some of the input to power the driver functions is 'waste'... but it's not nearly the same thing as what 7135s do with the excess voltage.
Some are better than others in the efficiency department but that's not the method by which a buck driver deals with the difference between Vin and Vout. Come on, you know this stuff.
I was getting the impression that the OP believed (and strangely, you seem to be saying the same thing) that buck drivers should be avoided because the greater the Vin/Vout mismatch, the greater the wasted energy. That's not how switch mode buck drivers do it. High voltage low current input gets converted to lower voltage higher current output. That's nothing at all similar to what a linear/CC driver does.
It sounded like you were saying that buck drivers are 100% efficient aside from control circuit usage which isn’t the case. Certainly you know that as well. Cheap buck drivers can have efficiencies as low as 50% in low modes and often barely hit 80% in high mode. A good one like a Bflex that gets 90% or better can also require more headroom than he has available after voltage sag.
The LD-2C might be an ok option as well in that low voltage kicks in at 5.5V which with voltage sag might be ~1V/cell. It has a 3A high/150mA low option. Efficiency may or may not be better than the Zener modded Qlite option, I couldn’t say, but that’s two options for you.
I might try the stock reflector, but if that doesn’t work I’ll probably turn my own if i have the time.
The main reason i was concerned about heat was because this emitter is going to produce a lot of heat, and i don’t want the driver to add a whole lot more to that. I know its not really all that much when compared to the emitter, but every little bit counts. Not to mention that this battery really doesn’t have a whole lot of extra energy to never see its way to the emitter.
I was never implying that bucks should be avoided, just in this particular situation efficiency is a necessity.
I didn’t think of the zener Qlite, but i think that might be the best option i had considered the LD-2C because of the lack of PWM. I despise PWM.
I had also thought about using a triple XM-L driver, but that would be a lot of current, and too high a voltage in.
I too hate PWM but you can’t see it from this driver even on video. For a heat sink, I don’t know who makes one for Mags but you should be able to hack something together out of copper. A good source for different diameter and thicknesses of copper slug Here That place has a big list of all kinds of copper stuff.
The Qlite is a pwm driver but at such a high frequency as to be unnoticeable. The LD-2C does not have pwm. Buck drivers convert some of the excess voltage to current but as I said above neither can possibly be 100% efficient and at 3A output both will start to dim before 6s nimhs are drained. Both seem to have the 2-mode option you seek which combined with the input voltage is why I mention them. OL has done numerous Maglite mods and now works with both custom sinks and a customized driver but other than having space for a lot of extra 7135’s it’s really a Zener modded Qlite(nanjg105C) in a new suit of clothes. MRsDNF has a thread running with different MT-G2 driver option but some of them are no longer available. Not mentioned yet is the knucklehead buck driver which combines the known firmware of the at tiny mcu with a buck driver. Choose one of those or if you find one you like better use it instead but in any please case share your results. Best of luck to you.
I meant that i was going to make my own heatsink. It will be aluminum and I might do a little testing with the thermal conductivity of a press fit copper rod down the center of the heatsink.
I have an entire machine shop, with three different lathes. In the future i might even machine my own hosts.
Works with 2-4 LiIon cells, 8.4-16.8v, outputs the same current with 2-3-4 cells (input current is lower with higher Vin). Will fit inside the D cell tube. There are others out there, but that's the only one I know of that has user selectable mode groups and will output enough voltage for a MTG2.
-----
The existence of some poorly designed buck drivers says no more about buck drivers in general than poorly designed linear drivers say about linear drivers in general. Even a fat piece of wire is not 100% efficient, anything you put electricity through is going to use up some of it. What's the efficiency of a typical clicky switch + PCB + spring? Check into that and it might help put the driver issue into a more realistic context.
If you do the math to calculate the waste heat from the driver at given power levels you'll see it's not really significant. It may be significant as far as the driver's lifespan/performance is concerned, but not to the overall package.
I understand that perfect efficiency is not physically possible, and that is not my point. I simply was asking for advice in choosing a driver that had decent efficiency.
This could be done with as small as a 2D or 2C mag if you use a linear driver such as Lflex or various import drivers and two 18650's or 26650's or 36650's. According to some charts the MT-G2 needs about 6.4 volts or so at 3 amps. (6.5V at 3.5A) In a 2 cell linear configuration, once the batteries drop to 3.2 volts a cell the light will start to dim (if on high) but the batteries are mostly dead at this point so it will serve as an appropriate low voltage warning. (if you are using a driver without one built in) On low output you wouldn't see the drop until the battery voltage was even less, but I think it would still be noticeable before the protection circuit of the battery kicked in. If you went with 18650's, the 4.35v cells should allow for more runtime than standard 4.35v cells, but there are new 26650's on the market that have over 5Ah of capacity. I'd go with a 2C mag if I were going to go linear with 26650's. Efficiency at 3.5 amp output (LFlex) would be about %77 with fresh 4.2v batteries, but this increases steadily until the battery voltage matches the output voltage, and at that point the efficiency is near 100% (direct drive).
With two lifepo4 cells, (3.2V nominal) like the 5Ah 36650's for instance, you could actually direct drive the MT-G2 and get around 3 amp output, and this would be near 100% efficiency. With a linear driver in this configuration, the highest output would probably not be regulated if it was set to 3A or greater, so it would essentially direct drive, but you would also gain low modes and battery voltage warnings. Lifepo4 batteries have a very flat discharge curve until their capacity is diminished, flatter than even nimh, and then the voltage tanks, so you wouldn't actually need low voltage warnings with two lifepo4's. The light would dim and shutoff. There are 26650 lifepo4's available as well if you wanted to go the 2C mag route instead of 2D with the 36650's.
With bigger mag's and a buck driver, battery possibilities also include "1/2 D" cells or sub C if you want to stick with nimh. Lifepo4 26650's or 36650's are an option in 3c configuration or greater as well, if you prefer lifepo4's. Efficiency with a good buck driver would be near 90%.
If you want a high end reflector, Phoenix Electroforms has a drop in replacement for mags, the PA11. The AQ coating is very nice and about 90% reflective.
For a good lens get one from flashlightlens.com. 52.1mm size. For plastic I like the AR coated lens, and for glass the UCL lens. You might want to go with plastic if you get a nice reflector because if the glass lens were to break it would damage your awesome reflector.
For the MT-G2 itself, get one of the Noctigon ones or similar mounted to a copper star.
Apparently the LD-2C won’t push the mtg but Here is the thread that has many of the mtg driver possibilities along with their limitations. If you haven’t ordered from China before be prepared for a 4-6 week lead time and products that don’t match specs. Some places can be faster or have upgraded shipping for a price.
