6D Mag or Custom build?

Tekno_Cowboy

Flashlight Enthusiast
Joined
Apr 2, 2008
Messages
1,680
Location
Minnesota
I'm looking to setup my 6D mag with 3 columns of 17650's, 17500's, or 18650's, with the 3 series columns in parallel to each other. I have a pretty nice charging setup, so I want to press some aluminum stock into the body and triple-bore it to fit whatever solution I go with. The aluminum would be used to seperate the batteries and to act as a heatsink for the bulb/LED. I'm not sure what kind of bulb I want to use, but it must be bright as all get out. I'd want to replace the stock switch with more heatsink and a higher quality switch, possibly with multiple settings. I was also thinking, since I have access to a lathe, high quality aluminum,brass, and stainless, and a seasoned toolmakeer for cheap/free, that I might be better off just making my own light. I'd maybe be interested in a drop-in bulb/led upgrade that still uses the 6 D cells. Suggestions?
 
Malkoffdevices makes this(1 LED) and this(4 LEDs). They both work with the Maglite 6D and have adequate heatsinking. They also run on normal D batteries.

You don't need a huge amount of heatsinking to keep the LEDs cool.

If you want to do something yourself, you can use the Shark Driver, 4 current gen LEDs, a good heatsink(like this one), and the right reflectors/optics to make a quad LED mod.
 
I am firmly against using multiple light sources from the same light, unless it's for color purposes (ie. red, white, or green). I might just grab a brighter bulb for my 6D now, and devote my money to building my own custom light.

What's the brightest bulb/LED I can easily run off of either 9 18650 in 3 parallel groups or 12 18500's in 3 parallel groups? I'm thinking that I want to go with an LED, but I don't know what's really bright on the LED end.
 
You'll need a driver for the LED to run well. White LEDs normally operate at around 3.7v.

Also, you can only fit 2 18650s side by side in the tube(more length wise, so you can fit more the 2 total).

Any reason why you want to use Li-ion batteries?

The current gen LEDs are at least twice as efficient as the last gen LEDs, so for the same power, the current gen ones will have twice or more output then the last gen ones.

Current gen LEDs:

Cree XR-E, P4-R2 bin
Seoul P4, U bin
Seoul P7
Luxeon Rebel, 80 and 100
Luxeon K2 TFFC, 200 and up

Last gen LEDs:

Luxeon I, III, V and K2

The brightest single die LED out of the current gen ones is probably the Cree XR-E R2 bin. It can hit 250 lumen when driven withing spec and over 300 if overdriven. The Seoul P4 U-bin works better with the Maglite reflector though.

The Dereelight DBS can at least match the stock 6D maglite in throw and is much smaller.
 
I'm thinking I'll end up putting a P7 in my 6D, just to keep it useful. I've decided to design my own light first though. For cost reasons I'll probably build it around a high-output bi-pin incan. I also know I want to run it off of 3 series columns of 18mm cells in parallel. Ideally for length I'd like to go with 3 x 18650 cells, but I would go 1-2 cells longer, or with more shorter cells, for a significantly brighter light. Direct drive would leave me the least work, but If there's a better way for relatively cheap, I'd be game.

Thanks for your help! The links you provided were a big help!
 
If you haven't already, check here for some data about some Bi-pin bulbs. I'd recommend 3 Li-ion batteries in series instead of parallel, you can find more bulbs that work.
 
I'll probably go with an osram 64458, some control circuitry to regulate the power, and since the voltage should stay the same if I put them in parallel, 12 18650's in a 4s3p or 3s4p lineup, one for runtime, and the other for output. I plan to build it so that the body is thermally connected, with the batteries shielded, but I want to isolate the electrical circuit from the body.
 
Well, since you'll be using Li-ions anyways, why don't you use Li-ion D sized cells? More capacity and you won't have to bore anything.

There are also many parts here that you can buy.
 
I'd love to use protected rechargeable D-size Li-ion cells, but I've not been able to find them. The NiMH only put out about 1.2V, so that won't cut it for me. ATM the closest I've come is a 6s4p, or a 6s5p of 18650's
 
Last edited:
Those D cells defeat the purpose of going to a D-cell with only 5Ah. I need my light to have a minimum of 11.1V/12V, and enough Ah's to last 45 minutes to 2hrs. I considered a battery pack, but the design calls for the batteries to be separated in the name of heat dissipation. Weight's not really an issue, but it needs isolated circuitry, waterproof to a minimum of 50ft, plenty o' lumens, and decent runtime on a Osram 64458 without overheating. I've solved pretty much everything except the battery config. I PM'd LuxLuthor, but havn't heard back yet.
 
3x18650 won't fit side-by-side in a mag, even bored. You coudl bore two channels and two two parallel sets of 18650s, or three parallel sets of 17650s. As 17650s tend to have only about 2/3rds as much energy as 18650s, that's actually a more expensive way to get the same amount of energy. In the 17650 size, you're acutally almost better off going with NiMH. Only slightly less energy, but better peak current, and much more tolerant to series-charging.

The kaidoman are actually the most energetic cells you can actually fit into a 6D, about ~110Wh for 6 of them, but the protection circuits are crap -- the short circuit protection is WAY too touchy, thus the inrush current for most incan lamps is enough to trip it, so that they don't light up without a soft starter.

Another option to consider is packing your 6D with AW C-Cells. You could probably squeeze in 7 if you use the extra room in the tailcap. That's 85Wh, but for running an incan, that's probably not a very convenient voltage -- and an odd number, so no series/parallel.

For high current mods, a lot of people are using drill pack cells, like the A123s or Emolis. These have the advantages of being inherently safe, and have ridiculously high power desnity, but their energy density is actually worse than other LiIons.

One nice perk of running a C-sized cell is that you can basically line the inside of the maglite with a copper pipe to act as a heatsink -- a significantly cheaper solution than custom machining a battery carrier from barstock like you were describing.


Hopefully AW will come out with a ~7Ah D cell. If he does, some truly amazing hotwire options will open up -- such as Philips 5761 lights with over an hour runtime, and ROP-HIs with almost two hours runtime (assuming you can get rid of that heat...)
 
Thanks for your response, 2xTrinity, it was very informative.

I should probably clarify though. I already decided to put a cree drop-in in my 6d maglite, with a MOP cammed/camless reflector from kaidomain. I'll keep it around for a very long runtime flashlight and even a cheap drop-in with alkaline D cells will give me improved light and crazy-long runtimes.

What I'm working on here is a super-bright custom light, with very long runtime, that isn't overly huge. so far I've come up with machining some round stock to fit my needs(high quality, but expensive due to my lack of equipment), using extrusion that comes very close to fitting my needs, or giving up on a round design and just making it all fit in a nice package.

I need at least 11.1v/12V, higher would work too. To get that with NiMH batteries I'd need about 10 in series, plus several sets of those in parallel to achieve the runtime I need. With Li-ion's I can use less batteries to achieve the same result, despite having a lower energy density. If I only needed 9V or 6V, I'd probably use NiMH.

Since my electronics class was a long time ago, I guess the question I need to ask is about how much runtime would I get out of 24 x 3.6V 2400mAh 18650's in a 6P4S configuration, with a 10A draw hooked up direct?

It's kind of a long term project, and I'm just in the research and design phase now, so there's no big rush. It's just a goal at this point: Make a flashlight as bright, or nearly as bright, as the "Torch" Mag, but with at least 3 times the runtime without getting too hot to handle. Whenever I get it done, it will likely go for sale on the forum market place, and I'll start a new project.
 
Top