I could possibly deliver this CAD design by the end of June. Got a lot of things happening though with a new house and work for the next 3-4 weeks. I think a reasonable expectation for this light would be 20mm x 85mm. 14500 saves 15mm, but single optic will add some. A proper deep pocket clip will be a top priority for the design as well.
Seems like identifying the optic might be a good first step, as it could affect the width and length dimensions. Looks like Ledil has some 16mm by 9-10mm options.
Or could this use the 10mm Carclo optics. Those are only 6mm deepā¦
Yajiamei also has 17*9.7mm 5050 TIRs in a wide variety of angles.
It looks like each of the optics in the Carclo triple used for the FW3A is 10mm. So using the Carclo 10mm would be like using one of those. Using one of these just feels right for this, but I donāt know how that square shape of the optic holder combined with the presumably larger circular opening would play out. Maybe Iāll order one of these and stick it in an FW1A with a FW3A bezel. Think thatās worth doing? Which one should I order: 10mm - Carclo LED Optics
Probably medium spot or frosted medium spot? The narrow spot cd/lm looks pretty throwy for a general use light. Or maybe I should just get all 3. If anyone else already knows this wouldnāt work well, Iād also appreciate that info :).
Looking at the proposal, I would just say that the 20mm diameter would be too thick. My opinion, of course!
Looking at the Tool AA V2.0, it measures 89.1mm x 18.5mm. I find that thickness ideal for an EDC light. I believe Eagletac has one that is similar to this one.
Making it 20mm seems to large.
About length, 85mm would be nice, but I guess we can save some more space too, maybe 2 or 3 mm (optimizing the tailswitch and the pill).
Donāt ask me how to do it, though, as I am no expert 
And thanks for your effort in advance too 
if you make FW1A ti, i will take this
The more Iāve thought about this over the last few weeks, the more Iāve realised I would buy a couple.
I live the Tool AA 2.0, but I love my fw3a and Fw1a lights much more.
While I like a smaller single led light, the driver, both button, firmware and power just canāt be beaten from the FWxA lights.
Iāve collected so many of them now, all for different LEDs, tube sizes, I would love one thatās 4mm thinner.
My favourite size is 18500, so the height would be similar and amazing for a 219b mod for around the house.
Iād be very happy if this happened!
well, aparently there is an Anduril option for 14500 (no AA)
There isnāt triple optic for made FW3A 14500/AA with E21A?
There really wouldnāt be much point in making it a triple.
A single LED is almost always preferable unless there are specific reasons to use more. Like, if one LED can only make 1000 lumens and you want to make 3000 or 4000, and donāt mind sacrificing throw, and are using a battery powerful enough to power it, that would be a reason to use more than one. Or if itās a lantern which needs light to come out at all angles, or if it uses tint ramping, those would be reasons to use more than one.
But for a 14500 light, the battery isnāt really strong enough to justify multiple LEDs. Also, if the whole point of the light is to make it small, multiple LEDs would work against that goal.
It can sometimes be useful to put four tiny LEDs under a single optic thoughā¦ like four XQ-E or E17A LEDs. This allows tint mixing to make the tint more pink. Or in the case of 4 x XQ-E, it has been used with RGBA emitters and a diffuser to produce a rainbow.
Those are pretty specialized cases though. For a general-purpose 14500 light, a single LED is the obvious choice.
(FWIW, Iām counting the Cree XHP series as a āsingle LEDāā¦ there are good reasons to use XHP35 HI and a boost driver, for example)
brilliant post!
thanks for taking the time to illuminate the fact that a triple, can have less Throw, and requires more power, than a single.
I think there is a commonly held misundertanding that more lumens equals more throwā¦ not true if the hotspot is less focused
Same. 14500 is already a compromise because I only really use that battery size for EDC or if Iām travelling. Iād want to be able to bung an easy to find AA in it if I took it on holiday or whatever.
But yes indeed, Iād be interested to see it, though the short tube 18350 option for the FW3 further weakens my need for a 14500 version. Still, needs are generally ignored in the world of the flashaholic, and desire is forefront. 
I asked because there isnāt any triple 14500 on market.So will be the first. Small leds like E17A can handle not much current.
New 14500 batt.can give the necessary current but of course runtime and throw will be a bit sacrified.
That driver is designed specifically for the Boruit D10 (probably works with most of the varieties of that light too, like Sofirn/77Outdoors D25), which takes an 18650.
So far, Anduril works with several ATTiny MCUs, which weāve made work with Li-Ion chemistries. Cell size doesnāt really matter at that point, just design a smaller driver to fit the host. Thereās not a whole lot of reason Anduril canāt fit on a 10mm driver besides space, which would allow for 10440 and even 10180 lights to run Anduril.
This thread has some information about the quest for AA Anduril support. To my understanding, thereās a prototype driver design being tested currently.
Yes, but Iām still waiting on a good 14/15mm 12v boost driver.
All the triple optics Iām aware of have a much larger diameter than makes any sense for a 14500 light.
But triple is more efficient than single and with the Vapcell H10 or Shockli 1000mah cells, it will be able to make use of triple emitters no problem. But Iām hoping for a 14500 version of the Manker MC13
What, uh, what do you gain with a 14500 over an 18350 in a light with a 41.7mm head diameter? Other than length.
Triples and quads have been great these past few years, but I miss single-emitter lights designed to be as small as possible. They generally had the most practical beams and were the easiest to carry. Extra lumens are neat, but on something small I carry around all the time I donāt really need more than ~100 lm or so.
So for a 14500 light Iād be pretty happy with a really thin light with a single E21A and a 500 mA buck driver or linear driver. Even just a single 7135 chip and a MCU would suffice.
But I probably wouldnāt buy a 14500 triple. Itād be wide enough that I may as well just use an 18650 light insteadā¦ and floody enough that Iād need to use twice as much light to see what itās pointing at.
But if efficiency decreases with power, shouldnāt the consumption for a certain lumen output go down with increased number of LEDs?
This is why multiple emitters are used instead of one, to increase
efficiency at high currents.
Problem the head size of 14500 host for put a triple, Donāt know if there are lens/mcpcb smaller than 20mm
If you use this lens size and also need to increase head diameter,yes there is no reason for do that.
I just launch idea
Efficiency, as measured in lumens per Watt, is usually higher with more LEDs. Like, at a given power level it may produce 5% more lumens or 10% or even 50%ā¦ and the greatest gains happen at the highest power levels. At low power, thereās very little difference.
Efficiency, as measured in candelas per Watt, is usually higher with just one LED. That is, assuming the overall diameter of the host remains the same. Take the FW3A vs FW1A as an example. The single-emitter version is about 500% as efficient, illuminating its target about 5X as bright at the same power level.
So youāre both rightā¦ youāre just talking about different things.
In the case of a 14500 version of the FW3A or FW1A, the topic of efficiency boils down to this: At a given power level, would you rather have a ~10% increase in lumens or a ~500% increase in candelas?
Basically, do you want an extreme flooder at ~2 cd/lm or a balanced beam at ~10 cd/lm?