FW3A, a TLF/BLF EDC flashlight - SST-20 available, coupon codes public

The poll is based on preferences. I prefer floody lights and I also prefer ones that might not heat up as high as others (although I’m not sure if it’ll be a noticeable difference). Having two lights with 219C emitters has shaped my preferences and I’d prefer this light to be able to handle heat a bit better than my D4 and I’d also like it to illuminate the area directly in front of me to the best of its ability. I’d love to see it supplied with an even floodier led as well.

All due respect Jason, I disagree. A person’s vote is their’s to do with as they like…. Whatever their reasoning, it does not invalidate the poll. Seems to me anyway.

It’s an arbitrary number representing the “score” for each cell. But the score itself means nothing, aside from its weight compared to other cells.

On the “all votes” maps, yes. But those don’t attempt to correlate the emitter type and color temperature, so all emitter types have higher scores at 5000K.

The more useful data is in the correlated maps. On those, LH351D has a slight bias toward 4000K. Perhaps I should run it again without votes where XP-L HI was ranked first, to get a better view… Here’s how that looks:

A good question. On the one hand, answering it properly might take one, or more, graduate dissertations.

On the other hand, I think there is a simple, pragmatic answer that applies to the present context, which is that the context of this poll is focused on emitter choice for the FW3A project. It isn’t a general poll about what emitter people thinks looks best, and it would be mistake to try and apply the results of this poll to a different or more general situation, at least not without interpretation and some added hypothesis testing.

Anecdotal information like PBWilson’s explanation for his reasoning can provide insight into interpreting poll results, and into designing subsequent polls. In this case, it reinforces the idea that at least some of the people voting in this poll are doing so in the context of their existing portfolio of flashlights.

My own considerations, as best I remember them now, a few days later:

  • I want at least 80+ CRI, and preferably 90+ because I already have enough lower CRI lights, most of which will probably be given away or receive emitter swaps. I don’t need another one.
  • Higher max output is good. Lower heat/ higher efficiency is good. Tighter beams are good in an 18650 tube-light triple.
  • The max output, high efficiency and throwier beam of the XP-L HI isn’t worth the added cost and the poor CRI to me.
  • The slightly throwier beam of the 219c isn’t worth the significantly lower max output and the lower efficiency wrt toe the LH319d.
  • My existing higher CRI lights already have Nichia (219c and 319at) emitters, and I have more such emitters for new builds or upgrades of existing lights.

My priorities would be different for a different light: single vs triple; 18650 vs 26650; ~20mm reflector/optic vs 25, 30, 40, 50mm, etc; general purpose beam vs thrower v s flood.

Anyway, I really hope this doesn’t end up being available with only 70+ CRI XP-L HI. If that were the case, I’d really like a version with everything but the emitters, at a lower price.

This raises some questions about the XP-L HI for me:

  • Will a 3x XPL-L HI draw significantly more current off a single high-drain 18650 than would 1x of an efficient low Vf emitter like an SST-40 or Luxeon V?
  • Whatever the peak draws and outputs are for these two possibilities, thermals and battery discharge curves are going to place severe restrictions on duration of peak power, so what do things look like at other power levels? 20W (which is also going to heat up fast)? 10W?
  • Will a 3x XP-L HI deliver a significant throwier beam than 1x of something like an SST-40, Luxeon V or XP-L2 at similar power consumption levels?
  • How different will the beam profiles be?

These may be pretty basic questions, but the answers aren’t all that obvious (to me, anyway). Have they already been addressed in this thread.

A big part of what I’m wondering about is that, given that this light is a triple, are there some emitter options that take better advantage of that tripleness than others?

It’s pretty clear that hitting the 3000lm goal in a 1s 18650 light without triple emitters (or an as yet non-existant, compact, high-power boost driver), isn’t really possible. So, certainly, any emitter that hits that output goal takes advantage of the tripleness. The goal of regulation to 1000lm doesn’t require a triple, though the lower Vf of a triple increaces the amount of time under which that regulation can be maintained.

There are other angles to be considered, but it kind of seams like taking advantage of trippleness isn’t necessarily a strong tie-breaker, at least not on its own.

It seems to me though that this project has a bias towards a bare, elegant, refinement laced with innovation. It isn’t an Emisar light (which is to take nothing away from any of the Emisar lights). The FW3A it isn’t pushing the limits of output nor compactness for an 18650 light (the D4 is), or compactness and throw (the D1). Which isn’t to say that it isn’t still trying to hit similar levels of output per-emitter, but it’s also seeks to provide regulated output at ~1000lm, a powerful but easy to use UI, a tail e-switch. For me, 90+ CRI is more refined and elegant than 70+ CRI. For me, using an emitter not yet seen in a large or limited-run flashlight is a bit of innovation.

In any case, it’s great to see this project moving forward. I’ve no interest in second guessing choices that have been made, I am interested in understanding them and how they might inform remaining choices for this project, and also how they might inform future projects I might want to undertake. I look forward to being able to buy a FW3A, light, even if it has XPL-HI that I want to replace :slight_smile:

Very nice research project TK, though I’m yet to cast my votes, if the poll is still open ?

Seeing the way it is going in advance is helpful, for some tactical decisions, which I won’t do, that would be wrong.

Now, are the two preferred choices actually going to be available (remembering the XPL-HI might cost 20% more, on the total delivered cost of the whole thing :open_mouth: if not absorbed by Lumintop), which some may not have appreciated, though it was clearly stated ?

Assuming there will be a choice of 2 (I hope).

Still, it’s fun, and anything would be better than XPG3s (seems to be recognised).

A shame proto2 has these. Any more feedback on it ?

Whilst we wait.

PS: why does an XPL-HI cost so much more than an XPL ? Perhaps the dies are cherry-picked from the best batches.

I’ll sign up for 1 unit, doesn’t matter for me which emitter wins (well, unless its really esoteric).

Put me on the list for 2 lights, please.

It doesn’t.

But XP-L (HI or HD) costs more than the other emitters being considered.

Inner diameter is 23.1 mm. It fits onto a D4, but it’s necessary to take the O-ring off first and then replace it after putting the clip on.

But it does cost much more than a plain XP-L (not HI or HD) which is why The Team chose it for the Q8.

Maybe that has changed (out of touch and a bit burned-out still, don’t let that happen to you).

We might have liked one of the others but it made no sense (given the budget). And we got a very nice LED, as it happens.

Still wondering about 20% more for this choice if that’s real, maybe that doesn’t matter to some.

@TomTom, the plain XPL is actually named XPL HD.

XP-L comes in two varieties:

  • HD: High Density (has a dome)
  • HI: High Intensity (no dome)

The BLF Q8 used XP-L HD.

There is no such thing as a “plain XP-L”, but usually when HD or HI isn’t specified, it’s safe to assume HD because it was created first.

I stand corrected.

I did say I was burned-out and can forget some things.


Would happily pay $10 each for 2 additional clips if that option ever pops up.

Samsung 5000k would be my vote. Seems like it would suit this host really well

Edit: Not a fan of the ranked choice voting. I wouldnt want any XPG in it, yet had to still include a vote for it. Also same with color temp. Wouldnt want anything higher than 5500k or lower than 4000k, yet had to vote for them…

Updated list : … Post moved to Page 130 , FW3A, a TLF/BLF EDC flashlight - SST-20 available, coupon codes public - #3870 by pepinfaxera

Short list, Last requests .

Thanks, TK!

Did you like the look of the FW3A clip on the D4?

This voting means that if you vote for them to have last place, you’re voting that any other option would be better. That’s… how ranked choice works. Just because you had them on the list doesn’t mean you’re voting “for” them.

Yep… it is easy for any of us to go brain dead & get forgetful of even the basics when we get “burned out”.

Get rested up…. you don’t want to end up ‘burned up’ from terminal “burn out”. :+1:


For example, consider this ballot:

  • 1st choice: 219c, LH351D, XP-L HI
  • 5th choice: XP-G2, XP-G3

This ballot says: Anything except XP-G. It votes “for” the first three, and votes “against” the last two.

If someone wants, they can further differentiate, stating which of the first three they like best or which of the last two they dislike most, but it’s not necessary. The ballot style allows a lot of flexibility in expressing one’s views, and the Condorcet vote-counting methods use all that data.

On a side note, Condorcet-based methods are not to be confused with IRV (instant runoff). The ranked-choice ballot looks pretty much the same, but IRV ignores most of the data and produces weird results. It gives the voter the illusion that the nuances of their views matter, but it’s a false illusion.