I’m fully confident I can swap the star if needed.
I may need to file down the edges of the replacement star. Possibly file a slot or drill a hole for the alignment screw. And I might need to stack disks of copper sheet underneath to make the replacement the right thickness. But I’m confident I can do it.
I have to check if I have a small one (10mm) from KD, and see if it would fit.
I will tell later on 
Other than that, given the structure/inner shape of the head/hole, I guess I would not be able to mod it.
Yup, would likely have to deal with a small hole in the reflector. I have a few copper mcpcb’s with 219b’s on them, should fit.
> would likely have to deal with a small hole in the reflector
you can do it 
the small hole is only in the centering ring… easy fix
.
Jon, do you think there would be room between the wires and the emitter for a 3535 footprint and some kind of centering ring?
If a 3535 LED was too big then would a Nichia E21A fit and be any good in 1 of these instead? Not as bright but you could get a much nicer tint and CRI.
3535 should fit just fine.
There are dozens of lights with narrower bezels that work perfectly and give very useful beams with 3535 stars (Eagletac DC25A, Sofirn SP10b, Lumintop Tool AA, etc.).
The problem isn’t the fit… rather it is Lumintop’s choice to go with a 3030 star that fits Oslon White 1 emitters and nothing else.
I’m sure an E21A would fit and work fine… but you would need to do the following:
What’s the MCPCB size?
I do have E21A’s… the star will be the issue. I think I could fit one in the odd cavity and would have to clear mask and then solder in a place that is not a solder tab. But, the output would be disappointing for an EDC. Gonna have to think on it a bit. Not having the light in hand, I am not sure what the opening in the base of the reflector looks like. May just have to come up with one to try and frankenlight it. I do love the form factor of the FWAA that I have, but it is just to much flood for some applications. Charging every other day to save pocket space is not an issue.
Eurekatronix/Clemence has a 10mm MCPCB for E21A and E17A , very high perf dielectric and can do 4A :
Still going to need a regulated driver, I just ordered boards for a FWAA buck driver and I was looking forward to this FW1AA but the weird MCPCB/shelf design is just very disappointing.
From the photo, the stock star doesn’t look very deep.
Another option might be to use an ordinary 16mm star and just have it sitting on the aluminum rim above the cavity for the stock star.
Here’s some more photos from the flashlight.
I took the MCPCB out, checked the dimensions, tried a 10mm MCPCB from kaidomain, tried different TIR optics, and also compared it with “my” FW1AA (described before).
Took out the red glue and the kapton tape. There was some thermal paste below the MCPCB, not as abundant as I initially thought, because it was more concentrated on the sides of the MCPCB.
No, those circles are not holes, just parts slightly deep than the rest of the shelf.
These are the dimensions of the MCPCB. The total height of the shelf (from the below up to the top, or what thefreeman mentioned as B in is drawing) is 3mm.
This is the Kaidomain MCBPC, that can fit IF you file it on the “corners”. It is not a straight fit. The original MCBPC needs to be put into that hole vertically, and will not fit diagonally or in other positions. There is virtually “no space” between it and the inner walls of the shelf, only enough space for it to “slide” down or up.
See the below of the shelf and the thickness.
And the driver…in case it is needed.
As for the optics, first, here’s a comparison with “my” FW1AA (LH351D, in a copper MCPCB from the EagleEyE X3R, a 3535 gasket from Simon (trimmed) and TIR from the S1R Baton I), on the right.
My modified flashlight has a gap between the bezel and the lower part of the threads, due to the height of the TIR and the MCPCB.
Also, with an FWAA:
Here’s a comparison of beam profiles.
VS “my” FW1AA (right)
VS “my” FW1AA & FWAA
And now, some experiences with different TIR Optics.
First, the one of S1R Baton I.
The glass lens was NOT used, the black gasket was not used. If the gasket was used, the TIR wouldn’t sit correctly, and if the lens was used (with or without gasket), there would be a gap between the bezel and threads.
However, since those 2 elements were not used, there was a slight gap between the bezel and the TIR, which makes the flashlight not waterproof.
The beam generated has artifacts, light circles around that intense and typical hotspot that the SIR Baton has. There was some spill.
The beam, compared to “my” FW1AA, has a stronger hotspot, a slight tint shift in the corona, and less spill but it still exists.
This one is with a 15mm Yajimei TIR optic that I had laying around.
This one is a clear lens, narrower beam, less large than the optic above and than the original reflector.
Since it has a small hole for the LED, the black gasket was not used, but since it is less larger than the bezel, the glass lens was used. Given the height of the optic, there was a small gap between the bezel and the threads.
The beam generated is not good, due to the bad centering (no gasket) and the clear surface of the optic. The center has a “hole” and is yellowish/greenish, and the overall beam is “yellowish”. Not good at all.
I could have tried a similar optic - 60º pebbled - , but since this is a “thrower”, it wouldn’t be worthy trying a completely floody optic, in my opinion.
@ JaredM
Given these experiences, and given that the 17mm Yajimei optics is taller than the 15mm optic, I would say that…. maybe it can be used without the glass lens and eventually without gasket, but without certainties about generating a gap in the bezel/threads.
The original is 1.8mm tall, and the suitable diameter - to do what you mentioned - would be 15.5mm, more or less. I tried to use a 16mm PCB before and it didn’t fit (at least without being filed/trimmed).
Is there a source for that Baton optic? I like that it has a frosted center. I find clear, smooth TIRs with this center cavity frosted make smooth but intense beams. Clear centers often create ‘ghost images’ of the emitter in the corona and can be pretty distracting in use.
https://www.aliexpress.com/item/727344383.html
I have a few of these. Best beam I’ve gotten from a small TIR, even with 1mm² emitters
I am not sure if the Baton TIRs are available out there. I’ve seen people modding Olights but with other optics, so I am not sure if they can be found out there.
I don’t actually like clear optics, I normally go with frosted or pebbled ones because they make the beam floodier, which I prefer.
I need to go back to my searches in AliExpress to see if any of them is compatible with this flashlight! If I find one, I let you know
I actually just wasted a solid hour combing through AE for better options… no dice.
The way I see it, the FWAA is the perfect little flooder. Compact, triple, and the tail e-switch is actually convenient such that I can operate the light with my tongue while holding it in my mouth.
To compliment it, I’d try to get as tight of a beam as possible with an SST20 or 219b, while minimizing artifacts. But the more I think about it, the pairing I have right now with my FW1A TIR SST20 4000k and FWAA sw45k is as close to perfect for EDC as I may get without designing my own light from scratch. And until there is an appropriate clip for these it’ll never be my primary anyway.
The original FW1A clip is pretty darn good. The double sided clips really are no good.
Yes, I think a 10mm mcpcb will work with a 3535 LED and the stock centering ring
optimistic I am
Agreed.
FW1A / FW3A clip is excellent.
Unfortunately, the latest fad among flashlight manufacturers seems to be 2-way clips. And all the 2-way clips are garbage. The FWAA’s clip is no exception.
Having a 2-way clip is ONLY useful for carrying the light on a ball cap. If you’re like me and never do that a traditional 1-way clip is always better … preferably with deep-carry.
I’m hoping the 2-way clip fad passes soon so we can get useful clips with our lights again.