TLF/BLF/Lumintop FW3A review (18650, 3x XP-L HI cool white)

FET driven lights make terrible calibration lights because there are too many variables involved.

Jason, he only asked that someone with a properly calibrated light tube take a reading on the 7A FW3A in comparison to the 1A FW3A so he can make a wise purchase decision. :wink:

Fair call. I’ll stick to gazing into the emitters on moonlight mode and going “oooh” then.

Here ya go HC, 18 Samsung LH351D W6 5000K with domes sliced off, in moon mode…

Future FW3A buyers can look forward to the day the Samsungs are available, or mod their new lights. :wink:

Oooh

You could have much more accurate readings if you want.
Maukka sells calibration lights for homemade Lumen Tubes/Integrating Spheres.

As a first approximation, expect 15% less at 3000K compared to the cool white

I’m basing this on what Cree specifies in the datasheet for the best flux bin of each color temperature.

Like DB Custom says, the datasheet doesn’t tell you everything, and even what it does varies. The worst bin of 3000K would rate at 25% less than the best bin listed of any XP-L HI color temperature. The best bin of the 3000K against the worst bin of the cool whites would only be 6% lower output.

Since max output of the FW3A depends on a FET, you can also get significant differences if the forward voltage is different.

awesome review, thanx!

I’ll take it. Do you need my shipping address?

Did shaving make them throwier (is that a word)? How much was the tint shift?

I’ve shaved all the LH351Ds I have, it made them throwier (more throw than a X-PL HD, less than an HI… maybe more throw if you cut very close to the phosphor?), the color temp goes down ~500-1000 degrees and the tint gets rosier depending on how much you slice off. I took off maybe 1/2 of the height of the dome to get the result I wanted.

Tests of them show about 10-15% output loss at most power levels, and I’d say it looks that way to my eyes as well. In the same host at the same current a shaved 4000K got pretty close to the throw of a stock 4000K SST-20 and had a wider beam.

I really need to learn soldering. Sounds exactly where i need it to be

If you’re bold/stupid/have steady hands you could partially shave some lights without removing the MCPCB. Not saying that it’s a good idea or that I would ever do that out of sheer laziness but it’s possible. :wink:

Are there any videos available showing how to change optics on the FW3A?

I haven’t found any, but you’d just have to unscrew the bezel, change the optics and screw it back again.

Received my production FW3A w/ XP-L HI 3D today. Works perfectly, even though I did tighten the driver retaining ring before even turning the light on for the first time. The finish is just slightly lighter and bluer than the proto.

The little protrusion on the clip ring is perfectly fine. It helps with the cigar grip, while the proto with a narrower ring is a bit slippery.

The emitters aren’t rosy for sure, but I’ve noticed that lower CRI emitters don’t look as green as for example CRI90 219Cs with similar duvs. This measured 4600K duv 0.0063 at 1x7135, but visually was alright. Gets closer to BBL on higher modes (8x7135, FET). The fine visual experience is also explained by my current ambient lighting conditions that are even greener (4800K duv 0.0085) because of UV blocking windows and lots of trees outside.

Integrated, the tint is not as bad as in the hotspot. So there is some tint shift in the beam but it’s not very visible to the eye. Output is 139 lumens in the 1x7135 mode.

Nichia 219b sw40 integrated at 1x7135, 83 lumens:

Your 3D doesn’t look rosy? Mine sure does. I like it.

I'm not sure what the graphics all mean, never did understand, but where can I find the results regarding above/below the BBL? What's the test result relative to the BBL?

I find my 3D FW3A surprisingly more blue/green than rosy.

Duv (delta uv) is the metric which shows how far and in which direction the source is in reference to the BBL. Positive duv is green (yellow on low CCT sources), negative is rosy/pink/magenta. At least to me, some emitters seem visually more neutral on a white wall than others even with similar CCT and duv. Even a slight tint shift within the beam also affects how the beam on average is perceived even if it’s not obvious on non-rainbow like emitters like the XP-G3.

I would consider duv ±0.0030 neutral. The 219C CRI80 or XP-L HI in CRI70 look neutral even above 0.0030, but most CRI90 emitters do look greener even when they are closer to the BBL. SST-20 for me is the worst in this regard, looking greenish even when it’s near the BBL.

Objectively, the 3D in my FW3A is greenish, but much less so when evaluated visually. Most high CRI emitters we use would look pretty horrible at duv 0.0060.

The 3D had to give way for some dogfarts (LH351D 5000K CRI90) and a slight minus green filter (Zircon 804) and Carclo 10507.

1x7135 output dropped from 139 lm with XP-L HI to 125 lm with just the emitter change. 10507 brought it up to 126 lm and finishing with the minus green got me 112 lumens.

FET output at turn on (30Q):
XP-L HI 3D: 3035 lm
LH351D dogfarts: 2626 lm
with 804: 2334 lm

Final results is under BBL with 1x7135 and full FET.

maukka, any chance you have some comparison beamshots?