Fireflies ROT66 Flashlight

How pleasant the beam looks really dependent on the tint, where it sits with respect to the BBL, instead of the CRI. I tried many 90CRI 219C 4000k and the LH351D 4000k 90CRI and could never get myself to like them no matter how much I try due to the yellow/green tint. However, I recently tried the 4000k E21A 9080 and 219B 4000k 9080 from Clemence, and they are just beautiful. I see absolutely no hint of yellow or green. Just pure neutral white. I don’t even notice any rosiness unless I compare it side by side with a 219c 4000k.

Hopefully this SST 95CRI they are using is really below the BBL otherwise at 4000k, it will be yellowish.

The nice thing about the ROT66 is that its bezel can be unscrewed, and you can just sandwich a Lee minus green Zircon filter between the lens and optics if the beam is slightly too green.

I really want a multiple optics light sporting E21A 9080 emitters.

I have values for calculated lumens for the ROT66 nichia. Can someone check it with real measurements?

calculated modes for Maukkas Nichia + Lexcel code
(doesn't value nonlinearity from the LED curve, doesn't contain Moonlight)
(1+7+FET)

ROT66 Nichia in Lumen
Modeset 1 3860
Modeset 2 102, 3860
Modeset 3 102, 705, 3860
Modeset 4 10, 102, 807, 3860
Modeset 5 10, 40, 102, 807, 3860
Modeset 6 6, 32, 102, 378, 807, 3860
Modeset 7 3, 80, 212, 433, 807, 2218, 3860
Modeset 8 807, 3860

The Nichia 219B has a pretty nonlinear output curve. It has a slow startup time at least, and seems like it may have a slow fall time too. To estimate the output, you may need to go from ~10/255 to ~245/255, instead of 1 to 255.

There’s a guy adamant on Twitter that Vinh has been in two movies

Has anyone got their replacement head for the ill-fated XP-L yet?

I received my SST-20-W40H ROT66 a few days ago. Excellent and pleasant tint, but I think it is of 7x7135 because its output is pretty close to the 219B version when both driven by 100% 1x7135 + 100% Nx7135 (level 3 of mode-set 4).

Alright, just checked with Fireflies. SST-20-W40H is of 7x7135, while the cool white SST-20 is of 14x7135.

That’s different than what Jacky told me, so it’ll be interesting to see what shows up.

I can’t open my 6500K ROT66

:person_facepalming:

Maybe you can switch to the mode-set #4, level 3 (moonlight is level 0) and measure its lumen number. Should be > 1400 lm if 14x7135 are used.

The main problem of the bump is that pretty much all carriers have onl spring loaded one side

Now if we want to fit different cell lenghts like from unprotected to protected the springs need to be relatively long
Also Copper alloy springs reduce their leght by time a bit as they are not as flexible as steel, but they are needed for the current

If flat top cells make a bump problem an easy fic is solder blobbed batteries, but this is not a bike light under normal use a bump is very unlikely

It is possible to glue some foam on the tail of the carrier if its not getting pressed enough to the front,
Astrolux MF02 has that problem one one batch the outer contact ring does not maker proper contact on some lights due to production tolerances

The carrier uses 3 rods and yes one of them carriers battery + 2 carry ground, this was decided early in the design process to get on both sides battery + availiable

Is it sst20 or xpl hi 6500??

With barely 67 mm space between positive contact plate and fully compressed spring I don’t know any protected 18650 which fits. Is there any at all?
I don’t think a solder blob will help in the long term, it will rather wear out the spring and the problem will recur somewhat later.

If this springs reduce their length over time I foresee more bump problems in future. They are simply too short - or the wrong selection at all.

It seems the only decent solution to the bump problem is adding a capacitor to keep the driver powered up when the batteries loose contact in a bump. This way the light still turns off in a bump, but it turns back on at the same level instead of resetting to “off”.

Did you take lumen measurements compared with the 219B? How does the tint compare with 219B or any 5A/5D Cree tint flashlights? Maukka tested the SST-20-W 4000k 95CRI from Kaidomain to have some greenish tint.

About 5% more lumens if compared with the 219B.

I’m not answering the specific tint question because I don’t have a spectrophotometer to analyze the spectrum. I’m satisfied with it though.

However these two ROT66’s I newly received both have the bump problem. They got reset when hit at the tailcap or shaken more quickly. Adding an additional plastic disk between the battery carrier and the tailcap can fix this and increase 10%+ Turbo output (well, poor contact), however I don’t like such a poor idea, so I’m returning them to Fireflies and asking for replacements.

Just an FYI Mike, it seems Sofirn have listened and their new version of the Q8 does include a chunky buffer capacitor. There is some information about this on Djozz’s Sofirn Q8 thread.

e.g. Sofirn Q8 group buy (gb closed, but try PM to Sofirn for discount) - #389 by ImA4Wheelr

Quite why this is so difficult to comprehend, and seems to be so resisted, by other designers, is a mystery to me. Just as selecting decent springs or simply making the battery compartment fit properly seems to be a lost art.

Puzzled why this thing needs a cell/battery carrier which just adds a load of complexity and potential problems for a basic parallel cell arrangement.

Many more contact surfaces in the current path, and bits to go wrong, in design, manufacture/assembly, and in service. Added cost, mass and volume too.

And complication when changing cells.

Edit: Trying to tug them out, perhaps with some fabric ribbons to help, if you remembered to tuck them behind when loading, or bash them out, or insert fingernails, penknife etc and damage the wrapper. Whilst tugging the springs sideways and further compromising their weak performance. If a plastic spacer put at the button end to prevent reverse polarity, even more difficult.

Edit2: If suffering poor performance and suspecting excessive resistance in the current path, well, there are so many places to suspect, look at, and try to fettle, good luck trying. Unless it’s just a hobby to muck about with.

Instead of faffing about with a carrier, just unscrew, drop out old cells, drop in new. run a finger around the top to check they are all the correct way up. Close.

The simple SRK/Q8 etc. arrangement is tried and tested. It also suffers bump disconnection because the contact ring in the head is solid, not spring loaded. But that is easily avoided with suitable circuit design.

Out of curiosity, what happens if you do this?

  1. Loosen the battery tube at least a full turn or two.
  2. Tighten the tailcap as far as it’ll go.
  3. Tighten the battery tube.

And if that doesn’t work, flip the battery tube around. I think it looks better backward anyway.

On mine, the parts can actually tighten more than necessary. When I tighten the tailcap first, then the tube, the only thing which stops it from turning farther is that the battery carrier is pressed firmly against the driver’s retaining ring. It’s so tight that the spring doesn’t even matter; there is no room to move either way.

Anyway, if it just needs a bit more thickness inside the tailcap, perhaps a printed copy of the UI cheatsheet would make a good spacer? Or perhaps a 3D puffy bubble sticker of a firefly.

Firefly puffy sticker you say?

Will stock them at my flashlight modding store…