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

I donā€™t expect tint-mixing to happen. Itā€™s not really in the Overton window yet, and it would mean tripling Lumintopā€™s investment. Instead of a single reel of 3000 emitters (1000 lights), it would mean committing to three reels of 3000 each, or 3000 total lights.

Even if Lumintop wanted to make three times as many units, theyā€™d probably just offer it in three different flavors.

You could create a ā€˜Team accountā€™, as was done in the BLT GT thread; and ask sb to change the original post ownership from the miller to the ā€˜Team accountā€™. That way the OP (original post) could be updated with the needed information.

It worked well in the GT thread once it was done.

Iā€™m in for two

Please count me in for 1. Thanks.

I guess you should be ā€œpromotedā€ to a team member!!!
Thanks for the work done so far pepinfaxera :+1:

Agreed! Thanks pepinfaxera :+1:

MascaratumB and Jiffy : Thank you
.
. Not being able to be a member of the team
.
I help what I can.

I ignore (not speak) (not understand) English language
.
Google translator

Pero, estas haciendo un trabajo muy bueno y de gran valor para lo seguimiento de este proyecto :wink: Gracias!!

(But, you are making a very good and valuable work to follow this project! Thank you!)

Put me down for one plz

In for (2)! :smiley:

EDIT: Letā€™s make it an even (2)ā€¦ hehe

Iā€™ll take one as wellā€¦Thanks BLF and all whoā€™ve worked on this design !!!

I am interested.

Forgive me for not going through 75 pages of messages, is there a lumen output approx. on this light? And any idea how long until completion?

Output of FET triples is pretty well established and tends to be between 2000 and 3500 lumens depending on emitters. Last I checked a 219C (90 CRI D240 flux) FET triple, it was about 2500. A low-CRI XP-G3 (or low-CRI 219C for that matter) would be higher.

I made some progress on the firmwareā€¦

Candle mode is updated and more realistic now, simulated using six oscillators ā€” three of which modulate the other three like an old analog synthesizer. The result looks similar to a candle with periodic disturbances in the air.

Fixed a bug in thermal regulation and LVP, which were both running about 1000X faster than intended. So, those work again.

Thermal regulation uses 512 internal steps while searching for a sustainable level, which makes the adjustments invisible by eye. A lux meter is needed to see what itā€™s doing.

Made thermal adjustment speed change depending on how far it needs to go. It can happen anywhere from 1 step every 8 seconds to 8 steps per second.

Added the ability to calibrate the thermal sensor to room temperature. This should help with the issues related to different attiny85 units disagreeing on how warm it is.

The UI diagram is updated too:

Very sweet ToyKeeper! Thank you!

Nice TK!! :+1: ā€¦ Thank you. :slight_smile:

Iā€™m in for a single FW3A.

Iā€™m down for 2. Thanks

Oh, I forgotā€¦ I also made lockoutā€™s moon mode use the current ramp floor. Previously it used whichever floor was lower. So, the momentary illumination in lockout mode can effectively have two settings depending on which ramp you left active.

In case this doesnā€™t ring a bell, lockout mode doubles as a momentary moon mode. That way, after locking it and tossing it in a bag, it can still be used to look around inside the bag without unlocking it.

There have been some complaints about the mid-ramp blink on the D4. I find it useful as a reference point, but some dislike it.

Should the FW3A have mid-ramp blinks?

Currently the default setup is:

  • Smooth ramp floor: Lowest setting which lights up.
  • Stepped ramp floor: About 10 lumens.
  • Both ramp ceilings: Highest regulated mode, 8x7135 or about 3 Amps or ~1000 lumens. (full FET still accessible via turbo)
  • Blink when passing 1x7135 boundary, ~150 lumens, or level 65/150.
  • Blink when passing 8x7135 boundary, ~1000 lumens, or level 130/150.
  • Blink when ramp hits ceiling.
  • No blink at moon, because itā€™s usually pretty obvious when it hits bottom.

A ā€œblinkā€ means the light shuts off for 8ms. Each ramp step takes 16ms. So, itā€™s off for half of that individual step, half of a 60fps ā€œframeā€, just long enough to see.

Soā€¦ which of the following reference points should blink when the user hits or passes them?

  • A. Floor
  • B. ~150 lm, level 65/150
  • C. ~1000 lm, level 130/150 (highest regulated level)
  • D. Ceiling

Some prefer ā€œnoneā€, some might like ā€œABCDā€. I like ā€œBCā€. What sounds good to you?