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

The more I look at this the nicer this looks. I like the quad better than triple. Perhaps Lumintop should make a FW4A since there isnā€™t a rear clicky quad in existence and it will be more efficient. Perhaps use 21700 batteries instead of 18650.

That would be great, after this one.

+1 :+1:

Itā€™s alive ā€¦. and twitching!
I want one.

Thanks mate, now there is a copper version :wink:

Yes, lets not distract them :smiley:

+1 . :wink:

Prototype 3 was sent to Germany, unfortunately not to Toykeeper.

I wouldnā€™t call this a prototype, only anodisation samples to deside the color.

Good call. And thank you TK so much for confirming it with Neal.

Not sure why doing a nice anodise should be proving so tricky, isnā€™t that what anodisers should just know how to do, itā€™s their job after all ? Perhaps itā€™s to do with the alloy used. ISTR that it also took Convoy a while to develop their clear anodise to their satisfaction.

A clear anodise on a ā€œrawā€ machined finish is also merciless in exposing the slightest defects in workmanship, when the machining marks are to be a design feature, not something to be covered up. Just handling the pieces, moving them around, keeping them almost surgically clean before processing, etc. needs care, individual bubble wrap etc. if you donā€™t want them dinged up, then a good hard anodise to then protect them. Unlikely to be real mil-spec HA3 stuff, but still something good, semi-hard.

It was probably unrealistic to hope for anything like that at this price-point.

Bead blast, a quick dunk in the tanks, then just chuck them all in a bin together is the production-engineering easy way out.

The solo artisan perfectionist, toiling at their lathe, can make things much more beautiful, and go from concept to execution in a few hours, but that doesnā€™t easily scale, as I think we are discovering.

Sitting behind a CAD screen, rendering, then perhaps coding for CNC (thatā€™s usually a different, more practical, and dare I say more skilled discipline), is not quite the same thing as making swarf, and sometimes pieces donā€™t turn out quite the way it was hoped, or are simply un-manufacturable, and another iteration is required, and things can become delayed interminably. The skill is not to let that happen. And to use a team to review, accept constructive criticism, involve the production people as early as possible, and not just live in an ivory tower.

It looks as if they have got a nice looking bland result.

For those interested, you could take a look at e.g http://www.metroplating.co.uk and study their ā€œservicesā€ options. Each provides a succinct summary of each process. Specialist finishers like these are crucial to the UKā€™s manufacturing competitiveness, particularly in motorsport, aerospace and defence, and metroplating one of the best and most nimble IME. There are hundreds more though to choose from, just Google e.g. ā€œanodising ukā€.

Or have a go yourself. Itā€™s not difficult. Research it, or just buy a kit from e.g. http://www.gaterosplating.co.uk/index.php?route=common/home or https://www.chronos.ltd.uk/acatalog/Anodising_Kits.html

Something for modders to experiment with, and a bit more creative than e.g. just baking your torches.

Start with something cheap, strip off the old with sodium hydroxide (e.g. some drain cleaners), maybe buff it up a bit, and get practicing. Some battery acid, inkjet printer ink, some sort of DC power supply, a cathode, some Al wire, and youā€™r good to go. Perhaps in time for the 2019 Old-Lumens challenge. Use titanium instead, and apply by brush, use masking, and you could make a really unique work of art.

They did not mention. I did ask though.

All I have is a couple of pictures, and the pictures donā€™t show any seams.

FWIW, Neal has asked to have a working sample made and sent to me for testing.

The firmware used should be this one: http://toykeeper.net/torches/fsm/anduril.2018-12-02.FW3A.hex

Recommended fuse values are:

  • Low: 0xE2
  • High: 0xDE
  • Extended: 0xFF

So, BOD is enabled. Without BOD, itā€™s very difficult to exit momentary mode because the standby drain is too low. It took several minutes to drain far enough to cause the driver to reboot, and with BOD itā€™s only a few seconds. This should also help avoid other theoretical issues with low voltage.

Factory reset isnā€™t included, because itā€™d cause too much delay making sure there are no bugs. I did a quick proof of concept but there were too many issues to move ahead with the change right before production. So I went back to the 12-02 version which, despite the date indicated, has been under test for about three weeks on various lights. Itā€™s virtually identical to the 11-11 version.

Anyway, it seems things are finally moving quickly.

Does BOD mean brown out detection?
the ability to stay on through super short power cycle?

If so I think this would be a good thing.

Is it a good thing?

Yeah the Emisar is watt I was comparing it to.

However there IS a nice subtle design aspect of the FW3A body I do like A LOT which stands out. The sleekly well-crafted ā€˜combat gripā€™ ridge right above the tail cap. Extremely useful and a feature I wish many of my rear switched single-cell flashes had incorporated.

Elegant simplicity.

I may buy this just for this brilliant point alone and the way it was executed.

Kudos to the brain that insisted it being there.

:+1: :beer:

PS. I wanna clarify that this finger grip ridge isnā€™t uncommon per se. Just that it appears the FW3Aā€™s is actually large enough for even glove wearers butt yet isnā€™t too large as to be ugly and out of proportion to the rest of the outer design. Plus it looks well smooth machined and thus comfortable. Sharp ridges and fingers pressing down donā€™t work for me. :sunglasses:

Thank goodness!! ā€¦ :+1: :+1: :+1:

Yes.

It is like LVP, but built deep into the attiny85ā€™s innards. Basically, if voltage drops below 1.8V at any time, even for a few microseconds, itā€™ll stop code execution completely. The main purpose of this is to make sure the MCU wonā€™t attempt to do anything when itā€™s out of its supported voltage range, because its behavior is unpredictable when voltage is too low. This helps prevent random eeprom corruption and things like that.

The benefits outlined in the attiny manual donā€™t seem to matter for flashlight purposes, but it doesnā€™t hurt. Iā€™m using BOD for its side effect ā€” increasing standby power. Because it makes no difference to battery life but it makes a lot of difference when attempting to exit momentary mode. Without BOD, it could sort of get stuck in momentary mode because there was no fast way to force residual power to drain.

Without BOD, standby power was so low I had to use a high-end 6-digit Fluke meter to measure it. So, basically zero. Or ~0.0001 mA, if I recall correctly. This is impressive, but too low for this purpose. So with BOD, standby power is about 0.02 mA. Low enough to be negligible for battery life, but high enough to make the light shut down only a second or two after power is removed.

Put me down for one!

Please add me to the list for one!

Put me down for one please!!

I would be interested in for sure. Thanks!

@Toykeeper
I redid the AndĆŗril UI graphic a bit and sent it to the BLf adress on Toykeeper dot net
Any comments if good or useless?

Hi there, cool looking project!

Put me in for one.