damn one day after mine was shipped... Well I think IĀ“ll have to flash Andruil on it myself.
Iām loving the improvements so far. 
Looks like Anduril option is now on the website
EDIT: Hank says heās using 2019-5-30 version of Anduril
I read on reddit that the D4 should be soon available in the same sand color of the D18! Is this true? I wish he also add a knurling on the tube like the titanium version.
One thing that I thought was weird is that my Anduril D4 Shipped on 5/28 has no muggle mode.
I wish we could get the smooth Green bodies back.
I also wish it, but hank said no
Aye. I would love Green or Titanium, but sadly I missed my chance for either of those options.
As the new D4 has flashing pads on the driver, will there be an adapter available, or will I have to get creative with some modelling clay and bare wires? After trying the improved 6/02 version of Anduril, I simply must have it on my flashlight.
It looks like you got one with a 2019-05-21 build, but I think the D4 should mainly ship with a 2019-06-05 version instead.
Muggle mode has been removed at Hankās request. It seems to cause too much confusion. Iām kind of happy to have it taken out, TBH.
Iām hoping Hank or someone will make the pogo pin adapters commercially available, but I donāt know what will happen.
Itāll definitely be required for the Anduril version of the D4 though. It seems those are using a new tiny1634 driver, so the pogo pin adapter is the only way to flash it. A SOIC8 clip wonāt work.
The RampingIOS version is still tiny85 though. That code hasnāt been updated to run on tiny1634. So if anyone wants to make sure they get tiny85, order the RampingIOS version.
Would a hex intended for tiny85 not flash to a tiny1634 driver? Is tiny1634 a significant improvement?
Tiny85 and tiny1634 are not binary-compatible, so a .hex for one doesnāt work on the other.
The new chip is a significant improvement in some ways. It has twice as much space for code, and it has quite a few more pins available for inputs and outputs. Past designs have often been rather limited because there werenāt enough pins, and that restriction mostly disappears with t1634.
Nice!!! Sounds like itās time for your to start writing firmware to take advantage of the new chip TK!!!
So what new features are you going to add support for to take advantage of the extra space and pins?
Some suggestions:
- Support for a speaker - so when we enter lightning storm mode in Anduril the light will play the sounds of simulated thunder with each lightning flash.
- Support for a sprinkler - Why just limit lightning storm mode to the sounds and sights lightning? The light should also provide rain so we get the full experience!
- A calendar - so the light can sing us Happy Birthday on our Birthday (using the previously referenced speaker)!
- An accelerometer - so the light can detect when itās been dropped and then tell us so through the speaker. Iām sure weād all like to hear our light yell out āOUCH! WATCH IT DOOFUS!!!ā when we drop it.
- Support for robotic legs - Why change the battery manually when our light can just sprout legs walk over to the charger and change the battery itself?
You joke, but Iād actually like to add sound and an accelerometer. What I have in mind wonāt fit on an attiny thoughā¦ itāll need something more like a raspberry pi zero. Iām hoping to eventually make a lightsaber driver which synthesizes its sounds on the fly, using an accelerometer (among other things) as one of the inputs to the synth engine. The lighting would be tied into the same engine too, so sound and light are generated in sync and both would be personalized for the individual person using it.
Also hoping I can fit a contact mic in there somewhere, attached to the blade, so itāll incorporate the sound of whatever it hits. This method is used on the Korg Wavedrum, and it works surprisingly well. That might be a bit ambitious though.
In the short term, more pins and more space means we can do things like a 7+1+7+1 driver on the next lantern, for adjustable color temperature with more fine-grained brightness control. Or a dual-constant-current driver, like two of what the BLF GT uses. Also, maybe the ability to have more than one switch, have both a flood and throw LED, or more than one aux LED channel, or ā¦ whatever.
The tricky part will be incorporating all that into FSM, and into Anduril, at the same time. Keeping the code and the interfaces clean will probably require some refactoring.
I wish Hank would make it so the tube for the 18350 would fit in the D1S 
AND make a 26350 tube for the d4s
Just got my emisar d4 xpl 3A with Anduril from Hank! Man, this thing is a light weapon. Lol!
Quick question. Iām assuming itās normal that thereās like a little ānotchā in the ramping. Iām assuming itās just as it switches from regulation to FET? Or itās just an anomaly because the firmware is expecting 3 7135 in the d18 as opposed to one 7135 in the D4?
Doesnāt really bother me, just curious.
you nailed it, the āflickerā is the transition from linear regulation via the 7135 to PWM via the FET. Or maybe better put, there is also PWM with the 7135 until it gets to 100% duty. Thatās when you will get the flicker. Higher outputs add PWM via the FET to the 100% duty 7135 current (nominally 0.35A IIRC).
I hope Hank (the real one, not me) will soon be selling a bundle ā flashlight and programmng hardware and instructions.
And that TK will be opening a Flashlight Programming for Dummies school ā¦.
Bummer, muggle is the main reason that I like Anduril over other ramping UIās.
Okay, done playing with it for now. The D4 has pretty much rendered all my other flashlights pointless. Lol! My custom Convoy M2 with FET driver and MGT2 got its ass whooped!
I really dig the candle and lightning modes!
Seems to like my LG HG2 more than my efest purple 18350. not surprising.
I have some pocket clips that work but I think I like it au natural. Haha.
I take that back a tiny bit. The color rendition of my MGT2 is more neutral. And my BLF A6 XPL 3D is more rosy. Thereās a touch of green at low levels with the D4. Totally nit pickyā¦ About on the level of my Astrolux SC and S2 XPL 3B.
The ānotchā in the ramp is only there as a reference point, to indicate that itās roughly halfway up the ramp and has just gone between a regulated mode and a hybrid mode. Itās basically a mile marker.
Speaking of the Anduril D4 thoughā¦ I spent almost all day and half of yesterday trying to tweak the thermal regulation response for it. It runs so hot that itās really difficult to make the light respond correctly. And while testing, I managed to break the hardware ā emitters started dying. So I think Iām done for now, at least until I get some new hardware. Then again, this was intentionally set up as a worst-case scenario for testing purposes. I was using a 219C model with the FET running at 100% power with a full 30Q cell. Hank normally sells those with the power artificially reduced, so my test case was theoretically worse than the real thing.
In any case, the results arenāt perfect but theyāre a lot better than they were before I started. Before, it was overshooting and bouncing pretty badly. It still does now, to some extent, but itās much milder and flattens out after one cycle, maybe one and a half at most.
To the right, the graph is basically just flat with occasional small step-ups to compensate for voltage sag as the cell drains.
Virtually every other light Iāve tested responds really well without bouncing, but the D4 is particularly stubborn because itās so ridiculously overpowered. The 219C version especially so, because the emitter type makes it run extremely hot.