I wonder if international outdoors or mountain electronics is going to have a driver/firmware to support a clicky version of the sandwich mcpcb.
I think these marker LEDs are a great idea, if well done with good efficiency (TK, are you using the trick with the internal pull up, plus an external resistor, to get the various levels from one pin ?).
Tritium vials are neat, but have a 12 year half life, even if the internal phosphor survives that long. Now 12 years might seem like an age, but it’s not. And you’ll notice the dimming long before then.
I have a small collection of British prismatic marching compasses, the most recent ones were surplus stock, sold off because the tritium vials had degraded and it was no longer cost effective to have them re-tritted, as used to be the practice, or they just didn’t need so many anymore.
These things: PRISMATIC COMPASS M-73
My most historic one, from WW1, was given to me by my grandfather and still has it’s radium paint, still glowing, but I keep it very carefully, and it doesn’t come out for use.
Similarly my father’s WW2 Negretti & Zambra pocket compass.
I really like the flashing through VIAs solution. I hope BLF driver designers will copy this idea in future designs. I really like to see fet+1x7135 drivers for E-switch lights like this and easy swap firmwares.
Yes. Make it a standard footprint please, then we can have a (3D printed ?) pogo pin adaptor to replace taking the thing apart, cobbling up a power source, and so much fannying about with dodgy clips, individual wires that have to be connected up the right (mysterious) way, etc.
MCUs are not going to have leads soon for clip attach, at least not the ones worth considering. True leadless SMD, BGA, …
We got past this years ago with I2C/SPI debugging. I’m a bit out of touch these days, but I’m sure there has been progress.
Done properly it could be liberating for those who are intimidated by the mystique of code development, nevermind simple re-flashing.
Exactly that is what I’m thinked about! 
Make the flashing pads position to standardized distance and orientation from driver center and I can 3D print a 18650 bar style pogopin holder and with a 26650 adapter tube it can put in also the D4S. the one thing it needs a hole or some orientation locking thing what prewents the flashing adapter to rotate where the pins doesn’t hit the pads. I can make the 3D printed part with pogopins and sell them completely built if it is a request later 
I can make the pins replaceble too.
I’m not queueing to get one largely because it is too large for my EDC while not matching any of my non-EDC uses well enough.
I find the aux LEDs a very interesting idea. It’s been mentioned before, but I haven’t seen it implemented before. Would like to see it….I guess I’ll have to wait.
Can somebody make something similar PCB design like this AUX led board for a D4? Lexel maybe? the pin is on the driver. just need two extra wires and an Andúril firmware.
Hahah yeah.
Still an EDC for a LEO in a belt holster is not quite the same as what I would consider an EDC in the pocket of my slacks.
I have a DQG Tiny IV 26650, which is smaller than the Emisar D4S, and I would not consider that DQG viable for pocket-EDC. It’s way too heavy and bulky.
I’d like to see a new aluminum version of the D4 that incorporates some of the improvements in the D4S, such as:
- recessed button
- square knurling on battery tube
- vias for reflashing driver.
I would also like to see square knurling on the d4. Much much prefer some form of knurling with smooth ano
Now I want those improvements on a D4v2…
Orange secondary LEDs would give a BOSS style on this budget light !
YES! D4v2 please! 
Excellent . I really appreciate the comparison at the end of the review.
Count me in, if D4V2 happens. I prefer 18650’s.
After thinking about this light some more, I have to agree that 20700 would be the better battery format since basically none of the big battery companies make a 26650 that I can see. But I guess it should be possible to adapt from 20700 to 26650 size slot to use one in a D4S. Might have to find someone to print an adapter though, I’m not finding anything commercially available.
Edit I’m an idiot, 20700 is longer. Damn.
Some of the protected 26650 cells are 700mm long. I doubt this light was designed to work well with protected cells but I wonder if some thought was put into accommodating some of the larger/longer cells and if they could work with a 21700 or 20700. The D4 doesn’t seem to have extra long springs, but perhaps this one will…
Nice detailed review TK.
This sounds like a first class light. I am looking forward to it. I like the idea of the "SAFE" blue LEDs.
You could simply glue the LED (s) to the MCPCB or solder one end of a LED to a power pad at the MCPCB as I did in my Green Lantern.
Very promising indeed. I’m in for one for sure.
The ~20 years runtime on aux LEDs (low) is already the dealbreaker for me…
Is there really a new version of the DQG avaible? I have only seen V3 up to now.
I also consider the D4s too big compared to the DQG, and would have preferred a 7 LED variant. Now I have to use the DQG for my planned RGBA + 3x white build - which is way more difficult due to the serial LED setup and the driver design.
It’s probably worth noting that the li-ion cell’s self-discharge rate is faster than that, so it probably won’t actually last 20 years per charge. But if the driver and LEDs were the only source of drain, it looks like it should last a couple decades based on how low the power draw is.
Effectively, there is no difference in terms of battery life between the low aux LED mode and off.