Sorry for the mess.
The idea with the BLF SRK FET v3 was for modding the Q8 DUAL footprint quad LED copper DTP board with a 8x reflector. So you can get everything in one light. Max output and throw with XP-L or XM-L LEDs driven by a FET and for the best Color Nichia LEDs driven by the 7135.
@Tom E
If you are willing to work on this firmware. Please let me somehow know if it is working! I am really interestet in this configuration! In German we would say this is: Die Eierlegende Wollmilchsau. That means one thing that can do everything 
Orsm work guys. :+1:
Double Orsm work guys!!
Has the driver design been sent to TF yet?
If not then we could consider adding the second channel 7135 pads to the driver as a future update for modders and they can flash an updated firmware to use them. The SRK boards are pricer then normal boards from oshpark so it would be nice if the driver already included the 7135 pads even if not the 7135’s themselves.
My drivers are also supposed to be in next week along with the digikey order with the parts to assemble them. I can’t wait to see what you work out for them, got a few e-switch lights I am looking forward to modding.
Although we would need to standardize the pinout between the Q8 driver and the Texas Avenger / Tripledown drivers ideally, there is one pin that is different IIRC, forgot which one off the top of my head.
I can also update the 2s2p and 4s version of the TA drivers with an agreed upon pinout so it would be a drop in option for the Q8.
I like the new LVP setup, simple and too the point.
Only thing I would note is that inevitably once China gets their hands on the design they will use far lower tolerance components then we spec out (hopfully they will at least use the correct values). As such 2.6v seems a bit close to me. Course if it proves to be a problem later people can always reflash the calibration or replace the components with better ones.
P.S. You can always tell a coder but quoting one of their posts and seeing raw code instead of shortcut code lol.
I put a 219C in both a Convoy S2 and S2+ with a FET driver and bypassed both springs. No problems with the 219C’s dying even running them with 30Q’s until they were too hot to hold. My guess would be that in a A6 that there would also be no issues.
Ditto, I am guessing that the MCPCB was either not DTP or not correctly mated to the pill thus breaking the heat path.
I run a triple 219C S2+ with 18+ amps and it has no issues except getting too hot to hold in about 1 minute.
About using 219C’s : yes they survive at least up to 10A, but the maximum output is between 6 and 7A. Direct driving them in the Q8 will run them at well over 7A, so the output will be under maximum and I foresee damage in the longterm.
Of course the mentioned ‘sub-optimisation’ like low-drain batteries, no spring bypasses, thin led wires or current limiting resistors on the driver can bring the current per led back to under 7A.
FYI, I believe the pinout for the TripleDown/TexasAvenger are the same, but they are different from the Q8 driver because it has the FET on Pin6, and Pin3 open for indicator LED
Hi, please add me for one unit
Yep, that was the difference. I copied the tripledown layout in order to allow bistro to work then added the new components.
Now that we know we can run PWM on 3 channels maybe the Q8 can be swapped to the tripledown layout so that the same firmware can be used across all the drivers?
@The Miller
I do not know if the idea is only a brainfart or not. But would you ask Thorfire if they are willing to sell a 8x reflector that fits the DUAL footprint quad LED copper DTP board (not instead of the 4x reflector, as an extra part for extra money).
TA, four times I started typing a reply to your post about major driver overhaul and every time I decided not to post it…
Fifth attempt 
Please read my previous post quoted below.
To add to that, we have received address confirmation a couple of hours ago so when Tom finds the time a V5 Narsil driver prototype FET + 1 is going to be send to Thorfire.
I really don’t get why pins must be allocated the same, I got a set of a clip, USB thingy and a flatcable with 20 individual wires to connect the USB thingy with the clip to flash the MCU
The cables slide on that clip and if I need to connect the USB thingy in another way I could connect the 8 cables to be used different without the use of a tool.
I am no programmer or driver designer and I trust Toykeeper, Pilotdog and Tom to have designed their work in such a way pin and board layout are the way to do it to run their respective firmwares.
Hence after all the discussion we had about the driver, Narsil and tweaking of the driver and its components must be the proper way to do it to run Narsil on the Q8, or at least this is what I trust all those people who have dedicated brainpower, time and testing for this project to have accomplished 
Ramping worked smooth on versions of the prototype made earlier, Narsil worked fine already on those too as a whole.
Now a lot of bugs are worked out.
The 1 - 3 - FET setup is for little more efficient modes between first 7138 on MAX (350mA) and modes with FET compared to a FET+1 config right?
Well the tests Tom did showed that runnning a Q8/prototype (this I call his Q8 driver using quad srk light ) for over an hour on 1700-1800 lumens only drained the cells a mere <0,2V.
So we are looking at the Q8 being able of producing so much light for 6 or 7 hours. This at a level the Manker/Astrolux S41/S14 can do shortly, and much brighter then all other single XML2 / XPL (hi) flashlights can produce at turbo, for such a long time this driver seems efficient enough for me.
I really am sorry you were not around when the driver discussion was taking place, but unless Tom is saying otherwise the last line of my quote (I corrected some typos to make it more clear ) should be clear.
I absolutely love all your input here and in other threads and it is the spirit of the Q8 all can give input and are taken seriously. But not everything can be done and a major possible driver and firmware overhaul a few hours before the prototype that is thoroughly tested and tweaked is send to Thorfire seems not doable
Tom gave this idea of 8 reflector tonfit a +1 and well it seems a pretty cool mod
Yes down the road when the Q8 is actually being sold I will ask for this reflector to be made for BLF.
A 8x reflector for the Q8 will look a bit weird. Because the leds sit more to the middle than a 9x SRK (which also has 8 leds in a circle, but more to the edge), the reflector cups will either be small-ish or have even more overlap than in the 9-led SRK.
Still a nice thought though…
Well, I just read the last 35 posts here, and when I saw the idea about the 8x reflector, my thought was a continuous ring instead of 8 pockets around the center. So, imagine taking a parabolic shaped router bit and cutting out an unbroken ring. Or, maybe the individual pockets centered around the LEDs at the MCPCB, but joining into a continuous ring coming out toward the face. Since there would be so much overlap with 8x pockets anyway, this would look a little nicer, I think.
No, not really, if you abandon the paraboloid shape for the reflectors, the focussing into a visible hotspot will hardly happen, the light becomes extremely floody, closer to a mule than to a reflector light with eight individual paraboloid cups. And I expect weird shapes in the beam too.
With 8x reflectors in such a small space, the overlap will ensure a very floody beam and lots of artifacts already. I mean, we’re talking about XP-L and XM-L2 LEDs here! 
My +1 to this idea was just for the idea. I would not want anything to further delay ThorFire at this point, but discussions are always fine.
Exactly, that why talking to Thorfire about this 8 reflector is not due before the Q8 is actually made and available