Yes, you can simply remove the retaining ring and tilt the board, flash it, re-assemble and you’re good to go. 
I think you reuse your end cap because you don’t get a new one. One set gets you 2 tubes and 2 carriers only.
i maybe missing something here but the Q8 with 8 cells running a 4s2p config, would make sense, to be using the xhp 35,50.2 and/or 70.2’s at 12v? even on turbo you’d have some kind of run time out if it…
e.g. xhp 35 running at 2 amps booste even, you could double up easily enough…
it would even be in the realms of an actual camping light that you could use to camp out for more than 20 minutes? maybe even the weekend.
xhp70.2 craving a 3 amp booste? 6 amp of capacity?
don’t get me wrong here, having some wampum with a hand torch is fun, but having that wampum for a good two hours, or a much better camp site torch for 6-18 hours….hell of a sales pitch.(add in a ramp ui and wu-hey you iz the popular guy with a very useful torch lol)…ok not as small as a C8…but.
(the watts turn out the same at 6 or 12v…and on learning amps burn hotter than volts…etc etc).
It wouldn’t make much sense with a FET driver to go the route of quad die LEDs.
Using a buck driver on the other hand would be excellent.
would you need a fet driver with a 14v supply, drawing less than 4 amps too. 
i noticed in the cree charts (correct me if i’m wrong), the higher up the scale (xhp 70.2) you go the less amps to equal the lesser LED’s luminum out put… all be it less throw, but, cranking out 4000 lum’s kinda makes up for a little dot a little further? eeeh horses for courses i suppose.
i’m also thinking lesser efficiencies in the drivers the farther the output voltage from the input voltages wouldn’t matter as much when you get more than double the run time for same light outputs… i’d take the hit on a 75% efficiency driver for twice the run time.
once you go past 2 cell flash-lights, the ‘compact’ compartment in the maths sort of goes out the window, and now we’re dealing with a bit more bulk longer run times is the new kid on the block. (else why carry that weight, baring those humongous throwers you see on you tube…sheesh sides of hills an all lol for nearly 15 minutes).
i havent lamped with a car battery in a number of years, i dont think i ever will again! eeeek…oh the kneck n back pains after 5 mile….6 mile……7 mile……. still got to get back home yet. ohh for a C8 back then :sunglasses:
see where i’m going with this one…i dont know all the details though.
I got the Thorfire BLF Q8 today and of course, the first thing I did was find a dark spot to test it.
A bit more in the yellow range then I thought it would. Not a problem.
A bit heavier then I thought it would be. No problem.
Pretty much met my high expectations and kudos to the forum for doing a great job on it.
One, just one tiny thing I would of done a bit different……I would of sunk the switch a bit. Just a bit. I had in in the glove compartment and as it was rolling around, I thought that if it hit the car manual just so, it may turn itself on, No big deal because I normally unscrew the back a bit to make sure there is no draw.
Question……why do i have to screw it all the way in order for it to work? The batteries are protected, which makes them a bit longer and it takes a bit of effort to screw it all the way. Is that normal?
Yes, perfectly normal for it be screwed all the way.
Yep, protected cells work but will be tight, always had been, even in the prototypes/pre-production ones. You can do the 4 clicks to lock it out as well. I'm not a fan of using protected cells in this light, but I can understand why some people want to.
Thanks guys. TomE……they are the first 18650 I bought. Will move on to IMR and the very high quality ones as soon I expand the collection.
A bit of a problem……now that I have 5000 lumens that make woman swoon at my manly lumen power…….where is the 20,000 lumen Q9? 
Well, knowing the BLF Q8 draws about 50W on max power, the Q8 2.0 outputting 20k lumens would have to draw 200W at minimum.
Unless you have active cooling, a vapor chamber, and some phase change material, you’re not going to be able to sustain even close to 20 000 lumens.
Still, if you want more lumens from the BLF Q8, here are some tips.
Get this AR lens: http://kaidomain.com/S027343-55mm-x-3mm-Multi-Layer-AR-Coated-Lens
Get these brass screws: https://www.ebay.com/itm/M2-M2-5-M3-Brass-Cross-Screws-Phillips-Pan-Round-Head-Screw/292442788167
Bypass the springs with some 22AWG/20AWG wire.
And get some of these nice 30Qs:
You’ll see a nice boost of performance all around.
Ahhh, been there, done that! The MT03 TA or MT09R (think?) is right there at 20K. I got one MT03 TA, and modded up another MT03 myself - both doing over 20K. This one I modded is somewhere over 30K:
On the right:
You know that it was a bit of tongue in cheek…right? Woman are not really swooning over my lumens!
The Q8 serve me fine for needs I can only conjure, never mind expect. The next light is a EDC or two. And of course 3-4 “safety lights’ if that project gets off the ground. Otherwise it will be TN4a’s,
The link on the Samsungs…….are they the best choice and is that the cheapest place I can find them?
Put me down for 1 set for the BLF Q8. Thanks.
OR, over 11,000 lumens by using LED4Power’s quad MCPCB and 4 SST-40’s, change the MOSFET to an 404, use the doubled Blue springs, voila! Even more output from Samsung 25S cells, even without changing to the UCLp lens.
But of course, the Q9 is feasible… 25,960 lumens
What is “sustainable”? I’ve been using this on a light stand firing into an umbrella to light portrait shoots, running it 20-30 minutes at a time.
First…I know less then I should about building flashlights….so please keep it in mind.
If one wants to build a custom, ultra high power flashlight, why not start with an ultra powerful led? The huge Crees like the cbx3590 come to mind.
Voltage, common battery configs, simple drivers, cheap optics.
Dale’s build uses all the same components as hundreds of other lights but with a lot of thought put into the arrangement and heat sinking.
That led is basically a bunch of smaller leds wired internally in a combination of series and parallel so that it needs 36 volts to operate.
Dale took a bunch of individual leds and wired them to run at a lower voltage (total 3 volts I think). So now he can use regular drivers to power it, like contactcr said.
Besides, what type of reflector or optic could you use on the cbx3590 to get any type of decent beam pattern? Dale’s method just makes more sense.
I saw a video of someone who took a huge thrower and remachined the body to take a huge led. I believe it was the led i mentioned. That is what led me to the question.
The given responses is what i thought the logical reasoning “may be” but given that i never build any flaslights, i thought that it may also be the cost of a single, large led. Or the higher 36/72 voltage.
When it comes to flashlights….best is to ask and buy.
I believe that was AdventureSportFlashlights. Impressive for sure but overall a lot of work for something a several other factory lights can replicate.
Look at something like this (same flashlight host) for perspective:
Yep, that was Matt at ASF who modded a BLF GT with the gbx3590. Voltage is a big deal. Luckily the GT uses carriers so Matt was able to mod them to provide about 30v, then the boost driver bumps that to over 40v. The biggest problem with that light is the lack of a driver user interface. He has to use a knob to adjust the brightness.
Another thing we don’t know is how much above spec can that led go? It’s supposed to be 3.6A, but can you push it more? Matt’s driver is limited to 3.2A, so we can’t really push it very hard.
On Dale’s light we have a lot of leds that we know can be pushed really hard, even double what they are rated and his more common low voltage drivers can push lots of amps. So he can really hot rod his light. Matt can not hot rod his.
Both are cool, though. Dale built his with the intention of using the extra battery carriers. Then he can push his light even harder. He might see double the lumens of the gbx3590 light.