Thanks, Miller. I was unsure if protected cells would react differently than unprotected when reversed.
I guess they can’t tell which way they’re facing either. 
I already do my best to ‘scare them straight’ when explaining lithium-ion safety to my friends.
Size comparison
Bottom is 0.5mm normal solder.
Each stick will do a spring with a bit to spare.
K, reviewed my notes. Are you sure you are thinking of domed XPL's and not de-domed? I've gotten ~1,800 lumens from a dedomed XPL V6 at turn on in the past. Also I had a modded SupFire M6 triple XM-L2 U3 that did about 5,200 at turn on, which is just over 1,700 per LED. I think the XPL V6's along with the really nice thick 4X MCPCB and the better modded up tail PCB, better integrated driver with 18 AWG wiring, etc. are all improvements in what I had in that modded M6.
So per LED, my best modded Q8 is 7,210 at turn on, 6,590 at 30 secs and the worse is 6,940 at turn on, 6,360 at 30 secs and that's out of 6 Q8's modded total, all on 30Q BT's from Banggood.
I did all the normal stuff to mine, tailspring bypasses, cleaned the screw holes, drilled and chamfered the driver mounting holes, etc. All that definitely improves output. I don’t have numbers to back up my claim, but I think we all know what these mods do. Its a helluva little light after just a bit of work. Here’s some pics of my modding, but its nothing you all haven’t seen before. Most of you have probably done it yourself at this point:
I did another thing though, and not entirely on purpose. See, I had some of those really rounded out screws in mine. And in the process of trying to remove my tailspring board I was unable to remove one the traditional way - with a screwdriver. I even tried to cut a slit in it and turn it out with a flatblade screwdriver, but even using very good tools I was unable to do anything except completely destroy the head of the screw. I’m pretty well convinced some of that “powder” in the screwholes is some sort of weird threadlocker/glue. Anyway, I removed the head from the screw entirely so I could get the board out, but that left me with a problem.
I tried to get a grip on that little nub of butter using some needlenose vice grips, but it just wasn’t happening. Since I didn’t have a 2.5mm tap in my set, I decided to use a different technique instead of drilling it out, something I’d done before on other projects — I tried to dissolve the steel screw in boiling alum.
I knew the aluminum wouldn’t react strongly with the alum so I thought this was a pretty good idea. What I didn’t count on was the ano being pretty unhappy about this treatment. 
I was able to remove my screw and replace all the fasteners, but I wound up stripping the ano off the entire battery tube and tail of my light after the coating was damaged by the alum.
I also stripped the ano from the little switch retaining ring, just to set it off a bit more. Looks pretty nice with the pink/blue LEDs I think.
Given the choice I’d probably strip the entire light, but since the reflector screw head inside my light is basically an empty cone of butter I’m not going to. I’m kinda over fighting with screws on this light. 
Besides, I think it looks nice enough just as it is.
I’ve ordered another light and hope it comes with better screws. I’ll probably strip the ano from the head of the new light and swap the two, leaving me with one solid black light and one solid stripped light. Eventually. But for now I’m just enjoying my Q8. These are really awesome little lights.
That looks real neat emarkd. It does need the rest doing to set it right of. 
So I posted in the Q8 thread a few days back when I received my Q8. Out of the box with well worn 2900mAh solder top Panasonic laptop pulls (grey ones) at 4.05V, I measured 4701 lm in my JoshK sphere.
Finally, after this busy work week I was able to get my Q8 on the bench and give it the proper BLF style test and tune session.
I started with the head to try and fix the tilted driver issue. I found mine to be pretty dirty and the screw holes were undersized, possibly half the issue with the tilted driver. Oh btw, all the screws in my Q8 were perfect. Actually the whole light was nearly perfect. I have absolutely no complaints with mine. This is just a tear it down and see what its all about because I can and I want to post.
For whatever reason I thought the sir800 FET was going to be used but found the old reliable 30L. I'll go ahead and swap that while im in here. Nothing wrong with the 30L but I have a 404 so why not.
This positive contact ring looks kinda rough so I decided to work on it a little bit. The screw holes were eventually opened up with a #30 bit.
It turned out to be sorta dished in. I started flattening it out with a DMT lapping plate and its very apparent that its not flat.
Here it is after flattening.
Not quite as rough now. 
Once I finished with the driver I moved on to the driver shelf and found the material around the screw holes to be mashed out and protruding above the plane of the actual driver shelf. This is the main cause of the tilted driver in my Q8. When I dragged my fingernail across the screw holes I could feel the material protruding up. My calibrated fingernail says it was probably .005" to .010" above the shelf. It was easily wacked down and smoothed out with that little rotary file end mill chucked up in my cordless drill.
I didn't get a before pic but you can see the shiny metal where the protruding material once was. My driver now sits flat and contacts the battery tube evenly all the way around. Notice the reflector retaining screw head is unmutilated. It was tight but a #2 ACR APEX bit does the trick every time.
moving on
The LED spacers are fine. I sanded .010" off mine to make the reflector sit closer to the MCPCB. I didn't take a LUX reading before I did this but afterwards my hotspot is definitely tighter and more defined.
I use PSA sandpaper stuck to CD's for sanding LED spacers. 320 grit works good...
The LED's look excellent, and so does the soldering and screws... Nothing to do here. I thought about dedoming them but.. NAH I like it the way is.
I have no idea how these 2 pictures became fused into one but neither one is very good. This is the driver end of the battery tube. I used my lapping plate to true up the contact area.
And moving on to the tail PCB. Mine was dirty but other than that there was not a thing wrong with it till I started messin with it. (see the right side where its in the vise. YEAH. These boards will deform when HOT. Easy fix though) The IMO mushy springs are getting replaced.
These are the stiff springs that RMM sells for 2s setups. I try to use them for all my builds for the added contact pressure. I dont care what the resistance is cause I always bypass them. In this case I used 20 AWG for the bypasses. IMO the added contact pressure is what matters most. It also cures the tail bump, reset situation. When I finished I slammed mine down on the rubber mat on my table as hard as I felt I could afford and it never missed a beat.
I scratched off some of the mask for a solder pad next to the screw hole. Path of least resistance...
Here it is finished. All cleaned up and straightened out. I completely filled in the screw holes with solder to the point that it was raised up higher than the rest of the board. I re drilled the holes and used a file to get the solder blobs flat. These are the contact points to the battery tube and IMO should be better than they were. The screws make no difference on the resistance of this contact. Whether its steel or brass or wood. If they're able to be torqued down then they're fine. Contact pressure is what matters here.
I applied Motorcraft XG-12 (NYOGEL 760G) grease to these contacts before reassembly. Corrosion X was applied to the battery tube screw holes to prevent any corrosion that may occur. Motorcraft XG-12 was also applied to the copper contacts on the springs, the positive contact ring on the driver and where the battery tube contacts the driver and of course all the threads. It WORKS. Just do it!
Now for the numbers...
Out of the box it made 4,701 lumens with old solder blobbed laptop pulls at 4.05V. Afterwards, with the same batteries at 4.09V it made 5,561 lumens at turn-on and 5,440 at 30 seconds. Not bad at all!
Enter SAMSUNG INR18650 30Q BT from bangood. I charged these and ran them down to 3.0V then recharged them again before the test. At turn-on it made 7,105 lumens and 6,716 at 30 seconds in.
This light is officially the brightest flashlight I own and it is AWSOME!
If I remember right highest current measured was about 22 amps, so 5.5 amps per led. Djozz measured 1900 lumens with this current. There must be very little losses to achieve 1800 lumens out of the head per LED, the Q8 doesn’t even have a coated lens. I’m very interested in knowing the current flowing in this high lumens light.
Quite some pages back someone asked is the switch could be made to “breathe”. I would just like to say this is an option I would like also. My version would be for a breath then a pause for about 20 seconds (or configurable), then another breath. With the processor in deep sleep in between to reduce parasitic drain. Hopefully someone implements this into the code so that I don’t have to but if not I will see if its possible for me to HACK it in.
Cheers,
Jayson
Wow, really nice modding emarkd and RotorHead64!
RotorHead64 - awesome work there, good to see output confirmation in the same realm as I'm getting. I have those same springs, though the old IOS "A" seem to be a perfect fit, but can't buy them any longer. I don't know what a DMT lapping plate is, or what that filing mill bit is, but I do know I want them! 
Did you do throw measurement? Interested how it does compare to mine with the filing of the centering rings.
For amps, I'm guessing in the 22-24A range. I think these mods squeak out a little more than what I've gotten before, based on measured lumens. Of course it's difficult to measure amps though with this tail PCB modding - you would have to rig up some sort of fixture, then the fixture will introduce some loss's.
Guess you could design a low resistance fixture and compare measured lumens against the modded light, and if in the same ballpark, then assume the amps are the same.
I must have missed that these are XP-L2 emitters. These (or XP-L W2) will indeed do the numbers, even a bit more. I just thought it was getting XP-L V6 3D. (an ad from Banggood says XP-L emitters)
emarkd, nice work, great results!
rotorhead64, also nice work, excellent finish on that brass ring! 
Seems pretty clear to me by the output numbers what emitter is in the Q8, but can someone give a look at the emitters and confirm this, as identified by the picture below?
The XP-L W2 2B is on the left, note the smaller flats on the side of the dome as compared to the XP-L V6 1A on the right. (both of these were supposed to be the new XP-L2, the second order from Cutter came as what you see on the right)
This smaller flat on the side of the new XP-L2 allows it to JUST fit under Carclo triple and quad optics. Older XP-L emitters had to be shaved or de-domed to fit these optics.
Dale - definitely they are listed and pretty certain they are XP-L V6 3D's. The tint seems to be right, and compared a bunch of Q8's, and seem to match well unit to unit.
Wow, that bare tube looks awesome! :+1:
I have never stripped anodizing but your Q8 is making me consider it. Could you elaborate on your ‘boiling in alum’ process?
How long did the initial boiling take? What is your mixture’s ratio? Does the ano simply fall off or is there scrubbing involved? Did you do any polishing afterward to get your final finish? Is there a way to ‘mask’ the threads to still allow lockout by loosening the head?
DMT makes diamond whetstones for sharpening and lapping tools. The lapping plate is the FLATTEST and most expensive, and would only be necessary for the largest and most precision jobs. I have a couple of the smaller ones that would be quite sufficient for lapping flashlight parts and one that is big enough to lapp bench planes, which I think is a 10 inch. Look around in Woodworking stores and shops, they are often cheaper there.
I measured 7K after a spring bypass. Back when I reported it, I assumed I must not have had my light tube quite dialed in. Now, I'm seeing that others are measuring the same. So I'm hoping my reading was accurate. I felt it was ballpark accurate before my 7K measurement.
DEL, I just noticed at Q8 modding - #739 by RotorHead64 RotorHead64’s photo of the Q8 production driver, where the three resistors are now labelled R1, R2, R3 and the old pads for previous R1, R2 are no longer present (boo hoo, I had an idea in mind for one of them).
Perhaps you could update your schematic to reflect the silk-screen on the production driver ?
Edit, please ignore this nonsense above (how do I make “strikethrough” work ?)
I still see empty pads for R1 and R2 there in those pics? R2, R3, and R4 are in a row, R1 is separate.
-I don’t see any unpopulated pads ? and I read the row of resistors as R1, R3, R2, though the “R1” could actually be R4 I suppose, though it doesn’t look like it to me ?.
Zoomed in, on this image: Imgur: The magic of the Internet
Edit, please ignore this nonsense above (how do I make “strikethrough” work ?)