Continuing with the upgrades to some lights, Emisar D4 18350 Rechargeable.
I still have some small tweaks to do with this but nearly finished.
Previous attempt.
Nichia 219C
Green glow bezel
Green 6mm x 1.5mm tritium vial in tail cap
Magnetic tail
USB charging unit (1A)
Green Lighted metal side switch
Max 3600lm
I got home from work and batteries fully charged.
Measured 9060 lumens 
But why is my led glowing dimly when the power is off?
Sounds like you have a bleed somewhere.
Could be anything reallyâŚâŚa chip that is barely bridged , or other connection , or a LED that is not properly seated on the board , I am sure others more knowledgeable than me will have more answers though.
Some pics of the light:
Left :9060 lumen XHP70.2 P2 3B 5000K MF02. Right: 7800 lumen XHP70.2 P2 5C 4000K SMO reflector L6
Both on turbo just setted the exposure low.
CRX that D4 is icredible ! such an awesome work on implementing charging !
Nice work Zozz ! I like xhp70.2 a lot too , especially in throwers 
I am no trying to start an all out off topic discussion (although it is on the modding track still) but have you tried the XHP 50.2 side by side in the same kind of lights , or changed out the 70.2 with the 50.2 in the SAME light to see what the throw looks like with both?
It may not be so for many lights and could be that it just worked better in the one particular light I have tried this with , but hereâs my storyâŚâŚâŚ
I have one of the Maxtoch 2X that came stock with the 21700 battery tube and 70.2 emitter in it , and although it didnât have the classic âdoughnut holeâ , it did have a slightly darker center to the beam instead of the other way around , with the center being the brighter hotspot.
I just decided one day to see what the XHP 50.2 would do in the same light so I flowed the emitter on an extra 32 mm Maxtoch MCPCB I had at the time. I DID NOT change anything else about the light and both emitters were as they came , with the dome ON.
I then put the light back together and checked the lumens and had lost approximately 600 lumens , so at first I was kind of let down. The light was making an average of 4500 lumens with the 70.2 in it (in Turbo Mode) , but the 50.2 was only making around 3900.
I took the light outside once it got dark and turned it on , just to see what the beam looked like and I was pleasantly surprised. Although I had 600 lumens less , the beam was tighter and had a lot more throw to it than the 70.2 did , but the 50.2 still has plenty of spill with it also. It seems kind of weird for the 50.2 with less lumens to look brighter than the 70.2 to my eyes , but I guess it is just the more focused beam that makes it appear that way. The light downrange is not even close from the 70.2 and once you get out past 125 yards or so , the 50.2 is still fairly bright at 400 plus yards , where the 70.2 in the same light was fading fast at the 175 to 200 yard or so mark , in comparison to what it looked like at 125 yards on in.
I need to revisit this simple mod with some dedomed , or sliced emitters and see what the result is with the 70.2 and 50.2 in the same tint and then put in the same light. The original 70.2 was in the 6500K to 7000K tint range while the 50.2 I put in it is a 5700K.
Here are the before and after beamshots of what it did to the beam and throwâŚâŚâŚâŚâŚ.
These shots were taken in different seasons , but from the same distance. If you look at JUST the tree in the 175 yard picture , you will notice how much brighter it is with the 50.2 emitter , as well as at the gate , with the 50.2 being brighter on the gate itself.
Original 70.2 ââ 100 yards
50.2 ââ 100 yards
Original 70.2 ââ- 175 yards
50.2 âââ175 yards
For comparison of the light on target and hotspot size , here is another 2X with a sliced SST40 emitter at the same tree at 175 yardsâŚâŚ.
Measured the MF02 XHP70.2
You already know the 9060 lumens and I got 226125 cd which is good for 951m throw. This is enough for me so I leave the dome on the led. I really like the beam!
Remove the dome and you will like the beam even more. Crees love it.
Very interesting comparison right there robo !
There is more light at the center clearly , which many times is nice to have âŚ
Clocks.
I needed a clock. Banggood had some neat clock kits. Soldering each pixel individually was a bit of a pain, but now that everything is assembled, it looks nice.
The dot-matrix one had 6 pads per LED, and 105 LEDs, or 630 total pads to solder on the front side. It took a while. The other one had 216 pads on the front, IIRC, which also wasnât the fastest process ever.
Stock photos for context:
DIY a nixie tube clock
Where you get it from, steel? I see many ussr-made components.
Today I got the Q8 MCPCB from led4power so I reflowed four of my favorite XM-L2 T3 7A3 CRI90 emitters in a my Q8:
â
I used a washer with two cut outs to clamp reflector to MCPCB :
The holes in the reflector was reamed out to fit this XM-L centering ring :
It was as reasonably straight forward mod, without any difficulties.
When I fired up the Q8 it wasnât the nice tint I expected instead the beam had a green hue completely unlike the tint of the XM-L2 T3 7A3 CRI90 in the other flashlights I put it in.
Convoy M2 XM-L2 T3 7A3 CRI90 (left) BLF Q8 âXM-L2 T3 7A3 CRI90â (right)
One of the emitters have the ânormalâ 7A3 tint the other three have a green tone.
What a huge disappointment, and the next problem will be how to explain to the seller I got three emitters with different color :weary:
That is a bummer khas, those emitters are supposed to be super-nice! 
One thing to check with your Q8, even though the tint difference between leds points to a led problem: is the lens AR-coated? That coated Q8 lens totally spoiled the tint of the LH351D leds in one of my Q8 lights, the non-AR coated lens of another Q8 made a world of a difference.
Ooooh, where??
I like nixies!
I have also tried it without the lens no difference, I donât know if itâs worth the hassle to try to get a refund
Yeah, I ordered some 2000K car-bulbs just to scrounge the LEDs from them⌠and didnât even to think to light âem up before wanting to use them.
Evil cold-white LEDs (probably 7000K), not 2000K.
Missed the pissânâmoan date on AX by about 3 days, so couldnât even grex about them, pissânâmoan to the seller, nothing.
@khas, ask for a refund. This looks like a bad U6 4C XM-L emitter.
@djozz, was it the original Q8 AR lens that some people received or the Kaidomain AR lens?
Was the seller KD? The ones I got from them were good sofar.
I referred to the stock AR lens that part of the Q8âs came with. I did order a KD replacement lens because it was said that it did not alter the tint as much but I did not receive it yet.