Like this maybe?
http://www.harborfreight.com/diamond-rotary-bit-set-4-pc-69664.html
Aweome!! I didn’t know those even existed… :person_facepalming:
Those flat ended bits are useful too but better with the round ones for your purpose.
Is diamond or tungsten carbide better?
They might be too aggressive, i haven’t used them, get both.
Diamond will be easier to keep steady, a tungsten carbide might want to grab and pull and leave you with a big gouge across the brass. Ask me how I know… 
On the TM16, I have two sets of readings… both after putting in the XHP-35’s. First stock driver, second bridged sense resistor…
Stock………………….Bridged
05.556 lumens………05.4855 lumens
311.535………………484.38
707.25………………1086.75
1966.5………………2939.40
4943.85……………7072.50
So the moon is controlled separately from the sense resistor and bridging the sense resistor bumps power across the board pretty evenly. Interesting, without that separate moon mode the lowest mode would be way high. As it is, being a Tiny Monster, this is livable. Would be nice to have something in the 100 lumens area but hey, it’s a Tiny Monster Quad, right?
Nice! I’m thinking about 4x XHP35 HI… can you take a picture of the positioning of the MCPCB and how did you cut the 26mm maxtoch? Thanks again
How do you like the beam compared to your L6 with the P-bin emitter? Should be pretty similar specs on the top end?
Interesting concept… The L6 with XHP-70 has 4 dies close together but not touching, in one big reflector. The TM16 now has 4 of the XHP-35, each emitter is 4 dies touching so there are 16 dies in 4 reflector’s. But both are making a bit over 7000 lumens. One is pushed to the max, the other is coasting. (the L6 is pulling right at 12A, for effectively 6A per die [2S2P 6V configuration], while the TM16 has about 1.10A to each emitter [based on the Cree Product Characterization Tool and the total lumens output divided by 4], and since the 4 dies of the 35 are in series then each die is seeing 1.10A)
If I have that correct (which would be a small miracle in and of itself) then the TM16 is coasting and should last a long time.
Yeah!
TurboS stock: 9350lm@5sec 9050lm@30sec
Modded with M2 3000K from Kaidomain: 7660lm@5sec, 7360lm@30sec
So 19% loss from going to warm white.
Btw, the shelf under the ledboard is Q8 thickness, and so are the ledboard traces :sunglasses:
The rings around the leds are not center pieces but pieces of foam sticking to the board to seat the reflector in.
The leds are in series.
Here’s a boring 10 minute video of the reflow:
Well done!
I’m curious, would the warm white emitter cause the X7 to have more, less, or similar runtime compared to CW?
Must be the nicest way to “lose” ~1700 lm I’ve ever seen, Djozz!
I would be just sick if I did a bunch of work and wound up with 1700 less lumens and an ugly orange beam color. :confounded:
I broke into the TM16 just to get rid of a color like that. They glued the heck out of it too! (Olight probably did on that one too, huh?) But we all have a different perspective on what constitutes perfection, which is what makes this place totally rock! 
I knew you were not going to like this Dale
I love this tint since the first time I set a forest on ‘fire’ with a S8 with floody 2000lm warm XHP50.
@unknown00101 : I would not expect a different runtime, AFAIK the tint difference is just made by the different phosfor layer on the same led.
Naaaa, it’s all right Jos, that’s an all right color, reminds me of the headlights in my ol ’51 Chevrolet Custom DeLuxe.
As a matter of fact, those old headlights were 6V too….
Nice for one two reasons djozz. One it worked when you were done and secondly you didn’t turn the MCPCB black. 
I’m digging the warmer tints, especially in combination with that kind of output. But I have to admit, I have no light with such kind of output. 3000-4000 lm is where it ends right now. I still restrict myself to single cell lights…
There’s no glue on the bezel of the X7, I checked that the first day I received mine, lol.