Nice job! That’s probably actually a preferable method for Nichia LEDs since they tend to have a slightly lower optical reference. What tool did you use to pop off the bezel?
Used this video as a guide with a utility knife, regular knife, and then a flat head screwdriver, tapping the head of the screwdriver on the back of the knives to slowly wedge the blades in. It was a lot easier than I had expected.
A knoife ? :-p
Thanks for the answers Lightbringer and CRX! I forgot about the pressure the battery may do. Will have to think it well!! Thanks again 
FourSevens Atom XHP50.2 3v K2 2A - H17-F - 16340 - RV Clicky Tail Sw - 2500lm.
That’s awesome. Do you have any idea what the turbo current is on the S1R? I am planning on swapping the emitter to the SST-20 3000K or 4000K, maybe to a Luxeon V2 4000K if the output suffers too much from the high CRI.
Reylight LAN Mokume Gane with electronic tail switch, D4v2 ramping firmware driver, emitter swap, extra copper heatsinking and all tritium vials installed.
LAN Mokume Gane XP-L HI v2 5000K - FET 15+1 D4v2 - 14500 - Electronic Tail Sw - 1300lm.
Verry nice! I like the red trits also, are they visible enough compared to ice blue or green?
Yeah I thought red would go nice with the light.
Might just be my eyes but in the dark the red trits are almost as bright as the green vials, really noticable to me.
Other colours are a lot dimmer.
e-switch in the tail? What, how? Do you have any details?
There is always more details in the links 
The red tritium vials are said by the manufacturer to be about 20% of the visibility in brightness of the green ones. I’m sure the observer’s particular eyes are going to perceive things differently sometimes but the normal list is like this…
mb-microtec in Switzerland has been making tritium vials for 50 years, they say the red is the lowest visibility with the green the highest, the above chart is taken from the mixglo site but I think it may have originally come from mb-microtec. I do see differences but they don’t seem as exaggerated as the chart would indicate.
Did you know that Tritium is also called Heavy Hydrogen? or H3? And that it’s unable to penetrate human skin? Or that it breaks down (decays) into Helium? In a watch with 14 Tritium vials, there is not enough radioactive material present to harm you any more than eating banana’s! (That’s IF you could ingest all the Tritium from the vials at once which is virtually impossible) The potassium in banana’s decays into radiation, so if you eat banana’s every other day you will have a low radioactive glow about you. 
The typical Tritium vials we see contain 25 millicuries of the radioactive isotope of Hydrogen. There are also versions in watches that have 100 millicuries, listed as T100 while the lower is T25. The T100 is said to be 4 times brighter and glow twice as long, up to 50 years…
I have all 9 of the colors listed installed in my Titanium X6. With 2 Green as separators.
Thnx! That was a interesting read.
I put a 219B in a Zebralight H53Fw. Tons of photos and measurements here: Zebralight H53Fw Emitter Swap and Review - Album on Imgur
Not totally sure, someone more knowledgeable could probably chime in. My best guess is somewhere between 3-5A. The Olight 16350 the S1R II comes with is a 550mAh battery rated for a 10C discharge, so peak output for the battery should be around 5A at full charge. The output isn’t 1000 lumens with the LH351D, but it’s not terribly far off judging by eye. I have no real means to measure current or output.
Yes on the D4 bosrd I use pads that are almost as small as regular 0402 pads, bigger does not fit, very tight fit to get all those LEDs on the tiny useable surface
Thanks! I’m hoping to get at least 800 lumens on turbo with the LH351D 4000K 90CRI. I ordered 14x1 mm mcpcb’s from AliExpress so then I’ll give it a go.
I fitted my NiteCore TIP Cu with a LH351D and it instantly transformed the light.
Cheers
Put a White Flat in a Cree Q5 version of the Jacob A60 for my buddy. Also fitted an LD4A set to 6A, used 22ga leads. Before I put the driver in I used a bench test LED to set the driver to 5.5A. After assembly I found that I could reset the driver to 5.75A for 5.45A at the tail, with the White Flat making 828 lumens. On the lux meter, at 5M distance, 800Kcd. 
Works for both of us! lol
Very nice mod! Whats the output like after the mod? Will you put a clear lens on it afterwards?
I received my first Emisar yesterday, a D1S. Phenomenal light! But it probably had less than a minute on the stock electronics before it was modded
Custom 5A linear driver running 5x CN5710’s (in dual channel: 1+4) and a Attiny412. Has programming pads for easy flashing. Running TomE/TK’s RampingIOS. Also put in a 1mm² White Flat from Hank with his gasket.
Can you tell I’m a sucker for symmetry?
I did a few before & after numbers, along with my B158 and L2 for comparison.
| Flashlight | Volts | Amps | Lumens | Lux |
| D1S, stock XP-L HI and FET+1 driver | 4.075V | 5.33A | 1540lm | 127,425lx |
| D1S, 5A driver and White Flat | 4.079V | 4.98A | 958lm | 272,880lx |
| B158, FET+1 and XP-L HI | 4.066V | 4.78A | 461lm | 185,925lx |
| L2, FET+1 and dedomed SST-40 | 4.188V | 9.48A | 1946lm | 256,250lx |
Observations / Notes:
- The beam, while a bit cool from the White Flat, is tight and beautiful.
- I didn’t play with the focus. It’s possible it could be better with some adjustments, but I’m pretty pleased.
- I knew the optics of the B158 would have quite a bit of loss, but I wasn’t imagining anything that drastic. It could be my lumen tube reading low, but the other measurements seem about right
- These CN5710’s don’t like very high frequency PWM. Spec is max 2kHz, which is what I have it running at. I tried them briefly at 13kHz and they really didn’t like that. If I put these in a light I planned on using on low very much, I’d use a digipot to dial in the amperage.
- Lux measured at 5 meters on a cheap UNI-T meter from Banggood.
- Lumens measured in a DIY lumen tube with a Adafruit #1980 / TSL2591 sensor
- Amps measured at tailcap with short, heavy wire and a clamp meter