I wouldn't bother dedoming a Nichia, they're low output to begin with, and already have a nice tint. I only do this to high-bin cool white LEDs, mostly to improve throw but it also takes off a lot of the blue tint the ~6500K LEDs have. They sometimes end up with a slightly more greenish tint but I only notice it if I'm comparing it side by side to a ~5000K part and on a white wall.
I finished the first S4 with a dedomed XML2 U2 1C, it's pretty nice. I think it's made the dedomed XPG2 version obsolete. Have yet to do the other S4 with a dome-intact U2, also finally have the empty S2 host in hand too.
I've stripped and polished the S2/5/6 reflectors and it doesn't make the spot any smaller, it just gives it a more sharply defined outer edge. Overall throw stays about the same - might be measurable on a lux meter, but not enough to notice by eye. The OP reflectors still throw just fine as you can see from the beamshots, but the softer center spot is more usable at close range.
Just a guess but I would think the heat from XML to XML2 should be the same at the same drive current, but the XML2 is more efficient so it will put out more light at the same current. Or, you can get the same light output with less current, which will make less heat.
Thank you Comfy, VERY informative post! I won’t put too much hope on the smooth reflector helping me out then. I’m really liking the wide beam that the XM-L is giving me in the S5 so I don’t think dedoming is the way to go. Sounds like putting an XM-L2 U2 in there would help the most. I don’t want to drive it any harder, it gets plenty hot as is. (It’s been in the mid 90’s here in OK when I ride at night) Does all this make sense?
Of course it makes sense. I'd have given up by now if you were acting like a nutter. :p
OK, with a dedomed XML at least, here's what it takes to get the best focus. I originally just used a single plain standard insulator, and it was good, so I left it alone.
There's two different thickness insulators out there, one on the left is .030" thick (at the part that touches the base of the reflector, not the raised locating ring), on the right is .018" thick. I used three of the thinner ones, sanded off the locating ring on the bottom two, and glued the stack together with a dab of superglue. Just one of the thicker spacers isn't enough, two thicker ones stacked is too much. Three thin ones stacked is I think still a tad too thick, could stand to have a little sanded off the bottom, but it's a lot better as-is.
I'm not going to bother with more beamshots until I have all the lights finished and can do them all in one go, so the conditions will be the same for all.
Disregard the previous info about insulators & focus for now... just had a dim little light bulb pop on and took the XML2 S4 apart again, the way the pill and reflector make contact means thinner insulators have no effect.. see pic
I will have to trim some off the top of the pill, can't shorten the reflector because of that silly o-ring.
I wouldn't bother with a first-gen XPG. There's a lot more improvement from XPG to XPG2, much more than from XML to XML2.
I'm still waiting to finish all the competitors, as of right now I'm mostly just confused. The two S4s are finished, the older one with the dedomed XPG2, and the newer one with dedomed XML2, both driven at the same current, both on copper soldered into the pills, both using unmodified reflectors, both using double-AR UCL lenses. And while the XPG2 version has a noticeably smaller beam, I can't see that the XML2 one is putting out any more light. It should be pretty obvious but it's just not there. So what has me confused is, is the XPG2 version really special, outperforming what the parts would be expected to do 'on paper'? Or is there something not right with the XML2 version? That one is on its second emitter, too - the first one was obviously not right, either damaged during the reflow or some other problem, something happened to it to jack the Vf way up and it wouldn't even light up on the Nlite's moonlight mode. Emitter #2 doesn't have that issue but the output similarities between XPG2 & XML2 are puzzling.
Or maybe the S4 reflector is just in a sweet spot with the XPG2? I ran the XML2 version for a while before taking the dome off and was not happy with it AT ALL, it wasn't particularly floody (compared to the S3, at least) but it didn't throw worth a crap, either. Dedoming made a huge improvement.
If I had a meter, I think I'd be a lot less confused! (maybe... sometimes the numbers don't correlate to what the eye sees and makes things more confusing)
Which driver are you using?
From FastTech, the 2800mA is using 8*AMC7135 but the 3rd pic from the left looks like not. I don’t see any drivers on the spring side with the retainer ring.
OK Comfy. I saw the empty host from Fasttech. I got an UltraFire S5 have moded it but I still need to replace the driver with a Nanjg style one. Can’t figure out yet how to put it in-place. With component on the spring side will not have threads for the retainer ring to hold it.
Use some copper wire formed into a circle to use as a spacer at the bottom of the threads, so the spring side of the driver is flush with the rear of the pill, then solder as usual. I never use those silly threaded rings.
I don't know lux numbers well enough to say what it's similar to, but XML2 with the dome intact doesn't throw at all, it's just a great big fluffy center spot. Not as bad as the flood monster S3, but not throwy at all. I'm still planning to do a new set of comparison pics if I ever get around to building the S2 (which I really like design-wise, maybe better than the S4) and the other S4.
I dislike the S3 so much, I may see how the stumpy little reflector does with a MTG2 + 2x 16340. It's not like I can make it any worse...
A dome-intact XPG2 is still pretty damn good. Feed it 3A and it should be doing 700L or so... though I don't know the warmer/high CRI XPG2 parts that well.
Gords you've built enough of these weird combos, am I going crazy or is there something wrong with the XML2 version? Shouldn't the extra, what, ~400 lumens be readily apparent to the eye? I could understand not seeing the difference if it was only 200L, but this should be a big difference, nearly 1/3 more. They look the same to me, like they're the same LED just with different focus, of course with the XML2 spot being bigger and slightly less intense.
Tailcap currents are 2.72A for the XPG2, and 2.78A for XML2. Same reflectors and everything else etc. etc.
