I just went out to the shop and got my good digital calipers. The XP-G3 die measures 1.4mm x 1.4mm. Then I tore the dome off a new XP-G2 and it measures 1.47mm x 1.47mm. So there you have it, destruction notwithstanding the G3 is a smaller die than the G2.
Well,based on all this chaos with xp-g2/xp-g3 transition,I wouldn’t be surprised that there are also two versions of xp-g3 
Now wouldn’t THAT get confusing! lol
I don’t understand how FMC got those measurements, I also used a good set of digital calipers and I actually measured the XP-G3 silicon wafer OFF the die. I got 1.47mm on the XP-G2 and 1.4mm on the G3. I could, conceivably, have gotten some phosphor or residual silicone dome with the XP-G2, which I must have if indeed the G2 is smaller.
But there is zero doubt, the G3 measures 1.4mm (well, zero doubt within the parameters of my Niko calipers)
Guys,
Die size is really not that important…
What matters that one more respective modder did test of de domed XP-G3 with well known host under same conditions and he got terrible results. And not to mention XP-G3 sags 8A of current!
Even bad(new production process) XP-G2 seems better than new XP-G3. So is there a reason to be enthusiastic about this emitter? I think there is no reason for that. My hat down to MEM if he will manage to improve it. My bet still stands MEM…
Even if it could driven at 8A reach performance of good old XP-G2(s42b for ex) driven at 4.2 –4.5A; same performance should not be considered as an improvement imho.
I want to say that at that current sag(8A) G3 would really have to be special and have more than 20% better performance than good old G2.
G3 and FET driver :question: :person_facepalming:
My god. Lol.
Luminarium, you and your “XP-G2 production changed” thread. :laughing: What I see is the popularity of a thread you made getting to your brain so bad that you just won’t let go and assess some certain realities. One reality is, that old emitter will be many times harder for you ever to come by anymore simply because you advertise it everywhere you post as being the undoubted champion of them all. At the same time, you’re begging people to get the emitter for you. I can’t help but find some humor there. I’m not disagreeing that it isn’t a good LED. It can make good intensity. We get that by now. It also has plenty of green phosphor ratio, so yes, it is a slightly green tint. When I bin the tints out, the S4 2Bs that aren’t so green, are fairly yellow. I do have strips of sealed S4 2B chips; they are the older smaller die. I have both series, though. You’ve said that the newer ones are more green, and that seems to be true in my testing. The interesting thing is that I have experienced opposite effect with newer generation 0 & 1 tint groups of the S2 bin. The newer generation of those S2s that are larger dies have had much greater white output in my limited testing. I can say I am very impressed with some newer XPG2 tint bins and their flux numbers.
What I’m most confused about though, is why the durability of the XPG3 keeps being labeled by you as “sag”, as if its ability to strike high amperage is something the emitter does wrong. Sag is something a battery does under load, plain and simple. When an emitter can pull 8amps on one single cell, that’s great for us, because the LED has a low enough resistance for the single cell voltage to still push through. The XPG3 makes nearly 1800 lumens at 3.8V! To actually compare the two emitters for efficiency, G2 and G3, you must compare power used, not amperage.
If a XPG2 at 6A and 4.5V makes 1400 lumens, that’s 27 watts of power being used. Dividing the lumens by the watts gives almost 52lm/W.
XPG3 at 8A and 3.8V is 30.4W. Divide those lumens and the result is almost 59lm/W.
I’m struggling to see why you seem to want the data to agree it’s a bad LED. Stock vs stock it’s great.
My dedomed XPL V6s are a lot more intense than V2 HIs. It’s almost like a V3, and in some few cases a V4.
I haven’t seen anyone else truly dedome a XPG3 yet, and so there is no bias because of me doing it, I’m going to send Dale one to test. (Sorry Dale, it’s the cost of free things around here.) 
What’s a dedomed XPG3 look like up close? This one below was done this morning.
EDIT: Oh and of course, here’s how that emitter’s getting ready to run tonight. :smiling_imp:
Oh goody! A freebie! I have a Jaxman X1 on the way, maybe this is the perfect compliment to that smallish thrower light! 
I made a light from scratch on the 17th of May, a small 18350 powered quad XP-G3. I like it very much for it’s tint and sheer power at 2680 lumens drawing 7.57A (at the emitters) from the little half cell. I am quite happy with the choice to use the G3 in this way, it doesn’t have to be “better” than something else, it’s a nice emitter, it works well, makes a nice tint, that’s enough for me.
So now I will see how it functions as a thrower in the X1, thanks for the opportunity MEM.
MEM do you have some throw measurements of the XP-G3? I mean lux or cd ratings of a dedomed emitter in a well-known host. Why is it so that it simply does not throw better, despite all the lumens and its size?? It really seems nobody has managed to make it perform better than its predecessor yet, the xp-g2. Have you figured it out yet?
In the second photo there is some reflexive collar installed. Are your lumen readings with or without that collar?
You see MEM…
Please don’t offend but I think you are one of the rare persons who thinks that XP-G3 is suitable emitter for thrower configurations.
You don’t have to read what I wrote… You never did… But please read what FmC wrote. He is respective modder, and his de domed G3 looks like successfully de domed.
I am sure you can improve things with your RA… That is ok but it will be informative but irrelevant data to most of mare mortals here that will not be able to get it or fit it.
So as Johny723 asked do you have lux or cd ratings of a de domed G3 emitter in a well-known host without RA.
If it is cool to have emitter that draws(is that right expression?) 8A of current to reach half the performance of old XP-G2 S4 2B at 4.2A than I obviously must change hobby….
I admit I may be very wrong here cause you really are expert and you probably have knowledge and technology beyond my imagination and I really wish you best luck with those G3 emitters.
I can’t believe I’m wading into this…and I’m reasonably sure it won’t matter…
What MEM’s post says, clearly, is that the XP-G3 is superior to the XP-G2 in MANY ways, except in the application you’re talking about, he uses the word “intensity” in his post.
Lower Vf, better phosphor conversion or tint, and output per total given input power, i.e. Lumens Per Watt. What this means is that in the majority of applications, with the same INPUT, a flashlight will be brighter, or possibly run longer, depending on circuit configuration, all while having better tint.
How many flashlights do you see for sale that are factory dedomed? How much of the total flashlight market do those represent? How much does that figure represent in the total LED global lighting market? I don’t know the exact percentage, but I know that its a minuscule number comparatively.
And yes, before you ask, I’m aware that Cree changed the production process of the XP-G2. I was before you asked me the first time.
-Michael
To make myself clear, I believe what MEM is saying. Still, I would like to have some numbers.
It’s clear the XPG3 looks like a high performance emitter with dome. But the discussion here in this thread has always been whether, when dedomed, it will have a high luminance like the XPG2, which leads to lots of throw. Based on two independent measurements (djozz and FmC), it appears that it does not have a high luminance when dedomed.
The only reason there is still debate about this is that MEM continues to suggest that it will in fact make a good thrower. The likely problem with your logic, MEM, is that the XPG3 will not maintain its high lumen output once it is dedomed. Djozz’s and FmC’s measurements of the not-very-high luminance indicate that the lumen output has dropped significantly upon dedoming.
I measured several times to be sure before I posted, but here it is again for all to see :)
Here's the G3, die only;
Here's the G2, still in tact, but easy to get a purchase on the side of the die;
Not much in it.... what's a few 'thou between brothers? :)
And if I show pics of my Digital Caliper showing a different story, where does that leave us? A few thousandths of phosphor, where my G3 had none, that’s probably what’s really happening here, or I could have an older G2 that I’m comparing, I don’t know. I would trust your Mitutoyo’s over my Nikko’s pretty much any day. I used to sell Mitutoyo and Starrett in an Industrial Supply House, so I have faith in their quality, even if I can’t afford em for my own purposes.
In the end I believe it’s fairly irrelevant. They are very very close to the same size, and the G3 makes more lumens off the same cell. For me, that’s really the bottom line. More from the same. I am not sure what exactly does it for everyone else, but for me, I like output, I could really care less about current and Vf and all that spec stuff, for me, repeat, for me it’s about the amount of light falling on a subject down range.
I do not have all the answers, in fact, I have very few answers at all. But I will build the X1 with a de-domed XP-G3 and I will then put a de-domed XP-G2 in it and write down all the ensuing test results. With pictures of the beam shots, and then we’ll have yet another little piece of the big puzzle in place.
From a modder’s standpoint, the G2 is the easier emitter to work with. All things being relevant, the G3 will certainly make a name for itself in small lights running smaller cells, like a 14500 or 10440, due to the difference in Vf and sheer output. I’ve built a lot of quads with XP-G2 and XP-L HI emitters, the XP-G3 is a pleasant surprise in total output numbers and I like the beam quality and tint. So they have a place, for me anyway, in the small lights that I like so much.
Looking forward to your results with the X1.
The disappointment is from the thrower guys, who were anticipating/hoping for good numbers from the new G3, but it seems to have fallen way short of the mark for that application.
I’ll definitely be using the G3’s in other applications, though
I’m still thinking it will meet or beat the G2, the only way to know is to assemble one and test it. Time will tell…
Exactly!
I personally don’t care about lumens. Only thing I really care about is lux intensity. So impressive lumens per watts can’t impress me.
And yes there is still no better emitter for thrower configuration(if you want achieve maximum throw performance) than old XP-G2 S4 2B.
So since Cree made a lot of other emitters for lumen output potential you would really expect that G3 as G2 successor carries better candela or lux performance.
So with the dome on what LED throws further?
Texlite you may correct me if I am wrong since English is not my native language but for me word “intensity” means lux or candela performance.
Word output for me means lumen output.
XP-G3 may have more output but it does not have better intensity than old G2 emitters. In other words it throws less.