Blew of the dome and phosphor and the led went dark. At 4A voltage difference is already 0,2 volts and as current goes up so does the difference in VF.
Good to know that it only takes an OP reflector to take care of the B35AM donut hole.
I can confirm as well, I tried several SMO and OP reflectors with B35A, with SMO there is always a darker yellow spot, with OP reflectors itās gone.
Took on the Sofirn IF25A as well, but with the 519A plan gone out the window, I made my 1st tint mix. Nothing special, just replaced two 4000K SST20 with 2700K. I was a bit sceptical at 1st, but my oh my, tint mixing is really nice.
Stock 4000K vs 2700K/4000K mix
Btw, worst out of the box thermal paste job Iāve ever come acrossā¦. replaced with properly applied Arctic MX-4 now.
Does the LH351D fit with the optic? Seem to remember reading conflicting reports. Mixing might actually make the greener ones bearable.
I built mine with 2 original 4000K SST20ās and 2 5700K LH351Dās sliced to 1mm. That way I brought duv down and achieved about 4500K color temp still with high cri and maybe a bit better performance.
So only sliced fits nicely then. Not done any slicing yet, so maybe I need to bite the bullet and just attempt it. Still have an ugly SST40 and XHP50.2 laying around gathering dust that could serve as practice before moving on to LH351Ds.
ā¦.like a sticker you saidā¦. I beg to differ 
Still works though!
As long as the phosphor is still on the led it shouldnāt be a problem. Otherwise you will get some nice blue. 
Iāve never been good with stickers
if it works, it counts :+1:
my first attempt had similar results
the next one, I pulled the dome off right after reflowing, while it was still hotā¦ came off much easier, in one pieceā¦
Ive been carrying the dedomed 519a since I built itā¦ more fun than a barrel of monkeys:
.
I was planning my next LED mod to be triple dedomed 4500K 519As, but Iām a little apprehensive now!
Yeah, fully understandable. But donāt let it deter you too much. Learn from our mistakes. And if youāll be putting it under an optic lens, any visual imperfections probably wonāt be visible. Different story when using triple reflector though.
Fwiw, I did it a minute or 2 after taking it of the hotplate, so it probably cooled down too much already. Will try again when warm, maybe that works better. It looks ugly on the outside, but the glass cover cleans up very nicely and is actually a nice reference point to see if the dome is fully off. Reflects brightly under certain light angles.
I went ahead and ordered 3 519As from Simon. Seems the risk is worth the reward. Theyāll be under a 10511 Carclo optic.
I wonder if it would be better to take a razor blade to the dome since thereās a glass cover to protect the LED? Keep shaving it closer and closer until the glass is reached.
Maybe between now and when I get them, someone will figure out the trick to dedome them. 
In my experience they are hard to not dedome right. Just slowly peel from corner.
Ouch !
The 2 i have dedomed, did come off exactly like shown in the pictures!
Good to hear yours still functions though
.
.
.Attempted slicing of a XHP50.2 today. Slicing went ok, although using a japanese kitchen knive might not have been the best choice. Very sharp = good, slightly curved = not ideal. I used two 1mm drills taped to the mcpcb as guide.
The main reason for this was that I really didnāt like the egg beam. So after slicing I put it back into the Astrolux FT03 it came from, expecting a massive improvement. Total failure on that front. Itās still by far the ugliest beam Iāve ever seen. And itās impossible to get it properly centered. Seriously disliking this emitter/host combo.
Usually people also scrape away the excess phosphor around dies. Slice and dice is the term I have heard used.
Thatās quite a lot that can be taken away around as well 
Thanks, good detailed info! Got nothing to loose with this one, so taking it out again to remove more material. Even if it fails, itās at least a good practice exercise.