Yep, it’s often on talkshows that the razor blades cost nearly as much as an actual razor. Kind of like how ink cartridges for printers cost 4x as much as the printer itself.
so, despite Agenthex’ rather obnoxious approach to discussion (try that at a conference and see how much people want to share knowledge with you…), it did prompt me to fix some glaring holes in my knowledge about how LEDs work which can only be a good thing.
Here’s one of the better references I found
http://www.digikey.com/us/en/techzone/lighting/resources/articles/Improving-the-Efficiency-of-LED-Light-Emission.html
although there are several others (many use excerpts from the wikipedia entry), which does a good job explaining things.
So, as I understand it, the dome exists to more efficiently extract light from the phosphor by reducing the difference in refractive index between the phosphor (3.7ish) and the air (1, I think) and by allowing the light leaving the dome to exit at as perpendicular an angle as possible. I could not find any references to the dome focusing the light in any way (if it did, why would we need optics?), only that it prevented light internally reflecting inside the phosphor. There are some prospective technologies to harness the remaining internal reflection, but right now it just ends up as heat. I would expect the decrease in total light output on de-doming to be the result of less efficient light extraction (or more internal reflection).
Why de-doming would effect the size of the hotspot, I have no idea. Perhaps because the light leaving the remainder of the dome is not doing so perpendicular to the dome and is therefore refracted away from the normal when it goes from dome to air, thus more light hits the reflector and is focused into the hotspot? It seems intuitive that the light isn’t leaving the LED in a way which the optic designer accounted for in the design. Perhaps it has the effect of moving the effective focal point of the reflector backwards, behind the emitter?
Either, my thanks to Agenthex for his inspiration for my education. I think I’ll leave the lesson on how to correct peoples’ misconceptions on the lab floor though, I can’t see any of my students benefiting from it 
My only experience dedoming a XM-L left it without phosphor and it was really annoying blue.
My measures(X7 XML):
Original: 97000 Lux
Dedomed with Blade, very good/clean surface: 116000 Lux
Complete without Dome: 182000 Lux
Thanks a lot man!, I’ve been wondering for a while now how a de-domed XML would behave in a Regina. How hard is that P60 drop-in being driven? That seems to be a ridiculous amount of throw considering it’s coming from an XML in a 19mm reflector!
………
Pulsar13 is dead on when he says stick to razor blade, like the one linked to in the OP, or an equivalent extremely thin really sharp blade that you can find in your area. When I say thin I mean thin enough to be flexible when bent… so knives are out . The method is quite simple as it is, and so a little effort put into getting the right tool will increase chances of success. If done correctly with one cutting motion or push, like LowLumen, AlexGT and others have stated, you will be left with a smooth clear surface that apparently won’t even need to be polished(this needs confirmation as it probably largely depends on the blade and technique used). LowLumen also had the idea of using a shim/spacer around the emitter to act as a guide when cutting, which makes this process far more accurate and dare I say foolproof.
…….
Regarding this method vs heating and pulling the dome off, if that works for some then great, it just didnt work so well for me when I tried it. Some of the phosphor ripped off and the center of the hotspot turned blue. I ended up doing an emitter swap and trying this method which worked. There’s another thread here listing problems people have had with the heating method which is what prompted me to put up this one. Maybe a certain amount of heat is needed to allow the bondwires as well as the phosphor to stay intact? The heat + fingernail method does seem to be getting some good results in this thread though lol.
Interesting, your de-dome with blade results are similar to what some members how reported here, yet the member the and others have achieved similar results to your complete de-dome with the blade de-dome method. This needs more investigation. How much of the dome was left after you sliced part of it off?
With blade-method, the rest of the Dome was ~0,2mm.
Thanks Andy, The only explanation that I can think of is that there were small imperfections after cutting which would have been helped by polishing. It’s interesting to see such a wide range of results using one seemingly simple method.
I fully dedomed the led in my UF D1. The dedoming methode was the following: I heated the side of the dome with my soldering iron directly, and I pushed very carefully. Only some sec and the dome softened, and it can be removed.
The led current somewere around 3.5A.
With original led was 52kcd, with dedomed led increased to 110kcd.
overall brightness decreased from 800-850lumen to 620lumen
original led
dedomed
I notice a tint shift in the pictures, what was the original tint and what is your best guess at the dedomed tint? Thanks!
The rest of my dome is coming off today. We shall see…
Well after 5 tries, I finally won one. And now I know what not to do with a pocket knife. But I did this Keygos KE-1600 that I have. Turned out pretty good. And I did it with a knife. You just got to know when to say when. Take it from me.
Well I killed my Z8 last night trying the heat up method. It looks like it broke the wires. I agree that method does work - the dome did get removed. But either I didn’t get the temperature right or I just wasn’t careful enough. It came off in a couple pieces. But it’s dead for sure.
Oh we’ll I guess it gives me an excuse to learn how to change out led stars. Guess I can upgrade it to a U2 this time. Anybody have a link for a US distributer of LED’s?
Thank you
the original tint was cool white, and the dedomed led has become warmer, tends to neutral white
viffer750, Yoy pussed the dome directly with the solder tip, or with tweezers?
I would like to give a try to this method, but would be nice to do it right in the first try hehe.
Just successfully dedomed the famed Manafont T6 3 mode . I used a brand new wide razor blade and used the edge of the pill as a guide . I probably removed about half of the dome .
I got a similar result as many of you , a tighter spot , a shift towards neutral ( in the hotspot , the spill retained it's purple hue ) and now there is a slightly darker ring around the hotspot .
I mainly did it because the dome had always had a defect in the dome that was visible on a whitewall beamshot .
With the solder tip. You will feel when the dome softens, but dont heat only one side of the dome, you have to heat from all side. Don’t force the dome when it is not enough soft, otherwise you will destroy the phosphor layer and the wire bond. The most dangerous area where the wires are, but you know that. Better, if you leave a little dome material around the wires.
I dedomed 2 more with great success.
I’m on a roll now.