Scalpel De-dome

So I bought some cheap scalpels off of amazon a while ago and figured I’d give de-doming a shot.

I had an old cool white xml and it made the perfect candidate. I should of taken some before color shots but it was very cool white.

I slid the blade between the silicone and the top of the board. It slid right through, and then I pried off the dome with quite little effort.

It looks very clean, but I can’t tell if I lost phosphor doing it this way. I don’t have a magnifying glass so my glasses are the closest thing, and my camera does not do macro too well.

It was surprisingly easy though, and I will be hooking it up in my c8.

It will still likely to have imperfections in the beam pattern. Hot and pry or gasoline methods prove to be better.
But good luck!

Using gas will remove all the silicone dome without messing up the Phosphors layer any, slicing it off like you did will leave a layer of dome on top but will also help protect the bond wires if you need to disassembel the light a lot for any reason.

Slicing through the dome leaves a pattern in the cut surface. Peeling the dome off completely will also remove the bond wires. If you haven't tested it yet, be prepared for disappointment.

Well I never sliced through the dome above the emitter, just around it where it contacts the board. (I should have been more specific)

The dome just peeled away from the emitter, and it looked like it took all of it with it. The dome was smooth where it came into contact with the emitter.

Yes, I know. The dome will peel off cleanly when cold. That's been known since forever. The issue is, peeling the dome off cold also breaks the tiny little gold bond wires, they are embedded in the silicone. Removing the dome hot softens the silicone so that the bond wires can cut through the silicone without breaking. Soaking in gasoline does the same thing. Peeling it off cold kills the LED.

Have you tested it yet to see if it still works?

Yeah I made sure to test it right away, the bond wires didn’t seem to suffer at all. It’s most likely just a game of chance and I got lucky with this one then.

I clarified that I didn’t cut above the emitter because it meant there wouldn’t be any kind of layer left to cause beam imperfections as the other guys pointed out, and I don’t understand what you meant by “leaves a pattern in the cut surface”.

I guess my first try was a lucky success, I’m gonna give it a shot on an xml-2 that I don’t have any plans for and see what happens there.

I pried it from the opposite side of the bond wires, so that it would allow the wires to slice through the dome gradually, at least that’s what my theory was.

Sounds like you got lucky, I’ve found gasoline so clean, quick and easy I don’t see any reason to do it any other way.

Good luck!

The bond wires will not cut through the silicone when cold, no matter how careful you are or which direction you go from. The only way they survived is if the silicone tore and the section around the bond wires stayed stuck to the substrate.

The thing about 'leaving a pattern in the cut surface' is in reference to when you slice the dome, instead of completely removing it. Some LEDs (XT-E, MT-G2) have the phosphor layer embedded in the silicone dome and can't be dedomed by just peeling it off, no matter if that's done with heat or gas or anything else. If the dome comes off it takes phosphor along with it. Those have to be sliced just above the die. It's about impossible to make that cut in one single motion without sawing back and forth, and that leaves a series of ridges and valleys in the top surface. That can be mostly cured with a single drop of LEDseal (a liquid, optically-clear RTV silicone spray).