What is the "Best" way to dedome a led ?

You can see the archived poll results on the Wayback Machine:

What is the "Best" way to dedome a led ? - LED Flashlights – General Info - BudgetLightForum.com

Crush?

Ive spent a few hours last night looking at different methods and guides.
Those were the 4 methods I found that people seemed to suceed with.

Heat . >80, <100 degrees … with a heat gun…. now I dont have a heat gun… does a soldering iron with a flat tip or a butane torch work ok ?
How would you go about doing this on a fully functioning light ? de-solder first and seperate ?

Disolve… people say petrol … what sort of petrol is never specified… unleaded ? premium ? 10% ethonol ? 50% ethonol ? Thinners ? Metholated spirits ? turentine ? acetone ? isopropanol ?
What is “Best” ?
What is the sucess rate ?
What are the negative effects ?

Crush… tap… grind… pliers ? vice ? Hammer ? … Yeah that doesnt sound to safe to me …
What is “Best” ?
What is the sucess rate ?
What are the negative effects ?

Cut … ok but after the cut ? is it just cut and pray ?
How high/low is “Best” ?
What is the sucess rate ?
What are the negative effects ?
Do you do anything to the protective coating after cutting some of it off ?

the two most accepted methods here are solvents to make the dome fall off, or let the led get hot, and carefully pull off dome.

Gasoline for me.

I don’t have to be precise or careful with a knife to cut it off or gently peel it off with the right amount of heat (though these are very fast methods).

I just hang in it a tiny bit of gas and let it sit. Come back a few hours later and Tada!

+1 i do the same but i hang my LED up side down so when the dome comes off it comes off all at once

Depends on the light, if it’s 4.5A or less then I use gasoline. Sometimes I drop the emitter in at bedtime and pull it out in the morning when I’m ready to build the light. Otherwise it’s a 2-4 hour delay which is a PITA if I didn’t know I was going to get the opportunity to build one.

If a light if over 5A (most of mine are over 6A) then I use the heat from the emitter itself running in Turbo to pop the dome in a couple of minutes. The only one that’s failed me to date wasn’t powerful enough at around 4A. I use a welders lens that I bought for taking pics of an eclipse to view the light while on, with a light trace around the circular base of the dome using a scalpel first…using the tip of the scalpel I pull up with a light but steady pressure and in about 30 seconds or so of run time you can see the separation on the die as the dome peels off. I like to start on the negative side and flip it towards the bond wires. Slow and steady, not getting ahead of the peel. Again, this is usually about a 2 minute job start to finish.

So tried unleaded petrol 10% ethanol.
Suspended the led at about 7pm last night. Took it out at 7 tonight. . Petrol had evaporated quite a bit and was no longer touching the led.
So side note… when putting led in petrol suspended about 5mm below surface seems to work well.
Did not need to cut anything… tipped the led on its side, touched it with the tip of the blade and the dome just fell off…
Easy as pie…
Slight greenish tint to the light now. Seems a bit brighter but i don’t have the gear to test so thats just observation. Will have to test with a better light.

Why do you only hot dedome LED’s when you are running them over 5A?

I have only de-domed two emitters…both with regular unleaded gasoline left overnight. Easy.

I just use the gas method, regular gas, in a capped jar to avoid evaporation. The time consuming thing is cleaning up the remnants of the silicone around the phosphor, if there is any left - sometimes they come out perfectly clean. To do that clean-up, I use toothpicks because they won't scratch the silver/gray surface. Metal tools will scratch it up. Must be careful to leave the 2 wires fully in tact, so usually I'll leave the silicone around those wires -- gets too risky. I use a 20X loop and a good light source (headlamp) when doing the toothpick clean-up.

The throw kcd value (not throw distance) will usually about double, if focused properly, while the total output (lumens) will drop between 10%-20%. The tint will turn a strong yellow with some green perhaps. A de-domed 1A will be more yellow than a 3C or 4D from what I've seen and others attest to. De-domed 3C's look even stronger yellow, so I limit the de-domed bins to 1A or 0D. Even 1C's are so-so.

I’ve de-domed 3 of them so far. :slight_smile: The heat from the emitter does the trick. You have to run some leads so the mcpcb is off the sink and put a rolodex card on the backside so the clips don’t short out.

Look at which side the small leads are on. There’s another lead that connects the 2 sides counterclockwise to the main ones. Found it takes ~8 seconds on and the mcpcb is almost too hot to touch. Switch the light off and lift one corner at a time on the side with the leads. Don’t lift in the middle though, you want to lift at the corners only with an exacto knife. You’ll see the dome kind of peel up as it hits the yellow emitter. Turn it on and do the other corner peel up a bit.

Now on the 3rd burn you hit it from the other side, and peel the whole thing up. There’s no leads on that end to mess up. It should come off with one peel clean. There maybe some left around the main leads but leave that there. A 3 dollar 40x jewelers loupe lets you inspect the work.

The gas method sounds good too, but I’m scared xd If you get anything like specs on the emitter before install after soldering. You can take cheap packaging tape, the kind that doesn’t stick well, and just touch it on top with a corner of the tape, comes right off. The loop helps a lot.

I'm 100% with gas - done dozens.

Forgot to mention - I always clean up by squirting isopropyl alcohol directly on the phosphor and all over the LED - again, never any damage, never any particles left. Seems 100% safe -- dozens, never a problem. My thread here: A Perfect Dedome?

I too have done dozens that way and it’s the most reassuring way. The toothpick step is almost a must.

After discovering the hot de-dome method, I really haven’t gone back to using gas. I don’t like the smell of it. Or the smell of the alcohol to clean it. yuck. :Sp lol

Since the way I use the hot method has the emitter in the light, all heat sinking is in place. Pretty sure it takes 5A or more to really get a safe release without breaking bond wires. The only one I’ve had fail was under 5A. For me, this is a happy medium of safety margins. Hot de-doming outside of the light leaves the possibility of overheating the emitter. I like the KISS rule, because I can find ways to destroy these lights that no one else has discovered yet out of my sheer forgetfulness. :smiley:

I use 2K thinners as my old man owns a panel and paint shop so I have plenty of it available. From what I have read about de-doming with petrol, the thinners is essentially the same just quicker. I find brand new LEDs take a bit longer than ones which have been used in the light for a while, however the longest I have waited for it to float off is about 2 hours.

Thats what I like to hear lsx, thinners just seems cleaner to me. Shall use it for next one.

I have a 1D in as throw. It’s not bad but it is more yellow when compared to my other lights. I should have went with 0D or 0A from the looks of the cool white chart. Where would you get those bins cheap? I don’t think I see them on on mountain electronics.

Has anyone noticed a difference in tint shift from using petrol than the peel method, which one creates less yellow tint afterwards? or is it all the same no matter what method you use? I wonder what original tint Vinh54 uses for his premium tint dedomes and what his rejection rate is.

Cheep hardware store plastic aspheric sacrificial lamb for thinners test..

!http://imgur.com/9TtTDtN!:http://imgur.com/9TtTDtN

Question... if dome does not just fall off.. above or below metal ring with knife ?

!http://imgur.com/VtvceS1!:http://imgur.com/VtvceS1

I got a Shocker modded by vinh once - most awful yellow dedomed tints I've seen. Think he dedomed 3C's or something. I've seen other dedomed lights by vinh and I don't notice any difference between his tints and my gas dedomed ones, but I think Dale has noticed differences. I've tried heat methods before and can't seem to get it right - gas is much easier for me.

Interesting to note now... djozz kind of proved our cheaper light meters (mine included) don't handle tint differences too well, and seems to record CW's higher, and NW/WW lower, so wonder if that's true for dedomed LED's, because I usually see a 20% loss in lumens after dedoming, so could be 5-10% loss simply from the poor quality light meter.