dedomed a xm-l with success

[Walterk]

To be pedantic, I'd say that it drastically increases throw rather than performance.
Since the total output is actually decreased after dedoming, I don't think its correct to use the word performance.

Yes we know, well I know from beamshots and experiences shared on CPF.

Talking only about the witnessed drop in lumen, please be clear on where it comes from:

Isn't it that the same lumen are emitted, but only in a viewingangle that is hard to collect by most reflectors and optics?
(And thus most experiments so far have improved throw but not lumen)

So we should regard dedomed led not anymore as an ' almost-point-source ' and thus have to look for an other lenssystem to maintain lumen...

I think dedoming invites to step off the straightforward aspheric, and think more like MegaRay and other light projection lenssystems.
Or using smaller diameter (led filling more of the aperture of the whole lens), multiple lenses like described in the Ledlenser.

[ergotelis]

I own you some photos and beamshots , sorry for not having already done this. On program it is to go out tomorrow in the same place i had taken some beamshots in the past, to show you how a dedomed led performs, among with other famous flashlights.
In general, a dedomed led has almost double the hotspot lux in the same reflector, but with a performance drop of 20-30%.
I will take beamshots of a dedomed P60 host (Xeno F8V5) and a Trustfire X7 dedomed xm-l mod.
I will include some very popular flashlights and another interesting mod, a Armytek Predator with a XT-E R5(from cutter) led.
I have another dedomed xm-l led(other than the first two mentioned), it is currently in a cheap trustfire 40mm host(of $12), i will include that too, just to show that in a very cheap setup you can get crazy lux numbers.

[saabluster]

To be pedantic, I'd say that it drastically increases throw rather than performance.
Since the total output is actually decreased after dedoming, I don't think its correct to use the word performance.

"Increased performance" is a perfectly acceptable term to use to describe the effect of removing the dome. Well perfectly acceptable given it is said in context. Context which I did not provide so I am glad you mention it lest we get people thinking dedoming creates more lumens. Thanks

[saabluster]

Yes we know, well I know from beamshots and experiences shared on CPF.

Talking only about the witnessed drop in lumen, please be clear on where it comes from:

Isn't it that the same lumen are emitted, but only in a viewingangle that is hard to collect by most reflectors and optics?
(And thus most experiments so far have improved throw but not lumen)

So we should regard dedomed led not anymore as an ' almost-point-source ' and thus have to look for an other lenssystem to maintain lumen...

I think dedoming invites to step off the straightforward aspheric, and think more like MegaRay and other light projection lenssystems.
Or using smaller diameter (led filling more of the aperture of the whole lens), multiple lenses like described in the Ledlenser.

The same lumens are not emitted.

The dome does two things that improve light extraction. It provides a more gentle transition in index of refraction which reduces reflection losses. They actually use silicones of varying indices going from higher index materials close to the die to lower index silicone as you move out away from the LED die. The LED die has a very high index of refraction and without this gentle transition more light would stay inside.

The other thing the dome does is move the escape surfaces farther away where the angle the light can escape the package is more perpendicular. Light transmission through materials of differing indices happens most efficiently with the light at a 90% angle to the refractive surface. The more you diverge from that the more that will be reflected internally. LED manufacturers could actually improve the performance(lumens) of their current LEDs by doing nothing more than making the dome larger. They of course also have other things like cost to factor into the equation.

[bshanahan14rulz]

SaabLuster, could you tell if the phosphor was suspended in silicone too, or if it might have been a different material? Perhaps a solvent or digester ought to be researched. The Nichia (not a cree, but I don't think I've seen this type of info on a cree DS) datasheet I have actually tells what the different parts' substrate materials are, generically, i.e. dome is silicone (doesn't say what it's doped with, or if it even HAS any additives), phosphor layer is phosphor suspended in silicone.

[Walterk]

In general, a dedomed led has almost double the hotspot lux in the same reflector, but with a performance drop of 20-30%.

Don't be sorry, its nice you show your experiences. When you take beamshots, could you make a white-wall beamshot with a dedomed led without reflector/lens, so we can see how the light distribution has changed?!

The same lumens are not emitted.

Thanks for the clear write up, I expected almost same lumen.

[ergotelis]

Don't be sorry, its nice you show your experiences. When you take beamshots, could you make a white-wall beamshot with a dedomed led without reflector/lens, so we can see how the light distribution has changed?!


Thanks for the clear write up, I expected almost same lumen.

I had promised to do the beamshots but i didn't , i am planning to go out tomorrow for this.
If i understood well, you need two beamshots, from a dedomed xm-l led and a normal xm-l , facing a white wall to see the difference?From what angle?
I clearly measured 20-30% lumen drop output. I think that, as current increases, the difference gets higher. It is logical up to one point, i believe that the energy from the lost lumens is converted into heat, so we get higher temperatures on the core and a another performance drop. I am waiting for saablaster to confirm this statement.

[psychbeat]

Im curious if dedoming helps as much in say a p60/D26 sized reflector as much
as it does with an Aspheric. Maybe Vinh or Dave would start offering this as an option?

thanks for posting your results!!

[Dioni]

I'm very curious to see the results in the beamshots !

[Walterk]

two beamshots, from a dedomed xm-l led and a normal xm-l , facing a white wall to see the difference?From what angle?

What I want to see, is how the light comes out from a dedomed led. How it flares away from the die.
Maybe half the beamangle, cut in half by a piece of paper or along the wall ? Something like this but then without optics:



With reflector shows what dedoming the led does for you.
I wonder what I can make from the dedomed. Half the beam indicates how the light and intensity is distributed, so we can find ourselves optics to match the led-characterisitcs.
Thats why I like to see the bare led without optics.

Der Wichtel posted some comparising beamshots in his post Another-dedoming-experiment sometime ago, but they are not that clear.
The picture above is from the most useful open dir archives from Newbie.

[saabluster]

SaabLuster, could you tell if the phosphor was suspended in silicone too, or if it might have been a different material? Perhaps a solvent or digester ought to be researched. The Nichia (not a cree, but I don't think I've seen this type of info on a cree DS) datasheet I have actually tells what the different parts' substrate materials are, generically, i.e. dome is silicone (doesn't say what it's doped with, or if it even HAS any additives), phosphor layer is phosphor suspended in silicone.

Yes Cree uses silicone as a binder for the phosphor. This means that any attempts to dissolve the dome away also dissolve the phosphor layer.

I had promised to do the beamshots but i didn't , i am planning to go out tomorrow for this.
If i understood well, you need two beamshots, from a dedomed xm-l led and a normal xm-l , facing a white wall to see the difference?From what angle?
I clearly measured 20-30% lumen drop output. I think that, as current increases, the difference gets higher. It is logical up to one point, i believe that the energy from the lost lumens is converted into heat, so we get higher temperatures on the core and a another performance drop. I am waiting for saablaster to confirm this statement.

It does increase heat but it hasn't affected the maximum drive levels as much as I had anticipated.

Im curious if dedoming helps as much in say a p60/D26 sized reflector as much
as it does with an Aspheric. Maybe Vinh or Dave would start offering this as an option?

thanks for posting your results!!

Yes it makes a big difference with reflectors as well. I am about to begin offering just this. Hang tight

[psychbeat]

Yes it makes a big difference with reflectors as well. I am about to begin offering just this. Hang tight

Keep us posted!

[mvyrmnd]

I realise this thread is quite old, but I'm about to replace the emitter in a light, and decided to de-dome the old one for giggles.

I was very, very careful, and picked apart the dome with a very, very sharp knife and a pair of tweezers. The last part of the dome broke off, and still remains covering the bond wires, and I'm not game to try and get it off, so there's a minor artifact in the beam because of it, but I'd call it a 99% success. I sliced straight through the top of the dome, leaving about 1mm of (now flat) silicon covering the die. Then I used the knife to lift the silicon on 3 sides (I didn't want to break the bond wires) then picked it off with the tweezers. It came off in one chunk, minus the bit still covering the wires.

Some pics





I have no means of measuring the output or lux increase or whatnot, but the spot is now half the size it used to be on my office wall, and it hurts even more to look at.

I'm pretty sure this scraggy Ultrafire HD2010 will now keep up with some of the big boys for lux, if nor sheer output.