Yes, highest specific heat for liquid (heated slower and also cooled slower). But you still need to get rid of that heat anyway. Silicone isn’t as good as water in both specific heat and heat transfer but also heated less by the trapped photons by releasing more.
I also noticed than domed LED are cooler than “HI” LED. The inefficiency of those domeless LEDs is converted as heat.
When it’s in an aluminum cup (the reflector) that should be fine as well.
…provided that there’s a thermal path from reflector to the outer body.
Does that mean it is more translucent than water?
Most of the photons reflected back into the die by total internal reflection because the sapphire has much higher refractive (RI) index than air. Silicone has RI between sapphire equates more photons extracted ~ more lumen. A lot has been done to reduce this TIR effect such as contouring the substrate to lessen the chance of photons bounced back into the substrate. This is what makes domeless LED less efficient. Optimum silicone dome geometry is one with x curvature so that each photon will exit in as low angle as possible below total internal reflection critical angle. The dome can be made from anything, not just silicone though.
By reducing the RI difference between sapphire substrate and air, we can extract more lumens out of an LED. Water has less RI compared to silicone.
- Air ~ 1
- Water ~ 1,33
- Silicone oil ~ 1,4
- Sapphire ~ 1,7
It’s a very complicated subject as even the best dome geometry (for a given emitting surface area) will increase the risk of tint shifting. Sliced dome like those found in Cree was made to extract more lumens while induce tint shift at high angle.
- Clemence
Yes, i understand the internal reflection is eliminated when the liquid has the same RI as the dome.
That’s like a piece of glass being “invisible” in certain vegetable oils.
But i still don’t understand how a dedomed / domeless (bare phosphor layer) LED has a lower light output.
…or do you mean the phosphor when you say sapphire? …or do you mean the dome?
The silicone “dome” extract more photons better than flat “HI” silicone dome due to it’s geometry. The phosphors usually just mixed or embedded in the base of the silicone as close as possible to the sapphire. Yes, I confused you, I meant the phosphors on top of the sapphire.
https://en.wikipedia.org/wiki/Light_extraction_in_LEDs
A bit of necroing, but I find it relevant.
First, even water is not a superb conductor. It doesn’t make much sense to fill entire head with it; in a non-edc light nearly all heat will just travel past the LED sides to MCPCB and from it to the body. So some smaller compartment would work nearly as good while having a fraction of weight. You can’t do this easily with reflector or TIR based lights, but with aspherics you could build a compartment on top of MCPCB. It’s actually quite natural if in the few precollimated lights that I’ve seem.
As to the heat conductor, water is a bad choice as it freezes below zero. You have to find some antifreeze that doesn’t destroy electrical or optical properties to make it usable.
Oil is better optically, thicker (and therefore less leaky), but bad thermally.
Solid compound would be better:
- doesn’t leak, no need for sealing
- compared to oils it’s a less developed market, so you can’t pick index easily
- solids are better thermally than fluids. May beat conductivity of water, will beat oil
Application:
1. Solder the LED to MCPCB
2. Mount a centring ring
3. (optional) Dissolve or slash the dome
4. Mount a pot on the centring ring
5. Pour in your potting compound
6. Use the centring ring to mount optics
You can make the pot in a dome shape for best light extraction.
Or, for bonus points, you can shape it as a TIR lens.
First of all, we were talking about heat build up by the photons unable to escape and phosphor layer that can’t be cooled (by design), not simple heat normally found on common LED.
I don’t think you get the idea or the complete pictures. My fault, the pictures link changed unintentionally. I’ll get them back tonight so you can see what I meant. See you later
- Clemence
Pictures in OP fixed!
As far as I see, what I posted above is relevant regardless of the heat source.
An additional dome like this would allow better light extraction than a stock one, reducing heat production. It would cool phosphor, slightly improving its efficiency and protecting it from overheating. And, indirectly, it would cool the die by taking part of the heat via another route, further improving efficiency.
Thermals aside, molding a TIR directly on the MCPCB would remove one air layer and related reflection losses.
Oils allow a better selection of index, so would enable a bit better light extraction. But they are worse in about every other parameter.