But wouldn’t the oil trash the beam? I mean, any liquid is going to diffuse light isn’t it? And while it might absorb some heat, it can’t conduct the heat away so it would soon become heat saturated and lose any cooling capacity. Would it not? The only way I can see for any liquid to have a cooling effect is if it is transferred to a location that can absorb and wick away the heat from said liquid, then the liquid can be returned for more heat absorption, just like a radiator in a car. An oil filled heater is supposedly energy efficient because once the electricity is expelled to heat the oil, the oil retains that heat for an effective period of time and radiates it slowly to the surrounding air, so once heated up the amount of electricity needed to keep it warm is minimal.
So what good would it do to have the reflector full of oil? Even in a fairly large light, that area is of sufficiently small capacity to allow rapid heat saturation, what would then remove this heat from the oil?
I guess it’s pretty obvious I don’t understand what you’re proposing.
Makes more sense to me to have a fan pushing air behind the heat-sink and out of the light. As has been said, once the heat-sink warms sufficiently the disparity between it and the ambient temperature is sufficient to pull heat off the copper or aluminum (preferably copper) to keep the emitter cool enough to be efficiently run for prolonged periods of time in an over-driven capacity.
I recently made a copper and brass light for O-L’s challenge. When working the metal (by hand, or hand held grinding with a dremel tool) the heat would fairly quickly get beyond hand holding levels. Blowing on it, for just a matter of seconds, immediately reduced the part so I could continue working. Didn’t take much moving air to pull the heat off of the copper or brass sufficient to allow handholding. So it wouldn’t take much air flow to keep the heat-sink saturation level in the efficient range of the emitter.
Am I way off base? Work copper for a couple of hours with it held in your bare hand, then tell me I’m wrong. And yes, I’m considering the level of heat an emitter can bring to the equation, but so can a cut-off wheel, or a carbide rotary bit. If a piece of copper is clamped in vise-grips and worked this way, it will get hot enough to melt things it comes in contact with (dropped a couple of parts, found out the hard way)
Not trying to be argumentative here, just trying to understand.
The mineral oil bottle I have in the cupboard for my skin is white, but previous ones have been clear enough to see clearly through bottle and oil. Slight clouding wouldn’t hurt much. The idea is that the oil picks up heat from the led, rises and carries the heat to the case. The case gets hot even quicker, but the led itself stays cooler. The oil is as poor a heat conductor as the plastic, but it swells which makes it rise and carry the heat to the metal case, where it doesn’t hurt anything till it gets so hot it burns your hand.
No, the overall efficiency of electric heaters is all exactly the same, unless they are heat pumps. Heat pump heaters are air conditioners working backward. They cool the outdoors and bring that heat inside. Two of the laws of thermodynamics say that energy is conserved and that it tends to end up as heat. So any electrical appliance has the same heating efficiency as any other, unless it moves some form of energy in or out. That is why the same led light bulbs that save electricity directly are the ones that save on air conditioning.
After a while the outside of the light would be the same temperature as if it were full of air, but the heat would get out of the led quicker so it would be cooler.
Yes forced air would work too. So would making holes in it, but that wouldn’t be good in the rain.
I am not sure how much cooling the air around the led in a normal flashlight provides. It must circulate by convection, but its density is so low that it has to be moving pretty well to carry much heat. The oil would circulate similarly to the air but take more heat with it.
Oh - crap!! Just did a C8 with de-domed XP-E2 with Nanjg at 3.15A (1 xtra 7135). So, I'm measuring 3.28A at the tailcap with a Pana PD, been tweak'n the LED position by unscrewing the pill and added a UCL lens. Got it up to 123 kcd now, but more tweak'n...
Edit: Ooops, only 89 kcd. Battery in the meter was goin, and replacement was drained too. Reading go up when that occurs with this meter, so, 89 kcd seems more realistic now. Seems it's the best I can do with a de-domed XP-E2 in this reflector at this amps. XP-G2's at higher amps can probably do better - my Convoy C8 does 125 kcd at 3.8 amps w/de-domed XP-G2.
I posted on the “100% dedome” about an XP-G2 I’m trying to dedome in gasoline (see: X-ML de-doming method with 100% success?). The dome came off easily, but left a thin uneven clear layer of something on the emitter surface.
I’m wondering if there is something about XP-G2s that causes this (different construction)?
The dome came off clean, but left a bunch of gunk on the wires going to the die. When I did the XM-L the other day and broke a couple of the wires, gluing them back down with electrical liquid solder worked out ok. But this one has the wires on the positive side. What’s up with that? I was very careful not to break these and left some stuff there this time but still cleaned the rest up with needle pointed tweezers.
This T6 warmed up a bit, but it’s still a nice tint, slightly warm and I like it. Also made a handmade Titanium reflector for it which also turned out well, it now has a large spot with a good even amount of spill, with the blend being smooth and creamy.
ohaya - read the posts, never seen that before with an XP-G2 (residue), but mine soaked longer. I noticed in your pic one wire is gone - the ESD wire side wire, but supposedly that's not needed.
Sorry to bounce between this and the “100%” thread…
FYI, I think the problem was not waiting long enough. I did a longer re-soak, and that residue on the emitter itself disappeared after that longer soak (about 6+ hours, I think). On the XP-G2, I think, as far as I could see, there were only 2 wires, on the side where there’s a “+” on the PCB. I know that I’ve seen references to an ESD wire on XM-Ls, but the XP-G2 looks like (the wires are tiny, and I can barely see them with a 10x magnifying glass) there’re only 2 wires.
Were you referring to the pic of the XM-L on the “100%” thread?
I dedomed 2 XM-Ls the last couple of days.
An XM-L T6 (from Amazon) went in my shorty STL-V6, and is now giving me 122 Klux,
The other XM-L U3 (from FT) is in my Jacob A60, and is giving me 87 Klux now.
Post #1019 in the other thread, close-up pic of the XP-G2 de-domed:
I just so happend to have a de-domed XP-G2 on a 20mm Noctigon around, and those 2 pts above the emitter in the pic above are connected by a wire on mine - yours appear to be missing the wire.
Ooops, sorry, 2 connections just above the yellow pad - one connection is to the left of the trace cut (small connection), the other is to the right (large connection). The wire is supposed to connect the 2 connections but it's gone.
Took a C8 pill with Sinkpad still JBWelded in place, made about a dozen shallow cuts all over the dome, tied it to a copper wire and dropped it in the who-knows-how-old gas (at least a year old, from last year's lawnmower gas, has been used for around 15 de-domes and I forget how many reflectors that needed the orange peel stripped). An hour later...
...and after 5 minutes with my homemade plastic toothpick...
Scoring the dome with a razor makes it happen A LOT faster.
As I see it, the main purpose of removing the dome is to eliminate a converging lens because it reduces the effective focal length and therefore enlarges the throw hot spot.
So, maybe, the next step in that direction is to introduce a diverging element to increase the effective focal length and make a smaller spot. A common device of this sort is a Cassegrain telescope or microwave antenna. As a transmitting antenna, it focuses a finite size source to a small angular range, as we want for a thrower light.
Classically this would be done by putting a convex, ideally hyperbolic, mirror in front of the led and surrounding it with a parabolic mirror as usual. The light that misses the diverging mirror would form the spill, being out of focus for the main reflector. Other possible configurations would be adding a diverging lens to an aspheric or putting a diverging lens between a backwards facing led and a parabolic reflector. Something close to a classical Cassegrain seems the most compact.
The spill is partly focused, which may reduce the chance of blinding people near you.
