Which begs the question: Why not active cooling for lights same as CPUs???
Most of the 2x and 3x XML lights work at full power for maybe 5-10 min before thermal protection reduces current. I was reading black shadow light reviews. One of them shows sustained output at 1/3 of max due to heat. What is the point of spending over 100 bucks on a light that can handle max output for only 5 min???
Of course, fan would reduce runtime; can’t be helped. Still it would allow higher output for longer by increasing cooling capacity a great deal.
So why don’t 2K and 3K flashlights have active cooling?
Besides reduced runtime and I suppose awkward design, body design issues?
Yeah, and i think it has to pretty efficient, cooling those 2000HP away
Not shure about your concept. Why active cooling a flashlight if passive is good enough?
Just make a few snap-on or screw-on cooling fins that are big enough.
But to get back to the OP’s question which pill is better for heatsinking.
Why is there even a pill ? Wouldn’t it be better if the Alu/Copper-Star was mounted directly
on the material of the body. Without any pill there wouldn’t be any problem selecting the material
and there wouldn’t be any thermal resistance in the heatpath (thread) to the outside of the body.
As far as i remember my ROOK there is no pill involved inside.
Unfortunatly until my LiIons arrive in a few days there is also no heat inside there
Mo
Edit : Just read NinjaBob’s posting (wasn’t there when i started writing)
For 2/3 or even worst 12 XM-L’s i think there is no other way than active cooling.
Snap-On Cooling fins that size would probably make a good parachute
Well, there are some flashlights where the "pill" is integrated in the body, but it makes assembly more complicated.
From what I have learned, copper can be really helpful when its close to the heat source, meaning: directly at the LED. After that it wont really matter. So an aluminium star with copper inlay would be great, too.
I’ve read through some notes made by Cree regarding thermal management for XLamp LEDs. Some very interesting points to consider. I’ll summarize a few:
From XLamp Thermal Management (direct link didn’t work)
- The main goal is to get heat away from the LED junction. Send it anywhere, just get it away. Ideally move it to where it can be transferred outside the product (ambient air).
- Two reasons to keep the LED cool; catastrophic thermal failure, and light output decreases as junction temperature (Tj) increases.
- XLamp LEDs are about 40% efficient, meaning of the energy supplied to the LED, 40% comes out as light, the rest is heat.
- The best method to get heat away from the LED is to reduce the number of materials in the thermal ‘stack’. In most lights, it consists of LED (heat source), solder, star, Thermal paste/glue, pill, flashlight head (heat sink).
- The last component in the stack should have excellent heat conductivity, convection, and emissivity (thermal radiation).
- A large surface area, particularly vertically oriented fins, is critical for convection. Tail-standing a light is not good for convection.
Anodized aluminum has very good emissivity (~0.8, compared to 0.09 for non-anodized aluminum).
I also checked elsewhere to see if anodizing affects convection, and it seems that the effect is minimal if any. Therefore, anodized aluminum is good for sending heat into the ambient air.
As for painted or coated aluminum, the jury is still out, but I’m guessing it’s not that great compared to anodized.
According to this table, painting makes a significant difference but not as much as anodizing.
Anodizing makes an order of magnitude difference in emissivity which makes an order of magnitude difference in the radiated power since Stefan-Boltzmann goes linearly with emissivity. I got 0.7 watt radiated from a 50°C non-anodized P60 flashlight-size piece of aluminum when I plugged the numbers into S-B; anodized aluminum would be more like 7 W radiated which no-doubt would make it much more important than convection and conduction as the final mode of heat transport into the environment.
edit - if you look through the whole chart it actually claims black paint is better than anodizing. odd they don’t include a color for anodizing. anyway, i think you were looking at ‘aluminum paint’ which i assume is referring to a color of paint?
And it’s also true that in real life the temperature of the flashlight isn’t fixed; it would run cooler if the emissivity were greater.
You’re right, I missed that. It looks like paint works really well (and my friend who has done some experiments confirmed that. He also said brushed aluminum is noticeably more emissive than polished).
Polished metal has almost no emissivity. That is why we use highly polished silver service to store either hot or cold liquids. Even though silver is an excellect conductor of heat, when there is a cold liquid inside, it stays cold because the surface does not absorb radiation (heat) from the room . When there is a hot liquid inside it stays hot because it does not radiate heat. Conversely, if you put a cold liquid in a metal container that is colored flat black, it will warm up to room temperature rather quickly because the surface will absorb heat. And if you put a hot liquid inside it will cool off quickly because it will radiate the heat.
This is also why the reflector is made from a shiny piece of metal.
In short, surfaces that are poor absorbers of energy are also (exactly) poor at radiating energy
and surfaces that are good absorbers of energy are also (exactly) good at radiating energy.
I am the opposite of a techie but I have been reading this thread for well over an hour with great interest. Good stuff in here! I love metal colors in my lights - I have several Stainless Steel torches and have a Titanium one on the way. I discovered many years ago that 3M Scotchbrite pads were the best way to remove nicks and scratches from Titanium (my road bike frame) and Stainless (my Weber grill/flashlights). Little did I know I was also improving the heat sinking at the same time by replacing the shiny finish with a brushed finish.
All I can say is that if your flashlight gets so hot you are concerned about heatsinking (or burning your fingers), walk proudly, you are a true flashaholic. The Maglite and Surefire guys never give it a 2nd thought.
I read all this, and even the linked earlier discussion. Full disclosure: I was playing with Peltiers when Peltiers weren’t cool. (Well, they were HALF cool, but I digress) :bigsmile:
It seems to me, the idea of “Heat Management” for the protection of our highest-tech crystalline structures (LED or CPU) involves more than just soaking up the heat produced.
Think of it this way:
(Not picking on anybody by quoting, just pointing out how closely we all seem to be dancing to that baleful light of Truth at the center of all this.)
In essence, this is what Peltiers, Heat Pipes, and my old neighbor Kryotech do. They use (extraneous properties enumerated elsewhere) to MOVE heat, not actually “sink” it.
LY, MI, but I’m not EDC-ing a refrigerator compressor, just to have a “wow” light!
When I think of the materials used, I tend to worry first about electrical conductivity, as the body of the flashlight is our path to Ground, usually. That has to account for part of the decision, as the LED won’t generate heat without that current. That’s why I think Brass and Copper are so popular. (If anyone doesn’t already know how to tin these for soldering, LMK.)
That same Ground Path is our “Thermal Via” (to abuse a term I learned from Cree). “Via”, of course, just means “Way”, so the pathway for heat is all you guys are talking about here.
Where does the heat go?
If you stand your torch on end (or stick it in an outside purse pocket to illuminate the entire campground by lighting up your tent from the inside — thank you SWMBO), you “conduct” and “absorb” heat into the torch body, right up to the point where the entire system (the torch in air) reaches that equilibrium of the LEDs temperature vs. (Convection + Radiation). (At this point, you balance the pain of the $6.00 worth of batteries about to catch fire in front of you vs. the ice it will take to soothe the burns you’re going to have to take anyway.) You could hang enough fins off it so it looks like a Maico 500, but at some point, with post-apocalypse LEDs, that “sink” will overflow. Then, expensive electronics die. Worst of all, Customers complain!! Maybe your sink has enough mass to outlast your battery… Then your “Hand Torch” will need hands like mine (Wells-Lamont size “Giant”) to hold it. At some point, Elvis has to leave the building.
OTOH, if you wrap your fist around it, the blood in your veins works amazingly well to sink that heat away from the source in the bright end.
I’m just sayin…
If you have a firm grip on things, the specifics of the material will usually work themselves out.
To me, and I did do the math in another life, the thermal-loading characteristics of the “sink” become much less important than the carrying capacity of the “drain”. Yes, that beautiful Pellet-head cooler will fry your CPU, if you don’t sink the hot side.
Sink and “sink” are making me dizzy.
It just seems to me that the Thermal Magic of a “Hand Torch” (to pull out the old name) is to conduct heat as quickly as possible into the user’s hand. You can’t do that if you’re holding on to that heat in a big Thermal Ballast like a coin or a lump of copper…
There has to be balance. In all things. Too little mass to absorb(sink) the heat generated by the led(s) and they will fry. A large chunk with poor contact to the host can be overloaded by a sufficient number of LEDs and they will fry. A lot of LEDs, a big hunk of copper with good contact to the host but the host has insufficient surface area to radiate the accumulating heat they will fry. For a hand held flashlight the ideal would be to minimize each component along the heat path to what is needed at that point in the path for the intended purpose of the light. An extremely bright but short duration signal light might get away with less thermally radiant surface if its initial sink were large enough to absorb the thermal bursts. Likewise a dive light meant for use underwater would need fewer, smaller fins than the same light used in air. The ideal light would have the rate of heat generated by the LEDs equal to the rate of absorbtion of the sink equal to the rate of transmition to the host equal to the rate of radiation to ambient. The benefit of using copper as a heat sink is that its high thermal conductivity allows heat to move away from the led and then to the next layer much faster. In a static test, a large chunk works well but for a flashlight we want a high transmition rate to the next layer and this requires surface area not mass. For example, I can cut a stainless steel screw with a grinder holding the screw with just my fingers and though hot enough to raise a blister instantly at the cut, is only warm where I hold it in my fingers. When cutting copper bits, the entire piece gets hot much more quickly. So obviously, stainless steel would be a poor choice for pill material. The next bottleneck is from the pill to the host and the amount and quality of the contact surface area. All the fins in the world won’t do a bit of good if the pill has poor contact with the host. Lastly the fins should be located as closely as possible to the area where the pill contacts the host not elsewhere to make it look cool.
These are about nickel sized coins but they also have one ounce coins about 30.6mm in diameter and2.6mm thick. It wouldn’t take too much effort to lap these perfectly flat and have pretty much perfectly pure copper heat spreaders. If you can manage to solder the dies directly to these I know they’d be able to suck the heat from the dies as effectively as possible.
For the nickel size at a little over 2 bucks a piece they’re worth it, plus you get some nice collectible coins out of it too. The one ounce ones are especially beautiful to have even more so when you polish them up.
A friend of mine bought a cheap Chinese clone of this Trustfire Z5
While the host was well machined, the screw in pill was horrendous.
It was completely hollow, with the X-ML LED star sitting on about .25mm of it’s outer edge (virtually ZERO heatsinking capacity)
So what we did was machine a Copper disc on the lathe about 2mm thick and press fit (very tight fit) it into the hollow pill so the LED star can sit directly onto the copper disc.
We lapped both the copper disc and the back of the LED star to a mirror finish with the help of a completely flat polishing stone to ensure good contact across the entire contact area.
We applied good quality thermal compount between the LED star and the copper disc.
For test purposes, to ensure sufficient heat transfer was as expect, we ran an X-ML LED at 5 amps direct drive using a power supply.
Heat transfer through the body of the torch/flashlight was pretty much instantaneous. The body became uncomfortable to hold after about 3 mins of runtime.
We stopped the test after about 5 or 6 mins and were happy that the modification was a success, as we knew that we could easily run 3-3.5 amps into the LED without worrying about insufficient heat transfer in the future.
