Has anyone used copper wire for heatsinking?
I have all sizes, the wire that goes from electric pole to houses, 10-2 and 10-3 used to wireinside houses.
It ought to work well as long as you can keep it where you want it. A spot of solder ought to do that. Wrap as tightly as possible obviously.
Try to find wire that is soft and not "work hardened". As you wind on a layer, like on a p60 drop in, use the smooth side of a spoon to flatten and tighten it.
copper tape yes (with excellent results too), but no not round cross-section wire. I don't think it would work "great" being that it would not have sufficient surface contact. Heat conducts from one material to another through surface contact area + applied force. The more area + force, the better the heat conduction will be. Copper wire could easily provide sufficient material and your host design could apply external compressive forces, but without surface contact area it would be limited.
That being said, it will draw heat better than an empty pocket of air.
You could flow solder to bridge the air gaps, and that might help a little. I have never tried though. My (perhaps incorrect understanding) is solder is generally a thermal insulator, relative to other metals.
You can get 3M copper tape pretty cheap on ebay, that stuff is great with P60 modules. that would be my recommendation first.
The thermal conductivity of normal lead-tin solder is given as 50 W/mK.
This means a cube of a meter solder will conduct 50 watts if there is a temperature difference of 1 K (which is 1 celsius or 1.8 fahrenheit).
A 2.5 cm / 1 in cube that has to transport 4 watts will have a temperature difference of: 50 / 40 = 1.25 watt for every degree, so 0.8 degrees for every watt. 4 watts * 0.8 degrees is 3.2 degrees K or C, which is nearly 6 degrees F.
Remember this is with the cooling behind the pill. Usually the path of the heat is shorter than its width. And if solder is only used as filler its influence will be less.
Copper sits at 400 (8 times as good) while aluminum sits at 250 (5 times as good). The difference between copper and aluminum is less than half a degree F in this scenario. That is pure aluminum; alloyed may be a lot worse.
Air, however, is at 0.024: conducting heat 2000 times worse than solder! Oil, 0.15. Thermal grease seems to be at 1 - 3. So in effect soldering is much better than a close fit with some thermal grease (solder can be 12 to 50 times as thick to get the same performance).
Most of the thermal resistance will be outside though, unless there is a brisk wind.
Thanks for that info... and welcome to BLF.
I was thinking about filling pills of LEDs completely with solder..
Way to jump in Burlap
It's very time consuming but works well.
Just remember all P-60 host differ in size so by permanently modifying the drop in for a specific host it might never fit in another host even the same model host from the same manufacturer.
I've got thick tin-bars from tin-soldier stuff.. so I dont think it would use that much time. Should I combine tin and copper bars/wire or just fill them up with tin?
Also, what do you recommend for a P60 dropin? I've read about alloy foil and copper foil. There is cheap 0.5mm copper foil on ebay.. and how do you mod a dropin permanently? And another question, I want to put some pieces together thermally conductive but electronically insulated.. any good possibilty for that?
@Tad: Sorry for hijacking your thread! :D
The real question is how big of a deal is heat-sinking? Most of the lights we buy here go to non-flashaholics, who probably never take them apart except to charge batteries. Is there a big failure rate of the drivers or emitters due to lack of heat sinking?
Unless you have a driver that is pulling more than 2.8A from your battery, I seriously doubt it.
I used to pack my P-60's with copper foil, but I have gotten lazy and have not done that for awhile. Still no failures, but I only have one light that is driven over 2.8A. I think that a lot of electronics can handle temperatures that your hand finds 'too hot'. Just think of some of the things that are operating in your engine compartment, and never seem to fail because of heat.
The other big question is, how expensive is the item you are trying to heat sink? If its a $14 drop-in, why would you put $30 worth of time, effort and materials into it? Buy two spares instead.
But I think I will test this my theory, and go through an 18650 with a non-heat-sinked S/F light on "High", and see if it still functions.
Anyone want to place any wagers on the result?
Even if it won't die, you shouldn't expect 50k hour lifespan from the LED either.. It gets lowered significantly if LED gets overheated often.
A big advantage of good thermal conductivity is that inside and outside temperatures are almost equal.
If the LED is at its limit of say 70 degrees C (160 F), the outside of the flashlight would be at say 60 degrees C (140 F). Which is painful to the touch and leads the user to cool down the flashlight exterior or reduce the power.
I just packed a pill with USA copper pennies! (pre 1982). It was an absolute PERFECT fit. I had to use a C-Clamp to get them to go. I ground the face of the first one to mate with to a 20mm star.
Amen Brother.
I for one will not discount the importance of thermal pathways but I have experimented with wrapped/unwrapped P60 lights and my (very) unscientific conclusion is;
. . . it ain't worth the trouble.
My solution is to simply turn the light off high when it gets hot and it seems to get just as hot with or without being wrapped. I don't have a single wrapped light and the only failures I experience involve Foy feeding an emitter too much current. I killed an SST-50 that Erik fixed, only to then be killed again.
toomanybatteriesFoy
Keep in mind that a (for instance) XM-L is cree rated to perform at up to 150 degrees C. At full drive, there's about 25C between the emitter junction and the pill. Even if we assume that a non-wrapped P60 has a horrible thermal pathway of 50C from pill to exterior (and I'm very confident it isn't that much), then the exterior of the torch would be 75C before the LED was in danger of damage. You can't touch 75C for long. With wrapping or without, I can assure you that your hand will be the thermal limit - not the LED emitter.
PPtk
That is sort of what I was thinking...
It performs a little better if the junction is cooler. But 4 degrees will not make a visible difference.