Heat being the enemy

[Track Terror]

Since it seems like heat is the enemy of the 50/90 why doesn't a custom flashlight builder build a light that has nicely machined holes. It doesn't have to be perfectly round holes and of course the builder could come up with his own variation of heat elimination. Not every flashlight need to be designed to work fine after being dropped to the bottom of the ocean. In fact I have only had water hit one of my flashlights, ever.

The end user who buys these would be well aware and would most likely have another flashlight to use if water was going to be an issue. They would also know that there might be a little more cleaning maintenance that would go with an "open" flashlight. It seems like there is so much heat retention because the battery and head are sealed keeping the heat in and making the host get rid of all the heat through basic heat sinking of the host instead of helping the host with some well positioned heat extracting holes.

 

[Tedfs]

Using the host as a path for heat transference is better than just air cooling the emitter.

More surface area = better heat transfer. That's why you see some designs with finned heads.

While it might seem logical that holes would aid in heat dissipation, unless there is a way to force air in or out of the holes, all you're really doing is taking away surface area from the host. Which is not what you want.

 

[Rod911]

I'm starting to get back to my previous hobby of PCs and watercooling, so this post reminded me of articles I read back in the day about how to cool stuff.

Opening up holes at the head of the light, or any where else really, is not a good idea. Air does not conduct heat away well at all as noted by Wikipedia articles here and here.

What you'd want to do is create the most efficient thermal path from the head to the body to take the heat away from the LED. From a practical sense, I can only see this being done passively at this stage (ie. the use of copper/brass/aluminium slugs contacting the [insert metal type here] head/body of the light). Others have noted here that heatpipes could be used to transfer the heat. However, when I picture the possible designs in my head, again it seems impractical. I do would love to be proven wrong though.

 

[jamesmtl514]

what these ^^^ guys said.

You want to add material to absorb the heat and then dissipate it. Also you want to use materials that are very good at absorbing the heat. Gold, Silver, Copper. What usually dictates the material of the heat sink is cost.


Like a heat sink on the CPU found inside your computer.
Pic found with this article (http://en.wikipedia.org/wiki/Heat_sink)

When money is not an issue, go with the best, as Maclaren did when making the ultimate sports car. Use 24k gold to get rid of the heat.

http://objetpart.blogspot.com/2010/02/part-of-day-mclaren-f1-engine-cover.html

 

[Jay R]

as Maclaren did when making the ultimate sports car. Use 24k gold to get rid of the heat.

As far as I know, the gold in the McLaren was to reflect the heat.

 

[Narcosynthesis]

Air is normally a fairly poor transmitter of heat, especially in something that won't have much airflow like a light.

What you want is something that will absorb the heat from the emitter, then distribute it away from the emitter. So havign a metal body means the metal nearest teh emitter absorbs the heat output from it, then the heat is slowly distributed through the body to the cool outer surface where it can be radiated out to the surroundings much more efficiently (as the outer body has more of a surface area to spread the heat out over).

As a wee primary school experiment, try taking a mug of boiling water, then place a metal and plastic spoon into the water. After a short time you will feel the metal is noteably hot, which means the metal has absorbed x amount of heat from teh water, and is then transferring it to both you by touch and radiating it into the air. The plastic on the other hand is much less efficient at heat transfer, so won't absorb the heat and will stay cooler, and the water will retain its heat slightly longer.
So if you want to cool your hot drink quicker, put lots of spoons in the cup... :thumbsup:

 

[Walterk]

It is no wild idea at all. Very KISS actually.

You can make the common cilindrical heatsink a little larger, and bore right through it from several angles to the axis.
Choose the borings right and you have a venturi effect, so it would vent itself in all situations. The sink is ventilated from the core, closer to the heatsouce.
A litle more then usual attention to the potting and it is watertight.
And not that much different from grooved heads, just creating surface exposed to air.

This is the same way insects like ants and flys exchange heat and oxigen to their body fluids. So its a tested concept.

 

[Track Terror]

It is no wild idea at all. Very KISS actually.

You can make the common cilindrical heatsink a little larger, and bore right through it from several angles to the axis.
Choose the borings right and you have a venturi effect, so it would vent itself in all situations. The sink is ventilated from the core, closer to the heatsouce.
A litle more then usual attention to the potting and it is watertight.
And not that much different from grooved heads, just creating surface exposed to air.

This is the same way insects like ants and flys exchange heat and oxigen to their body fluids. So its a tested concept.

That is more of what I was thinking. I totally understand the reason behind heatsinking and how it works but the airflow issue is a good point since there wouldn't be much if any.

It just seems like the minimal loss in heatsink material (holes) would be more than made up with not having the heat in a completely sealed area. I would think that some of the heat would find it's way out of the holes but I have been wrong before.

 

[jorn]

Using the host as a path for heat transference is better than just air cooling the emitter.

More surface area = better heat transfer. That's why you see some designs with finned heads.

While it might seem logical that holes would aid in heat dissipation, unless there is a way to force air in or out of the holes, all you're really doing is taking away surface area from the host. Which is not what you want.

Holes adds surface area Unless its adding holes to a paper or something else with a very two dimentional shape.
Air will flow trough it, because heated air rise.

 

[Noctis]

Heat is indeed the enemy. I think the point is to try and find a good balance between current draw, thermal mass, and thermal transfer.

Thermal transfer seems relatively simple to me. Leave as little empty space as possible in the head by filling it with copper tape(or with a nice fat copper heatsink). Then use a thin layer of Arctic Silver 5 to help move heat from one metal contact point to the other(from the module to the head).

Thermal mass seems to be the most important part. Aluminum would be like a sponge with large holes that looks like swiss cheese. Copper would be like a sponge of the same size, but made up of that memory foam they use in pillows. Point is, the copper will hold a lot more "water" despite being the exact same size. A copper heatsink would soak up the immediate heat from the emitter fast enough to keep it from blue shifting or dropping in lumens due to heat. Whereas a hollow brass pill would "overflow" in a matter of seconds.

I can't be 100% certain, but I suspect that if the Moddoo Triple XP-G drop-in had a solid copper heatsink rather than aluminum, that alone would be more thermal mass than the entire host itself.


I'm not sure if this is a point most people consider, but it also takes time for heat to move around and dissipate into the air. While I could stick a Malkoff M61 into an FM 1x26650 host with a Finned Bezel head, that would be complete overkill as the heat generation wouldn't even be noticeable after 4 hours. On the other end of the scale, driving an SST-90 to 9 amps in the same host would probably fry the emitter inside of 10 seconds since there simply isn't enough time for the fins to dissipate the heat.

Though if I had an SST-50 mounted on a solid copper heatsink in the same host driven at 4.8A, I could probably run that for a good 15 minutes before heat becomes an issue.

You really have to play around with everything in order to beat the heat:devil:.

It seems like heatsinks are readily available for Mags, but I still can't believe there aren't any for smaller lego lights(D26/36) considering how well the Moddoo drop-ins performed.

 

[asdalton]

Holes adds surface area Unless its adding holes to a paper or something else with a very two dimentional shape.
Air will flow trough it, because heated air rise.

On such a small scale, natural convection is negligible compared to conduction and radiation.