Heatsinking - is thicker better?

[CM]

That doesn't explain the large temperature differential I'm seeing on the L4 vs. the Arc or McLux bare aluminum flashlights. Maybe their definition of "good thermal conductivity" may need to be quantified and compared to bare aluminum to bare aluminum contact.

CM

 

[BentHeadTX]

CM,
It could be the threads have a tighter tolerance on the McLux with the L4 having a more loose tolerance. The tight threads would make a better thermal junction and transfer the heat to the body faster. Just speculation on my part...

 

[Rothrandir]

[ QUOTE ]
Rothrandir said:
i...why not have as much mass as seems reasonable too?...

[/ QUOTE ]

i didn't say as much as "possible", i said as much as "reasonable" /ubbthreads/images/graemlins/wink.gif
for that point, i was speaking mainly in regards to maglite mods, where there is sufficient room to plop in a bigger heatsink, but it also applies to smaller mods. if the room is there for more mass, then why not use it?
i can understand increased weight being a concern, which is another reason the hotlips rely's primarily on surface contact area, while retaining enough mass to conduct the heat efficiently to the body and not weigh too much.

if designing a custom light from scratch, the best thing one can do is find a good thermal path to the outside, either by eliminating thermal joints, placing fins or ribs on the body, enlarging mass, etc. some of my designs do all of these, some of them rely moreso on mass, or ribs, or joints, etc.
as i said before, you need to base your heatsinking on the application

 

[Doug S]

[ QUOTE ]
BentHeadTX said:
CM,
It could be the threads have a tighter tolerance on the McLux with the L4 having a more loose tolerance. The tight threads would make a better thermal junction and transfer the heat to the body faster. Just speculation on my part...

[/ QUOTE ]
Ditto that. Also, using some thermal compound on the threads can greatly reduce thermal resistance.

 

[Burnt_Retinas]

Kiessling,

Spot on. In short, mass/contact surface area (without going overboard) to absorb and transfer, surface area to dissipate. Got to get rid of that heat to atmosphere!

Good topic /ubbthreads/images/graemlins/smile.gif I see too many posts with finned heatsinks WITHIN the body of the torch, or too little heat transfer at all. Others rely on mass alone - useless for extended run-times.

*edit* Rothrandir,

Hotlips IMHO meets all criteria. It's effective without overkill, as light as possible (yet meeting heat transfer requirements) and transfers heat effectively. That's all that is needed, and it does it well. I'll keep buying them for any Mag mod.

Chris

 

[Kiessling]

hm ... that body/head heat difference of the L4 vs. the McLux at approx. the same load ... I do not have the impression that the threads of my McLux have less tolerance (=tighter) than my L4, I'd say it is more the opposite way ... yet the body of the McLux gets as hot as Chili while I can hold the L4 without any problems.

Could it be that the slightly higher load (660mA of the L4 vs. 750mA in the McLux) makes the batteries generate more heat, thus making the body where they are housed warmer?

bernhard

 

[Rothrandir]

burnt, i'm offended! /ubbthreads/images/graemlins/frown.gif

hotlips certainly is overkill /ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/twakfl.gif
just not to the point where the overkill sacrifices something else, size, weight, ease of use, etc /ubbthreads/images/graemlins/wink.gif

 

[Doug S]

Most of the old timers here have seen this thread but for the new folks, here is a thread with some discussion of heat sink issues. It gives you the tools to calculate actual heat sink performance and includes numerous examples.
Thermal stuff.

As an example of the thermal resistance of the pathway from the emitter to the flashlight body, if the pathway is pure aluminum with a crossection of 1 square cm and 2cm length, the thermal resistance in this 2cm long path is only 0.8 C/W. This is very small compared to the resistances in the total path, for example the die to bottom of slug for a 1W white emitter is 15 C/W.

EDIT: Recalculated the thermal resistance. While sitting on the crapper this morning I realized that I had used the wrong value for the thermal conductivity of pure aluminum in my initial calculation.

 

[AilSnail]

Some say that anodyze radiates heat better than alu. Convection is probably more important than radiation though, and the convection depends on the surface temp, in other words conductivity of the system. Conduction also rules the equation in flesh contact.

Kiessling, remember that conductivity is proportionate to the cross-sectional area, IE for conduction to the the tail it is proportional to the tube walls, AFAIK. Probably more important though, is the joint between the head and tail, like you said.

I don't know if it'd be "worth it", but i keep envisioning an inner shell and heatsink of silver glued inside a one piece body/reflector/head of aluminium. /ubbthreads/images/graemlins/smile.gif

 

[jtice]

I dont know why, but anodized metal trasfers heat to the air better. /ubbthreads/images/graemlins/icon3.gif

Thats why most computer heat sinks are black HA.

 

[unnerv]

I would imagine that anodizing imparts somewhat of a texture to the metal which thereby increases surface area exposed to the air.

 

[Wylie]

Sorry if this is redundant as I didn't take the time to read everything here.
I have just schooled myself on this a little and think the thermal path is the most important issue to consider. I see a lot of heat sinks sending the heat to the most vital parts of the flashlights. These Mag heads do have a bunch of alluminum but al only transmits heat half as well as copper and then the heat has to jump the threads to sink to the head so the most direct path is toward the batteries and the hotter the batteries get the less they are going to produce. This goes the same for the electronics and all that is housed in the battery cylinder. If it were possible to create a heat sink path by using a better transmitter of heat like copper into the head of the flashlight connecting to the head of the flashlight creating a longer path back to the battery cylinder I would think this would be far ahead of the game.
I just built something to test this with but still have the real testing to do and batteries to buy. From first inspection I can tell you I think I have done very well with what I have done.
Here you go, take a look. http://www.candlepowerforums.com/ubbthreads/showflat.php?Cat=&Board=UBB14&Number=416982&page=0&view=collapsed&sb=5&o=&fpart=1

 

[Burnt_Retinas]

Rothrandir,

OK, Hotlips is overkill by most peoples definition /ubbthreads/images/graemlins/grin.gif

For me though, it is 100% meeting all objectives perfectly with no compromise on effectiveness and without sacrificing size, weight etc. as you mention. You did well with that design /ubbthreads/images/graemlins/smile.gif

*edit* Re black anodizing of heatsinks. I think you'll find it's the black that does the trick, not the anodizing as such. Black radiates heat better than any other color. There's heaps of info out there, just do a google search if you want more info. For fan forced applications such as CPU, then it becomes a moot point.

Chris

 

[Kiessling]

so ... should the McLux get hotter than the L4 because it has a thicker body? that would mean that the thermal conductivity of the body needs a lot more mass than I thought to carry the heat away effectively. that would also mena that I was wrong in my first post and that long-running lights would need a lot of mass, too ...?????
bernhard

 

[Doug S]

[ QUOTE ]
Kiessling said:
so ... should the McLux get hotter than the L4 because it has a thicker body? that would mean that the thermal conductivity of the body needs a lot more mass than I thought to carry the heat away effectively. that would also mena that I was wrong in my first post and that long-running lights would need a lot of mass, too ...?????
bernhard

[/ QUOTE ]
No simple answer since there are other factors that may not be equal such as power input and surface area. If all other variables are the same, the peak surface temperature would be lower on the thicker bodied flashlight but the minimum surface temperature would be higher [at the point furthest away from the heat input in terms of thermal resistance].

 

[Pinter]

Is it a good idea to engrave paralell lines on the body, thus increasing effective area of dissipation?

[Why black]
Faq section from http://www.radianstore.com/support.html

Q: What difference does the color of the heatsink make?

A: In natural convection a black or dark colored heatsink will perform 3% to 8% better than an aluminum heatsink in its natural silverish color. This is due to the fact that dark colors radiate heat more efficiently. In forced air applications the color does not add any performance due to the increase in convection.


Q: What surface treatment should be applied to heatsinks?

A: Many heatsinks are black or color anodized. This is because of two reasons. 1.) It allows for the dark color to enhance the radiation effects and 2.) It protects the aluminum in harsh environments. Salt exposure is a good example. It is unnecessary in most electronic applications other than natural convection.

Gold or yellow chromate is also used to provide surface protection. It adds little or no radiation benefits.

Unfinished aluminum is typically used in most consumer applications where thermal radiation and harsh environments are not factors.

 

[Burnt_Retinas]

Pinter,

Any engraving/texture will increase surface area and if done right increase cooling. You'll need to make sure however that the grooving does get air circulation and make sure air can flow into the groove, or at least have turbulence with some flow, for this to work effectively. You may however reach a point where you get diminishing returns ie as you remove metal you also remove thermal conductivity, another critical factor. Convection or fan forced will probably require different solutions.

Chris

 

[RGBrainbow]

Hi,
just a question:
Is someone here using heat-conductive paste for the head/body threads?

joachim

 

[Burnt_Retinas]

RG,

Yes. I use it on MR-X, but my calibrated fingers can't decide if it made much difference. I figure that there's already sufficient contact in the Mag D threads, with the number of threads it has, that there's not much to gain. I used it anyway....it helps me sleep better at night /ubbthreads/images/graemlins/grin.gif The story could of course be different with other torches.

The stuffs so cheap, why not do it anyway.....and sleep easy.

Chris