Copper vs Aluminum vs Ect.

[TheDukeAnumber1]

Copper is only noticably better when air flow speeds are very high. We do not use finned heat-sinks in our hobby, rather use blocks of metal to absorb as much heat as possible before having to turn the laser off.

Copper does have an advantage, but the difference between aluminium is so small that it's not even worth it unless it's for a module. Copper is more expensive, harder to machine, looks awful and it corrodes.



What he is saying is relevant to this, this thread is about thermal dissipation after all. No need to be rude.

Wrong, when the heat is concentrated into a small spot like laser diodes, copper is better. And for duty cycle operations like handhelds, higher heat capacity copper is better. It's not a small difference either, comparing to aluminum copper runs 40% cooler, that's a pretty big difference.

He really wasn't very relevant, he was talking about active vs passive cooling systems. Everything I've said has been about materials for heat sinking diodes.

@Jander6442
I'm not saying we should all ditch aluminum, it's cheap and easy, just got tired of seeing here and there ppl say "Aluminum dissipates heat better" or "copper is barely better than aluminum" or "X metal is better for a heat sink for diodes"

 

[ARG]

Wrong, when the heat is concentrated into a small spot like laser diodes, copper is better. And for duty cycle operations like handhelds, higher heat capacity copper is better. It's not a small difference either, comparing to aluminum copper runs 40% cooler, that's a pretty big difference.

Read:

You're looking at the number wrong. Sure, one is twice the conductivity of the other, but they're already both very high. Take the reciprocal and you have thermal resistance. It's like comparing a 2mΩ battery spring contact with a 1mΩ battery spring contact. You'll see zero performance difference between the two.

 

[TheDukeAnumber1]

Read:

Comparing the volumetric heat capacities of Al and Cu shows that when energy is applied to similar sized heat sinks, for every 1°C copper rises aluminum will rise 1.4°C, and similarly for every 1°F copper rises aluminum will rise 1.4°F. Ignoring thermal conductivity and comparing duty cycles, copper will run 40% cooler than aluminum and give an overall +40% to runtime. When you consider copper's thermal conductivity that's about twice that of aluminum, copper is much better suited to protect laser diodes from heat than aluminum.

40% comes from volumetric heat capacity, not thermal conductivity.

 

[pullbangdead]

He really wasn't very relevant, he was talking about active vs passive cooling systems. Everything I've said has been about materials for heat sinking diodes.

No, you clearly didn't get my point.

My point was the difference you'll see is small, and it could be different in different hosts/applications/etc. And if you want to make a real difference, then you need to expand your arbitrarily-narrow set of conditions. You're also leaving out some very important variables. Namely: weight, price, corrosion, aesthetics.

You say that for Z, x is always better than y.

I say that if you actually, really care about Z, then you already missed the point because x and y aren't going to make that much of a difference. If you really want to help Z, then forget about x and y, and start looking at p and q.

 

[TheDukeAnumber1]

Lol, alright I do get what your saying, but only speaking of performance and taking out (weight, price, corrosion, aesthetics) within reason of course, copper stands out with the best performance.

 

[Things]

Wrong, when the heat is concentrated into a small spot like laser diodes

Unless you're bonding the die to something directly, your bottleneck is the dissipation the die has into it's mounting plate, not from the diode to module, or module to heatsink.

I love it when people go on about wanting long lifetimes .. then run them at double their rated power level

 

[TheDukeAnumber1]

Some things are easier to control than others. I think most of us know lifetime reduction is assumed when overrunning the current.

 

[flare09]

So. I think this can be settled if someone takes a nice flir thermal camera and compares the same diode, at same current in each raw heatsink. And measures the amount of material that heats in so many seconds, then cools in so many seconds.

Yes copper can absorb more heat, faster, but i thought the most desirable effect is the ability to absorb and release the heat faster. Which I thought, is why aluminum is better.

Plus copper builds up an oxide coating with raw uncoated metal, which does not conduct very well. The uncoated part is what will let heat out faster. And since the metal is constantly under heating and cooling stress, it speeds the oxidation.

 

[Cyparagon]

heating and cooling stress does nothing to malleable metals like Al and Cu, especially not at the measly 20C changes you'd expect in a pointer. Hotter metals do oxidize faster, but again, 20C above ambient has virtually no effect on this.

Diode efficiency varies. A better control would be to use mountable resistors and bolt them to the heat sinks instead. I've got a FLIR and other appropriate equipment if someone wants to supply identical Cu, Al, and brass cylinders. :beer:

 

[flare09]

Holy crap you own a flir?? :drool:

Yeah that is true that 20C isn't much difference, but overall copper is more prone to oxidation and is essentially inevitable. Especially down here in humid Texas.

Since we are own the subject, what does everyone think would be better, a TEC on a water block, or an attempt at a traditional heat pump system, higher pressure smaller tubes opening into low pressure area to be cooled, then repreassurized to pull heat off in a homemade radiator.

Or shall I start a new thread :/

 

[Cyparagon]

That REALLY depends on what you're cooling. A compressor system for a laser diode is colossal overkill, while cooling an ion laser with a TEC would be useless.

Phase change systems are better for heavy duty while TECs are better for smaller jobs.

 

[TheDukeAnumber1]

Aluminum forms an oxide on the surface with pretty bad thermal properties and copper tarnishes, but honestly the oxide layers formed are so thin they don't appreciably affect heat dissipation at all. Yes bring in the flir op:

 

[flare09]

Well. Its not exactly a phase change system, the pump im using is about the size of a quarter. So there isn't really that much compression. It would be really mild changes, and I would be running distilled deionized water with a little isopropol alcohol to lower the surface tension so the water really sticks to the walls.

I am cooling three diodes. An m140, mitsu 300mw and a dpss~200mw with fairly long duty cycles

 

[Things]

So not really a phase change system, but water(and alcohol) cooling

Phase change cooling generally refers to a refrigerant system, and if you can fit one of those into a pointer, I'll give you $100

 

[flare09]

So not really a phase change system, but water(and alcohol) cooling

Phase change cooling generally refers to a refrigerant system, and if you can fit one of those into a pointer, I'll give you $100

haha be careful, i might just accept that challenge and add a few more months to this project im working out a totally custom host, and welding is coming into the picture now. it was all going to be threads and o-rings but my machinist is wanting better seals.

i can get a little bit of pressure change from the tube sizing, but considering the size of everything...would it be negligible? if i add the tec, i would have to run it while the pointer is on, which means adding more batteries, upping the host size overall. otherwise its just an insulator not running. so there are some trade-offs

 

[Things]

IMHO anything but a normal heatsink is overkill for a pointer. If you're going for diode life and/or long runtimes, lab modules are a lot more suitable.