I want copper copper copper! (well... maybe not)

[sinner]

I've often thought that the whole aluminum is better at giving off heat vs. copper was a load of wash. However, I think the "truth" in the statement comes from the fact that the copper can store more heat due to its mass and SHC, so it doesn't heat up as quickly, and therefore it has more "heat inertia." The result would be that the mass of copper needs to absorb more heat before it creates a higher temperature gradient at the fins. Aluminum can't store much heat, so it heats up more quickly, which means there is a higher temperature gradient at the fins.

Ultimately, cost aside, I think using copper would be the best material for any heatsinking short of exotic materials, because it can both store heat, and transfer it. Aluminum would still be my choice as it cheap, conducts well, doesn't weigh much, and is easy to work with. It may also be good for having the host heat up more for a higher temperature differential at the surface.

Precisely! We do have to consider the practical side of it so it always comes down to copper and aluminium.. And using either of them with greater machining skills and designing is the better way to go..

 

[TheReddish]

So gold electroplated on copper is ideal? then if you wanted to go overboard just a tad? how thick do you need the plating?

 

[Dogs Best Friend]

Too bad diamonds have such a low specific heat...

How do you drill a 5.6mm hole in a diamond?

 

[Blord]

Find some carbon powder and press it around the diode. If you press carbon enough it will turn into diamond.

 

[Bionic-Badger]

So gold electroplated on copper is ideal? then if you wanted to go overboard just a tad? how thick do you need the plating?

The copper heatsinks I've seen for computers just use a thin layer of nickel for plating on copper. Gold would be overkill, and such a thin layer wouldn't make any difference.

 

[Cyparagon]

1)Thermal conductivity is the measure of how quickly an object can move heat from point A to point B.

Sorta. Let me compare thermal properties to electrical properties.

Like electrical conductivity, thermal conductivity can be thought of as the reciprocal of the resistance. If a metal has higher conductivity, it has less resistance to heat flow.

The current and the resistance will affect the voltage drop across a length of wire. Similarly, the heat flow in a heat sink (a function of the energy created) and the thermal resistance will affect the difference in heat.

When people propose using copper for a piece as small as an aixiz module, it is like proposing we use 10AWG for wiring laser diodes. The resistance is already low enough in standard supplies that any drops (in the wire or the heat sink) are already very low. Spending much more money (whether on a copper heat sink or 10AWG lead wires) will get you a miniscule performance increase at best.

Taking an equation from here. averaging the 5x2mm inner cylinder and the 12x8mm outter cylinder as a rough approximation of transfer area gives ~166mm² or 166*10^-6m², 6W for dissipated heat (4.5Vf*1.8A - 2W emitted), and material thickness of 3.2mm (0.0032m), I get

∆T=6*.0032/(k*166*10^-6)
∆T=115/k

where ∆T is the temp difference and k is the thermal conductivity in W/(m*K). Brass is 109, so the temperature difference between the inner and outer parts of the module should be somewhere around 1°C which is trivial. Why all the time/money/effort to bring that down to 0.2°C? The junction between the module and the heat sink will be much, much larger. The resistance to air will be larger still.

The back left wheel is squeaking just a tiny bit, and you guys are greasing the piss out of the front right wheel.

 

[Bluefan]

This thread really grows fast. It really surprises me that it took 4 pages for people to realize the whole "copper for transfer aluminum for dissipation" is nonsense, transfer and dissipation all comes down to the heat conductivity. The volumetric heat only dampens the response of the system.

I think we underestimate the significance of the conductivity to the air of most laser pointer hosts, but that's just my guesstimate.

To do this correctly would be to calculate the response of a system with the right differential equations taking into account the different materials, interfaces and geometry.
A numerical simulation could model the volumetric heat capacity as capacitors and the thermal conductivity as resistors, of course including the interfaces. The problem is finding the right simulation values without measurement, and the complex shape can make things overly complicated. Not to mention interpreting the results and getting the best design for a specific duty

I'd say copper if financially possible as the thermal resistance within your heatsink also plays, and to prevent things like 3 min on 15 min off duty cycles. If copper is too expensive go for aluminum, it's also easy to work with.

edit: this and what Cyparagon posted.

 

[Wolfman29]

I think Cyparagon brings up a good point we have all been neglecting... junction temperature differences :\ Would that imply that a press-fit gold-plated, copper heatsink would be better? That is why they gold plate diodes, after all - it's a soft metal, so it can more easily transform to fill in the air gaps when you press-fit it into something. So, theoretically, if we had a tight enough fit for a gold-plated copper heatsink, then the gold would mold into the gaps and leave us with a lower junction temperature difference?

 

[Sigurthr]

The back left wheel is squeaking just a tiny bit, and you guys are greasing the piss out of the front right wheel.

Thanks for the detailed explanation (I was being overly simple) and also for the amazing quote. That gave me a chuckle. With the math there though I've got to say you're right!

Now if only we had some highly (and deeply/thin) finned hosts...

 

[Wolfman29]

The problem with fins is that they make a host less durable (especially if they are thin and tall). A lot easier to break.

So couldn't we just drill a lot of through holes vertically through it? Then it would have higher air flow through it (especially when moving) and it would have more surface area, too. Further, it may make it easier for fans D:

 

[Bionic-Badger]

I think Cyparagon brings up a good point we have all been neglecting... junction temperature differences :\ Would that imply that a press-fit gold-plated, copper heatsink would be better? That is why they gold plate diodes, after all - it's a soft metal, so it can more easily transform to fill in the air gaps when you press-fit it into something. So, theoretically, if we had a tight enough fit for a gold-plated copper heatsink, then the gold would mold into the gaps and leave us with a lower junction temperature difference?

I always thought the gold plating was simply for corrosion resistance, as most laser diodes are not pressed into position, and you could always just mold other material around the diode (like in sleds).

 

[Wolfman29]

I suppose. Someone brought it up earlier though and it makes sense. Gold is a VERY malleable substance, so with a thick enough gold-plated layer, it should "fill in" gaps and thus make better thermal contact.

 

[Flaminpyro]

Yep the gold thing has been spoken about

Yes you are correct in your thoughts about the junction and that is why they gold plate the copper in expensive laser systems, the gold is softer and acts like thermal paste between the gold plated LD die and the copper of the heat sink.

And again you are right "O" rings are no way to take up the space inside of the host when a sloppy heat sink is made, really a correct heat sink for the host should be made.

If the heat sink is made correctly to begin with it will be within .001 to.0005" tolerance and fit in the host like your hand in a glove !

When you screw the bezel ring down it should grab the heat sink so it is held tight with a full metal to metal contact, there should be no insulating rubber inside the host in contact with the heat sink if the heat sink is made correctly !

This is the difference between mass produced $9.00 heat sinks like some make and custom heat sinks like I make

good you can get right now, The best takes a little longer and costs a little more

I suppose. Someone brought it up earlier though and it makes sense. Gold is a VERY malleable substance, so with a thick enough gold-plated layer, it should "fill in" gaps and thus make better thermal contact.

 

[tsteele93]

The back left wheel is squeaking just a tiny bit, and you guys are greasing the piss out of the front right wheel.

I disagree, I think the back left wheel is more than squeaking. We are trying to increase duty cycles. We know that we cannot run some of these lasers indefinitely without overheating so the problem we are attempting to address is real and worth addressing.

I don't know whether the copper modules make a big enough difference or not, but I feel better with a material that will conduct as much heat away from the diode as possible and toward the sink sitting closest to my diode. If I am over engineering or overspending by a few bucks per build, I'm ok with that.

I also want as big of a copper sink I can fit in the host and I'd like to look into hosts that would let me use nice finned head for a laser as well.

There is nothing wrong with shooting for lower diode temps (and the accompanying increase in diode life) and trying to increase duty cycles or eliminate them altogether.

That is more than just a wheel squeaking a tiny bit!

 

[tsteele93]

IWould that imply that a press-fit gold-plated, copper heatsink would be better? That is why they gold plate diodes, after all - it's a soft metal, so it can more easily transform to fill in the air gaps when you press-fit it into something. So, theoretically, if we had a tight enough fit for a gold-plated copper heatsink, then the gold would mold into the gaps and leave us with a lower junction temperature difference?

Yeah, but a good thermal paste is probably more at our price point.

Thanks for the detailed explanation (I was being overly simple) and also for the amazing quote. That gave me a chuckle. With the math there though I've got to say you're right!

I want to look at that math more. It is easy to have a small error in your model that affects your outcome.

Now if only we had some highly (and deeply/thin) finned hosts...

Do an eBay search for flashlight diode heat sinks. They are out there and no that expensive. The trick is making them look ok and work with a nice host.

 

[pullbangdead]

:: puts on materials scientist hat ::

Everybody is talking WAY too generally here. In general, for a heatsink you want high thermal conductivity and high heat capacity. You want both. Like anything else in the world though, you can't have both, so you have to make an engineering decision: higher heat capacity, or higher thermal conductivity.

And that decision will change in every application. Heat load, geometry, environment, it all changes the calculus, and you'll have to calculate it or measure it for YOUR application to find the best one. You can wave your hands around and say "heat capacity is more important" all you want, but until you either calculate it or measure it, then you're just blowing hot air.

It's all relative. You're comparing relative amounts of 2 different things in general terms and trying to draw specific conclusions, which you just can't do.

The first-order visualization tool to choose a material for a given application is called an Ashby chart. This problem would have thermal conductivity on one axis, and heat capacity on the other axis, and you would want a material from the top right-hand corner (high in both), but would have to choose something else based on other things because there is no perfect material. So calculate whether heat capacity or thermal conductivity is more important, then include cost/ease of use, and you'll start getting close.

But as of now, this is a thread full of hand-waving arguments with no real data, because every application is different.