Gold foil to maximize diode-heatsink heat transfer?

[Minamoto Kobayashi]

Hi. You have forgot some points.

- The thickness: between the copper diode heatsink and the alumium heatsink there is
a really thin gap. I have just done a little experiment with a domestic alumium sheet: it's enough a single turn to obtain a nice pressure between the walls. I think could be a 1/10mm thickness . You must consider that the heatsink has a round shape, so we must calculate 1/10 mm x2 --> (O)

- The malleability and the ductility: aluminum is too harsh! When You have done a coil, and leave the fingers, the coil roll out, because do not rest in place.
Gold is smooth, soft and pliable, and You can obtain more pressure between the walls. Is also lubricant, so it is easy to remove the diode with right pressure from the back.
Also, if I need to do more wraps around the diode, due to the thickness of the gold foil, surely the result will be better to have only air between the heatsink walls! More wraps = more pressure between the walls = less air between the sheets.
When I have the gold sheets in my hands I will do some experiments. I think 2 or 3 gold wraps could be more than enough to obtain a good pressure.

 

[COMMANDR]

Subscribed

 

[Sigurthr]

Also on the feasibility of aluminium foil;

All metallic aluminium exposed to the atmosphere forms an invisible coating of alumina - Al2O3 - Aluminium oxide. Alumina is a thermal insulator much like ceramic or porcelain with a very poor thermal conductivity (roughly 1/10 that of aluminium metal!).

Also, aluminium pressed against copper will undergo galvanic corrosion over time if exposed to the atmosphere. This will essentially degrade the thermal and electrical junction of the two metals as well as mechanically degrade the joint to the point that separation may be impossible without destruction of the parts.

 

[Minamoto Kobayashi]

Oh my God ... all the builds today are made with an heatsink copper pressed against an aluminum heatsink !! :thinking:

 

[Sigurthr]

Haha, it is very common practice. It takes a long time for significant galvanic corrosion to take place. If you live near the sea it might be worth storing your lasers in an airtight container with a desiccant to slow the corrosion as the high humidity and aerosolized saline greatly accelerates the corrosion.

Gold foil as an intermediary will essentially stop any galvanic corrosion.

 

[BShanahan14rulz]

You'll notice that the diode's heat spreader is almost always gold colored. Maybe even gold plated. I would infer, then, that there is some merit to coating things in gold :beer:

 

[djQUAN]

I may be wrong but I think having it gold plated makes a stable surface to solder the die onto. It also prevents corrosion which could contaminate the inside of the laser diode case.

 

[WizardG]

Use indium foil. It's far more malleable than gold or aluminum. Cheaper too! Indium foil is so malleable that any excess will 'squeeze' out like thermal paste as you tighten things together.

 

[Cyparagon]

All metallic aluminium exposed to the atmosphere forms an invisible coating of alumina - Al2O3 - Aluminium oxide.

Yeah, and it's 4nm thick - a few dozen atoms. Even if it was a thousandth the transfer rate of aluminium, the 4nm makes it insignificant. It's like running 10km, then crab-walking the last 2 meters.

Op: If you can afford a sheet of gold, why not eliminate the boundary altogether, and use a single block of metal?

 

[Sigurthr]

It's like running 10km, then crab-walking the last 2 meters.

 

[BShanahan14rulz]

quan, nah, that makes more sense, and is likely the primary reason behind the coating.

I think there are easier ways to address the thermal path issue. We already have direct-press heatsinks, and I've even seen some 12mm module heatsinks that are c-clamp style, I think. Much better than the crappy setscrew.

 

[TheDukeAnumber1]

Due to machining limitations, no two solid surfaces will ever form a perfect contact when they are pressed together. Tiny air gaps will always exist between the two contacting surfaces due to their roughness.

Mating metal won't fill those gaps as good as a paste will. I don't know if gold is ductile enough for the air to be pushed out but in this application it seems like it would be challenging. IMO... paste >> air = good enough

 

[Minamoto Kobayashi]

The thermal paste won't work if the two parts haven't enough pressure between them.
Tom's hardware has done some experiments a lot of times ago with a couple of different paste brands, and the results was poor heat conductivity if the two parts aren't in high pressure (like a heatsink tight to a CPU).
When You insert a diode with copper heatsink into a aluminum heatsink hole, and then tight it with a screw, only a small portion of the opposite diode heatsink wall has a real contact with other wall. The rest of the walls do not have any pressure, and here the thermal paste work more like an insulator rather than a heat conductor ...

 

[TheDukeAnumber1]

The thermal paste won't work if the two parts haven't enough pressure between them.
Tom's hardware has done some experiments a lot of times ago with a couple of different paste brands, and the results was poor heat conductivity if the two parts aren't in high pressure (like a heatsink tight to a CPU).
When You insert a diode with copper heatsink into a aluminum heatsink hole, and then tight it with a screw, only a small portion of the opposite diode heatsink wall has a real contact with other wall. The rest of the walls do not have any pressure, and here the thermal paste work more like an insulator rather than a heat conductor ...

Try to be a little bit more technical or specific in your jargin, pressure won't change anything since solids and fluids are conventionally incompressible, and using that logic the aluminum is insulating the copper and you should use a direct press heat sink. The point is to fill the gap with something that isn't air, and how are you sure wrapping the module in foil will be better if tiny air gaps could be there?

 

[Minamoto Kobayashi]

Hi.
What I have wrote is not an inference of mine, it is the results of a couple of experiments and measurements of Tom's Hardware technicians.
With some probes they have seen that the thermal paste won't work if is not applied a certain pressure between the parts.

Then, when You put a diode inside the heatsink hole, and illuminate the back part of the diode with a torch, You can see more than 3/4 of the edge of the diode illuminated (it seems the moon at the end of its last quarter).
This means that most of the part of the copper wall is not in direct contact of the walls of the aluminum heatsink.
I agree that there could be some bubble airs between foils gold and the walls of the heatsinks, but surely the sum of the microscopic air bubbles present in the gold-copper and gold-aluminum contacts will be much less of the original air present around the heatsink copper without applying any gold foil.

 

[TheDukeAnumber1]

I'm not surprised by those results the module is physically closer to the copper in that case, but 1 pascal vs 100 pascals of pressure won't make a difference if the module is in the same position. That would be a good test to compare thermal paste vs the foil idea.

Just seems like your going down a more difficult path that may not even work, would be neat to see you try though. If you came up with a good method of fitting everything together it could be a good alternative.