Gold foil to maximize diode-heatsink heat transfer?

[Minamoto Kobayashi]

Yep, when I will receive the foil i will do all the possible tests!

The thermal compounds are not a good idea, since they insulate the electrical path between the parts, and with high heat they tends to move away.
It is also very difficult to spread the paste homogeneously, especially the latest hi-tech compounds like Prolimatek nano silver.
I'm tempted to do the same experiment with indium foil, but I do not know where I can buy it.
Moreover, I do not know if indium foils can be thin as a gold sheet. I'm working with 1/10 mm gold sheets.
Finally, the indium is not a good conductor at low temperatures, so I expect some high resistance in Winter season

 

[BShanahan14rulz]

He is trying to fix a problem that many folks encounter but don't realize. I get the feeling that a large number of builds are 12mm module + set screw design. He's looking for a simple way to increase the effectiveness of this poor design, rather than designing something new.

It makes a noticeable difference in P60 hosts, whether or not you wrap the LED drop-in with aluminum foil or not. I mean, noticeable difference! I'm talking turn on the light and almost immediately start noticing the flashlight get warm, compared to no foil where the light would dim before you notice it heating up.

But here, we are talking gaps approaching 1mm. With 12mm module, some folks who turn their own heatsinks can make one that fits their specific module better, but to be safe, they might use looser tolerances to guarantee all 12mm modules fit.

If you think about it this way: you can't change the gap size, and the gap is too big for heat to jump. Well, if there were tiny paths of direct contact where foil touches foil touches heatsink, it will conduct much better than air. And the more pressure you use, the more air will be squeezed out from between the layers.

Now, whether gold foil vs aluminium foil, that is another question, and the answer may also want to take into effect the economic cost-effectiveness of the solution. At some point, you will want to ditch the 12mm module + setscrew all together and go with 12mm interference fit of module to heatsink, or diode directly interference fitted to heatsink.

 

[TheDukeAnumber1]

The thermal compounds are not a good idea, since they insulate the electrical path between the parts

I wouldn't say they aren't a good idea, having thermal paste there is much better than nothing, and I don't really see the electrical insulating being an issue, that's what the case pins are for.

@ BShan

Are you talking about aluminum vs thermal paste in the P60's?

 

[BShanahan14rulz]

@TheDukeANumber1: My P60 anecdote was with regards to aluminum foil to fill in the gap, vs just leaving the gap as it is with air.

With regards to the paste, I think the paste would be better than filling the gap with air, though still suboptimal compared to having the parts pressed together with enough force that they are touching, with paste to fill in the air gaps. Main downside is it can be messy. Not like foil wrapping isn't a pain in the rear too, though. After wrapping it perfectly, you go to insert and the foil just bunches up, wasting the time you spent wrapping it :tinfoil:

One trick I use to help reduce air pockets a bit is to flatten the foil. I usually fold it up once or twice to make each wrap 2 or 3 layers thick, and to get the wrap's width right. After that, I use a plyable, flat card to squeeze all the wrinkles out of the foil until it's nice and shiny and wrinkle-free.

Obviously, it wrinkles up a bit when you wrap the drop-in, but I do notice that I can stuff a few more layers of foil into the gap when I smooth it out first. And if there is more foil in teh space, that must mean less air.

 

[Minamoto Kobayashi]

So, I have found this:

Pure 99.995% 4N5 Indium Metal Foil Sheet 100mm x 100mm x 0.1mm 7.31g #ET2 | eBay

I'm in doubt if buy the 0.05 mm, the 0.1 mm or the 0.2 mm version:
if it is too thin, I risk to do more coils to obtain the right thickness, but if it is thicker, I risk that the diode do not enter in the hole anymore.

Also, I do not know if Indium is a good electricity conductor.
Here is the list of the best metal electricity conductors:

silver
copper
gold
chrome
aluminum
iridium
tungsten
nickel
iron
platinum
lead

... and I do not see mentioned Indio!

 

[Sigurthr]

Indium has about 5.5x the resistance of silver and three times that of aluminium at 84nOhms/meter. It is a slightly better conductor than iron.

 

[Minamoto Kobayashi]

Interesting!
So if I use it around a diode where the current involved is about 1.8A at 7.4V I could not have any significant loss of power, right?

 

[Sigurthr]

The added surface area that is conducting power would probably negate any increase in resistance of the material. This is an example where geometry overpowers chemistry.

 

[Bionic-Badger]

This really seems like a bandaid that isn't going to do much. If you can see that air gap, your foil is not going to do much good because you're not going to have enough material to really fill that space in. It'll still just have other gaps where the foil is.

You ought to just toss the piece-of-shit aluminum heatsink and spend the money on more precisely milled heatsink, or a better drill bit and reamer.

 

[Minamoto Kobayashi]

First thing, the indium foil can do the work very well.
If I have a copper diode heatsink of 12 mm that will be put in a 12.2 mm hole
made into an aluminum heatsink, 2x 0.1 mm indium foils can resolve the gap very well. I do not see any other gap possible, if the thickness of the foil is the same of the gap.
Second, the extra 0,1 mm added to the original 12.1 hole diameter is due to the removal of an anodization from the hole of a well done Zaser head. The head is not shit, and I suppose even the Zaser creator, MrCrouse.

 

[amkdeath]

Badger is right, the foil will only divert your problems elsewhere. This, on top of the fact that Indium/Gold foil seems to be excessive, you can achieve what you want with some simple copper tape. Get a thin roll, and tape a couple of rounds around the diode housing.

Back to the diversion, however, the airgaps will still exist, they will just be smaller now. You can find electrically conductive thermal transfer compounds online, they have suspended Al particles within the material.

Good luck.

Best,

amk

 

[Minamoto Kobayashi]

Copper oxydize, while indium or gold not.
Besides this, a 0.1 mm air thickness is much worst than a reducted air gap: nearest are the metal surfaces, better will be the heat transfer. If I obtain a strong pressure with indium foil, that is really soft and malleable, I will eliminate also every air bubble.

 

[BShanahan14rulz]

I don't think anybody claimed that foil was the best solution, but without buying new parts, that's the best solution for this problem. Sure, you can fix any heatsinking issue by saying "Buy a better heatsink." He has the heatsink he has, and it's good for what it does. Small airgap or large airgap, it doesn't matter because the contact is still miniscule compared to a non-module build.

And what if someone made a heatsink specifically for to pressure-fit generic aixiz modules? It wouldn't work, that's what, because aixiz modules are not all perfect. It's a compromise between convenience and performance. I think we all agree that the best solution is to desolder the LED chip and directly solder it to a large chunk of solid crystalline diamond, but that's not convenient at all (not to mention, it might be kinda pricy)

I liked the old meredith modules with screw in backs that press diode into a pocket. If people could make heatsinks like that, it would be a great compromise between performance and ease of use. Still have pressure like an interference fit, but can be removed easily without external tooling.

My two cents in the ring: I say just use cheap alu foil, unless you can get plain copper foil easily. Remove any adhesive, if present. Don't get too carried away with the type of foil, because most of the heat isn't jumping from foil layer to foil layer, it mainly moves between the layers through contact points, so it's still hard work for that heat to get out, just it's much easier than jumping a giant air gap.

pastes work the same way, but with small particles suspended in a viscous paste. Without pressure, those particles have no reason to touch each other, so it would not be suitable for this application. However, if you can do it cleanly, it would be alright to use in place of indium foil when affixing diode into press-fit socket.

Edit: maybe amkdeath was talking about using a combination of foil and paste? That sounds interesting. Messy, but interesting.

 

[Minamoto Kobayashi]

Foil and paste has again not sense, without pressure.
I have ordered a 0.05 mm indium quality foil from China.
Probably I need only one or two coils. Less coils are present, better will be the heat transfer. Indium is often used also between heatsinks and expensive c-mount diodes, so I think this could be the better solution.

 

[Sigurthr]

The idea behind indium is that it should squish to fill voids like a paste should. It isn't quite as good at filling voids as a paste is, but it has the added benefit that it is conductive and will not run when hot. Even "conductive" pastes are very poor electrical conductors, and no paste is engineered to fill a gap of more than a dozen microns or so.

 

[Minamoto Kobayashi]

I have just ended to do some tests with gold foil: it was a total disaster.
The layer is too thin and fragile. If You touch one surface with fingers, part of the sheet breaks and rests stuck on the tip.
And just a little breath of wind that once it all away ... it is not absolutely manageable!
Next week I will receive both indium foil 0.05 mm and pure copper foil 0.1 mm.
But keep in mind that indium is very soft and malleable. This could be a big problem when You go to tighten the screw: probably on the other side the indium will be compressed and will squeeze out thru the hole.
And on the screw side it will re-create an empty space.
In my modest opinion, if I will able to obtain a nice pressure against the walls, it will be better do not use any screw.
With copper foil this problem surely do not exist. I will do all the tests, stay tuned!