[rkremser]

Is there anyway to plate copper onto aluminum and if so how would one go about accomplishing this? I have aluminum heat sink extrusion but would like to solder the components onto the heat sink. Aluminum obviously can not be soldered to so i want to plate a thin layer of copper to it and solder to that. Is this possible or am i goign to end up with a multi-piece heat sink.

[martinw]

Dear rkremser,

The experts say it can be done!

http://www.finishing.com/472/50.shtml

Best wishes,

Martin

[martinw]

BTW,

You could well have a bit of difficulty soldering onto a heatsink....

Best wishes,

Martin

[Dr Pete]

Make the heatsink out of copper?
The link mentions using chemicals you would not normally want to use (cyanide)

We could vacuum evaporate copper onto an aluminum plate.
Safer IMHO but you would need a vacuum evaporator.
If you have the tools laying around would be easy to build a simple evaporator.

Still easier & cheaper to make the heatsink out of copper.

[Mariss Freimanis]

Dr Pete,

My concern would be the different coefficients of thermal expansion. Attempting to solder the copper plating would result in strong shear forces at the Cu/Al interface sufficient to delaminate the plating locally. This in turn would greatly increase the thermal resistance (deg C/W) and render its heatsinking properties ineffective.

Mariss

[Dr Pete]

Good point Mariss, besides evaporating cu onto al would result in a very thin ~5k angstroms thick.
Very easily melted while soldering besides delaminating issues.
I go back to making the heat sink out of cu in the first place.
Or they sell metal paste that might work?
Not sure if you can solder to this thou.
You might want to put the electrical contacts into the paste?
Ideas just bouncing around now
Good luck with what you try.

If you are in S. California and would like to try the evaporation send us the part and $180.00 I will put some copper on it for you with no guarantee that it won't fall off

[Geof]

BTW,

You could well have a bit of difficulty soldering onto a heatsink....

Best wishes,

Martin

Real BIG soldering iron.

Probably if it was possible to get it hot enough to solder the copper would start peeling.

Drill, tap and use stainless screws with the paste that is used for copper to aluminum connections.

[martinw]

Real BIG soldering iron.

.

Dear Geof et al.,

I would not be too worried about the state of the copper plating....what about the state of the component after it has been roasted by the Real BIG soldering iron for some considerable period?

Oh, I know...put a Real BIG heatsink on the component leads....

Best wishes,

Martin

[Mariss Freimanis]

Keep in mind solder dissolves copper. A 5k angstrom Cu thickness may very well completely dissolve into the melted solder like sugar in water. My apologies to Dr. Pete; I misread him as the originator of this thread.

Mariss

[Mariss Freimanis]

Heat will not damage the components if they are surface mount rated. Transistors are specified at 260C soldering temperatures; Pb63Sn37 (standard leaded solder) melts at 178C, Sn96Cu3.5Ag0.5 (lead-free solder) has a eutectic temperature of 228C. The common mistake is to use an undersized (too low Wattage) iron cranked up to a very high temperature (>300C) instead of using an adequate sized iron set to 210C and 260C respectively. A hot, undersized iron cannot heat things up quicker.

Keep in mind this is a heatsink so it's mission in life is to prevent you from doing what you have in mind; soldering to it. Find its rated thermal resistance, divide it into your delta-T. That's how many Watts of heat you will need. Example:

Thermal resistance = 0.5C / Watt
Delta-T = 235C this is 260C need to solder minus a 25C ambient.
Power = 470 Watts! (235C / 0.5C per Watt)

That is one hefty soldering iron.

Next, find the thermal time constant for the heatsink. Mutiply it by 3 to find how long it will take to reach 95% of the applied temperature. Example:

Say the thermal time constant is a very realistic 10 minutes. It will take 30 minutes to reach the desired temperature using your 470 Watt iron. That also makes you a very patient man hefting this substantial iron. Can't wait that long? A 1,500 Watt iron gets you there in 10 minutes, a 15kW iron gets it done in 1 minute.:-)

Mariss

[Geof]

.....That is one hefty soldering iron....... 15kW iron gets it done in 1 minute.:-)

Mariss

When I say BIG I mean BIG.

Of course you could get a multi hundred amp welder and just clamp the leads on each end of the heatsink.

[martinw]

Put the heatsink on your kitchen hob, and get it up to say 260 Centigrade .

Quickly place your flux paste, component, and solder. Clamp the assemply and chuck it into a very large container of liquid nitrogen. Then pray, (while you are being engulfed by clouds of nitrogen gas that may asphyxiate you) that the thermal conditions experienced by the component do not fall beyond the top and bottom limits.

It might work...

Best wishes.

Martin

[Dr Pete]

Mariss, no apology needed I love to bounce ideas off other people and learn.
Some real interesting projects and people here that WORK!

I still wonder why doesn't he make the heat sink out of copper in the first place?
I am missing something?
I could machine up something real quick (a day or so lol )

P.S. I love the idea of LN2 ha ha reminds me of the old days when we played harder

[Mariss Freimanis]

My guess would be copper is so expensive now. Unless there are overriding considerations to the contrary, I would urge using more traditional methods such as clamping or screwing the components to the heatsink.

LN2 was a lot of fun. It was interesting how brittle things became when immersed in LN2. A spider or grasshopper would become as brittle as glass when dropped on the floor.:-)

Mariss

[rkremser]

Thanks for all the input on this one, where do i start.
The component is a surface mount device actually a powerso-10 package (similar to http://www.icver.com/images/PIC/V/VB/VB025MSP.JPG) heat pad is on the bottom and temperature range permits soldering in reflow oven. I could actually do the copper sputtering at work as that is kind of what we do, i just first was only thinking of doing this at home. Also didn't think of the solubility of the copper into the solder. Didn't realize that was a possible issue. The counter point of just making this out of copper is sounding better and better, originally the cost was going to be preventative in my mind but the more i think about it, it may just be worth the headaches. I'd make a 2 piece heatsink, copper base maybe 1/4" thick with an aluminum finned traditional sink bolted attached with some tapped holes in the copper.

[Geof]

..... I'd make a 2 piece heatsink, copper base maybe 1/4" thick with an aluminum finned traditional sink bolted attached with some tapped holes in the copper.

Tap the aluminum, probably much easier to tap than the copper.

[rkremser]

the Tapped copper was mainly to help assembly, the machine screws would have to come through the pcb board and as i type that i'm thinking that would probably be an advantage to stabilize the entire assembly...... so ya tapped aluminum.

[Mariss Freimanis]

Another option is to use an aluminum core multilayer printed circuit board. Metal core boards are intended for use with high power dissipation surface mount components.

Mariss

[neilw20]

If the design was more efficient you might not need a heatsink.

Something obvious. I saw some poor soul trying to solder a hole in his car radiator.
Someone said he must get more heat into it to solder properly. (they were correct.
1. Poor soul had not drained the water.
2. He started the car to make it warmer after above advice.

He was never nice to me, so I just enjoyed observing the chaos:withstupi

[rkremser]

Neilw20, think you posted in another post of mine that I was discussing the application for this. Its a 300-350 amp 12v dc motor control, i've found mosfets that should keep the dissipated heat down but i'm still thinking its going to be more significant than being able to get away with no heatsink at all.