A friend is making a spacer to go between a tab on a device and a heat sink. The total thickness will be about 0.25 inch (6 mm?). Of course thermal grease will be used at both matings.
I've been asked what kind of aluminum.
What (in descending order of preference) are the aluminum types that would be good raw material for this?
Thanks.
Aluminum extrusions - may be not much less than that of "poured aluminum", which has heat conductivity of 161 w/m-k.
Die-cast zinc-aluminum alloy - 127 for "ZA-27", 115 for "ZA-8". Die-cast zinc "zamak" - 113 w/m-k Die-cast die-castable aluminum alloy - 96.2 w/m-k Annealed brass - 61 w/m-k
(There are heat-treated versions of 6061 including T4 and T6, which have mildly increased electrical resistivity, and I suspect also slightly decreased thermal conductivity.)
Otherwise, extruded aluminum, poured aluminum and 6061 are close enough to each other for "best aluminum", and I expect ordinary aluminum plate and bar stock (to be machined into shape) to be similar.
- Don Klipstein ( snipped-for-privacy@misty.com)
All of the above.
You'll lose more delta-T at the mating surfaces than in the aluminum, most likely. Matweb
All aluminum alloys have approximately the same conductivity (thermal and electrical). Pure aluminum (1199 is the purest) is maybe 20% higher conductivity than 356, 6061, 7075, etc.
Tim
Addendum: all metallic alloys have higher resistivity (thermal and electrical) than either pure metal. For instance, sterling silver is more resistive than either silver or copper. I suppose aluminum alloys are fairly surprising in that regard, as they are still fairly conductive.
Tim
Actually, I found it to vary quite a bit. Pure being best.
Copper is the one that I found remains the same, regardless of how wrought, and well into some alloys.
Regular 6061 T6 is 21% lower than pure Al in thermal speak.
Better look again. Even Al suffers a hit, once the lattice, she is split.
Do you have a choice ? If so, the alloy with the best thermal conductivity.
Graham
The mating surface should be as flat as possible. Many cheap extrusions fail this test. Using formed sheet material is very effective.
See Motorola's ( now On Semi's ) AN1040 IIRC for good thermal advice.
Graham
copper is better. aluminium is cheaper. pure is better than alloy.
the less gap you get at each join the better, so a flat mirror finish on all surfaces is is ideal.
copper is better. aluminium is cheaper. pure is better than alloy.
the less gap you get at each join the better, so a flat mirror finish on all surfaces is good.
Metal is much much better than heatsink paste so if you can solder or weld the spacer to the heatsink (with the apropriate solder and tools) so much the better.
I have seen devices with poor flatness. I have polished some things. Getting flat is hard.
greg
'Ya know, if you can clamp the assembly quite tightly (or even apply some persuation as a part of assembly), you could use pure aluminum to fill the space, no grease. It's soft stuff, about like lead. Oh, and use springy washers too, I suppose...
Tim
Isn't that exactly what he's asking?
Thanks, Rich
When I was in college I was able to achieve significant flatness fairly consistently, given enough beer and other things.
I don't like being that flat any more, though.
"> Addendum: all metallic alloys have higher resistivity (thermal and
Well except for super conductors where some alloys are better.
George H.
So long as the beer isn't flat that's fine :-)
We flatten worn oilstones, used for sharpening wood chisels and so on, by rubbing them with dry sand on a melamine board. Throw some sand on the board and rub the oilstone on top until all the hollows are gone - noisy and takes a little time but it does the trick.
Something similar ought to work for heatsinks
The best aluminum is copper!
After that, check the thermal conductivity of various alloys. They can vary a lot.
And make the spacer big to spread the heat laterally.
John
If you have access to a milling machine, it's easy to mill a surface flat. Aluminum machines really nicely.
Now they just show you how to use an inexpensive surface grinder without embedding the shards of a 3600 RPM stone in your forehead.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.