[Plat]

here's a variation on the debate.

i have the zalman 3000. now, the 3100 and 5000 are better but this is the one i have and i'm happy with it.

but here's my question/idea.

whether or not, aluminum does dissapate heat better than copper, let's say that aluminum does dissapate heat better than copper, is the difference enough to over come the advantage the zalman has of having continuous copper fins from base to the end of fins?

no matter what you use to fuse the pins/fins to the base, your going to lose efficiency, yes? vs the fins that go from base all the way to the end?

seems to me that the zalman 5000 really is the best of all worlds as far as hs. i've never owned one, but it looks to me like an efficient idea.

the Zalman's work really well with a larger fun(aka Delta). I wish the whole thing was a solid piece as each fin is pressed together, which does hurt some. But then again, milling that would cost a fortune.

[cjtune]

It's either you make the HS totally out of the higher thermally conductive metal or just use that metal for the base to spread the heat (as uniformly as possible to all fins). The reason for using lower conductivity metals for the fins is basically to save on material and/or manufacturing costs.....no use just plating the fins with the higher conductivity metal unless it is for corrosion protection or for looks....even IF aluminium 'throws heat away better' than copper. Go read that (super lengthy) thread "Copper and Aluminium".

thinking of the zalman, has anyone horizontally mounted a fan on that bad boy? If you come from the top you obstruct some surface area, but coming from the side gives you total access to all the fins. The problem is you are only exposed to a tiny portion of the fins. Hmm, anyone want to try this?

[most deaf]

copper abosrbs more heat then aluminium but aluminium gets rid of it quicker, the best design is having the fins made out of aluminium and the base out of copper

[Plat]

most deaf.

ok, so aluminum releases heat better. there is a huge debate as to whethor or not this is true, but you totatly ignored the real question here.

even if aluminum does release heat better, there will always be a loss of efficiency when trying to join the aluminum pins / fins to the copper base. it seems to me that this loss in efficiency will easily outweigh any advantage (if any) aluminum has over copper.

read my entire post.

[ManOfKnight]

i don't know if i believe all the aluminum, copper hype....every GREAT waterblock i have ever seen is copper....almost every GOOD HS is copper ( there are some exceptions )....silver or CVD would be the only things that i think would REALLY change a temp, if you had silver with copper fins, then of course it would perform better than a PURE copper hs, because it has silver...same goes for the copper/aluminum hs's.....i will get a pure copper 462 one of these days...if they get one

[cjtune]

most deaf (Jul 24, 2001 05:30 a.m.):
copper abosrbs more heat then aluminium but aluminium gets rid of it quicker, the best design is having the fins made out of aluminium and the base out of copper

Nope, aluminium absorbs more heat than copper. It's 903J/(kg.K) vs. 385J/(kg.K) of copper @ 300K (26.85'C).

[cjtune]

Plat (Jul 24, 2001 05:37 a.m.):
most deaf.

ok, so aluminum releases heat better. there is a huge debate as to whethor or not this is true, but you totatly ignored the real question here.

even if aluminum does release heat better, there will always be a loss of efficiency when trying to join the aluminum pins / fins to the copper base. it seems to me that this loss in efficiency will easily outweigh any advantage (if any) aluminum has over copper.

read my entire post.

I thought I have beaten this issue to death already in the thread "Copper and Aluminium". But anyhow, here it is again:

1. The best possible conduction within a heatsink can only arise with it being made totally out of a highest possible thermally conductive material.

2. Copper/aluminium hybrids are for lower costs. Well-designed heatsinks are more important than whether they were made out of aluminium, silver, copper, or flawless diamond -but most high-end heatsinks are made out of copper so appreciating this fact may be difficult.

3. Plating a HSF of one metal with another metal of different conductivity is like a kettle of hot water with metal inner skin and plastic outer skin or metal outer skin with plastic inner skin -no one the better, so why not a kettle just made out of metal? ...or plastic if the cost savings justify the slower cooling?

4. Convective (ie. by fluid movement and conduction to fluid) heat transfer does not depend on the material type of the solid surface in contact -just that of the fluid and the surface temperature(s). Conduction through the heatsink does. Convection + Conduction = heatsink heat transfer. So you cannot say that aluminium throws heat away better *just* because it is aluminium. Gee, and how do I know all of this?

5. ...and don't even bring up radiation. The amount of heat transfer through radiation is negligable compared to forced convection (with fans) and there's an article in the main page that (bare?) copper actually radiates heat better than aluminium. Anodizing a metal surface black increases the emissivity.

So let's end this aluminium "throws heat away better than copper" baloney once and for all, OK?

[CalCoolage]

Plat (Jul 23, 2001 09:15 p.m.):
no matter what you use to fuse the pins/fins to the base, your going to lose efficiency, yes? vs the fins that go from base all the way to the end?

How much conductivity do you lose? With good bonding or just squashing the fins into the attachment slot, it could be very small, even less than the microscopic corrosion or anodizing on the fins, which also has a low conductivity compared to aluminum/copper.

I see very no justification for the claim that alumuminum saves any noticeable money (or weight) over copper when the fins are so thin. It couldn't be more than a few pennies. (pun intended) They should charge the same thing.

[cjtune]

CalCoolage (Jul 24, 2001 11:13 a.m.):

How much conductivity do you lose? With good bonding or just squashing the fins into the attachment slot, it could be very small, even less than the microscopic corrosion or anodizing on the fins, which also has a low conductivity compared to aluminum/copper.

I see very no justification for the claim that alumuminum saves any noticeable money (or weight) over copper when the fins are so thin. It couldn't be more than a few pennies. (pun intended) They should charge the same thing.

All the more reason a 'good' HSF should be all copper. But generally, the more machining and assembly steps required, the more expensive a product is. Maybe a couple of years ago, molded or protruded one-piece aluminium HSFs were the most cost effective. Then someone got the idea of combining that with a copper base. But nowadays, the cutting, folding, and soldering of the many (hundreds?) copper fins to a copper base is more familiar to HSF manufacturers that it is the preferred performance-to-cost alternative. That is to say, the fabrication process has matured, and afterall, they are trying to manufacture HIGH-PERFORMANCE heatsinks, and the tolerances for molding or protruding thin fins is just not acceptable anymore.

BTW, the contact resistance between two different metals isn't as much a concern as the LOGIC of using dissimilar metals in a heatsink.

[cjtune]

William (Jul 24, 2001 05:16 a.m.):
thinking of the zalman, has anyone horizontally mounted a fan on that bad boy? If you come from the top you obstruct some surface area, but coming from the side gives you total access to all the fins. The problem is you are only exposed to a tiny portion of the fins. Hmm, anyone want to try this?

I think a model of Alpha heatsink has fans in crossflow config. Don't remember what model or how well it performs. In heat transfer, counterflow heat exchangers are the best (ie. requires less surface area than crossflow or parallel flow heat exchangers). But by mounting axial fans right above a HSF for counterflow, the spot directly under the hub of the fan and above the CPU core is virtually untouched by airflow and crossflow cooling may actually do better. To summarise: Someone try and mount a 80mm at the sides of the Zalman.

[Plat]

cjtune

it's the pep that has the fans blowing accross it.

w/ my zalman 3000 i rigged two 80 mm 2900 rpm fans at an angle (i'd say about 105 degrees from each other and placed it above the zalman so that the fans are blowing towards the CORE.

anyway, i get decent performance. not the same level as the swifty, about even w/ the gladiator or the SK6.

[cjtune]

Plat (Jul 25, 2001 06:44 a.m.):
cjtune

it's the pep that has the fans blowing accross it.

w/ my zalman 3000 i rigged two 80 mm 2900 rpm fans at an angle (i'd say about 105 degrees from each other and placed it above the zalman so that the fans are blowing towards the CORE.

anyway, i get decent performance. not the same level as the swifty, about even w/ the gladiator or the SK6.

Well, experimentation makes perfect.

[Mictlan]

I just posted some other information in the Swiftkiller HSF, please read it and make some comments about it