heatsink style

[Jerboy]

Are there any thermal engineers here who can explain the characteristics of different heatsink style? All the recent heatsinks are fin style and I haven't seen island/forest type heatsink since the Pentium I days and they were mainly used for passive cooling. Is it correct to assume island type is more efficient since there is more surface area?


If you don't know what I'm talking about here's a pic heatsink , the fourth from the top on the left.

 

[CTho9305]

fix the link? that pic isn't very good - do you have any others?

here are a lot of types...

 

[Evadman]

Custom design #6 looks like the heatsink that was used on the 1u server that Anadtech upgraded to, except a little taller.

 

[Evadman]

The greater the cooling surface + the greater amount ( mass 0 of airflow ) = better heatsink.

The fins cut down on air turbulance, thus keeping sections of lower airpressure form disrupting the heat transfer. The same amount of metal will transfer more heat if: #1, the air persure is increased or if there is more surface area.

There is a fine line tho, if the fins are too thin, they will not have enough surface area with the base to transfer heat into the fins. That is why most designs use fins that are thicker on the bottom, and get thinner on the way up.

Does that help a little?

 

[Belegost]

Hrmmm, wouldn't mind testing some of those spiral pattern sinks as the base for a waterblock.

 

[CTho9305]

<< Hrmmm, wouldn't mind testing some of those spiral pattern sinks as the base for a waterblock. >>



thats an interesting idea... have the water enter in the very center and work its way out (or other way round).

I think the main advantage of spiral heatsinks is that the air coming out of fans tends to spiral a little bit, and possibly on the intake side also, so having a spiral heatsink would mean smoother airflow.

 

[Jerboy]

<< The greater the cooling surface + the greater amount ( mass 0 of airflow ) = better heatsink.

The fins cut down on air turbulance, thus keeping sections of lower airpressure form disrupting the heat transfer. The same amount of metal will transfer more heat if: #1, the air persure is increased or if there is more surface area.

There is a fine line tho, if the fins are too thin, they will not have enough surface area with the base to transfer heat into the fins. That is why most designs use fins that are thicker on the bottom, and get thinner on the way up.

Does that help a little? >>



Thanks thats what I was looking for..

The island type in my photo is a passive cooling Pentium 200MMX cooler and the fin type came from some kind of force cooled industrial device.

What are other factors in choosing heatsink in engineering view rather than try this and try that overclocking?

 

[sitka]

<< What are other factors in choosing heatsink in engineering view rather than try this and try that overclocking? >>




LOOKS :Q

 

[CTho9305]

sweet heatsink
as long as the base is really well made, it could even work! (probably not for athlons though )

 

[Evadman]

<<

<< What are other factors in choosing heatsink in engineering view rather than try this and try that overclocking? >>




LOOKS :Q >>



That hurts my eyes just looking at it. I thought it was a bunch of twigs at first Is that your computer? I would think that it would not be too bad for transfering heat, but the way the copper is set, it probably concintrates the heat around the base, where you don't want it. Not enough airflow at the base.

What are your temps with that? ( assuming it is yours )

 

[Demon-Xanth]

The spiral heatsinks have a high pressure drop and work poorly except in the situations in which they were designed for (special order). A water cooler might work though since it has a higher heat carrying capacity.

 

[sitka]

Evadman
When this sink was in the machine there was a side panel mounted 120mm blowing down on it. I originally though the 120 was putting out around 60cfm because it was a cheapy but then looked up the specs it was over 100cfm. could have been getting funky readings from the insocket with all the air zomming around but it was decent enough for back then (~36C while SuperOrb showed 40C) with a duron@1000 1.75V. Still got the finer gauge wire pre cut (It really is a big mitful of copper) for the next version just don't have any more 1/4" copper bar stock. Anyways it was just for fun once water cooling was done but if ever there is a leak it is going back in on a Athlon@1.4 1.55V.
more LOOKS

 

[Jabroni]

That is the coolest thing I have ever seen. Nice copper cooling bush.