Building a CPU Heatsink. [Help]

I am senior mechanical engineering student and have taken a few classes in heat transfer. There is so much math that goes into the design of a simple pin/fin style heatsink it is ridiculous. If done wrong a heatsink will hurt more than it helps regardless of what it is made of. There is a whole section about pin efficecy and everything else. Also if you try to machine a heatsink with common household or shop tools you won't get good results. Heat transfers best from thin materials/high surface area, however using standard machining methods there is only so thin you can machine a fin before the cutting forces destroy the whole fin.

I love the idea but if you are going to make a custom heatsink you are better off going with making your own water cooling block. They can be alot more forgiving.

You know, if this was a waterblock, you'd have a better chance.

Plus, since using water, the block requirements would be alot smaller, if not more than half of what would be needed for an air block.

Drillpress/hand-held electric drill+vice, tap/die set, and mounting brackets.

Good luck if you are gunner attempt this .. I have often wondered how cool it'd be to build your own like perfect heatsink with like the best features/designs of all the top performing ones out there ..

Kasm said:

There is so much math that goes into the design of a simple pin/fin style heatsink it is ridiculous.

Click to expand...

I think that pretty much sums it up. Most successful HS/Water Block designs come from a series if design and testing steps. Without the knowing the math and the having the testing facilities there would be no way to accurately predict the effect of any design changes. So what would you do to improve on the first try if it wasn't as good as you needed it to be?

Without those facilities it's kind of like saying you're going to build a car to win the Indy 500, but only going to test it in your basement.

I'm not trying to discourage you from having some fun and doing some learning, just pointing out some of the pitfalls you'd have to overcome.

Buy a small copper cooking pot or pan, lap it, fill it with (ice)water, use several fans over the surface of the water. Instant win!

Im working on possibly building an oversized heatpipe heatsink, with a large radiator, such as a condenser used in phase change. Why, I dont know...

And echo, use salt with the ice in the pot, and mix the water with a pump ect. then post results

Still debating weather or not to make one, the high cost of the metal is a bit of a turn off but I love DIY work haha.

I have to agree with Kasm, the few thermo classes i had back in my college days (3 months ago) proved to me that more I thought i knew about head transfer the less i actually knew. I work with someone who worked for Thermacore. From what she tells me the process of R and D for heat pipes is insane. She said when the P3 came out it took them upwards of a year to develop a viable cooling solution to be used in laptops.

My advice would be to save the money and use it on something else.

I calculated the copper setup i was on about would be 15.3kg before cutting, imagine that on you mobo!!! Bit of flex!

Yeah this idea will only end horribly. Fins/pins have to be designed with as much surface area as possible but also maximizing airflow, while at the same time not being too heavy, and also they need to be corrosion resistant, oh and being cost-efficient is nice too.

Whoever is thinking about building their own heatsink for anything resembling a practical application is nuts. It will probably cool worse than a stock heatsink, weigh 5 times as much, cost $200, and take forever to build.

If you just want to do it to do it, that's fine, but don't expect good cooling. Be prepared for horrible results.

EchoTheDolphin said:

Buy a small copper cooking pot or pan, lap it, fill it with (ice)water, use several fans over the surface of the water. Instant win!

Click to expand...

LOL! I want to see this happen! You could use a copper skillet and just fill it with ice cubes. That would be hilarious, but might work. Probably would need a thinner skillet though.