Place 2 heatsinks on top of each other?

[rzandrew]

So i have a cheap 775 heatsink/fan(h/f) and also have a 478 h/f, now i took the fan and plastic clip off the 775 h/f so just left with heatsink and bottom bracket, and have placed the 478 h/f on top of that, will this make a difference stacking 2 heatsinks together with the 478 fan on top?

Might be wondering but i don't actually use a case i have the mobo sit fat on a workbench, so this is why i could do this, i could tape or somehow screw the 2 heatsinks together so they don't come apart..

Just a random idea, as my 775 h/f is quite a cheap one thought this might give better cooling?

 

[rubix_1011]

I don't really see any benefit to doing this. You are going to get similar results of passive cooling (with a junky HSF) on the bottom, and far lesser action with the 'active' HSF on the top with the fan. Not to mention that your heat transfer surface area from the passive heatsink will be horrible due to it only making contact with the 'upper' heatsink by the fins alone. You would be better off using the correct HSF (with fan) on the chip and board it was designed for. If you really want to get away from that, look into aftermarket air coolers.

I just don't see any benefit to the setup you described...I like the modder attitude, but I think it will be more of a detriment vs an improvement. One way to find out...try it out...I don't think your CPU will melt...it might go into thermal slowdown, though.

 

[Cranky1970]

Just get a better HSF would be my advice.

 

[evongugg]

You might like something like this:

Coolermaster Gemini II

http://www.xbitlabs.com/articles/coolers/display/coolermaster-geminii_2.html

 

[rzandrew]

I don't really see any benefit to doing this. You are going to get similar results of passive cooling (with a junky HSF) on the bottom, and far lesser action with the 'active' HSF on the top with the fan. Not to mention that your heat transfer surface area from the passive heatsink will be horrible due to it only making contact with the 'upper' heatsink by the fins alone. You would be better off using the correct HSF (with fan) on the chip and board it was designed for. If you really want to get away from that, look into aftermarket air coolers.

I just don't see any benefit to the setup you described...I like the modder attitude, but I think it will be more of a detriment vs an improvement. One way to find out...try it out...I don't think your CPU will melt...it might go into thermal slowdown, though.

You are probably right, but the 2 heatsinks don't make contact by fins alone, the flat top center of the bottom 775 hs mates nicely with the bottom flat center of the upper 478 hs, which has a copper bottom so good conductivity there....

But i am thinking an equilibrium will be met anyway, where both sinks will heat up and the fan will really be cooling the top heatsink and the top heatsink will be trying to 'draw' the heat out of the bottom one, but then again i dunno i am describing them as if they are 2 separate hs's whereas as i said they mate well so really act like one huge heatsink....

I guess i can just experiment and see what happens...

My reasoning was that the larger a heatsink the better a cooler it would be, but then once it heated up the fan would have more work to do to cool it down!!

 

[rubix_1011]

I think you are going with the 'bigger chunk of metal is a better heat sink' concept...not always the case. I think a higher quality heatsink would serve you far better...but like you said, half the fun is figuring it all out. Let us know if you find something different than what might be expected.

 

[cyberjock]

Being knowledgable in heat transfer, this thread intrigues me. I thought about this alot, and there really is no definitively prove what will happen. You say that they mate well, but I have to wonder what the heat transfer coefficients will be.

Overall though, I have to agree with everyone else. I can't imagine this being a net positive for cooling. If size was all that mattered then companies would be making them as big as possible(and I wouldn't have a g/f...). Also, the mounting bracket has weight limits, I think something around 500 grams. I'd bet that you are exceeding this limit. Some of the larger heatsinks exceed this limit, and they tell you not to use them in a computer that stands vertically.

My recommendation is the same as everyone else. Go buy a better quality heatsink if you are interested in better cooling. I've used stock heatsinks for 90% of my computers, and only in very rare situations have I been in a position where I "had" to upgrade. This is something that is primarily for overclockers or computers that are compact, yet make alot of heat.

IMO, buying something better than the stock is a decision that involves looking at your money as well as heat production. Boxed Intel CPUs usually are about $20 more expensive than buying the CPU and an aftermarket heatsink cooler. But the boxed versions come with a longer warranty. You can look at it from either spending $20 on a warranty and stock cooler and spending it on a nice solid copper(or whatever floats your boat) heatsink, or having a "lower" quality heatsink and longer warranty.

I say "lower" because I see it as buying gas. Just because the super unleaded is higher quality doesn't make it a better choice because of the cost. I'm sure there will be several people that will disagree with me, and that is fine. Some people swear by super unleaded and some wont.

To use the car analogy, do you buy super unleaded gas even though the cheaper unleaded works just as well?

 

[xxjudgmentxx]

no...just no.

 

[rubix_1011]

no...just no.

Was this intended for the OP or the previous post? Just clarifying...

I think this is an intriging concept just because no one has thought that trying this was a good idea, but we are interested in the actual outcome. Its kind of like telling your buddy to drive 60mph and throw it into reverse...just to see what happens. Its always a decent idea as long as its not your own hardware to replace.

 

[ram1009]

if you have this much time to waste I suggest studying heat transfer first.

 

[rzandrew]

if you have this much time to waste I suggest studying heat transfer first.

Well i do actually have a degree in Physics if that's any use, although it was a few years ago now and you start to forget a lot of things..

 

[noobe1981]

I disagree rubix. Doing that to a car.. You have a damn good chance at messing crap up. This guy idea, while right/wrong as I'm no expert.. Probably won't cause the cpu to fry on start up. He should have plenty of time to tell wither it works or not.

Now I do agree with the gas anology *w/e it is*. As some people's cars actually need super unleaded, and some even higher depending on what they actually do with the car.

Just try it out.. The way I see it.. It works, or it doesn't you screw up two cheap heatsinks, and have to buy a new one.

 

[rubix_1011]

Yeah, I know, but I did want to emphasize that people aren't always too keen on doing something questionable themselves, but always seem eager for someone else to do it.

I agree: The possibility for large-scale destruction and death vs. a CPU throttling down. Not exactly a good reference example.