How are passive cooling designs different from Air Cooling ones?

[Hunain]

Hey there, I recently read that the Ninja Copper keeps an Overclocked quad @ 3.2 pretty cool, passively at 64C on full load.

I thought id give my ASUS Silent Knight a try, plugged out the CPU fan, ran the PC, its all copper and lots of fins etc, still the CPU was pretty hot, wouldnt even run any benchmarks, they abort saying that the CPU is too hot, each core easily exceeded 65C threshold in 4 seconds, so I was wondering are there any fundamental differences in passive heatsinks? Or are they simply bigger? I expected my all copper silent knight to atleast keep the temperature to something useable, but it was a total failure

And how do Passive PSUs work? Same for Passive gfx cards, sorry if Im asking a stupid question,the truth of the matter is that I dont know when my younger bro asks me about cooling without a fan

[fredered]

One difference I bleive is that the fins are furthur apart so that air can move between them more easily only by convection, and then just being big I think is the other one. Case airflow will be important with a passive cpu cooler as well.

[hermeticist]

You are now venturing into the world of thermodynamics........Making something passively cooled should really start with the item being cooled, the more efficient the device the less energy is converted into heat.

Generally a passive cooler will have the largest surface area possible & will (by necessity) be larger than a fan fed cooler. The thermalright 120/IFX-14 are prime examples - they make good coolers because they`re designed to be passive, so when fitted with a fan they should outperform a cooler designed originally to require a fan (though optimisation of design through airflow - think windtunnel - is also important).

There are also considerations in the materials used - their heat (as opposed to electrical) conductivity. For instance, diamond is electrically non-conductive (I believe), but the best conductor of heat. Here we come to the crux of the matter, cost considerations. Larger surface areas require more material & are often more labour intensive, more conductive materials are generally expensive, silver, copper etc.

Cost considerations are the main reason we uses fans. Companies want to sell you a product in which they can maximise their profit (and be competitive). If it costs you a fortune in electricity over the life of the product that is not their problem & to be fair you have to way up the efficiencies across a whole. For instance: If a fan increases cooling by 50% and say costs £1 a year to run........

Other technologies utilised in passive cooling are heatpipes etc. where fluids in the tubes run a cycle of evaporation - condensation to dissipate or transfer heat (also look at the Akasa Revo, this does still use a fan and doesn`t work perhaps as well as people hoped, but the technology is interesting).

Finally, I wouldn`t recommend using a cooler designed for a fan, without the fan working. An overclocked Q6600 might use say 150W, the cooler is designed to handle say 165W, remove the fan and passively cooled that will drop to below the CPUs output and (hopefully) the processor will shut down to prevent heat damage.

I find it quite hard to believe in a typical household environment that the Ninja copper could cool a Q6600 @ 3.2GHZ under load @ 64 degrees centigrade.

Hope that helps - sorry for the long, rambling post......

[staffsMike]

That is pretty much it. When the fins are packed together you need a fan on the HSF to push air through them. They will more often than not give better results than passive coolers with or without fans, but the non passive coolers will probably struggle with low speed fans or no fan.

Thermalright ultima and ultras do work ok with low speed fans but are only at their best with a high speed fan, where the ninja doesn't really improve that much whatever the fan speed.

[Hunain]

Hmm, thanks Mike and thanks Hermeticist, very informative, Thanked you

[hermeticist]

Not a problem Hunain - sadly, I find it quite interesting!

[Hunain]

Not a problem Hunain - sadly, I find it quite interesting!

I find it interesting and wanted to ask since some time, forgot everytime

[Main]

Oh the joys of heat transfer... Yeah everybody already covered the jist of all this so I thought I'd just clarify the reasons for the differences

The difference between passive and active coolers is that passive coolers use natural convection and active coolers used forced convection as the primary means of getting rid of heat. Convection is the transfer of heat between a solid and a fluid as a result of localised movements of fluid as it is heated, expands, and rises, and is subsequently replaced by cooler air. The difference between natural and forced is whether convection occurs naturally by localised density differences, or as a result of the fluid being moved around by some external means.

For natural convection to take place the heatsink fins need to be sufficiently far apart to allow convection currents to form, but because the total amount of airflow is less there has to be a lot more surface area available. As a result passive coolers are much larger than active coolers with the fins further apart. Bigger is usually better, assuming you can get the heat from the source to the sink fast enough (like using heatpipes).

Forced convection is driven by a fan (maybe ionic wind if that ever makes it to market) so you can balance the size of the required heatsink against the power (noise level) of the fan you want to use. Tighter packed fin bundles will need a fan that can produce a higher pressure to get sufficient air through it. So to shift the same heat load an active cooler can be much smaller than the passive equivalent if it's fitted with a high power fan (assuming it's a good design...).

After all that, they're both just stacks of fins, with all the cost and weight balancing of different materials and the like. They can also both suck, no matter how big they are, if the design is bad (like using crap heatpipes).

IMHO 'passive' cooling is a bit of joke in computer cooling since you need to guarantee a supply of fresh air for them to work properly, so you still need fans - unless you've got one of those cases that is a giant heatsink and everything is heatpiped to it.

[Hunain]

Thanks Main, helpful additional info

[UrbanKiwi]

Based on the above, is it worth putting a fan on a large passive cooler like the Ninja rev B at all then ????

[Hunain]

Based on the above, is it worth putting a fan on a large passive cooler like the Ninja rev B at all then ????

You could use a bigger fan with a solwer rpm and try it, I think it would have positive effects, but not so drastic, since they pointed it out that the fins are spaced further apart etc. Try it and do let us know.

[Main]

No problem Hunain

Based on the above, is it worth putting a fan on a large passive cooler like the Ninja rev B at all then ????

This review of an Arctic Cooling Accelero S1 is a good example of adding fans to a passive cooler. Putting a fan on a passive cooler is much the same as uprating the fan on an active cooler, since in both cases you're just increasing airflow. As Hunain mentioned, it won't be an optimal design though.

As for the Ninja, I see it can take a 120mm fan? Although it looks like the designers just bolted it on as an afterthought when processors got more powerful... Can't say I've seen anything that says it's good for passive cooling of overclocked quads though....

[Hunain]

...Can't say I've seen anything that says it's good for passive cooling of overclocked quads though....

I think i read StaffsMike (sorry if im mistaken Mike) posting that he saw a video, or read somewhere, about the Ninja Copper, passively keeping an overclocked quad @ 3.2Ghz at 64C under full load..