Jump to content
Corsair Community

Push/Pull + Top Exhaust Configuration


TheTiesThatBind

Recommended Posts

So I was watching Bitwit's video once again showing that having a front mounted radiator actually cools the CPU better than a top exhaust configuration without affecting the GPU temp which is great but I came up with the following question.

 

If we move the radiator on top and use a push/pull configuration, wouldn't that mean that the CPU temp won't introduce huge thermal differences (like in Kyle's video) than a front mounted radiator? Also wouldn't we be able to see better internal temperatures as well (GPU mostly).

 

Kind regards

Link to comment
Share on other sites

There are a lot of videos out there. Some are fairly representational. Others have about as much scientific merit as the Pepsi Challenge. I have not seen that specific video, so this is mostly a general statement.

 

Part of a radiator's effectiveness is related to the temperature of the incoming air. Without doing calculations, cooler air in will result in cooler air out at the same level of wattage in the cooling system. Typically, we don't really care about the exhaust temp up top, but it also means in some instances with a climate controlled room, you are not going to cause a massive change in case temperature. If your coolant temp is 27C, the exhaust temp will be right around that as well. Most everything else in your case is warmer than 27C, so there is very little noticeable heat add on. Obviously your are adding heat into the case. It just doesn't show and this is where the front mount advocates claim victory.

 

In turn, the air above your motherboard is going to be warmer than external air. Even in a mild system, you are going to see VRM, RAM, and capacitors temps in the mid 30s at their lowest level. I was taking some thermal pictures of my PSU yesterday and was rather surprised to see the back of the motherboard as the warmest thing in that compartment -- more than the PSU shell right after a 600W load for 2 hours. That heat is going up and out. Is it going to be a negative on your cooling? Not much. But even in near perfect open cases with fans on every side, I usually have a 2-3C temperature difference between the top 5 cm of the case and the bottom (or front rail). That is the expected gain. If you have some temp probes or another way to measure air temperature, you can get a decent estimate of the improvement from the existing local temps. It's not a lot and would not make the difference for 99% of the builds out there. Other factors like fit and aesthetics may be more important... at least until you bring a big watt GPU into the equation.

 

It is the GPU that can really shift this from user choice to necessity. Results are always case specific, but it is not uncommon for people with top exhaust to see a +10C coolant delta with a moderate CPU load in a game vs +6C running a 100% stress test. That's +4-6C extra and I have seen a lot worse. When you are talking about adding 10C to your CPU temps, that is most certainly noteworthy. If you feel like your game CPU temps and specifically the coolant temperature are too high, but yet you get really nice CPU bench test numbers, then this is a possibility and a strong candidate for moving things around. Don't worry about the GPU temps with halving your front intake flow. The only thing that will cool your GPU is its own fans. The case fans have the same kind of impact the radiator fans do on coolant temp. However in this instance, it is the case ambient temp. Each +1C of case ambient is +1C to everything. So, there is a potential add on heat effect, but you still have free air top and rear fans to suck it out. It is usually not a problem. Multi-GPU set-ups are another matter. In those set-ups, you typically need a strong push from the front to get the air in between the cards out of there. Failure to do so usually has a strong, negative impact on the top card.

 

Ultimately, the final say in what's best is almost always determined by case design and layout. There are some cases that for whatever reason (solid front door, weird top baffle, etc) just kill radiator restricted performance. Some have permanent dust filters that add too much resistance. Many possibilities, so the bottom line is test for yourself and there is no golden rule about it. Right now I am running a full water 740 with a 8700K and 2080 Ti. I have dual 280mm radiators and the beauty of this case allows me multiple configurations. I spent half the Spring flipping fans and radiators trying to whittle down my coolant temp to match my old system. This one can put as much as 550W in the cooling system. I have done top/bottom intake exhaust, reverse flow, you name it. I finally settled on top/bottom radiators set to intake with dual side exhausts from 3x140. It is quite effective and it reduced my coolant temp another 2C down to the arbitrary number I had targeted. I cannot improve upon it any further. So, what do I think about most days now? "I wonder if the case would look better if I flipped the fans around to show the other side?". If you don't have to buy $120 worth of fans to try other configurations, do it. Don't like it? Switch it back. Decide what's important to you and remember 2C is not a life changing alteration.

Edited by c-attack
Link to comment
Share on other sites

I couldn't disagree less ;):

 

The only thing that I would add, is that hot air rises. So for me, exhausting at the top makes sense, which is why my radiator is at the front.

 

Internal case cooling can be optimised with well placed fans and temperature based profiles.

Link to comment
Share on other sites

Having the Coolermaster SL600M case which has the chimney effect cooling and my graphics card radiator having to be fitted to the bottom, I had no choice but to fit my H150i radiator to the top of the case as a exhaust.

I went with the push pull method, six fans running nearly in sync, and I have managed to keep my temps (under load) for both graphics card and CPU to around 45c

 

I'm happy with that.

i9 9900k - RTX 2080ti.

Link to comment
Share on other sites

I couldn't disagree less ;):

The only thing that I would add, is that hot air rises. So for me, exhausting at the top makes sense, which is why my radiator is at the front.

 

Actually it does not. Colder, heavier air sinks and displaces the warmer air, but this is only important when having a scientific discussion. Differences in local case temperature and the effects were discussed above.

 

 

Internal case cooling can be optimised with well placed fans and temperature based profiles.

 

Yes, I think that is what the entire discussion is about! Incidentally, even moderate fan speed will drastically overpower any natural convection, so that brings it back around to case layout and preference.

Link to comment
Share on other sites

I couldn't disagree less ;):

 

The only thing that I would add, is that hot air rises. So for me, exhausting at the top makes sense, which is why my radiator is at the front.

 

Internal case cooling can be optimised with well placed fans and temperature based profiles.

 

When not acted on my other forces and all other things being equal, this is effectively true. But like all super-simple statements, there's more to it than this. Very humid air, even if warmer, will sink under colder, drier air.

 

And, of course, in a system case, you have fans that completely and totally upend this. So from a case heat management perspective, it's pretty useless unless you are looking at an early to mid 90's era PC with no case fans.

Link to comment
Share on other sites

×
×
  • Create New...