Negative or positive pressure? Does it matter?

[AusWolf]

Then that's case dependent. Some cases are built for positive pressure, some for negative and some for balanced. On my 500DX I have four 140mm fans, two intakes & two exhausts. As it's the proper airflow configuration for this case.

I'd love to have 4x 140 mm fans as well, I'm just afraid that they won't fit. A new AIO is a bit expensive just for an experiment, but I guess I don't have much choice (but to experiment).

Well, that's even more fucked up then. Your only option is to have front intake and top exhaust. Two fans on each. Still, I think it's better to put AIO on front, but this needs testing. I really don't want it on top, because you are supposed to use pressure fans on intake and put obstructions on intake. Hot air is less dense, more viscous and is much harder to pressurize it, therefore using static pressure fans on hot air exhaust just looks like a bad idea and GPU almost always expels more heat than CPU, so you take it quite hot air and make it even hotter, that's probably bad for CPU temps.

To be honest, that's why I have my AIO as exhaust, as the GPU is more reliant on cool, fresh air to maintain proper temperatures. It peaks at 73-74 °C with the 2x 120 mm intakes. I'm wondering what the AIO pushing hot air onto it would do. Maybe something drastic. Maybe nothing. I'll have to test this.

 

[The red spirit]

To be honest, that's why I have my AIO as exhaust, as the GPU is more reliant on cool, fresh air to maintain proper temperatures. It peaks at 73-74 °C with the 2x 120 mm intakes. I'm wondering what the AIO pushing hot air onto it would do. Maybe something drastic. Maybe nothing. I'll have to test this.

Well, you only have two choices and only about where to mount AIO on top or as front intake. You can only put fans on front intake and top exhaust. Basically, you should just test that out in quick and dirty way like I did. I think that running prime95 small FFTs with Unigine Valley or Heaven on high preset (don't care about fps, it's only GPU creating heat) for one hour (should be enough for AIO temps to stabilize). Use HWiNFO64 to log data. Test one setup write down temp sensor results and test another. If you post your findings here, it would be nice. For AIO mount fans for push or push pull, pull alone is weak sauce. You should monitor at least CPU, CPU VRM, GPU temps, but the more data points the better. It's easier to visualize what's going on in PC with more data points. All in all, it would take you 3 hours to know things for sure. Those are likely the best cooling setups. If you want, you can also test with same layout, but all fans as exhaust or intakes, but those are silly setups and unlikely to yield good results and there fore are likely waste of time.

 

[Vayra86]

120x38s were used for cabinet cooling long before rads were a thing on the desktop.. The mills and lathes at work all use 120x38s and 140x38s for cabinet cooling as well.

A cabinet isnt a desktop.

For similar reasons we dont place 8000 RPM Deltas in our case... other metrics are of importance, like noise and its frequency/pitch.

 

[ThrashZone]

Hi,
Stereo volume control you'd never notice that b-52 on the runway

 

[freeagent]

A cabinet isnt a desktop.

For similar reasons we dont place 8000 RPM Deltas in our case... other metrics are of importance, like noise and its frequency/pitch.

A cabinet is a desktop, it’s another way of saying chassis. Because back in the day you usually had a thick exhaust and no intake fan. Or fancy setups had a thick intake and exhaust. Long before what we have today.

Hi,
Stereo volume control you'd never notice that b-52 on the runway

That’s why I gave up on desktop speakers in the early 2000s.. My trusty champagne face h/k was what drove my pc sound for years

 

[Vayra86]

A cabinet is a desktop, it’s another way of saying chassis. Because back in the day you usually had a thick exhaust and no intake fan. Or fancy setups had a thick intake and exhaust. Long before what we have today.

That’s why I gave up on desktop speakers in the early 2000s.. My trusty champagne face h/k was what drove my pc sound for years

Right. Lets not go over semantics and lets stop evading the point: noise is a metric too. And pressure fans will need higher RPM to move similar volume, because they are optimized for pressure and not flow.

Its the whole point of optimization; varying user cases (dat pun!)

 

[freeagent]

Thick fans don’t need high revs to push a lot of air, and thick fans don’t have to be loud either, that is an extremely common misperception.

 

[mtcn77]

Thick fans don’t need high revs to push a lot of air, and thick fans don’t have to be loud either, that is an extremely common misperception.

Thick fans have less flow separation.

 

[freeagent]

Uhoh.. I’m not familiar with that term

 

[Vayra86]

Thick fans don’t need high revs to push a lot of air, and thick fans don’t have to be loud either, that is an extremely common misperception.

But, they are thick, and still more noisy while doing so. If they were indeed better on all metrics we would not have left them behind... we also havent got wind tunnel desktops.

 

[freeagent]

But, they are thick, and still more noisy while doing so. If they were indeed better on all metrics we would not have left them behind... we also havent got wind tunnel desktops.

I'm not sure what to say.. I have had old computers with thick fans that were whisper quiet, they weren't 2500RPM shredders though. I have a 24v 120x38 as a rear exhaust in my case right now, and at a constant 12v it is literally whisper quiet. My other system fans are way louder when they spool up, and they are 25mm.. There is a lot of air moving in that scenario and you can hear it, but you mostly hear motor noise because 2500-3000 RPM is loud on 25mm and 38mm fans.

 

[thesmokingman]

I don't care about dust! My question is about airflow and internal case (and GPU) temperatures! Thank you.

Then just keep the ratio balanced or as close to it as you can. In my case I have 6 intake fans and 8 exhaust (rads).

@AusWolf Go for positive pressure. Doyll once explained that if possible, positive pressure is always better as it blows cold air onto components. Exhaust fans suck in air much weaker, than it exhausts. Since computers aren't sealed boxes, exhaust fans can't do much. They are only useful for creating flow from front bottom to top rear.

Here's the original article:

* Basic Guide to Improving Case Airflow*

How airflow works Airflow is simply displacement; for air to come into case, air must be leaving case .. or .. for air to leave the case, air must be coming into case. Think of the air around us as water and we are divers in it and a sunken van is a computer case. We can't move more water into...

hardforum.com

"But with good case intake fans we don't need exhaust fans, same as good cooler / radiator fans don't need pull fans."

But then you have sort of cubish case. I would remove front metal grille and use both stock intakes and maybe add a fan to front bottom as intake. It might not need any exhaust fans then. AIO would go to front intake and you can reuse stock fan and put it as bottom front intake. Taking side panel off more or less means that you have intake bottleneck.

Dislike threads like that. Part in bold I would avoid in a watercooling setup as you are pushing in warm to hot air that is already used up. And with no exhaust fans you will end up cooking every other component in the rig, not smart.

 

[The red spirit]

Then just keep the ratio balanced or as close to it as you can. In my case I have 6 intake fans and 8 exhaust (rads).


Dislike threads like that. Part in bold I would avoid in a watercooling setup as you are pushing in warm to hot air that is already used up. And with no exhaust fans you will end up cooking every other component in the rig, not smart.

What? Granted that case is decently open, if you mount AIO as intake, warmer air will be blown on board and card. Graphics card should heat it up more and then everything will be pushed out same intakes eventually through case holes. I tested that in this thread and well it works, but having exhausts helps, so what was said, is that intakes are more valuable than exhausts, but that doesn't mean that you shouldn't have exhausts. It's all about making air flow in case.

 

[thesmokingman]

What? Granted that case is decently open, if you mount AIO as intake, warmer air will be blown on board and card. Graphics card should heat it up more and then everything will be pushed out same intakes eventually through case holes. I tested that in this thread and well it works, but having exhausts helps, so what was said, is that intakes are more valuable than exhausts, but that doesn't mean that you shouldn't have exhausts. It's all about making air flow in case.

That's why writing strong points is stupid because no choice has a clear advantage.

 

[Vayra86]

I'm not sure what to say.. I have had old computers with thick fans that were whisper quiet, they weren't 2500RPM shredders though. I have a 24v 120x38 as a rear exhaust in my case right now, and at a constant 12v it is literally whisper quiet. My other system fans are way louder when they spool up, and they are 25mm.. There is a lot of air moving in that scenario and you can hear it, but you mostly hear motor noise because 2500-3000 RPM is loud on 25mm and 38mm fans.

Motor noise and 2500 RPM is certainly not going to fit your average (performance) desktop use case, so there it is then.

Noise is a thing of perception. Im used to zero motor or pump noise, all I hear is the actual air being moved, really. Whole rig is low RPM and even the GPU is using a slightly larger diameter fan than most. Motor spool up? What is that

Everything spins at 1000-1200 here and case fans arent even PWM anymore: its actually better to have a constant there, again for a good noise/perf middle ground.

 

[mtcn77]

That's why writing strong points is stupid because no choice has a clear advantage.

True and false. Intake radiator is still not as good as no internal exhaust setup which can be done with unobstructed intake side.

 

[claes]

If you’re seriously considering buying a new AIO then I would consider buying a new case. That seems like a better investment since your problem has nothing to do with CPU temperatures but with GPU temperatures and noise (something most of this this thread is neglecting).

There are some good and bad ideas about airflow here but I do not have the time, maybe @doyll has the energy :wave:

 

[maxfly]

I would suggest starting with a couple of these,

Phanteks Innovative Computer Hardware Design

www.phanteks.com

You cant get anything better for intake fans that need to overcome filter/grill restrictions while maintaining exceptional flow and still being quiet. If you find that they arent the answer to your high gpu temps they will certainly outperform your rad fans(if they will fit).

 

[AusWolf]

Well, you only have two choices and only about where to mount AIO on top or as front intake. You can only put fans on front intake and top exhaust. Basically, you should just test that out in quick and dirty way like I did. I think that running prime95 small FFTs with Unigine Valley or Heaven on high preset (don't care about fps, it's only GPU creating heat) for one hour (should be enough for AIO temps to stabilize). Use HWiNFO64 to log data. Test one setup write down temp sensor results and test another. If you post your findings here, it would be nice. For AIO mount fans for push or push pull, pull alone is weak sauce. You should monitor at least CPU, CPU VRM, GPU temps, but the more data points the better. It's easier to visualize what's going on in PC with more data points. All in all, it would take you 3 hours to know things for sure. Those are likely the best cooling setups. If you want, you can also test with same layout, but all fans as exhaust or intakes, but those are silly setups and unlikely to yield good results and there fore are likely waste of time.

Oh no... with a power limit unlocked 10 or 11th gen Intel CPU, you must avoid Prime95 like the plague! It pumps as much power into the poor thing as it can before it hits 100 °C and throttles. Not good. I know you didn't mention it, but I also advise people against Furmark for similar reasons. My test method is an unlimited run of Superposition 1080p Ultra, or a 3DMark Time Spy stress test for GPU, and a 10 or 30 minute Cinebench R23 loop for CPU.

I'll test some things nonetheless. My next plan is to put the AIO as intake, so that I can remove the front dust filter, and use the radiator to filter the incoming air. I also plan to remove the top filter, as I probably have no use for it anyway, thanks to the top glass, like you said it yourself. I'll report my findings as soon as I have anything. It won't be soon, though, as I'm going on a holiday.

If you’re seriously considering buying a new AIO then I would consider buying a new case. That seems like a better investment since your problem has nothing to do with CPU temperatures but with GPU temperatures and noise (something most of this this thread is neglecting).

It feels so refreshing when someone understands what my question is actually about! Thank you!

What I'm planning for now, is test different cooling setups in the same case. I bought it not so long ago, and I quite like it to be honest. Micro-ATX cases with similar functionality in such a compact size are rare. Some reviews said that this one works a lot better with the dust filters removed, so I have high hopes.

 

[The red spirit]

Oh no... with a power limit unlocked 10 or 11th gen Intel CPU, you must avoid Prime95 like the plague! It pumps as much power into the poor thing as it can before it hits 100 °C and throttles. Not good.

That just means that your power limit is set too high. Something like that shouldn't happen in prime95, as it is close to many BOINC workloads or encoding stuff. It's not unrealistic load and your hardware has to handle that for long periods of time. For thermal testing, you could just disable turbo at all and test at base speed and fixed fan rpms.

I know you didn't mention it, but I also advise people against Furmark for similar reasons. My test method is an unlimited run of Superposition 1080p Ultra, or a 3DMark Time Spy stress test for GPU, and a 10 or 30 minute Cinebench R23 loop for CPU.

I don't find Furmark malicious. I used all my cards with Furmark at some point and nothing spectacular happened. Neither in thermals or power usage Furmark wasn't much different from Unigine benchmarks or games or BOINC workloads.

As for Cinebench, my watt meter and CPU temperature sensor says that it's just as bad as prime95, maybe a little bit worse. A light CPU stress test is CPU-Z test, but only stable version, which doesn't utilize AVX. But it's a poor test for thermals and it doesn't check stability well, so I would avoid it.

 

[AusWolf]

That just means that your power limit is set too high. Something like that shouldn't happen in prime95, as it is close to many BOINC workloads or encoding stuff. It's not unrealistic load and your hardware has to handle that for long periods of time. For thermal testing, you could just disable turbo at all and test at base speed and fixed fan rpms.

My power limit is fine for everything except for Prime95.

I don't find Furmark malicious. I used all my cards with Furmark at some point and nothing spectacular happened. Neither in thermals or power usage Furmark wasn't much different from Unigine benchmarks or games or BOINC workloads.

The last time I tried Furmark was when I still had my 5700 XT, and it was hell. Prime95 GPU edition. Never again.

As for Cinebench, my watt meter and CPU temperature sensor says that it's just as bad as prime95, maybe a little bit worse. A light CPU stress test is CPU-Z test, but only stable version, which doesn't utilize AVX. But it's a poor test for thermals and it doesn't check stability well, so I would avoid it.

That's weird. What's the power limit on your CPU? Maybe it maxes out during Cinebench as well. As for CPU-Z, I agree. It has a nice little benchmark tool, but not good for stability/thermal testing.

 

[lexluthermiester]

It peaks at 73-74 °C with the 2x 120 mm intakes.

Those are not bad temps. Are you worried you're going to burn it out?

Oh no... with a power limit unlocked 10 or 11th gen Intel CPU, you must avoid Prime95 like the plague! It pumps as much power into the poor thing as it can before it hits 100 °C and throttles. Not good.

Prime95 is not the problem. It is designed deliberately to push a CPU as hard as it can so you can gauge compute performance and cooling performance. If it's spiking to 100C and throttling, you need better cooling, not to avoid Prime95.

 

[The red spirit]

My power limit is fine for everything except for Prime95.

Then it's not fine. It's set too high, lower it and it will be fine in prime and you shouldn't see any performance effect everywhere else.

The last time I tried Furmark was when I still had my 5700 XT, and it was hell. Prime95 GPU edition. Never again.

Not sure what hell is, but it's not much different from using your GPU for Collatz@Home or MilkyWay@Home or playing a video game that utilizes it to 100%.

That's weird. What's the power limit on your CPU? Maybe it maxes out during Cinebench as well. As for CPU-Z, I agree. It has a nice little benchmark tool, but not good for stability/thermal testing.

Well, I currently have set PL1 to 75 watts and PL2 to 80 watts, Tau is Intel spec so probably 28 seconds. Anyway, my cooling handles this chip with power limits lifted it's just that it tops out at high 70s, maybe low 80s. I prefer not to exceed 70C as is is recommended temperature limit for maximum turbo boost, else cooler with stock fan curve will be at 100% speed. And it's nice for VRMs too as I likely have one of those boards, that can't handle i9k.

 

[eidairaman1]

Thick fans don’t need high revs to push a lot of air, and thick fans don’t have to be loud either, that is an extremely common misperception.

That is true, fans with a larger cross section do not, also there is blade pitch/Angle of attack that play in moving air/noise.

I had a 60 mm delta that was thick, was a server fan and noisy!!!

Very high revs.

Ive seen other thicker larger fans move the same volume of air at lesser rpm.

 

[natr0n]

I dont even use doors/panels on my cases.

I feel like this is more of a peer pressure scenario.