Not really an official guide, but my method has always been set the minimum to the lowest rpm where the fans start to become noticably audible. I then plot a linear chart up to 100% fan speed at about 5c before the highest temperature on the component that I'm willing to tolerate. I find most of the time this keeps the fan at or close to the minimum while allowing a linear ramp up when thermals start to increase.
Yeah for sure, my old rig ROG strix B550-E fan curve at defaults was perfect but my new build Gigabyte Aorus Pro AX B650 even the silent curve sounds like a friggen server farm and custom curve was absolutely required. Thankfully that board saves the curve profiles separate for the main bios so if u have to flash or reset it it keeps it and even export to usb !
I'm an HVAC Tech. and its impressive how much you know about CFM and Positive, negative, neutral pressure. I know it's because your in the industry but it seems as tho you've also done your homework. And this is why I love your channel. Thanks for the video Jay. Very helpful for people who don't know how to do this. I HOPE people know this and it didn't have to be said but DON'T USE YOUR VAPE to test for the pressure in your system. It's vapor, it will stick to the components. Use an incent, it will not harm the system. Just saying this cuz I know how popular vape is right now and people might attempt this and not realize its a vapor...it's sticky...its not good.
Noticing this.. been vaping in my room the past couple of years. Noticed this molasses like stuff on the fans blades during quarterly clean up.. inside still looks mint though.
Just get a pc blower and blow it out once every couple of months to keep it fresh. I run no fan filters, they add noise, they cost airflow, they cause higher temps. Clean it!
@@jondonnelly3 I do that regularly but what I didn't notice was in-take v exhaust balance and I just finished setting it and it is cooling much better than before. I put in-take at 75% and exhaust at 60%. I'm still working on the exact numbers, but this gives an important insight about how to keep system cool with a little tweak. Thanks to Jay.
That tip about using an incense stick to check your vents is extremely useful, just found 1 tiny vent in the back of my case that pulls in about 2% of the smoke that goes by it. Time to completely redesign my setup!
A lot of people forget about their power supply fan. Depending on the PSU orientation, it may be pulling air from inside the case and pushing it out, so 2 in the front, 1 out the rear, 1 out the PSU, so it becomes close to neutral. Luckily some cases (like the one Jay is showing here) have it turned so the PSU intake pulls from outside and exhausts back out right away with minimal/no air from inside the case.
I call cases that have PSUs that suck air in directly from the bottom of the case vacuum cleaners as they suck in as much dust as possible, like a vacuum. (much worse if there are many fans at the bottom) It does limit that dirt contamination to the PSU though, but who needs a reliable PSU anyway. Use air cooling for the CPU and move all air from front to back like all the big boys do with their rack mounted equipment.
Would love to hear some more specifics on this topic. Such as: - Fan curve optimization - How an air cooler fan orientation would affect air flow - How each of the pressures affect the life of a fan - How RPM importance does or does not change with fans built for static pressure - And more I've always preferred air cooled systems over water cooled simply for the ease of maintenance. Water cooled systems can perform better and look amazing, but if something needs to be worked on, it seems that air cooled systems are easier. Thanks for the information, always appreciate it.
I went with an aio for the Intel build I did last fall. I'm re-housing everything into a slightly larger case (NR200 from Dan A4 H2O) and going back to air cooled. I'm a trucker by trade, and the aio just had me nervous as hell. The reduced airflow in the truck made passively cooled components run warmer, and I've managed to kill my boot drive (Samsung EVO 970 NVMe M.2 or whatever number) probably from the ambient temps being too high in that case. I'm going back to air cooling and even going with a bigger case than I'd prefer just to make it happen (and hopefully not kill any more components).
@@Just_Call_Me_Tim can just use a torrent nano, that single 180 gives insane positive airflow. Compact torrent is my recommendation tho as its more foolproof for upgrades.
"but if something needs to be worked on, it seems that air cooled systems are easier." My experience has been the exact opposite. I have my radiator mounted at the front panel, which is mesh. The only thing on my motherboard is the contact plate and the liquid housing overit, everything else is open for me to work on. So much easier than a massive air-cooler like an NH-D15. I don't even have to remove the contact plate to clean out the radiator.
Good topics. I've got a big dual tower air cooler but the weight Hanging off the mobo makes me a little uneasy. The problem is, air coolers never go bad and fans can be replaced in 15 minutes. AIO are awesome but knowing they eventually run low on fluid bothers me. Plus more points of failure.
I like and have run both but last build went with a Noctua nh-12a which is a twin tower but not as massive as the D15 and has good ram clearance. I worked computer infrastructure at a smelting company that ran 24/7 all of our servers and shop floor computers have air coolers.
I've always used cheap thin tissue paper to test pressure. Let it dangle flat in front of the gaps in your case and it will get pushed away or pulled towards the gap depending on which way the air is flowing. The hard part with case pressure is factoring in fan curves. Pressure may be very different under high load.
@@manuelcastle yes, you can, but you don't want to. besides looking fugly, if you happen to create negative pressure in your case, and seal off any gaps your case has, the only way in for air is via your intake fans, and the only way out is your exhaust fans. problem is, you already take more air out than you take in, because your intake is underperforming, and if pressure in your case gets too low, atmospheric pressure will push back against your exhaust fans more and more, until an equilibrium is created. since your underperforming intake will still suck air in, air in your case keeps moving and cooling, but you waste energy, because your exhaust is capped at the amount of work your intake performs. Now, that energy waste doesn't sound too bad, but it adds up, and additionally, the constant backpressure your exhaust fan needs to work against increases mechanic stress and might cause it to age faster / break sooner. None of this would be catastrophic for your system, but since it just takes a few mouse clicks for rpm curve adjustments and a couple of lines to read about airflow optimization, it would be a shame to not avoid it.
Slightly positive pressure is the best as fans are designed to push against a "load". Load in this case, being the air mass. You can find a fan spec called "back pressure" and that indicates how much pressure the fan can create or push against. When you have a positive pressure inside the computer case, or any other electronic device, it also means that the impedance match between fan wings and the air is better and this reduces noise, sometimes by a considerable margin. I work with this constantly on my field of pro audio devices and optimizing the cooling vs airflow vs noise is a constant battle.
I want to say something as a mechanical enginner who gets fluid dynamics and heat transfer lessons. Air flow type is very important for heat transfer. Turbulence will cause more heat transfer than linear air flow. I don't tested it with pc cases, I am talking with theories. Create a turbulance in your case with a bit possitive pressure and get lower temperatures. Why possitive pressure is better? Because high density air means more contact to hot surfaces. But very high pressure will cause more hot in the case air. I can recommend 3 in 2 out or 4 in 3 (maybe 2) out fans for possitive pressure and this fans must be a little randomised directions for creating a turbulence.
@@philosocio I wouldn't because I think there is already enough turbulent airflow even with the most perfect of setups. the one which moves the air out and over the components is the best imo.
I've been building my own PC for about 30 years, but I haven't updated my rig in about 5 years. I love all the great content around airflow, parts, and this video specifically was very helpful. Keep up the great (educational) work.
little side note about Positive Air pressure. it will turn your computer case into a mini clean room, provided you have proper dust covers on your intake. The advantage there is no particles can enter your system keeping it mostly dust and dirt free. The small particles and smoke (smoke is notoriously small) will still get in from the intake.
This is why I love having an Aquaero in my build. The level of control and monitoring it gives you really lets you experiment and monitor everything together to see the effect of little changes to RPM. They also have the ability to make synthetic sensors from multiple inputs and use those.
Same for the OCTO. Aquasuite and it's Playgound and PID controls is just a league beyond anything else. It's one of those things that once you've used it there's just no way back.
as an air cooling purist; this kind of information is SO useful to both people like me, who might need a refresher by the time we end up messing with our configs again, and new people in the PC game who want to make sure they get the most out of their system and have it as optimized as possible!
Im pretty satisfied with my airflow. Currently running 3 140mm intake fans in the front, one 140mm exhaust fan in the back, a NZXT Kraken Z63 with a pull configuration on top acting as exhaust, my front and back fans all run at max speed constantly, with the AIO being regulated on a case by case basis with different profiles, all depending on the load and CPU temperature. The case im using is a Corsair 7000D Airflow case.
Great content, as usual, Jay & Co! As someone who's tried to do the liquid cooling thing with 12th Gen Intel in a Dan A4 H2O case, I can tell everyone here that you still need good airflow for all the passively cooled components. Liquid cooling is badass for the components on the loop, but you still have to keep air flow over your board. This is especially true if you've got M.2 drives... I'm re-housing my system in an NR200 to have more room for airflow, and because I've managed to straight murder my boot drive. Live and learn, huh? That and stupidly (almost) lose almost 2 TB of stuff (had it all backed up already, but still miffed that I've got to go back to the drawing board). Having a computer building hobby as an OTR trucker isn't easy.
Good reminder, I will re-attempt water, thought board devices might be an issue. M.2 surprises me, though, I got one in a slot card under the gpu so not exactly in a well ventilated space, smart values are A-ok. Transcend TLC. May I ask, what make/model was yours? Did SMART readings confirm temp spiked?
An element I have went over is fan placement. The combination of intake and exhaust fans create an air current, which you want that air current to flow over the components. Air currents direct the air flow, which is a major case for neutral pressure. As such, you want your intakes at the front and bottom and exhaust over the back and top. This comes up to one little problem in fan placements on larger cases. The larger cases can place 3 fans across the top, which sounds good for exhaust, except that modern cases don't typically have external drive bays anymore, as optical drives are less used for transferring data and it is difficult getting a PC to play Blu-ray movies. That 3rd fan on the top is in the position of where the drive bays would be. Commonly, the motherboard is under the rear-top fan and middle-top fan, not the front-top fan. Placing an exhaust fan in that front spot on the top side pulls the air current from the top most front fan, directing its air pressure away from the motherboard. I have recommended leaving this spot vacant with the exception of using the spot for a liquid cooling radiator. At least with the radiator, the air will be cooling something. Another concern of fan placement is the video card. The high end video cards are rather large, and that size creates a wall inside the case. You will have to study the cooling fans to ensure where the video card's cooler's intake and exhaust are. You may have that bottom-front case fan bringing air that only diverts to the GPU and gets exhausted out the card slot. This means the upper front case fans are the intake for the CPU and motherboard. The concern of positive, neutral, and negative pressure gets more complicated when you have cards acting as walls inside the case, as you can have positive pressure on the bottom half of the case (assume PC is vertical) and negative pressure above the video card.
I am rather curious how a gpu with exhaust fans facing downwards vs a vertical mounted bracket gpu with fans facing outward affects the wall/negative airflow below?
The video is great for beginners, but if you're more advanced, here are a few key points to consider: 1. Case selection: The choice of case can make a significant difference. For example, I recently switched from the Antec 300 to the Torrent Compact, and the improvement in performance was remarkable. 2. PWM fans: It's important to have all fans equipped with PWM functionality, indicated by the presence of 4 pins. This allows for easy fan control through the operating system. 3. Fan control software: There are many free programs available that allow you to control your fans. Personally, I prefer to use a paid program called "Argus Monitor" for one specific reason: it allows you to customize fan control based on individual sensors. In my case, I use GPU temperatures to regulate the speed of all case fans. Overall, these are important factors to consider when optimizing your system's cooling and achieving better performance.
Speedfan has been amazing for me, allowed me to set curves up for non PWM Arctic F12's based on individual sensors and custom sensors as well. It's really cool, only downside is it doesn't work in BIOS but that's an upside to me as it's a lot more easily customized since it's running while the PC is up
7:15 - whether it's set up as intake or exhaust, you will still have a certain amount of flow eddy loss with the grate next to the fan. Thanks for the great content!
Nice. Feels like that was the part 1, should have a part 2 where you show the Bios settings, talk voltage vs PWM, consider to ramp or step fan speeds, etc. With some really good case fans (Noctua and Be Quiet!) you can get a lot of air movement without much noise, where as a cheapie 5 pack of Artic P12's can be a bit hummy if you don't manage the RPM around their resonances.
I really like motherboards that have a fan smoothing profile. Now that CPUs are boosting so aggressively, my old computer would go "Moo" every few minutes because it didn't have a fan smoothing profile. My newest build has it, and it's quiet throughout normal usage and only rams up when it's under heavy load. The heatsink can absorb the thermal spikes pretty effectively.
I hear you man. The Arctic P12's are very good fans for value but the resonance is horrible when using smooth curve and it keeps changing the rpm. So step speed is a must to get some peace of mind - something like 700rpm on idle up to 50C CPU temp and then step up to like 1200rpm from there and stay there up to 70C. Obviously it also depens on the specific CPU used and what are thats CPU temps unders certain scenarios. But generally with arctics the default smooth curve sucks.
@@brudel001 do you have a link to a good video on how to do this? I have five P12 fans in a Deepcool CH510 digital mesh with an Asrock B650E + Ryzen 7600X
@@oscargranath93 hey, first of all gongrats on having Ryzen 7'th series 😁 Unfortunately I don't have link to the video with specific tutorial. Basically with ASrock you should have in bios something like FAN-tastic tuning tab where you can do the setup - if you search the google you can find the image. So in that menu on the bottom horizontally are temps and left side vertical is fan percentage. At the top theres the drop down menu where you can select what fans you are wanna tweak - default is CPU so you should select something like SYSFAN1 or whatever you have the casefans connected to. Easiest would be if you have arctic PST fans so you can daisy chain the fans together and just use one fan header on the motherboard and alter all the fans together. Otherwise your case might have a fanhub or you can use splitter cables. As for specific settings it depends on your CPU cooler and behaviour - what are the usual idle temps - what are the gaming temps. I have an Intel 12700K that idles around 25C and when gaming it goes allmost to 60C. So my setup for arctic casefans is that on 20/30/40/50C its 35% of fanspeed. Then at 50/60/70C its 60% of fanspeed. And up from that its a smooth curve since the fans would be so loud anyway that it doesn't matter anymore but mine never reaches that so its unimportant. So the 35% is for desktop usage, surfing, watching videos etc. The latter is for gaming or other heavy usage. One thing that I would say is that I would use more fans in that case - either 3 intakes and 3 exhaust or 4 and 4. And what I would suggest to everyone is to use rear exhaust fan even if using top AIO because rear exhaust is the main fan to pull awat the heat generated by backside of the high end graphics cards. Also if possible use vertical mount for the GPU - its should lower GPU temps around 3-5C. It may not sound much but since GPU's are hot anyway it just might be the difference if the GPU ramps up the fans even higher under load or not.
2:15 most likely because of where the "official measurement" comes from. Just look at rim sizes for cars or display sizes. When an industry set a standard that has been widely used from the beginning it's kinda difficult to undo this. That being said, here in EU we also use m³/h (cubic meters per hour).
More intake than exhaust for less dust build up. Rule of thumb - 2 intake for every 1 exhaust. AIOs are ideal as intake in front of case. Consider aftermarket fans like 127 CFM $40 Icegale Xtra 3 pack only if you have high end build, otherwise find a semi decent case with 3 or more pre-installed fans. Set smart fan curves in BIOS. There saved you time
Hi Jay, I am a Canadian with a Mechanical Technology background, so I deal with HVAC design. Although I tend to use CFM for design, the official metric unit for airflow is L/s
One thing you forgot to touch on is fan selection. Although fan designs these days mostly tend to be a balanced type, chosing high air flow fans or static pressure fans plays a big role in a build. For chassis with open mesh panels, air flow fans are optimal to quickly move air in and out. With restrictive chassis designs or when adding radiators for liquid cooling, you want static pressure fans to brute force the air through a restrictive surface that air flow fan designs struggle with.
Don't forget that you can control the fan headers in the BIOS as well. If you have a smart enough motherboard, then you may have a Windows based software that allows for BIOS control, meaning that you can do this while it's running. When planning my last build, I made sure to match 1:1 my fans for in/out so that I knew it would be a constant flow of air.
I've had bad experience tuning it in BIOS, where for example my ASUS board lets me auto tune fans which sets the min fan RPM to 30% for example, but manually tuning it won't let me set the fans as low for some unknown reason and tells me any value below 60% is not valid. Not even BIOS update fixed that.
Glad you mentioned RPM. In doing research for my PC build I was surprised to see RPM not mentioned much, I always thought that was an obvious solution.
Thanks for all you do for us, Jay!!! (and team) Deliberately built a positive pressure case based on your other videos. Eventually ended up with 2 120 in on the front, 1 140 in on the side, 1 90 out on the back, and 1 120 out on the top. The temp difference was stunning.
@@Yengineered it took some time. wish I'd thought about using incense. But, matches and smoking paper work, too. I still tweak the fan curve from time to time...even after 2 years. but yeah. I have slight positive air flow.
I've been re-using the same case and fans for over 15 years now, and a few days ago I discovered that the top fans were turned the wrong way. I've been running all intakes for 15 years. Maximum positive pressure!
Talking so much about fans made me a fan. Not everyone is able to teach or at least to explain understandably. Even if i knew a lot, i did listen every word to the end. Rounded up my knowledge! Very constructive from a to z 👍
This is important to note when you have AIO rad as exhaust as well. The CPU cooler fans are going to ramp up, and the case intake fans need to as well.
As soon as you add WCing into the equation the subject becomes more divisive. If you want the coolest air (room ambient) to blow through your radiator then warm air will enter the case via the radiator (unless its an external to external rad or in a separate chamber). If you want the warm air to be exhausted out from the radiator then it will be using warmer case ambient air to remove the heat via the rad.
If you're using water-cooled cpu and air-cooled gpu, it's better for overall temps (when gaming) to have the aio as exhaust rather than intake. Sure the cpu will be a bit warmer, but the gpu will be several degrees higher than the other way around.
Linus bought one and it didn't work well for what they wanted to show. A good one is ridiculously expensive and he will be making it part of Labs when they are set up for it.
Those turbulent flow would distribute the smoke so well, you just can't see the flow that easily. Heck, even Gamer Nexus use contraption and negative filter just to see where the flow goes on RTX cooling testing
@@olandersnake Agreed, the pricing most likely will be insane.. I wonder what does Mr. Matt Lee use for his airflow visuals. He`s got a very thick and very visible smoke in his videos
Been vaping since 2011 and i do my own eliquid, never ever had anything sticky or any gunk build up on my system, no filters as well, clean it once a year with brush and air compressor, done.
A small panel (I used acrylic) between the GPU shroud and side panel can work wonders preventing GPU exh re-intake with horizontal mount. Easy way to tell if this is happening is feeling the heat soak pattern on your side panel. If the warm area isn't above the GPU, it's constantly cycling at least some % the same air. 12 case fans didn't prevent this in my 5000d air. 1 acrylic panel did.
@@Mr.Fall1n1 Flush with the GPU shroud, just below where the GPU exhausts. I cut a rectangle piece of cardboard the size of a horizontal cross section of my case interior, then cut my GPU's shape out of that. Remade it in acrylic after.
@@Mr.Fall1n1 gotta have sufficient intake below that plane for me to recommend this though. Your GPU can probably out flow 1 120 at low rpm. Might be worth playing around with some cardboard.
This is exactly the information I needed, I'm planning on getting a new case with extra fans since the one I have was used for a Windows XP machine... Thank you Mr Jay
I have the fractal torrent for almost 1 year now. It's a very good choice, if you go for an air cooled system. It has the top of the case closed off and positive pressure. This means you will have very little dust build up.
@@TheRobstar1983 they don't sell it here in Argentina, and buying it overseas is very expensive... I was thinking about getting "matrexx 40" and add 2x140mm or 3x120mm to the front along the 120mm rear fan, since it's the best thing I could find for a thight budget
I have a Corsair 4000X tempered front panel and notice a 10 degree increase when the front glass panel is on. Thank goodness they allow you to remove it.
I took some extra temp probes with my Corsair module and put them in the top of the case, bottom of the case, and wiring compartment. Its amazing how different they all are within 24" of each other.
I already knew all this, but Jay is fun to watch even on boring topics. Either way, I have my setup with 3 120 fans intake on front, 2 140 on top exhaust, and one 120 rear exhaust. Fan curves, however, help me keep the pressure slightly positive (as tested with a super high tech device called a piece of tissue), as the intakes rev up and stay up slightly faster than the exhaust fans.
Yeah it seems like with this case you'd want the rear fan as intake with 2 exhaust on the top, either that, or you're going to need bigger/higher RPM fans for intake.
All the HVAC dad gamers are grabbing their manometers and checking case static pressure 😂. And jay is spot on, as a general rule when balancing building pressure we shoot for a slight positive. Your case should be the same. Great video again jay.
I just built my first PC and these videos have been so helpful! This looks like the case I bought as well, Fractal Pop, so the demonstration was extra valuable for me
Top notch stuff here. For me this goes hand-in-hand with glass panels. I habitually use old fashioned side panels due to the fact that I need the front panel to have 5.25" bays for my expansions(hot swappable dual 2.5" SATA bay and extra front USB sockets) and like being able to stick my intakes on the side panel. So yeah, glass panels get a no from me.
Great video Jay! Something I learned over the years in building maintenance and dealing with HVAC issues now and then plus PC building is about what your air is flowing through. That case and most of the ones I've had over the years have that flat perforated metal partially blocking the air flow to or from the fans. I cut that out and install the chrome metal finger guards so the air can flow freely and turbulence noise is reduced. None of those areas are visible so from the outside the case looks fine but everything inside stays cool and I can sleep with the PC running and barely hear anything. I know, that would offend the flashy RGB crowd nowadays with their clear cases so there would have to be some compromise between looks and function.
This. I modded the front fan mount on a Corsair H500 case because the mount blocked like 15-20% of both intake fans' inlets. Given that the case shipped with 200mm fans, having a mounting plate that supports 120 and 240mm fans was pretty ridiculous. Switched to Fractal Torrent full tower for my Threadripper 3960 air-cooled CPU build. Sorted now.
A detailed fan curve tuning video with that awesome fan control software you showed us recently would be great! I have all my fans set to “mix” in that software but I’m not sure how I would separately rune the intake and exhaust fans to control the pressure.
I recently upgraded my PSU and GPU and decided I wanted to actually optimize my cooling apart from fan count and layout. I booted into my BIOS after watching this video and tuned my fan curves. My intake case fans were already running slightly faster than the exhaust fans. After testing, at idle I have almost completely neutral pressure, under load I have slight positive pressure. Really glad I didn't have to do hours of fine tuning to get a solid fan curve setup.
Great video. I have the Lian li lancool 3 with 3 120 front intake fans and an 360 aio top mounted with 3 120 fans as exhaust. I set my front fans to a higher fan curve profile and my exhaust fans to a lower fan curve profile. It helped a lot with cooling.
Apprechiate the video. I have two fans top mounted as exhuast, and they have that dust cover or whatever, and whn you said you can barley blow through it, i took mine off and didnt realize how much air it restrics. from now on im only puting the top dust filter on when i just the computer down. Thats why i put in top mounted fans for better exhuast, since heat rises, but now without the dust cover its at least over 50 percent more airflow. (and no im not an expert, look at the spelling im too lazy to correct in this comment) XD cheers jay and everyone else reading this.
I typically run my PC’s with no dust filters. I find that mesh front panels do a pretty decent job of keeping dust out and I prefer to let my fans draw in and exhaust air with as little impedance as possible, despite the increased dust ingress. I usually run my CPU radiator to exhaust through the top of the case with no filter and I run three intake fans in the front, also without a filter, to bring in as much cool air as possible at lower RPM’s. It may require me to clean the dust out of the system slightly more often but the trade off is worth it to me. Taking 20-30 minutes to thoroughly clean out my computer every couple of months just isn’t a big deal to me and in the grand scheme of things, it probably only amounts to one or two extra cleanings per year. If you already take good care of your machine and regularly clean it out, you probably won’t even notice if it’s getting dustier because you already tend to clean with enough regularity that you don’t really let much dust build up between cleanings anyway. I think running filterless likely also improves noise levels. Adding barriers for moving air to overcome always results in more noise. I still use the filter for the power supply fan but I made sure to buy a PSU with significantly higher specs than I need, thus the PSU fan rarely even turns on and when it does turn on, it runs quietly enough that I never notice it.
Excellent lesson professor. Would love to see more of those videos. Like JayzTwoLessons, where you would give two videos/lessons about topics like this, system optimisation, fan curve, fan orientation etc.
I think Jays point about neutral pressure is a good one, as long as you realize that you cant actually perfectly balance your flows. You can get them close, but as he pointed out, your case isn't air tight and so any imbalance will result in a slight positive or negative pressure. And since you can't actually measure your inflows and outflows (only infer them from RPMs and fan curves) you're always going to be slightly off perfect balance. Don't kill yourself trying to reach perfection, you wont even know if you get it anyway.
neutral isnt even perfect u want a bit positive pressure do minimize dust... wich means that when u have same fans all over u can start with intake RPM > Exhaust Combined RPM... lets say u have 3 front fans an AIO as an ex haust with 3 fans and 1 back fan... lets way min RPM is 300 and max 2100 ... u can run the intake min. at 600 RPM AIO at 300RPM min and Echaust 300min too wich means that u have 300 RPM more intake than exhaust... at minimum so u can twiggle it so at all points u have positive pressure while cooling is also good
This video was enlightening! I was sticking fans here there everywhere without really understanding the importance of negative and positive pressure. Thanks so much!
This is an excellent video on a subject that no one else seems to talk about or be bothered about. This one is being bookmarked. I hope you do more videos on how to choose and optimize layout of fans.
All the videos I've seen on optimising case fans seem to be based on the PC being sat on the desk in the open. I'd be interested to see how this advice changes for people with their PCs under their desks or similar limited airflow situations, and how heat build-up around the case effects cooling.
Thats why you shouldnt put your desk airtight to the wall. Let the exhaust of the PC approxiamately 20cm away from the wall, and there will be no problem. Fresh cold air is sucked in in the front, blown out in the back. There it "collects" a little bit, but can escape between wall and desk. An even if this is just 2cm wide, its accross the whole lenght of your desk, so its a real big hole at all.
I made a cheap "air hockey" table top for my stand beside my desk with pegboard replacing the top surface, and a plug in fan inside. Replacing parts of cabinet walls with cheap pegboard could probably help out a ton of ppl.
My friend used to get really high temps after a couple hours of gaming, we couldn't figure out why, did hours of troubleshooting over discord, only to find out his pc was in a closed cabinet lol
As one dude named Paul says: KEEP. PC. OFF. FLOOR!!! Seriously, put it sideways under the monitor like the old times even, but for the love of God don't put it on the ground.
The case itself is key here. I had a Corsair 400C, mid tower. I had no real choice but to mount my NZXT Kraken to the front, therefore pulling air in over the CPU and pushing it over the GPU, as a full mobo made it impossible to install a top push/pull. So be it. Had that case for like 3 years or something. 5600x would sit around 50-60c, GPU always around the 75c mark. Due to wanting a new design and a white case, bought the Lian Li Dynamic O11, of course. 3 pull at the bottom directly on to the GPU, Kraken CPU 2 pull at the top, 3 side exhaust. CPU never goes over 50c, and my GPU sits around 60-65c under load. It's crazy.
I'd be curious how temps would be if you had the 3 side fans as intake. I feel like it would be slightly better as intake, but I'm sure it wouldn't make that much difference.
@@willj3145 Well then I'd have 3 intake at the bottom, 2 at the top, 3 at the side and no venting...I thought the angle of the 3 side venting I have setup would be an issue as the air has to I suppose..."bounce" off the glass and be pulled out and that would reduce efficiency but it really doesn't. And I don't have the patience or desire to switch the fans around now, as it's such a nuisance to do.
For a more basic approach, I have always done this: You should be pulling out as much air as you are putting in. So if you have 3 x 120mm pull air in the front of the case, plus whatever the power supply does, then you should have a minimum of 3 fans doing the same to exhaust. So a common config (one I have used a lot too) is a top mount radiator with two fans, and one back panel exhaust fan, combined with 3 fans on the front (pretty standard case design). I prefer to top mount radiators for the simple reason that you get thermodynamics onto your side as a bonus. Heat rises, so they heat rises from the water block to the radiator, and then the fans push the heat out of the top in it's natural direction of flow. What you lose in pushing through the radiator balances naturally to what might be lost because of the mesh at the intake. Generally it means that all three exhaust fans (two on rads, 1 on back panel) work together to create a natural direction of flow on the case, front to back and front to top. The power supply puts a little into the bottom and that is just bonus at this point. Could I suggest jay that you get out a thermal camera and show how the air moves? That might make the point much more clear. Showing a running system and showing the temps in different places would explain it much better!
Not really. The rad will of course restrict it a bit but you want that fresh air going in for the gpu with a nice blow. Otherwise your gpu will run hotter as its circulating air already heated by it.
@@jepulis6674 The top mount fans on the radiator would be blowing hot air OUT, not in. Heat rises, why fight against nature? Front mounted rads and trying to force cool air in somewhere else is wasting a lot of energy to try to do what nature will do for you.
Really nice explanation! I thankfully already thought about RPM as a Factor so i found the following Solution: My Case Fits 5 140mm Fans, i have 2 as Intake in Front, 2 as Exhaust on the Top and 1 as Exhaust on the Back. Same Size Same Series Same Manufacturer. The Only difference: The Intake Fans are high-rpm variants. So i managed to get a slightly positive Pressure Setup despite having more Exhaust than Intake fans and the Temps are very good
i think the best air cooling design is making the build flipped up side down so the PSU being on top, top being exhaust (PSU fan + 2 fans = 3 fans total) & bottom being air intake (3 fans with filter) then other sides being sealed
I removed some of my PCI covers and added an exhaust fan to help pull heat out from under the GPU. Surprisingly, it works. I'm using the Helios case and have two intake fans in the bottom. They pull cool air up and the "PCI exhaust fan" pulls that cool air under the GPU wich makes cool air available for the GPU fans.
I feel like that aditional fun you placed on the PCIe covers maybe working aginst your goal if its exhaust. I think it's pulling the cool air that was suposed to be pulled by the GPU fans and sending it out of the case. But if the GPU temps are getting lower after placing it there then great!!
@simonrikhotso5816 The GPU gets plenty of air. I have 3 Noctua fans in the front of the case, 2 at the bottom, and three on the top. My GPU temps went down about 2 degrees overall. I still have positive pressure inside the case, so the air is not just being sucked in and out of the case. All of my fans are Noctua fans and we all know their reputation.
Another advantage of positive pressure systems is that if the air pressure in the case is higher, that means the air is denser, and so it can pick up more heat to be transferred out of the case. Dense air carries heat better than thinner air.
@@ScottGrammer gas cannot have higher density without being compressed. The very minor compression from a fan blade is immediately lost as the air moves away from the blade. Air is compressible and the only way to have "pressure" in the case is by compressing the air. For a point of comparison, the Noctua NF-F12 has 2.4mm H2O of static pressure the F12 industrial PPC at 3000rpm has 7.6mm, Corsair SP120 has 3.1mm... 1PSI is over 700mm of H2O. The density of the air inside the case will be exactly the same as outside the case because there is no compression. What makes a difference for cooling is air flow, and more specifically velocity. Velocity and pressure are also inversely related, Bernoulli's principle.
@@DragonHide94 If the pressure inside the case is not higher than the pressure outside the case, even if only a tiny bit, then why does air leave the case through every orifice when an intake fan pulls air into the case? It's true that a fan (as opposed to an actual compressor) raises the air pressure only a smidge. But that smidge is there. It's not a case of no compression, just a case of very little compression.
@@ScottGrammer not enough pressure to make a difference. Because air is flowing there is not actually any pressure. If you seal the case to restrict flow the pressure builds a bit, but again, not enough to matter. If you completely restrict the flow, the fan can't actually move air and it comes back through the gaps in the fan itself. Even if you could significantly increase the density of the air, it wouldn't positively impact cooling performance. Air flow cools things down. Not pressure.
One thing I have been messing with lately is covering vented portions with no fans. The dust is pretty insane after just 1 month, but it shows me that it's really moving air now.
I plan to buy this, what do you think about this setup? Would you add or change something? My budget is about 1000 euros. GPU:Gigabyte RTX 3060 OC dual 12GB DDR6 CPU: Ryzen 5 5600 wraith cooler POWER: TUF Gaming 650W MOTHER BOARD: Gigabyte b550m-ds3h CASE: MS Fighter S300 NVMe: NV2 1TB M.2 RAM: Kingston Fury Beast 2 x 16GB
In terms of fan position I like to have two intake fans in front, one exhaust fan above the cpu cooler and one exhaust fan at the back behind the cpu cooler. That way it's neutral, I have the chimney effect and the air going directly through the cpu fans also are directed towards outside lead by the exhaust fan at the back.
Super common problem I see with front fan positive pressure setups is large gaps to the case interior which makes it easier for the front fans to recirc air instead of pulling fresh air. Blocking off those gaps can be a pretty big performance improvement on a lot of cases with restrictive fronts.
Back in high school, I worked in a bar/restaurant, and the owner knew I liked to futz with electronics, and he was a cheap, cheap man. This was in the 90s, when smoking was pretty much treated as, "turn a blind eye if they're drinkin'," mentality. (At least in my podunk area) So he asked me if he could give me a tube TV to look at, that was hanging over the bar "for years," and he couldn't figure out why it stopped working. He was genuinely flabbergasted that cigarette nastiness-accumulation was the culprit. Your little, "vape kills fans fast," aside reminded me of this, and I got a chuckle.
I’m at the end of a week long binge of reviews for everything I’m thinking of buying for my new build. I am going to have 2 180’s at the bottom pulling cold into the GPU. 2 140’s attached to a liquid AIO cooler at the front of the case. Finally the best 120mm fan on the market for an exhaust on the back. There is no top exhaust, the case is fractal torrent compact. It has a fan controller built to the case so I will play around with that. Great video, really put my mind at ease.
Hey jay! That’s for the info! Id love to see a video about ddr5 ram and the performance at different speeds like you did a few years ago with ddr4 to know when you are really not getting any benefit of going any higher! I’m looking at doing a system upgrade from 9th gen intel and it’s probably the one area I’m really uncertain about making a decision on!
@@SrMorais 😂 why because it's a fraction of the cost? Or its a few chemicals compared to 1000s ? Disposable vapes could be dangerous but real vapours are at 1% risk currently compared smokers and after discussions with Professors and Drs who've looked into it. The biggest risk is inhalation of bacteria if you don't keep your vape clean and regular change of liquid and coils.
Heat rises , Radiator should always be on top , 120% to 140% in @ 900 rpm, 100% out @ 1100 to 1400 rpm out , note - out take pressures MUST include power supply , also use magnetic sheet cut out & placed around radiator to isolate positive heat load out , use Shin Etsu X-23-7868-2D thermal grease (buy direct from them or will end up with fake trash) have i9 13900k at 91C @ 100% stress @ 8hrs in , REMEMBER "HEAT RISES", good luck .
Thanks for the tips I asked my buddy because he told me I should build a pc how to improve my cooling he said put my Willy in my case and spin it. So I really appreciate the advice
I run the cte c700 Thermaltake case. Best air flow case I have ever used. Motherboard is sideways, so when a massive gpu is installed, there are no pcie slot issues at all. At a very good price too. My new 14900ks(Intel replacement cpu) runs at 85c on full load using a push/pull 480 aio. There is a spot to put a fan on the backside of the case to help cool the cpu socket. It just works awesome! Just an FYI, it is not quiet. Put your headphones on!!!! Can run up to 14 fans in this thing!
The app fan control that Jay talked about a couple videos ago works fantastic! After watching that video i immediately started using it. Its awsome you can tinker for hours. Set it up so the fans go up onder load etc. great stuf look up the video, wont regret it
I've looked into RPM vs CFM curves quite a bit and within spec (I think from 20% to 100% PWM modulation) the plots usually are pretty close to linear. Thus, you can usually get away with calculating max CFM IN and max CFM OUT (sum of the CFMs of your intake/exhaust fans) and set the PWM for the exhaust 5% below parity. Example: max IN = 400 cfm max OUT = 500 cfm Parity setting (neutral pressure): IN 100%, OUT 80% (400/500 = 0.8) Positive pressure settings: Set the intake fan curve as you want it, then set the exhaust curve points to 75% of the corresponding intake curve points. If you wanna make extra certain or use fine dust filters, go 10% below parity. Should work just fine in most cases.
Whenever I do a build I use pieces of tape of the appropriate color on the inside of the case to block every opening of any size that I don't want air entering or exiting. I usually replace the front mesh with a 1/4" thick foam open cell air conditioner pre-filter mounted in a "creatively engineered" bracket of some kind.
Ever since I switched to an Azza Cast 808w I haven't had anymore issues with ventilation. Dust collection is minimal. A breeze to work with really (no pun intended).
Nice video. Want to point a single (and minor) mistake. The temperature heating up rate is not 1:1. You can consider 2 divisions of air (actually would need to consider infinitesimal divisions then integrate, but that would not fit in one comment :D), one near the hot component, with the temperature from component, one from outside with a lower temperature. Because physics they will go towards the same value (one will get colder, one will heat up), in a particular amount of time:t = (m1 * c1 * ΔT1) / (k * A * ΔT1) = (m2 * c2 * ΔT2) / (k * A * ΔT2) (sorry if I missed anything in formula... been like 20 years since I used it last time). Now if you keep t constant (the time it takes to replace the air is the same), and calculate T2 in report with T1 you will notice they are not proportional at all, much less 1 to 1.
I have to say that the part about those "candles" it imho makes the core fo this video's topic - alltogether with all that stuff it makes sense, you have noted the most important stuff: the airflow - positive, negative, neutral, then the fact that gaps in a PC case matter, the RPM of fans also being important. I think that this might inspire a lot of people who might have neglected a bit their fan setup to option for your tips because who knows, maybe one has built a PC and and just a small RPM fine tuning might help with those hot spots.
This was such a “duh” moment for me. I hadn’t thought to consider whether the pressure was positive or negative. Nor had I considered the consequences of negative pressure (dust and/or circular airflow). Really helpful video!
To go along with your comment about not vaping in the room with your computer, I would add that you shouldn't use spray adhesives or lacquers in the room where your computer is. My son was an art student and use a sort of spray adhesive to finished charcoal drawings so the charcoal didn't smug. This finishing "adhesive" collected to surfaces, most notably fans, filters, and heat-syncs and in turn, as you would expect, this collected an unbelievable amount of dust. It ended up frying his CPU. I replaced the fans, filters and CPU, took alcohol to the heat-syncs and PCI slots and finally rebuilt the system. What a mess!! I actually may still have pictures of it.
Man that vaping thing is SOOO TRUE, 3 vapers in my house. It makes the dust moist and super sticky and a pain in the ass to clean. I had to take apart AIO and GPU to clean individual rows of aluminum fins from getting gunked up. No more vaping near my rig without a window wide open lol.
I just build my first PC since an AMD 64 3700+ and I have to say it changed a lil bit since then but your video really did help, at least strengthen what I already thought I knew
I really enjoy videos like this. Really explained everything quickly and clearly. That incense trick is smart, I used this to optimise my airflow. Thanks Jay!
Good advice! One thing though - convection is not that weak of a force. It can make a difference. That's why your exhaust fan is rear top. And that's why you should keep it as an exhaust, not intake.
Hey Jay fan curve tuning would be a big hit with novice builders.
He did that a short while ago already.
Not really an official guide, but my method has always been set the minimum to the lowest rpm where the fans start to become noticably audible. I then plot a linear chart up to 100% fan speed at about 5c before the highest temperature on the component that I'm willing to tolerate. I find most of the time this keeps the fan at or close to the minimum while allowing a linear ramp up when thermals start to increase.
How to tune fan curve. 1) go to bios, 2) find fan curve, 3) set to 100% 4) cewling 💨❄️
Yeah for sure, my old rig ROG strix B550-E fan curve at defaults was perfect but my new build Gigabyte Aorus Pro AX B650 even the silent curve sounds like a friggen server farm and custom curve was absolutely required. Thankfully that board saves the curve profiles separate for the main bios so if u have to flash or reset it it keeps it and even export to usb !
I've always wished I could set the fan curve for case fans by gpu temperature.
I'm an HVAC Tech. and its impressive how much you know about CFM and Positive, negative, neutral pressure. I know it's because your in the industry but it seems as tho you've also done your homework. And this is why I love your channel. Thanks for the video Jay. Very helpful for people who don't know how to do this. I HOPE people know this and it didn't have to be said but DON'T USE YOUR VAPE to test for the pressure in your system. It's vapor, it will stick to the components. Use an incent, it will not harm the system. Just saying this cuz I know how popular vape is right now and people might attempt this and not realize its a vapor...it's sticky...its not good.
You could say your components will come to a sticky end 😊! Greets from New Zealand 🍻
Noticing this.. been vaping in my room the past couple of years. Noticed this molasses like stuff on the fans blades during quarterly clean up.. inside still looks mint though.
@@SupraSav that’s from your bong bruh
I like to call vape pens, douche flutes.
@@rhoadsy this comment screams boomer lmao
I'm genuinely learning a lot more about my system dust issue today than ever before. Thank you, Jay
Just get a pc blower and blow it out once every couple of months to keep it fresh. I run no fan filters, they add noise, they cost airflow, they cause higher temps. Clean it!
@@jondonnelly3 I do that regularly but what I didn't notice was in-take v exhaust balance and I just finished setting it and it is cooling much better than before. I put in-take at 75% and exhaust at 60%. I'm still working on the exact numbers, but this gives an important insight about how to keep system cool with a little tweak. Thanks to Jay.
@@IndianGamer001 Yes!! well said...
@@mikem9536 that's what happening and with the help of this video I am able to resolve it by setting neutral pressure.
Postive pressure. No dust issues.
That tip about using an incense stick to check your vents is extremely useful, just found 1 tiny vent in the back of my case that pulls in about 2% of the smoke that goes by it. Time to completely redesign my setup!
XD
A lot of people forget about their power supply fan. Depending on the PSU orientation, it may be pulling air from inside the case and pushing it out, so 2 in the front, 1 out the rear, 1 out the PSU, so it becomes close to neutral. Luckily some cases (like the one Jay is showing here) have it turned so the PSU intake pulls from outside and exhausts back out right away with minimal/no air from inside the case.
Hehe my PSU is upside down. Cuz it suck air from the bottom of the case and blow it backward without affecting case airflow
@@za_pravdu1943thanks for repeating the second half of old mates comment. I needed to hear it twice for it to sink into my brain.
Just like with names.
@@za_pravdu1943 @cmdr_talikarni ,how would psu need to be oriented,mine is Cooler Master MWE 750 V2 750 Watt 80 Plus Gold SMPS
I call cases that have PSUs that suck air in directly from the bottom of the case vacuum cleaners as they suck in as much dust as possible, like a vacuum. (much worse if there are many fans at the bottom) It does limit that dirt contamination to the PSU though, but who needs a reliable PSU anyway. Use air cooling for the CPU and move all air from front to back like all the big boys do with their rack mounted equipment.
Would love to hear some more specifics on this topic. Such as:
- Fan curve optimization
- How an air cooler fan orientation would affect air flow
- How each of the pressures affect the life of a fan
- How RPM importance does or does not change with fans built for static pressure
- And more
I've always preferred air cooled systems over water cooled simply for the ease of maintenance. Water cooled systems can perform better and look amazing, but if something needs to be worked on, it seems that air cooled systems are easier.
Thanks for the information, always appreciate it.
I went with an aio for the Intel build I did last fall.
I'm re-housing everything into a slightly larger case (NR200 from Dan A4 H2O) and going back to air cooled. I'm a trucker by trade, and the aio just had me nervous as hell. The reduced airflow in the truck made passively cooled components run warmer, and I've managed to kill my boot drive (Samsung EVO 970 NVMe M.2 or whatever number) probably from the ambient temps being too high in that case.
I'm going back to air cooling and even going with a bigger case than I'd prefer just to make it happen (and hopefully not kill any more components).
@@Just_Call_Me_Tim can just use a torrent nano, that single 180 gives insane positive airflow. Compact torrent is my recommendation tho as its more foolproof for upgrades.
"but if something needs to be worked on, it seems that air cooled systems are easier."
My experience has been the exact opposite. I have my radiator mounted at the front panel, which is mesh. The only thing on my motherboard is the contact plate and the liquid housing overit, everything else is open for me to work on. So much easier than a massive air-cooler like an NH-D15. I don't even have to remove the contact plate to clean out the radiator.
Good topics. I've got a big dual tower air cooler but the weight Hanging off the mobo makes me a little uneasy. The problem is, air coolers never go bad and fans can be replaced in 15 minutes. AIO are awesome but knowing they eventually run low on fluid bothers me. Plus more points of failure.
I like and have run both but last build went with a Noctua nh-12a which is a twin tower but not as massive as the D15 and has good ram clearance. I worked computer infrastructure at a smelting company that ran 24/7 all of our servers and shop floor computers have air coolers.
I've always used cheap thin tissue paper to test pressure. Let it dangle flat in front of the gaps in your case and it will get pushed away or pulled towards the gap depending on which way the air is flowing.
The hard part with case pressure is factoring in fan curves. Pressure may be very different under high load.
Can I use painters tape to fill in the gaps?
@@manuelcastle yes, you can, but you don't want to.
besides looking fugly, if you happen to create negative pressure in your case, and seal off any gaps your case has, the only way in for air is via your intake fans, and the only way out is your exhaust fans. problem is, you already take more air out than you take in, because your intake is underperforming, and if pressure in your case gets too low, atmospheric pressure will push back against your exhaust fans more and more, until an equilibrium is created.
since your underperforming intake will still suck air in, air in your case keeps moving and cooling, but you waste energy, because your exhaust is capped at the amount of work your intake performs. Now, that energy waste doesn't sound too bad, but it adds up, and additionally, the constant backpressure your exhaust fan needs to work against increases mechanic stress and might cause it to age faster / break sooner.
None of this would be catastrophic for your system, but since it just takes a few mouse clicks for rpm curve adjustments and a couple of lines to read about airflow optimization, it would be a shame to not avoid it.
I use Argus monitor for fan curves, it's amazing
Jay uses 3x incense, I think that's overkill, 1x incense is sufficient :P
Slightly positive pressure is the best as fans are designed to push against a "load". Load in this case, being the air mass. You can find a fan spec called "back pressure" and that indicates how much pressure the fan can create or push against. When you have a positive pressure inside the computer case, or any other electronic device, it also means that the impedance match between fan wings and the air is better and this reduces noise, sometimes by a considerable margin. I work with this constantly on my field of pro audio devices and optimizing the cooling vs airflow vs noise is a constant battle.
Thanks for this info dude, got my new fans cooling my rig perfectly and silent as well
For beginners, don’t worry. You’ll figure it out in time.
So whts the slight positive setting for case with 3 front , 2 top and 1 back fan ??
@harrikirvesniemi9685 only the back one is smaller
Can I ask which pro audio company do you work for? It sounds like and interesting job
And, before using the incense, make sure it is, in fact, incense and not a leftover sparkler from the last 4th of July. :P
Lol... No harm in pointing that out... 🤣
I want to say something as a mechanical enginner who gets fluid dynamics and heat transfer lessons. Air flow type is very important for heat transfer. Turbulence will cause more heat transfer than linear air flow. I don't tested it with pc cases, I am talking with theories. Create a turbulance in your case with a bit possitive pressure and get lower temperatures. Why possitive pressure is better? Because high density air means more contact to hot surfaces. But very high pressure will cause more hot in the case air. I can recommend 3 in 2 out or 4 in 3 (maybe 2) out fans for possitive pressure and this fans must be a little randomised directions for creating a turbulence.
Are you sure about that
@@EpicBunty theory says that. test and see practical results.
@@philosocio I wouldn't because I think there is already enough turbulent airflow even with the most perfect of setups. the one which moves the air out and over the components is the best imo.
I've been building my own PC for about 30 years, but I haven't updated my rig in about 5 years. I love all the great content around airflow, parts, and this video specifically was very helpful. Keep up the great (educational) work.
It shouldn't take you 30 years to build a pc bro
@@badabing8152I bet it's the coolest PC ever
@@badabing8152 he is learning.
@@RainyFoxUwUits truly best. There is one person who has been taking care of his PC for 30 years.
I build the first computer for the nazis.
little side note about Positive Air pressure. it will turn your computer case into a mini clean room, provided you have proper dust covers on your intake. The advantage there is no particles can enter your system keeping it mostly dust and dirt free. The small particles and smoke (smoke is notoriously small) will still get in from the intake.
I remember my PC from 2001 with 2cm thick dust everywhere inside. There was no intake fans at all
This is why I love having an Aquaero in my build. The level of control and monitoring it gives you really lets you experiment and monitor everything together to see the effect of little changes to RPM. They also have the ability to make synthetic sensors from multiple inputs and use those.
Same for the OCTO. Aquasuite and it's Playgound and PID controls is just a league beyond anything else. It's one of those things that once you've used it there's just no way back.
as an air cooling purist; this kind of information is SO useful to both people like me, who might need a refresher by the time we end up messing with our configs again, and new people in the PC game who want to make sure they get the most out of their system and have it as optimized as possible!
Im pretty satisfied with my airflow.
Currently running 3 140mm intake fans in the front, one 140mm exhaust fan in the back, a NZXT Kraken Z63 with a pull configuration on top acting as exhaust, my front and back fans all run at max speed constantly, with the AIO being regulated on a case by case basis with different profiles, all depending on the load and CPU temperature.
The case im using is a Corsair 7000D Airflow case.
Great content, as usual, Jay & Co!
As someone who's tried to do the liquid cooling thing with 12th Gen Intel in a Dan A4 H2O case, I can tell everyone here that you still need good airflow for all the passively cooled components.
Liquid cooling is badass for the components on the loop, but you still have to keep air flow over your board. This is especially true if you've got M.2 drives...
I'm re-housing my system in an NR200 to have more room for airflow, and because I've managed to straight murder my boot drive.
Live and learn, huh? That and stupidly (almost) lose almost 2 TB of stuff (had it all backed up already, but still miffed that I've got to go back to the drawing board).
Having a computer building hobby as an OTR trucker isn't easy.
Good reminder, I will re-attempt water, thought board devices might be an issue. M.2 surprises me, though, I got one in a slot card under the gpu so not exactly in a well ventilated space, smart values are A-ok. Transcend TLC. May I ask, what make/model was yours? Did SMART readings confirm temp spiked?
An element I have went over is fan placement. The combination of intake and exhaust fans create an air current, which you want that air current to flow over the components. Air currents direct the air flow, which is a major case for neutral pressure. As such, you want your intakes at the front and bottom and exhaust over the back and top.
This comes up to one little problem in fan placements on larger cases. The larger cases can place 3 fans across the top, which sounds good for exhaust, except that modern cases don't typically have external drive bays anymore, as optical drives are less used for transferring data and it is difficult getting a PC to play Blu-ray movies. That 3rd fan on the top is in the position of where the drive bays would be. Commonly, the motherboard is under the rear-top fan and middle-top fan, not the front-top fan. Placing an exhaust fan in that front spot on the top side pulls the air current from the top most front fan, directing its air pressure away from the motherboard. I have recommended leaving this spot vacant with the exception of using the spot for a liquid cooling radiator. At least with the radiator, the air will be cooling something.
Another concern of fan placement is the video card. The high end video cards are rather large, and that size creates a wall inside the case. You will have to study the cooling fans to ensure where the video card's cooler's intake and exhaust are. You may have that bottom-front case fan bringing air that only diverts to the GPU and gets exhausted out the card slot. This means the upper front case fans are the intake for the CPU and motherboard. The concern of positive, neutral, and negative pressure gets more complicated when you have cards acting as walls inside the case, as you can have positive pressure on the bottom half of the case (assume PC is vertical) and negative pressure above the video card.
I am rather curious how a gpu with exhaust fans facing downwards vs a vertical mounted bracket gpu with fans facing outward affects the wall/negative airflow below?
@@centurion8446 Are you sure those fans are exhaust fans?
The video is great for beginners, but if you're more advanced, here are a few key points to consider:
1. Case selection: The choice of case can make a significant difference. For example, I recently switched from the Antec 300 to the Torrent Compact, and the improvement in performance was remarkable.
2. PWM fans: It's important to have all fans equipped with PWM functionality, indicated by the presence of 4 pins. This allows for easy fan control through the operating system.
3. Fan control software: There are many free programs available that allow you to control your fans. Personally, I prefer to use a paid program called "Argus Monitor" for one specific reason: it allows you to customize fan control based on individual sensors. In my case, I use GPU temperatures to regulate the speed of all case fans.
Overall, these are important factors to consider when optimizing your system's cooling and achieving better performance.
A lot of programs will come with fan control, I think my Asus software can customize on sensors but I don't remember
Speedfan has been amazing for me, allowed me to set curves up for non PWM Arctic F12's based on individual sensors and custom sensors as well. It's really cool, only downside is it doesn't work in BIOS but that's an upside to me as it's a lot more easily customized since it's running while the PC is up
@3meraldDoughnut Yes it does, the aura app can even control light depending on temperature
ChatGPT lol
@@edwinjaner5978😂😂😂
7:15 - whether it's set up as intake or exhaust, you will still have a certain amount of flow eddy loss with the grate next to the fan. Thanks for the great content!
Nice. Feels like that was the part 1, should have a part 2 where you show the Bios settings, talk voltage vs PWM, consider to ramp or step fan speeds, etc. With some really good case fans (Noctua and Be Quiet!) you can get a lot of air movement without much noise, where as a cheapie 5 pack of Artic P12's can be a bit hummy if you don't manage the RPM around their resonances.
I really like motherboards that have a fan smoothing profile. Now that CPUs are boosting so aggressively, my old computer would go "Moo" every few minutes because it didn't have a fan smoothing profile.
My newest build has it, and it's quiet throughout normal usage and only rams up when it's under heavy load. The heatsink can absorb the thermal spikes pretty effectively.
@@YengineeredThis is why I have my fans step up in 12 second intervals (the lowest amount of time on my chassis fans).
I hear you man. The Arctic P12's are very good fans for value but the resonance is horrible when using smooth curve and it keeps changing the rpm. So step speed is a must to get some peace of mind - something like 700rpm on idle up to 50C CPU temp and then step up to like 1200rpm from there and stay there up to 70C. Obviously it also depens on the specific CPU used and what are thats CPU temps unders certain scenarios. But generally with arctics the default smooth curve sucks.
@@brudel001 do you have a link to a good video on how to do this? I have five P12 fans in a Deepcool CH510 digital mesh with an Asrock B650E + Ryzen 7600X
@@oscargranath93 hey, first of all gongrats on having Ryzen 7'th series 😁 Unfortunately I don't have link to the video with specific tutorial. Basically with ASrock you should have in bios something like FAN-tastic tuning tab where you can do the setup - if you search the google you can find the image.
So in that menu on the bottom horizontally are temps and left side vertical is fan percentage. At the top theres the drop down menu where you can select what fans you are wanna tweak - default is CPU so you should select something like SYSFAN1 or whatever you have the casefans connected to. Easiest would be if you have arctic PST fans so you can daisy chain the fans together and just use one fan header on the motherboard and alter all the fans together. Otherwise your case might have a fanhub or you can use splitter cables.
As for specific settings it depends on your CPU cooler and behaviour - what are the usual idle temps - what are the gaming temps. I have an Intel 12700K that idles around 25C and when gaming it goes allmost to 60C. So my setup for arctic casefans is that on 20/30/40/50C its 35% of fanspeed. Then at 50/60/70C its 60% of fanspeed. And up from that its a smooth curve since the fans would be so loud anyway that it doesn't matter anymore but mine never reaches that so its unimportant. So the 35% is for desktop usage, surfing, watching videos etc. The latter is for gaming or other heavy usage.
One thing that I would say is that I would use more fans in that case - either 3 intakes and 3 exhaust or 4 and 4. And what I would suggest to everyone is to use rear exhaust fan even if using top AIO because rear exhaust is the main fan to pull awat the heat generated by backside of the high end graphics cards. Also if possible use vertical mount for the GPU - its should lower GPU temps around 3-5C. It may not sound much but since GPU's are hot anyway it just might be the difference if the GPU ramps up the fans even higher under load or not.
2:15 most likely because of where the "official measurement" comes from. Just look at rim sizes for cars or display sizes. When an industry set a standard that has been widely used from the beginning it's kinda difficult to undo this. That being said, here in EU we also use m³/h (cubic meters per hour).
m³/h is the SI unit for volumetric flow rate so as official as it gets. Last time I checked Noctua uses it, not sure about others.
There's also the scaled Litres per second. 36 m3/h is 10l/s ≈ 21.1888 CFM
@@v1nd1c4r3 Every fan I've ever seen uses CFM. I just put a PC together now. Everything was CFM.
It's litre per minute
More intake than exhaust for less dust build up. Rule of thumb - 2 intake for every 1 exhaust. AIOs are ideal as intake in front of case. Consider aftermarket fans like 127 CFM $40 Icegale Xtra 3 pack only if you have high end build, otherwise find a semi decent case with 3 or more pre-installed fans. Set smart fan curves in BIOS. There saved you time
Hi Jay, I am a Canadian with a Mechanical Technology background, so I deal with HVAC design. Although I tend to use CFM for design, the official metric unit for airflow is L/s
And yet metric computer fan specs seem to use m^3/hour. Ugh.
One thing you forgot to touch on is fan selection. Although fan designs these days mostly tend to be a balanced type, chosing high air flow fans or static pressure fans plays a big role in a build. For chassis with open mesh panels, air flow fans are optimal to quickly move air in and out. With restrictive chassis designs or when adding radiators for liquid cooling, you want static pressure fans to brute force the air through a restrictive surface that air flow fan designs struggle with.
Don't forget that you can control the fan headers in the BIOS as well. If you have a smart enough motherboard, then you may have a Windows based software that allows for BIOS control, meaning that you can do this while it's running. When planning my last build, I made sure to match 1:1 my fans for in/out so that I knew it would be a constant flow of air.
yes. Many motherboards can control the speed even for 3 wired fans, which used to run at fixed speeds.
thanks you for this comment I just changed my settings
Yup I like setting my fan curves in the BIOS 100 times better that some of that shitty motherboard software in windows, it has not let me down yet :-)
I've had bad experience tuning it in BIOS, where for example my ASUS board lets me auto tune fans which sets the min fan RPM to 30% for example, but manually tuning it won't let me set the fans as low for some unknown reason and tells me any value below 60% is not valid. Not even BIOS update fixed that.
Glad you mentioned RPM. In doing research for my PC build I was surprised to see RPM not mentioned much, I always thought that was an obvious solution.
Thanks for all you do for us, Jay!!! (and team)
Deliberately built a positive pressure case based on your other videos.
Eventually ended up with 2 120 in on the front, 1 140 in on the side, 1 90 out on the back, and 1 120 out on the top.
The temp difference was stunning.
Did you use anything to test for positive air pressure like a piece of paper? It must have been tedious to optimise.
@@Yengineered it took some time. wish I'd thought about using incense. But, matches and smoking paper work, too. I still tweak the fan curve from time to time...even after 2 years. but yeah. I have slight positive air flow.
I've been re-using the same case and fans for over 15 years now, and a few days ago I discovered that the top fans were turned the wrong way. I've been running all intakes for 15 years. Maximum positive pressure!
Talking so much about fans made me a fan.
Not everyone is able to teach or at least to explain understandably. Even if i knew a lot, i did listen every word to the end. Rounded up my knowledge!
Very constructive from a to z 👍
I always like when we revisits these topics because of all of the different case designs that keep launching. Thanks for this one!
This is important to note when you have AIO rad as exhaust as well. The CPU cooler fans are going to ramp up, and the case intake fans need to as well.
As soon as you add WCing into the equation the subject becomes more divisive. If you want the coolest air (room ambient) to blow through your radiator then warm air will enter the case via the radiator (unless its an external to external rad or in a separate chamber). If you want the warm air to be exhausted out from the radiator then it will be using warmer case ambient air to remove the heat via the rad.
@@superflyguy4488 What is WCing?
@@rawj1213 WC = water cooling.
@@rawj1213 👀
If you're using water-cooled cpu and air-cooled gpu, it's better for overall temps (when gaming) to have the aio as exhaust rather than intake. Sure the cpu will be a bit warmer, but the gpu will be several degrees higher than the other way around.
Hey Jay, as an idea for the airflow explanations - it may be useful to get a small smoke machine, to be able to visualize all kinds of airflow
Linus bought one and it didn't work well for what they wanted to show. A good one is ridiculously expensive and he will be making it part of Labs when they are set up for it.
Those turbulent flow would distribute the smoke so well, you just can't see the flow that easily.
Heck, even Gamer Nexus use contraption and negative filter just to see where the flow goes on RTX cooling testing
A crayon would be good too. Just kidding.. well not that much XD. Markers would be fine to draw the pathing.
Or use a CAD!
@@olandersnake Agreed, the pricing most likely will be insane.. I wonder what does Mr. Matt Lee use for his airflow visuals. He`s got a very thick and very visible smoke in his videos
Black case, white smoke, black light case lighting. Sounds like it’d be fun.
i just want to point out that the reaction to the dewalt electric leaf blower was incredible. Love your channel dude.
Been vaping since 2011 and i do my own eliquid, never ever had anything sticky or any gunk build up on my system, no filters as well, clean it once a year with brush and air compressor, done.
A small panel (I used acrylic) between the GPU shroud and side panel can work wonders preventing GPU exh re-intake with horizontal mount. Easy way to tell if this is happening is feeling the heat soak pattern on your side panel. If the warm area isn't above the GPU, it's constantly cycling at least some % the same air. 12 case fans didn't prevent this in my 5000d air. 1 acrylic panel did.
Can you elaborate please ? I don't really understand where the acrilic panel is situated .
@@Mr.Fall1n1 Flush with the GPU shroud, just below where the GPU exhausts. I cut a rectangle piece of cardboard the size of a horizontal cross section of my case interior, then cut my GPU's shape out of that. Remade it in acrylic after.
@@Mr.Fall1n1 gotta have sufficient intake below that plane for me to recommend this though. Your GPU can probably out flow 1 120 at low rpm. Might be worth playing around with some cardboard.
This is exactly the information I needed, I'm planning on getting a new case with extra fans since the one I have was used for a Windows XP machine... Thank you Mr Jay
I have the fractal torrent for almost 1 year now. It's a very good choice, if you go for an air cooled system. It has the top of the case closed off and positive pressure. This means you will have very little dust build up.
@@TheRobstar1983 they don't sell it here in Argentina, and buying it overseas is very expensive... I was thinking about getting "matrexx 40" and add 2x140mm or 3x120mm to the front along the 120mm rear fan, since it's the best thing I could find for a thight budget
I have a Corsair 4000X tempered front panel and notice a 10 degree increase when the front glass panel is on. Thank goodness they allow you to remove it.
The incense is a clever idea to test air flow. Thank you for sharing creative approaches.
I took some extra temp probes with my Corsair module and put them in the top of the case, bottom of the case, and wiring compartment. Its amazing how different they all are within 24" of each other.
I already knew all this, but Jay is fun to watch even on boring topics. Either way, I have my setup with 3 120 fans intake on front, 2 140 on top exhaust, and one 120 rear exhaust. Fan curves, however, help me keep the pressure slightly positive (as tested with a super high tech device called a piece of tissue), as the intakes rev up and stay up slightly faster than the exhaust fans.
Yeah it seems like with this case you'd want the rear fan as intake with 2 exhaust on the top, either that, or you're going to need bigger/higher RPM fans for intake.
@@mikem9536 that would work, but there is no dust filter on the back, so that just makes it worse faster
All the HVAC dad gamers are grabbing their manometers and checking case static pressure 😂. And jay is spot on, as a general rule when balancing building pressure we shoot for a slight positive. Your case should be the same. Great video again jay.
I just built my first PC and these videos have been so helpful! This looks like the case I bought as well, Fractal Pop, so the demonstration was extra valuable for me
11:25 "Neutral is ideal, slightly positive is best." - he actually said that sentence 🤣 It's all clear as day.
Top notch stuff here.
For me this goes hand-in-hand with glass panels. I habitually use old fashioned side panels due to the fact that I need the front panel to have 5.25" bays for my expansions(hot swappable dual 2.5" SATA bay and extra front USB sockets) and like being able to stick my intakes on the side panel. So yeah, glass panels get a no from me.
Yeah with glass panel it seems like you'd want the rear fan to be intake and exhaust out the top.
Great video Jay! Something I learned over the years in building maintenance and dealing with HVAC issues now and then plus PC building is about what your air is flowing through. That case and most of the ones I've had over the years have that flat perforated metal partially blocking the air flow to or from the fans. I cut that out and install the chrome metal finger guards so the air can flow freely and turbulence noise is reduced. None of those areas are visible so from the outside the case looks fine but everything inside stays cool and I can sleep with the PC running and barely hear anything. I know, that would offend the flashy RGB crowd nowadays with their clear cases so there would have to be some compromise between looks and function.
This. I modded the front fan mount on a Corsair H500 case because the mount blocked like 15-20% of both intake fans' inlets. Given that the case shipped with 200mm fans, having a mounting plate that supports 120 and 240mm fans was pretty ridiculous. Switched to Fractal Torrent full tower for my Threadripper 3960 air-cooled CPU build. Sorted now.
A detailed fan curve tuning video with that awesome fan control software you showed us recently would be great! I have all my fans set to “mix” in that software but I’m not sure how I would separately rune the intake and exhaust fans to control the pressure.
Hey man, what software is this?
I recently upgraded my PSU and GPU and decided I wanted to actually optimize my cooling apart from fan count and layout. I booted into my BIOS after watching this video and tuned my fan curves. My intake case fans were already running slightly faster than the exhaust fans. After testing, at idle I have almost completely neutral pressure, under load I have slight positive pressure. Really glad I didn't have to do hours of fine tuning to get a solid fan curve setup.
Great video. I have the Lian li lancool 3 with 3 120 front intake fans and an 360 aio top mounted with 3 120 fans as exhaust. I set my front fans to a higher fan curve profile and my exhaust fans to a lower fan curve profile. It helped a lot with cooling.
What kinda case do you use?
If I don't use aio for the top with the Lian Li 216... Can I just use regular fans?
Airflow is often overlooked. Thanks for taking the time to explain!
In Europe we use cubic metres per hour (or minute, usually hour) as an airflow measuring unit
Nice,
Really good and important basic information. Keep doing these, I bet there are a lot of people who can use this kind of tutorials. Keep 'em coming!
Apprechiate the video. I have two fans top mounted as exhuast, and they have that dust cover or whatever, and whn you said you can barley blow through it, i took mine off and didnt realize how much air it restrics. from now on im only puting the top dust filter on when i just the computer down. Thats why i put in top mounted fans for better exhuast, since heat rises, but now without the dust cover its at least over 50 percent more airflow. (and no im not an expert, look at the spelling im too lazy to correct in this comment) XD cheers jay and everyone else reading this.
I typically run my PC’s with no dust filters. I find that mesh front panels do a pretty decent job of keeping dust out and I prefer to let my fans draw in and exhaust air with as little impedance as possible, despite the increased dust ingress.
I usually run my CPU radiator to exhaust through the top of the case with no filter and I run three intake fans in the front, also without a filter, to bring in as much cool air as possible at lower RPM’s.
It may require me to clean the dust out of the system slightly more often but the trade off is worth it to me. Taking 20-30 minutes to thoroughly clean out my computer every couple of months just isn’t a big deal to me and in the grand scheme of things, it probably only amounts to one or two extra cleanings per year. If you already take good care of your machine and regularly clean it out, you probably won’t even notice if it’s getting dustier because you already tend to clean with enough regularity that you don’t really let much dust build up between cleanings anyway.
I think running filterless likely also improves noise levels. Adding barriers for moving air to overcome always results in more noise. I still use the filter for the power supply fan but I made sure to buy a PSU with significantly higher specs than I need, thus the PSU fan rarely even turns on and when it does turn on, it runs quietly enough that I never notice it.
How about doing an experiment with many different fan configs and measuring CPU and GPU temps. That would truly show the best setup.
Excellent lesson professor. Would love to see more of those videos. Like JayzTwoLessons, where you would give two videos/lessons about topics like this, system optimisation, fan curve, fan orientation etc.
I think Jays point about neutral pressure is a good one, as long as you realize that you cant actually perfectly balance your flows. You can get them close, but as he pointed out, your case isn't air tight and so any imbalance will result in a slight positive or negative pressure. And since you can't actually measure your inflows and outflows (only infer them from RPMs and fan curves) you're always going to be slightly off perfect balance. Don't kill yourself trying to reach perfection, you wont even know if you get it anyway.
neutral isnt even perfect u want a bit positive pressure do minimize dust... wich means that when u have same fans all over u can start with intake RPM > Exhaust Combined RPM... lets say u have 3 front fans an AIO as an ex haust with 3 fans and 1 back fan... lets way min RPM is 300 and max 2100 ... u can run the intake min. at 600 RPM AIO at 300RPM min and Echaust 300min too wich means that u have 300 RPM more intake than exhaust... at minimum so u can twiggle it so at all points u have positive pressure while cooling is also good
This video was enlightening! I was sticking fans here there everywhere without really understanding the importance of negative and positive pressure. Thanks so much!
This is an excellent video on a subject that no one else seems to talk about or be bothered about. This one is being bookmarked. I hope you do more videos on how to choose and optimize layout of fans.
All the videos I've seen on optimising case fans seem to be based on the PC being sat on the desk in the open. I'd be interested to see how this advice changes for people with their PCs under their desks or similar limited airflow situations, and how heat build-up around the case effects cooling.
Thats why you shouldnt put your desk airtight to the wall. Let the exhaust of the PC approxiamately 20cm away from the wall, and there will be no problem. Fresh cold air is sucked in in the front, blown out in the back. There it "collects" a little bit, but can escape between wall and desk. An even if this is just 2cm wide, its accross the whole lenght of your desk, so its a real big hole at all.
I made a cheap "air hockey" table top for my stand beside my desk with pegboard replacing the top surface, and a plug in fan inside. Replacing parts of cabinet walls with cheap pegboard could probably help out a ton of ppl.
My friend used to get really high temps after a couple hours of gaming, we couldn't figure out why, did hours of troubleshooting over discord, only to find out his pc was in a closed cabinet lol
@@squidwardo7074 😂😂😂
As one dude named Paul says: KEEP. PC. OFF. FLOOR!!!
Seriously, put it sideways under the monitor like the old times even, but for the love of God don't put it on the ground.
The case itself is key here. I had a Corsair 400C, mid tower. I had no real choice but to mount my NZXT Kraken to the front, therefore pulling air in over the CPU and pushing it over the GPU, as a full mobo made it impossible to install a top push/pull. So be it. Had that case for like 3 years or something. 5600x would sit around 50-60c, GPU always around the 75c mark. Due to wanting a new design and a white case, bought the Lian Li Dynamic O11, of course. 3 pull at the bottom directly on to the GPU, Kraken CPU 2 pull at the top, 3 side exhaust. CPU never goes over 50c, and my GPU sits around 60-65c under load. It's crazy.
I have the Fractal Torrent. 2x180's, 3x120's in. One 140 out the back. Does a good job.
I'd be curious how temps would be if you had the 3 side fans as intake. I feel like it would be slightly better as intake, but I'm sure it wouldn't make that much difference.
@@willj3145 Well then I'd have 3 intake at the bottom, 2 at the top, 3 at the side and no venting...I thought the angle of the 3 side venting I have setup would be an issue as the air has to I suppose..."bounce" off the glass and be pulled out and that would reduce efficiency but it really doesn't. And I don't have the patience or desire to switch the fans around now, as it's such a nuisance to do.
For a more basic approach, I have always done this: You should be pulling out as much air as you are putting in. So if you have 3 x 120mm pull air in the front of the case, plus whatever the power supply does, then you should have a minimum of 3 fans doing the same to exhaust. So a common config (one I have used a lot too) is a top mount radiator with two fans, and one back panel exhaust fan, combined with 3 fans on the front (pretty standard case design). I prefer to top mount radiators for the simple reason that you get thermodynamics onto your side as a bonus. Heat rises, so they heat rises from the water block to the radiator, and then the fans push the heat out of the top in it's natural direction of flow. What you lose in pushing through the radiator balances naturally to what might be lost because of the mesh at the intake. Generally it means that all three exhaust fans (two on rads, 1 on back panel) work together to create a natural direction of flow on the case, front to back and front to top. The power supply puts a little into the bottom and that is just bonus at this point.
Could I suggest jay that you get out a thermal camera and show how the air moves? That might make the point much more clear. Showing a running system and showing the temps in different places would explain it much better!
What if there's 3 fans pushing through the radiator?
@@ThaWiz11That might be a big enough difference to warrant testing having only two exhaust fans
Not really. The rad will of course restrict it a bit but you want that fresh air going in for the gpu with a nice blow. Otherwise your gpu will run hotter as its circulating air already heated by it.
@@jepulis6674 The top mount fans on the radiator would be blowing hot air OUT, not in. Heat rises, why fight against nature? Front mounted rads and trying to force cool air in somewhere else is wasting a lot of energy to try to do what nature will do for you.
Really nice explanation! I thankfully already thought about RPM as a Factor so i found the following Solution: My Case Fits 5 140mm Fans, i have 2 as Intake in Front, 2 as Exhaust on the Top and 1 as Exhaust on the Back. Same Size Same Series Same Manufacturer. The Only difference: The Intake Fans are high-rpm variants. So i managed to get a slightly positive Pressure Setup despite having more Exhaust than Intake fans and the Temps are very good
i think the best air cooling design is making the build flipped up side down so the PSU being on top, top being exhaust (PSU fan + 2 fans = 3 fans total) & bottom being air intake (3 fans with filter) then other sides being sealed
I removed some of my PCI covers and added an exhaust fan to help pull heat out from under the GPU. Surprisingly, it works. I'm using the Helios case and have two intake fans in the bottom. They pull cool air up and the "PCI exhaust fan" pulls that cool air under the GPU wich makes cool air available for the GPU fans.
I feel like that aditional fun you placed on the PCIe covers maybe working aginst your goal if its exhaust. I think it's pulling the cool air that was suposed to be pulled by the GPU fans and sending it out of the case. But if the GPU temps are getting lower after placing it there then great!!
Did the same in my system. It's a different case, but yeah, it works really well 👍🏻
Hey, it's great to hear I'm not the only one who jammed a 90mm fan on the unused PCI covers.
@simonrikhotso5816 The GPU gets plenty of air. I have 3 Noctua fans in the front of the case, 2 at the bottom, and three on the top. My GPU temps went down about 2 degrees overall. I still have positive pressure inside the case, so the air is not just being sucked in and out of the case. All of my fans are Noctua fans and we all know their reputation.
@@TheNetsrac hell yeah. I thought I was the only one. Would be nice if case manufacturers included a spot for a fan there considering SLI is dead.
Another advantage of positive pressure systems is that if the air pressure in the case is higher, that means the air is denser, and so it can pick up more heat to be transferred out of the case. Dense air carries heat better than thinner air.
True... Though I doubt anyone is going to be cooling their PC with fans that can actually compress air. Leaf blowers might work.
@@DragonHide94 Not a lot of compression, but the only way a fan moves air is to rarify the air behind it and compress the air in front of it.
@@ScottGrammer gas cannot have higher density without being compressed. The very minor compression from a fan blade is immediately lost as the air moves away from the blade. Air is compressible and the only way to have "pressure" in the case is by compressing the air. For a point of comparison, the Noctua NF-F12 has 2.4mm H2O of static pressure the F12 industrial PPC at 3000rpm has 7.6mm, Corsair SP120 has 3.1mm... 1PSI is over 700mm of H2O. The density of the air inside the case will be exactly the same as outside the case because there is no compression. What makes a difference for cooling is air flow, and more specifically velocity. Velocity and pressure are also inversely related, Bernoulli's principle.
@@DragonHide94 If the pressure inside the case is not higher than the pressure outside the case, even if only a tiny bit, then why does air leave the case through every orifice when an intake fan pulls air into the case? It's true that a fan (as opposed to an actual compressor) raises the air pressure only a smidge. But that smidge is there. It's not a case of no compression, just a case of very little compression.
@@ScottGrammer not enough pressure to make a difference. Because air is flowing there is not actually any pressure. If you seal the case to restrict flow the pressure builds a bit, but again, not enough to matter. If you completely restrict the flow, the fan can't actually move air and it comes back through the gaps in the fan itself.
Even if you could significantly increase the density of the air, it wouldn't positively impact cooling performance. Air flow cools things down. Not pressure.
One thing I have been messing with lately is covering vented portions with no fans. The dust is pretty insane after just 1 month, but it shows me that it's really moving air now.
I plan to buy this, what do you think about this setup? Would you add or change something? My budget is about 1000 euros.
GPU:Gigabyte RTX 3060 OC dual 12GB DDR6
CPU: Ryzen 5 5600 wraith cooler
POWER: TUF Gaming 650W
MOTHER BOARD: Gigabyte b550m-ds3h
CASE: MS Fighter S300
NVMe: NV2 1TB M.2
RAM: Kingston Fury Beast 2 x 16GB
In terms of fan position I like to have two intake fans in front, one exhaust fan above the cpu cooler and one exhaust fan at the back behind the cpu cooler. That way it's neutral, I have the chimney effect and the air going directly through the cpu fans also are directed towards outside lead by the exhaust fan at the back.
I don’t care who’s first.
You obviously cared enough to dedicate a whole comment about it.
The early bird gets the worm, but the second mouse gets the cheese.
Ha ha ha 53rd
How sad must your life be to get excited by having the first comment on a youtube video.
@@enthusiasticpaunch Actually, mice doesn't like cheese at all 🤓
Hi Jay, how many Inxhaust fans should I have to balance my Outtake fans?
2/3 exhausts 2 outtake
Super common problem I see with front fan positive pressure setups is large gaps to the case interior which makes it easier for the front fans to recirc air instead of pulling fresh air. Blocking off those gaps can be a pretty big performance improvement on a lot of cases with restrictive fronts.
I guess this can make having those mesh covers on top of the PC quite helpful
Back in high school, I worked in a bar/restaurant, and the owner knew I liked to futz with electronics, and he was a cheap, cheap man. This was in the 90s, when smoking was pretty much treated as, "turn a blind eye if they're drinkin'," mentality. (At least in my podunk area) So he asked me if he could give me a tube TV to look at, that was hanging over the bar "for years," and he couldn't figure out why it stopped working. He was genuinely flabbergasted that cigarette nastiness-accumulation was the culprit. Your little, "vape kills fans fast," aside reminded me of this, and I got a chuckle.
I’m at the end of a week long binge of reviews for everything I’m thinking of buying for my new build. I am going to have 2 180’s at the bottom pulling cold into the GPU. 2 140’s attached to a liquid AIO cooler at the front of the case. Finally the best 120mm fan on the market for an exhaust on the back. There is no top exhaust, the case is fractal torrent compact. It has a fan controller built to the case so I will play around with that. Great video, really put my mind at ease.
Hey jay! That’s for the info!
Id love to see a video about ddr5 ram and the performance at different speeds like you did a few years ago with ddr4 to know when you are really not getting any benefit of going any higher! I’m looking at doing a system upgrade from 9th gen intel and it’s probably the one area I’m really uncertain about making a decision on!
WAIT WHAT ? vaping is as destructive as smoking normal cigarettes ?? flush me... i just realized that i care more about my pc health than my own ...
Vaping is worse than cigarettes
Lol exactly.. I'm more worried about vaping near my tower than my health..priorities:P
@@SrMoraisare you saying vaping is worse than cigarettes for your own health, or for your PC?
@@anonymoose2474 For your health
@@SrMorais 😂 why because it's a fraction of the cost? Or its a few chemicals compared to 1000s ?
Disposable vapes could be dangerous but real vapours are at 1% risk currently compared smokers and after discussions with Professors and Drs who've looked into it. The biggest risk is inhalation of bacteria if you don't keep your vape clean and regular change of liquid and coils.
I'd be curious to know how this intersects with stuff like AIO radiator placement.
Heat rises , Radiator should always be on top , 120% to 140% in @ 900 rpm, 100% out @ 1100 to 1400 rpm out , note - out take pressures MUST include
power supply , also use magnetic sheet cut out & placed around radiator to isolate positive heat load out , use Shin Etsu
X-23-7868-2D thermal
grease (buy direct from them or will end up with fake trash)
have i9 13900k at 91C @ 100% stress @ 8hrs in , REMEMBER "HEAT RISES", good luck .
Thanks for the tips I asked my buddy because he told me I should build a pc how to improve my cooling he said put my Willy in my case and spin it. So I really appreciate the advice
I like cases with BIG fans, my current one has 2 200mm fans in the front.
I find two big boys is best for noise level !
"Circulate in a circle" - Jay 2024
I run the cte c700 Thermaltake case. Best air flow case I have ever used. Motherboard is sideways, so when a massive gpu is installed, there are no pcie slot issues at all. At a very good price too. My new 14900ks(Intel replacement cpu) runs at 85c on full load using a push/pull 480 aio. There is a spot to put a fan on the backside of the case to help cool the cpu socket. It just works awesome! Just an FYI, it is not quiet. Put your headphones on!!!! Can run up to 14 fans in this thing!
The app fan control that Jay talked about a couple videos ago works fantastic! After watching that video i immediately started using it. Its awsome you can tinker for hours. Set it up so the fans go up onder load etc. great stuf look up the video, wont regret it
Finally a proper instructional video. Have found so many ppl ignoring this. Knew some of the stuff he said, but not all, ie great video
I have the Coolermaster HAF 932 case and imao has the best airflow up to this day, with lots of customization options... I will never replace it!
I've looked into RPM vs CFM curves quite a bit and within spec (I think from 20% to 100% PWM modulation) the plots usually are pretty close to linear. Thus, you can usually get away with calculating max CFM IN and max CFM OUT (sum of the CFMs of your intake/exhaust fans) and set the PWM for the exhaust 5% below parity.
Example:
max IN = 400 cfm
max OUT = 500 cfm
Parity setting (neutral pressure): IN 100%, OUT 80% (400/500 = 0.8)
Positive pressure settings:
Set the intake fan curve as you want it, then set the exhaust curve points to 75% of the corresponding intake curve points. If you wanna make extra certain or use fine dust filters, go 10% below parity. Should work just fine in most cases.
😅.. not gunna lie I stopped watching after I found this ...thanks for the extra 10mins to go do thia
Holy crap I loved the DUST reference to Counter Strike Dust map hah!
Whenever I do a build I use pieces of tape of the appropriate color on the inside of the case to block every opening of any size that I don't want air entering or exiting. I usually replace the front mesh with a 1/4" thick foam open cell air conditioner pre-filter mounted in a "creatively engineered" bracket of some kind.
Ever since I switched to an Azza Cast 808w I haven't had anymore issues with ventilation. Dust collection is minimal. A breeze to work with really (no pun intended).
Boasts about using ft³/m, continues to use Celsius, good one Jay!
Also, static pressure is measured in mmH2O! (e.g. millimeter)
Positive pressure is the best with exhaust out the top. Keeps your case from building excess dust if all your intake fans have filters.
Nice video. Want to point a single (and minor) mistake. The temperature heating up rate is not 1:1. You can consider 2 divisions of air (actually would need to consider infinitesimal divisions then integrate, but that would not fit in one comment :D), one near the hot component, with the temperature from component, one from outside with a lower temperature. Because physics they will go towards the same value (one will get colder, one will heat up), in a particular amount of time:t = (m1 * c1 * ΔT1) / (k * A * ΔT1) = (m2 * c2 * ΔT2) / (k * A * ΔT2) (sorry if I missed anything in formula... been like 20 years since I used it last time). Now if you keep t constant (the time it takes to replace the air is the same), and calculate T2 in report with T1 you will notice they are not proportional at all, much less 1 to 1.
YES! FINALLY FOUND SOMEONE TALKING ABOUT IT! The only one that i saw talking about it was from 5 years ago..... thanks mate!
I have to say that the part about those "candles" it imho makes the core fo this video's topic - alltogether with all that stuff it makes sense, you have noted the most important stuff: the airflow - positive, negative, neutral, then the fact that gaps in a PC case matter, the RPM of fans also being important. I think that this might inspire a lot of people who might have neglected a bit their fan setup to option for your tips because who knows, maybe one has built a PC and and just a small RPM fine tuning might help with those hot spots.
About time i found someone who can explain PC stuff in mechanical terms an auto tech can understand
This was such a “duh” moment for me.
I hadn’t thought to consider whether the pressure was positive or negative. Nor had I considered the consequences of negative pressure (dust and/or circular airflow).
Really helpful video!
To go along with your comment about not vaping in the room with your computer, I would add that you shouldn't use spray adhesives or lacquers in the room where your computer is. My son was an art student and use a sort of spray adhesive to finished charcoal drawings so the charcoal didn't smug. This finishing "adhesive" collected to surfaces, most notably fans, filters, and heat-syncs and in turn, as you would expect, this collected an unbelievable amount of dust. It ended up frying his CPU. I replaced the fans, filters and CPU, took alcohol to the heat-syncs and PCI slots and finally rebuilt the system. What a mess!! I actually may still have pictures of it.
Man that vaping thing is SOOO TRUE, 3 vapers in my house. It makes the dust moist and super sticky and a pain in the ass to clean. I had to take apart AIO and GPU to clean individual rows of aluminum fins from getting gunked up. No more vaping near my rig without a window wide open lol.
I just build my first PC since an AMD 64 3700+ and I have to say it changed a lil bit since then but your video really did help, at least strengthen what I already thought I knew
I really enjoy videos like this. Really explained everything quickly and clearly. That incense trick is smart, I used this to optimise my airflow. Thanks Jay!
Good advice! One thing though - convection is not that weak of a force. It can make a difference. That's why your exhaust fan is rear top. And that's why you should keep it as an exhaust, not intake.