Since they sponsored this video it means that this is now a valid use case so now they need to make the mounts to add 14 fans on the cooler so everyone can get that sweet 1 degree lower temps.
Now, let's redo this project, but with delta fans. Y'know, the ones that come with grates because you would literally lose a finger in the blades due to the extremely high rpm
yep this i have a 120mm one the thing puts out pure power and sounds like a turbo jet it has easily cut my finger before luckily nothing serious just converting it from 5 pin to standard 4 pin.
@@virtualtools_3021 Yeah you would need to run power to the delta fans separately because those will try to pull a bit more power than what a PWM fan header is rated for.
Next try using blower style fans like you see in laptops... You could put quite a few together all stacked side by side instead of in front of each other thus preventing the blowers from blocking airflow from each other
He should indeed have put them in parallel instead of in series. After the first or second fan, the air is already going at the max speed, the other fans would have close to zero load moving air along that is already up to speed. In parallel tho a large area of air is being forced into a smaller area increasing the pressure which would help allot more.
Bet if you could get contra-rotating versions of those fans, running them in a giant stack would give you insane static pressure... Basically how server fan modules work but taken to ridiculous levels.
Not exactly how that works. Air flow is the volume of air that is produced by the fan measured by time. In this case, the air flow of a fan is measured in cubic meters per minute (m³/min) in metric units, or cubic feet per minute (CFM) in imperial units. In simplest terms, if you have a 5 ft x 5 ft x 5 ft enclosure, and a fan that produces 5 CFM, it will likely take 25 minutes for the fan to ventilate the hot air in the enclosure. (In actuality, it's not that easy.) Static pressure is the amount of air pressure that can be produced by the fan in an enclosure. In this case, the static pressure is measured in Pascals (Pa), or inches of water (inH2O). The pascal (Pa) is the SI derived unit of pressure used to quantify internal pressure, stress...etc. The unit is named after Blaise Pascal and is defined as one newton per square meter. Inches of water (inH2O) is defined as the pressure exerted by a column of water of 1 inch in height at defined conditions. At a temperature of 4 °C (39.2 °F) pure water has its highest density (1000 kg/m³). At that temperature and the standard acceleration of gravity, 1 inH2O is approximately 249.082 pascals. It's important to know that even though maximum values for air flow and static pressure are specified, the fan will not output both maximum values at the same time.
@@nightbirdds Really it's just however many fans it takes to go into the next room. He got a 1 degree improvement, because the air he's pulling in is now far enough away from the motherboard to avoid the ambient heat increase.
I remember OC3D doing this with 40 fans on an aio for an April fools video, that was silly and Dawid is even sillier so this will be a match made in heaven :D Edit: Thank you everyone for thumbs up! You're all way too kind
I had never heard of the video you were talking about, and I love the madness of it. The fact all the fans generated enough force to move the contraption forward was crazy
And in that scenario it would only make sense to use a bunch of 200mm fans instead of 120’s lol. Edit: or possibly a ton of delta fans stacked side by side to make the world’s loudest cooling solution
Fan stacking is usually done to compensate restricted air flow, for example to force air though a 1m long 1U server. Considering you've already had the ideal air flow with 2 fans, it wasn't really a surprise that adding more virtually made no difference.
Yo Dawid, can you try and get fan spacers between all the fans? if you space them out about 1-2 inches betweeen each fan, the temps have to be better, the fans were getting choked out and they probably lost air as there was just too little air and too much fan.
"we would like to sponsor a video of you checking out the Silent wing pro 4s" "oh wow, thanks... how many fans can I get for filming the video?" "uh, says here 14 is the maximum" "can I get 14 please" "sure, why?" "no... no reason"
I would think staggering the fan speeds from slow to fast, towards the middle, would prevent the air from “stalling” as it moved through fans of the same RPM
As Dawid discovered, this isn't going to increase airflow that much regardless of tuning. When you see fans can flow a certain CFM that is what they flow. The only advantage here is a slight increase in air pressure, but there are limits to how much a fan can increase static pressure.
The limiting thing is that all the fans spin in the same direction. After like the second fan, almost all the additional "input" will just go into spiralling the airflow more and more. That's why server fan assemblies use counter rotating fans and stator vanes.
This was super weird, entirely pointless and mindblowingly stupid. But absolutely fun and hilarious! Thanks for 8 minutes of this hilarity! And also : Bequiet! Rocks, been using their 750w PSU and 4 case fans in my rig for almost a decade now and they're still running just fine and just as quiet as they always have been. That speaks for quality, and I'm a big fan of quality, when I upgrade my rig next year I'm most definively buying Bequiet! products again.
Maybe next video you can try to flip the switch at the back to UHS to allow for maximum fan speed and see if it makes any difference. I think you had yours on medium or the lowest as 7:25 time frame. So, probably the rest of them are the same.
I want a part 2 where you do something similar with an aio. I remember seeing a more significant difference in altering fan speeds on an aio compared to a tower cooler. So adding more fans might make a bigger difference then tower would, since the rad fins are smaller and thus offer more resistance. Okay, so the difference between 2 and 10 fans wasn't that good. But still people don't really try putting 2 fans on a rad anyways.
In my experience push pull can work well. I was using 2 x stock EVGA 120mm fans on the 2080Ti Kingpin 240mm GPU AIO Radiator, with the card producing a ~280W load in a 19c room, I then upgraded to 4 x Noctua NF12 120mm fans placed them in Push Pull configuration so 2 on each side and when fans run flat out the cooling system reduced GPU temperature by about 10c compared to the stock 2 EVA fans.
@@micb3rd I was thinking it might work well, but hearing this makes me think it's actually correct. Push pull is what will enable 13900k overclocking 😜
@@Olav_Hansen Haha Yes, I'm looing forward to testing the 13900K on Noctua NH-D15 with Push Pull and then will try a 360mm AIO Push Pull. I'm also going try some undervolting. My current 10900K at 5.1 Ghz is fine on the NH-D15, it only get a bit hot when running 5.3Ghz and pushing heavy loads. (I managed 200Watts draw in CPU @ 85C on Cyberpunk 2077 with RTX Reflections on!) which is about the max temp I'm comfortable with long term.
I would say, use the fans to prop up the GPU in an ATX case with bottom intake to prevent it from sagging since these days GPU uses 4 slots and so freaking heavy . The fans may serve a better use that way.
Kind of a law of diminishing returns in play. The speed of the air being moved will be consistent, once you achieve the amount of fans needed to overcome the air resistance of the radiator. Still was fun to watch.
Man I love when my Silent Wing Pro 4(tm) fan start spinning and cooling my CPU, these Silent Wing Pro 4(tm) fans are running quiet(tm) and the Silent Wing Pro(tm) fans runs at low RPM, it really feels like these Silent Wing Pro 4(tm) are great.
As the experiment showed, stacking fans in series doesn't increase the airflow. One fan would have to push the next fan to spin faster in order to get more airflow thru it. If you built a frame to mount the fans side by side, and a plenum to direct the air into the cross section of the cooler, you would be able to move more airflow thru it.
the fans arent the limitation. its how much energy that can be conveyed to the heatsink fins themselves. you can add all the fans you want the temp wont change unless your able to get more transfer from the heatpipes and coldplate up to the fins themselves.
The biggest benefit I got from adding a second fan on my Be Quiet! Shadow Rock 3 CPU cooler was that under normal load you could shave off 2-300RPMs in fan speed and have the same performance or temps. The deeper the fins is the better the result. The one here is almost half as deep as the Shadow Rock3. FYI.
I have a Dell XPS 8940 with a 10th gen i7, 32gb ram and an EVGA 1660 Super. I added a the large Noctua heat sink to it and run it open case with an 8" desk fan blowing right into the front of it. It has a stock intake fan, a fan in the heat sink and a stock exhaust fan. It runs fairly cool when I am gaming on it. Can def feel the heat blowing out the back.
I am actually surprised that the temperature wasn't increasing again when you attached the third and fourth fan. Any difference in speed, because they will not all be running at exactly the same speed, will cause back pressure and a drop in airflow.
I was thinking using doweling in the holes would be great, slip a rod through all the fans and 'bracket' it to the cooler. It would also line up the fans well.
It would be a lot more efficient if every second fan would rotate in other direction or would have a spacer stoping the airflow rotating in the same direction as the fans.
that just proves that the CPU heatsink itself from Bequet is awesome if the thermal transferer to air is already at its best with just 2 fans, if it had gotten a lot better that would have been bad
Diminishing Returns do exist everywhere but that hasnt stopped a wild Dawid yet :) Great experiment. I also never liked how the CPU cooler has such a long way from the rear exhaust fan/ hole in the box, just make it one " fan- tube" :)
while stacking fans you should use counter rotating fans. One have to spin clockwise and the second should be counter clockwise. When using counter rotating fans you you increase static pressure which helps to cool better
Did you lock the CPU's clock speeds for testing? The 3700X Ryzen's are notorious for using ~70-74°c for TJ Max temperature and will just keep adjusting boost clocks to maintain that target. When I switched a better cooling solution on a 3700X, max temp was the same, but Cinebench score jumped ~400 points because the thermal advantage let it boost longer. That said, I don't think there will be any significant temperature difference or advantage beyond the two fan push/pull configuration. Stacking fans doesn't increase CFM, only static pressure (think rows of blades in turbine engines 'squishing' the air), which isn't really going to matter on that cooler.
I guess the moral of the story is one push and one pull is the most efficient setup for any configuration that uses more than one fan mounted onto the heatsink.
you need stator vanes in between fans, or they'll just spin the air around after the first couple, instead of accelerating it further. Providing that works, you could then produce a significantly higher pressure.
Very possible you reached the saturation point as to how much the cooler could actually move between the die and the heatsink fins. That'd be my guess as to why there wasn't any major gains after adding the first set.
Correct, what'd be really cool is if be Quiet made Dawid a one-off 6 tower air-cooler with 18 heat pipes and mounting space for 7 fans, it'd be insane, very heavy but omg how much TDP could it remove.
The cooling is affected by the Speed of the air going through the fins (and fin size, density, etc. The variables that are staying the same here.) Adding the extra fan to Push-Pull means maximizing that speed, pushing past the resistance. Just adding more fans at the same speed will do nothing. Needs FASTER air. So, say four fans spread out and a funnel of that air (maybe a cardboard contraption) would speed the air up.
- Have you ever heard of the BeQuiet Silent Wing Pro 4 fan? + No, I've never heard of the BeQuiet Silent Wing Pro 4, what's special about them? - Exactly, you don't hear the BeQuiet Silent Wing Pro 4's + Oh, that sounds like the BeQuiet Silent Wing Pro 4's are amazing - Additionally the BeQuiet Silent Wing Pro 4 don't drop performance to the competition + Wow! The BeQuiet Silent Wing Pro 4's are awesome :0 😂😂😂 LOVED IT!
Use fans only for pushing air not pulling, I think that will be more effective. Because these fans are designed to push air more effectively than pull air. You can see on the GPU the fan is like turbo fans that is designed to suck more air.
You would probably need some kind of air duct where you could put all those fans side by side in order to increase total airflow. I don't think that putting them in series increases airflow much.
I have an air cooler that supports 4 120mm fans, there are 3 seperated fin stacks with 4 heat pipes in each stack. But this is wow, lol. Under full load it doesn't get higher than 60c. But I have 7 120mm case fans front 3 intake, top 3 exhaust, 1 rear exhaust, then 4 on the cpu, and 2 on the gpu.
It a shame you couldn't use Velada gloves to bind them all together rather than duct tape, to break up the product placement. I have he Lian Li Mesh 2 case. I had originally ordered a bunch of BQ fans for my PC, but found them to be anything but quiet. I quickly replaced those with Artic fans, and haven't looked back. It an interesting experiment, I wonder if you'd have seen more change if you had varied the fan speed so that the fans run faster as you get deeper into the wind tunnel, to overcome the added obstruction you introduced.
These fans have a speed switch on the back. You should've tested them at their UHS setting to really see what they can do. Running a 3000 rpm fan at 33% and hoping for better performance is just weird
If you Mount fans like one normalny other backwards and spining in the other direction will gratelly improve air Flow. It has to do with fluid dynamics and stuff i tested something like that few years back check it out.
Another round for this video would be that you make 2 fan shrouds that put those fans as parallel to get more surface area and more airflow. With that you try to get same airflow that those industrial fans do but more quietly.
I think for that he only has to increase the fan speed of the ones already installed (the Pro version of the Silent Wings 4 goes up to 3000rpm, for whatever reason)
It's just a 4 heatpipe cooler, and thermally it's probably set up for close to max heat dissipation from factory with the single fan. Would try this again using a DARK ROCK PRO 4 instead.
It looks awesome! 😂 I think a common misunderstanding thing with air coolers is that adding loads of fans or dramatically increasing heatsink size/surface area only really increases the capacity of the heatsink to cool larger heat loads, rather than cause outright temp differences. The actual temperature they can cool down to (given a sensible heat load) will remain somewhat the same due to ambient temps/temp coming off the CPU. This is why we get 'my Arctic Duo 34 cools just as well as my mates NH D-15 with the same processor' where that processor is actually an I5 12400. Because the heat load is so low and the temp difference between the two will be relatively small it's easy to draw that conclusion. If you were to strap that stock be quiet cooler assembly to a heavily overclocked 12900K for example it would go over thermal capacity, heat soak and throttle damn fast. I used to use a dark rock 4 on my 8700k (12400 now, the easy life!) When it was sat at a comfortable 5.0ghz overclock. Before that, running the aforementioned freezer 34 it would commonly sit around or over 90 degrees with the overclocked 120-130 watt heat load. The DR4 brought the overclocked temps down to 70-75ish However with stock clocks both coolers were not dissimilar on temperatures. The real test would be to see if that high wattage heat soak with the DK2 cooler can be delayed/prevented with the eleventy million fans!
Anything over two fans is a deminishing return because in order to utilize all the static pressure the fins need to be more compactly packed like in 1U FF systems.
you would see a bigger difference i temp if you set the fans at a fixed voltage. if the fans are set te 2000 rpm the help from another fan would only let the first fan get to 2000 rpm easier but not spin fater so you would not get more airflow.
I figured out why your temps went down. There is recirculation normally of hot air from near the cpu. When there are 14 fans, the entry air will be very far away and so will have fresh air. It is like having a long fan shroud from outside the case, so that no hot case air can be reused. Similar concept. So it is not the number of fans, just the fact that you built a long tube.
Silly yet an important test to do and to know about. I would have duct taped the cooler itself on the sides to force the air through it and not on the sides also to see if something changes.
I wonder if the fan speeds were staggered and modified, if it would have any effect? Similar to a gearing system but to ramp up the air faster on the final output
Hmmm... how effective is optimally jetstreaming inside a PC case through passive coolers (CPU & GPU) compared to having the fan direct yelling at aluminium fins?
the only time i can say more fans is more better with a radiator double or quad stacking fans can make a HUGE dif depending on the setup... worked in a shop that made custom servers and workstations for a while and, sometimes we used super thick radiators and 2 high static pressure fans on each side in push/pull, a few builds had up to 4 fans on each side due to just how much resistance the radiator created.. the one i helped build like that was for a very specific use case and well, it was running dual chips at 4.7ghz 24/7 that produced enough heat that... under extended load it would still hit around 87c, and the loads it would be put under could last days at a time, renders and simulations... it was a beast of a box... the videocards also were on their own loop but just a normal thin copper rad used for that, with push/pull... nothing fancy needed for the gpu's... back then they didnt run nearly as hot as modern cards.
Since they sponsored this video it means that this is now a valid use case so now they need to make the mounts to add 14 fans on the cooler so everyone can get that sweet 1 degree lower temps.
I agree.
geez, spoiler
All they gotta do is swipe Noctua's rubber pull fastener idea, and sell them separately, *At participating, authorized dealers near you!* LMAO.
I think they should legitimately just supply a kit with a few feet of duct tape in it and call it a day, do it for the memes xD
imagine doing this with like a 12900K
Now, let's redo this project, but with delta fans. Y'know, the ones that come with grates because you would literally lose a finger in the blades due to the extremely high rpm
yep this i have a 120mm one the thing puts out pure power and sounds like a turbo jet it has easily cut my finger before luckily nothing serious just converting it from 5 pin to standard 4 pin.
@@ninjanolen4932 and frying your mobo
Yup.. The 120mm 4A one will really make your pc fly...
@@virtualtools_3021 Yeah you would need to run power to the delta fans separately because those will try to pull a bit more power than what a PWM fan header is rated for.
@@dragonfyre79 yep I got 5a deltas on my server but the headers only go to 2.5 so I got them powered from molex
Next try using blower style fans like you see in laptops... You could put quite a few together all stacked side by side instead of in front of each other thus preventing the blowers from blocking airflow from each other
Compound blowers! Tsttutut
Nah adapt a leaf blower, now theres one no one else will have tried.
He should indeed have put them in parallel instead of in series. After the first or second fan, the air is already going at the max speed, the other fans would have close to zero load moving air along that is already up to speed. In parallel tho a large area of air is being forced into a smaller area increasing the pressure which would help allot more.
@@andrewwatts1997 True, it'd be cool to see a 3d printed manifold for that
Haha!! That’s a great idea. Although the leaf blower idea is even better.
Bet if you could get contra-rotating versions of those fans, running them in a giant stack would give you insane static pressure... Basically how server fan modules work but taken to ridiculous levels.
If we are talking server fans he would need something that spins at 3k+ RPM.
also some air flow fans further in the stack could be better than static press. units
@@UhOhUmm so like noctua industrial fans
@@nathanlarson6535 i wonder if any industry actually uses those. I was thinking about Sanyo or Sunon.
Not exactly how that works. Air flow is the volume of air that is produced by the fan measured by time. In this case, the air flow of a fan is measured in cubic meters per minute (m³/min) in metric units, or cubic feet per minute (CFM) in imperial units. In simplest terms, if you have a 5 ft x 5 ft x 5 ft enclosure, and a fan that produces 5 CFM, it will likely take 25 minutes for the fan to ventilate the hot air in the enclosure. (In actuality, it's not that easy.)
Static pressure is the amount of air pressure that can be produced by the fan in an enclosure. In this case, the static pressure is measured in Pascals (Pa), or inches of water (inH2O). The pascal (Pa) is the SI derived unit of pressure used to quantify internal pressure, stress...etc. The unit is named after Blaise Pascal and is defined as one newton per square meter. Inches of water (inH2O) is defined as the pressure exerted by a column of water of 1 inch in height at defined conditions. At a temperature of 4 °C (39.2 °F) pure water has its highest density (1000 kg/m³). At that temperature and the standard acceleration of gravity, 1 inH2O is approximately 249.082 pascals.
It's important to know that even though maximum values for air flow and static pressure are specified, the fan will not output both maximum values at the same time.
Be Quiet: Lets send him 14 fans for a neat little giveaway
Dawid: It's only enough or one setup though.
This is the only TH-cam Content Creator I visit regularly to see just what crazy shit he may be up to. Glad to see he still doesn't disappoint.
I think you needed 16 fans for any real cooling gains
If it took 14 fans for a 1 degree improvement, then it stands to reason that to get a 5 degree improvement you would need 70 of them.
@@nightbirdds Really it's just however many fans it takes to go into the next room. He got a 1 degree improvement, because the air he's pulling in is now far enough away from the motherboard to avoid the ambient heat increase.
@@setcheck67 Oh, I know. But the idea of him trying to make 70 fans work just makes me chuckle. :)
Or maybe put those fans side by side in some sort of air duct to increase the total air flow.
@@Max_Mustermann That would turn into a gigantic monstrosity of a construction and Dawid totally needs to do it. :D
I remember OC3D doing this with 40 fans on an aio for an April fools video, that was silly and Dawid is even sillier so this will be a match made in heaven :D
Edit:
Thank you everyone for thumbs up! You're all way too kind
Oh my, i remember as well, thanks for the memories!
@@liarus You're welcome :)
I knew it was a long time ago, but holy cow it's over 10 years old that video
was the first thing that came to mind when the thumn nail
I had never heard of the video you were talking about, and I love the madness of it. The fact all the fans generated enough force to move the contraption forward was crazy
@@charredolive should look it up, its ten years old on the 29th of sep, i looked it up on youtube....cant believe its been that long mind
I am thoroughly convinced if you had just added two more fans we would have seen a significant improvement. 😄
this is what happens when you give Dawid to many fans🤣
if my math checks out you only need 816 more fans to achieve sub zero temps
You should build a cardboard adapter that uses all the fans surface area and funnels it down to the cooler. And then see what that does.
Think there was an older LTT that tried something to that degree
Sounds like a Jays2Cents video.
And in that scenario it would only make sense to use a bunch of 200mm fans instead of 120’s lol.
Edit: or possibly a ton of delta fans stacked side by side to make the world’s loudest cooling solution
Yes, that would probably have made more sense.
Fan stacking is usually done to compensate restricted air flow, for example to force air though a 1m long 1U server. Considering you've already had the ideal air flow with 2 fans, it wasn't really a surprise that adding more virtually made no difference.
Because he needed more airflow, but he increased pressure instead of airflow
Yo Dawid, can you try and get fan spacers between all the fans? if you space them out about 1-2 inches betweeen each fan, the temps have to be better, the fans were getting choked out and they probably lost air as there was just too little air and too much fan.
"we would like to sponsor a video of you checking out the Silent wing pro 4s"
"oh wow, thanks... how many fans can I get for filming the video?"
"uh, says here 14 is the maximum"
"can I get 14 please"
"sure, why?"
"no... no reason"
I would think staggering the fan speeds from slow to fast, towards the middle, would prevent the air from “stalling” as it moved through fans of the same RPM
As Dawid discovered, this isn't going to increase airflow that much regardless of tuning. When you see fans can flow a certain CFM that is what they flow. The only advantage here is a slight increase in air pressure, but there are limits to how much a fan can increase static pressure.
Doesn't really matter the bottleneck is the cpu heatsinks capacity to transfer heat
The limiting thing is that all the fans spin in the same direction. After like the second fan, almost all the additional "input" will just go into spiralling the airflow more and more. That's why server fan assemblies use counter rotating fans and stator vanes.
This was super weird, entirely pointless and mindblowingly stupid. But absolutely fun and hilarious! Thanks for 8 minutes of this hilarity! And also : Bequiet! Rocks, been using their 750w PSU and 4 case fans in my rig for almost a decade now and they're still running just fine and just as quiet as they always have been. That speaks for quality, and I'm a big fan of quality, when I upgrade my rig next year I'm most definively buying Bequiet! products again.
This is kind of "Just because you can, doesnt mean you should" type of thing
Would be fun to see if it affects coolers with even bigger heatsinks
So I took a shot of bourbon every time he said "Silent Wing Pro 4" and I'm currently writing this from the hospital. Also I'm out of Bourbon.
Dawid should be force-fed single channel ram in the next video.
What's he gona complain about when ddr5 is the norm
Ok I’m now convinced that Dawid’s entire life is sponsored by BeQuiet and LINODEEEE.
Also this is the amount of times he said “Silent Wing Pro 4”
⬇️
Maybe next video you can try to flip the switch at the back to UHS to allow for maximum fan speed and see if it makes any difference. I think you had yours on medium or the lowest as 7:25 time frame. So, probably the rest of them are the same.
This is a great example of how airflow only take you so far and at the end of the day you really need more thermal mass.
I want a part 2 where you do something similar with an aio. I remember seeing a more significant difference in altering fan speeds on an aio compared to a tower cooler. So adding more fans might make a bigger difference then tower would, since the rad fins are smaller and thus offer more resistance.
Okay, so the difference between 2 and 10 fans wasn't that good. But still people don't really try putting 2 fans on a rad anyways.
In my experience push pull can work well. I was using 2 x stock EVGA 120mm fans on the 2080Ti Kingpin 240mm GPU AIO Radiator, with the card producing a ~280W load in a 19c room, I then upgraded to 4 x Noctua NF12 120mm fans placed them in Push Pull configuration so 2 on each side and when fans run flat out the cooling system reduced GPU temperature by about 10c compared to the stock 2 EVA fans.
@@micb3rd I was thinking it might work well, but hearing this makes me think it's actually correct.
Push pull is what will enable 13900k overclocking 😜
@@Olav_Hansen Haha Yes, I'm looing forward to testing the 13900K on Noctua NH-D15 with Push Pull and then will try a 360mm AIO Push Pull. I'm also going try some undervolting.
My current 10900K at 5.1 Ghz is fine on the NH-D15, it only get a bit hot when running 5.3Ghz and pushing heavy loads. (I managed 200Watts draw in CPU @ 85C on Cyberpunk 2077 with RTX Reflections on!) which is about the max temp I'm comfortable with long term.
I would say, use the fans to prop up the GPU in an ATX case with bottom intake to prevent it from sagging since these days GPU uses 4 slots and so freaking heavy . The fans may serve a better use that way.
@@micb3rd cool stuff
Kind of a law of diminishing returns in play. The speed of the air being moved will be consistent, once you achieve the amount of fans needed to overcome the air resistance of the radiator. Still was fun to watch.
this is the content i am subbed for
can't believe you recruited and featured all of your fans for this video. stand up guy, dawid.
You should try the fans in parallel as well as series. Cardboard ducting maybe
Didn't know I was gonna watch a Big Black Cooler video this morning. And yet, it was just what I needed. Great job Dawid.
Now you need to do the same test with fans that make noise :) would love to see if fans that make more noise cool better or worse?
He could have switched them to gear 3 on the back for VERY noisy 3000 rpm.
@@Waldherz or use deltas that are 38mm thicc and 7000 rpm to actually cool it
Man I love when my Silent Wing Pro 4(tm) fan start spinning and cooling my CPU, these Silent Wing Pro 4(tm) fans are running quiet(tm) and the Silent Wing Pro(tm) fans runs at low RPM, it really feels like these Silent Wing Pro 4(tm) are great.
You should do this again, but with a Dark Rock Pro 4, so you have two heatsink towers to cool. Just for science, you know.
As the experiment showed, stacking fans in series doesn't increase the airflow. One fan would have to push the next fan to spin faster in order to get more airflow thru it. If you built a frame to mount the fans side by side, and a plenum to direct the air into the cross section of the cooler, you would be able to move more airflow thru it.
Ahhh I've always wondered this exact question. Excited to see this. Those are sexy fans by Be Quiet, I wish they'd make some RGB fans.
they did
are you living under a rock?
Lul they have RGB fans, and have for ages. I mean they are LITERALLY on the cooler in the video at the beginning.
🤔look at the picture on the Cooler Box
A spacer between the fan and the heatsink on reach side would help allowing airflow to reach where the motor of the fans block.
YES!!! IM FINALLY FIRST!!!!!!
Ok then
*second
ok and ?
the fans arent the limitation. its how much energy that can be conveyed to the heatsink fins themselves. you can add all the fans you want the temp wont change unless your able to get more transfer from the heatpipes and coldplate up to the fins themselves.
The biggest benefit I got from adding a second fan on my Be Quiet! Shadow Rock 3 CPU cooler was that under normal load you could shave off 2-300RPMs in fan speed and have the same performance or temps. The deeper the fins is the better the result. The one here is almost half as deep as the Shadow Rock3. FYI.
I have a Dell XPS 8940 with a 10th gen i7, 32gb ram and an EVGA 1660 Super. I added a the large Noctua heat sink to it and run it open case with an 8" desk fan blowing right into the front of it. It has a stock intake fan, a fan in the heat sink and a stock exhaust fan. It runs fairly cool when I am gaming on it. Can def feel the heat blowing out the back.
You can see the workmanship that Dawid has done, by the amount of cuts and scrapes on his hands and fingers
I am actually surprised that the temperature wasn't increasing again when you attached the third and fourth fan. Any difference in speed, because they will not all be running at exactly the same speed, will cause back pressure and a drop in airflow.
I was thinking using doweling in the holes would be great, slip a rod through all the fans and 'bracket' it to the cooler. It would also line up the fans well.
rhe be-quite tie fighter Darth Vader would be proud Dawid... the force is strong with this one
You have to respect Dawid for spending so much time with his fans!
Hahahaha too true
What if 9 fans were used on an enclosed full size case? 3 Pulling from the front 3 blowing air out the back, and 3 blowing up at the top?
I literally skipped to the conclusion at 8:18 and I have to say, this video is splendid.
Plot twist dawid and his 12 fans caused the hurricane in florida
It would be a lot more efficient if every second fan would rotate in other direction or would have a spacer stoping the airflow rotating in the same direction as the fans.
that just proves that the CPU heatsink itself from Bequet is awesome if the thermal transferer to air is already at its best with just 2 fans, if it had gotten a lot better that would have been bad
Based on vortex and eddy in fluid mechanics, it is impossible to synchronize each fans' air flow while minimizing interfering and blocking them
Diminishing Returns do exist everywhere but that hasnt stopped a wild Dawid yet :) Great experiment. I also never liked how the CPU cooler has such a long way from the rear exhaust fan/ hole in the box, just make it one " fan- tube" :)
while stacking fans you should use counter rotating fans. One have to spin clockwise and the second should be counter clockwise. When using counter rotating fans you you increase static pressure which helps to cool better
Gotta love when Dawid does something crazy like this just for science.
Did you lock the CPU's clock speeds for testing? The 3700X Ryzen's are notorious for using ~70-74°c for TJ Max temperature and will just keep adjusting boost clocks to maintain that target. When I switched a better cooling solution on a 3700X, max temp was the same, but Cinebench score jumped ~400 points because the thermal advantage let it boost longer. That said, I don't think there will be any significant temperature difference or advantage beyond the two fan push/pull configuration. Stacking fans doesn't increase CFM, only static pressure (think rows of blades in turbine engines 'squishing' the air), which isn't really going to matter on that cooler.
I guess the moral of the story is one push and one pull is the most efficient setup for any configuration that uses more than one fan mounted onto the heatsink.
I think the extra fans will come in good use when you actually push the PC performance during heavy workloads. Maybe..
You got a Fan in Me Now.
Now ask all the other Fan Manufacturers to sponsor you.
you need stator vanes in between fans, or they'll just spin the air around after the first couple, instead of accelerating it further. Providing that works, you could then produce a significantly higher pressure.
Very possible you reached the saturation point as to how much the cooler could actually move between the die and the heatsink fins. That'd be my guess as to why there wasn't any major gains after adding the first set.
Correct, what'd be really cool is if be Quiet made Dawid a one-off 6 tower air-cooler with 18 heat pipes and mounting space for 7 fans, it'd be insane, very heavy but omg how much TDP could it remove.
The cooling is affected by the Speed of the air going through the fins (and fin size, density, etc. The variables that are staying the same here.)
Adding the extra fan to Push-Pull means maximizing that speed, pushing past the resistance.
Just adding more fans at the same speed will do nothing. Needs FASTER air.
So, say four fans spread out and a funnel of that air (maybe a cardboard contraption) would speed the air up.
- Have you ever heard of the BeQuiet Silent Wing Pro 4 fan?
+ No, I've never heard of the BeQuiet Silent Wing Pro 4, what's special about them?
- Exactly, you don't hear the BeQuiet Silent Wing Pro 4's
+ Oh, that sounds like the BeQuiet Silent Wing Pro 4's are amazing
- Additionally the BeQuiet Silent Wing Pro 4 don't drop performance to the competition
+ Wow! The BeQuiet Silent Wing Pro 4's are awesome :0
😂😂😂 LOVED IT!
"Dawid's videos have been pretty normal lately, I wonder if-"
* looks at thumbnail *
"The king is back!"
we need more videos like this thanks made my day
Ive been here since a little ways into the channel. I took a break, but this vid got me on a dawid binge-o-thon
Use fans only for pushing air not pulling, I think that will be more effective. Because these fans are designed to push air more effectively than pull air. You can see on the GPU the fan is like turbo fans that is designed to suck more air.
Sponsor company: We'd like to sponsor one of your videos, don't do anything too crazy though
Dawid: Too crazy you say, I got you.
You would probably need some kind of air duct where you could put all those fans side by side in order to increase total airflow. I don't think that putting them in series increases airflow much.
I have an air cooler that supports 4 120mm fans, there are 3 seperated fin stacks with 4 heat pipes in each stack. But this is wow, lol.
Under full load it doesn't get higher than 60c. But I have 7 120mm case fans front 3 intake, top 3 exhaust, 1 rear exhaust, then 4 on the cpu, and 2 on the gpu.
It a shame you couldn't use Velada gloves to bind them all together rather than duct tape, to break up the product placement.
I have he Lian Li Mesh 2 case. I had originally ordered a bunch of BQ fans for my PC, but found them to be anything but quiet. I quickly replaced those with Artic fans, and haven't looked back.
It an interesting experiment, I wonder if you'd have seen more change if you had varied the fan speed so that the fans run faster as you get deeper into the wind tunnel, to overcome the added obstruction you introduced.
Take a shot every time Dawid says "Bequiet Silent Wing Pro 4s"
These fans have a speed switch on the back. You should've tested them at their UHS setting to really see what they can do. Running a 3000 rpm fan at 33% and hoping for better performance is just weird
If you Mount fans like one normalny other backwards and spining in the other direction will gratelly improve air Flow. It has to do with fluid dynamics and stuff i tested something like that few years back check it out.
Another round for this video would be that you make 2 fan shrouds that put those fans as parallel to get more surface area and more airflow.
With that you try to get same airflow that those industrial fans do but more quietly.
Now for a follow up video, find 2 fans with the highest fan speeds to see if a faster fan will lower temps.
I think for that he only has to increase the fan speed of the ones already installed (the Pro version of the Silent Wings 4 goes up to 3000rpm, for whatever reason)
I would describe this as “a weapon to surpass metal gear”
Someone did a video a while back and found how one (push) fan is all you need, nice to see this confirmed.
Bequiet: We're gonna send you a bunch of fans.
Dawid: Say no more...
People need to consider the added heat from each fan motor on the intake side, removing all but 1 on the intake side would produce best results.
Love all the engineers in the comments giving ideas -- very cool
Shoulda put some on the naked (wrong) sides of the cooler too, to fully sandwich it
I really like be quiet products. It's pretty cool they were happy to support a video like this.
It's just a 4 heatpipe cooler, and thermally it's probably set up for close to max heat dissipation from factory with the single fan. Would try this again using a DARK ROCK PRO 4 instead.
It looks awesome! 😂 I think a common misunderstanding thing with air coolers is that adding loads of fans or dramatically increasing heatsink size/surface area only really increases the capacity of the heatsink to cool larger heat loads, rather than cause outright temp differences. The actual temperature they can cool down to (given a sensible heat load) will remain somewhat the same due to ambient temps/temp coming off the CPU.
This is why we get 'my Arctic Duo 34 cools just as well as my mates NH D-15 with the same processor' where that processor is actually an I5 12400. Because the heat load is so low and the temp difference between the two will be relatively small it's easy to draw that conclusion. If you were to strap that stock be quiet cooler assembly to a heavily overclocked 12900K for example it would go over thermal capacity, heat soak and throttle damn fast. I used to use a dark rock 4 on my 8700k (12400 now, the easy life!) When it was sat at a comfortable 5.0ghz overclock. Before that, running the aforementioned freezer 34 it would commonly sit around or over 90 degrees with the overclocked 120-130 watt heat load. The DR4 brought the overclocked temps down to 70-75ish However with stock clocks both coolers were not dissimilar on temperatures.
The real test would be to see if that high wattage heat soak with the DK2 cooler can be delayed/prevented with the eleventy million fans!
All you need is 2 140mm intake Noctua A14 fans and I exhaust at the rear and one exhaust at the top. Noctua fans are hands down the best
Anything over two fans is a deminishing return because in order to utilize all the static pressure the fins need to be more compactly packed like in 1U FF systems.
This is why I'm a Dawid does tech stuff enjoyer, things I could only execute in my brain, Dawid brings them to reality
you would see a bigger difference i temp if you set the fans at a fixed voltage. if the fans are set te 2000 rpm the help from another fan would only let the first fan get to 2000 rpm easier but not spin fater so you would not get more airflow.
I figured out why your temps went down.
There is recirculation normally of hot air from near the cpu.
When there are 14 fans, the entry air will be very far away and so will have fresh air.
It is like having a long fan shroud from outside the case, so that no hot case air can be reused.
Similar concept.
So it is not the number of fans, just the fact that you built a long tube.
The fans just create extended turbulance. In order to really test this, you need alternating fan blades. That will smooth out the air.
Silly yet an important test to do and to know about. I would have duct taped the cooler itself on the sides to force the air through it and not on the sides also to see if something changes.
Dawid is the godly entity of creativity. Nothing else can explain this weird fun random stuff.
I wonder if the fan speeds were staggered and modified, if it would have any effect? Similar to a gearing system but to ramp up the air faster on the final output
Well that's way better than the "it's going to make it worse" that I thought might happen.
What a silly man you are. We love it!
Hmmm... how effective is optimally jetstreaming inside a PC case through passive coolers (CPU & GPU) compared to having the fan direct yelling at aluminium fins?
the only time i can say more fans is more better
with a radiator double or quad stacking fans can make a HUGE dif depending on the setup...
worked in a shop that made custom servers and workstations for a while and, sometimes we used super thick radiators and 2 high static pressure fans on each side in push/pull, a few builds had up to 4 fans on each side due to just how much resistance the radiator created.. the one i helped build like that was for a very specific use case and well, it was running dual chips at 4.7ghz 24/7 that produced enough heat that... under extended load it would still hit around 87c, and the loads it would be put under could last days at a time, renders and simulations... it was a beast of a box... the videocards also were on their own loop but just a normal thin copper rad used for that, with push/pull... nothing fancy needed for the gpu's... back then they didnt run nearly as hot as modern cards.
I especially loved that you did this on one of the lower TDP CPUs
I am simple man. I see Dawid. I watch Dawid. I subscribe, like, and hit the bell ... for Dawid. I am simple man.
Dawid actually showing us his thermal paste application POG