this is what im thinking aswell. comparing Noise, airflow, and pressure, on an rpm curve would give a more complete picture. Im never going to run a fan at anywhere near max rpm, so the performance there is useless to me. What is the performance at 500, 750 and 1000 rpm? is the "pressure" fans better at those rpms? or are they doing the same cooling, but quieter?.
@@kasslerify Performance at 1,000rpm and below generally levels out regardless of the type/design of the fan. Furthermore, the level of pressure in the range of speeds
@@FanPhys i am responding to your false claim of "Performance at 1,000rpm and below generally levels out regardless of the type/design of the fan. Furthermore, the level of pressure in the range of speeds
Why did you choose to normalize RPM? To me it makes way more sense to normalize either wattage or (more likely what people choose speed based on) noise level. It could be that I havn't thought this through enough yet, but don't understand why you went with the method you did.
Since my tests are purely performance tests and not noise tests or power tests, it makes sense to level the playing field by setting each fan to the same speed. It's also worth noting that fan motors have varying levels of efficiency, so one fan might draw three times the power of another fan for the same speed and performance. I will consider this for a future video however it's a big subject that would require a lot of time to adequately cover!
@@FanPhys Yeah, that makes sense, but I also feel like it's a bit weird since these fans with very different motor characteristics also are intended to run at different speeds, sometimes with completely different both upper and lower speed bounds. It could just be that the static pressure fans are intended for pressure at lower speeds compared to other fans, and that would be a nuance missed by this testing.
Makes more sense to test them at 12V. Because that's the maximum you'd have in your PC.... I mean they aren't hitting the RPM value at 12V because either the motor is crap OR the blade geometry is providing more resistance. And if they are spinning slower they probably can't actually perform as well...
@@insanemalThey need to be at the same RPM for consistency, IF the RPM was different, its not the design of the blade that would be the only factor giving diffrent results
They need to be at the same RPM for consistency, IF the RPM was different, its not the design of the blade that would be the only factor giving diffrent results
Someone please explain to me how the static pressure optimized Noctua fan building up more pressure and performing better in the thermal test than the airflow optimized Noctua is proving that static pressure fans don't exist?
You are cherry-picking two fans whose performance conforms to your own bias. The Cougar, eLoop and Gentle Typhoon all beat the NF-F12 in the thermal test despite their inferior pressure performance. There's no such thing as a "static pressure fan" because static pressure is a property of the fluid, not the fan. It's a nonsensical phrase conjured by marketing execs.
Static pressure is created by the fan. All fans are static pressure fans just to different degrees. The fluid doesn't move on its own. Fan design plays a part in amount of pressure created in a given space. The a12 is marketed as a do it all fan, not specifically a static pressure fan, ~21Pa at 2000rpm. F12, 23Pa at 1500rpm. The sp120 is marketed as a static pressure fan but it's only rated at ~13Pa at ~1400rpm. The p12 is rated at ~20Pa at 1800rpm. So your results are pretty much expected all those fans based on their claimed specs. Corsairs marketing was bad, but it usually is. Your test was flawed anyhow.
I was under the impression that the curved blade design was a life cycle consideration- that it allowed the motor to work against a higher load for longer. Could that be a part of it?
It's a good shout, but I quickly abandoned measuring power draw with my DMM because each fan motor is quite different. The NF-F12 for example is able to achieve a maximum static pressure of 17.5Pa at only 59mA of current, which is significantly lower than any of the other fans I tested. The Delta fans have beefy motors and draw four or five times as much current for the same performance. I am planning to look more closely at the motor designs, having already disassembled a couple of fans, so stay tuned!
There are exceptions to the performance rule. Some rare fans like bequiet Silent Wing 3 can be choked by a medium density heatsink and produce about half the expected performance per RPM. These particular are about half the static pressure of others while having 3/4 the airflow, trading it for much lower noise. Even in free air, they produce a shorter range "blow". This kind of a fan is meant to be used in high numbers distributed over a case in a negative pressure configuration and has problems even blowing through a thick or dense radiator. Most computer fans are not this design.
The Thermalright cooler I used has a higher fin density than most modern air coolers which is why I chose it for the tests. Even so, the actual resistance is still very low, so there's plenty headroom for each fan to handle a more demanding cooling application.
Hello! I want to do these tests here at home, I'm not sure which pressure gauge I should buy, 0-30pa or equal to your 0-60pa, I found a -60 - 60pa, I wanted to have the best precision, I'm going to test PC fans and use an industrial one just for reference, which is a Sanyo Denki San Ace 120 4A 12V, I'm waiting for a response 🙏
Well that's disingenuous to say there is no static pressure designs. Pressure is just a function of airflow and resistance. The steeply pitched blades of the airflow fans will develop more pressure than the curved blades of the static pressure fans at the same RPM as it will move more air at the same RPM. The static pressure optimized fans have curved blades that needs more RPM to move the same amount of air. The issue with the steeply pitched blades is that they make more noise at the same speed and the steeply curved blades are more likely to go into stall region against a high resistance which will cause even more noise. So, it will make way more noise trying to get the same static pressure as curve bladed static pressure fans. The term static pressure fans doesn't mean that it's making more static pressure, rather it is better optimized at making higher static pressure more efficiently (ie. less noise and motor power).
I didn't ever say that there are no static pressure designs. What I said is that there's no such thing as a "static pressure fan." The term is used very effectively to market a range of Gaming PC products to a consumer base that knows no better, which is precisely why I'm making this series of videos. Static pressure fans don't exist!
@@FanPhys Ok I don't understand how saying "Static pressure fans don't exist" is different to "no static pressure designs". Getting the static pressure designed fans for use on radiator still results in a much better end user experience than using the airflow fans on radiators so I don't see the issue that "The term is used very effectively to market a range of Gaming PC products to a consumer base that knows no better".
@@OTechnology The fans that are optimised for static pressure don't exist on the PC fan market. That's what I'm getting at. And there is also no such thing as an "airflow fan" because ALL axial fans are airflow fans. These catchphrases are marketing hooks which make you think you are buying a special product... you are just buying a fan, and they all do the same thing regardless of design, as I am demonstrating in this series.
@@FanPhys What do you mean pressure optimized fans don't exist? These pressure optimized marketed fans literally work better for higher static pressure requirements since they produce less noise. Your argument is odd to say the least.
One revolution is one revolution, regardless of size. Some fans needed a boost in voltage to reach the standardised test speed of 1600rpm; testing a 1200rpm fan against a 1600rpm fan is not an apples-to-apples comparison.
Isn't the pressure differential across the fan the reason for the airflow in the first place? Doesn't this make pressure and flow essentially the same thing?
It's a causal relationship - the low pressure in front of the fan is what causes the airflow at the rear of the fan. In a purely linguistic sense it helps to think of pressure and flow separately, but it gets complicated pretty quickly because we're dealing with fluids which, unlike solids, can have "velocity pressure", which is not the pressure of the fluid itself (static pressure) but the pressure exerted by the fluid flow. In simple terms, "airflow" is volume over time, and "pressure" is force over area... but you need pressure before you can have flow. Yay physics!
@@FanPhys This is my only benchmark for fans now, I have literal boxes full of fans I've collected over the years and the only thing i care about is "does it do enough to keep my PC running without annoying me". I recently picky up the Deepcool Assassin IV and immediately shelved it because the fans generate a noticeable non-airflow whining sound.
My theory: Curved blades are better at focusing the airflow, and a shallower angle of attack twists the air less. Heatsinks are basically flow straighteners, so the straighter the flow is before the heatsink, the less losses there are when the air passes through. Something like the Corsair SP120 Elite seems like a very good design when considering this; the blades have a pretty shallow angle of attack, and the stator vanes gently straighten the flow. It also avoids the acoustic issue that the Noctua NF-F12 has, where the straight blade trailing edge lines up almost perfectly with the straight leading edge of the stator vane, and the interaction creates a droning sound. The trailing edge of the SP120 Elite blade is curved, and the stator vanes are also curved, so they are closer to perpendicular rather than parallel like the Noctua.
From what I can find in research papers and articles, it looks like turbulence can improve heatsink performance where there is a pressure deficit, which is probably why the Gentle Typhoon is such a good heatsink fan despite it's low pressure performance. In the case of the best all rounder, the Delta WFC, it has large petal-like blades and a strong motor. The NF-F12 does make a distinct and obvious noise where most of the other fans don't, however I really don't think that the stator design does much of anything at all for performance, not on a 1,500rpm fan. The PPC version would be a different story!
From my not-so-scientific testing (smoke machine in front of the fan) this is true, the Noctua A12x25 "directs" the airflow a lot more than an S12B or A. I can't say anything about how (and if) this effects performance on a heatsink though.
What it actually proves is that static pressure is not really a good metric when comparing a fan's cooling performance, because pressure isn't the only thing happening when a fan blows air through a heatsink; it's significantly more complex than that.
MajorHareware "fans" would be very interested in this video.
What about the static pressure curve over the whole speed range of the fan? Would be interesting how much it changes depending on rpm.
this is what im thinking aswell. comparing Noise, airflow, and pressure, on an rpm curve would give a more complete picture. Im never going to run a fan at anywhere near max rpm, so the performance there is useless to me. What is the performance at 500, 750 and 1000 rpm? is the "pressure" fans better at those rpms? or are they doing the same cooling, but quieter?.
@@kasslerify Performance at 1,000rpm and below generally levels out regardless of the type/design of the fan. Furthermore, the level of pressure in the range of speeds
@@FanPhys No - that is exactly what they are made for.
@@ABaumstumpf I don't follow - what were you responding to?
@@FanPhys i am responding to your false claim of "Performance at 1,000rpm and below generally levels out regardless of the type/design of the fan. Furthermore, the level of pressure in the range of speeds
Why did you choose to normalize RPM? To me it makes way more sense to normalize either wattage or (more likely what people choose speed based on) noise level.
It could be that I havn't thought this through enough yet, but don't understand why you went with the method you did.
Since my tests are purely performance tests and not noise tests or power tests, it makes sense to level the playing field by setting each fan to the same speed. It's also worth noting that fan motors have varying levels of efficiency, so one fan might draw three times the power of another fan for the same speed and performance. I will consider this for a future video however it's a big subject that would require a lot of time to adequately cover!
@@FanPhys Yeah, that makes sense, but I also feel like it's a bit weird since these fans with very different motor characteristics also are intended to run at different speeds, sometimes with completely different both upper and lower speed bounds. It could just be that the static pressure fans are intended for pressure at lower speeds compared to other fans, and that would be a nuance missed by this testing.
Makes more sense to test them at 12V. Because that's the maximum you'd have in your PC.... I mean they aren't hitting the RPM value at 12V because either the motor is crap OR the blade geometry is providing more resistance.
And if they are spinning slower they probably can't actually perform as well...
@@insanemalThey need to be at the same RPM for consistency, IF the RPM was different, its not the design of the blade that would be the only factor giving diffrent results
They need to be at the same RPM for consistency, IF the RPM was different, its not the design of the blade that would be the only factor giving diffrent results
Someone please explain to me how the static pressure optimized Noctua fan building up more pressure and performing better in the thermal test than the airflow optimized Noctua is proving that static pressure fans don't exist?
You are cherry-picking two fans whose performance conforms to your own bias. The Cougar, eLoop and Gentle Typhoon all beat the NF-F12 in the thermal test despite their inferior pressure performance. There's no such thing as a "static pressure fan" because static pressure is a property of the fluid, not the fan. It's a nonsensical phrase conjured by marketing execs.
Static pressure is created by the fan. All fans are static pressure fans just to different degrees. The fluid doesn't move on its own. Fan design plays a part in amount of pressure created in a given space. The a12 is marketed as a do it all fan, not specifically a static pressure fan, ~21Pa at 2000rpm. F12, 23Pa at 1500rpm. The sp120 is marketed as a static pressure fan but it's only rated at ~13Pa at ~1400rpm. The p12 is rated at ~20Pa at 1800rpm. So your results are pretty much expected all those fans based on their claimed specs. Corsairs marketing was bad, but it usually is. Your test was flawed anyhow.
I was under the impression that the curved blade design was a life cycle consideration- that it allowed the motor to work against a higher load for longer. Could that be a part of it?
Did you pull enough data to denominate by (or control by) wattage instead of RPM?
It's a good shout, but I quickly abandoned measuring power draw with my DMM because each fan motor is quite different. The NF-F12 for example is able to achieve a maximum static pressure of 17.5Pa at only 59mA of current, which is significantly lower than any of the other fans I tested. The Delta fans have beefy motors and draw four or five times as much current for the same performance. I am planning to look more closely at the motor designs, having already disassembled a couple of fans, so stay tuned!
@@FanPhys Yeah that's fascinating that the pressure-per-power could be so different. Looking forward to more!
@@macspresso yup, we need pressure vs power vs dBm sound level 3D plot
I got super spooked at 3:32 as I live alone and thought someone was in the other room
I am glad I wasn't the only one, I was like "is someone yelling outside?"
I need to do a better job duct taping my kids...
There are exceptions to the performance rule. Some rare fans like bequiet Silent Wing 3 can be choked by a medium density heatsink and produce about half the expected performance per RPM. These particular are about half the static pressure of others while having 3/4 the airflow, trading it for much lower noise. Even in free air, they produce a shorter range "blow". This kind of a fan is meant to be used in high numbers distributed over a case in a negative pressure configuration and has problems even blowing through a thick or dense radiator.
Most computer fans are not this design.
This is fine on a brand new radiator or immediately after cleaning dust out, but what about once there's some resistance/blockage?
The Thermalright cooler I used has a higher fin density than most modern air coolers which is why I chose it for the tests. Even so, the actual resistance is still very low, so there's plenty headroom for each fan to handle a more demanding cooling application.
Hello! I want to do these tests here at home, I'm not sure which pressure gauge I should buy, 0-30pa or equal to your 0-60pa, I found a -60 - 60pa, I wanted to have the best precision, I'm going to test PC fans and use an industrial one just for reference, which is a Sanyo Denki San Ace 120 4A 12V, I'm waiting for a response 🙏
Well that's disingenuous to say there is no static pressure designs. Pressure is just a function of airflow and resistance. The steeply pitched blades of the airflow fans will develop more pressure than the curved blades of the static pressure fans at the same RPM as it will move more air at the same RPM. The static pressure optimized fans have curved blades that needs more RPM to move the same amount of air.
The issue with the steeply pitched blades is that they make more noise at the same speed and the steeply curved blades are more likely to go into stall region against a high resistance which will cause even more noise. So, it will make way more noise trying to get the same static pressure as curve bladed static pressure fans. The term static pressure fans doesn't mean that it's making more static pressure, rather it is better optimized at making higher static pressure more efficiently (ie. less noise and motor power).
I didn't ever say that there are no static pressure designs. What I said is that there's no such thing as a "static pressure fan." The term is used very effectively to market a range of Gaming PC products to a consumer base that knows no better, which is precisely why I'm making this series of videos. Static pressure fans don't exist!
@@FanPhys Ok I don't understand how saying "Static pressure fans don't exist" is different to "no static pressure designs".
Getting the static pressure designed fans for use on radiator still results in a much better end user experience than using the airflow fans on radiators so I don't see the issue that "The term is used very effectively to market a range of Gaming PC products to a consumer base that knows no better".
@@OTechnology The fans that are optimised for static pressure don't exist on the PC fan market. That's what I'm getting at. And there is also no such thing as an "airflow fan" because ALL axial fans are airflow fans. These catchphrases are marketing hooks which make you think you are buying a special product... you are just buying a fan, and they all do the same thing regardless of design, as I am demonstrating in this series.
@@FanPhys What do you mean pressure optimized fans don't exist? These pressure optimized marketed fans literally work better for higher static pressure requirements since they produce less noise. Your argument is odd to say the least.
Can manual RPM readings be consistent with the varying hub sizes though? I don't get the point of overvolting, since your motherboard can't do that.
One revolution is one revolution, regardless of size. Some fans needed a boost in voltage to reach the standardised test speed of 1600rpm; testing a 1200rpm fan against a 1600rpm fan is not an apples-to-apples comparison.
Isn't the pressure differential across the fan the reason for the airflow in the first place? Doesn't this make pressure and flow essentially the same thing?
It's a causal relationship - the low pressure in front of the fan is what causes the airflow at the rear of the fan. In a purely linguistic sense it helps to think of pressure and flow separately, but it gets complicated pretty quickly because we're dealing with fluids which, unlike solids, can have "velocity pressure", which is not the pressure of the fluid itself (static pressure) but the pressure exerted by the fluid flow. In simple terms, "airflow" is volume over time, and "pressure" is force over area... but you need pressure before you can have flow. Yay physics!
best high static pressure and silent fan?
You can't have both, sorry!
I like quieter fans. Thanks for testing!
I'm with you on that one - when I'm done with testing I usually throw the P12 back on the heatsink because it's super quiet and doesn't bug me x)
@@FanPhys This is my only benchmark for fans now, I have literal boxes full of fans I've collected over the years and the only thing i care about is "does it do enough to keep my PC running without annoying me". I recently picky up the Deepcool Assassin IV and immediately shelved it because the fans generate a noticeable non-airflow whining sound.
I love the U.S.A Ice Gales Xtra .... probably would break your "pressure" meter
2:29 and 4:14
My theory: Curved blades are better at focusing the airflow, and a shallower angle of attack twists the air less. Heatsinks are basically flow straighteners, so the straighter the flow is before the heatsink, the less losses there are when the air passes through. Something like the Corsair SP120 Elite seems like a very good design when considering this; the blades have a pretty shallow angle of attack, and the stator vanes gently straighten the flow. It also avoids the acoustic issue that the Noctua NF-F12 has, where the straight blade trailing edge lines up almost perfectly with the straight leading edge of the stator vane, and the interaction creates a droning sound. The trailing edge of the SP120 Elite blade is curved, and the stator vanes are also curved, so they are closer to perpendicular rather than parallel like the Noctua.
From what I can find in research papers and articles, it looks like turbulence can improve heatsink performance where there is a pressure deficit, which is probably why the Gentle Typhoon is such a good heatsink fan despite it's low pressure performance. In the case of the best all rounder, the Delta WFC, it has large petal-like blades and a strong motor. The NF-F12 does make a distinct and obvious noise where most of the other fans don't, however I really don't think that the stator design does much of anything at all for performance, not on a 1,500rpm fan. The PPC version would be a different story!
From my not-so-scientific testing (smoke machine in front of the fan) this is true, the Noctua A12x25 "directs" the airflow a lot more than an S12B or A.
I can't say anything about how (and if) this effects performance on a heatsink though.
All this proves is your methodology for testing static pressure is poor.
What it actually proves is that static pressure is not really a good metric when comparing a fan's cooling performance, because pressure isn't the only thing happening when a fan blows air through a heatsink; it's significantly more complex than that.