Id like to argue the use of a tray before you dinning, even a small one to remind you of the right way. How your grandma told you we dont eat like bacteria, by just bumping into things, see your way tomthat or over a table so you get the cut above where to take it
I am the designer of Hold My Beer. All of my tests were done on a rig that was fully sealed so I can confirm the thats where the differences in measurements came from. On my rig "Hold My Beer " got almost double the preasure that "The Wonder from Downunder" did. Thats why I predicted 14mm h20. Still, I think 10.5mm h20 fully sealed is good! Just for fun(bonus episode), can you test the top 5 at the end of the season in a fully sealed test rig?
Every season he should definitely upgrade his setup to try and get higher quality machinery and tests going. It's a good reason to explain "the numbers in this season will not be 1:1 comparable with last season". Plus, it's extra content to bring old fans into the new season, and see how they fare against the higher accuracy machines/tests.
@@weeveferrelaine6973 What I think would be really interesting, is to test the top 5 fans as you suggest, but on 2 variables, absolute static pressure (fully sealed as humanly possible) vs. flow (completely unrestricted). Based on Centrifuge Madness static number, I would wager that Hold My Beer would still be top spot for fully sealed static pressure, but CM has much more flow, allowing it to reduce it's static loss in the test rig.
Definitely one of those I'd be excited to see rad performance or unconstrained. Anything that does reasonably well in the 80mm form factor (within 2 or so inches of height as well) gives me that real-world potential feeling.
For next season, why not directly measure how well each fan cools a CPU heatsink? Heat the heatsink to a standard temperature, then remove the heat source, and turn on the fan. Measure how long it takes for the fan to drop to a specific finishing temp. Fastest wins. The setup could be as simple as mounting each test fan to the heatsink, then popping the whole assembly into a preheated oven (to say, 120C) for a specific period (e.g. 20 minutes). Remove from oven, attach to power supply, and start the clock (should be at the same elapsed time since removal from oven for all fans - e.g. 2 minutes). Stop the clock when the heat sink temp has dropped to, say 1ºC above ambient (use the same ambient temp in all runs). I don't know if an IR thermometer is adequate for measuring heatsink temps. If it is, then you're good to go! It might be fun to do this for the top 10 fans already tested just to see how well cooling correlates with static pressure and volume of air moved.
for the most part, more pressure -> less air volume. And also, more pressure -> more air lost in leaks. so, the imperfections in your setup are basically holding back the good fans.
I really hope that in future seasons we have a complete package of tests, rating on both static pressure and airflow to find the best overall fan, not just hyper-specialised ones. Also it would be neat to have a table for "traditional" fans, as well as the more outlandish contraptions. The 2in1 looks like it could be a fantastic blend of the two criteria, which is all the more awesome because its still pretty much a "traditional" fan in terms of form-factor.
It has been fun watching your waterloop in that system on the right slowly getting lower over the years. Also, shout out to the guy just sitting in the air @ 5:50, lmao.
might be interesting to take the top ten at the end of the season and do a best of the best showdown with everything sealed up tight (my thoughts are that at lower pressure of used to calibrate the test bench the leaks are not as noticeable but as you go up in pressure the loss to air leaks is not linear)
We've definitely gone from the realm of fans to pumps. LOL. I love it. I've been learning how to do more precision modeling in Blender for a properly optimized and "mutated" version of the Biomass fan I submitted and was tested a few years back. 😎
Since your test system is not sealed, you are not actually measuring static pressure because the pressure isn't static when it's leaking! You're measuring a combination of CFM and static pressure. The amount of air leaking makes your setup bias towards fans with both high static pressure and high CFM, since you have a leak of a fixed amount in the system. If something moves less air but with much more static pressure, it would be a disadvantage. You should re-test all the fans with it properly sealed.
No, it's the static pressure of this test setup. Static pressure is always relative to the system and environment it's being measured in. I could take a sealed system, measure it on the seafront, then measure it on top of a mountain, and get two different measurements because the air going in has different pressure and humidity. Static pressure isn't some constant of a fan, it's always relative to the system the fan operates in.
@@habilainNo to what? This isn't a sealed system, therefor not measuring static pressure of the fan since static pressure measurements require a sealed system. This isn't measuring the static pressure of the system, because the system isn't static, it's leaking into earth's atmosphere and earth isn't a sealed system. The static pressure of the system would be the same as the fan without the leaks, since testing the system would require it be sealed and therefor the fan could hold at full pressure. You can say you are measuring pressure at equilibrium, but it sure isn't static. Static pressure is a constant given a set frame of reference relative to outside air density, composition, temperature, etc.
@@BlakeEM I think you're misunderstanding something. There are two components for pressure: dynamic pressure, which is pressure which is attributed to the velocity of the fluid (in this case, air), and static pressure which is exerted independent of the velocity of the fluid. Every system has both static and dynamic pressure. A good example is an aeroplane - static pressure comes about from the aeroplane moving through air, and dynamic pressure from wind hitting the aeroplane. Also note that the aeroplane is a completely open system, and still has static pressure. So just because the air is moving or a system isn't sealed, it does not mean you cannot measure static pressure. I'm not sure what exact measurements MajorHardware is doing, but my assumption was that he's got a tool designed to measure static pressure (and not total pressure) and therefore is measuring the velocity of the air, calculating dynamic pressure and subtracting this from total pressure to solve for static pressure.
Oooh, so the pressure tests this season are a pressure test with a CFM floor that must be passed to be effective. That's not a good test then as you're actually testing 2 things at once. I agree with maker of "Wonder from down under" that a bonus episode at the end of the season should be made of the top performing fans with a fully sealed test rig to get their true peak static pressure results!
The test setup not being fully sealed is quite a flaw in the testing methodology, especially since now we're measuring pressure vs airflow. The airflow is now tainting the test results for pressure. Ofcourse you'd never actually use a fan with a 1x2cm opening, but still 😅 quite a shame this wasn't sealed from the start, especially when it was known to leak
Yeah this is such a dichotomy. In a real PC case it won't be sealed so it's not realistic. However... this isn't about being super realistic it's about seeing what is possible. So I like that you included both results. Thanks!
The leaks definitely wouldn't be consistent, more pressure would have a higher flow through the leak. Can't halfass the testing and expect designers to wholeass participation
About the helicopter fan - it's actually funny how you spent 30 minutes on it, and then almost 3 minutes talking about it in your video, which at the moment of writing it, adds up to almost 90 000 minutes in total that humanity spent on that one innocent endeavour of your. That's already more than 62 days of humanity's time.
Just imagine all the days wasted in all universes of all aliens and simulated or not, god beings or not, just imagine what could have been done for Africans or the soon extinct brumblebuss-folk in universe AI2. SMH. :P
@6:11 - so kind of felt like how I just wanted to watch a fan showdown and instead got 1 minute of advertisement for Factor and nearly 3 minutes of you complaining about AI.....
Love this content. On the Hold-My-Beer. I look at it like having a 1HP water pump pushing water through a 1" pipe, which will have amazing pressure, but when it suddenly dumps into a 10" pipe the pressure will drop drastically as the water expands to fill the available space. This lower pressure stays constant as long as it freeflows through the 10" pipe. The output would have to be stepped back down through smaller pipes to achieve the origial pressure. Also, air density, which corresponds with static pressure, varies with environmental temperature, humidity and altitude which may explain the differences in the designer's test results versus yours.
Ok, straight talk. Here is what you need to do: you need to seal up your rig and re-do the tests for all the fans submitted this static pressure season, and then update the leaderboard with the new numbers. If it changes the order, then so be it, they should have been scored that way all along. Doing a static pressure test in a rig that isnt sealed defeats the entire purpose of doing a static pressure test. You're determining winners and losers here from your viewers' submissions. You're asking people to take the time to design and submit fans for the channel, but who is going to do that if they can't take your results seriously bc youre doing static pressure tests, and determining winners on a rig that leaks air? You can still fix this, my man! Just gotta do it right and fair for everyone!
I think. That the setup is suitable for its purpose. yes we could improve the seals, but the test really is 'how much can you build given these conditions' rather than a scientific measurement of perfection. The result thus, is valid, and penalizes low flow high pressure setups, and encourages a useful volume of air for the purpose of static pressure, instead of fixating on max pressure alone. You are supposed to be doing something productive with a static pressure fan after all, not merely inflating balloons.
I think comment comes off a bit too aggressive - he's putting on a fun competition, not deciding someone's fate. But I do think we would all appreciate if you could redo the tests with a properly sealed unit so that everyone has a fair chance being measured against the stated criteria. Maybe it's not practical to do that for every fan, but the top contenders would be nice to see.
@@glenmcgillivray4707 It's not because it's not static pressure if it leaks. The test is already "How much can you build given these conditions" but it unfairly disadvantages fans with low CFM. It shouldn't be called a static pressure test if the pressure isn't static
No changes or upsets would occur with the rankings on the leaderboard. The only way changes would occur is if the fans are tested one by one while updating for every fan. The "sealed" leaderboard would still look identical (except for the slight increase to all fans in the measured result) compared with the "leaky" leaderboard. All the fans are strong enough to overcome the outside pressure, hence the leak. But they are not strong enough to make the "hole" bigger.
I feel like the 2-in-1 might perform better with a velocity stack (aka a venturi) on the intake side, to get a little more performance out of the axial part of the fan.
Since you mentioned you weren't getting that many fan submissions, maybe next season you can revamp everything and do tests sealed from now on! It'll give you a chance to retest some older fan submissions as well.
based on your explanation of what is happening, your setup for testing has a theoretical limit for the maximum testable limit. Once a fan exceeds the maximum limit its going to produce a static value which is the rate at which air is being forced out of the system and no longer creating pressure. anything that is able to exceed this threshold will produce a static value at or near the limitation.
My opinion on 8:45: by pure sight, I think the 2N1 and offset rotor would be the best intake options because of the amount of intake and overall airdistribution.
Air loss through your tube may increase non-linearly as the pressure rises - so that fans with higher static pressure experience more air loss than the other fans do (making them appear less effective than they are). This may account for some differences in measurement as pressure increases, even if the Noctua measurements stay the same. Pretty sure there is a square root somewhere in the compressible and incompressible fluid flow equations as a function of pressure but you'd have to research it, just my initial feeling watching this.
This works pretty good if you actually want that helicopter image. NightCafe's cinematic style on the Google Imagen 3.0 Fast model. Prompt "boeing ah-6 helicopter shooting machine guns at a wildfire"
Hi mate. Would like to propose a scoring system for next season. I'm sure you've heard of Bernoulli's principal before - a simplified way of using conservation of energy to analyse laminar fluid flow. Basically, it says that pressure (P) + kinetic energy (density*v^2/2) + potential energy (density*go*h) of a flow remains constant (along flow lines where no energy is input or extracted from the flow field). Measuring the amount of energy that a fan adds to the air could be a good way of scoring its performance - you have tools to measure pressure and flow output already. The gravity term can be neglected since they're all going to be tested with the same rig. Would be interesting to see if there's any correlation between this and noise output too. Keen to hear thoughts.
Don't worry I'm working on a real contender. It's in the testing phase right now. A quick refresh I'm the guy who sent an entire file with 52 fans. You chose tangent.
a large problem I see with “hold my beer” is the blade curvature. Straight or forward curving blades will improve airflow but reduce pressure. Backward curving blades do not have as much airflow but they produce higher pressure.
I asked grok to create a picture of that firefighter prompt and it had a firefighter sitting inside of a cockpit of an AH6 holding 2 literal mini guns. Like really small glocks
Because most people trying to optimize cooling are using water coolers nowadays, and for radiators it's more important to be able to push through the resistance of the fins rather than at a high velocity when unimpeded. Even when pushing through an air cooler there is some amount of resistance which requires the fans to be able to produce a decent amount of static pressure
its usually because computer fans are having to 'push' that air through something that obstructs the flow of air be it a water-cooling radiator or the fins that are attached to heat-pipes. Velocity matters but if that velocity drops drastically the moment it encounters the fins it doesn't help all that much in 'pushing' that now hot air out. CFM is much the same, though probably better then raw velocity, the volume of air a fan can move is helpful to know but in a restricted environment such as a computer case that likely has limits to how much air can be 'pulled' or 'pushed' out of the case upon which it will run into needing to work harder, ie pressure, to overcome that.
Had the same problem. I needed some pictures for my DND compain. But I have found a way. Get some food (homemade or takeout doesn't matter), go to your nearest art school/college, look for some kids drawing outside (under trees, inside gazebos, on rooftops, under cars etc), then throw food at them and draw them into your house. Now you have sandwich operated image generator.
The lesson is obvious. Stop feeding the AI machine, hire some dude of Fiver or ask one of your fans to make the image with photoshop! I'm sure there are plenty of creatives that would jump at the chance to help you out!
I know you put it on screen, but you should also say what frequency the decibel peaks are at. You don't have to of course, but it would help mobile users.
Big FAN of your videos. Just one thing and this is not a criticism or anything like that to anyone. The hold my beer has an unfair advantage because the rotor is larger than others. again its possible that this is how it looks in the video, and again no criticism to Stephan or any one.
As somebody who runs a local copy of Stable Diffusion, it's heavily influenced by the model and prompt you use. Some models are just completely unable to do specifics, while others are so specific you can't get anything else.
Another way to look at it is Newtons 3rd law. For every action there is an equil and opposite reaction. meaning that fans can only ever push back what they are pushing against. If it leaks, the fans will never show the full pressure they can produce.
The V-6 pump for a carpet cleaning job when I was a teen could pull over 40 feet of Hg. We'd joke about turning ourselves inside out with it. You talkin' about mm of h²o is very funny.
Could you do 1 vs 2 from the same creator. Properly sealed so we could really see how good they are. Also would the 2 work better with a bigger opening?
I tried your prompt idea in nightcafe using flux... it got the helicopter right every time, only had the forest fire in one of them, and only had the guns firing in one of them, but overall better than what you showed here.
Use code 50HARDWARE to get 50% OFF plus free shipping on your first Factor box at bit.ly/3OrRVCP!
hidden behind info wall marketing bullcrap.
Id like to argue the use of a tray before you dinning, even a small one to remind you of the right way. How your grandma told you we dont eat like bacteria, by just bumping into things, see your way tomthat or over a table so you get the cut above where to take it
can you to reduce the lenght of the blades by 1 centimeter or 2/5 of an inch?
Forget ChatGPT, if you want good AI images of belugas go to civitai.
Bing image generator is one of the best on the net. I suggest that to play aground with. Cheers mate.
I am the designer of Hold My Beer. All of my tests were done on a rig that was fully sealed so I can confirm the thats where the differences in measurements came from. On my rig "Hold My Beer " got almost double the preasure that "The Wonder from Downunder" did. Thats why I predicted 14mm h20. Still, I think 10.5mm h20 fully sealed is good!
Just for fun(bonus episode), can you test the top 5 at the end of the season in a fully sealed test rig?
I'm wondering, did your setup open up in the same manner like the one used in the tests here?
Every season he should definitely upgrade his setup to try and get higher quality machinery and tests going. It's a good reason to explain "the numbers in this season will not be 1:1 comparable with last season".
Plus, it's extra content to bring old fans into the new season, and see how they fare against the higher accuracy machines/tests.
I agree, a bonus ep at the end of the season would be good to see with a fully sealed test to allow the fans to show their true peak pressures.
yes :) (apart from the fact that a bonus episode is always a good idea)
@@weeveferrelaine6973 What I think would be really interesting, is to test the top 5 fans as you suggest, but on 2 variables, absolute static pressure (fully sealed as humanly possible) vs. flow (completely unrestricted). Based on Centrifuge Madness static number, I would wager that Hold My Beer would still be top spot for fully sealed static pressure, but CM has much more flow, allowing it to reduce it's static loss in the test rig.
Man that 2n1 may be the first wildly successful attempt at a double fan. And it looks freaking sick! Bravo to the creator!
Definitely one of those I'd be excited to see rad performance or unconstrained. Anything that does reasonably well in the 80mm form factor (within 2 or so inches of height as well) gives me that real-world potential feeling.
@Varadiio hehe 'rad' performance
Because it looks like a radish 😂
top 4 deserves a sealed rerun before ending the season
Upvote
I appreciate that GPT just picked a random helicopter and wrote "AH-6" on it... multiple times. Shockingly lifelike.
For next season, why not directly measure how well each fan cools a CPU heatsink?
Heat the heatsink to a standard temperature, then remove the heat source, and turn on the fan. Measure how long it takes for the fan to drop to a specific finishing temp. Fastest wins.
The setup could be as simple as mounting each test fan to the heatsink, then popping the whole assembly into a preheated oven (to say, 120C) for a specific period (e.g. 20 minutes). Remove from oven, attach to power supply, and start the clock (should be at the same elapsed time since removal from oven for all fans - e.g. 2 minutes). Stop the clock when the heat sink temp has dropped to, say 1ºC above ambient (use the same ambient temp in all runs).
I don't know if an IR thermometer is adequate for measuring heatsink temps. If it is, then you're good to go!
It might be fun to do this for the top 10 fans already tested just to see how well cooling correlates with static pressure and volume of air moved.
Could mount the heatsink to a hot plate. Have a thermometer between the hot plate and heatsink.
Nah, the tests should be useless for real-world performance, showing only half of the picture
I’m not sure a pc fan would survive 120c for 20 minutes tbh
for the most part, more pressure -> less air volume.
And also, more pressure -> more air lost in leaks.
so, the imperfections in your setup are basically holding back the good fans.
I really hope that in future seasons we have a complete package of tests, rating on both static pressure and airflow to find the best overall fan, not just hyper-specialised ones. Also it would be neat to have a table for "traditional" fans, as well as the more outlandish contraptions.
The 2in1 looks like it could be a fantastic blend of the two criteria, which is all the more awesome because its still pretty much a "traditional" fan in terms of form-factor.
It has been fun watching your waterloop in that system on the right slowly getting lower over the years.
Also, shout out to the guy just sitting in the air @ 5:50, lmao.
might be interesting to take the top ten at the end of the season and do a best of the best showdown with everything sealed up tight (my thoughts are that at lower pressure of used to calibrate the test bench the leaks are not as noticeable but as you go up in pressure the loss to air leaks is not linear)
Especially with something that has pretty high static pressure with minimal CFM
Even top 5 would be worth it I think.
I'd like to see that, the designs cant make full pressure if there is not full pressure pushing back!
Revisit the top fans with the fully sealed static test.
We've definitely gone from the realm of fans to pumps. LOL. I love it. I've been learning how to do more precision modeling in Blender for a properly optimized and "mutated" version of the Biomass fan I submitted and was tested a few years back. 😎
"Beluga"
I saw it as well :D
"Beluga"
Beluga
Since your test system is not sealed, you are not actually measuring static pressure because the pressure isn't static when it's leaking! You're measuring a combination of CFM and static pressure. The amount of air leaking makes your setup bias towards fans with both high static pressure and high CFM, since you have a leak of a fixed amount in the system. If something moves less air but with much more static pressure, it would be a disadvantage. You should re-test all the fans with it properly sealed.
No, it's the static pressure of this test setup. Static pressure is always relative to the system and environment it's being measured in. I could take a sealed system, measure it on the seafront, then measure it on top of a mountain, and get two different measurements because the air going in has different pressure and humidity. Static pressure isn't some constant of a fan, it's always relative to the system the fan operates in.
@@habilainNo to what? This isn't a sealed system, therefor not measuring static pressure of the fan since static pressure measurements require a sealed system. This isn't measuring the static pressure of the system, because the system isn't static, it's leaking into earth's atmosphere and earth isn't a sealed system. The static pressure of the system would be the same as the fan without the leaks, since testing the system would require it be sealed and therefor the fan could hold at full pressure. You can say you are measuring pressure at equilibrium, but it sure isn't static. Static pressure is a constant given a set frame of reference relative to outside air density, composition, temperature, etc.
@@BlakeEM I think you're misunderstanding something. There are two components for pressure: dynamic pressure, which is pressure which is attributed to the velocity of the fluid (in this case, air), and static pressure which is exerted independent of the velocity of the fluid. Every system has both static and dynamic pressure. A good example is an aeroplane - static pressure comes about from the aeroplane moving through air, and dynamic pressure from wind hitting the aeroplane. Also note that the aeroplane is a completely open system, and still has static pressure.
So just because the air is moving or a system isn't sealed, it does not mean you cannot measure static pressure. I'm not sure what exact measurements MajorHardware is doing, but my assumption was that he's got a tool designed to measure static pressure (and not total pressure) and therefore is measuring the velocity of the air, calculating dynamic pressure and subtracting this from total pressure to solve for static pressure.
Oooh, so the pressure tests this season are a pressure test with a CFM floor that must be passed to be effective. That's not a good test then as you're actually testing 2 things at once.
I agree with maker of "Wonder from down under" that a bonus episode at the end of the season should be made of the top performing fans with a fully sealed test rig to get their true peak static pressure results!
I love how many times you remixed Chat GPT lol GPP GTP etc lol
Was about as consistent as the "AI" itself :D
The test setup not being fully sealed is quite a flaw in the testing methodology, especially since now we're measuring pressure vs airflow. The airflow is now tainting the test results for pressure.
Ofcourse you'd never actually use a fan with a 1x2cm opening, but still 😅 quite a shame this wasn't sealed from the start, especially when it was known to leak
I wouldn't have designed it the way I did if I had known it leaked. But yea, you aren't going to make pressure if there is no pressure pushing back.
Fan Showdown: Where Fans make Fans to Compete with other Fans in producing the best Fan ever
I love the fan showdown! Looking forward to more
Yeah this is such a dichotomy. In a real PC case it won't be sealed so it's not realistic. However... this isn't about being super realistic it's about seeing what is possible. So I like that you included both results. Thanks!
The bunny trail of the helicopter chatGPT fiasco was great.
Awesome. I don't know a thing about any of this but I enjoy the creativity of all the submissions. Excellent stuff!
What a Beluga of a time !!
Bro i though no one could see it lol
Always good to see a new video from Major Hardware. I hope you're doing well, and thank you!
Can you do a special episode where you take some top fans and actually properly balance/stabilize them before testing again?
The leaks definitely wouldn't be consistent, more pressure would have a higher flow through the leak.
Can't halfass the testing and expect designers to wholeass participation
"Radish Cavity" will be the name of my new punk vegan band.
About the helicopter fan - it's actually funny how you spent 30 minutes on it, and then almost 3 minutes talking about it in your video, which at the moment of writing it, adds up to almost 90 000 minutes in total that humanity spent on that one innocent endeavour of your. That's already more than 62 days of humanity's time.
Just imagine all the days wasted in all universes of all aliens and simulated or not, god beings or not, just imagine what could have been done for Africans or the soon extinct brumblebuss-folk in universe AI2.
SMH.
:P
I'm just wondering if using Disturbed's "The Guy" as the face of ChatGPT was a conscious choice
YEA FAN SHOWDOWN TIME
@6:11 - so kind of felt like how I just wanted to watch a fan showdown and instead got 1 minute of advertisement for Factor and nearly 3 minutes of you complaining about AI.....
Watch something else. Go away.
The 2n1 looks just like the impeller in a Dyson AM09 bladeless fan...and I have to respect that.
Love this content. On the Hold-My-Beer. I look at it like having a 1HP water pump pushing water through a 1" pipe, which will have amazing pressure, but when it suddenly dumps into a 10" pipe the pressure will drop drastically as the water expands to fill the available space. This lower pressure stays constant as long as it freeflows through the 10" pipe. The output would have to be stepped back down through smaller pipes to achieve the origial pressure. Also, air density, which corresponds with static pressure, varies with environmental temperature, humidity and altitude which may explain the differences in the designer's test results versus yours.
i would like to see the difference between between ax12 and a 3d printed copy
and maybe other models
the gpt voiceovers had me rollin! 🤣🤣
and the " screenshot from fs24" !!!🤣🤣
Ok, straight talk. Here is what you need to do: you need to seal up your rig and re-do the tests for all the fans submitted this static pressure season, and then update the leaderboard with the new numbers. If it changes the order, then so be it, they should have been scored that way all along.
Doing a static pressure test in a rig that isnt sealed defeats the entire purpose of doing a static pressure test. You're determining winners and losers here from your viewers' submissions.
You're asking people to take the time to design and submit fans for the channel, but who is going to do that if they can't take your results seriously bc youre doing static pressure tests, and determining winners on a rig that leaks air?
You can still fix this, my man! Just gotta do it right and fair for everyone!
I think. That the setup is suitable for its purpose. yes we could improve the seals, but the test really is 'how much can you build given these conditions' rather than a scientific measurement of perfection. The result thus, is valid, and penalizes low flow high pressure setups, and encourages a useful volume of air for the purpose of static pressure, instead of fixating on max pressure alone.
You are supposed to be doing something productive with a static pressure fan after all, not merely inflating balloons.
🙄
I think comment comes off a bit too aggressive - he's putting on a fun competition, not deciding someone's fate.
But I do think we would all appreciate if you could redo the tests with a properly sealed unit so that everyone has a fair chance being measured against the stated criteria. Maybe it's not practical to do that for every fan, but the top contenders would be nice to see.
@@glenmcgillivray4707 It's not because it's not static pressure if it leaks. The test is already "How much can you build given these conditions" but it unfairly disadvantages fans with low CFM. It shouldn't be called a static pressure test if the pressure isn't static
No changes or upsets would occur with the rankings on the leaderboard.
The only way changes would occur is if the fans are tested one by one while updating for every fan.
The "sealed" leaderboard would still look identical (except for the slight increase to all fans in the measured result) compared with the "leaky" leaderboard.
All the fans are strong enough to overcome the outside pressure, hence the leak.
But they are not strong enough to make the "hole" bigger.
I feel like the 2-in-1 might perform better with a velocity stack (aka a venturi) on the intake side, to get a little more performance out of the axial part of the fan.
I have just finished the design and will be submitting the STL. I don't have high hopes for me to get another chance.
Beluga.
Since you mentioned you weren't getting that many fan submissions, maybe next season you can revamp everything and do tests sealed from now on! It'll give you a chance to retest some older fan submissions as well.
Ive submitted a fan a couple of times. Ill guess Ill keep trying. Love the series.
So why wasn't the test set sealed from the beginning of the series?
@@mk-wz2sh I'm guessing he thought it was and that it wasn't noticed until the higher pressure fans came along.
that water-pump style fan is one of the fans I've ever seen!
based on your explanation of what is happening, your setup for testing has a theoretical limit for the maximum testable limit. Once a fan exceeds the maximum limit its going to produce a static value which is the rate at which air is being forced out of the system and no longer creating pressure. anything that is able to exceed this threshold will produce a static value at or near the limitation.
Ai BS ends at 6:30 for all you who came to see fans
my saviour. Fuck ai bullshit
The Lord's work lol
saved 6min of my life
Actually, listening to his bit on AI made me not as afraid of AI taking over the world since it couldn't get the picture right, lol. :)
My opinion on 8:45: by pure sight, I think the 2N1 and offset rotor would be the best intake options because of the amount of intake and overall airdistribution.
Air loss through your tube may increase non-linearly as the pressure rises - so that fans with higher static pressure experience more air loss than the other fans do (making them appear less effective than they are). This may account for some differences in measurement as pressure increases, even if the Noctua measurements stay the same. Pretty sure there is a square root somewhere in the compressible and incompressible fluid flow equations as a function of pressure but you'd have to research it, just my initial feeling watching this.
"Hold my beer" HAHA. Classic
I think for testing, you should go back to air volume moved, but make the fans push through a radiator.
"I was in too deep. I wasn't just going to stop asking it to do this... I wanted a picture."
XD
This works pretty good if you actually want that helicopter image.
NightCafe's cinematic style on the Google Imagen 3.0 Fast model.
Prompt "boeing ah-6 helicopter shooting machine guns at a wildfire"
Hi mate. Would like to propose a scoring system for next season.
I'm sure you've heard of Bernoulli's principal before - a simplified way of using conservation of energy to analyse laminar fluid flow.
Basically, it says that pressure (P) + kinetic energy (density*v^2/2) + potential energy (density*go*h) of a flow remains constant (along flow lines where no energy is input or extracted from the flow field).
Measuring the amount of energy that a fan adds to the air could be a good way of scoring its performance - you have tools to measure pressure and flow output already. The gravity term can be neglected since they're all going to be tested with the same rig.
Would be interesting to see if there's any correlation between this and noise output too.
Keen to hear thoughts.
"Raddish cavity" is not a combination of words I thought I'd hear today.
another good one! can't wait for the next one! happy holidays !
lol love the Disturbed face in the back
Hi! The bear fan is completely airtight, while the others have gaps where air escapes, and the result is low pressure
GPT: "I tried my best!"
Gronk: "Gronk."
That GPT fan is really good! Good performance with low noise
Don't worry I'm working on a real contender. It's in the testing phase right now.
A quick refresh I'm the guy who sent an entire file with 52 fans. You chose tangent.
The best part about the GPT rabbit hole is that Gemini is such an also-ran/bad it wasn’t even thought of while he was on Google.
Well... The unfair advantage lies with the lower pressure fans. With a higher pressure differencial the leakage climbs exponentially.
a large problem I see with “hold my beer” is the blade curvature. Straight or forward curving blades will improve airflow but reduce pressure. Backward curving blades do not have as much airflow but they produce higher pressure.
You have to allow for it being 3d printed, curved blades mean longer blades and more flex. By design there is only 1mm clearance .
curious how it stacks up against the centrifuge madness with a fully sealed rig. maybe at the end of the season you can compare top 4 and the noctua
I asked grok to create a picture of that firefighter prompt and it had a firefighter sitting inside of a cockpit of an AH6 holding 2 literal mini guns. Like really small glocks
That gpt yes made me lol so hard 😂😂😂
As a fellow Michigander... this is quite the late in the day upload lol
Oh is he? Michigander checking in. What is daylight?
Next season should be a water pump. Gotta create the best 3d printed water pump to move the move the most volume of water.
The season after that you could even do head pressure.
that hold my beer SCREAMED
jesus
Can someone explain why static pressure is a better metric for cooling fan performance than CFM or velocity?
It's not better it's just what he's testing this season becuse he already has done multiple seasons on flow volume and velocity
Because most people trying to optimize cooling are using water coolers nowadays, and for radiators it's more important to be able to push through the resistance of the fins rather than at a high velocity when unimpeded. Even when pushing through an air cooler there is some amount of resistance which requires the fans to be able to produce a decent amount of static pressure
It's only a better metric if the fan is fighting against restrictions like a radiator.
it knocks the dust off a the radiators!
its usually because computer fans are having to 'push' that air through something that obstructs the flow of air be it a water-cooling radiator or the fins that are attached to heat-pipes.
Velocity matters but if that velocity drops drastically the moment it encounters the fins it doesn't help all that much in 'pushing' that now hot air out.
CFM is much the same, though probably better then raw velocity, the volume of air a fan can move is helpful to know but in a restricted environment such as a computer case that likely has limits to how much air can be 'pulled' or 'pushed' out of the case upon which it will run into needing to work harder, ie pressure, to overcome that.
Had the same problem. I needed some pictures for my DND compain. But I have found a way.
Get some food (homemade or takeout doesn't matter), go to your nearest art school/college, look for some kids drawing outside (under trees, inside gazebos, on rooftops, under cars etc), then throw food at them and draw them into your house. Now you have sandwich operated image generator.
I'm pretty sure thats illegal in most countries, but I like the concept.
Was cool to see a 2 in 1 do so well! beluga
I love the 4 blade fan. Wish people made more of them
The lesson is obvious. Stop feeding the AI machine, hire some dude of Fiver or ask one of your fans to make the image with photoshop! I'm sure there are plenty of creatives that would jump at the chance to help you out!
Beluga! You may want to add a bit of coolant to your PC bc I've noticed it's evaporated a bit compared to a few seasons ago.
I burst out laughing at your AI exp[erience, merely cause its exactly my experience as well. loved it.
Defiantly consistency! I was going to make you your bird, but because of copyright issues it's probably not a good idea aha
"I wanted a picture."
It already _gave_ you a picture!
Can you teat the CFM on the static pressure fans as well?
I'd be curious to see the correlation.
I know you put it on screen, but you should also say what frequency the decibel peaks are at. You don't have to of course, but it would help mobile users.
Even numbers of fan blades this time. Eight on the GPT fan was weird to behold.
Please, make a highlight episode with previous fan favourites, like fanfan-fan. Please and thank you! I need more fanfans in my life!
Holy cow, James, how did the Offset Rotor go over TWELVE thousand RPM?!
Your thumbnail broke my brain with the Jacko Jackson reference.
Big FAN of your videos. Just one thing and this is not a criticism or anything like that to anyone. The hold my beer has an unfair advantage because the rotor is larger than others. again its possible that this is how it looks in the video, and again no criticism to Stephan or any one.
At some point the rig will leak so much that the pressure won't increase no matter what the fan can produce
That is my AI experience too. I have gotten some funny stuff, but if you want anything really specific. No, No, not going to happen.
As somebody who runs a local copy of Stable Diffusion, it's heavily influenced by the model and prompt you use. Some models are just completely unable to do specifics, while others are so specific you can't get anything else.
5:25 - That helicopter self-identifies as an AH-6.
Wow. The return of the Beluga. Quite unexpected
BELUGA
love these showdowns! I need to upgrade my rig sooo bad it makes me sad
You're right, I'm swimming over right now
As an Aussie I’ve been coming down under for quite some time
The ChatGPT fan would have done VERY GOOD in the old format of testing.
I came for the fan content, but stayed for the chopper content! 🚁
I can't wait to see what happens on this next episode of man versus fan 😂
Another way to look at it is Newtons 3rd law. For every action there is an equil and opposite reaction. meaning that fans can only ever push back what they are pushing against. If it leaks, the fans will never show the full pressure they can produce.
Time to make the setup more perfect and retest all of the fans. lol
its hot asf in Australian rn, so it's evolutionarily advantageous to design good fans
The V-6 pump for a carpet cleaning job when I was a teen could pull over 40 feet of Hg. We'd joke about turning ourselves inside out with it.
You talkin' about mm of h²o is very funny.
Thanks for choosing to test "Hold My Beer" !
Could you do 1 vs 2 from the same creator. Properly sealed so we could really see how good they are. Also would the 2 work better with a bigger opening?
so the static pressure rig needs a minor overhaul with some built in seals then? at least for next season
I tried your prompt idea in nightcafe using flux... it got the helicopter right every time, only had the forest fire in one of them, and only had the guns firing in one of them, but overall better than what you showed here.