3D Printer Fans - AXIAL vs BLOWER (MUST KNOW)

Get $5 OFF for your first order at the PCBWay - bit.ly/PCBWay5off
👉🏻 Today I am testing axial versus blower fan performance for a 3d printer's print cooling system. After watching this video you will know the definitive answer with real data as to what fan to use in this specific scenario.
🧡 Support my work at PATREON - / diyperspective
📋 RELATED ITEMS TO THE VIDEO (Affiliate)
Fans:
- 5015 blower amzn.to/3Ppy4m8
- 4020 axial amzn.to/3uNJoAY
3D printers in the video:
- Kingroon KP3S amzn.to/3yDvHWe
- Longer LK5 Pro amzn.to/3chMvud
🕗 Timestamps:
00:00 - The problem
01:03 - Testing airspeed & noise
04:49 - Support
🔗 YOU CAN FOLLOW ME:
Twitter: / diyperspective
Instagram: / diyperspective
Instructables: www.instructables.com/member/...
#3dprinter #test #fans

[@802Garage]

I'd say the main reason for all the designs using axial fans is that they repurpose stock fans that came with printers.

[@Jakob3xD]

Did you watch the vid? Axial fans are optimised for high flow and radials for static pressure.....

[@802Garage]

@@Jakob3xD I'm not sure if you misunderstood my comment or something. I'm saying the reason fans are used in non-optimal ways in many designs is because people who don't know are designing around fans that came with printers. Just a guess.

[@Jakob3xD]

​@@802Garage oh nvm then

[@802Garage]

@@Jakob3xD No worries.

[@seebuch93]

been printing for almost 4 years now as a hobby and have done almost every upgrade to my ender 3 i could find. never considered this at all but incredible information! i love the rigorous testing.

[@jimmym2719]

You are right, many just use tiny shrouds for axial fan. Even using radial fan I use the same mm2 surface area of this outlet when lofting 3D shroud in cad and smoothed it’s air flow path. Thanks for sharing 💕

[@fireheadpet2039]

I just discovered your channel and subscribed right away. Great approach, science and empirical, and concise. Very well done and amazingly informative.

[@PrintingPerspective]

Welcome aboard! I really appreciate the comment, these are the fuel and motivation for enthusiast content creators. :)

[@Nilbyte]

Thanks for the information. It's living and learning! I never thought about the differences between radial and axial fans for cooling in the 3d printer.

[@PrintingPerspective]

Yeah. I bet blower fans would be even better when cooling hotend's heatsink, I will have to test that.

[@Schuessel]

Nice Video, thanks! I also loved the design of the TUSH-Style Spool Holder - can you please give me a Tip where to get the Files for this special one?

[@jackle3002]

Would have been good to know the electrical comparisons of the fans. Voltage but mainly rated current. The amperage & RPM greatly effects the flow rate.

[@lanzer22]

For argument’s sake I would say that for this application the size and airflow are the two main aspects to consider. The third is usually noise which is mostly governed by the RPM. There is almost no scenario where wattage is an issue in an FDM printer setup.

[@Sevalecan]

Based on your nozzle pictures and the cheap E3d V6 nozzles I got off aliexpress, and the fact that I am getting lots of buildup on the nozzle, I should try polishing them.

[@silent_woolf.]

Wow, new video :) That's always like a little holiday :)

[@PrintingPerspective]

Hope you enjoyed it!

[@coxyofnewp]

jUST RECENTLY GOT INTO 3D PRINTING, and just this vid well worth a like and sub.. Can't wait to see your other vids..

[@alainthire]

Just found out videos and they are GREAT!!!!!!

[@stefanguiton]

Great video!

[@uss-dh7909]

This is an extremely interesting find to me. Just last week I picked up an old enterprise nvidia tesla gpu for ai image creation. It's passively cooled because normally these are stuck into racks with a lot of fans forcing air through them. Got a fairly big case so I thought I'd just duct one of my 140 front mounted fans with a cobbled together cardboard shape to guide the airflow and I'd be off to the races. Large fan pushing air through a small volume ought to create high enough pressure to push air across the entire tesla card right? Even a fairly slow rpm should be enough to keep things manageable so no noise... Or at least that's been my thought process since I pulled the trigger on this gpu last week. Been precipitating for the past few days so the cardboard bins contents would be soaked and I don't want to tear apart the box my new static pressure fans came in should I want to return them. Going to get that duct made, but I'll absolutely be keeping these blower fans in mind. Never thought to use one since my laptops fan and it's just... meh...

[@chrismay2298]

Using the radials allows you to turn down the RPMs to cut down on noise. Even if you don't need dual 5015s, running two turned down is quieter and is offers room if you DO need to crank em! 👍

[@Karaon]

i just recently came up with the idea of using old laptop fan and this i just what i needed to see. off to design 360 degree airflow "duct" i guess. I want to make it cool from all directions simultaneously

[@austinvickymoore3656]

I have a repurposed radial fan from an all in one pc that doesn't work anymore and want to use that for my printers part cooling but I think it's a bit big and I'll have to make my own custom design so it's going to be a while before I have something mocked up

[@CediKayable]

I did not know this is still a thing in 3D printing. Back in 2018 with Sovol entering the market, the german community was changing to radial fans with amazing fan ducts.

[@paul_wiggin]

What about combining high performance axial fan with high pressure radial fan? First one will provide lots of air with minor pressure, second will make use of the big flow with little pressure.

[@robertohernandez1642]

Link to the filament roller with adjustable width? Plz and Thank you!

[@THotSouce]

I just want to know, do you need to use a termoresistant material or pla is K?

[@racing_mntage1584]

Which type is better for task whoch requires more suction power through small vents?

[@StumblingBumblingIdiot]

If you go back and forth with the first 2 sounds you let us hear, it will sound like an old style dial up modem :) Nice look at the fans. Would be neat to see your take on what ducts actually work better than the stock setup on most printers. I found that most of them do worse than the factory setup on my Ender 5 probably due to the print speed not getting high enough to make the massive airflow worthwhile...at least that is my theory...

[@PrintingPerspective]

Yeah, I have a test like that in mind, but before that, I need to test multiple-size blower fans - 5015, 4020, and 4010 to determine which ones are the most optimal choice considering cooling power (how good they force the air though ducts) and the noise. One-side blower fans like on Ender 5 or anything similar is pretty effective, especially if you orientate harder spots in the print (like overhangs) towards the airflow from the fan. The only problem is if you want equal cooling from all sides. This is when all the testing needs to be done comparing different ducts and how they affect print cooling, which is also on my "want to make" videos list.

[@StumblingBumblingIdiot]

@@PrintingPerspective Excellent! Yea it is so hard to test fans because of the ducts being sized differently on almost every model so it changes the airflow between them. I am looking forward to more videos as you got my subscription ;)

[@PrintingPerspective]

That is why I will just make same size ducts for all the fan types. :) This is the easy part :D

[@shadeofsound23]

Honestly, it's because they never thought of it. Axial fans are the ones that they're most familiar with, they're light, and thin, and easily mounted directly onto a printhead platform without significant changes to power and weight load requirements.

[@andreyl2705]

nice work)

[@wilkimhernandez]

What cooling end are you using? 4:18

[@Kinoko314]

The always-on fan on my Ender 3 was noisy and driving me crazy, so I had to get a buck converter to lower the voltage going to that fan.

[@silent_woolf.]

And yeah, I finally bought all parts to make my own "THE BEST DRILL GUIDE with CNC parts". :))) Another few months and it's gonna be done :D

[@PrintingPerspective]

Nice, it is better late than never :)

[@phiranaguy]

What is the paper referenced at 2:50?

[@pettre6016]

I see what you're saying but I like using 4 of my 40x40x56mm counter-rotating fans at 23000 rpm for part cooling.

[@PrintingPerspective]

Nice :D Do you need to clamp the printer to prevent a liftoff? HA HA!

[@percival5771]

my guy started a bridge test... its still bridging and hasnt reached its end yet.... its been a month. what are we at? couple kilometers? :D

[@PrintingPerspective]

Here are some more data from my testing: 4020 axial fan to 30mm hole = 6.3m/s airspeed (no restrictions) 4020 axial fan to small duct to 30mm hole = 2.6m/s (small duct restriction) 5015 blower fan to 30mm hole = 4.1m/s airspeed (no restrictions) 5015 blower fan to small duct to 30mm hole = 3.6m/s airspeed (small duct restriction) 59% vs 12% airspeed loss when comparing 4020-axial vs 5015-radial.

[@FratDede]

Are they in same wattage? We can find many different energy level fans at that patterns so we should compare with same or similar watt levels.

[@Tigrha]

@DIYPERSPECTIVE Hi, do you by any chance have some data that would compare 4010 / 4020 axial with 4010 or 4015 radial fans? i´m not interested in adding bigger fan just if it would be viable to consider change of 4015 radial with 4010 or 4020 axial .. cause i think to compare 5015 blower with 40mm axial was destined to be loss for the smaller fans but the end performance difference of same size fans might bring bit different results (like i can´t open link to the 4020 axial, but when i tried to find similar design in offers that are on amazon now it seems its CFM might be around 8 while noctua 4010 has CFN 8,9 (4020 9,4 so no that big increase over double size - but pressure is better with 4020 and noise 5dB lower) .. as far as i was able to check CFM of 4015 blower might be around 3-3,5 and 5015 you compared with that axial fan is probably about double the CFM of 4015 ...

[@matthewszn7257]

can someone please give me the file for the orange one

[@darkshadowsx5949]

I can confirm noctua although a quiet fan, it sucks at cooling prints sufficiently even with the offset duct design.

[@mikoaj2323]

I was always wondering, isn't it more silent just because it's not that strong? If you decrease power on regular fan, it will also be quieter.

[@morbus5726]

i use an axial 4028 server fan for my printer. Spins up faster, has more airflow, and almost as much static pressure as a 7530 flower.

[@PrintingPerspective]

When I was buying branded blower fans and checking their datasheets for CFM and static pressure, 7530 blowers had lower static pressure and higher airflow than 5015 fans. 0.380 in H2O vs 0.880 in H2O as for example. 4028 server fans had from 0.880 to 1.325 in H2O... but that is achieved at 13k-16k RPM. So if you don't mind the noise and you optimized the duct for axial fan design in CAD simulations then maybe it could be on par with the 5015 fan, but I am not fully sure.

[@dtibor5903]

Use a Delta server fan, it will blow your print off the bed plate :D bonus: sounds like a jet engine

[@morbus5726]

@@dtibor5903 i use a sanyo for the quicker ramp up times, but delta fans are great too!

[@jamonserrano2906]

How do you control the fan rpm? Tried and it starts after 80%. I played with klipper settings with no success

[@AM-pi7jy]

What you're doing at 2:09 doesn't look like the way you should use an anemometer?

[@PrintingPerspective]

Why not?

[@scaletownmodels]

That is interesting information. Best way to find out the real life effects of something is to just test it in reality. Now I have to rethink my design.

[@Enjoymentboy]

I wish fans provided a cfm per watt rating.

[@PCBWay]

Welcome B-A-C-K! ✌✌

[@kilianlindlbauer8277]

You can also step up the 4020 axial fan to a 4028 axial fan. Even 70x35 radial fans look pathetic compared to good 4028 axial fans. For example ebm papst sells a 4028 fan which is still fairly quiet at rated 50dB but has over 4 times the air flow and over twice the static air pressure of the most powerful controllable sunon or gdstime 5015 radial fan. 38m3/h vs 9/9,1m3/h amd 520 to 550 (graph was awful to read on smartphone) Pa pressure vs 250 Pa on the radial fans. So in every case the axial fan will perform better as at every pressure value it has higher flow and at each flow value higher pressure. The downside of the axial fan is the hefty price of 55 euros and the sometimes difficult packaging

[@PrintingPerspective]

I honestly don't understand why so many people talking about the 4028 axial fans. They are insanely loud because they spin at ~15k RPM to get that high pressure and airflow. 50dB for a fan is a lot and when you add it into a duct the noise will increase more. Also, you can't compare noise levels from manufacturers' datasheets to my testing results, the testing methodology is very different. The whole point of this video is to show that regular 4020 fans perform considerably worse at similar noise levels compared to 5015 fans. The only scenario where you would consider a 4028 fan is for high-speed printing because you actually need the performance and the noise levels become irrelevant. Well, this is my opinion about all this.

[@kilianlindlbauer8277]

@@PrintingPerspective most will be fine with 5015 radial fans like you said and they are easier to package. But after a certain point of print speed you will need such cooling systems like some fat 4028 fan, 7035 radial (horror to package), berd air or similar air pump setups. Beyond the 300 to 350 mm/s mark (let it be travel or infill, doesn't matter) your steppers will sound hella ugly to the point where spread cycle is quieter/less annoying tone. At this point cooling fans will be the least noisy thing. Im aware that the rated noise levels are under most optimal circumstances with just the fan and no ducting behind it, real life noise levels will be probably more in the 60 to 65 dB range. The truth is that if you want powerful part cooling its getting loud, thats just physics. Air pressure always generates noise as sound is basically a wave of different air pressure. Some fans are just more efficient at it than others (quality vs knockoff fans). Some are quiet because they have no power (noctua, pc fans in general). The quietest part cooling solution that provides really good cooling suitable for print speeds beyond 300mm/s is remotely mounted fans/bed blasters as you can use large, high volume fans that dont generate much noise combined with radial fans on the toolhead. Bed blasters are only suitable for medium to small printers as the effectiveness decreases with distance and you will have to add more of them to cover a wide area of the bed. 15k rpm is at the medium to lower end of th 4028 fans, at least if we talk without ducts, the fastest ones even reaching 25k rpm without load. One extreme example is 38k rpm and 31 watts of power, 2300 Pa static pressure and 63m3/h with insane 71 dB rated sound

[@s.sradon9782]

That's really loud, atr that point just put your printer in a foam box.

[@ajingolk7716]

A difinet yes from me but the video is short😢

[@marckart66]

I upgraded my ender 3 v2 with 2 5015 fans. They blew so much air that bridging was difficult. For the speeds an ender 3 v2 can acheive, a single 5015 fan will suffice. Lighter toolhead is more advantageous on a ender style printer than 2 5015 fans.

[@PrintingPerspective]

A more accurate indicator for how much cooling you need is cubic millimeters per second. The more plastic you extrude the more thermal mass it will have and the more cooling it will require. Speed doesn't say a full story. And if you have too much cooling you can turn down fans to 50%, you know that right? It is a good "problem" to have. Plus running two fans at a lower speed also reduces the noise at the same airflow. You should check the video about that if you didn't see it - kzbin.info/www/bejne/iZLWoqeYgJ2an6c

[@NexGen-3D]

3rd Option, Berd-Air

[@testtesting5088]

You are creating restriction with the venturi type duct you test. If so you should test something like the 4028 fan which surpass most fans in these sizes. How ever it is not that simpel. You would want a certain air speed at the throat. And that should be what defines the ducts. Not the looks or what ever fits. Look into bernoullis princip and formula for pressure drop. Would give you more knowledge. Normally it is first or second semester knowledge on the engineering study 🤟

[@PrintingPerspective]

This video idea is based on people using axial 4010 or 4020 fans to reduce the noise of the cooling fan or fans. There are so many 3D printer "upgrade" designs that perform horribly. This is the whole point of the video, to show that those regular fans (and not 4028 server fans that run at ~15k RPM and sound like a jet engine) are a bad idea. If you want to make a video about the best-performing ducts with server fans explaining all the principles and showing real-world application scenarios, be my guest. :) I personally, like giving practical testing examples as those are the most useful to the majority, you know.

[@testtesting5088]

@@PrintingPerspective You misunderstand med. Your testsetup is what is not correct. The hourglass type you are using is creating a venturi effect meaning loss. E.g why i wrote look into Bernoullis principal. The video you made are very clear and mostly correct in all parts. But people need to see that it is air cooling and not pressure cooling. So flooding with a lot of air is better than point a blowgun at only one spot if you want to cool a surface ;) It was not meant as anything else than an input for you to look into.

[@PrintingPerspective]

​@@testtesting5088 I just showed how bad the hourglass duct can be but I tested with a modified version that represents other designs we can find online. But what if the surface is just like a tiny cylinder of 5mm in diameter? Then we need more concentrated airflow at higher speeds to exchange heat in that small area more effectively. If we are printing a big object then yes I completely agree with you, but a lot of 3D printing consists of printing smaller objects. So I am not sure what you are trying to say. Cheers!

[@desnou]

​@@PrintingPerspective But you didn't dig deeper than simple hourglass/off center designs and concluded that using axial fans is a bad idea altogether. Read more about why exactly axial fans are bad at creating static pressure and why there are fixed blades in axial compressors (you do know, that axial compressors exist, do you?). Single Noctua 4020 with fixed blades duct is good enough cooling for printing at 75mm/s with .5 nozzle. With two of those you can print at 120mm/s with 0.6 HF nozzles, no problems. And all of this at 37-40dB, just above background noise levels. At those speeds servos and even silent drivers are more noisy than my cooling setup. Most cooling designs on the web suffer from the same fundamental flaws - multiple 90 degrees bends in the duct, asymmetrical division of airflow, overly long ducts etc. Using radial 5015s is simpler, but axial 4020s, when used properly, are totally appropriate for typical hobby printing applications.

[@sierraecho884]

It is incompetence. Looks at most printer frames. They copy the popular Prusa i3 frame design, which is inferior to a XY cube for instance. Why ? Just because they have always does that and it has worked so why change it ?

[@LoosiuFlying]

Hi! In terms of fans and fan ducts testing there is a giant on YT he's name is Alex : kzbin.info . One can learn a lot form both authors.

[@PrintingPerspective]

He definitely has some good videos that everyone can learn something new but with all due respect, the channel is kinda a mess because it is just like a blog. But that is only my opinion and everyone is entitled to have a different one. :)

[@LoosiuFlying]

@@PrintingPerspective that is 100% true. I wish more ordered content. That's why I'm here.

[@victoryjoe01]

Is mine Give it back

[@agustinsida8877]

just finished printing my extruder with axial fans........fuck

[@PrintingPerspective]

Look at the bright side, at least you are not one of those who bought a setup like that for $30 from AliExpress :) Back in the day, I printed one too and quickly realized that it was waste of time and filament.

[@maxkool0076]

Been saying this for years, but this stupid designs like the brissfang exist and solve no problems. Stacking fans doesn’t do anything in this application…..

[@mickeysimmons6310]

your "review" is bullshit. you never tested the fans with no restriction. you tested one fan against the other without knowing if they moved the same amount of air with no restriction. You didn't post the specs of the fans are they the same rpm and wattage? A higher wattage axial will have more static pressure. the blade pitch on the axial fans can drastically change the air pressure produced. I replaced my blower with a 4020 noctua and it pushes out so much air that i generally can't run it at 100. at 100% it will push the filament out sideways when loading filament. the stock 5015 can't come even close. i have to dial it back to 60% and then its a third the noise of the stock 5015.even at 100% its still half the noise. You obviously don't have a clue what your doing. You need to take into account the specs static pressure, rpm, air flow, stock noise and wattage of the fan you are selecting. Not just compare two rando fans to each other. there is a reason people choose noctua. they move alot of air at a good static pressure with less noise. go watch major hardware videos on the fan show down

[@PrintingPerspective]

"Push the filament sideways" now that's some factual data :) Literally every fan does it. You come here all swinging with swords and this is the main data you bring? :D I am sorry to say but your noctua 4020 fan has only static pressure of 2.26mm H2O and 5.5 CFM. The 5015 blower style have ~14mm H2O and 4 CFM, while running at same speed of 5k RPM. And this info are from datasheets. So when you do actual testing that provides you real data you will see how much you really know about all this :)

[@sergeb7945]

the only reason your axial fan is not generating air flow is your bad duct design. Google for "axial fan guide vanes" or "vaneaxial fan" this is THE proper way to design a duct for axial fan. Quite surprised you didn't find this during the preparation, investigations (if any) and making of this video, as this is really the basics.

[@webslinger2011]

Great video!