The spacer you added between the blower and the window cover to compensate for the thickness of the gasket creates space for air to bypass the blower, so you're losing performance there. You need a tight gap axially between the front and back faces of the blowers and the case, not just radially around the outer edge.
+1 for this. The first thing i thought of when he added the spacers. This needs to be corected and tested again. You definatly loose some pressure there.
He is right we do need a screw type (most commonly associated with Whipple) As they are completely different from the centrifugal "procharger" brand superchargers
I wonder how this charger, the other one and the old E-Turbo compare to each other with a high-end brushless motor. (He already did a "procharger", hence it's the "old one") I really want to see the pressure capabilities of printed chargers
You should make this its own sub-series for 'Fan-Showdown' videos. You can call it 'Fan-gineering' where the whole premise is to take failed attempts of previous competitors and re-engineering their fan design to create a 'best case scenario' if the concept of their idea could actually work. Kinda like a Mythbusters episode, taking the concept of the myth with the premise to prove or disprove their validity by replicating the intended results either by natural chance or by excessively manipulating the circumstantial factors to make it happen.
I was thinking the same thing no holds bared with and upper rpm limit for the air moving compressing part say 5 to 10k rpm max speed but everything else is fair game even multiple motors for multiple stages of compression.
It would be interesting for the 'fan-gineering' to also be community driven. Just release the stl files to the community to redesign/modify the base concept. Anything to far off from the original design gets disqualified.
btw, sewing machine oil is friggin amazing for reducing friction(for the open bearings) with just a couple drops and isn't effected by high heat. edit: amazing design and surprised it worked so well. thought it would have trouble with the back pressure and start fighting the motor.
I use Lew's fishing reel oil...It is incredibly thin and works great for bearings and other stuff that isn't going to be spinning for long periods of time...
Glad to see a fellow sewing machine oil believer. It is my go-to generic lubrication. Unless the equipment needs very specific lube then sewing machine oil most likely will do it. It is made for sliding and spinning things and finely meshing gears. It also has good wetting properties, it stays on as a thin film for a LONG time. Doesn't have any particles, like for ex PTFE lubes have. It is wonderful stuff. It was in my childhood always there, my grandpa repaired sewing machines so it kind of became the "family lube", no puns for fucks sake intended.. Another that works well and even works well with sewing machine oil is.. good old vaseline. You can create the film with vaseline and then thin it with SM oil (no puns..) to get really good stay with good viscosity, But also: if the application needs special lube.. don't use SM oil.... i've learned that too, it is not good for absolutely everything, like high heat: it ain't good at that.. Things need to stay below 150C, which isn't that big of a limitation.
@@ghomerhust it doesn't really need to handle any load, so even skate oil will end up too viscus. For this application you really want something as thin as possible and only if you're going to run it long term.
This is a really cool concept, the lobes could have been printed in primary colors so when they spin the middle combines them into their complimentary colors.
6:09 This is why next time, maybe try using skateboard bearings. They're designed specifically for speed and smooth operation. Also, they're 608 bearings so they have very common measurements.
The limits that the fan showdown is constrained by is what makes this series so interesting. That being said, it wouldn't be less interesting if the motor limit was removed or at least modified by allowing something a bit more torquey.
I agree with this. Then it could have 1 lobe set powered from the back and the other from the front. This would make the only use of the gears to be keeping the timing.
Was going to say the exact same thing. Allows for two bearing for each lobe, geardrive would still suck some power but as both lobes would be powered it be 1/4 - 1/2 (?) as much
As a thorough engine head, I can tell you the roots pushes air around the outside of the case (why the smoke show was unexciting) and was actually originally designed to ventilate mines and blow air into forges. Also I unironically want this on my computer but I'd definitely have to put some kind of temp switch on it for the noise, so it only kicks on when I absolutely need it. Could I possibly get the STL files?
@@theelmonk Roots blowers are loud, even if you quiet all the mechanical noise, just due to the way they move air. In a lot of applications, like truck mounted carpet extraction equipment, they use a silencer to make the noise tolerable.
11:39 ooohhh baby... that fade in of the most chill/epic music in the whole TH-cam universe is what I come for every time. Turutututuuuuu tutuuuutu turututuruuuu!!! 🤣🤣🤣
Very cool! If nothing else, you demonstrated a lot of good tricks for improving performance and making it easier to assemble. I thought the smoke test still looked great, even if you couldn't see the smoke inside the chamber. Seeing it flow cleanly into the top and slowly building as it spread out the bottom was almost cinematic. Good work!
Straightening vanes are great. They need to have a slight curve into the flow to receive the air initially... they should immediately start turning that air in the desired direction greatly reducing eddy currents that will disrupt airflow if you only make them straight. I worked for a fan design company and this was an addition we put on inline axial fans to increase their static pressure ability.
Yep, they do make a big difference to efficiency and static pressure. They tend to add a lot of noise though, I guess that's why we don't usually see them on PC fans!
I think i speak for all of us that we want to see you build a PC utilizing a bunch of these crazy fans! Would be cool to get some case temps and stuff for the video and obviously flow smoke through it.
So.... next season is water cooling pumps? Or wait! Let's go full circle with this You water cooled an air cooler Now it's time to pump air through a water block
I think it'd be cool if you attached a low-speed motor to that and just had it slowly twirling in the background of videos. Maybe just a little pwm controllable one with just enough torque to kick it off and then cap it at a lower speed once it's going. Could put some RGB in there too. Really gamer-ify it up.
Regardless of the actual fan performance, that thing is just hypnotizing to watch run. If there were a way to make it quieter and be even better, but as it stands, it’s still absolutely incredible to watch.
I bet if you stuck some like, Lucas red N tacky grease on the lobes you'd get even more performance and a lot less noise. Just kinda seal it up a bit by 'simulating' apex strips.
I wonder if adding a thin flexible "spline" along the flats and the lobe ends would increase the static pressure even more. As another commenter said, finer teeth on the gears would help decrease the noise level, or even going to a belt drive between the two yellow gears.
Belt drive ftw. And tighten up the tolerances inside to eliminate gaps for improved performance. And the smooth delivery of power through quiet belts.. Yes...
So really theoretically you could print a 1 inch by 1 inch (or smaller) roots blower and a gear reduction on the a12x25 and it would make pressure very slowly (provided no leaks
I know high viscosity micropumps work exactly like this. They are just two rotating gears inside a small capsule wich you connect to the plumbing. I think they are generally used for constantly, slowly pushing grease into the port of big machines to reduce maintenance work.
This was amazing!! Most of the machinery I have been around for work are both positive displacement and centrifugal pumps. As others have said, we are getting into the supercharger and turbo fans/pumps of the show. I should build a fan that is like the turbine engines we use on aircraft carriers. One, day!!!! Caffeinator, out.
I hope that during the next seasons of the showdown we get an 'Assembly' category or Open Division separate from the hub/frame based designs from what I feel was the showdown's original intent. Not denigrating people's crazy constructions, I think it's WILD! But I'd like to see what an end user could realistically install inside their machine in lieu of a factory fan.
This design could be a lot more efficient, if you reference a modern Eaton unit for the rotors. The main changes that were made since the 1940s when this type of blower was in use in aircraft engines is this: They went from 2 wanes to 3, and they added a rotational twist to the wanes. So using 3 bladed rotors that have a say 60 degree twist in the rotor blades would give better performance. The roots blower was designed to move lots of air, but not for generating lots of pressure - this is why they are called blowers. They were originally used in things like mine ventilation equipment where the volume of air moved was more important, than the pressure generated.
It is, I wonder if you put sound dampening foam in the corners and around the edges of the case if that would reduce the noise levels. Perhaps a needle bearing would be quieter too.
Lobe pumps are used a lot in food processing, I work in an ice cream plant and all of the ice cream freezers use them, so do the revel pumps and our fudge pump. Those suckers can pop a quarter inch thick hose with the pressure they can push. Granted your version could use some revisions to tighten the gap between each rotor and between the rotors and the sides of the enclosure.
Graphite on bearings. Angle cut gears would run way quieter, transfer load more efficiently. Perhaps a belt drive? That would eliminate the gear chatter altogether. Side clearance matters. There was no doubt it would work... Like you said, it's a supercharger. That was AWESOME! Thanks for sharing!
Awesome concept. There are a few ways to cut down on noise, too. One would be to use helical gears with a cover over them and a small bit of oil. A performance boosted would be felt gasket material for apex seals on the lobes. Or UHMW plastic and a coating of the Silicone-Teflon spray.
I made a wind powered spinning cap for an air vent 4 years ago, it has two 10*30*9 bearings with all the grease removed and replaced with thin coat of gun oil. Its still spinning like new after four years with zero maintenance. Ball bearings really don't need much lubrication if the loads are minimal.
If you tighten up your rotor end clearances to the cover/case (you could do this by omitting the gaskets all together and using a thin layer of silicone to seal the cover), you could easily double or even quadruple that number. Positive displacement superchargers are capable of massive pressures if you keep tight clearances.
Smoke test wasn't a miss, it was spot on, you could see it being sucked into the blower at the top and being blasted out the bottom. That is bad ass, shame about the noise though.
I'd be curious to see how much CFM this little roots blower puts out. Kind of a tangent, but a few air raid sirens back in the day actually used a roots-style blower to force air into a small fan to make it much louder than what would be otherwise. Many have made 3D printed models of these sirens, albiet with a Shopvac or a centrifugal blower due to complexity. Despite it being super loud, I'm sure this video could act as a basis for more people to try a roots blower at a small scale.
would be cool to see a gearing that ramps up the gear ratio more, and some kind of gear/motor enclosure to hold everything and hopefully the noise in. Interested to see how far this can be taken with some more tweaks.
I know straight cut gears are easier to find. But if you can print slant cut gears, they'll be much quieter. And some kind of graphite, moly or Teflon lube in tiny amounts will help it spin freely. My guess is the pressure died off because of gear mesh issues. Also, an option to seal the front better is to make the front out of the acrylic sheet and then mount your bearings to the outside using tapered screws from the inside to hold them. Then the whole sealing surface is flat, uninterrupted and within a much tighter tolerance to the lobes. Way cool results though! Maybe I can supercharge my lawnmower...
That thing looks similar in nature to the persitaltic pumps we use in lab and medical equipment, except with a bigger displacement and without the hoses. The issue with these, aside from the hoses which need replacement and sometimes need to be speciality models coated with platinum, is that if you don't use tubes to contain the gas or fluid you pump through it, they need to be super low tolerance to not leak.
If you were ever going to use bearings for any other projects I would suggest the Bones speedcream 1 to 2 drops might help the Performance over longer duration tests
Consider using timing belts instead of gears for the gearbox, to reduce friction and noise by a lot. The Voron 2.4 Z-axis gearbox gives a 4:1 (or 1:4) ratio, uses high-speed bearings throughout, and can easily be "borrowed" for this design.
Adding some lithium grease onto the pla gears can make them both more efficient and a ton quieter. I did this for a water pump contest in one of my freshman engineering classes and it got us something like 40% more water from the same current draw on the motor
It put out 50% more pressure than the leader. Wonder if scaling it back smaller, you could use the fan motor to drive it. Hollow or very little in fill on the rotors making them as light as possible.
I recommend a bit of machine oil (sewing machine, 0w, etc..) on the bearings, then a bit of ultralight lithium grease on the lobes (Tetra gun grease would work well and it's slick and clean to apply). You could mix hexagonal boron nitride HBN with either or both for a film that would remain even as oil moved off or evaporated. You could enclose the gear drive assembly (make it a window of course) and set it up to where there was a small oil bath pit in the bottom that allows part of the lobe mounted gears to "splash" the oil across the other components in the same way many lawn mowers and small 4 stroke engines do. I think that would quiet it down some, would tighten up the lobe tolerances a bit, and maybe even add a couple hundred rpm to the speed.
I had a watercooled setup with an automotive radiator and fan in a shrouded box, that was very loud, but the radiator was so big it could run completely passive with two GPUs in the loop.
As some as stated about lubrication on the gears, a looser mesh would lower noise but quicker wear. Also try making gears not straight cut as they are loudest kind.
That 'gasket' is killing whatever performance this design could have. It creates a gap between the lobes and the side wall/ window that shouldn't be there for this design to work properly.
Nice. As for the noise, what about herringbone gear teeth? I cant see well, but putting a small chamfer on the leading edge of the gear teeth could reduce some of the friction in them as well, and a little smoother may be a lot quieter.
Seeing this, well air-compressor you could call it I guess, creates mostly pressure, perhaps it could be used to move air via Bernoulli effect, i.e. shoot output air from a nozzle through a duct and get air flowing that way? A bit like these "bladeless" fans. That may give nicer visuals for the smoke test as well.
One thing i would like to see is a quick comparison between 120 and 140 fans, particularly how much more air or static pressure you can get from the wider fans.
you could probably eke another couple mmH2O out of that design by tightening up the gap between the rotors, as well as the gaps between the rotors and housing. Maybe look at doing something similar to the wipers used in commercial designs, but instead of having full contact (as getting the surface finish on PLA to the standard that requires is not happening...) you have a T slot at the apex of each lobe withe the bottom of the T pointing straight out from the center of the rotor. This way, the same way you can slot bolts into the channels in extrusions, you can use a TPU wiper printed so it almost makes contact. since the wiper would be removeable, you can just swap it for a longer or shorter one to check different gaps. Alternatively, you can try using other materials like paper or cardstock which may work for a full contact design. of course, that all depends on if you want to optimize the design.
It seems like your gasket and spacers are creating a larger gap between the rotors and the window. I would be interested to see if you could get a higher pressure with them removed if it will run.
so. several improvements, some of the comments, use graphite or sewing machine oil to lube the barrings, use ring gaskets on the shafts to prevent bypass, and use helical gears to reduce gear noise. (as well as using a different material to avoid the 3d print rub sound)
If we're dealing with positive displacement, might be able to beat it by copying a slip-vane rotor style setup such as a Judson. Might have to be liberal with Teflon coating to get it to work however, given the motor expected to be used has next to nil for torque.
Now what would be nice is if you did a scale model of the A12X25 to fit that motor. That way, we can see if the design is even worth pursuing. It definitely produces more pressure, but it's got a stronger motor. Does it produce more pressure than a comparable fan could? So we either need to see a design like this one that works on the A12X25 motor - maybe just a smaller scale of everything - or a larger scale fan-style comparable to what the Noctua people could do if they were using an engine with that kind of torque
I wonder just how much of a difference a better, longer meshing between the gears would have on the sound. I mean these are "straight cut" gears so naturally they would be loud. I would imagine using belts where possible would make it much, much quiter and reduce drivetrain losses
Since you kind of already go outside the realms of the norm... if you could belt drive instead of gear drive, it would be so much quieter and possibly provide even smoother flow. I really like seeing all the different fan designs.. I didn't even think this was going to perform at all, but it blew my expectations away. (Pun intended) Thanks, James! These are fun videos and we learn along the way!
if you used two fan motors mechanicly linked I wonder if it would still work, have one on the 'front' the other on the 'back' each powering a different... spinny thing, whatever they are called, while still being linked via a meshed gear acting as a syncroniser.
Funny. I was in the same place about 6 months ago. Same motors. Same power supplies. Design in fusion. 3D printed (and desktop CNC’d!). In my case, it was a very cool knife sharpener… I feel like I’m watching myself lol
Not surprised that a fluid pump has higher static pressure than a fan. It is a shame that it can't really work with a standard fan motor, though. Perhaps a scaled down version would work?
I'd be interested to see what you could do with the A12x25 motor, just geared down until it had enough torque to spin the lobes. Might only be 500 or 250 RPM love speed, but it's be interesting
Now how do we make it quiet and use it to super over pressure a PC case! for a one of a kind cooling expereince :D The supercharger set up that way on the top front of PC case would be amazing.
Hallo, I really love your videos. I'm no ingeneer, but maybe, with all those compressor fan designs, there's time to add output temperature readings to your showdowns... 😅 Maybe intercooler for turbofans 😂
In the automotive world people are usually told Superchargers are inefficient & draw engine power when turbos do not, but in reality not only do they create boost throughout the entire RPM range they also require good planning to be working in an efficient boost range for a specific engine. Most supercharged engines people make are just some V8 they slap an old Detroit Diesel supercharger on without much thought into the drive ratio; those superchargers aren't meant to create ANY boost at all as they are just meant to make a 2-stroke cylinder flow so the start of the build is already no good but fun for Hot-Rodding! When designed correctly and for the right engine even the most basic Supercharger, a 2-love Roots which you have here, can create better performance & even efficiency than a Turbo and Centrifugal Superchargers are just the cold end of a Turbo pretty much so they have the same efficiency or better (depending on rotor design) but tap power from a different source. I know this isn't super related to the video but I just wanted to let people know a bit more about a device most think is just antiquated & inefficient, even your test shows they are capable & you're using pretty much the worst type, a short 2-lobe Roots.
Looks really good! Personally I'm not too impressed by the performance, I'm also not surprised that you got it to work, with a much more powerful motor. Positive displacement pumps has a huge advantage over fans, when it comes to static pressure, even at the same amount of power. But again, it looks really good, and it's definitely great for showing the working principle of a roots blower. If you can get it reasonably quiet, even if that requires running it at very low rpm, it should be great for decoration.
I'm sure you get this suggestion all the time (and I have no idea how much they cost), but have you considered getting a slow motion camera for the smoke test shots?
You know what would be full brutal? Using two individual fan motors and trying to electronically synchronize them. Can you imagine the amount of precise position sensors and microcontroller programming required?
AFAIK, a pump like that isn't meant to move a lot of air, but to create pressure. As for other commenters suggesting that more be done with bigger motors: where do you stop?
The spacer you added between the blower and the window cover to compensate for the thickness of the gasket creates space for air to bypass the blower, so you're losing performance there. You need a tight gap axially between the front and back faces of the blowers and the case, not just radially around the outer edge.
Bump
yep, you need the front and back of the rotors to be super close to the edge
+1 for this. The first thing i thought of when he added the spacers. This needs to be corected and tested again. You definatly loose some pressure there.
@@cannesahs I did write it all out again and then I saw the comment was already there. 😅
Look at y'all! Bunch of smart peeps in here!
Great suggestion.
"this one's heavier" :D Now that you made a roots style supercharger you have to make a screw type and procharger. It's for science.
I think he's done a few centrifugal style fans
He is right we do need a screw type (most commonly associated with Whipple) As they are completely different from the centrifugal "procharger" brand superchargers
I wonder how this charger, the other one and the old E-Turbo compare to each other with a high-end brushless motor.
(He already did a "procharger", hence it's the "old one")
I really want to see the pressure capabilities of printed chargers
@@madmod he has, but not with that same motor.
He did the "turbo" supercharger (procharger) already
Put graphite on the gears...less noisey and more spinny.
Could also sprinkle some graphite powder into the chamber so surfaces inside are lubricated too
dry teflon spray.
graphite is nasty going everywhere on electronics.
Graphite is harder than the gears, it's not the right lube for plastics.
Also frequently overlooked is tooth size, smaller is quieter.
@@lasskinn474 which he already used 6:35
You should make this its own sub-series for 'Fan-Showdown' videos. You can call it 'Fan-gineering' where the whole premise is to take failed attempts of previous competitors and re-engineering their fan design to create a 'best case scenario' if the concept of their idea could actually work. Kinda like a Mythbusters episode, taking the concept of the myth with the premise to prove or disprove their validity by replicating the intended results either by natural chance or by excessively manipulating the circumstantial factors to make it happen.
I second this
Third this! This would be awesome!
I was thinking the same thing no holds bared with and upper rpm limit for the air moving compressing part say 5 to 10k rpm max speed but everything else is fair game even multiple motors for multiple stages of compression.
Love the name
It would be interesting for the 'fan-gineering' to also be community driven. Just release the stl files to the community to redesign/modify the base concept. Anything to far off from the original design gets disqualified.
With positive displacement pumps, static pressure depends mostly on how airtight the pump is.
btw, sewing machine oil is friggin amazing for reducing friction(for the open bearings) with just a couple drops and isn't effected by high heat.
edit: amazing design and surprised it worked so well. thought it would have trouble with the back pressure and start fighting the motor.
grease the lobes too with it, it doesnt have to last very long
@david -- maybe print the lobes out of a harder TPU verrrry slightly oversized? But the friction losses may make that a bad idea.
I use Lew's fishing reel oil...It is incredibly thin and works great for bearings and other stuff that isn't going to be spinning for long periods of time...
Glad to see a fellow sewing machine oil believer. It is my go-to generic lubrication. Unless the equipment needs very specific lube then sewing machine oil most likely will do it. It is made for sliding and spinning things and finely meshing gears. It also has good wetting properties, it stays on as a thin film for a LONG time. Doesn't have any particles, like for ex PTFE lubes have. It is wonderful stuff. It was in my childhood always there, my grandpa repaired sewing machines so it kind of became the "family lube", no puns for fucks sake intended.. Another that works well and even works well with sewing machine oil is.. good old vaseline. You can create the film with vaseline and then thin it with SM oil (no puns..) to get really good stay with good viscosity,
But also: if the application needs special lube.. don't use SM oil.... i've learned that too, it is not good for absolutely everything, like high heat: it ain't good at that.. Things need to stay below 150C, which isn't that big of a limitation.
This.
I think removing grease from bearings is fine, but maybe look at using ceramic oil used for skate bearings instead :)
thats what i was thinking, like a gun oil or skate oil type of thing. thinner but still handles the loads and speeds
@@ghomerhust it doesn't really need to handle any load, so even skate oil will end up too viscus. For this application you really want something as thin as possible and only if you're going to run it long term.
Use use break cleaner and remove all of it. They will spin better and longer. They will wear out faster but this won't be used often.
Fully ceramic bearings with no grease would be even better
@@ratgreenthat's more like it, ceramic balls and races.
This is a really cool concept, the lobes could have been printed in primary colors so when they spin the middle combines them into their complimentary colors.
Great suggestion
6:09 This is why next time, maybe try using skateboard bearings. They're designed specifically for speed and smooth operation. Also, they're 608 bearings so they have very common measurements.
And they are cheap. Even the crappiest ones still would handle the load of this blower with ease.
The limits that the fan showdown is constrained by is what makes this series so interesting. That being said, it wouldn't be less interesting if the motor limit was removed or at least modified by allowing something a bit more torquey.
I agree except in this case, 2 of them would be good, while not within the limits of a single A12x25, at least it is a known amount.
Maybe the fan showdown could be expanded to allow dual motors / fans? Awesome print btw, hope it goes on display.
I agree with this. Then it could have 1 lobe set powered from the back and the other from the front. This would make the only use of the gears to be keeping the timing.
Fully agree, then you can have two leader boards for people to work towards!
Vote up! But within another category haha
Was going to say the exact same thing. Allows for two bearing for each lobe, geardrive would still suck some power but as both lobes would be powered it be 1/4 - 1/2 (?) as much
Season with multiple hub or a larger motor would be amazing
I reall love fan showdown.
Let us all appreciate for a moment that a wonky fan vid got over 31k views in just 14h.
Keep up the good work!
As a thorough engine head, I can tell you the roots pushes air around the outside of the case (why the smoke show was unexciting) and was actually originally designed to ventilate mines and blow air into forges.
Also I unironically want this on my computer but I'd definitely have to put some kind of temp switch on it for the noise, so it only kicks on when I absolutely need it.
Could I possibly get the STL files?
Maybe use rubber belts instead of gears to reduce the noise ?
@@theelmonk Roots blowers are loud, even if you quiet all the mechanical noise, just due to the way they move air. In a lot of applications, like truck mounted carpet extraction equipment, they use a silencer to make the noise tolerable.
11:39 ooohhh baby... that fade in of the most chill/epic music in the whole TH-cam universe is what I come for every time. Turutututuuuuu tutuuuutu turututuruuuu!!! 🤣🤣🤣
Very cool! If nothing else, you demonstrated a lot of good tricks for improving performance and making it easier to assemble. I thought the smoke test still looked great, even if you couldn't see the smoke inside the chamber. Seeing it flow cleanly into the top and slowly building as it spread out the bottom was almost cinematic. Good work!
Straightening vanes are great. They need to have a slight curve into the flow to receive the air initially... they should immediately start turning that air in the desired direction greatly reducing eddy currents that will disrupt airflow if you only make them straight. I worked for a fan design company and this was an addition we put on inline axial fans to increase their static pressure ability.
Yep, they do make a big difference to efficiency and static pressure. They tend to add a lot of noise though, I guess that's why we don't usually see them on PC fans!
I like the idea of the next season being maybe a silent motor and then no limitations fan designs!
You sound more and more like Mythbuster: busted but what would be needed for it to work. Awesome man! And glad to see you back!
I think i speak for all of us that we want to see you build a PC utilizing a bunch of these crazy fans!
Would be cool to get some case temps and stuff for the video and obviously flow smoke through it.
So.... next season is water cooling pumps?
Or wait! Let's go full circle with this
You water cooled an air cooler
Now it's time to pump air through a water block
If you pump strong enough you can even heat up the system!)
Make fan shaped pumps. Dunno about the torque, would probably need a lot.
Oh, and water cool the radiator - FROM THE OUTSIDE
Just hook up a 3/4 ton water source heat pump and install a ground water loop in your backyard
YOU TOOK MY IDEA!!! lol jk
That was AWESOME!!! THANK YOU for doing that build.. it was WAY COOLER than I expected!!! GREAT IDEA JAG!!!
The giggle when it jumped to 13 the first time got me, I LOL’d as well.
maybe try oiling the gears bearings / axels. might make it a bit more quiet. the wait was worth it imo btw.
I think it'd be cool if you attached a low-speed motor to that and just had it slowly twirling in the background of videos. Maybe just a little pwm controllable one with just enough torque to kick it off and then cap it at a lower speed once it's going. Could put some RGB in there too. Really gamer-ify it up.
Roots blowers by roots brothers were originally air pumps for forges, supercharging was derivative of those
8:47 - That edit cut was epic! :D
Regardless of the actual fan performance, that thing is just hypnotizing to watch run. If there were a way to make it quieter and be even better, but as it stands, it’s still absolutely incredible to watch.
I want him to belt drive them.. that would make them so quiet and run smoothly.
ya DONT USE STRAIT CUT GEARS lol.
I bet if you stuck some like, Lucas red N tacky grease on the lobes you'd get even more performance and a lot less noise. Just kinda seal it up a bit by 'simulating' apex strips.
I must admit. I did not expect anything near that. Not at all, actually. Very impressed!
2:48 how it feels to turn on your PC at night when your parents are sleeping
The rabbit hole you went down for this was DEEEP 🤣🤣🤣
Great job. I love the channel
So this is like the 6 million dollar man version. We can rebuild it! We can make it better! 😆
I wonder if adding a thin flexible "spline" along the flats and the lobe ends would increase the static pressure even more. As another commenter said, finer teeth on the gears would help decrease the noise level, or even going to a belt drive between the two yellow gears.
Belt drive ftw. And tighten up the tolerances inside to eliminate gaps for improved performance. And the smooth delivery of power through quiet belts..
Yes...
So really theoretically you could print a 1 inch by 1 inch (or smaller) roots blower and a gear reduction on the a12x25 and it would make pressure very slowly (provided no leaks
I know high viscosity micropumps work exactly like this. They are just two rotating gears inside a small capsule wich you connect to the plumbing. I think they are generally used for constantly, slowly pushing grease into the port of big machines to reduce maintenance work.
This was amazing!! Most of the machinery I have been around for work are both positive displacement and centrifugal pumps. As others have said, we are getting into the supercharger and turbo fans/pumps of the show.
I should build a fan that is like the turbine engines we use on aircraft carriers. One, day!!!!
Caffeinator, out.
You're back!! This was a fun video.
I hope that during the next seasons of the showdown we get an 'Assembly' category or Open Division separate from the hub/frame based designs from what I feel was the showdown's original intent. Not denigrating people's crazy constructions, I think it's WILD! But I'd like to see what an end user could realistically install inside their machine in lieu of a factory fan.
You been watching the channel doing the boat propellers?
there is channel doing that?
@@ThunderClawShocktrix rctestflight.
This design could be a lot more efficient, if you reference a modern Eaton unit for the rotors. The main changes that were made since the 1940s when this type of blower was in use in aircraft engines is this: They went from 2 wanes to 3, and they added a rotational twist to the wanes. So using 3 bladed rotors that have a say 60 degree twist in the rotor blades would give better performance. The roots blower was designed to move lots of air, but not for generating lots of pressure - this is why they are called blowers. They were originally used in things like mine ventilation equipment where the volume of air moved was more important, than the pressure generated.
It is, I wonder if you put sound dampening foam in the corners and around the edges of the case if that would reduce the noise levels. Perhaps a needle bearing would be quieter too.
Lobe pumps are used a lot in food processing, I work in an ice cream plant and all of the ice cream freezers use them, so do the revel pumps and our fudge pump. Those suckers can pop a quarter inch thick hose with the pressure they can push. Granted your version could use some revisions to tighten the gap between each rotor and between the rotors and the sides of the enclosure.
Graphite on bearings. Angle cut gears would run way quieter, transfer load more efficiently. Perhaps a belt drive? That would eliminate the gear chatter altogether. Side clearance matters. There was no doubt it would work... Like you said, it's a supercharger. That was AWESOME! Thanks for sharing!
That window joke paired with the synth song made me chuckle 😂 now every time im watching through a window that song will ple playing in my head 😂
Awesome concept. There are a few ways to cut down on noise, too. One would be to use helical gears with a cover over them and a small bit of oil. A performance boosted would be felt gasket material for apex seals on the lobes. Or UHMW plastic and a coating of the Silicone-Teflon spray.
I've seen several performance and noise improvement suggestions. I would love to see all of these.
I made a wind powered spinning cap for an air vent 4 years ago, it has two 10*30*9 bearings with all the grease removed and replaced with thin coat of gun oil. Its still spinning like new after four years with zero maintenance. Ball bearings really don't need much lubrication if the loads are minimal.
Not bad, given the massively wide gaps involved. I'm not sure how long it would last, but it is darn fun.
If you tighten up your rotor end clearances to the cover/case (you could do this by omitting the gaskets all together and using a thin layer of silicone to seal the cover), you could easily double or even quadruple that number. Positive displacement superchargers are capable of massive pressures if you keep tight clearances.
So glad you are back, I was craving more fan content.
Smoke test wasn't a miss, it was spot on, you could see it being sucked into the blower at the top and being blasted out the bottom. That is bad ass, shame about the noise though.
I'd be curious to see how much CFM this little roots blower puts out. Kind of a tangent, but a few air raid sirens back in the day actually used a roots-style blower to force air into a small fan to make it much louder than what would be otherwise. Many have made 3D printed models of these sirens, albiet with a Shopvac or a centrifugal blower due to complexity. Despite it being super loud, I'm sure this video could act as a basis for more people to try a roots blower at a small scale.
That was far better performing than I expected.
would be cool to see a gearing that ramps up the gear ratio more, and some kind of gear/motor enclosure to hold everything and hopefully the noise in. Interested to see how far this can be taken with some more tweaks.
I know straight cut gears are easier to find. But if you can print slant cut gears, they'll be much quieter. And some kind of graphite, moly or Teflon lube in tiny amounts will help it spin freely. My guess is the pressure died off because of gear mesh issues.
Also, an option to seal the front better is to make the front out of the acrylic sheet and then mount your bearings to the outside using tapered screws from the inside to hold them. Then the whole sealing surface is flat, uninterrupted and within a much tighter tolerance to the lobes.
Way cool results though! Maybe I can supercharge my lawnmower...
That thing looks similar in nature to the persitaltic pumps we use in lab and medical equipment, except with a bigger displacement and without the hoses.
The issue with these, aside from the hoses which need replacement and sometimes need to be speciality models coated with platinum, is that if you don't use tubes to contain the gas or fluid you pump through it, they need to be super low tolerance to not leak.
That looks sick! Perfect colors for visuals as well
Wow. Just wow. 😅
I'm glad I'm not the only one that goes a little nuts on projects sometimes ❤ this was super entertaining and worth it dude
If you were ever going to use bearings for any other projects I would suggest the Bones speedcream 1 to 2 drops might help the Performance over longer duration tests
Consider using timing belts instead of gears for the gearbox, to reduce friction and noise by a lot. The Voron 2.4 Z-axis gearbox gives a 4:1 (or 1:4) ratio, uses high-speed bearings throughout, and can easily be "borrowed" for this design.
Adding some lithium grease onto the pla gears can make them both more efficient and a ton quieter. I did this for a water pump contest in one of my freshman engineering classes and it got us something like 40% more water from the same current draw on the motor
It put out 50% more pressure than the leader.
Wonder if scaling it back smaller, you could use the fan motor to drive it.
Hollow or very little in fill on the rotors making them as light as possible.
I recommend a bit of machine oil (sewing machine, 0w, etc..) on the bearings, then a bit of ultralight lithium grease on the lobes (Tetra gun grease would work well and it's slick and clean to apply). You could mix hexagonal boron nitride HBN with either or both for a film that would remain even as oil moved off or evaporated. You could enclose the gear drive assembly (make it a window of course) and set it up to where there was a small oil bath pit in the bottom that allows part of the lobe mounted gears to "splash" the oil across the other components in the same way many lawn mowers and small 4 stroke engines do. I think that would quiet it down some, would tighten up the lobe tolerances a bit, and maybe even add a couple hundred rpm to the speed.
I had a watercooled setup with an automotive radiator and fan in a shrouded box, that was very loud, but the radiator was so big it could run completely passive with two GPUs in the loop.
As some as stated about lubrication on the gears, a looser mesh would lower noise but quicker wear. Also try making gears not straight cut as they are loudest kind.
As others have mentioned, a light oil would work wonders on the bearings, but helical gears will be a lot quieter.
I love it. I'd love to see a separate bracket for these kinds of designs.
That 'gasket' is killing whatever performance this design could have.
It creates a gap between the lobes and the side wall/ window that shouldn't be there for this design to work properly.
Nice. As for the noise, what about herringbone gear teeth? I cant see well, but putting a small chamfer on the leading edge of the gear teeth could reduce some of the friction in them as well, and a little smoother may be a lot quieter.
Seeing this, well air-compressor you could call it I guess, creates mostly pressure, perhaps it could be used to move air via Bernoulli effect, i.e. shoot output air from a nozzle through a duct and get air flowing that way? A bit like these "bladeless" fans. That may give nicer visuals for the smoke test as well.
One thing i would like to see is a quick comparison between 120 and 140 fans, particularly how much more air or static pressure you can get from the wider fans.
you could probably eke another couple mmH2O out of that design by tightening up the gap between the rotors, as well as the gaps between the rotors and housing. Maybe look at doing something similar to the wipers used in commercial designs, but instead of having full contact (as getting the surface finish on PLA to the standard that requires is not happening...) you have a T slot at the apex of each lobe withe the bottom of the T pointing straight out from the center of the rotor. This way, the same way you can slot bolts into the channels in extrusions, you can use a TPU wiper printed so it almost makes contact.
since the wiper would be removeable, you can just swap it for a longer or shorter one to check different gaps. Alternatively, you can try using other materials like paper or cardstock which may work for a full contact design. of course, that all depends on if you want to optimize the design.
It seems like your gasket and spacers are creating a larger gap between the rotors and the window. I would be interested to see if you could get a higher pressure with them removed if it will run.
so. several improvements, some of the comments, use graphite or sewing machine oil to lube the barrings, use ring gaskets on the shafts to prevent bypass, and use helical gears to reduce gear noise. (as well as using a different material to avoid the 3d print rub sound)
If we're dealing with positive displacement, might be able to beat it by copying a slip-vane rotor style setup such as a Judson. Might have to be liberal with Teflon coating to get it to work however, given the motor expected to be used has next to nil for torque.
Nice 'out of the box' idea - I will keep watching to see were it goes. Lots of resonance with automotive designs.
Now what would be nice is if you did a scale model of the A12X25 to fit that motor. That way, we can see if the design is even worth pursuing. It definitely produces more pressure, but it's got a stronger motor. Does it produce more pressure than a comparable fan could?
So we either need to see a design like this one that works on the A12X25 motor - maybe just a smaller scale of everything - or a larger scale fan-style comparable to what the Noctua people could do if they were using an engine with that kind of torque
Glad you put a thin oil on the bearings, as they turn without any lubricant, but they will get louder, and increase friction as it heats up
I wonder just how much of a difference a better, longer meshing between the gears would have on the sound. I mean these are "straight cut" gears so naturally they would be loud. I would imagine using belts where possible would make it much, much quiter and reduce drivetrain losses
But how would your version do WITH the A12X25
Would be interesting to see you redo the list sorted by performance per watt and find out who comes out on top given the same input power.
Since you kind of already go outside the realms of the norm... if you could belt drive instead of gear drive, it would be so much quieter and possibly provide even smoother flow.
I really like seeing all the different fan designs.. I didn't even think this was going to perform at all, but it blew my expectations away. (Pun intended)
Thanks, James! These are fun videos and we learn along the way!
@Major Hardware, have you tried making a slowmotion video? Perhaps it would show the smoke moving inside beter.
if you used two fan motors mechanicly linked I wonder if it would still work, have one on the 'front' the other on the 'back' each powering a different... spinny thing, whatever they are called, while still being linked via a meshed gear acting as a syncroniser.
Funny. I was in the same place about 6 months ago. Same motors. Same power supplies. Design in fusion. 3D printed (and desktop CNC’d!). In my case, it was a very cool knife sharpener… I feel like I’m watching myself lol
Have you ever considered applying graphite paint/paste (sometimes called slipplate) to the parts where they slide against each other?
Not surprised that a fluid pump has higher static pressure than a fan. It is a shame that it can't really work with a standard fan motor, though. Perhaps a scaled down version would work?
to make it better you can either scale the size make the sides seal better. it's much simpler to scale it in 'height'.
I'd be interested to see what you could do with the A12x25 motor, just geared down until it had enough torque to spin the lobes. Might only be 500 or 250 RPM love speed, but it's be interesting
between the gears and the body you could put some type 2 plastic (like a milk jug), or at least a washer of... to make it spin smoother.
Now how do we make it quiet and use it to super over pressure a PC case! for a one of a kind cooling expereince :D The supercharger set up that way on the top front of PC case would be amazing.
This was a fun rabbit hole.
I wonder if printing the lobes out of LW-PLA (Colorfabb, e-sun etc.) would help?
Hallo, I really love your videos. I'm no ingeneer, but maybe, with all those compressor fan designs, there's time to add output temperature readings to your showdowns... 😅 Maybe intercooler for turbofans 😂
In the automotive world people are usually told Superchargers are inefficient & draw engine power when turbos do not, but in reality not only do they create boost throughout the entire RPM range they also require good planning to be working in an efficient boost range for a specific engine. Most supercharged engines people make are just some V8 they slap an old Detroit Diesel supercharger on without much thought into the drive ratio; those superchargers aren't meant to create ANY boost at all as they are just meant to make a 2-stroke cylinder flow so the start of the build is already no good but fun for Hot-Rodding!
When designed correctly and for the right engine even the most basic Supercharger, a 2-love Roots which you have here, can create better performance & even efficiency than a Turbo and Centrifugal Superchargers are just the cold end of a Turbo pretty much so they have the same efficiency or better (depending on rotor design) but tap power from a different source.
I know this isn't super related to the video but I just wanted to let people know a bit more about a device most think is just antiquated & inefficient, even your test shows they are capable & you're using pretty much the worst type, a short 2-lobe Roots.
If you want it to be more quite you will have to replace the gears for belts. Or redesign the mesh on the gears to help reduce lashing noises.
Looks really good! Personally I'm not too impressed by the performance, I'm also not surprised that you got it to work, with a much more powerful motor. Positive displacement pumps has a huge advantage over fans, when it comes to static pressure, even at the same amount of power. But again, it looks really good, and it's definitely great for showing the working principle of a roots blower. If you can get it reasonably quiet, even if that requires running it at very low rpm, it should be great for decoration.
I love your videos. They are always entertaining.
I'm sure you get this suggestion all the time (and I have no idea how much they cost), but have you considered getting a slow motion camera for the smoke test shots?
You know what would be full brutal?
Using two individual fan motors and trying to electronically synchronize them.
Can you imagine the amount of precise position sensors and microcontroller programming required?
I wonder if reducing the diameter of the lobes while increasing the length would help in the torque department.
AFAIK, a pump like that isn't meant to move a lot of air, but to create pressure. As for other commenters suggesting that more be done with bigger motors: where do you stop?