i like your take on using the curves. i was using 8 layer pcb to make the windings and using through holes to get the magnetic flux lines short enough. www.ata.org.au/wp-content/uploads/2018/12/marand_high_efficiency_motor.pdf
Oh, you didn’t mention it but reluctant motors (induction) motors require slip speed between the field and the steel to create the eddy current and hence the back emf. Your magnet plate was matching the speed of the rotor so no slip speed and no drive.
I'm probably not the first person to say it - but we learn more from our mistakes than from our successes. Your next design will be better - and the process repeats. Well done for completing the project - which is a major achievement, regardless of how you perceive the outcome. Thanks for the videos. Stay safe out there.
so true... my projects never even get to the stage that looks that pretty. maybe id finish more if i made them look nice and complete, even if prototype
yeah I spend more time trying to prove myself wrong than just about anything else and I've actually become really good at proving myself wrong over time.
Bro didn't even touch the commercially produced switch reluctance type axial flux. Bro went straight for the holy grail design and made a *Permanent magnet (or not), Axial Flux, Synchronous Motor!!!* Only thing left is to sandwitch the coil winding, by magnets. And make a yokeless (coreless/no iron core) design. And proper sophisticated cooling.. And you'll have the best that can ever be made! In all honesty I think I have an/few ideas to improve the thermals and coincidentally energy density even further. But we'll cross the bridge when we'll reach it.
I really like your involute stator design. The reasoning makes sense and seems like a good way forward. I would like to point out however that every BLDC motor, permanent magnet or reluctance, is a synchronous motor. They all rely on the stator magnetic field rotating at the same speed as the rotor magnetic field. The stator field is always advanced in relation to the rotor field, as if they were aligned the torque would be zero. As you well know, on asynchronous motors the stator magnetic field always have a higher speed than the rotor magnetic field, and the slip is what produces torque. Commercial motors are called "Synchronous reluctance motor" or "permanent magnet synchronous motor" just to differentiate them from asynchronous (inductance) motors. In reality, every PM BLDC motor is a PMSM. If that was not the case, Field Oriented Control would not work. But regardless of nomenclatures, please keep up with your ideas!
Thanks for pointing out the definition mistake! Yes, it's a little bit confusing. I have added this point to the pinned comment along with the explanation of what I considered as synchronous in the video instead.
Basically, the amount of attention to details editing this presentation alone has had deserved my subscription.... Thank you for sharing your process in such detail. As it has greatly helped me along mine (Albeit mine is not really a motor per se). I have grown curious of the rest of your videos... This one was definitely a treat. Thank you
Just a word of appreciation of your experiments with motors. It is interesting and hoping you'll get nice efficient motor one day out of these experiments
Loved this video! I'm quite understudied on all of this but watching you process through this was really exciting and I love the design you created. Thank you for including your mistakes too, they offer a ton of encouragement toward pushing forward despite the innumerable and infuriating setbacks that are bound to occur in one's projects. My penchant for mistakes is seemingly unrivaled. Learned so much. ^^
Please dont stop, dont dumb things down. Even if it seems like dumbing it down would make it seem like you'd get more views in the short run, leaving it as is will build you a fanbase of people that like the complexity and believe me, you dont see alot of videos on youtube that go into this much detail. You're fulfilling an unmet need that so many people like myself have. You could leave the more complicated titles (so people like me can find videos like this easier and more directly) but put a dumbed down thumbnail with a more appealing title (still in the thumbnail) to also draw in people who may like the content but got scared by the title like I almost did. The best of both worlds! Keep doing what your doing and doing what you enjoy even if it isnt what youre already doing. Youre doing great, I hope you keep doing great, and I luv u. Kisses.
Hey Birdbrain, make a note of this... the thinner the iron core, the less eddy currents you will have to deal with. Eddy currents will cause you to lose efficiency and will produce more heat as a result. Overall, you've made an AMAZING build for an axial flux synchronous motor. Keep up the great work!
That is true, yep! My goal wasn't to minimize eddy currents this time... but for the future I have a stator design in mind that should be great at minimizing eddy currents.
Also, the eddy currents can be used to move something like an aluminum wheel. Just as an example to illustrate the point, if one were to rotate a set of magnets arranged in alternating north/south alignments near enough to an aluminum plate/flywheel mounted on an axle that plate will rotate and there will be less overall heat in the original system from the eddy currents. An eddy current is basically an opposing force in the opposite direction and can be used to do work. I've been looking at this as a sort of capture system to capture the extra force generated when a magnetic gear slips due to too much input torque. Here is a video demostrating the basic priciple - ua-cam.com/video/wYNMpRUAuyY/v-deo.html
Nice video! I think it'd be possible to replace the iron parts with iron powder-infused resin. That way, you could just pour the iron resin where it needs to go instead of cutting and bending steel. You'd also minimize eddy currents since the resin is non-conductive. You could also take it further and replace the resin with metal-filled 3d printer filament, although it'd be more expensive and wrecks brass nozzles.
The magnetic filaments for 3d printers are very poor magnetic materials - the steel laminations I made are far superior in terms of magnetic permeability. Any powder-based core is going to have poor magnetic permeability in general. In practice, as far as I am aware, powder-based magnetic cores are used in high-frequency inductors and inductors where you want a very clean release of energy... basically in places where you want to minimize eddy currents but don't care too much about the other magnetic characteristics of the core material. I already have an idea for how to make the stators better in the future.
@@BirdbrainEngineerno eddy currents, yes, but more importantly in hf applications you NEED low hysteresis (which metallic iron and steel lacks), ferrite cores are a necessity in hf applications, the eddy currents are more of a bonus.
Ah yes, hysteresis losses was another reason for ferrite cores... couldn't think of it at the time. But yeah, I'm absolutely not well versed in high frequency applications, there's a whole load of new stuff I'd have to learn first.
@@BHARGAV_GAJJAR PCB stators are not suitable for high load applications due to fairly poor thermal characteristics and low current density compared to wire windings or better yet, hairpin windings.
Just discovered you through the recommendation gods known only as "The Algorithm" and frankly I'm annoyed I hadn't found you sooner. This is exactly the type of stuff I love to see and your designs, prototyping, presentation and editing is all amazing. I hope more people are able to stumble across your works, not just because you deserve bigger recognition but also because it's genuinely great content that people would enjoy if they knew how to find it. I look forward to seeing how this progresses, but for now please excuse me while I go watch your entire backlog ᵃˡˢᵒ ᶦ ˡᶦᵏᵉ ʸᵒᵘʳ ᵖᵘʳᵖˡᵉ ⁿᵃᶦˡ ᵖᵒˡᶦˢʰ :)
epic video. I think it's great that even if things go south with the project, you can still make a video about it and somehow it's even more interesting than a success story.
I am certainly no expert and really learned a lot from the clip. Thanks! One of the issues I thought happened about 3:18, 3:28, 3:37. All these pictures show cores with metal wrapped with a wire strand. Each core is designed to fit next to another using a ratio of winding/steel core. No spaces are left in between to maximize magnetism. When you built yours, it seemed the winding/steel-core ratio was way different, and the plastic housing took up 50% of the space. Next time, take the design at 3:18 and measure the weight/mass of one steel plate. When you build your curved plate, ensure you have the same amount of steel per core. Then when making your wheel, build a 2mm curved slot for the steel sections to sit down into. Remove all the plastic barriers. I would recommend making a smaller easier-to-build motor with fewer cores. Failure is normal, make it easy to build and learn, then build again. After you get it working well, then make a bigger, more core model. Anyway, my ideas...
Very nice! I've been making a motor for a pc cooling pump, and landed in about the same axial configuration. I used permalloy sheet cut into iron slots, but it worked without core as well. It took me about half a year to build a working pump. Now I am building a blower turbine for air cooling, rotor printed already, but had no inspiration in developing it further. Now I might have it, thanks to you.
Awesome design! I haven't seen involute coils like that before, but it looks great. Much easier to DIY than radial flux motors. Especially if you use thinner steel sheet so can stack them up with glue and clamp between two cauls to create the shape rather than having to individually press bend them. And make non-stick spacers that half-fill the slots so you just pack as much wire in as you can, hit with CA, and then pull the spacers, leaving perfectly sized slots for the neighboring coils.
I learned so much in 15 minutes thanks to your years of effort exploring your curiosity. I have finally built my electric bicycle and it is fun passing people on the highway while going 45-50mph (70-80kph) I have my motorcycle license for this amount of power. I just wish I had a motor controller that had regenerative braking, not necessarily for recovering energy but primarily for stopping. My bike with its 3000 watt hub motor and its 72volt 40amphour battery weighs in at an earthquake generating 101lbs (46kg) and that with my 200lbs (91kg) traveling at those speeds the brakes for bicycles are being consumed like warm butter on a hot sidewalk... They don't make breaks for bicycles that reliably dissipate that amount of kinetic energy.
If you put resistors on the motor terminals it will brake, the lower the resistance the more it will brake but do not forget that this can heat up the motor
This take me back when I was on college, I did as well a generator that uses eolic energy as input to the shaft, in that time I made the parts using a laser cutter and wood and bought a 3D printer to make it better, but the wood prototype was enough, I graduate from school, and 2 years later I finally work again on this project with cleaner parts from the 3D printer, I'm not an electrical engineer but definitely found fun in doing this type of projects, great video, good day to you.
This is nuts that this is able to work, I just wanna share my admiration. I'm just sitting here knitting (very slowly) and impressed by how much people can make with 3D printers
I can't believe I'm seeing a dude on UA-cam design his own axial flux motor. Im happy I'm living in a time of great technological change. This decade is going to be a wild ride and it's going to be exciting unless the nukes start flying.
It would, the laminations were made simply because it was easier to construct the parts that way. But I do have a lead on how to construct the stator next time to reduce eddy currents by a lot.
Awesome! I stumbled across this type of motor for the first time today, and this video did a great job of explaining the guts. Hearing this style of motor was lighter, it got me thinking about 3D printer extruders, and how the fastest printers tend to have the lightest hotends. It got me wondering how small folks have made them. While I didn't find a NEMA17 sized motor, this was still awesome to see. Thanks for sharing. ☺️ PS: 💅♥️
that loop/curve is also present in a lot of celtic knot art thanks for making this its quite interesting its similar to a idea i have had for a telsa turbine shaped ionic continual pulse engine
Your magnetic flux paths is through or normal to the pancake surface, so it will make eddy currents in the plane of the windings creating the magnetic flux, offset by gap. So the thin sheet, or not so thin sheet, is thin in the wrong dimension for breaking up eddy current areas into smaller, less lossy ones. Giant toroid cores can be cut into sections that are approximately the shape you have. Like a 1/3 of a toroid. The powder core does not need orientation like soft steel layers do. If the material can be had, and a way to fire the shape, you can make your own custom core shapes. Liked this video, shows curious not afraid of failing, just want to learn mode. :)
This is awesome . . . the achilles heel to the Axial flux motor IS the dependance of large mafnetic requirements . . . . there just isn't enough know deposits of approriate rare earth to powe the potential need of this motor type . . . . . a great achievement. The next step is to find ways to increase the efficiency as reluctance motors are always less efficient.
Distributed vs concentrated coils depends on stator/pole ratios. It shouldn't be decided based on other goals. It's simply what needs to be a full pitch span of coil.
Wow. Awesome project really, the way you set up the coils and ferromegnetic look very nice but I bet they were a pain in the a... to set up. I am actually trying to build a wind turbine generator and came up with faulhabercoils in the stator and a double rotor (on top and underneath). Those faulhabercoil windings are very nice, because you don't come up with lots of wire outside the magnetic fields, the downside is, that the magnetic fields do not cross the wire in 90° angle. This is (maybe) the cause why you have to 'kickstart' your motor. Love your work!
Excellent work. Very novel. Since you're using a lot of 3D printed parts have you ever tried the iron composite PLA from ProtoPasta? It's not as good as using real iron, but it's far better than just plain plastic. It's about 40% iron by weight, which translates to about 8% iron by volume. Magnets will actually stick to it fairly well, but it's not magnetic itself. For generating stators, or for making iron backings for rotor magnets, it's just so convenient. All those volute shapes could just be 3D printed instead and would immediately do the job, and there is, in essence, no air-gaps between the iron in the PLA and whatever you're wanting to generate your magnetic fields.
Haha thanks! It's much appreciated and needed... UA-cam's algorithm seems to be working in a very flawed way nowadays - previous 3 videos I have uploaded have been doing awesome for *exactly* the first 21 hours after upload, and then the impressions rate suddenly gets nuked to be like 10x+ lower than it was literally minutes before, despite all of the metrics staying good and promising... I am guessing that the algorithm is currently built to literally push a video hard for only the first day as it's "fresh", and then it discards the video in favour of new "fresh" content. Pretty bullshit if you ask me, especially since this doesn't seem to be the case for large channels - their videos stay being recommended for weeks usually...
Building magnetic stators that work & work efficiently is no easy task. From taking stators apart the main rules are that the laminations have to be packed tightly, are not able to move against each other, and of course are varnished to create the electrical isolation barrier. They also need to be reasonably thin and made out of the right kind of steel.
the gap created by the print between your windings and the laminations probably causes a lot of reluctance. reluctance motors in general are lower torque and are more sensitive to air gap also. for your KV: you should do the opposite test, drive the motor at a set RPM and measure the output voltage under open circuit
:D I think, the uh, iron inserts for the rotor would be making noise since it's loose. buuuuut, I think you could make a dye press out of wood levers and 3d printed parts. also I think designing your own esc with visual feedback, as hard as it is, would help you see what's happening, and adjust everything!
I considered making a diy FOC unit... but fuck me there's a lot to it if I want to make something that actually works properly haha. Definitely would have to be a video on its own and I'm not sure if and when I'd do it.
Thank you for making a video of you fabricating a 3d printed motor. Seeing your video is inspiring me to dust off my original Radial Flux design and see it through to completion. Stay classy.
Wow! What a video, you are quite the teacher, keep on going with your amazing and very hard work my friend, I would absolutely love and appreciate new videos. I’ll make sure to share this one around :)
If you're looking to do without earth magnets, have you considered maybe a shaded pole motor design? Of course you'd need an inversion system of some sort to convert it from AC to DC, but I recently pulled one out of an old microwave, and it seemed to work well without the need for magnets. I'm new to all this, but an AC motor of some sort would seem like the way to go
14:55 so you made the world's biggest motor for a 7,200 rpm hard drive, well done. One suggestion is to find a local metal shop to cut your metal strips.
I do all the initial tests for motor max speed behind a sheet of metal and not looking at it directly. On other motors I usually find that the speed is not too bad so I ditch the protection but not with this one haha.
I bet two of those doubled up as one one motor would do well on an ebike. It would need decent torque to move one, though. I don't know too much about them, though, to really speak on it.
This design in concept, really is genius! It looks like a double Fibonacci spiral…like a sunflower. The problem is with it is twofold. First is that the magnetic flux needs to have a continuous flow path from opposite sides so that it magnifies the torque as much as possible (basically oscillating from left to right, back and forth for each ‘petal’), and the wires of the coils need to be right next to the stator blades (as close as possible without shorting out) with no plastic in between. The plastic of the mold and the air gaps are essentially turning your motor into a giant spiraling capacitor instead of making torque. The plastic and air is a dialectic, which turns the coils into coiled capacitors which are charging and discharging into each other, instead of generating torque, because the flux and the coils are not perpendicular and opposing, like in a normal axial flux motor design. One way this might work is if you could somehow print the sunflower stator and petals so that they could be one contiguous conductive core instead of conductive ‘leaves’ being held by a non-conductive/dialectic core. Then the current wire/coil configuration might work at driving the flux around continuously from left to right so that it creates torque instead of capacitance. Just a thought….but I could be wrong.
No matter the motor design, the coils are always insulated from the cores, and the cores are left floating (exactly for this reason of reducing capacitive losses), so there really is not any appreciable amount of capacitance there. But it is true that the coils are supposed to be as close to the cores as possible - unfortunately in this case it would have damaged the insulation of the coil wires, and it would have been very difficult to keep the metal strips that make up the cores stable while winding the coils, so the plastic enclosure for the cores was necessary. As for the continuous flux paths, that is exactly what is going on here.
@@BirdbrainEngineer Also the second thing, which I was scratching my head at is the coil winding pattern itself. You are using an overlapping serpentine pattern, which may or may not work at all. In the original vanilla axial flux motor you have very simple coils of copper wire around ovular or rectangular metal neodymium magnet stators. It’s simple and works. This design you made is so much more complex, and probably needlessly so. I would start with a working plane vanilla axial flux design and slowly change it by adding one idea at a time, rather than going all in at once with many unconventional ideas, and that way it becomes easier to tell if an idea works or not. But I commend you on your efforts and your ingenuity. So again nice job!
@@BirdbrainEngineer The serpentine winding pattern + large air gaps and dialectic stator = massive capacitor instead of moving stator. That’s the gist of it.
@@BirdbrainEngineer Here is how I would suggest doing a redesign using your initial basic ideas… 1) Simplify! Limit the number of petals to 6 - 12 (max) total around the entire stator. 2) Make the shape of the petals like a hysteresis curve with no sharp or pointed ends (like a real sunflower petal shape). (Sharp or thin ends will focus all of the magnetic flux to those points while killing the flux in the body, which is not what you want in a motor design.) For flux to be even you need a regular, smooth shape (like a platonic solid) with no thin or sharp ends. 3) Don’t use a serpentine winding pattern. (use regular self contained and separate coils of wire instead)…Have distinct and separate coils with no overlap. 6) Keep the coils as close to or wound on top of the stator magnets as possible without it shorting - (ie no air gaps and no dialectic material in between with the sole exception of the tiny amount of coating on the copper magnet wire itself.) 7) Definitely use permanent magnets, because it won’t work as a core-less design unless the entire thing is fabricated as continuous conductive metal. That’s it! I think if you make those changes the motor should work as intended.
its amazing how axial or rotor still 365 herts AKA 360 degrees or a circle equates to 120 which is the current degree seperation of a 3 phase power producion system today. which is 1/3. I would try using air core as in no core.
You can bias magnets into focusing their fields on one side by a halbach array, what´s not possible with coils. So, the stator has to be electromagnetic, and in the middle, in order to exploit its both sides, while the rotor is made of 2 discs with magnets, in halbach array config, embracing the stator between them. Otherwise it´s a huge waste of magnetic field, and thus, power/energy.
I'll be going for the e-bike motor and e-bike build based on the motor only some time in the first half of next year, as biking in the winter is a bit meh :P
@@BirdbrainEngineer yeah that's fair, even in the winter i still wanna start biking more though, and having an ebike i think would really help a lot with that
should use those green magnetic viewing sheets to verify fields. helps tremendously when making electromagnets :) Cool project. wish u much success ! Also, digital simulations can be misleading! More often than not, simulations merely get a person to believe in falsities.
I actually used a linear hall effect sensor as a really basic gauss meter, and the stators did generate 3 north poles and 3 south poles in alternating polarities.
So, you made an axial brushless universal motor, kind of. If i understant it correctly, the 2 winded faces have the same winding and are both static. The core is ment to position itself something like a magnetic transmission does? Soo to work, the phases of one side would need to be very slightly offset from the phase of the other side, and the magnetic field should be only directed by the laminated core. Im not sure it can work if the laminated core is parallel to the axis. I think you might need the phases to be in line and have the core try to redirect the magnetic field to produce he offset. If im right, you would need a new core but could test it kinda quickly.
Looking at your rotors, one cant really see much reluctance variation when they spin, regarding the coils. Low reluctance variation means almost no torque. A way to get stronger reluctante variation is to redesign the rotors and the wiring, to imagine having like U electromagnet and a U steel core facing each other. The orientation of the rotors laminations is paralel to the ones in the coils, which might not produce the highest reluctance variation.
The stators generate opposing magnetic poles facing the rotor. As an example a single flux path "starts" from stator1 pole, through the rotor, to the opposing pole on stator2, through the back iron of stator2 to an adjacent pole on stator2, through the next set of laminations on rotor, to the opposing adjacent pole on stator1, through the back iron of stator1 back to the "start". These generated poles rotate rather smoothly thanks to the distributed windings and so they should keep the rotor nicely in sync. The problem is that since I only managed to fit 9 turns on each stator, then there was just very little flux to begin with, and some other reasons were also said in the video.
There in Cars and Trucks GMC uses them in there Hybrid Trucks Since 2003/2004 and Many other Cars and Trucks have them and Aircraft use them these kind of Motors these are NOT any New kind of Motor.
I wonder if using electromagnets vs permanent magnets and using some sort of pulse/timing based approach to rotate that last rotor would work? What I mean is that maybe one could replace the permanent magnets in that last rotor with steel wrapped in copper and then simply feed it power when the device needs to run. One could potentially even have that rotor be stationary and alternate the charges in such a way that it is effectively as if the rotor was spinning. That might work as both a sort of torque and speed regulator, meaning if one increases/decreases the power there will be a corresponding increase/decrease in magnetic field strength and if one were to increase/decrease the speed a which they are alternating the electromagnets that would effectively be like controlling the overall speed of rotation. Also, I've been doing a lot of research into static electricity and potential methods for harvesting it to do useful work since it is fairly straightforward to generate extremely high voltage static electricity. Throughout that research I found ion thrusters and magnetohydrodynamic drives to be really interesting. An extremely oversimplified explanation of how they use the lorentz force to generate thrust is that they basically take an electric charge, hit a magnet with it at the appropriate angle(90 degrees) and that causes a force that is perpendicular to the velocity and magentic field strength and is basically the product of the velocity of the charge and magnetic field strength(it acts like a multiplier when the angle is right). My basic idea was to take advantage of the phenomena of short circuits to power such a device. When there is a short circuit there is a sudden drop in resistance which causes a corresponding spike in amps which can be framed in such a way that it is analogous to increasing the velocity of the charge, or like if one were to suddenly remove the support holding up a massive object. One could potentially create a pretty extreme amount of force when combining relatively low voltage dc power combined with sort circuits and lorentz force.
So a difference between SynRM motors and IPM/SPM motors is their Id, Iq current scalars. In an internal PM machine the Id current is always going to be negative while the Iq current will be either positive for positive torque values or negative for negative torque values. In a Synchronous Reluctance Machine the Values of both Id and Iq are positive. This could be why you kept getting a stuttering effect and no motion, the controller was braking the machine rather than accelerating it. I'm not sure what settings you had enabled in the Inverter but that could be something to look into. Also the Field weakening algorithm is different between the two motors.
yo cool bird dudette! 1:10 Now I might be flying a little to high, if you catch my drift, but could you like idk combine them or smt? Surely more means better, right?
IIRC, that would be called a "raxial flux" motor, and they do exist, and they are basically the highest power density motors in the world. Maybe one day I will try to make one.
awesome video! really good production quality, extremely informative, one of the most informative Iv seen in a while, and just ingenuity at its finest with all that math and good understanding of the motor and its physics! EDIT: also Im just curious and i doubt this comment will get seen but how do reluctance motors induce a back emf/ resistance of some sort to the input voltage. all motors of some sort have some sort of back resistance because as rpm increases output power increases and thus input voltage and current draw should also change. or if there is no back emf or anything how would a motor like this work or what shape power chart would it produce?
Increase the number of phases in the coiling and base your curve on the golden ratio then your initial design will work. You may need a controller board to switch the fields on and off precisely
I don't think the golden ratio curve matters, but more turns (than the 9 I had) would have made it more likely for the reluctance design to work, even if poorly.
About driving it, maybe working with true sine power. The Spanish youtuber Electronoobs made a video about driving bldc motors that way. About the variable reluctance motor, I might think that the steel alloy isn't magnetically soft enough, having too much histeresis renders the variable reluctance not working
Yeah the steel is not good for this purpose... One day I might try to make a properly working reluctance motor as I now have some ideas on how to make better stator- and rotor cores. Since wave driving is nearly always preferred, yeah.
What kind of calculations did you do in order to determine the design of the reluctance structures for the stator and the rotor? Have you used any electro-magnetic simulation tool like FEMM or MAXWELL? I use FEMM for a good approximation of axial flux motors / generators. It is a little bit tricky because it calculates more in 2D and I have to build the e-machine in a way to get a good approximation of the data. So I simulated a linear motor version of the build circular structure. But this is ok. The data came out in 10 to 15% approximation.
Cool project! Maybe laser cut parts could be a decent intermediate scale alternative for widely available and affordable production access besides 3D printing
The next step that would be significantly better than 3d printed parts (and thus worth going for) would be metal parts. So either metal laser cutting, plasma cutting, or (cnc) milling. None of those methods are that widely available nor affordable unless you yourself have or know someone who has such a machine. The e-bike motor I will still try to make in such a way that nearly anybody could replicate it "easily". Though at one point I do intend to build a cnc mill capable of cutting mild steels.
I randomly found this video and enjoyed it a lot! Always fun to see a maker doing something interesting and it's a bonus when the voice is as lovely as yours. ^_^
An easy way of making mediocre performance iron cores in impossible shapes is filling the mold with fine iron filings then impregnating it with very thin glue, preferably epoxy. Better if you can vacuum impregnate them.
Very cool stuff! Like you said, bot having proper bulky magnetic cores is probably the biggest problem. But besides that, I wanted to point out that i think you should be able to measure the rotor position using the existing copper coils. Using an hf signal, you should be able to see different inductances at the channels, depending on their alignment. I'm not sure if there's a good way to get an absolute position this way, but at least getting the (electrical) position of the coils relativ to the stator should be possible (and sufficient for efficient operation). If you don't feel like modulating a pieer wire, you can probably also add some extra sense coils. Using the symmetry of the design you could place them in a way that they would be mostly unaffected by the voltage induced from the poeer coils. Dang, noe I kinda want to try that...
Back-emf based control scheme is used in sensorless speed controllers, however, since the initial idea was to make a reluctance motor, which produces no back-emf, so this control scheme would have been impossible. Generally hall effect sensored position feedback is superior to back-emf based position feedback, especially at low speeds, which is why basically every synchronous motor that has any significant power output has hall sensors in them.
Not sure if I'm correct (automation engineer by trade, so I think I might be onto something, but I don't know much about synchronous motors), but doesn't the reluctance rotor also need windings? Like... Except the magnetic flux core, should it also not have a conductive "cage" in between? I know, that asynchronous industrial motors have something, that is called a "cage winding", in English I think it's called a "squirrel cage winding", which is - most of the time - made from aluminium for cost-saving measures, and it is just a bunch of aluminium rods stuffed between the metal rotor core, and then welded with two aluminium rings on the opposite sides of the rotor. To achieve the same thing, I think one would have to wind copper windings around the metal core and then solder the ends together, so it makes a continous wire loop, so the current induced has a path to flow.
What you are explaining is called an induction motor :) Induction motors are famously non-synchronous, as the rotor speed is not coupled with the rotation of the electrical poles due to what's called rotor slip. The non-synchronity of induction motors is also why they are often preferred in industry - you can usually just plug them into 3-phase grid energy and they'll work... no controller needed.
Hi, Love you video man! I've worked on some ac and BLDC motors in the past and I'm curious if you looked into your ESC at all. I've had major problems running normal motors with commercial ESC's and having the control algorithms work correctly. (Even in sensored mode which you seem to be using). Usually it involves a lot of PID tuning and tinkering. Have you tried changing your ESC when trying to run your induction motor? It also seems like the ESC is made for higher speeds which could also cause problems, I remember using some motor controllers called SOLO or something with a nice web interface and they had a different mode for low RPM motors which you probably need. (I think it has to do with the speed it runs the PID loops or something)
Yeah, the ESC tries to spin the motor pretty fast but it did sometimes try to sync up to the extremely low rotation rate (you can hear it thanks to coil whine). In addition, the ESC I used does work just fine with a high-torque low speed RC car motor.
Unfortunately pure reluctance motors can not be used as generators. However if you meat whether a proper involute design brushless motor could be used as a generator then I'd imagine so, however it might not be as good as a generator with the standard layout... would have to do some simulations.
Hello friend! I've played around with turning car alternators into bldc motors. You can see what I came up with on my channel. But I was working on a design for a helical toroidal flux bldc motor, but had to move. Link to my printables designs in my channel page. I am really loving your 3D printed designs! And your need to use distributed windings for smooth operation had me thinking of my helical toroidal flux motor.
Yes did use the one loop at the loop connections would have been way better you and running them side-by-side like that as long as they're touching might as well loop them
I hope you enjoyed the video!
You are brilliant. Maybe the design can use CRnGO laminated sheets and ferrite magnets ?
UA-cam has again deleted my subscription from you ....
i like your take on using the curves. i was using 8 layer pcb to make the windings and using through holes to get the magnetic flux lines short enough.
www.ata.org.au/wp-content/uploads/2018/12/marand_high_efficiency_motor.pdf
Oh, you didn’t mention it but reluctant motors (induction) motors require slip speed between the field and the steel to create the eddy current and hence the back emf.
Your magnet plate was matching the speed of the rotor so no slip speed and no drive.
Reluctance and induction motors are not the same type of motor.
I'm probably not the first person to say it - but we learn more from our mistakes than from our successes. Your next design will be better - and the process repeats. Well done for completing the project - which is a major achievement, regardless of how you perceive the outcome. Thanks for the videos. Stay safe out there.
Engineering is an iterative process! It also humbles one to not think that everything is easy to do.
The greatest teacher;
Failure is.
-Yoda
so true... my projects never even get to the stage that looks that pretty. maybe id finish more if i made them look nice and complete, even if prototype
stay safe out there is really good advice which i wanted to give too
yeah I spend more time trying to prove myself wrong than just about anything else and I've actually become really good at proving myself wrong over time.
when it comes to axial motors, at least on youtube - very few try a new design. great job!
Bro didn't even touch the commercially produced switch reluctance type axial flux.
Bro went straight for the holy grail design and made a *Permanent magnet (or not), Axial Flux, Synchronous Motor!!!*
Only thing left is to sandwitch the coil winding, by magnets. And make a yokeless (coreless/no iron core) design. And proper sophisticated cooling..
And you'll have the best that can ever be made!
In all honesty I think I have an/few ideas to improve the thermals and coincidentally energy density even further.
But we'll cross the bridge when we'll reach it.
I really like your involute stator design. The reasoning makes sense and seems like a good way forward.
I would like to point out however that every BLDC motor, permanent magnet or reluctance, is a synchronous motor. They all rely on the stator magnetic field rotating at the same speed as the rotor magnetic field. The stator field is always advanced in relation to the rotor field, as if they were aligned the torque would be zero.
As you well know, on asynchronous motors the stator magnetic field always have a higher speed than the rotor magnetic field, and the slip is what produces torque.
Commercial motors are called "Synchronous reluctance motor" or "permanent magnet synchronous motor" just to differentiate them from asynchronous (inductance) motors.
In reality, every PM BLDC motor is a PMSM. If that was not the case, Field Oriented Control would not work.
But regardless of nomenclatures, please keep up with your ideas!
Thanks for pointing out the definition mistake! Yes, it's a little bit confusing. I have added this point to the pinned comment along with the explanation of what I considered as synchronous in the video instead.
I'd like to thank you for a wonderful comment. My faith in an intelligent Humanity has been restored.
Basically, the amount of attention to details editing this presentation alone has had deserved my subscription.... Thank you for sharing your process in such detail. As it has greatly helped me along mine (Albeit mine is not really a motor per se).
I have grown curious of the rest of your videos... This one was definitely a treat.
Thank you
Just a word of appreciation of your experiments with motors. It is interesting and hoping you'll get nice efficient motor one day out of these experiments
Loved this video! I'm quite understudied on all of this but watching you process through this was really exciting and I love the design you created. Thank you for including your mistakes too, they offer a ton of encouragement toward pushing forward despite the innumerable and infuriating setbacks that are bound to occur in one's projects. My penchant for mistakes is seemingly unrivaled.
Learned so much. ^^
Please dont stop, dont dumb things down. Even if it seems like dumbing it down would make it seem like you'd get more views in the short run, leaving it as is will build you a fanbase of people that like the complexity and believe me, you dont see alot of videos on youtube that go into this much detail. You're fulfilling an unmet need that so many people like myself have. You could leave the more complicated titles (so people like me can find videos like this easier and more directly) but put a dumbed down thumbnail with a more appealing title (still in the thumbnail) to also draw in people who may like the content but got scared by the title like I almost did.
The best of both worlds!
Keep doing what your doing and doing what you enjoy even if it isnt what youre already doing.
Youre doing great, I hope you keep doing great, and I luv u.
Kisses.
Hey Birdbrain, make a note of this... the thinner the iron core, the less eddy currents you will have to deal with. Eddy currents will cause you to lose efficiency and will produce more heat as a result. Overall, you've made an AMAZING build for an axial flux synchronous motor. Keep up the great work!
That is true, yep! My goal wasn't to minimize eddy currents this time... but for the future I have a stator design in mind that should be great at minimizing eddy currents.
Also, the eddy currents can be used to move something like an aluminum wheel. Just as an example to illustrate the point, if one were to rotate a set of magnets arranged in alternating north/south alignments near enough to an aluminum plate/flywheel mounted on an axle that plate will rotate and there will be less overall heat in the original system from the eddy currents. An eddy current is basically an opposing force in the opposite direction and can be used to do work. I've been looking at this as a sort of capture system to capture the extra force generated when a magnetic gear slips due to too much input torque. Here is a video demostrating the basic priciple - ua-cam.com/video/wYNMpRUAuyY/v-deo.html
The simple explanation of axial flux and synchronous motors followed by "reluctance is.... uhhhh........" was so real
UA-cam recommendations never cease to amaze me with amazing videos like this one!
Nice video! I think it'd be possible to replace the iron parts with iron powder-infused resin. That way, you could just pour the iron resin where it needs to go instead of cutting and bending steel. You'd also minimize eddy currents since the resin is non-conductive. You could also take it further and replace the resin with metal-filled 3d printer filament, although it'd be more expensive and wrecks brass nozzles.
The magnetic filaments for 3d printers are very poor magnetic materials - the steel laminations I made are far superior in terms of magnetic permeability. Any powder-based core is going to have poor magnetic permeability in general. In practice, as far as I am aware, powder-based magnetic cores are used in high-frequency inductors and inductors where you want a very clean release of energy... basically in places where you want to minimize eddy currents but don't care too much about the other magnetic characteristics of the core material.
I already have an idea for how to make the stators better in the future.
@@BirdbrainEngineerno eddy currents, yes, but more importantly in hf applications you NEED low hysteresis (which metallic iron and steel lacks), ferrite cores are a necessity in hf applications, the eddy currents are more of a bonus.
Ah yes, hysteresis losses was another reason for ferrite cores... couldn't think of it at the time. But yeah, I'm absolutely not well versed in high frequency applications, there's a whole load of new stuff I'd have to learn first.
You don't need iron core at all by converting windings into PCB stator then there is no problem of Eddy currents and Steinmetz losses
@@BHARGAV_GAJJAR PCB stators are not suitable for high load applications due to fairly poor thermal characteristics and low current density compared to wire windings or better yet, hairpin windings.
THIS, is engineering. the application of science to solve a problem. if more people realized what engineering is, im sure many would be intersted.
Just discovered you through the recommendation gods known only as "The Algorithm" and frankly I'm annoyed I hadn't found you sooner. This is exactly the type of stuff I love to see and your designs, prototyping, presentation and editing is all amazing. I hope more people are able to stumble across your works, not just because you deserve bigger recognition but also because it's genuinely great content that people would enjoy if they knew how to find it.
I look forward to seeing how this progresses, but for now please excuse me while I go watch your entire backlog
ᵃˡˢᵒ ᶦ ˡᶦᵏᵉ ʸᵒᵘʳ ᵖᵘʳᵖˡᵉ ⁿᵃᶦˡ ᵖᵒˡᶦˢʰ :)
You're so good at explaining and making things easy to parse, plus your nails are so pretty
epic video. I think it's great that even if things go south with the project, you can still make a video about it and somehow it's even more interesting than a success story.
I am certainly no expert and really learned a lot from the clip. Thanks! One of the issues I thought happened about 3:18, 3:28, 3:37. All these pictures show cores with metal wrapped with a wire strand. Each core is designed to fit next to another using a ratio of winding/steel core. No spaces are left in between to maximize magnetism. When you built yours, it seemed the winding/steel-core ratio was way different, and the plastic housing took up 50% of the space. Next time, take the design at 3:18 and measure the weight/mass of one steel plate. When you build your curved plate, ensure you have the same amount of steel per core. Then when making your wheel, build a 2mm curved slot for the steel sections to sit down into. Remove all the plastic barriers. I would recommend making a smaller easier-to-build motor with fewer cores. Failure is normal, make it easy to build and learn, then build again. After you get it working well, then make a bigger, more core model. Anyway, my ideas...
Holy smokes!! This is one of the coolest engineering + 3d printing things I’ve ever seen!!
Congrats!!
Very nice! I've been making a motor for a pc cooling pump, and landed in about the same axial configuration. I used permalloy sheet cut into iron slots, but it worked without core as well. It took me about half a year to build a working pump.
Now I am building a blower turbine for air cooling, rotor printed already, but had no inspiration in developing it further. Now I might have it, thanks to you.
Awesome design! I haven't seen involute coils like that before, but it looks great. Much easier to DIY than radial flux motors. Especially if you use thinner steel sheet so can stack them up with glue and clamp between two cauls to create the shape rather than having to individually press bend them. And make non-stick spacers that half-fill the slots so you just pack as much wire in as you can, hit with CA, and then pull the spacers, leaving perfectly sized slots for the neighboring coils.
Amazing work, No matter what's the result, the honor to make the experiment it's a success
keep it up
amazing project! i am really looking forward to your bike motor video!
I learned so much in 15 minutes thanks to your years of effort exploring your curiosity.
I have finally built my electric bicycle and it is fun passing people on the highway while going 45-50mph (70-80kph) I have my motorcycle license for this amount of power. I just wish I had a motor controller that had regenerative braking, not necessarily for recovering energy but primarily for stopping. My bike with its 3000 watt hub motor and its 72volt 40amphour battery weighs in at an earthquake generating 101lbs (46kg) and that with my 200lbs (91kg) traveling at those speeds the brakes for bicycles are being consumed like warm butter on a hot sidewalk... They don't make breaks for bicycles that reliably dissipate that amount of kinetic energy.
If you put resistors on the motor terminals it will brake, the lower the resistance the more it will brake but do not forget that this can heat up the motor
@@BryanPino-l8t thank you.
This take me back when I was on college, I did as well a generator that uses eolic energy as input to the shaft, in that time I made the parts using a laser cutter and wood and bought a 3D printer to make it better, but the wood prototype was enough, I graduate from school, and 2 years later I finally work again on this project with cleaner parts from the 3D printer, I'm not an electrical engineer but definitely found fun in doing this type of projects, great video, good day to you.
This is nuts that this is able to work, I just wanna share my admiration. I'm just sitting here knitting (very slowly) and impressed by how much people can make with 3D printers
I can't believe I'm seeing a dude on UA-cam design his own axial flux motor. Im happy I'm living in a time of great technological change. This decade is going to be a wild ride and it's going to be exciting unless the nukes start flying.
I really like how you demonstrated the different type of commutations. Also great job on the rest. That's a lot of work!
This is a really neat project. I imagine a layer of insulation between the sheet metal plates would reduce eddy currents.
It would, the laminations were made simply because it was easier to construct the parts that way. But I do have a lead on how to construct the stator next time to reduce eddy currents by a lot.
This is so cool, thank you for sharing your work with the world!
This project is awesome, its fun to see it progress. Cant wait for the ebike video!!!
Awesome! I stumbled across this type of motor for the first time today, and this video did a great job of explaining the guts. Hearing this style of motor was lighter, it got me thinking about 3D printer extruders, and how the fastest printers tend to have the lightest hotends. It got me wondering how small folks have made them. While I didn't find a NEMA17 sized motor, this was still awesome to see. Thanks for sharing. ☺️ PS: 💅♥️
awesome video! I am looking forward to your next designs sis!
Thanks! ❤️ Will be doing a couple other projects during winter but early spring will be the time I make a real effort for making that e-bike motor :D
I think this is very impressive. Well done!
that loop/curve is also present in a lot of celtic knot art
thanks for making this its quite interesting its similar to a idea i have had for a telsa turbine shaped ionic continual pulse engine
Just subscribed. I can't wait to see more. Great video❣️ Thanks.
Your magnetic flux paths is through or normal to the pancake surface, so it will make eddy currents in the plane of the windings creating the magnetic flux, offset by gap. So the thin sheet, or not so thin sheet, is thin in the wrong dimension for breaking up eddy current areas into smaller, less lossy ones. Giant toroid cores can be cut into sections that are approximately the shape you have. Like a 1/3 of a toroid. The powder core does not need orientation like soft steel layers do. If the material can be had, and a way to fire the shape, you can make your own custom core shapes. Liked this video, shows curious not afraid of failing, just want to learn mode. :)
This is awesome . . . the achilles heel to the Axial flux motor IS the dependance of large mafnetic requirements . . . . there just isn't enough know deposits of approriate rare earth to powe the potential need of this motor type . . . . . a great achievement. The next step is to find ways to increase the efficiency as reluctance motors are always less efficient.
Distributed vs concentrated coils depends on stator/pole ratios. It shouldn't be decided based on other goals. It's simply what needs to be a full pitch span of coil.
Wow. Awesome project really, the way you set up the coils and ferromegnetic look very nice but I bet they were a pain in the a... to set up. I am actually trying to build a wind turbine generator and came up with faulhabercoils in the stator and a double rotor (on top and underneath). Those faulhabercoil windings are very nice, because you don't come up with lots of wire outside the magnetic fields, the downside is, that the magnetic fields do not cross the wire in 90° angle. This is (maybe) the cause why you have to 'kickstart' your motor. Love your work!
Excellent work. Very novel. Since you're using a lot of 3D printed parts have you ever tried the iron composite PLA from ProtoPasta? It's not as good as using real iron, but it's far better than just plain plastic. It's about 40% iron by weight, which translates to about 8% iron by volume. Magnets will actually stick to it fairly well, but it's not magnetic itself. For generating stators, or for making iron backings for rotor magnets, it's just so convenient. All those volute shapes could just be 3D printed instead and would immediately do the job, and there is, in essence, no air-gaps between the iron in the PLA and whatever you're wanting to generate your magnetic fields.
many thanks for this valuable information
Wonderful process. I don’t know much on this but it’s fascinating.
I don't know how this channel hasn't got hundreds of thousands of subscribers. Let me feed that ignorant algorithm.
Haha thanks! It's much appreciated and needed... UA-cam's algorithm seems to be working in a very flawed way nowadays - previous 3 videos I have uploaded have been doing awesome for *exactly* the first 21 hours after upload, and then the impressions rate suddenly gets nuked to be like 10x+ lower than it was literally minutes before, despite all of the metrics staying good and promising... I am guessing that the algorithm is currently built to literally push a video hard for only the first day as it's "fresh", and then it discards the video in favour of new "fresh" content. Pretty bullshit if you ask me, especially since this doesn't seem to be the case for large channels - their videos stay being recommended for weeks usually...
Building magnetic stators that work & work efficiently is no easy task. From taking stators apart the main rules are that the laminations have to be packed tightly, are not able to move against each other, and of course are varnished to create the electrical isolation barrier. They also need to be reasonably thin and made out of the right kind of steel.
12:52 Yaaay!🎉 We have a working long-name motor😂😅
Nice work birdbrain
the gap created by the print between your windings and the laminations probably causes a lot of reluctance. reluctance motors in general are lower torque and are more sensitive to air gap also.
for your KV: you should do the opposite test, drive the motor at a set RPM and measure the output voltage under open circuit
:D I think, the uh, iron inserts for the rotor would be making noise since it's loose. buuuuut, I think you could make a dye press out of wood levers and 3d printed parts.
also I think designing your own esc with visual feedback, as hard as it is, would help you see what's happening, and adjust everything!
I considered making a diy FOC unit... but fuck me there's a lot to it if I want to make something that actually works properly haha. Definitely would have to be a video on its own and I'm not sure if and when I'd do it.
@@BirdbrainEngineerhave a look at simpleFOC of you haven't already, it's pretty cool!
Thank you for making a video of you fabricating a 3d printed motor. Seeing your video is inspiring me to dust off my original Radial Flux design and see it through to completion. Stay classy.
Wow! What a video, you are quite the teacher, keep on going with your amazing and very hard work my friend, I would absolutely love and appreciate new videos. I’ll make sure to share this one around :)
If you're looking to do without earth magnets, have you considered maybe a shaded pole motor design? Of course you'd need an inversion system of some sort to convert it from AC to DC, but I recently pulled one out of an old microwave, and it seemed to work well without the need for magnets. I'm new to all this, but an AC motor of some sort would seem like the way to go
its a great concept and i believe in your brilliance
Instructions unclear; I accidentally created a black hole.
14:55 so you made the world's biggest motor for a 7,200 rpm hard drive, well done.
One suggestion is to find a local metal shop to cut your metal strips.
I was worried about delamination during your RPM test. You're braver than I getting an FDM part to 7200 rpm, ha. Great video!
I do all the initial tests for motor max speed behind a sheet of metal and not looking at it directly. On other motors I usually find that the speed is not too bad so I ditch the protection but not with this one haha.
I bet two of those doubled up as one one motor would do well on an ebike. It would need decent torque to move one, though. I don't know too much about them, though, to really speak on it.
This design in concept, really is genius! It looks like a double Fibonacci spiral…like a sunflower. The problem is with it is twofold. First is that the magnetic flux needs to have a continuous flow path from opposite sides so that it magnifies the torque as much as possible (basically oscillating from left to right, back and forth for each ‘petal’), and the wires of the coils need to be right next to the stator blades (as close as possible without shorting out) with no plastic in between. The plastic of the mold and the air gaps are essentially turning your motor into a giant spiraling capacitor instead of making torque. The plastic and air is a dialectic, which turns the coils into coiled capacitors which are charging and discharging into each other, instead of generating torque, because the flux and the coils are not perpendicular and opposing, like in a normal axial flux motor design. One way this might work is if you could somehow print the sunflower stator and petals so that they could be one contiguous conductive core instead of conductive ‘leaves’ being held by a non-conductive/dialectic core. Then the current wire/coil configuration might work at driving the flux around continuously from left to right so that it creates torque instead of capacitance. Just a thought….but I could be wrong.
No matter the motor design, the coils are always insulated from the cores, and the cores are left floating (exactly for this reason of reducing capacitive losses), so there really is not any appreciable amount of capacitance there. But it is true that the coils are supposed to be as close to the cores as possible - unfortunately in this case it would have damaged the insulation of the coil wires, and it would have been very difficult to keep the metal strips that make up the cores stable while winding the coils, so the plastic enclosure for the cores was necessary.
As for the continuous flux paths, that is exactly what is going on here.
@@BirdbrainEngineer Also the second thing, which I was scratching my head at is the coil winding pattern itself. You are using an overlapping serpentine pattern, which may or may not work at all. In the original vanilla axial flux motor you have very simple coils of copper wire around ovular or rectangular metal neodymium magnet stators. It’s simple and works. This design you made is so much more complex, and probably needlessly so. I would start with a working plane vanilla axial flux design and slowly change it by adding one idea at a time, rather than going all in at once with many unconventional ideas, and that way it becomes easier to tell if an idea works or not. But I commend you on your efforts and your ingenuity. So again nice job!
@@BirdbrainEngineer The serpentine winding pattern + large air gaps and dialectic stator = massive capacitor instead of moving stator. That’s the gist of it.
@@BirdbrainEngineer Here is how I would suggest doing a redesign using your initial basic ideas… 1) Simplify! Limit the number of petals to 6 - 12 (max) total around the entire stator. 2) Make the shape of the petals like a hysteresis curve with no sharp or pointed ends (like a real sunflower petal shape). (Sharp or thin ends will focus all of the magnetic flux to those points while killing the flux in the body, which is not what you want in a motor design.) For flux to be even you need a regular, smooth shape (like a platonic solid) with no thin or sharp ends. 3) Don’t use a serpentine winding pattern. (use regular self contained and separate coils of wire instead)…Have distinct and separate coils with no overlap. 6) Keep the coils as close to or wound on top of the stator magnets as possible without it shorting - (ie no air gaps and no dialectic material in between with the sole exception of the tiny amount of coating on the copper magnet wire itself.) 7) Definitely use permanent magnets, because it won’t work as a core-less design unless the entire thing is fabricated as continuous conductive metal. That’s it! I think if you make those changes the motor should work as intended.
its amazing how axial or rotor still 365 herts AKA 360 degrees or a circle equates to 120 which is the current degree seperation of a 3 phase power producion system today. which is 1/3. I would try using air core as in no core.
You can bias magnets into focusing their fields on one side by a halbach array, what´s not possible with coils. So, the stator has to be electromagnetic, and in the middle, in order to exploit its both sides, while the rotor is made of 2 discs with magnets, in halbach array config, embracing the stator between them. Otherwise it´s a huge waste of magnetic field, and thus, power/energy.
Ok keep grinding your halfway their
really excite to see where this goes ! i want to make an ebike myself ^~^
I'll be going for the e-bike motor and e-bike build based on the motor only some time in the first half of next year, as biking in the winter is a bit meh :P
@@BirdbrainEngineer yeah that's fair, even in the winter i still wanna start biking more though, and having an ebike i think would really help a lot with that
should use those green magnetic viewing sheets to verify fields. helps tremendously when making electromagnets :) Cool project. wish u much success !
Also, digital simulations can be misleading! More often than not, simulations merely get a person to believe in falsities.
I actually used a linear hall effect sensor as a really basic gauss meter, and the stators did generate 3 north poles and 3 south poles in alternating polarities.
@@BirdbrainEngineerSo it's an AC motor.
So, you made an axial brushless universal motor, kind of. If i understant it correctly, the 2 winded faces have the same winding and are both static. The core is ment to position itself something like a magnetic transmission does?
Soo to work, the phases of one side would need to be very slightly offset from the phase of the other side, and the magnetic field should be only directed by the laminated core. Im not sure it can work if the laminated core is parallel to the axis. I think you might need the phases to be in line and have the core try to redirect the magnetic field to produce he offset.
If im right, you would need a new core but could test it kinda quickly.
Looking at your rotors, one cant really see much reluctance variation when they spin, regarding the coils. Low reluctance variation means almost no torque. A way to get stronger reluctante variation is to redesign the rotors and the wiring, to imagine having like U electromagnet and a U steel core facing each other. The orientation of the rotors laminations is paralel to the ones in the coils, which might not produce the highest reluctance variation.
The stators generate opposing magnetic poles facing the rotor. As an example a single flux path "starts" from stator1 pole, through the rotor, to the opposing pole on stator2, through the back iron of stator2 to an adjacent pole on stator2, through the next set of laminations on rotor, to the opposing adjacent pole on stator1, through the back iron of stator1 back to the "start". These generated poles rotate rather smoothly thanks to the distributed windings and so they should keep the rotor nicely in sync. The problem is that since I only managed to fit 9 turns on each stator, then there was just very little flux to begin with, and some other reasons were also said in the video.
I hope you return to the involute design some time in the future, it looks like an intriguing design that deserves to be explored further.
There in Cars and Trucks GMC uses them in there Hybrid Trucks Since 2003/2004 and Many other Cars and Trucks have them and Aircraft use them these kind of Motors these are NOT any New kind of Motor.
Loved this video
I dont understand this, but it looks really cool. Good luck with it, this reminded me of the "dark matter" electric motor Koenigsegg made.
Thats cool ma dude, I am suscribing
cool! i like the bend, I don't understand it much but i feel that a cycloid drive is supposed to give & take power/torque with stepping the magnets
I wonder if using electromagnets vs permanent magnets and using some sort of pulse/timing based approach to rotate that last rotor would work? What I mean is that maybe one could replace the permanent magnets in that last rotor with steel wrapped in copper and then simply feed it power when the device needs to run. One could potentially even have that rotor be stationary and alternate the charges in such a way that it is effectively as if the rotor was spinning. That might work as both a sort of torque and speed regulator, meaning if one increases/decreases the power there will be a corresponding increase/decrease in magnetic field strength and if one were to increase/decrease the speed a which they are alternating the electromagnets that would effectively be like controlling the overall speed of rotation. Also, I've been doing a lot of research into static electricity and potential methods for harvesting it to do useful work since it is fairly straightforward to generate extremely high voltage static electricity. Throughout that research I found ion thrusters and magnetohydrodynamic drives to be really interesting. An extremely oversimplified explanation of how they use the lorentz force to generate thrust is that they basically take an electric charge, hit a magnet with it at the appropriate angle(90 degrees) and that causes a force that is perpendicular to the velocity and magentic field strength and is basically the product of the velocity of the charge and magnetic field strength(it acts like a multiplier when the angle is right). My basic idea was to take advantage of the phenomena of short circuits to power such a device. When there is a short circuit there is a sudden drop in resistance which causes a corresponding spike in amps which can be framed in such a way that it is analogous to increasing the velocity of the charge, or like if one were to suddenly remove the support holding up a massive object. One could potentially create a pretty extreme amount of force when combining relatively low voltage dc power combined with sort circuits and lorentz force.
Very cool! Just makes me realize I still need to learn so much more, hahaha.
So a difference between SynRM motors and IPM/SPM motors is their Id, Iq current scalars. In an internal PM machine the Id current is always going to be negative while the Iq current will be either positive for positive torque values or negative for negative torque values. In a Synchronous Reluctance Machine the Values of both Id and Iq are positive. This could be why you kept getting a stuttering effect and no motion, the controller was braking the machine rather than accelerating it. I'm not sure what settings you had enabled in the Inverter but that could be something to look into. Also the Field weakening algorithm is different between the two motors.
yo cool bird dudette!
1:10 Now I might be flying a little to high, if you catch my drift, but could you like idk combine them or smt? Surely more means better, right?
IIRC, that would be called a "raxial flux" motor, and they do exist, and they are basically the highest power density motors in the world. Maybe one day I will try to make one.
@@BirdbrainEngineer❤
such a joy to learn this stuff! thanks
Hello BirdBrain : i enjoy the video. You use a controler for BLDC. For reluctance motor a very special controler is needed : asymtric half bridge
awesome video! really good production quality, extremely informative, one of the most informative Iv seen in a while, and just ingenuity at its finest with all that math and good understanding of the motor and its physics! EDIT: also Im just curious and i doubt this comment will get seen but how do reluctance motors induce a back emf/ resistance of some sort to the input voltage. all motors of some sort have some sort of back resistance because as rpm increases output power increases and thus input voltage and current draw should also change. or if there is no back emf or anything how would a motor like this work or what shape power chart would it produce?
This was worth a subscription!
You did a great job.
Nice Work
You Sound Like Buk Lau ❤ 😂
Increase the number of phases in the coiling and base your curve on the golden ratio then your initial design will work. You may need a controller board to switch the fields on and off precisely
I don't think the golden ratio curve matters, but more turns (than the 9 I had) would have made it more likely for the reluctance design to work, even if poorly.
About driving it, maybe working with true sine power. The Spanish youtuber Electronoobs made a video about driving bldc motors that way. About the variable reluctance motor, I might think that the steel alloy isn't magnetically soft enough, having too much histeresis renders the variable reluctance not working
Yeah the steel is not good for this purpose... One day I might try to make a properly working reluctance motor as I now have some ideas on how to make better stator- and rotor cores. Since wave driving is nearly always preferred, yeah.
Good job, congratulations. It is inspiring.
What kind of calculations did you do in order to determine the design of the reluctance structures for the stator and the rotor? Have you used any electro-magnetic simulation tool like FEMM or MAXWELL? I use FEMM for a good approximation of axial flux motors / generators. It is a little bit tricky because it calculates more in 2D and I have to build the e-machine in a way to get a good approximation of the data. So I simulated a linear motor version of the build circular structure. But this is ok. The data came out in 10 to 15% approximation.
Cool project! Maybe laser cut parts could be a decent intermediate scale alternative for widely available and affordable production access besides 3D printing
The next step that would be significantly better than 3d printed parts (and thus worth going for) would be metal parts. So either metal laser cutting, plasma cutting, or (cnc) milling. None of those methods are that widely available nor affordable unless you yourself have or know someone who has such a machine. The e-bike motor I will still try to make in such a way that nearly anybody could replicate it "easily". Though at one point I do intend to build a cnc mill capable of cutting mild steels.
@@BirdbrainEngineer Laser cut parts are pretty easy to order online.
Amazing!and very educational
5:40 thanks for credits given to bg music
I randomly found this video and enjoyed it a lot! Always fun to see a maker doing something interesting and it's a bonus when the voice is as lovely as yours. ^_^
I loved your GD moment!!!
That is one strong motor you made
Very nice effort! Welldone.
I would need two of them for my wheelchair. 😂
Would probably have to be a bit more robustly made as well :P
An easy way of making mediocre performance iron cores in impossible shapes is filling the mold with fine iron filings then impregnating it with very thin glue, preferably epoxy. Better if you can vacuum impregnate them.
Impressive work ! 👍🏻
Very cool stuff!
Like you said, bot having proper bulky magnetic cores is probably the biggest problem. But besides that, I wanted to point out that i think you should be able to measure the rotor position using the existing copper coils. Using an hf signal, you should be able to see different inductances at the channels, depending on their alignment.
I'm not sure if there's a good way to get an absolute position this way, but at least getting the (electrical) position of the coils relativ to the stator should be possible (and sufficient for efficient operation).
If you don't feel like modulating a pieer wire, you can probably also add some extra sense coils. Using the symmetry of the design you could place them in a way that they would be mostly unaffected by the voltage induced from the poeer coils.
Dang, noe I kinda want to try that...
Back-emf based control scheme is used in sensorless speed controllers, however, since the initial idea was to make a reluctance motor, which produces no back-emf, so this control scheme would have been impossible. Generally hall effect sensored position feedback is superior to back-emf based position feedback, especially at low speeds, which is why basically every synchronous motor that has any significant power output has hall sensors in them.
Not sure if I'm correct (automation engineer by trade, so I think I might be onto something, but I don't know much about synchronous motors), but doesn't the reluctance rotor also need windings? Like... Except the magnetic flux core, should it also not have a conductive "cage" in between? I know, that asynchronous industrial motors have something, that is called a "cage winding", in English I think it's called a "squirrel cage winding", which is - most of the time - made from aluminium for cost-saving measures, and it is just a bunch of aluminium rods stuffed between the metal rotor core, and then welded with two aluminium rings on the opposite sides of the rotor. To achieve the same thing, I think one would have to wind copper windings around the metal core and then solder the ends together, so it makes a continous wire loop, so the current induced has a path to flow.
What you are explaining is called an induction motor :)
Induction motors are famously non-synchronous, as the rotor speed is not coupled with the rotation of the electrical poles due to what's called rotor slip. The non-synchronity of induction motors is also why they are often preferred in industry - you can usually just plug them into 3-phase grid energy and they'll work... no controller needed.
Hi, Love you video man! I've worked on some ac and BLDC motors in the past and I'm curious if you looked into your ESC at all. I've had major problems running normal motors with commercial ESC's and having the control algorithms work correctly. (Even in sensored mode which you seem to be using). Usually it involves a lot of PID tuning and tinkering. Have you tried changing your ESC when trying to run your induction motor? It also seems like the ESC is made for higher speeds which could also cause problems, I remember using some motor controllers called SOLO or something with a nice web interface and they had a different mode for low RPM motors which you probably need. (I think it has to do with the speed it runs the PID loops or something)
Yeah, the ESC tries to spin the motor pretty fast but it did sometimes try to sync up to the extremely low rotation rate (you can hear it thanks to coil whine). In addition, the ESC I used does work just fine with a high-torque low speed RC car motor.
Loved your painted nails ❤
Thanks
Very innovative project👌. Can you suggest this concept can be converted to make axial generator with maximized magnetic field distribution?
Unfortunately pure reluctance motors can not be used as generators. However if you meat whether a proper involute design brushless motor could be used as a generator then I'd imagine so, however it might not be as good as a generator with the standard layout... would have to do some simulations.
Pretty awesome! Thanks
i only understood 15% of the process how the motor works but i loved it haha. :D
Hello friend!
I've played around with turning car alternators into bldc motors. You can see what I came up with on my channel.
But I was working on a design for a helical toroidal flux bldc motor, but had to move. Link to my printables designs in my channel page.
I am really loving your 3D printed designs! And your need to use distributed windings for smooth operation had me thinking of my helical toroidal flux motor.
Yes did use the one loop at the loop connections would have been way better you and running them side-by-side like that as long as they're touching might as well loop them