Problems of heating: 1. Eddy currents, which might be heating the iron within the PLA, typically motors use laminated steel to minimise this. 2. PLA being a good insulator and not being able to dissipate heat generated from eddy currents or the resistive heating of the copper. 3. Friction from bearings Efficiency: Due to the lack of conductive material required for magnetic permeability achieving high efficiency will be difficult. Very cool though, I may have to print one :)
I assume that making a laminated core from steel sheet will make this way more efficient and less heated. It's a must for the design. You can make the layers with the CNC then glue them together.
@@martindieux The PLA with iron powder works like an SMC. Infact the eddy currents generated in this setup are likely far lower than laminated steel sheets. Whats most likely the issue is PLA's mechanical properties degrade as the stator winding temperature increases (I^2*R losses). Since the windings are in tension (from the winding process), this crushes the stator teeth when hot.
A lot of your efficiency losses will be from the motor being unbalanced. Every rotation you can think of how much energy each wobble is exerting into the side of the bearings and not into rotational energy.
A lot is doubtful, the energy from the vibration will result in energy loss but they will be comparatively low to other losses. My reasoning is if all that energy was wasted in the bearings the bearings would be melting the plastic and not the stator. It's quite likely the stator is saturating or has lots of Eddy current loss or other things and that's why it's getting hot enough to melt and "nothing" else is getting that hot. Also the fact that the winding with more current (test 2) has lower efficiency is probably a good indicator of this.
@@jakbain1337Idk man, an unbalanced motor dumps power to more than just radial loads. It limits the max RPMs significantly, introduces all sorts of turbulent flexes into the blades, keeps a nice tail vortex from forming. You lose a lot of thrust from these three issues; I've seen a an unbalanced motor cause a ~50% thrust loss all by itself. I do agree that a saturated stator sounds very likely, given the heat distribution you described. The lower efficiency on the higher current run could also be caused by the unbalanced motor though, as many of those issues scale nonlinearly; bad tale vortex, and your blades start having to push through oscillating pressure regions, which kills momentum and causes more loss.
@@stevewells20 Absolutely the vibrations and flow on effect on that on the performance of things can be significant especially if the system isn't very stiff etc. In this scenario that isn't "that relevant" to motor efficiency as the thrust, motor mechanical power is being measured independently.
@@jakbain1337 Ah I see what you mean, I didn't think too much into the testing method; given that, the dominant effector is going to be in the electromagnetic system! Thanks for pointing that out, I missed the forest for the trees lol
I can provide a little bit of insight into waterproofing your electronics pod, if that's the thing you need to do again. It's possible for water to infiltrate the pod between the insulation of a wire and the wire itself like a tiny plastic hose, so at the point where the wire goes through from the dry side to the wet side through the resin you want to strip off a bit of insulation, so there's a short section of exposed copper inside the resin, then the resin will effectively plug the little pipe created by the insulation on the wire, and because the wires aren't touching the resin becomes the new insulation that separates them inside the joint. It's not too big an issue just to be rainproof, but when you're dealing with water pressure it always finds its way in. You can also or additionally pressurize your electronics package, install a Schrader valve like on a tire to the PVC pipe or the cap on the opposite end and pump it up to pressure higher than what you're going to encounter at the bottom of your operating depth.
Give the enclosure a bilge and put some water-absorbent material in it, eg water beads, desiccant packs, calcined clay/gypsum. That'll give some margin for minor error.
Agree, what @AlRoderick said. There are a couple more helpful options. 1) You can make electrical passthrough connectors with threaded rod, washers, and silicone. The trick is to silicone the joint between the pvc and threaded rod, let it set up, and then compress it with a rubber washer and nut. The washer and nut compress the silicone against the threaded rod, stopping migration down the threads. 2) Fill the PVC pipe with a non-conductive fluid like mineral oil, like the immersion PC people do.
I ran into this doing an underwater ROV. We just could not figure out the water ingress, doing seal tests in a bathtub worked perfectly, but then 15 feet under and we'd get water. No water in the Oring channel, but somehow water in the chamber. We ended up making a passthrough of bare wire with a scotchkote coating, then potting that in epoxy. Probably cheaper at that point buy aircraft style passthrough connectors.
For the stator, may be cut a metal strip into rectangles and then stack them up into several blocks. You could then 3D print a structure with iron-filled PLA to hold them together. Altenatively, insert holes in the existing stator so you can insert nails into them. They'd have to be inserted from the inside though so they don't risk sliding out.
One of the things I rarely see done in these 3D printed motors is a magnet backiron as well as a stator yoke. The backiron will increase the power output a lot without sacrificing efficiency
@@noanyobiseniss7462 I've never even left the rabbit hole 😅Spent a few years doing research on motors and motor drives and I've built a few PCB motors myself
@@newagerc5356 The strength of the magnetic field will be determined by the strength of the magnets and the resistance of the magnetic path. You can diminish the resistance (and therefore increase the magnetic field strength) by adding a steel ring behind the magnets. This way the magnetic field increases and you'll get more torque out of the motor
There's a rule of thumb to have your pole count as close as possible to your slot or winding count (but not equal cause that can't work). So for 24 windings you're better off with 22 or 26 magnets instead of 20. Though this would improve like 1% so not a big deal.
It's not a problem to have equal number of windings and magnets, it's actually the most efficient setup for power output, but they don't do it because it causes cogging at low speeds - the magnets are attracted by the ion cores opposite them and it stutters, even sticks into this position when stopped. Otherwise, yes you need them as close as possible, just 1 or 2 off to prevent cogging.
Very cool. I love the fact that the motor performed better underwater than the camera and enclosure with millions of dollars in development which overheated submerged in cold water.
High voltage will be your friend. You can do higher windings to reduce Kv. For example, a 320 kv 6s motor vs a 160kv 12s motor with the same load, the 12s motor will run with a tempt delta 1/3 less than the 6s.
Ok, I've been planning on building my own small 3 phase motor and design is definitely a great starting point for me. As for improvements using a cnc laser cutter or similar device, cut out the stator from thin steel sheets and stack them in high temp resin. That is my plan partially, Ill end up tweaking your design a bit more but not for improvements but to fit my purpose. Dude thanks and keep up the work!
1:33 You can also print transparent parts on an FDM printer. Use transparent PETG, print dead slow (5-10 mm/s) and with 100% linear infill and some (101-103%) overextrusion. Not as perfect as resin, but it does the job.
You need to have a iron (or steel) core. The 3D filament you use never works efficiently. Not only have I tried it but other TH-camrs have tried it and we all have the same result. However you have your own milling machine. I think it should be possible for you to get some really thin steel sheets, and either cut them out on that directly and laminate them together yourself or, maybe cut squares quickly and spray some kind of insulating layer on them and stick them together one on top of the other until you get a thick piece and then mill out that thick piece to the shape you need. Either way you end up with a laminated core like all of them have. This will not only make it more efficient, but give the coils a heat removal path.
I had an idea to use flat roofing sheets. They're about 0.5mm, I think, and they come with an insulating coating from the factory 🙂. I've never gotten around to trying it though.
Print some fan like scoops into the housing or coils to force water or air through the coils as the motor spins. Thus creating active cooling to raise your performance cap before you melt things!
Sweet project! I'm surprised that you didn't do the Halbach array to begin with. Balance likely will help with with your project as well. Would love to see where this goes.
I can recommend looking up on three phase Y and Delta connections for motors. Generally (if i remember correctly) you start the motor in delta connection and then switch over to Y connection after the motor is started. This gives the motor better stability at the start and becomes more efficient after switching to Y connection. The draw back is that the circuits become more complex. Additional notice: Tesla is just started working with six phase motors. But I speculate that they just drive both side of the coils instead of doing a Y/Delta connection.
You did amazing Bruv I love the resin being used! I also love how you make it easy for everybody to understand. This is some innovative thinking! We need more inventors like you to progress in this world! I love it! we can make so much progress with bldc motors! Your design looks soooo well balanced! Personal bldc vehicles are coming!!!! Try tilting the magnets 25’. Looking forward to the next project! A/c free energy, LOL
Nice job man! You could further improve your motor by balancing the rotor (it seemed a bit wobbly in one of the shots). Another way to improve it is by improving your winding technique (I know it's hard to make a nice straight and evenly spaced out winding by hand).
Mold the core from Smooth-On EpoxAcast 670 HT. With proper post curing it can take up to 177C or 350F. Mix in some powdered iron to make it like the 3D printing filament with iron.
With a bit of steel wire, a rig could be made to use the router table to wind the coils and maybe also do them more evenly, or maybe even get more raps or at least more balanced, neat raps.
I actually created a 3D printed Brushless DC, Areal flux Pancake style, 6 coils, 8 magnets, 15AMP ESC. I had the same issues you did with melting coils, so I "upgraded" to PetG, which helped to some degree. Issue is though, I couldn't find any metal version of PetG, so I also experimented with core metal screws to clamp the coils to the stator, which helped some with the tork and the starting issues, that i saw you had issues with too. For heat dissipation I designed a version where the rotor assembly had small lamels shaped like wings to create airflow through the rotor and stator, in an effort to cool down windings... It kind of requires a pancake design to make room for the airflow in a direction that would help, and so now you can also get this in commercial BLDC's, which I think is a testimony that it actually would be a way to go too... I however never had the tools to get this version "good enough" to run.
Well done! The DIY BLDC is really impressive -- the benchmarks at the end are super helpful and will make me think twice about buying a COTS motor for some upcoming projects! (Also, what lake were you testing in? You were starting to hit Lake Erie depths! 😅)
A SUGGESTION FOR ACHIEVING HIGHER OVERALL PERFORMANCE: the slot setup you used to mount the Permanent Magnets increases the distance to the coils. The tighter you make that gap the better! almost all of the motor's performance will improve by making that gap smaller.
' Will it work?' Yes, of course it can basically work. The real problem is increasing gaps, etc to make a waterproof motor. You're generally better off to have a sealed bubble for a dry motor, and then just a magnetic coupling for the shaft through the wall to spin the propeller. There's very little gain in trying to make the motor itself waterproof, and so many ways to lose by trying to do it.
Michael, please put the loop in the Qsea’s tether around the handle!! The connector was never designed to take a load. When looped around the handle it has a 200lb rating.
Regarding the heat dissipation problem there are many good suggestions in the comments. One less good suggestion: consider making a variation of the components which melted that can be actively water-cooled (i.e. there are empty channels through the hot parts which you pump water through). Or run your tests in a fish tank, at least.
I think the most obvious suggestion would be to use a higher temp plastic for the stator. I'm not sure how the windings work, will have to check out that website. If used underwater, that actually wouldn't be a problem. - Try that same design, on your sub? No way it can overheat. You could probably overdrive it. The issue is heat, and you have an entire lake of watercooling.
Really impressive that you've gotten a DIY brushless DC motor to work. And pretty cool that it can spin a propeller pretty well. I'm guessing this not intended as a practical use item - more of an interesting project right? It spinning so slowly means your kv is pretty low. The main problem here (besides efficiency losses) is that normal brushless dc motors can get up to 90 degrees Celcius. 3D printing the stator means that it melts way before that, thus you can't get the amount of power throughput that you need. Making the stator out of metal would help...
In my underwater vehicle, I usually use Chinese motors with a sealed stator. On my channel there is a video about my underwater vehicles and how we launched them in the Arctic. If you're interested, come in.
How do you solve the problem of the bearings? They should require many replaces with sea water and dirt. This is not talked about in the video, but this is the only relevant topic, the rest is just a 3d printed normal BLDC...
Go ahead and use the high current windings, but only run it submerged. As long as there's some flow near the windings, the water will carry the excess heat away.
I suspect the iron filled PLA has a really low % of iron. Else it wouldn't print well. Also, it should be a high permissivity alloy able to quickly switch pole polarity.
The principle is sound, but you shold use laminated plates of a ferrous material in the coils to keep it permanently working. This shold also provide a lot better effiviency.
You need to use a thinner copper wire to minimize resistance and also you need to add cooling holes and maybe something that doesn’t insulate as well as pla
Radius. If you're not space constrained, wider means more torque. Litz wires (braid) to reduce EMF. Coils can spiral around the axis instead of being parallel
Cnc the stator out of metal and print the rotor. Most all motors generate heat, its benefital to add internal air fins incorporated into the inner rotor.
Some people suggest eddy currents as a problem - this is most certainly not correct as the iron filings in the metal pla are separated by... pla (famously an insulator)... thus the eddy currents in the grains of iron are incredibly small and would not produce basically any heating. Having said that, the PLA with iron filings is not that good of a material to use for a core, as the magnetic permeability is literally not much better than just air and pla on its own. It basically just adds weight and lower temperature stability for no gain. You might want to use PP filament if possible next time, as it melts at a much higher temperature, but with low thermal conductivity, any plastic is going to be fairly poor as core (and casing) material in the end. I noticed the very low efficiency of the motor, a fair bit of it is probably to do with how large the air gap between the electromagnets and permanent magnets are. The larger the air gap the less of the available flux from permanent magnets you use. You definitely need to reduce it down by a lot, especially if you want to use a Halbach array. In addition, the motor looks quite wobbly with rather large bearings so there's probably some friction- and other big mechanical losses there. You might also want to consider a back-iron for the permanent magnets as well. The winding configuration is probably fine, as the calculator uses the best practices. Another thing to think about is that because your core material is very much sub-optimal, then long electromagnets like radial flux motors often go for, are going to be less efficient. You might want to look into making an axial flux motor instead, which is better suited for making out of low magnetic permeability materials.
He's me at nearly 20 years in engineering not understanding how this lad seems to have 90 years of engineering knowledge in the body of a 20 something year old
I have a bit of future waterproofing for you: Stranded wire has little itty bitty holes in the wire and water will creep through the wires SLOWLY so you will think everything is fine and then suddenly water gets into places you never thought it would get into and cause issues.
This is just ground breaking. There have been suggestions for the heating and balancing issues My advice is to make a metal housing for it (like a tube) which lets water through (for cooling) then plate the inside with a diamagnetic material to confine the magnetic field to the stator and maximize flux density
increase the space between magnets to half the width of the magnet. Your magnetic fields are fighting each other which is why you have to force it to spin at the start.
Changing the rotor mechanical design will help for submersible project cause water is denser than air, make the rotor a circular shaft rather than whole exo rotating this will reduce torque required and reach rpm fast and reduces current consumption.
The compact design seems to have left no room for cooling the trapped air inside the stator housing . (Under water and above ground designs will differ here) Directional cooling fins and good heat dissipating materials should help further your proof of concept . Nice work .
1. Quality of winding is crucial for efficiency - each deviation from wires going in paralel in same coil weakens magnetic field very much. Also make sure that centrifugal force does not mess up coil shape. 2. If possible - move magnets closer to stator. Magnetic field decreases by quare of distance, so each change, especially on small scales, makes big difference. 3. For heat - you can re-shape your rotating motor case holes so that each such hole acts as small fan blade that moves air inside the motor to cool it
Are you able to minimize the air gap between magnets and the stator? IIRC, the rate of field strength decrease is greater for Halbachs as you move away from the magnets vs a N-S-... configuration and/or standard magnet. If you're going to put the effort into Halbach arrays, minimizing the air gap will be critical to utilizing the benefit provided by Halbachs.
i'm imagining that the most efficient way to move thru water is pulling/pushing thru it without disturbing it much, like fish and whales do. but your motor is good too :)
It's interesting that your stator melted, but while eddie currents are often cited as the cause, i'd wager it's more likely to be simply the winding current you were seeing at full power. I think the more important thing to consider is what part of a BLDC motor needs to be waterproofed. With a bit of enamel or some manner of coating, the steel parts of a cots BLDC motor could be corrosion proofed. Making the stator out of steel laminates, and in fact making the rotor out of a steel ring, are very important to a motor's efficiency as those materials and their shapes help focus the EM fields in the most optimal locations. The amount of incohesive iron or steel in the plastic is likely not very useful, and the plastic rotor will do nothing to focus the M fields inward. The biggest issue with waterproofing a BLDC that I can see is the bearings. That is where something that can withstand corrosive emersion at depth is important. in your motor, you used unsealed bearings, so you were able to wash them afterwards, but even then, they will struggle to retain lubrication and in your post-water test, you could already see rust forming on one of the races. Sealed bearings do a much better job of retaining lubricants and keeping out debris, but even they are not typically fluid proof, nor able to withstand hydraulic pressures experienced at depth. I still think this is really cool, and an awesome way to learn more about motors. I wonder how someone could make a BLDC motor with bearings that could withstand prolonged cycling in and out of water/saltwater. Keep it up, thanks for the video!
Yep. Need to terminate the insulation "hose" in the epoxy so that it doesn't bypass the seal. Strip the wires in the middle of the epoxy so that only the copper wire passes through all the way. Stranded wire? Epoxy or solder in between the strands will take care of that.
print the stator in thin layers,get some vinyl or something else that insulating and put between stator layers, you could use thin layers of insulating enamel used for winding wires, then epoxy the stator together, this will get cut down on the eddycurrents or just send your stator file to send cut send get sponsored, so its free and laminate it together with electrical enamel.
I think I can give you some help with fluids, as that's closer to my area of expertise. You should really avoid spinning parts from free waterflow and there's plenty of effects messing the flow down the line, especially if the propeller comes after it, so put it in some casing with cooling ducts as necessary. You could also make the whole assembly smoother by making the cap fit perfectly to the propeller and have some rounded edges. Invest some time into CFD.
Electric motors have a little problem: magnets are double sided,not one sided. That means that an electric motors are actually not having a full power but around 50% depending on design. But,if you use bell electromagnets (they are used in electric cranes) you will get full 100% of magnetic field power. As a matter of fact,a 20 mm magnet with 100-200 rounds of 0,1 mm wire powered by 3 AA sized batteries connected in series can lift ~140 kilos.
Ok, so you want to make an underwater propulsion system? Why not make a super duper rim thruster?! I'd like to see one DIY made. There are companies that already make them but I'd like to see somebody like you make a smaller version. I'd actually like to see somebody make a rim driven lens for GoPros to spin the water off of the lens of the camera for water sports. Have fun! Cheers, John
Wow! Wow! Wow! really cool. It's like the motor on the "Make Sea" TH-cam channel. I 3D printed 8 so far. However, the magnets in my motor are much bigger than yours and you never want to go over 12v with a 3D-printed motor because they will always melt. 😎😎😎😎😎🤩
I'm not sure how similar electric motors are to solenoids (in physical & power calculations, but I think they are pretty similar) but I remember being amazed at the difference in force the solenoid could generate by reducing the distance (space) between the coil & the magnet/iron core. You want to do absolutely everything possible to get that gap as close to zero as possible and I'd suspect the same thing should be considered when designing the distance between the rotor to stator. IDK if you would be able to turn either on a lathe to make the fit as close as you dare, I guess even using a drill press & sanding the piece might be a substitute for the lathe. Would the motor benefit from any type of flux shielding?
You can't really expect an electric motor to be efficient and powerful with all these magnetic loses. Not sure about the performance of this iron filled PLA in this application but the rotor doesn't has any metal at all which leaves a big air gap (or plastic gap in this case) between the outer side of the magnets. The magnetic flux needs to run in a closed loop through a stator finger, the magnet , the rotor casing and then back via the neigbouring magnets and stator fingers. Every millimeter in this path that doesn't run through a ferromagnetic material adds to the losses. It may be worth and try to print the rotor case from iron filled PLA also. But you would probably also have to change the mounting of the magnets because the slots for the magnets would create a kind of shortcut between poth poles of the magnets. I'm pretty sure it's possible to improve efficency a lot but probably not to the level of a motor build from "proper" magnetic materials.
efficiency will never be anything unless you use laminated transformer armature or stater as they will direct the magnetic field most efficiently to the opposing magnets... Also bearings will get damaged underwater as there are many particles which will wear the bearings out even if they are ceramic apart from the friction of the water in between all the motor parts which are very close together generally....
Perhaps using 3D printing to make a mold to lost wax cast a pure iron stator as you’re obviously not getting optimal field off a iron infused pla filament.
Too bad you can't use PCB windings - but I think it would work fine. Fishman depends on their stacked PCB tech and they're quite literally a magnetic transducer that requires fidelity - a coil emulator, if you will.
What about journal bearings/that also act as a cooler ? Essentially a turbo charger. Would still need something I guess for thrust and loads or extreme motion. It's a thought though.
seems you should copy the design of a alternator ie has built in fan so win it spining it cooling it at the same time or something to help keep temps lower so you have air flow over the Stator windings
My drone motors work fine in the water til the bearings give out from the lubrication being displaced. I dunno if this was addressed in the video no time to watch currently but yeah brushless motors are good in water generally as long as the power source itself isn't being shorted which is a pretty easy thing to accomplish. So in reality I think all you did was make a custom motor that isn't as powerful or durable as motors you can buy. Pretty neat and fun stuff to do though.
Problems of heating: 1. Eddy currents, which might be heating the iron within the PLA, typically motors use laminated steel to minimise this. 2. PLA being a good insulator and not being able to dissipate heat generated from eddy currents or the resistive heating of the copper. 3. Friction from bearings
Efficiency: Due to the lack of conductive material required for magnetic permeability achieving high efficiency will be difficult. Very cool though, I may have to print one :)
I assume that making a laminated core from steel sheet will make this way more efficient and less heated. It's a must for the design.
You can make the layers with the CNC then glue them together.
But can it be liquid-cooled? Maybe have coolant channels?
@@swishiestmetal6215 not worth. Changing the solid core for a laminated core will eliminate eddy's currents, therefor will eliminate the heat source.
@@martindieux The PLA with iron powder works like an SMC. Infact the eddy currents generated in this setup are likely far lower than laminated steel sheets.
Whats most likely the issue is PLA's mechanical properties degrade as the stator winding temperature increases (I^2*R losses). Since the windings are in tension (from the winding process), this crushes the stator teeth when hot.
full ceramyc bearings
A lot of your efficiency losses will be from the motor being unbalanced. Every rotation you can think of how much energy each wobble is exerting into the side of the bearings and not into rotational energy.
A lot is doubtful, the energy from the vibration will result in energy loss but they will be comparatively low to other losses. My reasoning is if all that energy was wasted in the bearings the bearings would be melting the plastic and not the stator. It's quite likely the stator is saturating or has lots of Eddy current loss or other things and that's why it's getting hot enough to melt and "nothing" else is getting that hot. Also the fact that the winding with more current (test 2) has lower efficiency is probably a good indicator of this.
@@jakbain1337Idk man, an unbalanced motor dumps power to more than just radial loads. It limits the max RPMs significantly, introduces all sorts of turbulent flexes into the blades, keeps a nice tail vortex from forming. You lose a lot of thrust from these three issues; I've seen a an unbalanced motor cause a ~50% thrust loss all by itself.
I do agree that a saturated stator sounds very likely, given the heat distribution you described. The lower efficiency on the higher current run could also be caused by the unbalanced motor though, as many of those issues scale nonlinearly; bad tale vortex, and your blades start having to push through oscillating pressure regions, which kills momentum and causes more loss.
@@stevewells20 Absolutely the vibrations and flow on effect on that on the performance of things can be significant especially if the system isn't very stiff etc.
In this scenario that isn't "that relevant" to motor efficiency as the thrust, motor mechanical power is being measured independently.
@@jakbain1337 Ah I see what you mean, I didn't think too much into the testing method; given that, the dominant effector is going to be in the electromagnetic system! Thanks for pointing that out, I missed the forest for the trees lol
You have to make the stator out of iron core for enhancing the magnetic strength and better torque
I can provide a little bit of insight into waterproofing your electronics pod, if that's the thing you need to do again. It's possible for water to infiltrate the pod between the insulation of a wire and the wire itself like a tiny plastic hose, so at the point where the wire goes through from the dry side to the wet side through the resin you want to strip off a bit of insulation, so there's a short section of exposed copper inside the resin, then the resin will effectively plug the little pipe created by the insulation on the wire, and because the wires aren't touching the resin becomes the new insulation that separates them inside the joint. It's not too big an issue just to be rainproof, but when you're dealing with water pressure it always finds its way in. You can also or additionally pressurize your electronics package, install a Schrader valve like on a tire to the PVC pipe or the cap on the opposite end and pump it up to pressure higher than what you're going to encounter at the bottom of your operating depth.
Give the enclosure a bilge and put some water-absorbent material in it, eg water beads, desiccant packs, calcined clay/gypsum. That'll give some margin for minor error.
Agree, what @AlRoderick said. There are a couple more helpful options. 1) You can make electrical passthrough connectors with threaded rod, washers, and silicone. The trick is to silicone the joint between the pvc and threaded rod, let it set up, and then compress it with a rubber washer and nut. The washer and nut compress the silicone against the threaded rod, stopping migration down the threads. 2) Fill the PVC pipe with a non-conductive fluid like mineral oil, like the immersion PC people do.
I ran into this doing an underwater ROV. We just could not figure out the water ingress, doing seal tests in a bathtub worked perfectly, but then 15 feet under and we'd get water. No water in the Oring channel, but somehow water in the chamber. We ended up making a passthrough of bare wire with a scotchkote coating, then potting that in epoxy. Probably cheaper at that point buy aircraft style passthrough connectors.
For the stator, may be cut a metal strip into rectangles and then stack them up into several blocks. You could then 3D print a structure with iron-filled PLA to hold them together.
Altenatively, insert holes in the existing stator so you can insert nails into them. They'd have to be inserted from the inside though so they don't risk sliding out.
One of the things I rarely see done in these 3D printed motors is a magnet backiron as well as a stator yoke. The backiron will increase the power output a lot without sacrificing efficiency
Well this post didn't just send me down the rabbit hole! ;)
@@noanyobiseniss7462 I've never even left the rabbit hole 😅Spent a few years doing research on motors and motor drives and I've built a few PCB motors myself
@@J.A.Robots Care to share a good link for a decent finished model?
Can you explain this a little more.
@@newagerc5356 The strength of the magnetic field will be determined by the strength of the magnets and the resistance of the magnetic path. You can diminish the resistance (and therefore increase the magnetic field strength) by adding a steel ring behind the magnets. This way the magnetic field increases and you'll get more torque out of the motor
There's a rule of thumb to have your pole count as close as possible to your slot or winding count (but not equal cause that can't work). So for 24 windings you're better off with 22 or 26 magnets instead of 20. Though this would improve like 1% so not a big deal.
It's not a problem to have equal number of windings and magnets, it's actually the most efficient setup for power output, but they don't do it because it causes cogging at low speeds - the magnets are attracted by the ion cores opposite them and it stutters, even sticks into this position when stopped.
Otherwise, yes you need them as close as possible, just 1 or 2 off to prevent cogging.
Very cool. I love the fact that the motor performed better underwater than the camera and enclosure with millions of dollars in development which overheated submerged in cold water.
High voltage will be your friend. You can do higher windings to reduce Kv. For example, a 320 kv 6s motor vs a 160kv 12s motor with the same load, the 12s motor will run with a tempt delta 1/3 less than the 6s.
Not guna lie u probably the only youtuber to make a real life arc reactor!
What I like in your experiments are the rigs and especially all the data you collect and interpret. A lot to learn!
Ok, I've been planning on building my own small 3 phase motor and design is definitely a great starting point for me. As for improvements using a cnc laser cutter or similar device, cut out the stator from thin steel sheets and stack them in high temp resin. That is my plan partially, Ill end up tweaking your design a bit more but not for improvements but to fit my purpose. Dude thanks and keep up the work!
1:33 You can also print transparent parts on an FDM printer. Use transparent PETG, print dead slow (5-10 mm/s) and with 100% linear infill and some (101-103%) overextrusion. Not as perfect as resin, but it does the job.
You need to have a iron (or steel) core. The 3D filament you use never works efficiently. Not only have I tried it but other TH-camrs have tried it and we all have the same result.
However you have your own milling machine. I think it should be possible for you to get some really thin steel sheets, and either cut them out on that directly and laminate them together yourself or, maybe cut squares quickly and spray some kind of insulating layer on them and stick them together one on top of the other until you get a thick piece and then mill out that thick piece to the shape you need.
Either way you end up with a laminated core like all of them have. This will not only make it more efficient, but give the coils a heat removal path.
I had an idea to use flat roofing sheets. They're about 0.5mm, I think, and they come with an insulating coating from the factory 🙂. I've never gotten around to trying it though.
Print some fan like scoops into the housing or coils to force water or air through the coils as the motor spins. Thus creating active cooling to raise your performance cap before you melt things!
Sweet project! I'm surprised that you didn't do the Halbach array to begin with. Balance likely will help with with your project as well. Would love to see where this goes.
I can recommend looking up on three phase Y and Delta connections for motors. Generally (if i remember correctly) you start the motor in delta connection and then switch over to Y connection after the motor is started. This gives the motor better stability at the start and becomes more efficient after switching to Y connection. The draw back is that the circuits become more complex.
Additional notice: Tesla is just started working with six phase motors. But I speculate that they just drive both side of the coils instead of doing a Y/Delta connection.
this is for big AC motors.I dont think anyone is doing it with BLDCs which are ESC driven.
Awesome build! Looks really nice and seems pretty interesting to make.
You did amazing Bruv I love the resin being used! I also love how you make it easy for everybody to understand. This is some innovative thinking! We need more inventors like you to progress in this world! I love it! we can make so much progress with bldc motors! Your design looks soooo well balanced! Personal bldc vehicles are coming!!!! Try tilting the magnets 25’. Looking forward to the next project! A/c free energy, LOL
Nice job man! You could further improve your motor by balancing the rotor (it seemed a bit wobbly in one of the shots). Another way to improve it is by improving your winding technique (I know it's hard to make a nice straight and evenly spaced out winding by hand).
This has got to be one of my favourite TH-cam channels.
Mold the core from Smooth-On EpoxAcast 670 HT. With proper post curing it can take up to 177C or 350F. Mix in some powdered iron to make it like the 3D printing filament with iron.
With a bit of steel wire, a rig could be made to use the router table to wind the coils and maybe also do them more evenly, or maybe even get more raps or at least more balanced, neat raps.
I actually created a 3D printed Brushless DC, Areal flux Pancake style, 6 coils, 8 magnets, 15AMP ESC.
I had the same issues you did with melting coils, so I "upgraded" to PetG, which helped to some degree. Issue is though, I couldn't find any metal version of PetG, so I also experimented with core metal screws to clamp the coils to the stator, which helped some with the tork and the starting issues, that i saw you had issues with too.
For heat dissipation I designed a version where the rotor assembly had small lamels shaped like wings to create airflow through the rotor and stator, in an effort to cool down windings... It kind of requires a pancake design to make room for the airflow in a direction that would help, and so now you can also get this in commercial BLDC's, which I think is a testimony that it actually would be a way to go too... I however never had the tools to get this version "good enough" to run.
Thanks a lot for providing the calculator I have been searching for that for a long time
Well done! The DIY BLDC is really impressive -- the benchmarks at the end are super helpful and will make me think twice about buying a COTS motor for some upcoming projects! (Also, what lake were you testing in? You were starting to hit Lake Erie depths! 😅)
A SUGGESTION FOR ACHIEVING HIGHER OVERALL PERFORMANCE: the slot setup you used to mount the Permanent Magnets increases the distance to the coils. The tighter you make that gap the better! almost all of the motor's performance will improve by making that gap smaller.
Ok but can we just talk about how deep the lake he's testing it in is? Like geez dude!
' Will it work?' Yes, of course it can basically work. The real problem is increasing gaps, etc to make a waterproof motor.
You're generally better off to have a sealed bubble for a dry motor, and then just a magnetic coupling for the shaft through the wall to spin the propeller.
There's very little gain in trying to make the motor itself waterproof, and so many ways to lose by trying to do it.
Excellent idea using the viewing film to check for field homogeneity
Michael, please put the loop in the Qsea’s tether around the handle!! The connector was never designed to take a load. When looped around the handle it has a 200lb rating.
Regarding the heat dissipation problem there are many good suggestions in the comments. One less good suggestion: consider making a variation of the components which melted that can be actively water-cooled (i.e. there are empty channels through the hot parts which you pump water through).
Or run your tests in a fish tank, at least.
I had dinner last night where you tested the RC car and I hiked this morning where you tested the motor underwater. Weird.
I think the most obvious suggestion would be to use a higher temp plastic for the stator. I'm not sure how the windings work, will have to check out that website.
If used underwater, that actually wouldn't be a problem. - Try that same design, on your sub? No way it can overheat. You could probably overdrive it. The issue is heat, and you have an entire lake of watercooling.
Really impressive that you've gotten a DIY brushless DC motor to work. And pretty cool that it can spin a propeller pretty well. I'm guessing this not intended as a practical use item - more of an interesting project right?
It spinning so slowly means your kv is pretty low. The main problem here (besides efficiency losses) is that normal brushless dc motors can get up to 90 degrees Celcius. 3D printing the stator means that it melts way before that, thus you can't get the amount of power throughput that you need. Making the stator out of metal would help...
Excellent job. Beautifully done motor and test tand.
first time watcher and instant sub and like just for your contribution to the 3d printing library! thank you!!
Those large diameter bearings will have far more resistance to turning than more standard bearings.
In my underwater vehicle, I usually use Chinese motors with a sealed stator. On my channel there is a video about my underwater vehicles and how we launched them in the Arctic. If you're interested, come in.
If you're interested, these are motors F2838-350kv
How do you solve the problem of the bearings? They should require many replaces with sea water and dirt.
This is not talked about in the video, but this is the only relevant topic, the rest is just a 3d printed normal BLDC...
@@faultboy In such motors, fluoroplastic gaskets are usually used instead of bearings. In an aquatic environment they provide sufficient gliding.
@@PolarUVTeamExactly what i plan to use, custom made on a lathe, thanks!
Copper bearings are also a good choice, as they are less afraid of sand in the water. Not easily oxidized even in sea water.
Go ahead and use the high current windings, but only run it submerged. As long as there's some flow near the windings, the water will carry the excess heat away.
I suspect the iron filled PLA has a really low % of iron. Else it wouldn't print well. Also, it should be a high permissivity alloy able to quickly switch pole polarity.
The principle is sound, but you shold use laminated plates of a ferrous material in the coils to keep it permanently working. This shold also provide a lot better effiviency.
Try to make axial flux electric motor, with your set of 3d printers it is doable task.
You need to use a thinner copper wire to minimize resistance and also you need to add cooling holes and maybe something that doesn’t insulate as well as pla
11:40 wow more than 100% efficiency for the second motor!
Inertia :)
Well I understand your desire to have a magnetic stator, I think I would also resin print that to help with the melting problem.
Radius. If you're not space constrained, wider means more torque. Litz wires (braid) to reduce EMF. Coils can spiral around the axis instead of being parallel
Cnc the stator out of metal and print the rotor. Most all motors generate heat, its benefital to add internal air fins incorporated into the inner rotor.
Some people suggest eddy currents as a problem - this is most certainly not correct as the iron filings in the metal pla are separated by... pla (famously an insulator)... thus the eddy currents in the grains of iron are incredibly small and would not produce basically any heating. Having said that, the PLA with iron filings is not that good of a material to use for a core, as the magnetic permeability is literally not much better than just air and pla on its own. It basically just adds weight and lower temperature stability for no gain. You might want to use PP filament if possible next time, as it melts at a much higher temperature, but with low thermal conductivity, any plastic is going to be fairly poor as core (and casing) material in the end.
I noticed the very low efficiency of the motor, a fair bit of it is probably to do with how large the air gap between the electromagnets and permanent magnets are. The larger the air gap the less of the available flux from permanent magnets you use. You definitely need to reduce it down by a lot, especially if you want to use a Halbach array. In addition, the motor looks quite wobbly with rather large bearings so there's probably some friction- and other big mechanical losses there. You might also want to consider a back-iron for the permanent magnets as well.
The winding configuration is probably fine, as the calculator uses the best practices.
Another thing to think about is that because your core material is very much sub-optimal, then long electromagnets like radial flux motors often go for, are going to be less efficient. You might want to look into making an axial flux motor instead, which is better suited for making out of low magnetic permeability materials.
He's me at nearly 20 years in engineering not understanding how this lad seems to have 90 years of engineering knowledge in the body of a 20 something year old
move the magnets on the rotor and the coils on the stator so you can use some transformer oil to cooldown the coils when the motor it is powered
Chris Rosser just designed one of the most powerful and efficient 5" quad motors. Id love to see him give you tips.
I have a bit of future waterproofing for you: Stranded wire has little itty bitty holes in the wire and water will creep through the wires SLOWLY so you will think everything is fine and then suddenly water gets into places you never thought it would get into and cause issues.
This is just ground breaking. There have been suggestions for the heating and balancing issues
My advice is to make a metal housing for it (like a tube) which lets water through (for cooling) then plate the inside with a diamagnetic material to confine the magnetic field to the stator and maximize flux density
resin is usually better tasting than distillate
Print the bell also from metal, it will reduce the magnetic field going outward from the motor
increase the space between magnets to half the width of the magnet. Your magnetic fields are fighting each other which is why you have to force it to spin at the start.
Changing the rotor mechanical design will help for submersible project cause water is denser than air, make the rotor a circular shaft rather than whole exo rotating this will reduce torque required and reach rpm fast and reduces current consumption.
The compact design seems to have left no room for cooling the trapped air inside the stator housing . (Under water and above ground designs will differ here) Directional cooling fins and good heat dissipating materials should help further your proof of concept . Nice work .
1. Quality of winding is crucial for efficiency - each deviation from wires going in paralel in same coil weakens magnetic field very much. Also make sure that centrifugal force does not mess up coil shape.
2. If possible - move magnets closer to stator. Magnetic field decreases by quare of distance, so each change, especially on small scales, makes big difference.
3. For heat - you can re-shape your rotating motor case holes so that each such hole acts as small fan blade that moves air inside the motor to cool it
Are you able to minimize the air gap between magnets and the stator?
IIRC, the rate of field strength decrease is greater for Halbachs as you move away from the magnets vs a N-S-... configuration and/or standard magnet. If you're going to put the effort into Halbach arrays, minimizing the air gap will be critical to utilizing the benefit provided by Halbachs.
i'm imagining that the most efficient way to move thru water is pulling/pushing thru it without disturbing it much, like fish and whales do.
but your motor is good too :)
for easier windings since your motor hub is so large you can just wind every pole by itself and attach them in the middle as a tri-star
It's interesting that your stator melted, but while eddie currents are often cited as the cause, i'd wager it's more likely to be simply the winding current you were seeing at full power. I think the more important thing to consider is what part of a BLDC motor needs to be waterproofed. With a bit of enamel or some manner of coating, the steel parts of a cots BLDC motor could be corrosion proofed. Making the stator out of steel laminates, and in fact making the rotor out of a steel ring, are very important to a motor's efficiency as those materials and their shapes help focus the EM fields in the most optimal locations. The amount of incohesive iron or steel in the plastic is likely not very useful, and the plastic rotor will do nothing to focus the M fields inward. The biggest issue with waterproofing a BLDC that I can see is the bearings. That is where something that can withstand corrosive emersion at depth is important. in your motor, you used unsealed bearings, so you were able to wash them afterwards, but even then, they will struggle to retain lubrication and in your post-water test, you could already see rust forming on one of the races. Sealed bearings do a much better job of retaining lubricants and keeping out debris, but even they are not typically fluid proof, nor able to withstand hydraulic pressures experienced at depth.
I still think this is really cool, and an awesome way to learn more about motors. I wonder how someone could make a BLDC motor with bearings that could withstand prolonged cycling in and out of water/saltwater. Keep it up, thanks for the video!
6:10 = I can already predict that water is going to seep in via the cable covering...
Yep. Need to terminate the insulation "hose" in the epoxy so that it doesn't bypass the seal. Strip the wires in the middle of the epoxy so that only the copper wire passes through all the way. Stranded wire? Epoxy or solder in between the strands will take care of that.
Amazing, i appreciated so much!!!
How it gets dark going underwater is terrifying to me.
bldc motors have no issues working under water, however, their bearings do !
print the stator in thin layers,get some vinyl or something else that insulating and put between stator layers, you could use thin layers of insulating enamel used for winding wires, then epoxy the stator together, this will get cut down on the eddycurrents or just send your stator file to send cut send get sponsored, so its free and laminate it together with electrical enamel.
what a time for a good video!
Have you thought about using that new unique propeller design to create a waterproof portable generator ?
Wow. I was looking for this content a week or two ago, and had to settle for watching a ton of different multiple youtube videos.
You've earned yourself a patreon
Fantastic work, Michael! Really well done! 😃
Stay safe there with your family! 🖖😊
Suggestions: you've got a nice Bambu, us their PC filament. It has a MUCH higher glass transition temperature.
I think I can give you some help with fluids, as that's closer to my area of expertise. You should really avoid spinning parts from free waterflow and there's plenty of effects messing the flow down the line, especially if the propeller comes after it, so put it in some casing with cooling ducts as necessary. You could also make the whole assembly smoother by making the cap fit perfectly to the propeller and have some rounded edges. Invest some time into CFD.
The pressure at 100 ft deep is almost 43 psi, so you can say that the motor can withstand this pressure condition
use your CNC to make a Iron stater, and replace all the other plastic 3D printed parts with high temp filaments.
wow this is amazing , i loved.
Electric motors have a little problem: magnets are double sided,not one sided. That means that an electric motors are actually not having a full power but around 50% depending on design. But,if you use bell electromagnets (they are used in electric cranes) you will get full 100% of magnetic field power. As a matter of fact,a 20 mm magnet with 100-200 rounds of 0,1 mm wire powered by 3 AA sized batteries connected in series can lift ~140 kilos.
Ok, so you want to make an underwater propulsion system? Why not make a super duper rim thruster?! I'd like to see one DIY made. There are companies that already make them but I'd like to see somebody like you make a smaller version. I'd actually like to see somebody make a rim driven lens for GoPros to spin the water off of the lens of the camera for water sports. Have fun! Cheers, John
Wow! Wow! Wow! really cool. It's like the motor on the "Make Sea" TH-cam channel. I 3D printed 8 so far. However, the magnets in my motor are much bigger than yours and you never want to go over 12v with a 3D-printed motor because they will always melt. 😎😎😎😎😎🤩
I'm not sure how similar electric motors are to solenoids (in physical & power calculations, but I think they are pretty similar) but I remember being amazed at the difference in force the solenoid could generate by reducing the distance (space) between the coil & the magnet/iron core. You want to do absolutely everything possible to get that gap as close to zero as possible and I'd suspect the same thing should be considered when designing the distance between the rotor to stator.
IDK if you would be able to turn either on a lathe to make the fit as close as you dare, I guess even using a drill press & sanding the piece might be a substitute for the lathe.
Would the motor benefit from any type of flux shielding?
A steel ring around the outside of the bell of the motor. And a way to increase the amount of iron in the stator.
You can't really expect an electric motor to be efficient and powerful with all these magnetic loses. Not sure about the performance of this iron filled PLA in this application but the rotor doesn't has any metal at all which leaves a big air gap (or plastic gap in this case) between the outer side of the magnets. The magnetic flux needs to run in a closed loop through a stator finger, the magnet , the rotor casing and then back via the neigbouring magnets and stator fingers. Every millimeter in this path that doesn't run through a ferromagnetic material adds to the losses. It may be worth and try to print the rotor case from iron filled PLA also. But you would probably also have to change the mounting of the magnets because the slots for the magnets would create a kind of shortcut between poth poles of the magnets. I'm pretty sure it's possible to improve efficency a lot but probably not to the level of a motor build from "proper" magnetic materials.
Instead of iron PLA, perhaps use inserts? Iron or laminated steel bars throughout the body with a petg or nylon frame.?..
Maybe JLCPCB can mill you some plates so you can assemble your stator for better magbetic conductivity
Magnets could be glued to the rotor rather than slotted in so they are touching to improve performance.
efficiency will never be anything unless you use laminated transformer armature or stater as they will direct the magnetic field most efficiently to the opposing magnets... Also bearings will get damaged underwater as there are many particles which will wear the bearings out even if they are ceramic apart from the friction of the water in between all the motor parts which are very close together generally....
Video idea: try building a B-17 esk plane with the same motors you used on the qaudcopter ?
Perhaps using 3D printing to make a mold to lost wax cast a pure iron stator as you’re obviously not getting optimal field off a iron infused pla filament.
Too bad you can't use PCB windings - but I think it would work fine.
Fishman depends on their stacked PCB tech and they're quite literally a magnetic transducer that requires fidelity - a coil emulator, if you will.
What about journal bearings/that also act as a cooler ? Essentially a turbo charger. Would still need something I guess for thrust and loads or extreme motion. It's a thought though.
How did you come up with your Slot to Pole ratio??
Always interesting, and thought provoking. Thanks.
You should use this for a regular electric jet engine, that would be dope af.
seems you should copy the design of a alternator ie has built in fan so win it spining it cooling it at the same time or something to help keep temps lower so you have air flow over the Stator windings
for the stator frame,
what if you 3d printed thinnish slices with that iron PLA and seperated those with something like aluminum foil.
Great video. Could you make a video about your data test rig and how to record that information in real time?
I didnt see if it already was or not but the effiency might be better if it had sensors
My drone motors work fine in the water til the bearings give out from the lubrication being displaced. I dunno if this was addressed in the video no time to watch currently but yeah brushless motors are good in water generally as long as the power source itself isn't being shorted which is a pretty easy thing to accomplish. So in reality I think all you did was make a custom motor that isn't as powerful or durable as motors you can buy. Pretty neat and fun stuff to do though.
7:16 you have your paddles the wrong way round :D Your less efficient while paddling.