Well, it's blue diode that took most time and effort to develop and overall brought us into the modern LED world, but with such logic you can't say that was a real invention since there were already red and green ones
@@SimbiontS95 the F are you even talking about, BLDC motor field is mature established industry, and this motor is nothing spectacular, interesting or out of the ordinary, blue LED by comparison required decades of refinement and netted a Nobel prize. 'Invention', blergh, this word just cheapens engineering effort.
@@VEC7ORlt yeah a brushless dc motor that is literally printed into a circuit board, not an invention and required no engineering or thought. ffs get real
@@tvcomputer1321it required plenty of engineering and thought. Just not much inventiveness. It's an interesting engineering challenge withoutmany applications as of yet.
When you showed the price I said, out loud, "Holy shit that's cheap". I can't believe you got the price that low for such a small scale product. You are doing amazing work!
@ I don’t think his product is outlandishly great on its own, but I was surprised at the price because he isn’t a million dollar company. Usually small scale products like this are a lot more expensive.
@@nikkiofthevalley i love what carl is doing but if your project require something l like this you could get 0802 fpv motor which is smaller in diameter and has much more power
Your aluminium rotors are black which means they are anodized. Anodization make surface spongy and make dimensions little bit bigger. Final dimension is unstable after anodization. Not only every batch, but every run in anodizing bath gives you another results. It expands aluminium because raw aluminium is converted in new chemical compound. If you want exact tolarance, you should make hole smaller, anodize part and resurface critical dimension up to desired tolerance. You really need dynamic balancing after completing rotors.
I was thinking this, and also eliminate the bearing to use a bushing. Tolerance can be more controlled, reliability and longevity increase, can have a much lower rotor height too.
This would be a great time to pitch JLCPCB and the other half dozen mainland manufacturers for sponsorship in an area that is clearly going to explode like GLP-1 analogs
@@CarlBugejaWhat training courses do you recommend? Also, do you know where one can find a EU certified course for led light panel fixtures for electricians? I"m stuck since my searches haven't yielded any good results.
>Spends countless hours designing >Spends countless hours testing >Spends countless hours refining >Makes it open source >Sells it for 5.99 >Calls it expensive >leaves >Gigachad energy
Did you ever look at Hard Disk Drive motor bushings? These also do not use a locknut as the magnetic field keeps it in place. Bushings are probably even better than bearings in this case.
This is peak RUclips! Showing young inventors and engineers creating the future right in front of our eyes. I’m currently working on ESP32 based platforms so will look into purchasing a few of your motors to have a play with!
I think this could be very useful for integrating small fans onto existing PCB designs. Place it right next to your heatsink or hot parts. Just need a good "fan rotor" now.
@roboman2444 6 hours ago - There are already multiple RUclips channnels focussed on PC fan design contests. How long before we see a similar motorcell fan competition?
Generally an amazing step forward in micro mechanical technology. You did an amazing job documenting the journey and giving us an excellent idea were we can take this. Thank you for all this hard work and I look forward to improvement and any of your next projects.
That light motor would be great with a light propeller, on a paper airplane. Use a small CAP for power, charge with USB. Kids would have a blast in class with those. Would be a good STEM experience for kids.
IRB60 and IRB6 - one of first industrial robots, made by Swedish ASEA (ABB) - they used solid magnets in rotors and flat coils in stators, the same construction - 1972! Technics RS-B100 deck 1984 - the same idea. Nihil novi sub sole ...
Move from knurled to toothed so you can have gear reduction. There's enough diameter change available to get a considerable torque multiplication. The added mass and frictional losses will hurt performance of course but there's always tradeoffs. Alternatively, 4 motors could be placed centrally with a reduction belt drive to radially placed rotors. Though, you might find it difficult to minimize PCB flex due to the required belt tension.
to get a torque increase you would need an much bigger second wheel / gear which would make this entire thing pointless as the whole appeal is the very small package size to get a 5x increase in torque you would need 5 times more teeth on the secondary gear and that would be very big indeed
@@SianaGearz yes that should be possible provided one can make all the parts for a reasonable price with reasonable accuracy it would still cost 3-4 more to make unless there is an off the shelf planetary that could be adapted to this application .
@@lelsewherelelsewhere9435 Harmonic drives are great, if you're not trying to spin a propeller. The main advantage of a harmonic drive is that you get an incredibly high reduction ratio in an absolutely tiny form factor. That's pretty much the only advantage they have, if you make one out of polymers the need to have parts flexing for basic functionality (which is how the whole thing works to begin with) means you waste a good portion of the already tiny amount of power you have available from such a tiny motor as this PCB-windings one. Might still be worth it if you're trying to make an ant-robot that has actual closed-loop positioning of it's limbs rather than a gait determined purely by the arrangement of a mechanical linkage, and doesn't just opt to use a much simpler and cheaper ready-built solution like a servo motor to do so. Neat, but impractical, esp. since the motor itself can already operate a "skid steer" type mobility system via direct drive, as demonstrated in the video itself.
Given the force/torque levels, it should be possible to replace the screw+nut bearing shaft by a soldered or clinched part. Its a light weight yet convenient enough option. If needed, wave springs can be used to handle backlash (for a thin double bearing assy)
Connect a small harmonic gear drive to it to increase the torque! It is super simple, like 3 pieces, its inline with the axis, and has a crazy high gear ratio!
At my last job, the predecessor had mounted ordinary ball bearings in the cnc tools. They said that the ball bearings break all the time, checked them and ordered high-speed bearings, they lasted for a few years instead of a few weeks. The right bearing with the right tolerance is important. He had also tried to put shims so that the bearings would fit where the old one had locked and spun, I don't need to tell you that I placed those tools in the scrap container. He had only looked at the price of bearings and taken the cheapest, but his poor knowledge cost the company a lot as the holders cost about 1800usd each. 6 holders were completely worn out due to locked bearings that spun on the holders. The right material for the right applications is important for durability and safety.
you'd think the company would have a specification sheet for maintaining the machines and a supply of parts on stock to replace worn out parts. rather than leaving it to presumably the operator, who hasn't necessarily been trained in machine maintenance... both my parents worked in a bakery factory. my mum was an operator, so she knew how to operate the machines to produce whatever they were producing, and she was a trainer so she could train other people to operate the machines. but she had no training in fixing anything, and so if anything ever went wrong with anything she'd call for an engineer or an electrician. seems like this would be standard operating procedure in most places.
(not an affiliate) Igus has triibological materials that are long wear life and self lubricating - and they also make a lot of them in 3D printer filament as well as bar/sheet stock, tapes, and powders for sintering applications. They make a big point of replacing traditional bearings in heavy industrial applications, so they'll probably work great in smaller applications. I've used them in interesting ways to make components that really simplify and strengthen designs/projects/automation fixturing in the lab at work and they've super attentive sales/field engineers who enjoy working on interesting projects.
I'm sure you've already thought about this, but what about mulit-layer PCBs so the coils can be stacked one on top of another for more windings and a stronger magnetic field.
It would be a balancing act as stonger field results in more torque but reduces top speed by increasing back EMF. It will also increase resistive losses, and inductance, not sure about the ratios and where the sweet spot would be. Actually balancing act is not quite the right word, it's an act of matching the motor characteristics to your load case. Not many things need a 36 kRPM motor, so trading some top end for more torque would make sense for a lot of use cases.
I'd also query whether the coils are done on both sides with an additional magnet on the other side thats attached to chassis to see if double torque can be produced?
This is absolutely fantastic! I find that this is why youtube started and what youtube core should be. I can only imagine that there were hundreds of hours that went into this video...Others would have made 100 videos about this instead of making a single video. I appreciate you doing this and sharing this with everybody.
Probably, we'll never see a flying drone with pcb motors because they have low energy conversion efficiency due to lack of a metal core directing magnetic fields from windings
You are a young man with initiative. I have been following your work for a long time and for this I sincerely say to you: RESPECT Soon your new inventions will find a special role in microrobotics. I wish you much success!
it's incredible how much engineer could involve just an haptic little motor and how much strive you have to overcome to setup little details like bearings and tolerances, good job!
As another person who attempted spinning things for a long ass time. My suggestion instead of a bearing with multiple balls with points of failure like dust and just wear use a needle and a jewl like in a mechanical watch or a screw for the pivot on the disc. The disc pivot can be a screw just make sure its a similar metal to the needle so galvanic reactions don't occur with it balanced right and up dust cant fall into it plus the only wear would be the needle it would take forever though and even if it wore out it would keep going and just be rounded off. Plus this approach still would keep your idea of a pcb motor intact all you would have to solder into place is a needle. "I know there would be like no torque nor can it be used like a wheel but if you just wanna spin something kinda fast and upright its a old school method but it works." Don't mean to talk about my build but plan to have a video on the thing I described soon too if my comment is a bit confusing to read.
@@MBunn-uf1we Brilliant, but expensive. The air bearing I saw was nearly frictionless, but it required compressed air to be fed into the interface. They're made from extremely precise parts, and precision is expensive.
@@MBunn-uf1we Graphine bearing is also good but you need a strong magnetic field and it floats like a super conductor with out sub freezing temperatures. Only problem you need a equal north south magnetic field then having to fit in his case a pcb coil under that. But a ring magnet and I hope I remember it right graphine can produce solid levitation that does not need power but whether his coils can push it after would need trial and error.
@@machinemaker2248 some designs do not require compressed air; they are shaped in such a way that they can trap air as they move, with the caveat that they have high friction at rest.
This whole series of videos has. Even incredible! The creativity, methodology, engineering, excitement... it's all so pleasing to me. So many cool potential applications. The idea of some fairly strong haptic feedback integrated right into a PCB is really interesting!
I met with a few PCR stator motor/axial flux motor companies at my previous company and the tech is the real deal. I can’t wait to see this stuff used full scale in the future.
+/- 0.05 is crazy for a press for like that on a bearing with radial clearances in the microns. There is no reason you can't get +/-0.01 on a 5mm bore economically. I also don't see why your other dimensions are so tightly toleranced, that's potentially adding unnecessary cost.
That's a good point, boring the center is easier than getting tight tolerances on the outsides but there is no real need for the outside to be that tight since it's just getting a 3D print slapped over it anyways.
I 've seen all your work along years with these motor, and it's awsome! May I make a humble contribution: 1 - I saw you using a caliper, but for this project is better change for a micrometer; 2 - all the alumínium disk must be balanced and test before endurece testing; 3 - low tolerances will be costly, an adaptative solution like a epoxi glue or a compression sleeve betewn the disk and the bearing maybe will be more suitable; 3 - groves at the disk maybe need to be hand made to improve gripping. Someone told above the problem with the anodizing. Consider too a design that reduce this weight, like 2 or 3 arms wheel. By the way these disks are casting or machining? 4 - insertion of bearings it's tough! Try long freezing for one piece and heating the other with lubrication (aluminiun will expande more than stell...); Good Lucky and dont give up!
for the aluminum rotors, if you have the ability to put the finish you need back on the surface (anodizing it or whatever), you could intentionally order rotors with a smaller inner diameter than you want and then just finish it up to a really tight tolerance yourself, and then you could have more flexibility in what manufacturers you can use
Ive seen youtubers sell cad files for more than this motor, AND the stator is open source??????? Gold standard! If I can think of something cool to build with it I'll have to make a video.
I remember back in 2000 a buddy designed a highly controllable micron scale oscillator . He was seriously thinking about a vibrating toothbrush as his first product when someone mentioned that fiber optic wires were really hard to align. That idea alone took the project in a completely different direction.
Could also work as a control moment gyro. Even if the torque is low, over time it'll be able to spin up a decent mass to high speeds. Then you could use a higher power servo to rotate the entire PCB
@VEC7ORlt I'd argue having both would be useful. You could use the reaction wheels or CMGs for quick large movements, and use the magnetorquer to desaturate.
@@VEC7ORlt can't use magnetorquers for precision pointing or fast movement, and they also require fairly accurate knowledge of position around the globe and a good Earth magnetic model on board (it's way messier than most people think).
It's an interesting idea. The Panasonic Twin Rotor Motor was a very slim motor, only a few millimetres in thickness, in compact cassette players made by Panasonic in the 90s, that had very similar looking PCB-based coils on them. While there isn't much information on them anymore, they appeared to operate in a similar way to yours.
If you're chasing tighter tolerances i would not measure them with the back of a caliper like that at 8:16 there is a flat side there that will interfere with getting a correct dimension. IMO you should consider investing in "small hole gauges" and using a micrometer as the gauges are rounded to get the very tangent of the hole. Best of luck in this project.
I recently jumped to studying PCB design. I thought making a circuit more compact was the only way to utilize PCB. this video really opened my eyes to what PCB can do
A few questions, if you have the time: - What does the back-EMF look like? Trapezoidal or sinusoidal? - It looks like you made early design decisions based on no-load RPM. Have you gone back and tried e.g. the 8-pole design with a load and measured dynamic response? - Have you considered sensorless FOC? If so, what made you choose trapezoidal? BOM cost?
I notice in your application you use a conventional fan. You should look at centrifugal fans as their airflow is the square of their rotational speed. This is why Dyson uses miniature brushless motors to achieve VERY high airflow.
PCB motors have been in existence for decades. Floppy disk drives used them, video cassette recorders used them on the head drums, CD players among other things. They're not as new as you suggest. Nevertheless you're doing some interesting stuff. Good luck with your projects.
Very thorough R&D! Nicely done. Unless you already tried and found some source of failure, one way to lower the profile is to replace the lock nut with a flatter nut and just put some thread locking compound like "locktite". Nail polish or varnish would work too but might make it harder to remove the nut later.
A few ideas that came to my mind after watching your video. 1. Why ball bearings and not needle bearings, or a pair of needle + flat thrust? Needle bearings can be much smaller in size because they can use the shaft itself instead of an internal part. 2. How about using it where you need to hold and vice versa, low speed? For example, tuning knobs with feedback, or micro-mounts for a micro-camera? 3. What about increasing the thickness of the tracks, how does this affect efficiency?
Motors like this could be very useful in space applications, especially when weight constraint are one of the biggest problems. Think "reaction wheels in cubesats" or one time applications like extending booms and other deployment tasks after launch. With cheap motors like this that can be directly integrated into existing PCB designs (and similar inventions by other garden shed inventors), this will allow hobby groups and small universities to afford building more complex spacecraft within cubesat specs.
Nice! I designed a stirrer motor like this in the early 1990s using wound coils and discrete logic. It's good to see how technology has progressed; I have often thought of recreating my motor for home use.
Use a pully and send the cable back around the generator so it generates power when the branch rebounds, so you don't lose half of the potential power generation. 1. From branch to pawed wheel 2. From pawed wheel to reversing pulley 3. From reversing pulley back to a second pawed wheel but wrapped in the opposite direction, 4. From second pawed wheel to your weight. This will ensure it generates power 100% of the time, rather than wasting the potential power generation when the branch rebounds.
Fantastic work and congratulations on your wedding! Now I'm eager to mate a couple of your motors onto a micro'Bot like you demonstrated, but with some feathers/tassels attached and possibly a chirping/tweeting piezo-speaker. My cat will go nuts over this! Hell, go ahead and market some of these as cat toys...you'd make a killing in no time.
I started watching the video and thought to myself "that accent is so familiar!" Then I saw the name :) It feels good to see something so cool coming from our country :) 🇲🇹 Keep up to good work!
I am super excited to see how far this has come and once I get situated in a few months I am definitely going to be experimenting with these (or their successors) to try my hand at a micro-drone. Bravo and keep it up!
Great work. Consider putting a stator on either side of the rota to double the torque. It would limit it's application but would still be useful in many areas.
Bearings have a limited amount of turns in them, in maintenance we calculate how long they will life given their load, rpm and pre-load from the factory. Manufacturers have guides for this
thats not expensive atall ...im a starving artist and i could even get some at that price...also the inbeded driver is a godsend to me ...i cant code so making it as simple as possible is so nice
I do enjoy seeing firsthand why certain items are expensive (the bearing) and how the quality actually does matter, even in "hobbyist" applications (aka I see it on RUclips instead of never hearing about it and it just goes into some military machine)
Looks very promising Carl. Interesting to apply in one of my prototype projects. Just ordered 2 pieces. Looking forward to get practical experience with them.
In astrophotography we are using special cameras that need to be cooled at -30 degrees Celsius, this is done by gluing a peltier cooler to the sensor or the pcb. Maybe a good ideea to try it out in order to cool the motor at a manageable level
Your PCB motors remind me a lot of the approach that Fishman uses for their Fluence pickups (electric guitar pickups are pretty much just motors). They use stacks of 48 PCB coils, that I'm pretty sure are connected but vias or similar, wrapped around a ferromagnetic core.
OMG, I just watched a video here on RUclips about using analog signals to help boost clock speeds in digital applications as nano-manufacturing costs keep rising. Having a billion resolutions per second must have some exciting applications in this research. Small robots for espionage and exploration might actually become a reality!
This stuff is awesome! And I really, really advise looking into licensing this technology for commercial use. That is to say, in addition to parenting what you can, get a senior engineer from industry that can sell licenses of this tech to existing companies. Fans, robots, industrial equipment, HMI, and a hundred other companies would benefit from being able to buy permission to use what you've spent so much time and money learning how to do. Best of luck!
Stack two, one on top and one on the bottom slightly rotated for physical phase offset. This might be enough for lift then, especially if you tweak the phase/amplitude over time across the top and bottom for adaptive climbing rpm.
Good to see actual inventors at work, hope your project will be used in many awesome real-world applications, it deserves recognition
'inventors' I hate this word with passion.
This is not an invention, just a garden variety BLDC motor with an unusual winding.
Well, it's blue diode that took most time and effort to develop and overall brought us into the modern LED world, but with such logic you can't say that was a real invention since there were already red and green ones
@@SimbiontS95 the F are you even talking about, BLDC motor field is mature established industry, and this motor is nothing spectacular, interesting or out of the ordinary, blue LED by comparison required decades of refinement and netted a Nobel prize.
'Invention', blergh, this word just cheapens engineering effort.
@@VEC7ORlt yeah a brushless dc motor that is literally printed into a circuit board, not an invention and required no engineering or thought. ffs get real
@@tvcomputer1321it required plenty of engineering and thought. Just not much inventiveness. It's an interesting engineering challenge withoutmany applications as of yet.
"I decided to study PCB stators on my honeymoon" 😂😅
A girl who is OK with that is truly a dream one )
you dirty minda I'm watching you 😅
@@ccflan Make for a set of gloves with vibrating fingertips.... Happy wife, happy life methodology.
Plot twist his wife suddenly become a geek after first night lol
I mean going by this guys content yeah no it sounds like she’s the right one for him. I bet she supported it
When you showed the price I said, out loud, "Holy shit that's cheap". I can't believe you got the price that low for such a small scale product. You are doing amazing work!
Those are 13 dollars each. You can buy normal motors for less than that.
@ I don’t think his product is outlandishly great on its own, but I was surprised at the price because he isn’t a million dollar company. Usually small scale products like this are a lot more expensive.
@@Anon-te6uqYes, but buying normal motors this compact...?
If you were to buy a million of them, the price would probably be 50 cents
@@nikkiofthevalley i love what carl is doing but if your project require something l like this you could get 0802 fpv motor which is smaller in diameter and has much more power
Your aluminium rotors are black which means they are anodized. Anodization make surface spongy and make dimensions little bit bigger. Final dimension is unstable after anodization. Not only every batch, but every run in anodizing bath gives you another results. It expands aluminium because raw aluminium is converted in new chemical compound. If you want exact tolarance, you should make hole smaller, anodize part and resurface critical dimension up to desired tolerance. You really need dynamic balancing after completing rotors.
Great point!
I wonder if it outweighed by the slight boost in heatsinking efficiency the black coating gives.
Could have gotten a reamer to fix the size also
I was thinking this, and also eliminate the bearing to use a bushing. Tolerance can be more controlled, reliability and longevity increase, can have a much lower rotor height too.
@@inkrpen It will work but only few times. Aluminium oxide is very hard and dulls standard HSS hand tools very quickly.
That is the greatest PCBWay ad i ever seen. the amount of designs you pump out is truly incredible.
This would be a great time to pitch JLCPCB and the other half dozen mainland manufacturers for sponsorship in an area that is clearly going to explode like GLP-1 analogs
You do never stop to amaze us, Carl! We just can't thank you enough for your support throughout the years❤! Happy Holiday🎄🎄🎄
what a pathetic corporate message LOL
Oh i see a communist triggered. north korea is a nice place to live @chimpo131
Relax bro @@chimpo131
Reduce shipping cost in asia, come on, do something
Are you a patreon? Are you supporting Carl's projects? If not, save it. @@chimpo131
Well done! I was wondering when we would see this. Totally worth the wait.
Thanks! It was really hard to summarise everything into one enjoyable video but I'm happy with how it turned out!
@@CarlBugejaReally impressed by the results, i def buy it in the future if our startup works out
@@CarlBugeja noob, you made high speed low torque motor why???? make high torque low speed motors, more useful for robots
@@CarlBugejaWhat training courses do you recommend?
Also, do you know where one can find a EU certified course for led light panel fixtures for electricians? I"m stuck since my searches haven't yielded any good results.
>Spends countless hours designing
>Spends countless hours testing
>Spends countless hours refining
>Makes it open source
>Sells it for 5.99
>Calls it expensive
>leaves
>Gigachad energy
bruh
Did you ever look at Hard Disk Drive motor bushings? These also do not use a locknut as the magnetic field keeps it in place. Bushings are probably even better than bearings in this case.
Sintered bronze impregnated with oil.
Either that or magnetic/ air bushings which have zero contact between surfaces
@@joefish6091 Exactly! This is the key in my opinion. Does require a hard chrome smooth shaft though for longevity.
A Japanese company dominated that space. I don't know if they still are in business but that gets you started.
FD bearings
This is peak RUclips! Showing young inventors and engineers creating the future right in front of our eyes.
I’m currently working on ESP32 based platforms so will look into purchasing a few of your motors to have a play with!
I think this could be very useful for integrating small fans onto existing PCB designs. Place it right next to your heatsink or hot parts. Just need a good "fan rotor" now.
@roboman2444
6 hours ago - There are already multiple RUclips channnels focussed on PC fan design contests. How long before we see a similar motorcell fan competition?
not sure why we'd use that obsolete crap considering we already have solid state chips that blow air.
There is a fan design specifically for these. The files for printing them are on the microbots page where these motors are for sale.
@@ermakers1297 there are also piezo fans, as well. i'm still buying myself these for xmas, though
Danke!
Amazing work, Thanks for showing some of the R&D details. It showcases the rigor of creating a good product.
Generally an amazing step forward in micro mechanical technology. You did an amazing job documenting the journey and giving us an excellent idea were we can take this. Thank you for all this hard work and I look forward to improvement and any of your next projects.
That light motor would be great with a light propeller, on a paper airplane. Use a small CAP for power, charge with USB. Kids would have a blast in class with those. Would be a good STEM experience for kids.
Not the right dimensions for a paper plane.
@@GelloMello-j9zwell, pack it up then, not like there are any engineers around here that can change form factor of a motor
11:55 not really "new". I have a cassette player from 2001 which motor has its coils printed onto PCB. It was done in mass production ~20 years ago.
IRB60 and IRB6 - one of first industrial robots, made by Swedish ASEA (ABB) - they used solid magnets in rotors and flat coils in stators, the same construction - 1972! Technics RS-B100 deck 1984 - the same idea. Nihil novi sub sole ...
Move from knurled to toothed so you can have gear reduction. There's enough diameter change available to get a considerable torque multiplication. The added mass and frictional losses will hurt performance of course but there's always tradeoffs. Alternatively, 4 motors could be placed centrally with a reduction belt drive to radially placed rotors. Though, you might find it difficult to minimize PCB flex due to the required belt tension.
to get a torque increase you would need an much bigger second wheel / gear
which would make this entire thing pointless as the whole appeal is the very small package size
to get a 5x increase in torque you would need 5 times more teeth on the secondary gear
and that would be very big indeed
@@Crushonius What if it was a planetary and the rotor was the sun gear.
@@SianaGearz yes that should be possible
provided one can make all the parts
for a reasonable price
with reasonable accuracy
it would still cost 3-4 more to make
unless there is an off the shelf
planetary that could be adapted to
this application .
Even better, harmonic drive. This has insane gear ratio, and is super simple.
@@lelsewherelelsewhere9435
Harmonic drives are great, if you're not trying to spin a propeller.
The main advantage of a harmonic drive is that you get an incredibly high reduction ratio in an absolutely tiny form factor.
That's pretty much the only advantage they have, if you make one out of polymers the need to have parts flexing for basic functionality (which is how the whole thing works to begin with) means you waste a good portion of the already tiny amount of power you have available from such a tiny motor as this PCB-windings one. Might still be worth it if you're trying to make an ant-robot that has actual closed-loop positioning of it's limbs rather than a gait determined purely by the arrangement of a mechanical linkage, and doesn't just opt to use a much simpler and cheaper ready-built solution like a servo motor to do so.
Neat, but impractical, esp. since the motor itself can already operate a "skid steer" type mobility system via direct drive, as demonstrated in the video itself.
Given the force/torque levels, it should be possible to replace the screw+nut bearing shaft by a soldered or clinched part. Its a light weight yet convenient enough option.
If needed, wave springs can be used to handle backlash (for a thin double bearing assy)
Connect a small harmonic gear drive to it to increase the torque!
It is super simple, like 3 pieces, its inline with the axis, and has a crazy high gear ratio!
That sound at the end, of the motors spinning down, that's super cool!
I've watched your channel for a long time. It's great to see you keep going until you reached a useful outcome! Well done!
At my last job, the predecessor had mounted ordinary ball bearings in the cnc tools. They said that the ball bearings break all the time, checked them and ordered high-speed bearings, they lasted for a few years instead of a few weeks. The right bearing with the right tolerance is important. He had also tried to put shims so that the bearings would fit where the old one had locked and spun, I don't need to tell you that I placed those tools in the scrap container. He had only looked at the price of bearings and taken the cheapest, but his poor knowledge cost the company a lot as the holders cost about 1800usd each. 6 holders were completely worn out due to locked bearings that spun on the holders. The right material for the right applications is important for durability and safety.
you'd think the company would have a specification sheet for maintaining the machines and a supply of parts on stock to replace worn out parts. rather than leaving it to presumably the operator, who hasn't necessarily been trained in machine maintenance...
both my parents worked in a bakery factory. my mum was an operator, so she knew how to operate the machines to produce whatever they were producing, and she was a trainer so she could train other people to operate the machines. but she had no training in fixing anything, and so if anything ever went wrong with anything she'd call for an engineer or an electrician. seems like this would be standard operating procedure in most places.
(not an affiliate) Igus has triibological materials that are long wear life and self lubricating - and they also make a lot of them in 3D printer filament as well as bar/sheet stock, tapes, and powders for sintering applications.
They make a big point of replacing traditional bearings in heavy industrial applications, so they'll probably work great in smaller applications.
I've used them in interesting ways to make components that really simplify and strengthen designs/projects/automation fixturing in the lab at work and they've super attentive sales/field engineers who enjoy working on interesting projects.
The friction is too high in IGUS plane bearings
I'm sure you've already thought about this, but what about mulit-layer PCBs so the coils can be stacked one on top of another for more windings and a stronger magnetic field.
It would be a balancing act as stonger field results in more torque but reduces top speed by increasing back EMF. It will also increase resistive losses, and inductance, not sure about the ratios and where the sweet spot would be.
Actually balancing act is not quite the right word, it's an act of matching the motor characteristics to your load case. Not many things need a 36 kRPM motor, so trading some top end for more torque would make sense for a lot of use cases.
I'd also query whether the coils are done on both sides with an additional magnet on the other side thats attached to chassis to see if double torque can be produced?
This is absolutely fantastic! I find that this is why youtube started and what youtube core should be. I can only imagine that there were hundreds of hours that went into this video...Others would have made 100 videos about this instead of making a single video. I appreciate you doing this and sharing this with everybody.
That's epic man! Congratulations! I've been following your work on pcb actuators since you started, so wonderful to see what you've acheived
Hope to see the drone soon!
Probably, we'll never see a flying drone with pcb motors because they have low energy conversion efficiency due to lack of a metal core directing magnetic fields from windings
I actually came across with PCB motors when I bought supersmall 10x10mm fans, they use exact same motor design
You are a young man with initiative. I have been following your work for a long time and for this I sincerely say to you: RESPECT Soon your new inventions will find a special role in microrobotics. I wish you much success!
it's incredible how much engineer could involve just an haptic little motor and how much strive you have to overcome to setup little details like bearings and tolerances, good job!
Excellent work, great presentation/explanation. Please wear safety glasses. :)
glad you enjoyed it - thank you!
Love to see this done. Amazing work!
This is really cool. Thank you for your hard work and time spent on making such a neat PCB motor and sharing the process with us! :D
As another person who attempted spinning things for a long ass time. My suggestion instead of a bearing with multiple balls with points of failure like dust and just wear use a needle and a jewl like in a mechanical watch or a screw for the pivot on the disc. The disc pivot can be a screw just make sure its a similar metal to the needle so galvanic reactions don't occur with it balanced right and up dust cant fall into it plus the only wear would be the needle it would take forever though and even if it wore out it would keep going and just be rounded off. Plus this approach still would keep your idea of a pcb motor intact all you would have to solder into place is a needle.
"I know there would be like no torque nor can it be used like a wheel but if you just wanna spin something kinda fast and upright its a old school method but it works."
Don't mean to talk about my build but plan to have a video on the thing I described soon too if my comment is a bit confusing to read.
The jewel bearings is a great idea.
I wonder if a gas/air bearing could work instead?
@@MBunn-uf1we Brilliant, but expensive. The air bearing I saw was nearly frictionless, but it required compressed air to be fed into the interface. They're made from extremely precise parts, and precision is expensive.
@@MBunn-uf1we Graphine bearing is also good but you need a strong magnetic field and it floats like a super conductor with out sub freezing temperatures. Only problem you need a equal north south magnetic field then having to fit in his case a pcb coil under that. But a ring magnet and I hope I remember it right graphine can produce solid levitation that does not need power but whether his coils can push it after would need trial and error.
@@machinemaker2248 some designs do not require compressed air; they are shaped in such a way that they can trap air as they move, with the caveat that they have high friction at rest.
Thank you for helping humanity inch forward on our technological journey, you are doing important work my friend!
High respect for passion and sustainability in pursuit of perfect result 💪
Це просто дивовижно, скільки терпіння та часу ти вклав в цей проект. Натхнення та яскравих ідей тобі. Ти просто супер.
This whole series of videos has. Even incredible! The creativity, methodology, engineering, excitement... it's all so pleasing to me.
So many cool potential applications. The idea of some fairly strong haptic feedback integrated right into a PCB is really interesting!
I met with a few PCR stator motor/axial flux motor companies at my previous company and the tech is the real deal. I can’t wait to see this stuff used full scale in the future.
+/- 0.05 is crazy for a press for like that on a bearing with radial clearances in the microns. There is no reason you can't get +/-0.01 on a 5mm bore economically. I also don't see why your other dimensions are so tightly toleranced, that's potentially adding unnecessary cost.
That's a good point, boring the center is easier than getting tight tolerances on the outsides but there is no real need for the outside to be that tight since it's just getting a 3D print slapped over it anyways.
I 've seen all your work along years with these motor, and it's awsome!
May I make a humble contribution:
1 - I saw you using a caliper, but for this project is better change for a micrometer;
2 - all the alumínium disk must be balanced and test before endurece testing;
3 - low tolerances will be costly, an adaptative solution like a epoxi glue or a compression sleeve betewn the disk and the bearing maybe will be more suitable;
3 - groves at the disk maybe need to be hand made to improve gripping. Someone told above the problem with the anodizing. Consider too a design that reduce this weight, like 2 or 3 arms wheel. By the way these disks are casting or machining?
4 - insertion of bearings it's tough! Try long freezing for one piece and heating the other with lubrication (aluminiun will expande more than stell...);
Good Lucky and dont give up!
0:32 Bless you both! Romance is not dead!
for the aluminum rotors, if you have the ability to put the finish you need back on the surface (anodizing it or whatever), you could intentionally order rotors with a smaller inner diameter than you want and then just finish it up to a really tight tolerance yourself, and then you could have more flexibility in what manufacturers you can use
Ive seen youtubers sell cad files for more than this motor, AND the stator is open source??????? Gold standard! If I can think of something cool to build with it I'll have to make a video.
Have you thought of using it to drive a flywheel for Cube Sats.
I remember back in 2000 a buddy designed a highly controllable micron scale oscillator . He was seriously thinking about a vibrating toothbrush as his first product when someone mentioned that fiber optic wires were really hard to align. That idea alone took the project in a completely different direction.
Could also work as a control moment gyro. Even if the torque is low, over time it'll be able to spin up a decent mass to high speeds. Then you could use a higher power servo to rotate the entire PCB
Why do you even need flywheels for - just use a magnetorquer.
@VEC7ORlt I'd argue having both would be useful. You could use the reaction wheels or CMGs for quick large movements, and use the magnetorquer to desaturate.
@@VEC7ORlt can't use magnetorquers for precision pointing or fast movement, and they also require fairly accurate knowledge of position around the globe and a good Earth magnetic model on board (it's way messier than most people think).
I love seeing updates on your projects. Always fascinating. Great work!
It's an interesting idea. The Panasonic Twin Rotor Motor was a very slim motor, only a few millimetres in thickness, in compact cassette players made by Panasonic in the 90s, that had very similar looking PCB-based coils on them. While there isn't much information on them anymore, they appeared to operate in a similar way to yours.
potential use cases. fly wheels, mini drone motors, reaction wheel
If you're chasing tighter tolerances i would not measure them with the back of a caliper like that at 8:16 there is a flat side there that will interfere with getting a correct dimension. IMO you should consider investing in "small hole gauges" and using a micrometer as the gauges are rounded to get the very tangent of the hole. Best of luck in this project.
I recently jumped to studying PCB design. I thought making a circuit more compact was the only way to utilize PCB. this video really opened my eyes to what PCB can do
A few questions, if you have the time:
- What does the back-EMF look like? Trapezoidal or sinusoidal?
- It looks like you made early design decisions based on no-load RPM. Have you gone back and tried e.g. the 8-pole design with a load and measured dynamic response?
- Have you considered sensorless FOC? If so, what made you choose trapezoidal? BOM cost?
Dental tools have amazing bearings. 80k rpm is common. Have you looked into them?
Dental tools has far more than 80k rpm. Newest models can beat 500k! and dental drills are powered by air turbine.
I notice in your application you use a conventional fan.
You should look at centrifugal fans as their airflow is the square of their rotational speed. This is why Dyson uses miniature brushless motors to achieve VERY high airflow.
PCB motors have been in existence for decades. Floppy disk drives used them, video cassette recorders used them on the head drums, CD players among other things. They're not as new as you suggest. Nevertheless you're doing some interesting stuff. Good luck with your projects.
i see some vhs with planar motor like this pcb motor
i love projects like this because its very obvious all the things this can change
This man single-handedly is making a PCB motor manufacturing company
AMAZING video guys!! Integrated driver is awesome! knurling was a brilliant solution! LIKED AND SUBSCRIBED!
Very thorough R&D! Nicely done. Unless you already tried and found some source of failure, one way to lower the profile is to replace the lock nut with a flatter nut and just put some thread locking compound like "locktite". Nail polish or varnish would work too but might make it harder to remove the nut later.
Incredible work! I would love to see a tiny self balancing demo. Seem like they would make good gyroscopes
A few ideas that came to my mind after watching your video.
1. Why ball bearings and not needle bearings, or a pair of needle + flat thrust? Needle bearings can be much smaller in size because they can use the shaft itself instead of an internal part.
2. How about using it where you need to hold and vice versa, low speed? For example, tuning knobs with feedback, or micro-mounts for a micro-camera?
3. What about increasing the thickness of the tracks, how does this affect efficiency?
Wow really inspiring to see this process end to end and congratulations on shipping! You should be super proud.
Motors like this could be very useful in space applications, especially when weight constraint are one of the biggest problems. Think "reaction wheels in cubesats" or one time applications like extending booms and other deployment tasks after launch. With cheap motors like this that can be directly integrated into existing PCB designs (and similar inventions by other garden shed inventors), this will allow hobby groups and small universities to afford building more complex spacecraft within cubesat specs.
You are really close to a fully fledged product. Congratulations.
Nice! I designed a stirrer motor like this in the early 1990s using wound coils and discrete logic. It's good to see how technology has progressed; I have often thought of recreating my motor for home use.
this would be SUPER useful for cubesat reaction wheels
It's crazy how tiny changes can make the difference between useless and useful
I love seeing RUclipsrs actually advance our technology.
How about Gyroscopes for low gravity? Also have you considered any gearing or combining/stacking of motors through a common shaft?
"I love how these DIY ideas make everyday life so much easier!"
These motors seem like a great option for a micro gyroscope, or for reaction wheels.
Use a pully and send the cable back around the generator so it generates power when the branch rebounds, so you don't lose half of the potential power generation.
1. From branch to pawed wheel
2. From pawed wheel to reversing pulley
3. From reversing pulley back to a second pawed wheel but wrapped in the opposite direction,
4. From second pawed wheel to your weight.
This will ensure it generates power 100% of the time, rather than wasting the potential power generation when the branch rebounds.
thank you for making it both open source and available to buy to give the option
Fantastic work and congratulations on your wedding! Now I'm eager to mate a couple of your motors onto a micro'Bot like you demonstrated, but with some feathers/tassels attached and possibly a chirping/tweeting piezo-speaker. My cat will go nuts over this!
Hell, go ahead and market some of these as cat toys...you'd make a killing in no time.
Have you considered using the pcb motor with a built in cycloid gear to get more torque?
i have but still not sure if its feasible
I started watching the video and thought to myself "that accent is so familiar!" Then I saw the name :) It feels good to see something so cool coming from our country :) 🇲🇹 Keep up to good work!
I am super excited to see how far this has come and once I get situated in a few months I am definitely going to be experimenting with these (or their successors) to try my hand at a micro-drone. Bravo and keep it up!
Great work! It must have been very frustrating at times
It was but I am proud with the end results 🙂
What a journey Carl! 🤯
I think it might be time to start building your own PCBs just to experiment with what is possible. Congratulations on the successful tests🎉
Great work. Consider putting a stator on either side of the rota to double the torque. It would limit it's application but would still be useful in many areas.
Bearings have a limited amount of turns in them, in maintenance we calculate how long they will life given their load, rpm and pre-load from the factory. Manufacturers have guides for this
Pastizzi powa!! Well done.
Amazing R&D work Carl! Great job!
"Printed Circits, Mozarella, PEPPERONI"🤣🤣
Bravo from Morocco , i think your pcb motor will be the perfect joint actuator for the future insect robots
thats not expensive atall ...im a starving artist and i could even get some at that price...also the inbeded driver is a godsend to me ...i cant code so making it as simple as possible is so nice
I do enjoy seeing firsthand why certain items are expensive (the bearing) and how the quality actually does matter, even in "hobbyist" applications (aka I see it on RUclips instead of never hearing about it and it just goes into some military machine)
Have you considered running a RUclips competition to encourage innovative new use cases for the motor cell?
Looks very promising Carl. Interesting to apply in one of my prototype projects. Just ordered 2 pieces. Looking forward to get practical experience with them.
In astrophotography we are using special cameras that need to be cooled at -30 degrees Celsius, this is done by gluing a peltier cooler to the sensor or the pcb. Maybe a good ideea to try it out in order to cool the motor at a manageable level
Your PCB motors remind me a lot of the approach that Fishman uses for their Fluence pickups (electric guitar pickups are pretty much just motors). They use stacks of 48 PCB coils, that I'm pretty sure are connected but vias or similar, wrapped around a ferromagnetic core.
OMG,
I just watched a video here on RUclips about using analog signals to help boost clock speeds in digital applications as nano-manufacturing costs keep rising.
Having a billion resolutions per second must have some exciting applications in this research.
Small robots for espionage and exploration might actually become a reality!
beautiful work, and thanks for sharing!
Great Job Carl 👍. Well done, well thought trough, great contribution to DIY-Projects.
This stuff is awesome! And I really, really advise looking into licensing this technology for commercial use. That is to say, in addition to parenting what you can, get a senior engineer from industry that can sell licenses of this tech to existing companies. Fans, robots, industrial equipment, HMI, and a hundred other companies would benefit from being able to buy permission to use what you've spent so much time and money learning how to do.
Best of luck!
This is great. I imagine scaling up the size could enhance torque.
RUclips nailing the awesome tech recommendations lately.
Very nice to see your advencement!
Thanks for sharing
Get yourself some safety glasses with all this spinning things :) Crazy work, keep it up !
This is incredible! I hope that your sales go well, and that we see some really cool applications for this!
Stack two, one on top and one on the bottom slightly rotated for physical phase offset. This might be enough for lift then, especially if you tweak the phase/amplitude over time across the top and bottom for adaptive climbing rpm.