600 Watt, 3d-printed, Halbach Array, Brushless DC Electric Motor [video] | Hacker News Reader

600 Watt, 3d-printed, Halbach Array, Brushless DC Electric Motor [video]

youtube.com

147 pointsdonquichotte49 comments

We really need a cheap online laser cutting magnetic metal lamination service. There are more and more online laser cutting services, but they don't stock magnetic lamination steel. The only online specialized service is incredibly expensive.

We are in a golden age of electric motors, but access to custom laminations is a limiting factor for tinkerers (now that the magnet are widely available).

What would go into building something like this?

stock management, and expedition. I couldn't find the special steel distribution in less than 50 metric tons rolls from China. Need a few thickness and a few "qualities" in store.

I think the issue is that anything involving custom magnetics is a niche area for hobbyists right now. I think most hobbyists and tinkerers prefer off-the-shelf parts to one-off motors, transformers, inductors, etc.

That being said, I've already seen quite a few videos of people winding their own magnetics, so perhaps there is a market to be capitalized on.

no, I think the RC modeling people are quite numerous, and quite into custom motors, I think they are the ones who made the electric skateboards and scooters possible, BLDC control was only for serious business before them. Then, they started rewinding their motors. And after that they started putting them backwards (outrunners), by recycling brushed DC rotor laminations for the stator.

But I really feel like the RC community is screaming for custom lamination. And with them will come the homemade electric car people, skateboards, bikes and wind turbine, all those tinkerers.

Next step I guess will be the magnets, they will want custom magnets whose magnetization surface is not planar. And also custom arc circle magnets (for halbach array pancake motors).

Interesting, I'm not familiar with the RC market at all. I've looked into e-bike, electric car, and wind power projects before, and I definitely found the choice of motors was often one of the most limiting factors. I've followed projects where people have built there own motors, but I've never seen anyone build a custom motor using custom lamination. I've seen more motors made out of wood than I can count though.

Honestly, I think this is one of those things where the market doesn't exist until it exists. Someone could probably create a market for it by creating some cool projects in certain fields using custom cut metal lamination. If you offered to make custom lamination at cost for people who were building their own projects, they'd end up posting about it, and soon enough people would realize that you could make things with custom lamination.

That being said, I definitely don't have the technical expertise to judge whether inexpensive custom lamination is even possible, or if it's one of those things where you build a super cool company that makes just enough money to keep you in poverty.

Well, sounds like you've got a niche market for a small-scale startup. Get to it!

If used for anything substantial, such as an ebike that goes fast or up slight grades, this motor would simply melt. Its really cool though. A lot of ebikers fool around with liquid cooling but ive never seen it pulled off. This guy could take advantage of the 3d printing method to add liquid cooling and besides being really cool it might allow the motor to do useful work without melting.

looks like a good way to iterate, just disregard the melting because the real one will use metal

It must be very application specific what kind of air cooling you get? If you have a motor enclosed on an e-bike vs. if you have it naked on some kind of aircraft with a propeller blowing on it must make a huge difference?

A problem I guess with plastic is that it doesn't conduct the heat to the outside like a metal construction would, so it will always dissipate less than metal, regardless of environment? That issue is still there if you add liquid cooling - the plastic won't transfer heat to the cooling medium either.

What's the state of the art in high-temperature resistant 3D-printed materials?

That carbon printer would move the heat around very well.

Not really, like most micro particle or fiber filaments the thermoplastic resin is the weak part.

Any thermoplastic extruded filament that does not require secondary treatment is going to be heat sensitive.

The filaments that are not heat sensitive are either ones that require substantial heat to extrude in the first place or at then chemically treated to not be thermoplastics anymore.

If you think about this way it's simple, the melting point and the thermoplastic point of the end product needs to be substantially higher than it's operating temperature this cannot be done without additional treatment or using a process other than thermal extrusion.

Resins that use chemicals or light to harden are can have thermal resistance properties, metals and other materials that are extruded or bonded at very high temperatures are also resistant to relatively high operating temperatures.

Makes sense, Although I was trying to infer that the carbon material would transfer the head through the object better. Which would make using a liquid cooling channel useful. As opposed to just outright higher thermal operation.

But this is only evidenced from the carbon on my XPS laptop becoming untouchable in full-sun and much of the base heating evenly from the processor.

Ultem from Stratasys (a grade of PEI) is used frequently in aerospace applications where high temperature / low flammability are required. http://www.stratasys.com/materials/fdm/~/media/83DA2BBEE7DE4...

There's always metal too.

They print titanium and ceramics ive heard. Must be expensive. I would be very interested in seeung someone print a ceramic motor, beefed up to account for poor strength of ceramics. Yes most hub motors are closed but its very common for people to drill holes in the walls of the motor to get some air cooling. Less common is actually filling the closed hub motor with oil -- ive seen blog posts where people report that if you get that to work it does wonders for cooling. But there are huge problems with it, mainly related to accounting for the expansion of the oil as it heats up. Even less common is real liquid cooling where coolant is channelled down the axle and into the stator. Doing it with plastuc would mean having some metal inserts for heat conduction to the coils.

Some guys from Berlin demoed heat resistant printer filament at the Make Munich. After printing, you put it in the oven and then its heat resistant up to ???°C.

The content-vs-time density for this video is extremely high and it was a pleasure to watch all of the video. I print in PLA rather than ABS right now so the video isn't directly applicable to my life but it's nice to see somebody working on the designs. Cheers to the author!

I would recommend PETG instead of ABS next. It is easier to print, the fumes don't smell as bad and it is a very impressive material mechanically

Not only do ABS fumes smell bad, they can actually be quite dangerous. I believe PETG is much safer in that regard.

The enclosure looks like it's leaving a lot on the table in terms of cooling channels and structure. Given the printing can do all kinds of shapes it looks like it's replicated the kind of body that would come out of a mill or mold.

Are there any printers that can put down glass fibre in the process?

The MarkForged 3d-printers can put down carbon or kevlar fiber with Nylon binder.

That's a nice project. The author claims 80% efficiency, which is reasonably good. Most amateur motor designs are far worse.

The plastic motor shell has a very smooth finish, and screw threads. What process did he use to 3D print those fine screw threads? I didn't think any available 3D printing process was good enough for that.

I think he printed an undersized hole, drilled it to the right diameter and then used a tap to thread it.

Look at him screwing on the end bell of the motor at 222 seconds in. [1] That's a big diameter thread, and the close-ups of it show a very clean thread. You can see the marks of the 3D printing layers on the outer motor casing, but not on the thread. I suspect that thread was turned on a lathe.

[1] https://youtu.be/NFvMC3l3fGY?t=222

I've printed finer threads on my cheap 3D printer, although it took a few tries to get them the right size.

It's possible that he used acetone vapour to smooth out the surface of the threads.

that would be a huge tap. the motor looks to be about 4-5" wide. you'd need a $500+ tap for that, no?

Probably not tapped, but single-pointed on a lathe or thread milled: https://en.wikipedia.org/wiki/Threading_(manufacturing)#Sing...

At 1m33 [0], you can see the thing in an Ultimaker 3D printer, but as other commenters say, the threads are likely tapped. I have seen 3D printed threads, but not that fine, at least not with FDM

[0] https://youtu.be/NFvMC3l3fGY?t=1m33s

This is very impressive, but keeps me wondering - how is this engine better than the stuff you can just go and buy in the shop? This is an honest question, I have no clue what are the benefits here (besides obvious coolness factor!).

We're one step closer to a 3d-printer printing itself :)

It is already happened I believe. http://reprap.org/

I've never understood reprap's cult-like fascination with self-reproducing 3D printers. This is such an absurdly impossible goal that it's more of a category error than a current technological limitation: 3D printers are not organisms with ancestors, they are manufacturing tools. The things that CNC mills, injection molds, and 3D printers are made of are very different than the things they make.

Well, aiming for the possible isn't as much fun as the impossible.

Also the results are less interesting, usually...

Not in the sense that parent post is alluding too. reprap printers use "vitamins" - the electronics, motors, hardware - in addition to the printed parts.

Being able to print a motor would remove the need to buy motors.

>> Being able to print a motor would remove the need to buy motors.

Not every part needs to be "printed". Motor laminations can be CNC cut or laser cut from sheet stock. Then you stack them, bolt them together and add windings. For PM machines you slide some magnets into slots on the rotor. Done. This is fairly easy for hobbyists to do given a good design.

To build a 3d-printer you don't need to print a single part. But the goal of the Reprap project is to explore and develop maximally self-replicating machines. Some of the results are useful, some maybe not so much.

The idea can be applied outside 3d-printers as well. Here is a CNC machine you can CNC (including motion/drivetrain) https://github.com/fellesverkstedet/fabricatable-machines/bl...

One of my personal interests is in making it possible to replicate an full makerspace/fablab by visiting an existing one. Bootstrap one or two digital fabrication machine by borrowing access, then build everything else with those in your own lab. This benefits from high degree of self- or co-replicability.

(Perhaps) people similarly asked about personal computer kit by Steve Wozniak back in 1976: How is this better than a computer we have in that cooled back room in our company?

The benefit of this motor is the possibility of customization without steep price ... that our particular time allows to do it now. If you need a motor with a particular characteristics and it is already produced, go buy it. If not - here you go.

What's the type of paper called that he puts on the motor at the 2:15 mark? Doesn't look like simple iron dust.

https://en.wikipedia.org/wiki/Magnetic_field_viewing_film

Magnetic field viewing film http://a.co/5s8f0Kz

So in brushless motor the outside part with the magnets is the part that rotates? is that way it's brushless? You don't need to supply electrify to the part that moves?

Or do I have this all wrong?

Exactly, here is a video that explains it:

https://www.youtube.com/watch?v=bCEiOnuODac

Compared to brushed-motors they can be more powerful for the same weight (which is why they are popular in the RC world), quieter, last longer, and produce less EM interference.

The downside being they need powerful rare earth permanent magnets to work, making their construction more expensive than induction or brushed universal motors.

In a brushless DC motor, the rotor is a magnet, and the stator is a series of electromagnets that can be powered independently, yes. But the rotor is still usually at the center. You're dead right on why it can be brushless - the bit that moves is the permanent magnet. One other important difference is that the coils need to be controlled by a computer of some sort to get the timing right to make a rotating magnetic field and keep the magnet spinning.

And just for reference, at least in the hobby community, these are called in-runners and out-runners. An in-runner has the rotor on the inside and just a shaft that sticks out; an out-runner has the stator on the inside and the outer body of the motor rotates.

Impressive on so many fronts, it's so cool what's possible nowadays with modern prototyping tech. Thanks for sharing!

I hate the trend that's been going on for a few years now where they try to imply that all of a given product is 3D printed.

For example this motor is easily 80% metal and copper wire for the poles. The remaining 20% is the plastic case. That is the 3D printed part. Not exactly very impressive.

Having built a motor from scratch, I've learned that the hand-winding of the coils that's impressive. Mine had no torque and was terribly inefficient but there's a local motor rewinding shop where they produce real art.

It is not really implied. Like, if a road bake frame is 3D-printed it is called 3D-printed bike and doesn't imply that its wheels are not aluminium and rubber.

Reply on news.ycombinator.com