Building a diy friendly motor

I've been thinking about building a motor for a while. I've learned a ton from this site and still have a lot more to learn.

I have a large SLA printer hopefully arriving soon (Phrozen Mega) and have a seat of Solidworks for my day job that I will be using for CAD.
It needs to be easy to assemble, build, and use as of the shelf components as possible.

I'm not sure on coreless yet, not clear yet to me on the trade offs. What terms can I look up to better understand this?

I'm also unsure on axial vs radial design yet, I'd be ok with either. I think magnet sourcing /availability will drive this.

My thought is to use 22.5° arc magnets (16) that seem to be somewhat readily available. I'd like to do a Halbach setup but I'm not sure if the magnets that redirect the flux need to be the same size (physical or magnetic strength) as the magnets doing the motor work. My current thinking is arc magnets with smaller rectangular halbach magnets between them.

This part might seem silly but it seems like it could work out really well. 15 slot / 16 magnet (22.5°) seems sensible in a lot of ways besides looking really wrong at first glance. am I missing something?



This would allow for 3 modular (identical) pieces of stator and seems like they would be really easy to wind and swap between delta and wye.

My last questions currently are on litz wire size. I've seen ac frequency mentioned in choosing the wire diameter. This would be targeted to the ac frequency of your expected normal motor speed?
I've also seen some huge claims on current capacity using litz wire. Does anyone have any references / calculations to help size the stands, number of strands, and current capacity?

I found this last night. Wish I'd found it ages ago, i learnt the contents of it the hard way.
https://things-in-motion.blogspot.com/2019/01/selecting-best-pole-and-slot.html?m=1

15n isn't a nice number because it results in asymmetric force generation. You really want symmetric phases either side of the rotor or the magnetic forces load your structure and cause warping and potentially vibration and rub.

18n16p and 24n22p or 26p are better stator slot numbers.

I've found getting good stators a pain in the neck. That pushes me towards coreless for DIY. I finally found a 36 slot stator I'm going to wind with 30p probably single layer, but I can't see it being readily available long term. Core less much more so.

That was an extremely useful link, thanks for sharing it!

I keep coming back to 3D printing these parts and using different methods to keep it simple but functional and in reach for hobbyists.
Probably means litz wire is out or maybe optional.

For stator teeth I've been thinking about using insulated iron powder and epoxy. We could find an ideal ratio by weight to mix. Then cast in place with vacuum infusion (or individually then assembled.)
I suppose insulated iron powder and resin could be used as back iron as well if a consistent void can be left to fill.
My thought is a 3D printed frame that can hold different parts where they need to be that stays in place. If needed, removable frames / fixtures to help with concentricity and other consistency struggles as it cures.

I think the hardest part to fab for most will be the hubs. I have the tools to do this but I really don't want to. It puts it out of reach of most hobbyists - only cnc owners would embrace it I think. I guess it could become a cnc yourself or a purchased part.

its fun to experiment, and I encourage you to explore. That being said...

For a DIY motor, I am a fan of the Lebowski Axial. Can be made with one, two, or three stators. Look in the motor section, at the stickey index...

Thanks for the recommendation. It is definitely worth a read. I do really like the idea of it being stackable.
https://endless-sphere.com/forums/viewtopic.php?f=30&t=46476&hilit=Axial

My printer arrived yesterday, I haven't opened the box yet. I'm going to get it up and running next week.

I appreciate the offer. I'd be more than willing to collaborate but not sure our motor wants align and from what I'm seeing I don't think I will like openscad. I've struggled to get anywhere with open source cad softwares.

A few suggestions I have for you to consider.
The halbach magnets being taller means the od is larger and makes back iron not easy or practical.
A steel tube could be the back iron and give the outside some strength.
Look at using something better than pla. There are filaments with iron in them that could help. https://www.proto-pasta.com/products/magnetic-iron-pla?variant=27771130888&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic
Should be some other options too.
Raidiused corners will help strength & fatigue.

Solidworks has a hobby license for $100. That's well worth anyone's time that's struggling with inferior cad software. Solidquirks isn't perfect either.

There are lots of casting / molding / "forging" methods that can be assisted by 3D printing. There are high temp SLA resins that can be used to make low run injection molding tooling. Laser cut metal isn't too bad price wise either.

My thought was to build a motor that could:
- power a lawnmower blade (3.5hp / 2.6kw peak ballpark).
- since reading Lebowski's motor build, stackable is on the wish list.
- modular, easy to wind, common parts, etc.
- be hobby level buildable. 3D printed, low cost, off the shelf parts, and preferably no / minimal cnc parts.

I think this means a pancake shaped outrunner or an axial. Hallbachhero's motor is pretty close to what I was thinking and is worthy a read if you haven't.

I've been thinking about this again lately. My thought is to use pcbs for as much as possible. I'd love to be able to do the work in design, order the parts and quickly bolt it together. Planning on around 48v @ 60A.

One idea was to use unshielded inductors on a 2oz pcba - soldered in place. Could use 1 or more pc boards to lock the coils all together. I really don't want to wind a stator. There are some patterns that are decent looking - over .8 winding rato and have the same winding direction to allow this to possibly work. 30 slot, 40 poles and some others. I haven't found any info on anyone trying using inductors in a stator online - I'm taking that as I might be missing something pretty important.. Coilcraft has some good looking options for this. Not sure if multiple inductors per "coil" approach would work magnetically to spread out the power and get the inductance and power handling reasonable.

I was also thinking of using pcb trace coils as the stator. Looks like 3mm, 2oz, 2 layer boards spread over 7 boards for 2600w...might work. Also sounds like a great way to cook a meal. There are some surprising claims for pcb coil motors on the web that seem inflated. I'm also trying to think about how to cool it. 10 coils per phase, per board with 7 boards. 210 coils, 12.5 watts per coil (and water cooled) sounds much more reasonable .
Edit
It will likely be using 2 stators (maybe 3 with a second rotor) so the load per stator could end up a lot less.
End edit.
Might make slots around the coils with some thermal bridges off each of the coils for a couple water tubes.

The rotor could likely be 2x or 3x 2mm aluminum or fibreglass pcbs stacked, magnets inserted and glued.

Are .5uH and .8 - .6 ohms appropriate targets for a 2600 watt 48v motor?

I think the thing your missing is that inductors are normally made from ferrite and are optimised for low Eddie losses at ultra high frequency.

The trade off is a low saturation point, so the torque limit is probably 1/4 of using electric steel of the same dimensions.

mxlemming said:

I think the thing your missing is that inductors are normally made from ferrite and are optimised for low Eddie losses at ultra high frequency.

The trade off is a low saturation point, so the torque limit is probably 1/4 of using electric steel of the same dimensions.

Click to expand...

Air core inductors would work, no?

This is what I was thinking, https://www.coilcraft.com/en-us/products/rf/air-core-inductors/high-current-flatwire/1010vs-1212vs-2014vs/

In this layout,

I guess air core does not have this limitation, so it may well do, noticing of course that the iron does generally allow for higher field strengths than air for similar amounts of magnet.