DIY Toroidal Axial Flux PM

Hummina Shadeeba said:
without a dyno or load you can add to test torque i guess just go for max speed! the motor not having any iron makes it seem especially a high speed motor and where its power will be and then youd gear it down.

maybe before going full speed balance the rotor. Ive heard you could try a laser pointer fixed to the stator thats pointing at a wall and add or subtract material with the goal of getting the laser point stable as you remove vibrations. sounds fun. never done it.

eventually i guess a magnet will fly off like a missile though and surely dangerous, surely you know. peak power maybe could be way higher if you can fix the magnets better and then ..double the speed. I think you should add long kevlar strands wrapped around the outside of the magnets!

Yeah I will spend some time balancing things before I let it rip, and obviously Ill collect some data too. I had an early version blow up in my work shop. I only ever found about 30% of the rotor

TorontoBuilder said:
As I said I have contacted a few sources that supply low volumes typically for experimental uses. One company in the USA will sell to individuals without a corporate entity which is great. I expect a quote within the next few days, and I shall let you know. It may require a larger volume purchase than a couple of kilograms. In such a case I'd see if we could perhaps do a group buy.
Wow, all of that was awesome information. thanks for sharing. Depending on the price, I could probably swing a group purchase. Curious what you hear back.




First a few glamor shots.
PXL_20220127_014456565_small.jpg
PXL_20220127_014520231_small.jpg
PXL_20220127_014626044_small.jpg


I made a new hub for the stator and a new spacer ring. The tolerances are really tight now. I am currently printing a mount for it so I can test it as a motor, but I was able to spin it with a drill while holding it by the machine screws. I believe I was able to calculate a Kv of ~181 (~337Hz @ 9.3V). This is slightly less than the lowest I measured for any of the other motors, and this one doesn't drag on anything, and is smaller in diameter and thickness, and is lighter than the last couple (408g with all the screws), and still making nice sine waves.
Overall this is a great improvement, very happy with the results. Looking forward to trying more/less turn stators and seeing how they compare.
 
TorontoBuilder said:
..lots of SMC info..
Thanks for sharing all of your legwork TorontoBuilder! If they only want to sell to a company I could probably help us with the group buy depending on cost / interest. I might go the die, compaction, sinter route.

Hummina Shadeeba said:
How are you going to balance it?

At the very least a lawnmower blade or an rc airplane prop balancer could work. A jig with a couple ceramic or air bearings Could work too.
 
Jrbe said:
TorontoBuilder said:
..lots of SMC info..
Thanks for sharing all of your legwork TorontoBuilder! If they only want to sell to a company I could probably help us with the group buy depending on cost / interest. I might go the die, compaction, sinter route.

Hummina Shadeeba said:
How are you going to balance it?

At the very least a lawnmower blade or an rc airplane prop balancer could work. A jig with a couple ceramic or air bearings Could work too.

Thank you for that offer. I may very well need to take you up on that.

Despite speaking to someone on the phone about ordering from this supplier and the sales person telling me we should not need to have a company place the order, last night I got an email asking for the name of my company and the required specifications for the product we are seeking. The email did not specifically state that only registered companies can order from them but I got that impression from the email.

I'd also already informed them that I believed that their specific product met all of our specifications.

I hate when there is such a disconnect between the people you speak to at a sales department and the people who follow up on calls. I also believe that I had filled out a form on their website requesting a quote a few months ago but I never received a reply to that inquiry. I believe this may be because I stated I did not have a company.

Getting back to the specifications.

The process of compacting metallurgic powders imparts high internal stresses in parts which in turn create high hysteresis losses. For this reason parts require annealing at high temperatures to relieve these internal stresses. Any coating must survive the annealing process in order to be effective in the manufacture of die pressed stator elements. Of course this does not matter to those wishing to experiment with such powders as an epoxy additive.

The only firm specifications I have is that the powder features an exterior oxide layer that is thick enough to
provide an insulating layer which aids in the magnetic properties AND that remains highly electrically resistive post-processing. Effectively, the product shall provide a level of electrical resistivity in excess of 100000 micro-Ohm-cm when compacted to 800 MPa and cured at a temperature above 800 deg C. Additionally inductor components manufactured with the product shall have low core losses at frequencies >1 kHz.

It is my understanding that the product I am seeking from the supplier provides all these features, but there is no guarantee of that until I hear their response. The powder I am seeking is an Iron-silicon chromium (FeSiCr) powder manufactured using an advanced technique called water atomization. Unlike gas atomization which produces spheroid particles that do not compact well due to the lack of surface irregularities which can lock together, water atomization produces irregular jagged flakey particles that compact very well.

The silicon in the alloy helps form the oxide insulative layer that provides both high electrical resistivity and that resists degradation during high temperature sintering. The chromium in the alloy is supposed to improve "the environmental reliability of inductance" and rust resistance. I believe what environmental reliability of inductance means that the inductance values of parts made of the smc material will have little variance with changing environmental factors such as heat and humidity.

Compared with other soft magnetic alloys, FeSiCr alloy powder has lower core loss at high frequency and better anti-rust performance.

I'll keep people posted and if we can get an order I'll start a thread to judge the level of interest people may have
 
If you are just going to mix iron powder into the epoxy, the insulation won't be much of an issue as the epoxy is a good insulator. It will make the epoxy more like a paste and not flow well if there is a lot of iron. Pressing the powder into wedges or whatever shape will be expensive. There are lots of cylindrical pieces you can get off the shelf.
 
fechter said:
...There are lots of cylindrical pieces you can get off the shelf.
Have any sources on where to find them? I've been searching and have found nothing.
 
Google "ferrite rod" or "powdered iron core" and use the image search. There are all kinds of other shapes but most not suitable. Tons of rings and tubes. There are also cores made to go around flat ribbon cables that are rectangular in cross section. The SMC stuff APL was using was fairly easy to cut. Some types are super hard and not easy to cut. You want the powdered iron stuff.

Rod example:
https://static1.squarespace.com/sta...c/1502264287692/IRON+POWDER+CORE+ROD+TYPE.PDF
 
fechter said:
Google "ferrite rod" or "powdered iron core" and use the image search. There are all kinds of other shapes but most not suitable. Tons of rings and tubes. There are also cores made to go around flat ribbon cables that are rectangular in cross section. The SMC stuff APL was using was fairly easy to cut. Some types are super hard and not easy to cut. You want the powdered iron stuff.

Rod example:
https://static1.squarespace.com/sta...c/1502264287692/IRON+POWDER+CORE+ROD+TYPE.PDF

Hahaha hahahhahaha right. You can't improvise when it comes to high performance.

Most of the materials used in such are not appropriate for my usage or anyone who is seeking to make experimental motor that pushes the efficiency envelop. I'm not trying to build a cheap soap box racer that will simply make it down a hill without falling apart.

Real motors have precise engineered stator sizes. In that search of yours you wont find a single actual "stator core" for a reason.

Sorry HH for putting this on your thread. i will refrain from further smc comments here
 
Hummina Shadeeba said:
How are you going to balance it?

That's a great question, I got the mount printed, and yeah its a bit out of balance. Also ive had this issue with several versions now, but if the bearing is not a tight enough fit, the stator will slip from one side to the other, innevitably dragging on one side since one side has nothing physically stopping it from sliding that far. I just added a grove to the rotor hub and made a little E clip to hold it in place. I am fine tuning tolerances, but it seems to be doing the trick. Once I have this working well. I will need to think about how to balance it. I really am not sure how I am going to. and not only figuring out how, but then im not sure what to do to get it right. I dont think I can add weight since it will likely fly off if not mechanically attached... Ive been thinking on this for a bit now. I believe it needs dynamic balancing, as its pretty good on its own, but once I spin it it shakes the stand quite a bit

Jrbe said:
At the very least a lawnmower blade or an rc airplane prop balancer could work
These are intended for static balancing correct?

TorontoBuilder said:
For those interested in smc powder.

viewtopic.php?f=30&t=114869&p=1698735#p1698735
Thanks for posting that. I will definitely follow along

fechter said:
If you are just going to mix iron powder into the epoxy, the insulation won't be much of an issue as the epoxy is a good insulator.
This is about the extent to which I will likely take this project. Looking forward to seeing the difference

TorontoBuilder said:
Sorry HH for putting this on your thread
Yeah that was uncalled for
 
HalbachHero said:
Jrbe said:
At the very least a lawnmower blade or an rc airplane prop balancer could work
These are intended for static balancing correct?
Yes static but you can get decent results if you use one with almost no friction.
You could look at dynamic balancers too. They are $$$ or a few diy builds using microcontrollers. I wonder if a car tire balancer would work at all..

You might be able to use a wave washer on each side to help keep it centered if there are some thermal expansion elements to deal with.
Otherwise some shim washers are ideal to not load the bearings and won't give it the opportunity to move. But if it does have thermal expansion issues it might start to bind up or might get loose.
 
Edit. Higher RPM's make balancing harder. You have the six case bolts to work with though, maybe just adding or subtracting a few shim washers can help pinpoint.
 
Hi I found this thread very interesting I'm at the beginning of my own AF motor and learned a lot, thanks for that.

I don't know if someone already asked but aside from improving the design on various metrics, what is the end goal once your motor is performant enough for you? Is the end goal the journey perhaps? I am wondering what is the future you want for your motor?
 
Hey all, I just wanted to chime in since I have been radio silent for a bit. Life was taking priority. I have not made much progress since the last update, but I was playing around with FEMM again and have some findings with that I would like to share and get some feedback on once I collect some more data.

I still need to balance the motor, even though it happily takes 16V at full throttle which comes out to 2880 RPM @180Kv. I was having a difficult time getting a good reading on RPM with the tachometer I have since the rotor faces are all metal and reflective now.


Jrbe said:
I wonder if a car tire balancer would work at all..
I was actually wondering this as well. wonder if a local shop would let me use it

Jrbe said:
You might be able to use a wave washer on each side to help keep it centered if there are some thermal expansion elements to deal with.
Otherwise some shim washers are ideal to not load the bearings and won't give it the opportunity to move. But if it does have thermal expansion issues it might start to bind up or might get loose.
If I am understanding correctly, this would go on a threaded shaft and hold the stator down correct? The e-clip I made actually works quite well, but I think that I will need something more robust in the future.

APL said:
Edit. Higher RPM's make balancing harder. You have the six case bolts to work with though, maybe just adding or subtracting a few shim washers can help pinpoint.
Yeah this is the only thing I could think of as well, but I haven't done much with this yet.

Joannot said:
Hi I found this thread very interesting I'm at the beginning of my own AF motor and learned a lot, thanks for that.

I don't know if someone already asked but aside from improving the design on various metrics, what is the end goal once your motor is performant enough for you? Is the end goal the journey perhaps? I am wondering what is the future you want for your motor?
Welcome and thank you for you interest. I encourage you to post your journey, it's great to get feedback from this community (or others).
As for what I intend to use this for. I still don't really have a good answer for that. I have been thinking of many different things.
Just a few:
- RC car or airplane
- A small e-scooter for my son
- Wind turbine / Flywheel generator
Or if I can keep improving this, maybe an E-bike or something similar
 
HalbachHero said:
Jrbe said:
You might be able to use a wave washer on each side to help keep it centered if there are some thermal expansion elements to deal with.
Otherwise some shim washers are ideal to not load the bearings and won't give it the opportunity to move. But if it does have thermal expansion issues it might start to bind up or might get loose.
If I am understanding correctly, this would go on a threaded shaft and hold the stator down correct? The e-clip I made actually works quite well, but I think that I will need something more robust in the future.
I'm not exactly sure how your bearings are. Along the lines of this is what I meant, https://www.smalley.com/blog/how-and-why-you-should-use-wave-spring-bearing-preload
 
Maybe try a torque converter shop for balancing. Their balancers are horizontal instead of vertical. They are centered by a bushing (which adapters are easily made on a lathe that they will have) and held on the plate by a flat rubber mat.

https://youtu.be/grKLUXkDeXg
 
Nothing major to report life has still been a priority lately, but I posted a new youtube video. it shows the thrust bearing and how I was trying to use it.

[youtube]1K8LQPfMmJk[/youtube]

I have been brainstorming a balancing rig for a bit now, and found a video of what might be the simplest solution so Im going to try it. It going to have an accelerometer on either side of the motor with a sensor that I can use to measure the rotation. With this information I should be able to figure out where on the face of the rotor its out of balance.

I also made a torque test arm and started playing with measuring torque, but It damn near impossible with the controller I have, it jerks it all over the place. I imagine an FOC controller would be much better suited for this task. Finally time for the FlipSky?


drother said:
Maybe try a torque converter shop for balancing. Their balancers are horizontal instead of vertical. They are centered by a bushing (which adapters are easily made on a lathe that they will have) and held on the plate by a flat rubber mat.
Thanks for the suggestion. I actually called around to a few places, and none of them were too keen to the idea, but I'll try out a DIY method, if I don't have any success I will reach out to some machine shops for help.

Jrbe said:
I'm not exactly sure how your bearings are. Along the lines of this is what I meant...
Thanks for sharing this. I could probably use these in mine. I am hoping to experiment with different stator hubs to get it machinable. with a flat surface combined with the preloading. I think it will do the trick
 
a bit of progress.

I ordered an ODESC ODrive controller to test torque. I also got my accelerometers and position sensor. Just need to model and print some test stands. But im still battling with things not being straight which seems like an obvious thing to fix first.

I have started communicating with some local machinists and drive shaft/electronic motor shops to help with manufacturing and balancing if my method doesn't work out.

Most of my time has been spent on this though.

I started making a double helical planetary gear set. It seems to work so far, but its realllly tight, so it puts a pretty good load on the motor without anything extra.

Its a 5:1 from axle to the ring. 50 teeth on the ring, 20 on the planets and 10 on the sun gear. each planet has a bearing.
I had to modify the stator hub a bit to fit the planet carrier.

PXL_20220309_025700484.jpg
 
If it was metal I'd say to throw in some jewellers rouge powder and let it run. I'm guessing it won't work out so well on plastic though.
Is it just print artifacts making it tight? Have you tried vapor smoothing on the printed parts?

Are any of these gears available manufactured in nylon or brass?
 
Jrbe said:
If it was metal I'd say to throw in some jewellers rouge powder and let it run. I'm guessing it won't work out so well on plastic though.
Is it just print artifacts making it tight? Have you tried vapor smoothing on the printed parts?

Are any of these gears available manufactured in nylon or brass?

I have heard of using some silica powders to do similar things. but usually for removing layer lines. I believe its tight because there is nearly 0 clearance in the modeled parts. I usually leave like a 0.3mm gap between parts to make them fit nicer.
I could play with horizontal expansion settings on the printer, but I haven't tried to print too many gears yet. I think I can feel the ring gear warping slightly between the planet gears. I have run it for a while with the motor to loosen it up and you can watch the current draw drop over time.

I have never heard of vapor smoothing, but seems simple enough. I could certainly give that a shot, but I think ultimately I just need to provide a very small gap and re-print the ring gear.

I could probably print in Nylon with some small improvements to my current set-up, but the gears are fairly custom. I have not searched for off the shelf parts, but I doubt I would find a direct replacement.
 
Silica is not great for your lungs. Research it quickly before deciding to buy some.

I looked quickly for off the shelf stuff. I'm not sure what your plans are but a ring gear setup from battery tools, starter motors, kitchen mixer - lots of other applications that are completely unrelated but could work well and be cost effective.
https://www.zoro.com/bosch-gear-set-1617000665/i/G2500736/#specifications For an example.
You could also source a custom setup but I imagine you could find similar off the shelf components very close to what you have now from a gear manufacturer. This is likely an expensive route though.
https://shop.sdp-si.com/catalog/?cid=p130

I can't quite see how your gear set tooth profile is. Helical or V shaped teeth? If helical do you have something to deal with the axial load? It can be an issue depending on the setup when using helical gears.
 
It's very hard to get printed gears to run smooth and efficiency wont be great but I'd try the rouge with it, plastics and abrasives don't really play nice but with some playing around you'll likely be able to get everything to lap together even if it means running it unloaded for a few days straight and drip-feeding it an abrasive solution.

Tooth sizes and profiles may help, the biggest teeth possible would probably make sense to reduce the effect of the inherit tolerance limitations of printing and helical might not be helping, even that very slight overhang can effect finished dimensions badly. Not sure if you're using a raft, the steep temperature curve immediately above the bed really screws around with dimensions for me but that's probably just my settings, materials etc. Often thought alternative tooth profiles and methods could be better for 3dp, similar to a sprocket with the mating gear using small bearings as rollers for example. Looking really good :)
 
Jrbe said:
Silica is not great for your lungs. Research it quickly before deciding to buy some.
Yeah Im aware of this fact. My wife does pottery. nasty stuff. But thank you.

Jrbe said:
I looked quickly for off the shelf stuff. I'm not sure what your plans are but a ring gear setup from battery tools, starter motors, kitchen mixer - lots of other applications that are completely unrelated but could work well and be cost effective.
thank you for the sources. I will explore this option. I went the route that I did because I thought I could package it nicely. but I think if I want efficient gears, I have to accept that I will not easily accomplish making them on my own, nor is that the goal of the project. Ill see what I can find off the shelf.

The tooth profile is a double helical, so V shaped. It's self aligning and only possible in a planetary gear set because its printed and I can bend the ring gear to get it on. Obviously it could be split in two, but yeah I just forced it into place. With helical I will need to worry about the axial load yeah, maybe thrust bearings if its an issue.



stan.distortion said:
It's very hard to get printed gears to run smooth and efficiency wont be great but I'd try the rouge with it, plastics and abrasives don't really play nice but with some playing around you'll likely be able to get everything to lap together even if it means running it unloaded for a few days straight and drip-feeding it an abrasive solution.
Yeah I question if this is even worth it, but ill explore the idea a bit.


Lately my thought has been around making new stators, possibly with iron cores if I can get them made for a reasonable price with the contacts I am in touch with. I am modifying the mold slightly to make it not rely on clamps, and just screw into place on both the inner and outer diameter.
Going to try to make thinner stators, and replace more parts with metal, as I'm starting to feel a bit more confident with some of the parts.
 
Back
Top