APL's DIY axial-flux motor

It looks like there's enough space for another layer of windings with the 14AWG wire. You could lay another 11 winds over the originals and connect them in parallel on each core. That would give you half the copper losses and maybe allow you to run about twice the power, with the same amount of voltage.
 
Looks like my next move is to hook the Cycle Analyst back up, I need to do it anyway. I'll get a no-load
current reading then.

I'll also work on getting the motor in the bike, and take it for a spin,.. should provide 'some' interesting
data, and at the very least, complete what I set out to do in the first place. Design, build, ride.

Once the new stator plates are made, at some point I wouldn't mind trying different core materials/ideas
again, since the laminations were by far the hardest part of this build.

But thats down the road, first,.. the CA.
 
Cycle Analyst is working again, and I took a no-load reading of 152 watts at 38.7 volts, full throttle.
The number seems to keep dropping little by little as the as time goes on, and the motor heats up.

Also, as a follow up on single coil volts,.. if I divide 6.5v by 14 coils, I get .464 volts per coil, average.
 
APL said:
Once the new stator plates are made, at some point I wouldn't mind trying different core materials/ideas
again, since the laminations were by far the hardest part of this build.

But thats down the road, first,.. the CA.

I've seen stuff called Metglas that may be better than laminated iron. May not be better too, but on paper looks great. Sometimes you can find surplus pieces of it on eBay for not too much. It's basically already laminated for you but really hard to cut so you want to find pieces that don't need cutting. Supposed to saturate at about the same level as steel but much lower losses.
 
APL said:
Cycle Analyst is working again, and I took a no-load reading of 152 watts at 38.7 volts, full throttle.

So the noload current is about 4A at 1700 rpm. My previous revolt motor with same kV had 5.7A at 4000rpm and the losses go up by rpm squared --> Motor has some extra losses due to the plates.

Efficiency with these losses added would go from max 66% to 61%.
 
APL said:
Cycle Analyst is working again, and I took a no-load reading of 152 watts at 38.7 volts, full throttle.


150W no-load is in the ballpark or better say not bad for a motor this size spinning at 1000RPM.
Did you measure the Crystalyte?

The problem will be the lack of torque as it seems because the winding resistance it very high considering the kV.
Whats the airgap width now?
 
The air gap is questionable, since I just threw the motor together for the first spin test. I need to take it
back apart and make the fine adjustments & remeasure parameters. Might make all the difference.
But design-wise, it's about one millimeter, if I remember right. Devils in the details,.. so I'll check.

The core adjustment system I made for the stator sucks. The set screw's make divot's in the core and act as a
memory, so it doesn't have the fine adjustment that I wanted. Fact is, I don't need adjustment, just all held
in the same spot, or plane.

On the new fiberglass plates, I'm hoping to mill the slots, and then bond a thin .035" piece of fiberboard over
them on one side, to form a 'cup' that just holds them all in place. It will make the slots stronger as well.
I'll have to use some spacers across the two plates to clamp them. We'll see what kind of fiasco this turns into.

The Crystalyte that I used for a ballpark model on this build is in the bike I'm riding, and once I switch the two
around, I can get some test stand measurements from it. At the moment, I'm taking some CA readings from it
on various hill's and flat's, so I can compare it with the axial, once it's in there, on the same sections of road.

Fechter, I like the Metglas, it makes a lot of sense to me, super thin foil bonded under high pressure into a glass
like core is awesome. To bad it's so hard to cut, and work with. Also very expensive, I could buy a motor for the
cost of some of that stuff.

I found an article on a product called Somaloy, a soft magnetic composite that has each particle individually
coated, that has lamination type properties. It's made for prototyping, and can be machined fairly easy.
The only problem is that it's not available for the hobbyist that I could find. Figures.
Then theres 'fully processed steel', which I've read is a good steel for prototyping?

Somaloy info. use page down box on top.https://docplayer.net/20823110-Iron-core-material-somaloy-unique-magnetic-properties-high-purity-iron-powder-electrically-insulated-surface-result-in.html
 
Somaloy has decent machinability, but is very prone to cracking. You have to be very careful with it or it'll just fall apart. Super cool stuff, though.
 
Thanks for the info,.. anybody know where I can get my grubby little hands on some? (Somaloy)

Also heres the info I found on fully processed steel,.. maybe the guy doesn't know what he's talking about though.
Happens more than not.

IMG_1968 copy.jpg

Taken from Practical Machinist; https://www.practicalmachinist.com/vb/general/need-source-electrical-steel-also-called-lamination-transformer-steel-etc-246370/
 
Not sure if this is what its called in the states but what about HPL compact plates? (High pressure laminate) I work with it from time to time and it's a pretty strong material. It's not that difficult to machine and cut as well.

csm_Max_Exterior_Fassadenplatten_2f4acd760b.jpg
 
It seemed to me that the aluminium was still causing some eddy loss, he did mention fiberglass and all.
 
I think the 'correct' material might be phenolic sheet, as it's made for electrical purposes, and can be found in
many configurations and types. Goes by the name of G10 or FR4, and is made from canvas or linen as well as others.
Comes in thin or thick sheet, and very thick 2 - 3" chunks.

Phenolic sheet.jpg

Phenolic sheet, etc.; https://www.google.com/search?hl=en&tbm=isch&source=hp&biw=1680&bih=1012&ei=CKxUXfy7I8awtgWgrpTQBw&q=phenolic+sheet&oq=phenolic+sheet&gs_l=img.1.0.0j0i5i30j0i8i30l5j0i24l3.13547.21467..29147...0.0..0.300.2214.0j15j1j1......0....1..gws-wiz-img.....0..35i39j0i10.yJquMVhA0HA

Other offerings of the same thing are Garolite, Micarta, Glass board, PCB sheet, and epoxy glass. I'm sure there are
others. Each one has slightly different properties, but all would work. Excellent electrical properties and heat
resistance, as well as burn through. Fairly cheap too. Not that fun to machine though, I think,.. glass dust.
Lots of colors to chose from. :)

G11-G10-FR4 Garolite-Laminate-Sheet.jpg

I've used some 1/4" thick green Garolite in amplifier chassis before, and it was pretty tough stuff.

Plastics would work as well though, there are some really strong plexiglass offerings, and I don't expect any high heat
situations. As a hobby motor, we can get by with a lot less than production motors can. (safety-wise)

Aluminum is perfect though,.. dang pesky eddy currents!!! :cry:
 
Also, it's still unclear as to whether carbon fiber sheet would be eddy free,.. it's the strongest, and probably has
good thermal transfer? Not sure what the cost might be though.
 
APL said:
Also, it's still unclear as to whether carbon fiber sheet would be eddy free,.. it's the strongest, and probably has
good thermal transfer? Not sure what the cost might be though.
I think I remember hearing carbon fibre has interesting magnetic properties and could be used as a lamination material but don't quote me on that, it's a very hazy recollection and probably a false one.

Is there any reason rolled laminations couldn't be used? ie. get (for ex.) an inch wide by yard long steel strip and roll it up into a cylinder? Should be easy enough to form into a rectangular or wedge shape but guessing there's some reason it's not used, only ever seen anything like that in old car ignition coil cores and then it was zig-zag folded (not rolled up).

Congrats on getting it running, bet it was a big relief after all the trials and tribulations :)
 
The rolled lamination core has been used for certain things, but for a stator tooth the laminations must
all go from left to right, and no vertical's. I have seen a large roll of a lamination strip, 10" round by 2"
deep or so, that was bonded, and then cut into wedges, for axial teeth.

Then theres 'segmented stator teeth' which are very interesting. Most of the work is done, and would only
need a little more machining for a project like this. Or, the motor could be redesigned to use them.
Trying to find some that are the right size, shape, and cost, might be next to impossible though.

Segmented stator tooth..jpg

I keep looking at a big 21 lb. chunk of powdered iron core, on E bay, for $20. that might be machinable?
I don't know to much about the stuff, but for the price, I can't go too wrong. Could make a good door stop.

E bay, Micrometals iron core; https://www.ebay.com/itm/Micrometals-Iron-Powder-Block-Core-40-Mix-12-x-3-x-2-5-21-5lbs/282910437725?hash=item41dec6b15d:g:KlwAAOSwlJlaxNxX

Then last but not least, I still have the idea of using stacked steel wires, or rods, for a core. They could be
coated, and easily cut. Plus, being stacked and bonded, they will conform to any shape, and wire is cheap.

The idea is that the super small individual ends facing the rotor, would contain the eddy currents, much like
the thin laminations do. It's like a shredded lamination.

The problem is finding silicon wire or rod, although I have seen steel Mig welding wire with silicon content. :idea:
A shot in the dark, but worth a try. And kinda what this motor is all about,.. being able to try different things. :)
 
The idea of round or flat wire transformer cores is not new, there has been some research done,
and amorphous wire is not that uncommon either, it is used as cores for small devices like sensors.

Research gate transformer wire article;https://www.semanticscholar.org/paper/Silicon-Steel-Wire-With-Low-Magnetic-Losses-Yonamine-Fukuhara/2d9a13c554d16eac8e062d95fc6eb4e946b11e99

Aichi amorphous wire Co.;https://www.aichi-steel.co.jp/ENGLISH/products/amorphous_wire/product_01.html

The problem, as usual, is getting ahold of it, and at what cost. More research is in order.

But the potential is there, and being able to make a straight core into any shape is a real plus. (axial DIY)
Although, you can't beat lamination material for the technology and metallurgy that is put in to it, the
amount of effort that goes into making it is simply staggering. It's just damn hard to work with, and purchase.
 
That would be nice, thanks for the info. I'll have to see if I can put together an inquiry for them, although
it might be tough, because I know they're going to want specifics.

Proto Lam Company web site; http://www.protolam.com/index.html Small run lamination cutting.

This motor is still in a turbulent design stage, so it would probably be a little more cost effective to use their
services a little later, when a little more stable design is achieved.

It's a good link though, and might be just the ticket, as this is just what they do, and this post definitely needs
more options like that.

A while back in this build, amberwolf suggested using 'The Big Blue Saw' which is a 'small run' water jet cutting
company. I contacted them to see if they could cut lamination material, and what it would cost. They said (at
that time), that they didn't have lamination material, but could probably cut it. There would be a $200. charge
for cutting 'your' material, plus, the normal service charge for set up and cutting. So I imagine a cost of four,
to five hundred dollars, would be the ball park figure on that,.. just a guess though.

The big blue saw link; https://www.google.com/url?client=internal-uds-cse&cx=012839003547054862050:rcui7bzn6g8&q=https://endless-sphere.com/forums/viewtopic.php%3Ft%3D48811&sa=U&ved=2ahUKEwjm87b4gYjkAhXpqlkKHT7yCkUQFjAAegQIAhAB&usg=AOvVaw3EDwGkgfRuoEuueqM4g3cX

Thats about what this motor project has cost 'so far' (estimate), so between the two, it might be a $1000. motor.
In my humble opinion, thats seems a little extreme for DIY. But I'm a poor boy, so it's a matter of opinion.
And not really that it's such a bad price, it's just that in early, early, prototyping, you want the cost as low as
posible, because your design is going to change overnight sometimes,.. like this one is doing.

Keeping in mind, that cost for parts made, goes down with numbers. So once a design is finalized, one could build
cheaper and cheaper DIY motors, and lamination cost's would look much better, providing that you at least kept the
core, or tooth, design the same.

I definitely wouldn't mind spending the money for 10,000 laminations to build several motors, as long as I know I am
going to use them all.
 
I can quote having laminated plates cut, either waterjet or machined. No $200 "your material" fee. Shoot me a pm if you end up needing it done.
 
Thanks coleasterling, that's awesome! As soon as I get to that point, I'll get in contact, and see if we can work it out.
Good to hove a knowledgeable source that understands this build. :)
 
Back
Top