Magnetic field and coil depth

[SimonMester]

So as the magnetic field gets weaker with distance exponentially, how would you go about calculating the depth of coil that it will still go through and produce a worthwhile current?
I have looked at many designs and they are wildly different. Some have coils that are about as deep as the magnet height, and some that are 2 or 3 times that.

[Mary B]

For an axial flux machine you want the coils no thicker than the magnets, and the magnet rotors as close to the coils as is safe/without scraping when the magnet disc flexes with wind direction changes.

[SimonMester]

What about radial machines?

[MattM]

Same principle.  The goal is creating as quick of changing magnetic field as practical.  Axial is easier to use magnets on both sides, where weaker magnets can compliment each other for best results.  Not so easy to do the same on radials.

[SimonMester]

Same principle.  The goal is creating as quick of changing magnetic field as practical.  Axial is easier to use magnets on both sides, where weaker magnets can compliment each other for best results.  Not so easy to do the same on radials.

I will have a bunch of cheap ceramics next month, I could use that to play with axial designs.
I suppose the one annoying thing there is the coils have to be a triangular shape. I'm not sure how that effects the magnetic fields crossing it.
Also, the two rotor magnet plates, would they have the same poles to the opposing rotor plate? Or would that alternate as well?

[SimonMester]

Ah! Bonus question: How does stacking magnets compare to actually having a magnet one piece? If you glue them together for security, surely that must be worse than having it in one piece because of the extra gap the glue takes up?

[MattM]

Its more of a bitten piece of pie shape you are aiming for, not really triangle.  Your coil legs should lay straight out from center.  One piece magnets are better than stacking.  You can always run several layers of alternating layers of magnet holders and coil stators.  Magnet holders should all be arranged with the magnets aligned so that poles align in the same direction. 

https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-rpg.2018.5311

[SimonMester]

Its more of a bitten piece of pie shape you are aiming for, not really triangle.  Your coil legs should lay straight out from center.  One piece magnets are better than stacking.  You can always run several layers of alternating layers of magnet holders and coil stators.  Magnet holders should all be arranged with the magnets aligned so that poles align in the same direction. 

https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/iet-rpg.2018.5311

Thanks. I'm definitely giving some thought to designs. As I'll have plenty of different magnets to play with, and to be honest I'm growing appreciation to ceramic magnets, they are not all that bad.
Also, what is the reason for the air gap is most ferrite cores/iron cores I see? Wouldn't it be better to have a solid iron core filling out the space? (As long as its made of proper material to minimize eddy currents of course)

[MattM]

As long as you aren't using solid pieces in your stator, and instead using powder isolated with epoxy, you shouldn't get strong eddy currents.  Unless of course your epoxy is conductive.

[SimonMester]

As long as you aren't using solid pieces in your stator, and instead using powder isolated with epoxy, you shouldn't get strong eddy currents.  Unless of course your epoxy is conductive.

At the moment I only have solid pieces of mild steel. I plan to get magnetite and epoxy later on. I'll try to figure out what ratio is best, as currently I could only guess.
I doubt the epoxy is conductive, just bog standard clear epoxy people use for decorative stuff.

[SimonMester]

So I made a little illustration.
I hope the colours make it obvious what is what. Alternating pole magnets, and a single coil to demonstrate my question.
Isn't it very problematic to properly align coil "legs" and the faces of the magnets?
Especially since you want each leg to be over the opposing pole, as symmetrically as possible?
Wouldn't somehow making the coils "curving" outwards to better match the faces of the magnets be worth the effort?
Is this why its often recommended for the gap between the legs to be 'slightly bigger than the magnet width'?

Is my brain expanding?

[SimonMester]

Probably the closest approximation I could do:




Would that work reasonably well, or the potential of the magnets is 'wasted' due to the coils not really being optimally covered/filling out the space?

EDIT:

So I worked around a bit, and figured to try and design a curved coil design.
The plan is the make a mold with the shape of the curved core.
Print the mold out of resin on my 3D printer, then mix epoxy and magnetite to cast the iron core.
Winding the coil around it might be a challenge, but I'll try make the mold as accommodating as possible.
Should probably use hot glue to temporarily stick 2 flat sheets on both ends to make the winding easier, and still be able to break those supports off when the coil is wound.

[JW]

https://www.fieldlines.com/index.php?topic=139752.0


So I worked around a bit, and figured to try and design a curved coil design.
The plan is the make a mold with the shape of the curved core.

[MattM]

You really want to avoid matching the number of coil legs to the magnet count.

[SimonMester]

You really want to avoid matching the number of coil legs to the magnet count.

Why is that exactly?
Gets really tricky aligning the two to overlap very well otherwise.

[SimonMester]

https://www.fieldlines.com/index.php?topic=139752.0


So I worked around a bit, and figured to try and design a curved coil design.
The plan is the make a mold with the shape of the curved core.

I'm not sure if what they say there will be very applicable for me, I do not have a coil winder, I do everything by hand.

[MattM]

You really want to avoid matching the number of coil legs to the magnet count.

Why is that exactly?
Gets really tricky aligning the two to overlap very well otherwise.

The resistance will be higher than if they mismatch.  Adrian has demonstrated many ideal designs to reduce cogging.

[SimonMester]

You really want to avoid matching the number of coil legs to the magnet count.

Why is that exactly?
Gets really tricky aligning the two to overlap very well otherwise.

The resistance will be higher than if they mismatch.  Adrian has demonstrated many ideal designs to reduce cogging.

Hmm. Well I'll try working something out, it's just the pointers I got so far seem to somewhat contradict each other:

- Legs should be as wide as magnet
- Opposite poles should be directly over the two legs at the same time
- Distance between magnets and coil legs should be the same as magnet width
- Should be as small air gap as possible

With all those, not having them match up in number seems quite difficult.

[joestue]

With all those, not having them match up in number seems quite difficult.

pull an alternator out of a junkyard and take it apart. most of them are 3 slots per pole and they have a 1:1 ratio of the coil width = magnet width.

now imagine removing the iron core of the stator, leaving the copper where it is. 3 coils, overlapping produce 3 phase without a cogging problem because there is constant energy flow out of the system. (no cogging without a core), but a single phase air core motor will have a cogging torque issue. it may not matter.

now, you can make a 3 phase winding from a single layer of coils.. you just space them out in the correct places and you still connect the coils in separate groups of 3. or 2, for a two phase machine.

anyhow you can fit nearly twice as much copper into a machine with overlapping coils as you can the single layer coils most people here build axial machines around.
you won't get twice as much power out of twice as much copper, because the coils have to be a bit bigger so they can flow around each other. more like 1.7 times as much power.

[SimonMester]

With all those, not having them match up in number seems quite difficult.

pull an alternator out of a junkyard and take it apart. most of them are 3 slots per pole and they have a 1:1 ratio of the coil width = magnet width.

now imagine removing the iron core of the stator, leaving the copper where it is. 3 coils, overlapping produce 3 phase without a cogging problem because there is constant energy flow out of the system. (no cogging without a core), but a single phase air core motor will have a cogging torque issue. it may not matter.

now, you can make a 3 phase winding from a single layer of coils.. you just space them out in the correct places and you still connect the coils in separate groups of 3. or 2, for a two phase machine.

anyhow you can fit nearly twice as much copper into a machine with overlapping coils as you can the single layer coils most people here build axial machines around.
you won't get twice as much power out of twice as much copper, because the coils have to be a bit bigger so they can flow around each other. more like 1.7 times as much power.

I wish we had junkyards like that.
I was going to build a single phase prototype starting out for simplicity while learning.
Are there any pictures you can maybe link me so I can understand better what you mean?

[MattM]

I don't remember anyone here suggesting to "Opposite poles should be directly over the two legs at the same time."

Generally people assert multiple phases wired in series or in star.  Three phase is pretty popular.

[SimonMester]

I probably misunderstood from here: https://www.fieldlines.com/index.php/topic,150770.msg1062236/topicseen.html#msg1062236

To quote:
"Looking at the pics of both the magnets and wire along with the information about magnets.
The ceramic ones are 2x+ wider than the Neos, this is why you're seeing a "higher" number.
Also the reason why it was stated the spacing between the Neos was too wide.
Just as one leg of the coil is passing over the N side of the mag the other should be passing the S side of the next mag."

I took to understand this as they need to lineup the legs with the alternating sides of the adjacent magnets.
But if that is not actually a criteria, then I can work with a non 1:1 ratio much easier.