A bunch of questions regarding radially magetized ring magnet.

[TinselKoala]

Well, it depends...  See this hard disk picture with its back plate behind it (I took the pictures
from this site: http://www.reuk.co.uk/wordpress/wind/hard-disk-drive-magnets-for-wind-turbines/ )
You know such magnets are quadrapole ones, they act like two magnets stuck together side by side.
On one face there is a South pole on the left and a North pole on the right, on the other face there is
a North pole on the left and a South pole on the right. And the back plate connects the N and S poles
that are at the bottom face. Some time ago I checked with an office pin how magnetic the back side of
such a back plate + magnet assembly had been. The pin was not attracted to the back of the backing plate,
this means the plate surely had high permeability and was thick enough to avoid saturation so a negligible
amount of flux could exit (i.e. leak) from the back of the plate.

I have not had a chance to have such unmagnetized material so I do not know, I assume they should have
a high magnetic permeability before the magnetization.  Such materials should have the same rigid, hard
and next to impossible to form_to_shape properties like the Neo or any other permanent magnets have.
I think even if one had say metglas material, he would have a hard time to shape it.

One more notice: if you use magnets to complete a magnetic circuit, you need to stack several of them
to reduce their own leakage too.   Also, at most uncovered surfaces of a magnet the lines of flux can freely
come out because all the body of a magnet is in an almost complete saturation, while the same lines of flux
'have a hard time'  to come out from a highly permeable soft iron material like the backing plate of a HD magnet. 

Gyula

The metal backing plate shown is made of "mu-metal" and indeed is very high permeability. This is how it acts as a "magnetic
shield", by "sucking up" the magnetic field into itself and making the connection between the opposite polarities with very very
little flux leakage. The "mu-ness" of the alloy is created by a special heat-treatment and annealing schedule, and any working
of the finished piece, such as the bending shown in the photo, or sawing or other cold-working, reduces or even destroys
the "mu-ness" of the material.

Here's an interesting motor design that bears some studying and thoughtful consideration.

[gyulasun]

Hi TinselKoala,

Thanks for the additional comments on the backing plate. The guy showing how he removed the magnets from the backing plates by bending the plate coners with vice grips may have been interested to scavenge only the magnets for building wind generators and may not have needed the plates and / or was unaware of the ill effects of abusing the plates mechanically.

Indeed you show an interesting motor design. Surely the backing plates do their job and keep the stator magnet fields in a closed magnetic circuit for all the outside poles, and the radially arranged copper stripes (i.e the rotor "coils") are sandwiched between the inner facing poles, within a narrow air gap. This way the rotor can have a very low self inductance hence the brush sparking can also be at a minimum due the low counter emf. Due the total lack of any ferromagnetic material in the rotor, there can be no cogging at all, no rotor core loss and there can be very little  I²R heat loss in the copper stripes.

Gyula

[Magluvin]

I  have one of those motors, and we talked about them before. They are unique in that hey are coreless armatures and have advantages as such that core losses are reduced and they can change speed very quickly, in this case a car cooling fan that accelerates very quickly. I wanted to make a new holder and mount neos in place of the poles you see in the multi pole ceramic in the pic above. Neos are stronger than the ceramic shown and the return path casing may have difficulty containing the fields as well as it handles the ceramic. What I wanted to do was mount neos on both sides of the armature and only have outer core return like a large washer on each side which would link all the outer poles to opposing poles of each side.

Similar to speaker motors, the ceramics rely on large surface area and thickness to obtain the desired field  density at the gap. There are guys trying to copy some companies speaker motors where the ceramic is replaced with multiple neo disks between the top and bottom plates. They take the speaker apart and separate the plates from the ceramic mag and install the disks and put it back together.  I think the mistake is that the plates will have a harder time containing the field than the ceramics.

I hadnt tested the fan disk motor before I took it apart to see if there was any outside leakage. And it is a different monster compared to the speaker motors as the disk motor casing is a closed system but the speaker motor is not. Most speakers back when crts were popular had mag shield cups covering the speaker mags and they did well with leakage containment.

I have some pics of fem analysis of a couple speaker motors. Ill get those up when I have time in a day or so. Im swamped.  One shows a very thick core return and another shows 2 different core changes that maximized the fields in the gap compared to the other.  It is a little misleading when you see the cross section of the mag having like 10 lines in the mag and having most all taking the core path to the gap, with only a few strays out of the core. But as many of us have seen in the past, that leakage can be detrimental to a color crt at a distance. If the field lines were shown to be denser than 10 field lines, the leakage lines would be way more than the fem shows in those examples.

Ill get those up soon.

mags

[Magluvin]

Here are the fem pics.

Ill get back on later to talk about them

Mags

[Lakes]

I've never head of speaker motors before now, are they some sort of actuator?

Do you have a link?