Speaker magnetic Circuit

[ZeroResistance]

Following are few pictures of speaker construction





I would expect every speaker to be constructed like the top one. What I have practically noticed is that almost every speaker that I have seen so far has the magnet of larger diameter than the steel plates similar to that shown in the second figure.

I would have thought that
1. You would want to use every mm of flux that the magnet generates and therefore sandwitch the magnet completely like the first image.
2. The speaker that I have seen have the magnet jutting out by 2 -3 mm from all sides of the steel plates. I would guess that this is the area where you have maximum magnetism too since it is the outside edge of the ferrite ring magnet .

Why would the speaker manufacturer's use a bigger diameter magnet that the steel plates?

[Zero999]

All the speaker designs I've seen look like image 2. I suspect it's to make manufacturing easier, more than any other reason.

[T3sl4co1l]

Cost?  Steel is probably a bit cheaper than ferrite.

There's not much flux gained by extending the pole pieces around (to or past) the magnet; indeed there may be more lost as a result, because the fringing flux will extend out farther.

For a counter-example, consider old school AlNiCo magnets.  These have higher flux density (ferrite ~0.4T, AlNiCo ~1.2T), and higher cost too, I suppose (I mean, hey, they're not used anymore).  Double reason to surround them with steel.

Tim

[ZeroResistance]

All the speaker designs I've seen look like image 2. I suspect it's to make manufacturing easier, more than any other reason.

Agreed! Yes I thought so too initially but on second thoughts I wondered would'nt it be a waste of that magnet area.

Cost?  Steel is probably a bit cheaper than ferrite.

There's not much flux gained by extending the pole pieces around (to or past) the magnet; indeed there may be more lost as a result, because the fringing flux will extend out farther.

For a counter-example, consider old school AlNiCo magnets.  These have higher flux density (ferrite ~0.4T, AlNiCo ~1.2T), and higher cost too, I suppose (I mean, hey, they're not used anymore).  Double reason to surround them with steel.

Tim

I was thinking on the same lines as your second point, regarding the fringing flux. If the steel is extended till the edge of the magnet then the flux would probably leak out along the outer sides of the magnet (remember that the ferrite magnet is magnetised thru the thickness)
Speaker manufacturers would want to concentrate all of it on the inside where the voice coil is located.
However I have not recieved any definite confirmation on this, I wonder if this would be mentioned in some speaker design books.

[Seekonk]

A friend worked for a speaker company an they used ferro fluid in the gap.  They touted how great it was in advertisements.  He said the real reason was it made centering of the voice coil much easier and it was a cost savings procedure.

[Audioguru]

Ferro fluid is used to pass heat from the voice coil to the huge magnet structure. It also damps high frequency resonances.

[b_force]

Ferro fluid is used to pass heat from the voice coil to the huge magnet structure. It also damps high frequency resonances.

The first part is very true (hence air is a very good insulator)
The second part not so much.
Ferrofluid dampens the whole movement. This results in a lower Q factor on the lower end.
Sometimes it also results in a slightly lower BL factor (force factor), which gives a slightly lower top end frequency response.
That's actually easy to compensate with a little better force factor (magnet, or more efficient construction)

Ferrofluid can't be used in speaker with to much excursion, so therefore it's only being used in tweeters.

Going back to the first post.
I have seen thousands of different designs. It all depends how the magnetic field is being distributed and cost saving vs performance.
I would suggest reading; Loudspeaker handbook by John Eargle & loudspeaker and headphone handbook by John Borwick

[PChi]

Another book worth  looking at is the 'Loudspeaker Design Cookbook' http://www.cc-webshop.com/Loudspeaker-Design-Cookbook-AA-BKAA068.htm which despite the title does have some information on loudspeaker magnet design.
Unfortunately HiFi has gone down the marketing route rather than good engineering (in my opinion) so cost of the loudspeaker drive unit is important along with the bullshit. Performance is less important.

[Zero999]

Smaller speaker designs often don't have a spider, coil former or dust cap. The diaphragm is one piece of vacuum formed mylar, with the voice coil attached directly to it. The coil is made my winding the wire round a suitable rod and coating it in resin. When the resin has cured, the coil is removed from the rod and glued to the speaker's diaphragm. This simplified speaker design is common in headphone transducers, dynamic microphones and small speakers used in toys and mobile phones.

[b_force]

Smaller speaker designs often don't have a spider, coil former or dust cap. The diaphragm is one piece of vacuum formed mylar, with the voice coil attached directly to it. The coil is made my winding the wire round a suitable rod and coating it in resin. When the resin has cured, the coil is removed from the rod and glued to the speaker's diaphragm. This simplified speaker design is common in headphone transducers, dynamic microphones and small speakers used in toys and mobile phones.

This is also used in many speakers <1 inch, like bluetooth speakers/boom boxes.
In that case a spider is also not really useful anymore, since the stiffness of the other materials can keep everything centered.
Klippel has some very interesting articles about the fact that with these small sizes, air compression in the design is actually a significant parameter.
(something you don't really care about with 6" or so)

[RandallMcRee]

If you model the circuit in femm, or another magnetic modeler, you find that the flux concentration is highest at the voice coil (and fringing on the outside is a minimum) if the steel poles do not go all the way to the edge. So, yes, extending the pole to the edge makes the fringe path 'easier' and there is then less flux across the voice coil.

You can make the steel poles pieces much much thicker (or out of much more expensive alloy with better magnetic properties) but the cost-benefit ratio favors the designs you have shown.

Thicker pole pieces would also lead to an increased length dimension of the voice coil. This is not desirable for structural reasons, so that is another balancing constraint.

[b_force]

In some cases more 'steel' is also needed for thermal purposes.
More thermal mass (= equivalent to a capacitor), means better power performance.
This is one of the reasons why most PA drivers are made with ferrite magnets and not neodymium.
(besides the costs of course)