Great link.
The cutout in the center of the pole piece is exactly how we shim pole pieces. They did a simple cut. So the Field does wiggle a bit. They could easily optimize that by a few iterations.
What I wanted to point out is how poorly they use the iron of the pole piece. Their design crowds the flux at the corner where the back plate meets the pole. That crowding carries all the way to the gap area. If they necked the pole right below the first gap, they will make the two gap fields more equal. If they added iron at that corner, they could reduce the flux intensity as well.
Because they go higher up the curve there, the iron will begin to lose permeability.
Jn
I had to look that up. Ended up being slightly disappointed as expected something like pantograph but a bit more steam punk.
![afeed[1].gif](https://www.diyaudio.com/community/data/attachments/613/613888-20059e811cc7cbc0f63ea2cdf4233512.jpg?hash=IAWegRzHy8 "afeed[1].gif")
Perhaps drawing it on the left side of the schematic made a difference??
I assume the 500K could be lowered if needed.
jn
Last edited:
Fast aren't you, In hindsight didn't want to get into a discussion of this circuit.
jneutron, is it better for the magnet assembly to have high permeability or low permeability? Wouldn't you want the permeability to approach the permeability of air? If the iron saturation is increased at the poles, wouldn't that reduce certain distortions? A shorting ring can make it lower than air, so it's possible to go too far.
The permeability is used to control where the flux goes. When the iron starts to saturate, its permeability reduces, the flux starts to leak outside of the circuit. The tradeoff is, how much iron is needed for the level of field required.
The VC inductance tends to make the flux in the pole change, which changes the permeability, which causes non linear B as a result of excitation currents.
jn
The VC inductance tends to make the flux in the pole change, which changes the permeability, which causes non linear B as a result of excitation currents.
jn
Here is an example of a speaker with the XBL**3 motor, with a Bl graph:
http://www.tektron-italia.com/various/A3StatsGraph.pdf
I view the XBL motor sort of the same way I view trying to reduce the distortion of a BJT output stage by tweaking the emitter resistors. It's never going to be right.
For a normal motor the distortion level is monotonic with output because the Bl curve doesn't turn up as output increases. For the XBL motor, they are trying to herd cats into a straight line and while the increase in headroom is useful, I wonder about the possible high order intermodulation this results in.
Compared to the BL curve for an underhung coil, it seems like the XBL is mainly useful for headroom extension, it's actually possible the performance at low levels could be worse than for a well designed conventional motor.
http://www.tektron-italia.com/various/A3StatsGraph.pdf
I view the XBL motor sort of the same way I view trying to reduce the distortion of a BJT output stage by tweaking the emitter resistors. It's never going to be right.
For a normal motor the distortion level is monotonic with output because the Bl curve doesn't turn up as output increases. For the XBL motor, they are trying to herd cats into a straight line and while the increase in headroom is useful, I wonder about the possible high order intermodulation this results in.
Compared to the BL curve for an underhung coil, it seems like the XBL is mainly useful for headroom extension, it's actually possible the performance at low levels could be worse than for a well designed conventional motor.
The question is where on the saturation curve do we get the most linear behavior. I think that either very light saturation or very high saturation would be those points, although perhaps the light saturation scheme would be worse because the magnitude of the nonlinear inductance would be much greater.
I believe one of the aims of the Ryu field coil project is to saturate the steel to the greatest extent possible (don't quote me on that).
Project Ryu - DIY Field Coil Loudspeaker
What's really interesting is that neo has a permeability about 1.05 give or take. That means that a neo magnet speaker path reluctance is dominated by the dimensions of the magnet itself. For a ferrite, the gap is the dominating reluctance.
The gap field of 1.6 tesla is pretty much what we use as a limit for copper coil magnets. Pole tips start to misbehave when saturation starts. For speaker magnetics, I'd have to think on it a while..not sure of the answer.
jn
The gap field of 1.6 tesla is pretty much what we use as a limit for copper coil magnets. Pole tips start to misbehave when saturation starts. For speaker magnetics, I'd have to think on it a while..not sure of the answer.
jn
Yes. With some ac bal trim pots in strategic places, does very well .... Can be put into IC package excepting OPS.
-RNM
Last edited:
JNeutron ----
To get an appreciation of how old I am .... when i was a kid, we had a playback system with one huge Electro-Magnet speaker in the console.
How could we do better today using electro-magnet windings for speaker?
Or even a combination perm and electro magnets? Right now, if you add a big value Cap coupled series output to speaker, you can put DC bias (+ or -) on speaker to null Harmonic distortion. An electromagnet bias coil could do same without the need for Dc blocking cap.
THx-RNMarsh
To get an appreciation of how old I am .... when i was a kid, we had a playback system with one huge Electro-Magnet speaker in the console.
How could we do better today using electro-magnet windings for speaker?
Or even a combination perm and electro magnets? Right now, if you add a big value Cap coupled series output to speaker, you can put DC bias (+ or -) on speaker to null Harmonic distortion. An electromagnet bias coil could do same without the need for Dc blocking cap.
THx-RNMarsh
Last edited:
I had one in my bedroom. I loved the spider geometry it used, none of this corrugated stuff.
I'd just stick to perms myself. Why make it more complex?
We could rube Goldberg something where we look at where the coil is and goose the main field to linearize BL, but to me less active is better...
And yes, I'm still talkin bout speakers....😀
jn
I had a friend whose grandfather had one in the door of a closet driven by 3 PP pairs of 45's and interstage transformers. He tested oil caps for ESR by charging them to 500V and listening for the loudest spark when shorted (he was an MIT EE BTW). No caps in the signal path and no servos either.
- Status
- Not open for further replies.