Using FEMM and how much flux density for mid driver - Page 9

Baseballbat · 17th February 2013, 09:53 AM

Dave,

Quote:

Originally Posted by speaker dave

The "flux" rings deal primarily with the high 2nd harmonic distortion throughout the mid frequencies (not just tied to high excursion).

yes, but they also reduce inductance variations due to high excursions (Le(x)). These variations are the cause for high intermodulation distortion.

Quote:

I know that saturated top plates or pole pieces will reduce this effect but I don't think they reduce it as much as a large aluminum ring around the core pole, as used by JBL and McIntosh.

It does, but the principle is completely different. A shorting ring cancels flux modulation by shorting them (the name implies). A saturated pole piece does not even allow flux modulation. Of course, it is not so simple and cheap as a shorting ring. And it is quite easy to add a shorting ring to an existing design. Bringing an existing design into saturation is, at least, very difficult.

But the design process itself is not so difficult. Take the desired flux density in the gap (say 1T), design the magnetic circuit to have something around 1.2 T, and find material for the pole piece which is safely saturated at A_gap/A_pole * 1.2T, and you're fine (don't argue about these numbers, they're purely fictional). Of course, these 0.2 T are lost, so it is not ideal. And this means you have to put great efforst in optimizing the magnetic circuitry, and the magnet is slightly bigger (in this examples factor 1.2x). But look at the benefits:
- very good suppression of flux modulation
- very low inductance (close to an air coil)
- linearized Le(x)

To get the same benefits with copper/aluminum in the circuit, you'll need
- a shorting ring at the bottom of the pole piece (reduced Le(i) and Le(x); the overall inductance is not so much influenced)
- a copper cap (reduces inductances, but also flux)
- probably an additional ring directly below the gap

Is this cheaper? I don't think so.

Jef Nuyts · 17th February 2013, 11:52 AM

Quote:

Originally Posted by Kindhornman

Jef,
There is a really good paper I have but I am not sure of who posted the link to the paper discussing the various factors in loudspeaker nonlinearities. The title is as follows: Loudspeaker Nonlinearities-Causes, Parameters, Symptoms by Wolfgang Klippel, Klippel GmbH, Dresden, Germany, Klippel@Klippel.de

Thx glad I found it. This confirmed what I thought. You have primary distortion issues and all the rest. But first you need to eleminiate (or as good as it gets) the primary distortion issues before you get to the rest.

Kms(x) Bl(x) Le(x) Le(i) etc...
As first they mention the nonlineair suspension. => that is why i replace it with a full lineair suspension. So I think if you can improve the main problem and you have to sacrifice (not as much) a less critical problem then it would be better. Today I tested the lineairity of the EM suspension and it is 95% lineair till Xmax. As it is current driven it will not distort (also not improve) anything (of the primary distortion parameters). It eleminates also the resonance freq of the surround and also of the spider (two peaks).
Question. since it is current driven (DC) would it generate a second resonance frequentie (peak) on it's own? I think it would shift the primary resonance frequentie but not a different one. Am i correct?

nonlineair suspension.JPG

Kindhornman · 17th February 2013, 05:43 PM

Jef,
You have to remember that each speaker has its own set of compromises. Most speakers have a limit to the maximum excursion and this is commonly handled by using a rising rate spider that will slow the cone movement as it approaches the limits of the excursion. You also have to make sure in most designs that your former does not hit the back plate and this is the intention of the bumped back plate. The surround is also a contributor to the non-linearity and with common surrounds there isn't a lot of room for change. You have to have some form of surround for centering the outer diameter of the cone and also to attenuate the wave traveling along the cone as it reaches the outer diameter and wants to be reflected back at this point. So the choice of surround is at least as important as the spider selection for multiple purposes. Most 1/2 round surrounds will have a fairly limited excursion limit unless the outside diameter is increased to allow more cone movement. But then you have now introduced another problem, this surface area of the surround is often in antiphase to the cone output, the larger the surface area the greater the phase problem. If you use an M-roll or W-roll these suspensions are often much stiffer than the 1/2 round surround and now you add added stiffness to the suspension and need to add mass to the cone to counter this affect. It is all a series of tradeoffs, what can you live with and what is important to you. There are no magic bullets here.

Jef Nuyts · 17th February 2013, 06:25 PM

Quote:

Originally Posted by Kindhornman

Jef,
There are no magic bullets here.

I think there is. no spider and no surround

I have to thank everyone for the suggestions, learnt alot. I know what I have to know. I will post the results once it's finished.

Regards,

Jef

Kindhornman · 17th February 2013, 06:42 PM

Jet,
Though you say there is no surround that is probably a miss statement of fact. If there is no way for the outside of the cone to move in and out how can it move period. If you are using the cone itself as the surround, the same material that forms the cone itself you will find the problem when you actually get to the point of testing for resonant problems with the cone.

Jef Nuyts · 17th February 2013, 08:16 PM

@ kindhornman

maybe this will help to visualize.

dome.JPG

Kindhornman · 17th February 2013, 08:21 PM

Jef,
Okay I get the idea now. So you are counting on the dc current source to do everything then. I am not so sure how that is going to work as not only the centering force but also the damping force for the moving mass. I don't know how that is going to work and what actually sets the centering of the voice coil as I think you said the dc coil is the long coil short gap design? Perhaps I have that backwards but still it seem that the location for center will not have a precise positioning. Very strange indeed.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to use FEMM?	Spasticteapot	Planars & Exotics	23	11th February 2011 04:44 PM
Flux Density in Car SMPS Toroid	Preamp	Power Supplies	10	24th November 2010 01:32 AM
TS parameters when changing Flux Density	f4bok	Multi-Way	4	24th December 2008 11:33 AM
TS parameters when changing Flux Density	f4bok	Pass Labs	2	23rd December 2008 02:46 PM
Core saturation and flux density calculations	machinehead	Power Supplies	8	17th January 2006 07:45 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

17th February 2013, 05:43 PM	#83
Kindhornman diyAudio Member Join Date: Aug 2012 Location: Los Angeles, California	Jef, You have to remember that each speaker has its own set of compromises. Most speakers have a limit to the maximum excursion and this is commonly handled by using a rising rate spider that will slow the cone movement as it approaches the limits of the excursion. You also have to make sure in most designs that your former does not hit the back plate and this is the intention of the bumped back plate. The surround is also a contributor to the non-linearity and with common surrounds there isn't a lot of room for change. You have to have some form of surround for centering the outer diameter of the cone and also to attenuate the wave traveling along the cone as it reaches the outer diameter and wants to be reflected back at this point. So the choice of surround is at least as important as the spider selection for multiple purposes. Most 1/2 round surrounds will have a fairly limited excursion limit unless the outside diameter is increased to allow more cone movement. But then you have now introduced another problem, this surface area of the surround is often in antiphase to the cone output, the larger the surface area the greater the phase problem. If you use an M-roll or W-roll these suspensions are often much stiffer than the 1/2 round surround and now you add added stiffness to the suspension and need to add mass to the cone to counter this affect. It is all a series of tradeoffs, what can you live with and what is important to you. There are no magic bullets here.

17th February 2013, 06:42 PM	#85
Kindhornman diyAudio Member Join Date: Aug 2012 Location: Los Angeles, California	Jet, Though you say there is no surround that is probably a miss statement of fact. If there is no way for the outside of the cone to move in and out how can it move period. If you are using the cone itself as the surround, the same material that forms the cone itself you will find the problem when you actually get to the point of testing for resonant problems with the cone.

17th February 2013, 08:16 PM	#86
Jef Nuyts diyAudio Member Join Date: Jan 2013	@ kindhornman maybe this will help to visualize. dome.JPG

17th February 2013, 08:21 PM	#87
Kindhornman diyAudio Member Join Date: Aug 2012 Location: Los Angeles, California	Jef, Okay I get the idea now. So you are counting on the dc current source to do everything then. I am not so sure how that is going to work as not only the centering force but also the damping force for the moving mass. I don't know how that is going to work and what actually sets the centering of the voice coil as I think you said the dc coil is the long coil short gap design? Perhaps I have that backwards but still it seem that the location for center will not have a precise positioning. Very strange indeed.