Page 4 of 4 FirstFirst ... 234
Results 46 to 56 of 56

Thread: alnico/ferrite drivers

  1. #46
    whgeiger
    Guest

    Not Likely

    Quote Originally Posted by scott fitlin
    But that JBL did take the time, and invest the money, into making a differing diaphragm says to me JBL felt it made an audible difference, and did what they may have felt was more suitable for each diaphragms intended application.

    I am of the opinion that minute design changes in both driver assemblies, and diaphragms and materials and shapes and ribs, etc, do make audible differences.
    SF,

    The motivation totem for driver design changes looks something like this:

    1) Reduce field returns, particularly for those units still under warrantee. Diaphragm fracture, surround fatigue, voice coil failure (due to heat dissipation problem) are typical causes.

    2) Reduce manufacturing cost (acoustics enhanced or otherwise 'negligibly' affected).

    3) Design has nasty resonances (response irregularities) that need to be suppressed, or unit drifts out of specifications during the warrantee period.

    4) Match driver to new horn (or enclosure) design. These changes result in new driver models rather than variants to those already existing.

    Note that 'how it sounds' is at the change totem bottom.

    Regards,

    WHG

  2. #47
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,720
    Hey, for them it's a business not a hobby.

    Widget

  3. #48
    whgeiger
    Guest

    AlNICo vs. Ferrite Magnet Oxymoron

    To All,

    There is sufficient evidence to suggest that different magnet materials will yield different sounding drivers, since design of the entire magnetic circuit is dependant on the magnet material used. The proposition, that two drivers are identical in all respects, except for the use of different magnetic materials, is at best, an oxymoron.

    For example, the operating point for ferrite magnets, with their steep sloped demagnetization curves, is radically shifted as the voice coil moves back and forth through the magnetic gap. This shifting is manifested as a modulation to the magnetic flux ‘seen’ by the voice coil when it is in motion. To counter this effect, a flux-stabilizing (copper or aluminum) ring is wrapped around the pole piece body to suppress such action. This design, under these circumstances, will not be functionally or physically identical to a driver motor designed around an AlNICo V magnet nor any other magnetic material.

    Regards,

    WHG

  4. #49
    Senior Member frank23's Avatar
    Join Date
    May 2005
    Location
    The Netherlands
    Posts
    356
    Quote Originally Posted by whgeiger
    To All,

    There is sufficient evidence to suggest that different magnet materials will yield different sounding drivers, since design of the entire magnetic circuit is dependant on the magnet material used. The proposition, that two drivers are identical in all respects, except for the use of different magnetic materials, is at best, an oxymoron.


    Regards,

    WHG
    That's what I heard once someone telling me, that the magnetic gap properties of an alnico magnet driver were different from those of a ferrite magnet driver

    isn't is also so that the magnet itself is in a whole different location in an alnico driver [in front?] than in a ferrite driver [around?]

    or am I typing nonsense right now?

    does someone have a cut through picture of a ferrite magnet compression driver? to compare with this alnico driver?

    frank
    Attached Images Attached Images  

  5. #50
    Member
    Join Date
    Sep 2004
    Location
    Australia
    Posts
    98
    Quote Originally Posted by whgeiger
    For example, the operating point for ferrite magnets, with their steep sloped demagnetization curves

    I'm reasonably sure that the common view of this is incorrect so I'll state my idea here for confirmation or criticism.

    The voice coil gap geometry sets the reluctance of the magnetic circuit and the designer's specification sets the flux.
    To minimize the size (and cost) of the magnet there is a definite operational condition that specifies both B and H. This is the point at which the BH product is maximum. There must be a match between the magnet and the gap -similar to the way impedance is matched to maximizes power. The optimum magnet dimensions are matched to the gap with the field return structure.
    The magnet thickness and area will vary for different materials but the B/H relation as seen by the voice coil should be the same. The material properties will vary but the actual magnets will be equivalent because they both must match the same gap.
    So it's not because ferrite material has a steeper slope that it had more flux modulation problems (at least to first order).

    I believe the main reason is that there are flux stabilizing current flows in Alnico drivers. These are much lower in ferrite drivers due to the lower conductivity of ferrite and possibly the different field return structure required by the ferrite's different magnet proportions is also less able to support stabilizing currents.

    [QUOTE=whgeiger]
    is radically shifted as the voice coil moves back and forth through the magnetic gap. This shifting is manifested as a modulation to the magnetic flux ‘seen’ by the voice coil when it is in motion.[/QUOTE]

    I don't think that the motion of the voice coil is a factor here. Flux modulation would still happen if the VC was blocked. Of course it wouldn't matter as much


    Dave

  6. #51
    whgeiger
    Guest

    Refutation

    [QUOTE=Dave Zan][/size][/font]
    I'm reasonably sure that the common view of this is incorrect so I'll state my idea here for confirmation or criticism.

    The voice coil gap geometry sets the reluctance of the magnetic circuit and the designer's specification sets the flux.
    To minimize the size (and cost) of the magnet there is a definite operational condition that specifies both B and H. This is the point at which the BH product is maximum. There must be a match between the magnet and the gap -similar to the way impedance is matched to maximizes power. The optimum magnet dimensions are matched to the gap with the field return structure.
    The magnet thickness and area will vary for different materials but the B/H relation as seen by the voice coil should be the same. The material properties will vary but the actual magnets will be equivalent because they both must match the same gap.
    So it's not because ferrite material has a steeper slope that it had more flux modulation problems (at least to first order).

    I believe the main reason is that there are flux stabilizing current flows in Alnico drivers. These are much lower in ferrite drivers due to the lower conductivity of ferrite and possibly the different field return structure required by the ferrite's different magnet proportions is also less able to support stabilizing currents.

    Quote Originally Posted by whgeiger
    Quote Originally Posted by whgeiger
    is radically shifted as the voice coil moves back and forth through the magnetic gap. This shifting is manifested as a modulation to the magnetic flux ‘seen’ by the voice coil when it is in motion.[/QUOTE]

    I don't think that the motion of the voice coil is a factor here. Flux modulation would still happen if the VC was blocked. Of course it wouldn't matter as muchhttp://audioheritage.org/vbulletin/i...lies/smile.gif


    Dave
    DZ,

    These statements, while mostly true for static conditions, obfuscate, deny or otherwise ignore what takes place within the magnetic circuit under dynamic conditions when the voice coil wire ‘cuts’ through that magnetic gap. So, the claim remains that the acoustic signal so generated, contains distortion products whose signature is unique to each circuit configuration and that such signatures are audible to a discriminating listener.

    So that there is no further equivocation on the matter, the following excerpts from Douglas Button’s paper [1] and John Eargle’s book [2] are provided forthwith. For those that want to persue magnetic circuit design for loudspeaker drivers, an additional reference to Eric Guarin’s work [3] is also provided.

    Regards,

    WHG

    Quotations:

    “3. The Magnetic Assembly” ([1], p. 9)

    “The magnetic assembly is probably the most taken for granted part of a loudspeaker assembly. It is this author's opinion that it happens to the most exciting part of the design. The magnetic structure is in fact not a simple static field that is merrily supplying the voice coil with needed flux lines to modulate the cone. The materials used and geometry of the magnetic structure can radically alter the impedance curve and completely dominate the 2nd. and 3rd. harmonic distortion products in mid and lower frequencies. The interaction of the voice coil and it's surrounding is very complex and can change the sound of a driver dramatically.”

    “3.4.1 Flux Modulation” ([2], p. 52)

    “This occurs when the magnet is alternatively over-magnetized and reversely magnetized by the cyclic signal current in the voice coil itself. This effectively amounts to shifting of the instantaneous operating point up and down the demagnetization curve. It is minimal when the demagnetization curve itself is fairly flat, as in the upper portion of a typical AlNiCo V demagnetization curve.”

    “However, in the case of a ferrite magnet, with its high-sloped demagnetization curve, no such solution is available. When JBL, like all other manufacturers, was forced to develop a new ferrite motor structure when cobalt became too expensive, they added a large conductive ring, made of aluminum, at the base of the pole piece. Through transformer action, the voice coil signals that are induced into the magnet structure will set up a countercurrent in the aluminum ring. The ring, a single turn with a cross-sectional area of about 1 cm^2, has an extremely low resistance; consequently, there is substantial current flow through the ring at high signal levels. As in all induction phenomena, the action of the induced current is to counter the effect that produced it in the first place, thus the tendency for flux modulation is greatly reduced at the price about 1 dB of overall magnetic efficiency”

    References:

    [1] File: AESP3192.pdf
    Title: Design Parameters and Trade-Offs in Large Diameter Transducers
    Author: Douglas J. Button
    Affiliation: JBL/Harmon Loudspeaker Manufacturing, Northridge, CA
    Publication: AES-P, No. 3192, Cnv. 91, Sep-1991
    Abstract (1): In the design of high level large format transducers, a variety of interactive parameters define how a loudspeaker will sound at different power levels; voice coil diameter, former and wire material; cone material, geometry and mass; magnet size and material; magnetic gap geometry; dome geometry and material; cone surround linearity and damping; spider linearity; and the mechanical integrity of the entire moving system.
    Abstract (2): These factors define; inductance; hysteresis and eddy current losses; spectral response and high frequency extension; cone break-up and modal behavior; power compression; flux modulation; low frequency linearity; and of course, sensitivity. This paper will investigate the trade-offs and design elements that influence the sound of modern high-level transducers by evaluating current designs.
    Abstract (3): The purpose of the paper is to shed light into some commonly held preconceptions as to why a given engineer may have made the design choices he or she has made.

    [2] Title: Loudspeaker Handbook
    Author: John M. Eargle
    Affiliation: JBL/Harmon International
    Publication: Book, Kluwer Academic Publishers, Norwell, MA, 1996
    ISBN: 0-412-09721-4
    Abstract: A well-balanced treatment covering both the technological and practical aspects of the basic component design and construction of loudspeakers is presented. Rigorous mathematical/graphical explanations concerning the function of loudspeakers are provided and insight into modern testing methods is reviewed.

    [3] File: AESP2979.pdf
    Title: Design and Optimization Considerations for Speaker Magnets
    Author: Eric Guarin
    Affiliation: Armadillo Audio, Anaheim, CA
    Publication: AES-P, No. 2979, Cnv. 89, Aug-1990
    Abstract: A production ceramic magnet needs to be downsized for possible automotive use. Simultaneously, high energy magnet materials are to be evaluated as alternatives. Magnet design basics, pole and plate shaping, and production feasibility considerations combine to yield more efficient designs. Computer software implementation aids in the design process.

  7. #52
    Member
    Join Date
    Sep 2004
    Location
    Australia
    Posts
    98
    As someone once said "What we have here is a failure to communicate" so I'll expand my thesis - somewhat tediously!
    Proposition 1.
    I think most people have read statements like the one in Eargles book and conceive the demagnetization curve mentioned, and hence the flux modulation, is inherently a property of Alnico - similar to the way the conductivity of copper alters with temperature -essentially the same (percent) for any copper voice coil.
    They therefore expect the flux modulation to be much the same for any Alnico magnet.
    (To be pedantic this is actually an opinion about what people think. It's really by way of introduction rather than essential to the thesis.)

    Proposition 2.
    I believe this is a misconception and that what determines the response to flux modulation is a property of the entire component - very similar to the _resistance_ of the voice coil - and therefore that this can be built to practically any value.
    (This is a statement of physics.)
    If Prop. 2 is true then it raises an obvious question. If the "magnetic resistance" (in fact called the reluctance) can be built to specification then why were ferrite magnets initially specified that had more flux modulation than Alnico and that required a new structure? The obvious comparison is between ferrite and Alnico magnets where both are "equivalent" in the sense that they are the cheapest (minimum amount of magnetic material) that will create the same flux in the same gap. This leads to

    Proposition 3.
    A ferrite magnet "equivalent" to an Alnico magnet will have the same reluctance and hence incremental sensitivity to flux modulation - to first order and small values of delta.

    Now I realize there are complications beyond this. The curves of Alnico and ferrite are non linear and different. I just want to establish the basics before I consider the complexities because I don't think even the basics are well known and I want to understand what the physics really are.
    If the propositon is true then we can better understand what _really_ causes the differences (Which I don't deny - JBL didn't reengineer the transducers just for fun). The two most obvious explanations are -
    1. That the magnets are operated dynamically so far from the static conditions that the distinctive shape of the BH curves does indeed matter.
    This is most people's intuitive expectation but I don't think it likely to be a major factor. A magnet operated significantly into the non linear zone will tend to be irreversably demagnetized.
    2. That the _conductivity_ of the Alnico meant that significant currents flowed in these tranducers that stabilized the flux. This is less obvious but I believe that this is actually the case for several reasons.
    First of all is that it seems that the JBL engineers were a bit surprised by the differences. The shape of the BH curve is their daily bread so I expect it was the less obvious.
    Secondly, one of Don's comments about the Alnico/ferrite work practically stated as much (I meant ask him to confirm this but now can't find the quote - hello Don?).
    Thirdly is that (as Button says in his quote) the magnetic distortion is most present at lower frequencies. The magnetic curve of ferrite doesn't alter much over this frequency span. If the slope of the BH curve doesn't alter then nonlinearities dominated by this source would not decrease with frequency. But eddy currents do increase with frequency so a frequency dependent decrease in flux modulation would be expected.
    Finally. The MMF from the voice coil in a woofer is quite substantial. I calculate it would be sufficient to significantly demagnetize the Alnico if there were not substantial flux stabilization currents. This is substantiated by the discussion of the development of the 1500AL woofer which needed the Aluminium flux stabilization coil in addition to internal currents.

    Do we have any competent physicists or Elec.E.s to comment on the thesis? Because if I'm correct then I would say the Eargles book is not as clear as it could be.(by coincidence I ordered a copy of the 2nd Edition a few weeks back but it hasn't arrived yet so I can't comment on the full text).



    Quote:
    Originally Posted by whgeiger

    These statements, while mostly true

    What of the _mostly_ true bits aren't true?

    for static conditions,


    It's precisely the dynamic flux modulation that I modelled.


    obfuscate, deny or otherwise ignore what takes place within the magnetic circuit under dynamic conditions when the voice coil wire ‘cuts’ through that magnetic gap.

    I didn't think I obfuscated! As I said - There will be dynamic (time variable) flux modulation even if the voice coil is stationary. The effect of the voice coil wire as it cuts the magnetic field is reflected in the back EMF. This is usually modelled as a linear circuit element and it is other circuit elements that reflect the nonlinear magnetic properties (frequency dependent resistive losses and inductances) Do you have a reference for this idea?

    So, the claim remains that the acoustic signal so generated, contains distortion products whose signature is unique to each circuit configuration and that such signatures are audible to a discriminating listener.
    So that there is no further equivocation on the matter

    I'm an equivocator and an indiscriminate listener too?! In fact, as I said above, this is not my position at all.

    “3. The Magnetic Assembly” ([1], p. 9)

    “The magnetic assembly is probably the most taken for granted part of a loudspeaker assembly. It is this author's opinion that it happens to the most exciting part of the design. The magnetic structure is in fact not a simple static field that is merrily supplying the voice coil with needed flux lines to modulate the cone. The materials used and geometry of the magnetic structure can radically alter the impedance curve and completely dominate the 2nd. and 3rd. harmonic distortion products in mid and lower frequencies. The interaction of the voice coil and it's surrounding is very complex and can change the sound of a driver dramatically.”

    No dispute here! This is why I want to analyze dynamic flux modulation.
    I don't understand the 3rd harmonics yet.
    Structure is so important it can even trump materials.

    “3.4.1 Flux Modulation” ([2], p. 52)

    “This occurs when the magnet is alternatively over-magnetized and reversely magnetized by the cyclic signal current in the voice coil itself. This effectively amounts to shifting of the instantaneous operating point up and down the demagnetization curve. It is minimal when the demagnetization curve itself is fairly flat, as in the upper portion of a typical AlNiCo V demagnetization curve.”

    “However, in the case of a ferrite magnet, with its high-sloped demagnetization curve, no such solution is available. When JBL, like all other manufacturers, was forced to develop a new ferrite motor structure when cobalt became too expensive, they added a large conductive ring, made of aluminum, at the base of the pole piece. Through transformer action, the voice coil signals that are induced into the magnet structure will set up a countercurrent in the aluminum ring.

    I think this duplicates the situation that tends to occur in the conductive Alnico itself.

  8. #53
    Member
    Join Date
    Sep 2004
    Location
    Australia
    Posts
    98
    Looks like everyone's lost interest in this but just for the archive.
    I've had a look at John Eargle's book and there are definitely mistakes in the chapter on magnetics. I'm finally sure that the analysis I posted above is correct - which is a relief, because at first I couldn't believe that it was his mistake rather than mine!
    That's all unless anyone wants the tedious details.

    David

  9. #54
    whgeiger
    Guest

    MFR

    Quote Originally Posted by Dave Zan
    Looks like everyone's lost interest in this but just for the archive.
    I've had a look at John Eargle's book and there are definitely mistakes in the chapter on magnetics. I'm finally sure that the analysis I posted above is correct - which is a relief, because at first I couldn't believe that it was his mistake rather than mine!
    That's all unless anyone wants the tedious details.

    David
    DZ,



    When considering the differences between pros of the author and that of the critic, suspect that what follows this post will constitute yet another inept exercise in ‘nit-picking’ an example rather than dealing with the issue at hand. When doing so, see if you can be as brief as the author you criticize. Irrespective of allegations to the contrary, the previous submission remains flawed and not worthy of the effort that would be required to correct the sophomoric errors it contains.



    WHG

  10. #55
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Quote Originally Posted by Don McRitchie
    There shouldn't be a significant difference, if any, assuming that the throat and phase plug geometry is the same in both the ferrite and Alnico versions. I believe this is the case with the 2420/2425 and 2440/2445, but this is worth checking. Of course the big issue that will have to be checked is whether an Alnico, used for comparison, still meets the factory specs for flux density.

    In bass drivers, there is a difference is sonic character because of differences in flux modulation and temperature curves of the different magnetic materials. In a compression driver, the coils are generally underhung so that flux modulation is mitigated to low levels. The lower power handling of compression drivers results in lower operating temperatures so that the temperature curve difference has minimal effect. Nonetheless, I would be interested in the results of any experiments.
    I think Don's explanation says it better than any faded white paper.

    Ian

  11. #56
    whgeiger
    Guest

    Non Sequitur

    Quote Originally Posted by Ian Mackenzie
    I think Don's explanation says it better than any faded white paper.

    Ian

    IM,

    The reference is to a book, now available in a most recently published edition. So, the characterization "faded white paper" remains a fallacious attribution of no consequence.

    Regards,

    WHG

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. The Great Alnico / Ferrite Debate
    By Don McRitchie in forum General
    Replies: 1
    Last Post: 07-16-2007, 04:53 AM
  2. Model 14 drivers T/S
    By WildWest in forum Lansing Product General Information
    Replies: 9
    Last Post: 04-04-2004, 04:02 PM
  3. Compression drivers with cracked throats
    By Doogster in forum Lansing Product General Information
    Replies: 1
    Last Post: 12-16-2003, 01:11 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •