About the use of ferrofluid in commercial drivers: Monacor seems to use it quite much! I just started taking those sp-200x full range drivers apart and what was in the air gap - ferrofluid! Now I believe that there's ferrofluid also in my friend's sph-300tc 12" subwoofers, because it read "ferrofluid" on the side of the box tey came in. And Monacor doesn't say anywhere that sp-200x has the stuff in. I wonder how many models are there which actually have ferrofluid in the gap 
If a phase plug addresses the issue of interference at high frequencies due to cone-wall reflections, can this also be alleviated with a flat diaphram?
So far, the only problem I see with a flat diaphram is its rigidity due to it's shape, as opposed to a cone.
So far, the only problem I see with a flat diaphram is its rigidity due to it's shape, as opposed to a cone.
Cheers Bill. There's enough material there to keep me occupied for a while
Here's a pic of the latest motor. 2T in the air gap. Any ideas on how I can model the effects of back EMF? I'm not sure how to implement a farady ring into the design.
The problem I'm having with this motor is the diameter of the former that it imposes (>50mm), and that's with a pole vent of only ~6mm. What's a safe vent size to avoid noise?
Here's a pic of the latest motor. 2T in the air gap. Any ideas on how I can model the effects of back EMF? I'm not sure how to implement a farady ring into the design.
The problem I'm having with this motor is the diameter of the former that it imposes (>50mm), and that's with a pole vent of only ~6mm. What's a safe vent size to avoid noise?
Attachments
I haven't attempted that with FEMM yet. You might want to put that question to the FEMM yahoo group (and then share your findings with me!
).Depends on frequency/SPL. I'd suggest sidestepping the issue with a phase plug--no pole vent needed. (Another side benefit of ferrofluid: it will provide an air seal around the VC.)
Bill
As far as faraday rings (a good idea, BTW), I'd suggest the overkill approach: Face the the entire gap--top plate and pole--with copper. It means you'll have to spec more magnet (your effective gap spacing will be wider), but you'll be largely rid of B-field modulation distortion products.
Bill
Bill
I finally got round to making a crude basket so that I could better test the suspensions ideas with a proper cone. Everything is aluminium as it was free .
It's simply two rings (where I'll sandwhich the surround) held up by some long bolts. The thin strips attached to the magnet assembly are also aluminium - a bonus as they won't carry magnetic flux away. It feels quite rigid.
I also had the insane idea of making a paper mache diaphragm. I used a funnel as a mould (even though the cone now looks more like a horn) and the amusing part is that I replaced the flour with cement! It turned out quite rigid and light. Shame about the shape - but on to mark II...
Is there an optimum cone angle, or formulas to model the effects of different cone angles?
. It's simply two rings (where I'll sandwhich the surround) held up by some long bolts. The thin strips attached to the magnet assembly are also aluminium - a bonus as they won't carry magnetic flux away. It feels quite rigid.
I also had the insane idea of making a paper mache diaphragm. I used a funnel as a mould (even though the cone now looks more like a horn) and the amusing part is that I replaced the flour with cement! It turned out quite rigid and light. Shame about the shape - but on to mark II...
Is there an optimum cone angle, or formulas to model the effects of different cone angles?
Attachments
SWEET!!!
I suspect that the best shape is torodial, such that the hole of the donut is the same size as the voice coil, but the outer diameter is really, really big, and the shape is truncated to fit your application. Now, how you would fashion a torus that large for casting paper mache, I don't know...
BTW--orthodontic elastics make great suspensions. Call local orthodontists and ask if they'll just give you some.
I suspect that the best shape is torodial, such that the hole of the donut is the same size as the voice coil, but the outer diameter is really, really big, and the shape is truncated to fit your application. Now, how you would fashion a torus that large for casting paper mache, I don't know...
BTW--orthodontic elastics make great suspensions. Call local orthodontists and ask if they'll just give you some.
As far as the suspension goes, has any one ever tried dual foam/rubber/cloth surrounds? Its just an idea, but maybe you could eliminate the spider by having one conventional surrond at the edge of the cone and a smaller one as some point closer to the voice coil. The basket would have to be modified to accept the inner surround. This seems like it could certainly keep the voice coil from rubbing. I don't know how much centering force it could provide though...
Vikash said:
I just had a look through my collection of various drivers (about 20 odd) and all of the midrange drivers had a curved cone profile, while the larger "bass" drivers had straight cone profiles, mabie the curved profile helps limit/mask cone breakup? where the straight profile on the bass drivers is more important for strength?
That sounds right to me, and yes, I think you've nailed the underlying issues on the head.
That drawing isn't quite what I meant when I described the torodial cone shape--where the cone meets the surround, the cone will be at a roughly 45 degree angle; where it meets the voice coil former, it's at a 90 degree angle (so it's parallel (collinear?) with the VC former) and the shape is a circle. Here, lemme draw a picture:
(Hmm. In retrospect, maybe the angle *is* greater than 45 degrees. This sure does look like a very "deep" cone to look into...)
That drawing isn't quite what I meant when I described the torodial cone shape--where the cone meets the surround, the cone will be at a roughly 45 degree angle; where it meets the voice coil former, it's at a 90 degree angle (so it's parallel (collinear?) with the VC former) and the shape is a circle. Here, lemme draw a picture:
(Hmm. In retrospect, maybe the angle *is* greater than 45 degrees. This sure does look like a very "deep" cone to look into...)
Attachments
I'd be interested in how to create a proper phase plug is order to maybe horn load the kind of cone design posted by Nappy lady.
I've been working lately on a simple phase plug I could add a to a full range driver with a whizzer cone, to avoid wave collision and to somehow horn load the whizzer.
I could not find a resource where I could benefit from someone else's diy experience. So I started making my own r&d..
Started from the idea that if I had a side cutout of the whizzer cone, and if theoritically every point on that wizzer radiates in a 2d Pi field, I would simply have to opose symetricaly the exponantial function of the whizzer cone, to reflect wave in a proper way.
Anyone can please tell me that I'm wrong, It cant be that simple.
And anyone here has ever seen complete resources on phase plugs? patents? anything...
I've been working lately on a simple phase plug I could add a to a full range driver with a whizzer cone, to avoid wave collision and to somehow horn load the whizzer.
I could not find a resource where I could benefit from someone else's diy experience. So I started making my own r&d..
Started from the idea that if I had a side cutout of the whizzer cone, and if theoritically every point on that wizzer radiates in a 2d Pi field, I would simply have to opose symetricaly the exponantial function of the whizzer cone, to reflect wave in a proper way.
Anyone can please tell me that I'm wrong, It cant be that simple.
And anyone here has ever seen complete resources on phase plugs? patents? anything...
Just to muddy the waters a bit...
Nappy Lady's cone concept looks pretty good in terms of stiffness and breakup control (although actual results will depend heavily on materials) but it also looks like its treble response would suffer from phase cancellation.
At high frequencies, the coherent wave will be launching from the central part of the cone around the VC (everything else will be in breakup). The trouble arises from the fact that, because of its extreme angle, this inner cone surface has a high rate of path-lenth change over its effective radiating area. An inch or two is quite a significant path-length difference at high frequencies--two inches is a full wavelenth at 6.6kHz.
Remember that cone designers have to balance geometric stiffness, which would logically call for a profile like Nappy Lady's cone, with phase-cancellation concerns, which ideally call for a flat cone.
Of course, you can bend the rules a bit by tweaking mechanical wave speed with materials engineering, but that's another whole can of worms...
Bill
Nappy Lady's cone concept looks pretty good in terms of stiffness and breakup control (although actual results will depend heavily on materials) but it also looks like its treble response would suffer from phase cancellation.
At high frequencies, the coherent wave will be launching from the central part of the cone around the VC (everything else will be in breakup). The trouble arises from the fact that, because of its extreme angle, this inner cone surface has a high rate of path-lenth change over its effective radiating area. An inch or two is quite a significant path-length difference at high frequencies--two inches is a full wavelenth at 6.6kHz.
Remember that cone designers have to balance geometric stiffness, which would logically call for a profile like Nappy Lady's cone, with phase-cancellation concerns, which ideally call for a flat cone.
Of course, you can bend the rules a bit by tweaking mechanical wave speed with materials engineering, but that's another whole can of worms...
Bill
JBL Differential Drive....attractive for woofers
Differential Drive® Design
(U.S. patent no. 5,748,760 and other patents pending)
http://jbl.com/car/products/gti/GTi_tech_info_3.asp
http://manuals.harman.com/JBL/CAR/Owner's%20Manual/W101215%20OM%20FINAL%20(revised%2092000).pdf
From JBL Literature.
Historically, the pro sound industry has gone to larger diameter voice coils and more massive magnet structures to increase loudspeaker output capability within desired degrees of linearity. AT JBL, we are addressing additional challenges in areas of driver fit and function as we strive to make them smaller and lighter with no compromise in performance. JBL's Differential Drive technology is a step in this direction.
Differential Drive technology uses a pair of separate, reverse-wound voice coils on a single voice-coil former and cone. The two coils operate in opposing magnetic fields to accomplish performance similar to a conventional design but in a considerably smaller and lighter structure. Although the dual-coil approach is not new, JBL has improved on the design through the application of two new features.
I will first explain how Differential Drive works by comparing the new design with the standard approach. For the sake of making an apples-to-apples comparison, assume that both designs have the same total flux density in the gap and that the amount of copper and moving mass is the same in each design. In the traditional JBL structure, magnetic flux B crosses a gap in which a coil of copper has a total electrical resistance of R[e]. These quantities establish the value of the electromechanical coupling coefficient, (Bl)2/Re.
In Differential Drive topology, there are two magnetic gaps with opposing flux. The two voice coils are connected in reverse polarity so that the mechanical forces they produce will add. For the moving mass to remain the same, the two voice coils must have the same height and half the thickness as in the standard design. The value of B will remain the same.
When these changes are made, the total length of the voice coil wire will be doubled, and the resistance per-unit length of wire will be halved. The total resistance of both voice coils in the series will then be four times what it was with the standard approach. Since the length has doubled, the quantity (Bl)2 will now be four times what it was in the standard approach. This results in a coupling coefficient value of 4(Bl)2/Re. Canceling out the fours yields the previous value of (Bl)2/Re.
In other words, we have exactly the same coupling coefficient as before, but we have picked up several important advantages relative to the traditional design. The new voice coil assembly now has twice the surface area of the traditional one, and this means that it will have twice the heat dissipation of the traditional single coil, which translates directly into twice (+/-3 dB) the power input capability for a given operating temperature and observed amount of power compression. The new dual voice coil structure will have less effective inductance than the standard one because the reverse-wound coils will have negative mutual inductance between them. This translates into a flatter impedance curve at higher frequencies, producing more output for a given drive voltage. Finally, the new design is generally more compact, and when used with neodymium magnets, it requires less steel to complete the magnetic circuit assembly. Consequently, it is much lighter. The design, however, is not limited to new magnet materials and can be used with standard ferrite magnets with benefits one and two above still applicable.
The two important design features referred to earlier deal with overall system linearity. First, the two voice coils are not placed at the axial center points of their respective magnetic gaps; they are symmetrically displaced axially outward so that the overall net distribution of flux density in the combined gap space is most linear. This ensures maximum system displacement linearity for the moving system.
In high-level operation, low-frequency high-displacement signals often tend to drive the voice coils out of their linear operating region. While traditional designs rely on progressive suspension designs to constrain this motion mechanically, the Differential Drive transducers make additional use of a shorted electromagnetic braking coil. This is shown in Figure 1. The coil is located mid-way between the two driving coils, and at normal excursions, it is virtually inert. On high excursions, the shorted coil enters each magnetic field alternately, and current is induced into the coil. That induced current, by Lenz's law, acts to oppose the motion that causes it. The result is additional braking on cone motion, resulting in lower distortion.
Differential Drive® Design
(U.S. patent no. 5,748,760 and other patents pending)
http://jbl.com/car/products/gti/GTi_tech_info_3.asp
http://manuals.harman.com/JBL/CAR/Owner's%20Manual/W101215%20OM%20FINAL%20(revised%2092000).pdf
From JBL Literature.
Historically, the pro sound industry has gone to larger diameter voice coils and more massive magnet structures to increase loudspeaker output capability within desired degrees of linearity. AT JBL, we are addressing additional challenges in areas of driver fit and function as we strive to make them smaller and lighter with no compromise in performance. JBL's Differential Drive technology is a step in this direction.
Differential Drive technology uses a pair of separate, reverse-wound voice coils on a single voice-coil former and cone. The two coils operate in opposing magnetic fields to accomplish performance similar to a conventional design but in a considerably smaller and lighter structure. Although the dual-coil approach is not new, JBL has improved on the design through the application of two new features.
I will first explain how Differential Drive works by comparing the new design with the standard approach. For the sake of making an apples-to-apples comparison, assume that both designs have the same total flux density in the gap and that the amount of copper and moving mass is the same in each design. In the traditional JBL structure, magnetic flux B crosses a gap in which a coil of copper has a total electrical resistance of R[e]. These quantities establish the value of the electromechanical coupling coefficient, (Bl)2/Re.
In Differential Drive topology, there are two magnetic gaps with opposing flux. The two voice coils are connected in reverse polarity so that the mechanical forces they produce will add. For the moving mass to remain the same, the two voice coils must have the same height and half the thickness as in the standard design. The value of B will remain the same.
When these changes are made, the total length of the voice coil wire will be doubled, and the resistance per-unit length of wire will be halved. The total resistance of both voice coils in the series will then be four times what it was with the standard approach. Since the length has doubled, the quantity (Bl)2 will now be four times what it was in the standard approach. This results in a coupling coefficient value of 4(Bl)2/Re. Canceling out the fours yields the previous value of (Bl)2/Re.
In other words, we have exactly the same coupling coefficient as before, but we have picked up several important advantages relative to the traditional design. The new voice coil assembly now has twice the surface area of the traditional one, and this means that it will have twice the heat dissipation of the traditional single coil, which translates directly into twice (+/-3 dB) the power input capability for a given operating temperature and observed amount of power compression. The new dual voice coil structure will have less effective inductance than the standard one because the reverse-wound coils will have negative mutual inductance between them. This translates into a flatter impedance curve at higher frequencies, producing more output for a given drive voltage. Finally, the new design is generally more compact, and when used with neodymium magnets, it requires less steel to complete the magnetic circuit assembly. Consequently, it is much lighter. The design, however, is not limited to new magnet materials and can be used with standard ferrite magnets with benefits one and two above still applicable.
The two important design features referred to earlier deal with overall system linearity. First, the two voice coils are not placed at the axial center points of their respective magnetic gaps; they are symmetrically displaced axially outward so that the overall net distribution of flux density in the combined gap space is most linear. This ensures maximum system displacement linearity for the moving system.
In high-level operation, low-frequency high-displacement signals often tend to drive the voice coils out of their linear operating region. While traditional designs rely on progressive suspension designs to constrain this motion mechanically, the Differential Drive transducers make additional use of a shorted electromagnetic braking coil. This is shown in Figure 1. The coil is located mid-way between the two driving coils, and at normal excursions, it is virtually inert. On high excursions, the shorted coil enters each magnetic field alternately, and current is induced into the coil. That induced current, by Lenz's law, acts to oppose the motion that causes it. The result is additional braking on cone motion, resulting in lower distortion.
JBL Differential Drive Motor Pciture
JBL Differential Deive Motor.
The bottom coil is wound in the reverse direction of the top, so one coil is "pushing-up" while the other is "pulling up" . This clever arrangement of having two identical coils wrapped in opposite directions also nulifies most coil inductance. Differential drive only became feasible for woofers with NdFeB magnets which can provide strong fields in a narrow cylinder. A ceramic magnet of similar strength would have made the distance between the two pole very long.
The "neo radial magnetic" motor from Aura Sounds seems to have superior properties for long linear underhung woofers where you want both the ultimate in detail plus long excursion. A heavy Faraday ring is required with the Aura radial motor for obtaining the low Le necessary for wide bandwidth with flat frequency response.
JBL Differential Deive Motor.
The bottom coil is wound in the reverse direction of the top, so one coil is "pushing-up" while the other is "pulling up" . This clever arrangement of having two identical coils wrapped in opposite directions also nulifies most coil inductance. Differential drive only became feasible for woofers with NdFeB magnets which can provide strong fields in a narrow cylinder. A ceramic magnet of similar strength would have made the distance between the two pole very long.
The "neo radial magnetic" motor from Aura Sounds seems to have superior properties for long linear underhung woofers where you want both the ultimate in detail plus long excursion. A heavy Faraday ring is required with the Aura radial motor for obtaining the low Le necessary for wide bandwidth with flat frequency response.
Attachments
This is the design that JBL uses in their EON G2 series speakers. The system seems to work very well. Of all of the high power JBL speakers I service, I have the fewest problems with the JBL EON G2 LF driver. usually the spider will separate from the basket before the VC bottoms out or jumps out of the gap. I have only seen one G2 with a damages VC, and that was caused by excessive abuse by the customer.
Cheers,
Zach
Cheers,
Zach
Aura Neo Radial Magnetic Motor
This image is the Aura Neo Radial Magnetic motor. The center cylinder is the steel pole piece that runs down and is attached to the outer cylinder-ring for the complete magnetic circuit. The next concencentric ring is a copper Faraday ring to reduce inductance. The ten arc segments outside the gap are the NdFeB magnets. NdFeB is brittle and difficult to machine. It would be difficult to manufacture an NdFeB ring to tight tolerances, and hence multiple arc segments are used. This topology offers a very uniform magnetic field over a long underhung coil length. The physical former and coil can be shorter than the JBL Differential Drive, and this can save mass. Also, the radial field and having only one magnet to metal interface gives good efficiency and field uniformity.
Seas Millenium tweeter uses a similar radial design with 6 rectangular bar magnets.
This image is the Aura Neo Radial Magnetic motor. The center cylinder is the steel pole piece that runs down and is attached to the outer cylinder-ring for the complete magnetic circuit. The next concencentric ring is a copper Faraday ring to reduce inductance. The ten arc segments outside the gap are the NdFeB magnets. NdFeB is brittle and difficult to machine. It would be difficult to manufacture an NdFeB ring to tight tolerances, and hence multiple arc segments are used. This topology offers a very uniform magnetic field over a long underhung coil length. The physical former and coil can be shorter than the JBL Differential Drive, and this can save mass. Also, the radial field and having only one magnet to metal interface gives good efficiency and field uniformity.
Seas Millenium tweeter uses a similar radial design with 6 rectangular bar magnets.
Attachments
Hi all,
Does anybody have an updated version of this spreadsheet (posted here many years ago)?
http://www.diyaudio.com/forums/multi-way/17936-diy-dynamic-driver-2.html#post212875
The person who posted does not accept PM's... The version on the link seems to have a glitch.
Thanks for any help.
Does anybody have an updated version of this spreadsheet (posted here many years ago)?
http://www.diyaudio.com/forums/multi-way/17936-diy-dynamic-driver-2.html#post212875
The person who posted does not accept PM's... The version on the link seems to have a glitch.
Thanks for any help.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.