Mechanical isolation of driver from cabinet

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi there,

while experimenting i fell over a method useful for mechanical
driver - cabinet isolation. This is not necessarily restricted to FR drivers.

The driver is not directly attached to the baffle anymore, there is
only some soft material as a gasket between the drivers basket
and the baffle.

The driver is magnet mounted using a spring (e.g. damped metal beam),
which is connected at one or both ends rigidly (optional with some damping)
attached to the cabinet. (even a spiral spring with a damper would be
possible or an air spring).

The resonant frequency fd of the driver and the suspension is adjusted
well below the usable range of the driver/cabinet combination.
(e.g. one octave below fs for a closed cabinet).

Above fd the excursion of the driver basket suspended in
its "Pendulum" is "mass inhibited". At fs things looks pretty much like
the driver was mounted conventionally ...

The absence of a rigid connection between driver chassis and cabinet
reduces vibration of the cabinet and thereby structure born sound
radiated unwanted from the cabinet walls, which typically
intoxicates the middle to upper bass region ...

I do not know by now if this is "new" entirely, since magnet mounting is
known practice. But in many cases structures supporting the magnet
are so stiff (and underdamped), that fd is within the usable range of
the speaker, which is counterproductive.

Defined "fd" and "Qd" using a well designed mass spring configuration
and a damping mechanism would IMO improve matters.

Opinions ?


Kind Regards
 
Administrator
Joined 2004
Paid Member
I'm completely out of my element here, but I had always understood it was desirable to tightly couple the driver to the cabinet in order to minimize the energy stored in the driver components.

A lot of current full range designs feature heavily braced cabinets and drivers that are tightly coupled to the cabinet via notched braces in tight contact with the magnet assembly. (Like the HCC I recently built)

How do you damp vibration in the basket, flange and other areas of the driver? Based on your comments I assume the driver support/suspension compliance is tuned to some frequency way below the driver/box resonance. Does the compliance of the suspension damp the driver structure in its operating range effectively?

I'm just curious - I don't have sufficient knowledge or experience to know whether this is advisable or not, and hence have no sense of the merit of your idea. :D
 
Last edited:
Disabled Account
Joined 2007
I did some experimentation on this idea HERE and I think the results were good.
I'll need to be a bit more thorough in my testing to determine if it is worth the effort.

Sigfried Linkwitz recommends isolating the driver frame from the baffle while rigidly mounting the driver magnet.
 
frugal-phile™
Joined 2001
Paid Member
The FujiTEN Eclipse uses a sysyem where the driver is connected to a mass-loaded base (blue & black) and the shell is floated on that with no rigid contact. Something like the attached sketch.

http://www.eclipse-td.net/

dave
 

Attachments

  • eclipse-mount.gif
    eclipse-mount.gif
    12 KB · Views: 1,174
I've also thought about this for quite a long time. Manufacturers are trying to decouple the drivers since forever. One of my first real hifi loudspeakers was a Van Medevoort CD 2.5 (around 1985), it had the drivers glued in place using silicon without any screws. Just recenty i saw a picture of some monitor that had a midtone compartment embedded in foam in a hollow space inside the loudspeaker box.
My own experience is that everything must be as rigid as possible, i find that the speaker sounds less muddy that way. Ofcource it's possible to decouple and retain detailed sound but at a certain point it's just no longer economical or plain silly. You could think of increasing the mass of the driver by adding weight to the magnet but this would put additional stress on the frame, to avoid this you can make a heavy metal ring to attach to the rim of the driver but having that custom made costs a fortune. Another way to dampen vibrations is to use a damper, like the ones on an archery bow, i think a car manufacturer used isomething similar to dampen the vibration of a car engine.
The thing is if you decouple the drive from the cabinet then the cabinet becomes lighter so you have to make the cabinet heavier otherwise the soundpressure will just push away the cabinet. Cost, material and total weight will spiral out of control.
Olympics+Day+6+Archery+Iu6ks78yG9ll.jpg
 
I was an engineer at KEF in the 80s when we were having good success decoupling drivers from the cabinets.

The issue came up with the original KEF 105 (not the 105mKII) that was found to have a fairly nasty cabinet resonance in the middle hundreds. Cabinet stiffening moved the resonance up but never got rid of it so rubber mounting the driver was tried. We found that we could get a strong reduction in the amount of "drive" to the cabinet resonances via decoupling. It was easy to measure and you could hear the difference, especially on test impulses or impulsive sound.

The trick is to get the decoupling resonance low enough. Soft glues, silicon, etc. don't really do it. The 105 II used rubber motor mounts and soft foam tape. We even went to 3 mounting bolts to get the resonance a little lower. You had to be careful that mounting bolts didn't touch the cabinet and "short circuit" the effect. If I remember right the accelerometer measurements showed about a 20dB reduction in resonance levels with decoupling. Even with perfect decoupling their will be an acoustical drive to the cabinet, only the direct vibration coupling is reduced. The later 105 MkII used two woofers in opposition (plus decoupling) for even greater vibration reduction. I think people have even played with dummy motor structures with cone mass substitues that vibrate out of phase and cancel vibration (read about antivibration countershafts in 4 cylinder engines).

Audiophiles will tell you that dirvers must be rigidly mounted because this is a well ingrained part of accepted dogma. Our tests showed that this wasn't the case.

David S
 
The FujiTEN Eclipse uses a sysyem where the driver is connected to a mass-loaded base (blue & black) and the shell is floated on that with no rigid contact. Something like the attached sketch.

speaker | ECLIPSE TD series speaker

dave

designed/patented by Yoshii Hiroyuki
decoupling seems to be technically much easier to achieve with upward firing speakers
TIMEDOMAIN �F�S‚̃I�[ƒfƒBƒI �`Ž©‘R‚ȉ¹‚̃Xƒs�[ƒJ�[
Speaker apparatus equipped with means for producing complicated waveform of low frequency with higher improved fidelity - Patent 6796401
 

Attachments

  • 008_yoshii9_td.jpg
    008_yoshii9_td.jpg
    6.5 KB · Views: 983
Naim and Mordaunt - Short (440?) both gave decoupled cabinets/speakers a lash. IIRC, the MS has a rigid skeleton to which the drive units were affixed, and the cabinet floated around this, sealed/supported with rubber gaskets. Roksan also decoupled tweeters. No doubt Paul Voigt/ Emil Berliner/ Gilbert Briggs/ your own name here suggested a similar system.
 
No doubt Paul Voigt/ Emil Berliner/ Gilbert Briggs/ your own name here suggested a similar system.

oh yes, because decoupling is good, it's just good engineering

on the other hand it raises some issues of DFM - additional costs and manufacturing complications

fortunately in DIY where we are free from DFM considerations decoupling is quite easy to achieve, and in case of upward firing speaker it becomes easy as pie
 
Last edited:
...
If I remember right the accelerometer measurements showed about a 20dB reduction in resonance levels with decoupling. Even with perfect decoupling their will be an acoustical drive to the cabinet, only the direct vibration coupling is reduced.
...

David S

First of all gentlemen i thank you very much for the detailed
input and the examples of constructions which had already
been realized.

@ David: If the reduction of cabinet vibration was that good,
one could ask how much effort would have been necessary in
improving cabinet damping to achieve the same effect without
decoupling ...

I am convinced, that it may be an effective and economically
reasonable method to implement high quality cabinets.

If you have got a set of components that work well in effectivity,
durability, manufacturing effort, ...

you can easily customize them to meet the parameters needed
for different speaker/cabinet situations.

Since in audiophilia there have been invented
(and sold) very much gadgets to compensate minor effects, i
wonder why this has not found a wider application ...

Cabinet vibration - to me - is NOT a minor effect. It is one of
the nastiest hurdles for real musical enjoyment.

A speaker meant to be "high end" which in an appropriate listening
environment can be convicted to exhibit disturbing cabinet
resonances is IMO disqualified.

edit: For shure, direct excitation from the drivers motor is
not the only source of unwanted cabinet vibration.
But in many cases that mechanism is underestimated IMO.


Kind Regards
 
Last edited:
...
Does the compliance of the suspension damp the driver structure in its operating range effectively?
...

For sure the mass/spring system has to be damped sufficiently.

What inhibits movement of the drivers chassis relative to the
cabinet above the resonance frequency of the
(whole drivers - not the cone's ) "pendulum like" suspension ,
is the mass component of the driver and the suspension components
moving with the driver: Simply inertia will reduce movement with
increasing frequency.

Decoupling rising with frequency is important because cabinet
vibrations are radiated more effective with increasing frequency
and our auditory system is getting more sensitive with
inceasing frequency, thereby multiplying the disturbing effect.

The proposed suspension is in fact no difference to
every speaker available ...

Just in conventional speakers there are no serious constructive
measures to bring that resonance(s) under control.

To control a resonance means to me, to put its frequency and its
Q to where i want it to be. To ignore something does (mostly) not
lead to having it under control.

I am rather old fashioned, construction means to me bringing a
systems behaviour under control to meet some specifications,
e.g. a specified maximum fraction of the total acoustic power
radiated by the cabinet walls.

Oops, there is no spec ? So there's no Problem ?

Kind Regards
 
Last edited:
I have built subwoofers with moving mass compensation,
and i must admit, that an enclosure that stays calm is a very
fascinating thing. In case of a subwoofer one of the main advantages
is IMO, that there is no movement of the cabinet as a whole body,
so the bottom of the room (or any wall the sub is mounted to)
is not excited directly. You can use it as a a side
table if you like and put a glass on it without problems ...

But vibration of cabinet walls itself is not such a big problem with
subwoofers since for low frequencies you can win the fight with pure
stiffness of the cabinet.

Once the lowest resonance is in the mid hundrets in a mono sub
there will be no coloration from vibrating cabinet walls.

Although i had the impression that a mass compensated configuration
plays very "tight" and "clean".

Moving mass compensation is for sure in many cases an alternative for
decoupling and may often be the more cost effective one ...

Kind Regards
 
I went the same way one a couple of my speaker projects. My Hawthorne Audio Duets use a polystyrene baffle with the drivers held rigidly by the
magnets.

baffle_mods4.jpg


Another (experimental) project used a single full-range drive held in a skeleton but surrounded by a baffle of polystyrene.

Hi-fi DIY - Building full-range open baffle loudspeakers.

An externally hosted image should be here but it was not working when we last tested it.
 
First of all gentlemen i thank you very much for the detailed
input and the examples of constructions which had already
been realized.

@ David: If the reduction of cabinet vibration was that good,
one could ask how much effort would have been necessary in
improving cabinet damping to achieve the same effect without
decoupling ...

I am convinced, that it may be an effective and economically
reasonable method to implement high quality cabinets.



Kind Regards

Cabinet damping is an important factor. In my experience most companies deal with cabinet mass and stiffening but few actually trouble with damping. While at Snell we had good luck with 2 layer cabinets and Noise Killer Yellow in between. Definitely sounded deader with the 'knuckle rap test". Just stiffening cabinets moves resonances up in frequency but may or may not improve matters sonically.

Decoupling is actually pretty cheap to do, once you have worked out a viable scheme, and probably gives the most effective reduction of cabinet energy over any other technique.

Correction for my other post: the double opposed woofers were in the 104.2. This really knocked down the vibration drive to the cabinets, and makes more sense than a vibration oppositional motor/mass would make, since such a mass has most of the cost of a second woofer with no acoustical output.

If you want to ponder the theory of this remember that F = M x A, force equals mass times acceleration. The current through the voice coil makes the force but both the cone and the magnet structure accelerate in opposite directions. If the magnet is 10 times the cones mass then it will have 1/10th the acceleration, but acceleration none-the-less. If the magnet structure accelerates in opposition, then the cabinet comes along for the ride and, at frequencies of resonance, the cabinet becomes essentiall transparent. The most interesting test is with a woofer with a double cabinet: a conventional cabinet behind and a second similar cabinet in front of the woofer (both cabinets seal together at the baffle). Play voice through it and listen to what comes off the cabinet walls.

David
 
One reason manufacturers shy away from decoupled cabinets aside from custom seals, is handling by customers. If speaker driver and cabinet are only connected by a sealing gasket, how do you move the speakers? Both parts are liable to be weighty so you have to figure out a way to shift them without ripping the seals. Like picking up a deckchair wrong - suddenly you're in a world of pain and trouble.
 
Yeah, being predominantly degreed electrical and/or mechanical engineers, the pioneers of audio that developed PA, cinema sound apps decoupled drivers either via mass loading in the case of horns or isolation mounting for point source drivers, but once they moved away from large cinema sound OBs where drivers could be rigidly mounted to a massive floor and installs were no longer done by the sound system manufacturer, they quickly switched to enclosed cabs with lighter driver assemblies and used many bolts to baffle mount them with a gasket designed to both seal it for reflex loading and somewhat damp spurious driver related resonances, so once again, what was once old is yet new again/those who don't learn from history are doomed to repeat it, yada-yada. ;)

GM
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.