Geddes on Acoustic Lever

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


Can you give some insight into practical Acoustic Lever (AL)? It is meant to be an alternative to bass horn. Speak software that simulates such enclosure doesn't work with non-US Windows and even then it does not load databases and projects with AL so it's useless.

How should we calculate those three volumes, compliances and masses?

What is known driven diaphragm should be slightly smaller then driver and radiating slightly bigger. Maximum ratio 2:1. I'm most interested in rear AL so that we extend bandwidth rather than gain efficiency.
 

Attachments

  • Clipboard01.gif
    Clipboard01.gif
    33.2 KB · Views: 1,350
Dear Dr. Geddes,


Can you give some insight into practical Acoustic Lever (AL)? It is meant to be an alternative to bass horn. Speak software that simulates such enclosure doesn't work with non-US Windows and even then it does not load databases and projects with AL so it's useless.

How should we calculate those three volumes, compliances and masses?

What is known driven diaphragm should be slightly smaller then driver and radiating slightly bigger. Maximum ratio 2:1. I'm most interested in rear AL so that we extend bandwidth rather than gain efficiency.

Fundamentally, it is a transformer, so it should function similar to the bass horn.

I believe Dr. Geddes' research with regards to Acoustic Levers can be found in his book "Audio Transducers".

Here is a link to an article published by Dr. Geddes ("The Acoustic Lever Loudspeaker Enclosure").
http://www.aes.org/tmpFiles/elib/20091201/12118.pdf
 
Last edited:
Disabled Account
Joined 2008
From what I can see, the device above appears to operate in a similar fashion to how a contemporary automotive hydraulic brake system works. The trade-off is excursion at the input for increased force at the output. An interesting concept, perhaps its goal is to better couple the driver's cone to the air-filled universe. A sort of mechanical impedance matching.

I would wonder at the mechanical losses imposed by such a system? That schematic, which I think is supposed to be the equivalent circuit expressed in electrical terms, is missing some resistors!

Jim

edit: having just Googled the device name, I now realize that it is Mr. Geddes design. I will shut up then and let him explain it, should he chose to do so! :p
 
Last edited:
Earl come on ;)

Impatient aren't you. Mind if I go out for the evening? And "useless" - SPEAK works fine for me, and I don't "sell" it any more, so I don't support it. Finally, everyone else is correct, there is plenty of literature out there about it. Read that first and then ask your questions.

But extending the bandwidth won't work, that's not the levers strength. Improved efficiency is and trying to make it do anything else is going to be a dead end.

Nothing is going to beat a direct radiator for bandwidth. All designs are efficiency/bandwidth tradeoffs.

And it is just like a brake system and if you used fluid inside instead of air it would work even better!
 
Impatient aren't you. Mind if I go out for the evening? And "useless" - SPEAK works fine for me, and I don't "sell" it any more, so I don't support it. Finally, everyone else is correct, there is plenty of literature out there about it. Read that first and then ask your questions.

But extending the bandwidth won't work, that's not the levers strength. Improved efficiency is and trying to make it do anything else is going to be a dead end.

Nothing is going to beat a direct radiator for bandwidth. All designs are efficiency/bandwidth tradeoffs.

And it is just like a brake system and if you used fluid inside instead of air it would work even better!


Earl,

It really looks to me to be a mass/spring resonator coupled to the driver by another spring. The electrical analog makes it look more complex that I think it really is. Ignoring the air spring behind the driver, the double disk is a mass, the air in the double disk chamber is a spring, and the air in the chamber between the small disk and the driver the coupling spring.

It would not work (very well) with an incompressible fluid though. An incompressible fluid in the coupling chamber would rigidly couple the driver and small disk on the compression stroke, on the expansion stroke there would be the possibility of cavitation or rigid coupling, depending on how fast the driver pulled back. Additionally, the small and large disks must move in unison but since they have different areas the swept volume of each is different. This would attempt to compress or expand the incompressible fluid and that indicates that the disks won't move. They could move on the expansion cycle but again, that would lead to cavitation, or sucking a vacuum.

I understand why you call it an acoustic level, but it looks to me like a PR with an additional compliance element coupling the PR to ground.
 
Earl,

It really looks to me to be a mass/spring resonator coupled to the driver by another spring. The electrical analog makes it look more complex that I think it really is. Ignoring the air spring behind the driver, the double disk is a mass, the air in the double disk chamber is a spring, and the air in the chamber between the small disk and the driver the coupling spring.

It would not work (very well) with an incompressible fluid though. An incompressible fluid in the coupling chamber would rigidly couple the driver and small disk on the compression stroke, on the expansion stroke there would be the possibility of cavitation or rigid coupling, depending on how fast the driver pulled back. Additionally, the small and large disks must move in unison but since they have different areas the swept volume of each is different. This would attempt to compress or expand the incompressible fluid and that indicates that the disks won't move. They could move on the expansion cycle but again, that would lead to cavitation, or sucking a vacuum.

I understand why you call it an acoustic level, but it looks to me like a PR with an additional compliance element coupling the PR to ground.

Hi John

I agree about the fluid coupling, I didn't think that through, yes, without a compliant coupling there is no chance for a resonance and the associated gain. The coupling does have to be compressible.

But I don't agree with your assesment about it being simply a passive radiator. When you simplify the circuit, it comes out to being a simple transformer with a turns ratio equal to the quotient of the two areas of the level. Since transducers tend to be poor impedance matchs to air - especially at low frequencies - there is gain to be had from this transformer. Many people who casually look at this circuit conclude that there is not much to be gained, but we dispelled that rumor. We did a fairly complete study of the level in an actual sound system in a car and found that the claims made about its incresed efficiency were completely correct. This was published as an SAE paper not an AES paper and so it was not widely seen.

Of note is that a horn, in the end, also looks like a transformer when simplified. But the horn has the turns ratio equal to the square root of the area ratios, while the lever is simply the ratio. This makes, in the most simplified cases, the lever and the horn comparable in efect, but the lever is more efficient because of the linear versus the square root ratio. In terms of gain for a given volume of enclosure the lever is much more effective than a horn. Why don;t we see them? No one makes a lever and they would not be easy to build (by a DIYer that is).

The downside of the lever, as I said above, is bandwidth. You simply cannot get gain AND bandwidth at the same time. The horn has a bandwidth advantage, but the lever has an efficiency advantage.
 
Would it be possible to write an akabak script to simulate one?

Of course, but that's all been done. It's been studied and quantified completely. There is nothing left but making them. I would gladly design a lever system if someone would fabricate the lever component itself and agree to sell it at a fair price (TBD). But with tooling etc. that's a big load to carry for a risky business.
 
Ex-Moderator R.I.P.
Joined 2005
Maybe DIY does have the advantage

The acoustic gain is interesting
Bandwidth concerns me less, quite the opposite

Somehow it does look like one big massive passive radiator
But quite different because its in front of the driver
That ofcourse is what limit bandwidth
4th order bndpass with pretty advanced PR instead of a port ?

But I would like to know, at which frequency does this gain happen
At the same frequency where a passive would be "active"...my that was funny

But velocity/compression means delay time, but a port have that too
 

Attachments

  • Geddes acoustic lever.JPG
    Geddes acoustic lever.JPG
    29.5 KB · Views: 772
Last edited:
Maybe DIY does have the advantage

The acoustic gain is interesting
Bandwidth concerns me less, quite the opposite

Somehow it does look like one big massive passive radiator

It is and it isn't. In a lever, the volume velocity of the output is greater than the volume velocity of the input. With a passive radiator it is one to one.

4th order bndpass with pretty advanced PR instead of a port ?

But I would like to know, at which frequency does this gain happen
At the same frequency where a passive would be "active"...my that was funny

But velocity/compression means delay time, but a port have that too

Who cares about "delay time" in a sub. Yes, the system is "like" a PR bandpass, and yes the gain comes at the bandpass passband.

You can also port the back (this is covered in the patent) which adds a little, but makes a LF dipole.
 
Ex-Moderator R.I.P.
Joined 2005
Yes, the system is "like" a PR bandpass, and yes the gain comes at the bandpass passband.

Hmm, this means the passive system more or less replaces what would have been upper roll off port in 4.order BP
I understand that whats make it more like a horn without horn
No real low bass extension, only upper bass gain
But but but, might still have a very clean and natural upper roll off, and less cone ressonances
Ehh, might there be some kind of counter reactive influense from the passive system onto active driver
 
Last edited:
Whats the difference between this lever and JBL's driver that divided
a large cone concentrically with a compliant section in the middle?
I forget now what that one was called...

Aside from the central portion of the driver being able to radiate
directly, I'm asking about the function of the outer cone as a
compliantly coupled lever.
 
You might be thinking of the Altec 419-8B and 420A "Biflex" drivers.
Consider the compliant section as a crossover and model it as such. It's quite different than an acoustic lever as defined by Earl. For example, the resonance between the main driver and the acoustic lever diaphragm is missing.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.