Beveridge vs traditional ESL?

[markusA]

What would you say are the pros and cons of the Beveridge way compared to the "regular way" of building ESL's?
The Beveridges are supposed to be perfect line sources for example.
Still I've heard people liking the sonic signature from the standard ELS better?
What's your take on the matter?
Building an acoustic lens and a cabinet isn't really all that hard compared to some of the builds you see on this board?
Being able to work with a flat panel instead of a curved one would make the panel construction easier.

[Mike Beck]

Pros:
Can be placed against the wall.
It has good dispersion.
Cons:
It has a cabinet and all the associated problems.
The enclosure is tuned for an LF boost (it's a Helmholtz resonator)-this might result in a less-defined bass compared to a dipole ESL. I have not heard a Beveridge ESL so I'll leave that for others to address; I have heard their owners love them.
Being a monopole, it will excite more room modes than a dipole.
Building this enclosure and acoustic lens would not be trivial..Beveridge's patent is well-written and with patience and diligence you could replicate it. It has a self-contained direct-drive valve amp that would be another project altogether.

[geraldfryjr]

IMHO,I would think that building a beveredge lens system would be quite labor intensive and would reqiure a lot of material to build the lens system and it is basicaly a multicelled horn.

I know very little about horn design, But multicell horns have a whole, another set of issues on their own.

However horn loading a small width panel is somthing that I have been contemplating for quite some time.

I have done some some experiements on trying to limit the backwave of the panel to get more bass.
I have found that even the soft fluffy side of fiberglass insulation does slightly cut back the high frequency detail when placed within close proximity of the back side of the diagphram.

It my belief that this pressure loading is limiting the movement of the diagphram slightly or it could be just a small amount reflections off of the isulation causing cancelations back through the front side of the diagphram.

This seems to only be an issue when in close proximiity.
But even being as far away as 2 to 4 inches still causing this effect does not support the refelction/cancelation theory.
It is something I wish to possibly investigate further.

But I don't think this would be as much of an issue with an open horn design as a multicell design would be.

I have never heard the beveridge system, But the design is very interesting.
But, By some of my experiments a can see were some of the comments are coming from.

I have always wanted to try a curved panel build and I have listened to a few Martin Logan ESL's.

But, I personaly have to agree with Roger Sanders views about the perspective of detail using a flat panel verses a curved panel.

This is not a biased opinion at all since he was the one whom had invented the curved panel.

It is just my personal pereference . jer

[LineArray]

Originally Posted by markusA

What would you say are the pros and cons of the Beveridge way compared to the "regular way" of building ESL's? ...

Long time ago i listened to that "ton shaped" Beveridge
model with integrated subwoofer and remember it as a
homogeneus sounding system with stable and
"believable" imageing.

I heard an ESL 63 in the same room at that time.

Cannot say one of them being "better", just different.

With that integrated subwoofer the low bass of the
Beveridge was superior, but i guess this is not the
question here.

The uppermost end will be somewhat compromised
with that lens approach, due to path differences in the
lens not being neglegible.

But also depending on the listener's age this may not
be an issue at all ... i did not recognize it as being compromised
auditively although my ears were somewhat younger those days.

[a.wayne]

Beveridge had responded to the Alloy Mylar before and for the record a unique SOTA speaker back then and maybe today...

2005

"I want to make it perfectly clear that all Beveridge transducers
were, and still are, made with fully conductive aluminized Mylar.
Your observations regarding the pros and cons of differing membrane
conductivities fail to consider the impact of differing "stator"
designs. Most stators use metal as a conductor, protected by an
an insulating coating. This design causes problems that our Epoxy
composite stators do not experience.

In some stators, like the Accoustat, the metal is in the form of an
insulated copper wire. Other designs use a copper layer on a circuit
board and are insulated either by another circuit board or by a layer
of insulating material which is applied by silk-screening or spraying.
The most common method uses perforated metal (usually steel), coated
with an insulating coating.

Since he decided to build an electrostatic driver in the early
1950's, my father chose to use Aluminized Mylar. In order to do so,
it was necessary for him to develop an ingenious design for what is
now commonly called a stator. He never referred to them as "stators".
He preferred to call them "electrodes" because he felt that the term
was more precise. He also never referred to the drivers he made as
"panels", but called them "transducers".

He developed an electrode with very unique properties. It was made
from a highly resistive material. It had a highly conductive coating on
the side away from the Mylar membrane. This electrode material also
has a very high dielectric constant. It is this resistive-capacitive
material which carries the full force of the field to the air gap
without threatening to burn the Mylar, even if it is fully conductive
aluminized Mylar. Specifically, the resistive component handles the
DC polarizing component, while the RC "mesh" handles the AC component.

One benefit of the capacitive property of this electrode material is
that it doesn't lose its punch when a heavy demand is temporarily
made upon the transducer. This is not true of resistive membranes.
With medium- to high-resistance membranes, a heavy demand will
severely reduce the charge stored in the membrane. Under the best of
circumstances, after a high voltage demand, a high-resistance
membrane will take several minutes to restore the membrane to its
full operating voltage.

If the circumstances are not so ideal, like when humidity is high,
this can take hours to restore. If the humidity is very high, and the
membrane resistance is also very high, the membrane will never
achieve its full charge. The loss of charge to the humid air exceeds
the charging capability of the membrane. This is not a problem with
fully conductive aluminized Mylar membranes.

Several popular electrostatic speakers on the market today, which use
high resistance membranes, can only maintain half of their normal
membrane charge on humid days. This causes the electrostatic driver
to lose half of its power. To make matters worse, these speakers
usually have the bass produced by standard dynamic cones; having the
electrostatic element varying its output against a constant bass
output throws off the balance between them.

Besides humidity, another problem encountered with conventional
membranes is that they collect dust. This is because they are given a
positive or negative charge which is maintained at a constant level
(if humidity doesn't interfere). The membrane then attracts any
particles with an opposing charge. Electrostatic air filters are very
effective. This attraction of dust is a common cause of arcing and
failure of convention systems.

A fully conductive membrane is driven like the electrode. Both the
voltage and the polarity change with every half cycle. The only dust
that collects in a Beveridge transducer is dust that has no charge
and does not settle onto the membrane because the membrane is moving.
Neutrally charged particles are shaken off and charged particles are
not attracted.

Another nasty problem solved with fully conductive membranes is that
of total voltage. In conventional electrostatic panels, the membrane
is passive. It takes on a charge, but that charge remains at a
constant, or is at least supposed to do so. The membrane is moved by
applying a high voltage charge to one or more electrodes near the
membrane. All electrostatic speakers commercially made today are
made with one electrode on either side of the membrane.

With a fully conductive membrane, it can, and is, driven along with
the electrodes. It is not passive, but is a working member of the
unit. The same total sound pressure level can be achieved with half
the total voltage. This ability to produce the same SPL at half the
voltage of conventional designs has many benefits.

As you say, fully conductive membranes can not be used with normal
stators. However, using a design such as ours is not an easy "fix".
Making the electrodes the way we do is difficult, requiring skill,
knowledge, and experience.

As a result, our transducers are neither easy or cheap to produce.
Despite this method's obvious benefits, no other electrostatic
speaker company has yet tried to accomplish it. It certainly does
not lend itself to the low cost and ease of production required by
the DIY market.

In an effort to reduce the cost and difficulty of production, my
father sought to produce a more conventional electrode. These were
his circuit-board electrodes. They were a total disaster. They would
have worked with more conventional membranes, but they did not work
well with the aluminized Mylar. This was especially true when used
in conjunction with step-up transformers and conventional amplifiers.
Although the circuit-board electrodes would have worked with more
conventional membranes, they would have gained all of the problems
of conventional transducers.

As far as low frequency capability is concerned, in the model 2SW,
the single electrostatic transducer plays from 100 cycles on up.
There is no need for a "woofer panel". Of the 30 plus pair of model 2
SW's sold by our Singapore dealer, all were made prior to 1980. Most
or all of them are working in a very humid environment, but not one
pair of them has yet had a reported transducer failure.

It is the combination of my father's original electrode, working with
highly conductive aluminized Mylar, that produced a truly superior
electrostatic speaker. This is borne out by the fact that there are
presently about 175 pair of Beveridge speakers still playing in at
least 17 countries around the world. The last ones made are at least
20 years old, with the largest majority of them being between 25 and
30 years old and using our unique electrodes and aluminized Mylar.
After more than a quarter century, less than 15 percent of the speakers
which use our Epoxy composite electrodes and fully conductive
aluminized Mylar have needed work on their transducers.

I have decided to continue to produce them, with their aluminized
Mylar membranes, in spite of the difficulty and cost. As Harry Pearson put
it so prophetically, "They are not for everyone".

After 30 years, I would add that, for some, they are perfect.

I hope that this will clear the air about aluminized Mylar membranes."

Rick Beveridge

[DaveG]

I would say the trick to the lense would be to design it to NOT be a horn. Just a guide.

[CharlieM]

I've not heard the Beveridge speakers either but I've read only great reviews about them. I figure it's strengths would be a very wide sweet spot and a lack of the low frequency dipole cancellation that can make open ESL's sound thin if they're not EQ'd properly to compensate.

On the flip side, no wide-dispersion ESL (lensed or curved) could ever hope to match the magical, pristine imaging you get from a line-source flat-panel ESL. A flat panel has superior imaging because it's ultra-directional. The downside, of course, is its miniscule sweet spot-- move just a foot or two outside the focal sweet spot and the treble energy falls off a cliff.

I'm with jer and Roger Sanders, though. I like the flat panel sound, even if it puts my head in a vise

[DaveG]

My Acoustat 1+1's sound blurry to me after listening to the Bev Model 5's for awhile. I listen in a dedicated room, with couch and speakers all positioned to better than an 1/8 inch and all additional excess speakers and furniture also balanced, and a wall of drapes behind the speakers to control the backwave off the Acoustat's.
I listen with my head in the vise all the time, and I enjoy listening that way.
But the Beveridge Model 5's, when setup like the Acoustat's, angled, so I can see my reflection in the mylar, have an even tighter apparent location of sounds.

2 things about the Beveridge's. The highs remain quite consistant when walking around the room, unlike the Acoustat's where the highs drop off when moving off axis. But there is still an ultimate sweet spot as far as the best imaging, just a small as the Acoustat's.

"On the flip side, no wide-dispersion ESL (lensed or curved) could ever hope to match the magical, pristine imaging you get from a line-source flat-panel ESL. A flat panel has superior imaging because it's ultra-directional."

in response, it is rather like the center beam of on axis information is still there, but the additional information from the curved portions of the lense supports the main information beam, giving it more stability and shape, rather than distract from it.

Also, the rear wave is not bouncing all around causing whatever damage it can do.

Result is they seem even more focused - almost like a haze has been removed around each sound...from the ability to locate an individual sound.

One more thing: I enjoy listening to each speaker-they each have plusses and minuses

[Calvin]

Hi,

charmingly Rick doesn´t mention in all his praise of the electrodes, that the design gives up on on of the most desirable properties of ESLs - the linearity of a constant charge drive system.
The high-K additives to the epoxy-electrodes function to increase the dielectric constant value -i.e the capacitance of the insulating layer- to reduce the losses in this part of the AC-voltage divider which is comprised of the insulative layer and the air gap. But all high-K materials have certain undesired nonlinear properties.
It also doesn´t mention that only field strength counts when it comes to a flashover. Any panel can be made to arc, no exception! The dissipative volume resistivity surely helps in preventing flashovers in that it works to reduces Charge-/voltage-buildup on the electrode surface. But the same mechanism works with any other stator electrode coated with a dissipative layer too. The difference here only beeing the lower voltage levels due to the drive system, which reduces the chance of flashover. I rather assume the epoxy electrode needs to be quite thick simply because of mechanical stability reasons. This of course creates the demand for high Epsilon-r values and reduced volume resistivity in first place.
But in case of a flashover a aluminized membrane might burn off in a quite spectacular way, because of too low associated impedances, that allow the buildup of ´hot´ flashovers. An effect that might not occur with high resistive membrane coatings and self-healing PET-diaphragms.
The recovery time Rick talks of can also be made small with any insulator material using an dissipative additive. Its the same reason James Strickland suggested PVC as insulation with his wire-stators (comparably high epsilon-r values and low resistance values for an insulating plastic material, similar to Nylon and polyurethane). There´s no reason for wire or sheet metal stators not to be insulated with a coating in the dissipative range and not to achieve similar behaviour as the epoxy stators under similar voltage conditions.

The charme of the constant chage ESL is that it takes the membrane out of the active system (think of electret ESLs). It behaves like a passive probe charge acted upon by the electrical field spanned between the two stators. Even though everbody knows about the problems regarding longtime stability of coatings and especially the contacting points. Making the membrane part of the ´active system´ puts the membrane coating and the contacting points under considerable stress (high voltage and current flow) and can lead to early failure. Afaik this was the main reason that forced the Final company out of business. Their ´patented´ inverter-principle worked in a similar way as the Beveridge drive system.

Since the fabrication of the epoxy electrodes and the chosen drive system don´t allow for easy electrical segmentation as with wire stators, mechanical means are used to shape the distribution character, hence the lens system. Consequently the rear side sound waves are ´controlled´ by the closed cabinet, though I have serious doubts that the cabinet is large enough and of a shape that early reflections from the cabinet´s inside can be avoided (the thin menbrane is acoustically transparent, so the reflected rear sounds could do whatever damage). Still though a very starightforward and logical design, which of course must also lead to a different sonic impression.

jauu
Calvin

[david yost]

The Beveridge 2 cabinets I had were made from lightly braced 3/8" (or thinner) veneered plywood and the rear of the transducers were only about 4-6" from the back of the boxes, which had a layer of 2-3" acoustic foam as absorption. The only other absorption inside the boxes were layers of the same material inserted between the lenses and the sides of the boxes. The lenses were painted with what looked like automotive undercoating on the outsides. The assemblies were surprisingly light and could easily be moved by one person. The hard part was getting the transducer assembly "lined-up" with the amplifier modules--there were two tall alignment guide pins and a total of 5 bananas that had to be mated--3 for the transducers and 2 for a safety interlock.

I posted a few pictures of the internals on audioasylum several years ago, and they should show up in a search, if you are interested.