TWEETER TECHNOLOGY

Ribbon Loudspeakers by Newform Research.

BACK 1 2

Tweeter technology gets most of the attention when it comes to advances in loudspeaker design. Whereas their larger mid-range and woofer siblings may have been refined tremendously over the past 4 decades, tweeters have been completely transformed. From the early cone and horn drivers have evolved an almost unrecognizable array of high frequency devices. These new types of tweeters range from the field tested mainstays of the industry to brilliant but fatally flawed flights of theory.

What does a tweeter do and why is it so important? A tweeter reproduces the higher frequencies of the audio spectrum, generally from 2000Hz to 20,000Hz. This is where the life and delicacy of most music resides. Every area of the audio bandwidth is important but high frequencies form the emotional link with most music and are probably the hardest to get right.

Hence the huge diversity of design approaches. These designs all begin with the knowledge that the objective is to create an air pump not a musical instrument. An instrument implies a character which will create its own music. A good audio air pump will limit itself to accurately reproducing previously created music.

Air is such a compliant gas, it can be pumped in many different ways. It can be pushed, squeezed, levered, compressed and heated until it moves according to the dictates of the motivating pump. Now let's put a few names and descriptions to the various types of tweeters.

Under the Dome

The most widely recognized in audiophile circles is the dome tweeter. A dome consists of a hemispheric diaphragm whose perimeter is attached to a voice coil. As the audio signal from the amp moves through the coil, it reacts to the constant magnetic field established by the magnet in the magnetic gap in which the coil is suspended. As the coil moves, so moves the diaphragm and so moves the air. Conceptually, domes and cones (which are attached at their apex to the coil) are very simple devices.

What makes them perform differently is an infinite combination of factors, two of which are diaphragm material and shape. There are round domes, shallow domes and almost flat domes. Domes can be inverted, kind of like belly buttons, except that inverted domes do more than collect lint.

The dome can be made from virtually any material. Cloth, silk, boron, titanium, paper, Kevlar, flat aluminium honeycomb, you name it. If it can be glued to a voice coil it has been tried. But why try? Because different materials react differently to the stresses of being ripped back and forth 20,000 times a second pushing a mass of air each and every time. The domes respond to this stress by changing shape. This is called breakup. When a diaphragm ceases to act like a perfect piston, it is adding its own sound to that of the music.

This will happen to every speaker at some frequency and at some power level. The trick is to push the point of breakup to beyond the threshold of audibility or to make the breakup so benign it is less offensive. Combinations of laminations of different materials produce a combination of characteristics and when all else fails, we can coat the surface, or part of it, with "doping" compound to further dissipate the primary materials fundamental resonances. Soft materials breakup in a pattern of gradual decay. Hard, brittle materials breakup in glorious resonant peaks and precipitous valleys. Glorious on a graph (unless one wishes to maintain his employment) but tough on the ears if it happens anywhere this side of 22,000Hz. The art of tweeter design is in the management of these inevitable resonances.

Keep this in mind the next time you are listening to a system with a great high end. You are not listening to a perfect tweeter, you are listening to a collection of controlled resonances so skillfully blended, you do not even recognize their presence.

Critical Mass

This is probably the time to mention the weight (ok, moving mass) of the dome and the voice coil. As the frequency increases, the harder it is to keep a given mass under control. That is why light weight is so prized in a dome material. Lighter means more extended frequency response but it also means more poorly controlled breakup modes. Actually, the idea is force over mass. You can motivate a cannon ball to 20,000Hz but you would need a bungalow size motor and a direct power plant feed to run it. Also, I would not want to toe them in, in case the suspensions failed and they reverted to their original mission in life.

Horns, for our purposes, are a basically domes or cones that have been loaded with a better control of the air via the horn structure aligned before them. They produce higher efficiencies but at the price of colourations as the higher frequencies echo down the throat of the horn.

So much for now, for the conventional drivers. Let's get on to something a little more creative.

From Pistons to Panels

Panel or film drivers (ribbons and electrostatics) are probably the next largest group of high frequency reproducers. These drivers consist of very thin films (between 1/3 and 1 one thousandth of an inch thick) suspended in either a magnetic field (ribbon) or an electrostatic field (electrostatic). An audio signal is applied to the film and the entire film surface moves which pushes the air. This is very different from the dome which is driven only at the perimeter. So we have a very large surface of very light film driven consistently over the great majority of its radiating surface. Hence very different frequency response and very different breakup modes and very different sound from domes! Film diaphragm material is typically Kapton or Mylar and can have vapour deposited conductive metal coatings or etched aluminium laminates. Different approaches produce among other things, different weights, different diaphragm tensions and, of course, different frequency responses.

To add to the complexity, most film drivers are dipoles (they send both a front wave and a back wave into the room). All drivers produce a back wave since the diaphragm has to move back before it can move forward again, but the back wave is trapped by the magnet structure in any moving coil tweeter.

Film drivers tend to inhabit the high end of the loudspeaker market but they are quite simple in the way they work. Don't mistake this to mean it is simple to make them work well.

Weird Tweeting

Now for some fun exotica. Revolutionary drivers from ESS Heil loudspeakers had a pleated Mylar diaphragm with a conductor running up and down the pleats. When active, the pleats of this diaphragm squeezed together motivating the air to "shoot out like cherry pits" according to the sales literature of the day.

Not resting on their laurels, Ohm created the Walsh driver which had a conventional voice coil attached to the top of a large metal cone. As the voice coil moved down, the cone rippled. The ripple carried on to the large base of the cone. On its way down, it pushed the air. A similar design from MBL in Germany used a segmented cylinder so that motion of the coil caused the segments to bow outwards, moving the air.

The Lineaum tweeter, currently available on some models from Radio Shack, uses a long voice coil at the apex of two bowed sections of plastic to produce sound. As the voice coil moves out, the entire bowed surface area moves out pushing the air ahead of it.

My personal favourite is the plasma tweeter which heated the air to produce sound. Think of it as an arc welder driven by an audio signal. No moving parts, therefore no resonances. As fast as the air itself. And now the fatal flaw. The developers left the driver on overnight and returned the next morning to find everything in the room white from the ozone produced by the process. If this tweeter had proved otherwise practical in the early 60's, we would probably all be dead of skin cancer by now.

A new technology for audio sound reproduction is called the "air" speaker. In effect this creates a sound source in mid-air by the intersection of two very powerful ultrasonic sound beams. Audio frequencies are produced as the different frequency beams (say 200kHz and 205kHz) intermodulate and the difference tone appears at 5kHz. Sounds neat but in order to get 100dB in the audible range the room would have to be filled with beams of 200kHz + signals with levels of at least 150dB. A fatal flaw if there ever was one.

Diffraction, Imaging and Directionality

But sound quality is more than just producing perfect (or good enough to fool you) linearity at the diaphragm. How the sound wave moves out from the tweeter, into the room and to the listeners ears is also of prime importance. You are listening to not just the tweeter but the entire sound delivery system which also includes the enclosure and the room.

Even as the waveform moves off the diaphragm, it is still not clear of the enclosure. As the front part of the wave moves out into the room, the side part of the wave must negotiate its way over the baffle and off to the side of the speaker. It must pass over the baffle where it may have received reinforcement or may have bounced and then as it moves along the baffle, it encounters the edge of the speaker cabinet which it then tries to bend around. The result is baffle bounce and diffraction, two effects which will afflict even the perfectly produced waveform.

They produce anomalies which arrive at the ear slightly after the primary wave. This causes a "smear" or lack of precision and detail. Rounded edge baffles, felt coverings, foam rings (Imagers) are all techniques to reduce these effects. They all work to some degree, but it is better to eliminate the problem in the design stage rather than trying to partially fix it later on. You will see more and more sculpting, narrowing or virtual elimination of baffles in the future as production techniques allow. A sphere or an egg shape enclosure minimize diffraction and have a very beneficial effect on sound quality. All things being equal, the more refined the baffle, the more refined the sound. What you see is indeed partially what you get.

The hot objective several decades ago was to emulate a "pulsating sphere" where all frequencies would be radiated evenly in all directions. Now the consensus seems to be swinging towards very controlled vertical dispersion and broad horizontal dispersion. Why? Most rooms have ceiling bounce. You want coverage and integration between speakers but you don't want reflections from the ceiling or floor. Broad and even horizontal dispersion produces superb soundstaging, vertical reflections degrade it. Important in stereo reproduction, it becomes even more critical in surround sound where the localization of sound sources requires much more control of the propagating wave.

Why do tweeters have different radiation patterns? Dispersion is a function of frequency and diaphragm diameter. A loudspeaker will tend to radiate well until the wavelength is smaller than the diaphragm. After that, beaming starts to set in and increase. Beaming is the term used to describe a response which is highly directional, the pattern of dispersion decreasing into an ever tighter beam of acoustical energy as frequency increases.

A Ribbon speaker that has a diaphragm 3/4" wide and 15" long will disperse horizontally like a 3/4" dome tweeter and vertically like a 15" woofer. So, when you stand up with your ear above the plane of the Ribbon, you can ask "what happened to the highs?" Ribbons produce a "strata" type dispersion pattern which, in my humble (and biased) opinion offers the best interface for the home listening room.

Dome arrays are being used by some major manufacturers to emulate the limited vertical dispersion patterns of film drivers. With an array, a pattern of cancellations is set up which effectively reduces vertical radiation. However, cancellation patterns carry a price as well and it is better to eliminate the problem than to replace a big problem with a number of smaller ones.

If you want to get the most out of your tweeter, eliminate the destructive reflections that are compromising it's performance in your room. Do everything you can to reduce reflections from the baffle, immediately adjacent reflective surfaces (most people can forget about reducing, in any practical fashion, ceiling reflections) and reflections from any surface close to the listeners head (i.e. the back wall). Do this and you will be rewarded with the cleanest wave your tweeter was designed to produce in the first place.

As if there is not enough to consider with the forward moving wave, some manufacturers have sought to either make use of the existing rear wave (with an open back called dipole radiation) or create a rear wave with a driver on the back (bipolar). These designs are called dipolars (using the existing rear wave, say from an electrostatic panel) or bipolars which have additional drivers in the back and fire in-phase with the main front drivers. The presence of a rear wave in the room, whether in or out of phase with the front wave, produces an airy effect which has been regarded as desirable by many for stereo reproduction. It tends to make speaker placement and system setup more difficult and room dependent. Also, as we move to a surround sound standard with totally discreet channels (Dolby Digital) the presence of a reflected rear wave arriving several milli-seconds after the front wave can only diminish the coherence and precision of the sound field.

Newform's Outlook

When we set out to design a superior loudspeaker, we settled on film drivers very quickly since the diaphragm is driven over virtually 100% of its radiating surface (yielding tremendous control) and the diaphragm is flat and automatically in phase with itself over its entire bandwidth. We settled on a narrow diaphragm Ribbon which would provide wide horizontal dispersion and limited vertical dispersion.

We heavily beveled the Ribbon structure and made it very narrow to keep its acoustic profile to an absolute minimum for the most consistent off axis response possible. The results of this approach show up in soundstaging depth and focus. We went to a mono-polar configuration for reasons of ease of setup and the need to get the diaphragm at the very front of the magnet structure to avoid cavity resonance which afflict many Ribbon designs. As well, we view any design having a rear wave as having very limited applications in a high quality surround sound application.

The future will see the refinement of our current tweeter architectures and perhaps a new technology will come to dominate the industry by virtue of zero reproduction errors and perfect room interaction combined with a complete absence of "fatal flaws". And then again, maybe not.

BACK 1 2

Text Size

Email: ribbons@newformresearch.com Tel: 705.835.9000