." -----
Daniel von Recklinghausen's most recent major technical project was the KLH series of Computer Controlled Loudspeakers, introduced at the June, 1979 Consumer Electronics Show. They are interesting not only on purely technical grounds, but also for the intriguingly different assumptions used as the basis of their design, and for their significance to KLH in the marketing context. Plagued by heightened competitiveness from European and Japanese loudspeaker makers, KLH has suffered from anemic sales throughout the '70s; the Computer Controlled Speaker concept is their long-researched effort to re-establish a strong position in the loudspeaker field.
Detailed in a paper presented by von Recklinghausen at the 65th Audio Engineering Society Convention (available as Preprint 1617 from the Society), the Analog Bass Computer concept is comprised of three basic elements: A variable gain equalizer with equalization slopes that are also dynamically variable, a thresh old circuit to determine the levels at which gain and slopes are altered, and a transducer analog circuit that examines the power amplifier signal returning from the loudspeaker for evidence of thermal overload or mechanical fatigue.
The analog bass computer is in essence, therefore, an equalizer that changes its equalization curve continuously over a wide range of levels to protect the loudspeaker from damage that might result if a simple bass-boost circuit were used. The loudspeakers themselves--there are three units in the series--are vented designs developed in accordance with the now-classic Thiele/ Small speaker alignments (see "A.N. Thiele: Sage of Vented Speakers" by Ray J. Newman, Audio, Aug., 1975, p. 30). But while Richard Small's elaboration of Thiele's "speakers as filters" thesis discusses system types where bass boost and system roll-off are achieved with a relatively simple electrical equalizer, von Recklinghausen's version uses a dynamic equalizer that changes parameters as it goes along--altering the amount of bass boost by more than 30 dB, depending on signal level, and removing the bass boost entirely during very high-level passages. As shown in the family of curves, which apply to the smallest system in the series, the speaker's-3 dB limit may be as low as 40 Hz at moderate signal levels or may be as high as 160 Hz at peak levels.
The big question, of course, is whether a loudspeaker that uses dynamic compression of the bass signal to protect itself and to re duce low-frequency distortion is going to sound realistic playing back wide-range musical material.
KLH President Denis Wratten notes that "I would rather listen to some subtle alteration of the deep-bass dynamics than hear the speaker choke up completely on high-level bass passages." Many speaker de signers agree that this is the way to build a truly wideband small speaker, though the proof is naturally in the listening; meanwhile, DvR is at work on new projects.
-G.S.
-----
-- We have proven that very narrow response peaks are audible. --
Janszen. He then begat--not necessarily in correct sequence--the KLH Model 9, the Acoustech units, and he has recently gotten involved with Dennesen. All this history aside, one major problem with electrostatic loudspeakers is that in order to get reasonable efficiency, you have to operate the speaker very close to the breakdown field intensity that exists in air. In the process, too, you become a dust collector. When the air is not terribly clean, as in New York City, you have to periodically send your speakers out for professional cleaning.
Audio: What about if you place them in an envelope of high-dielectric strength gas, like Dayton-Wright in Canada has done?
DvR: It's a reasonable approach. In fact, it's an approach that dates back to Van der Graaf and his electrostatic accelerators. But it's also expensive and prone to leakage. Not exactly a broad-appeal consumer item.
Audio: What about piezoelectric plastics in large-sheet form?
DvR: Piezoelectric plastics in large-sheet form work on basically the same principle as ceramics. Due to the internal electric field of the material, and the externally applied electrical field of the signal from the amplifier, one side of the piezoelectric material expands while the other one contracts, causing the whole surface to bend.
That bending is used in a sort of secondary way as a sound transducer. The approach has virtues. Where it lacks virtue is in its necessity to convert this bending motion-really a form of flapping-into a uniform acoustic out put. You end up with a transducer that in its most practical form is perhaps a cylinder. And, as a cylinder, you end up with a transducer that does not radiate uniformly over its circumference. You must support the piezoelectric sheet in some way, and the supports do not move.
Audio: What about other variations on the flat-sheet principle?
DvR: There are moving conductors, ribbons; there are flat sheets with conductors zigzagging across their backs. For both, you need substantial and ex pensive supporting magnet structures. Another practical point here relates to the scientific law prescribing that cur rent, direction of motion, and direction of magnetic force in a magnetic transducer of any kind are all mutually perpendicular. Now your motion should be in and out, which means that both the motion and the electrical current flow must be within the plane of the radiating surface. Then you must create a magnetic field perpendicular to those two, and here is where the difficulties come. It's very hard to get a high-intensity magnetic field in this plane, because you are really trying to create sound with the stray energy in an air gap. The larger your air gap, the more losses and the lower the intensity you have. Result: An extremely expensive and inefficient magnetic system, no matter what you do.
You can put the radiating surface between two magnetic structures and operate it as a push-pull device. But you also have a cage of some kind in front of the radiating surface--a cavity resonator--and this will affect frequency response. You may be able to use these resonances positively, for in stance, to augment the very high frequency output by placing the cavity resonances in the high frequencies, but they remain resonances and there fore a source of coloration.
Audio: If you had to predict future trends in speaker design over the next ten years, what would you project as the most promising approaches? Do you, for instance, see integration of the loudspeaker and the amplifier being a natural approach in the '80s, as a great many engineers do?
DvR: Integration is one thing that will look good. The use of more reproducible materials as radiating surfaces is another clear trend-away from paper and other fibers and into plastics whose time is coming. We've seen the light here at KLH, and others will as well, eventually. Paper cones, as used in speakers, have the undeniable ad vantage of having a very low velocity of sound in the material, coupled with a reasonable amount of mechanical damping. But cones are just paper fibers, and 'anything that involves fibers always involves some black magic, as well as some degree of inconsistency from one cone to the next.
==============
( Audio magazine, Feb. 1981)
Also see:
The Audio Interview: Rudy Bozak (May. 1982)
Interview with Les Paul (Dec. 1978)
The Audio Interview: Willi Studer (April 1981)
= = = =