Revisiting the Acoustic Research AR9 4-Way Legendary Loudspeaker

In 1978, Teledyne Acoustic Research introduced the AR9 (initially called the “AR-25”), a brand-new, technically advanced four-way loudspeaker system that was a radical departure from the firm’s familiar bookshelf-type loudspeakers—well-known to music lovers and audiophiles for many years. Now, forty-four years later, we revisit the flagship AR9, symbolic of the halcyon years of high-fidelity audio, just as the industry entered the digital age. This speaker was a remarkable achievement in loudspeaker design and came when audio enthusiasm was peaking; a time when audiophiles eagerly awaited each new issue of Stereo Review, High Fidelity or Audio magazine to read new-equipment reviews, admire the “Installation of the Month,” or laugh at Charles Rodrigues’ clever cartoons.

After it was publicly introduced, the AR9 was presented to Boston Audio Society members in a detailed overview by AR’s Director of Engineering Tim Holl, Chief Engineer Alex DeKoster and other AR personnel. As a part of the presentation, a pair of AR9s was compared side-by-side with a pair of AR-11s—both sets positioned close to the front wall. Many members commented on the improvements of the AR9 over the AR-11 (an improved version of the original AR-3a).

Early AR History

Acoustic Research, Inc. was incorporated in Cambridge, Massachusetts in August, 1954 by cofounders Edgar Villchur (President) and Henry Kloss (Vice President), and over the next decade was best known to the high-fidelity community for its pioneering inventions: the AR-1 acoustic-suspension loudspeaker, the AR-3 speaker with dome tweeters and the AR-XA Turntable.

 

Acoustic Research dramatically altered the course of hi-fi history with the huge success of its bookshelf loudspeakers—with superior bass reproduction and smoother, more “lifelike” midrange and treble reproduction, compared with most of its competitors—nearly obsoleting the huge, expensive horn and infinite-baffle speakers so popular with audiophiles after WWII.

From 1955 until 1973—the “golden years” of AR—the company operated out of an old four-story factory at 24 Thorndike Street in Cambridge, Massachusetts. Most of AR’s early accomplishments resulted from work by inventor-researcher Edgar Villchur, but there were also many contributions from other talented AR personnel, including Roy Allison, Henry Kloss, Jerry Landau, Chuck McShane, Harry Rubenstein, Victor Campos, Maurice Rotstein, Manny Maier and Abe Hoffman. As a result, AR prospered and its industry reputation and success grew because of product quality, innovation, durability/reliability, customer satisfaction and for the company’s industry-leading workplace environment and employee benefits. AR’s leadership in the high-fidelity industry grew steadily for over fifteen years. For example, by 1958, AR’s sales had grown significantly, and earnings-per-share dramatically increased to $89.89 from $21 for 1956 and $28 for 1957 (Annual Report, Acoustic Research, Inc., 1958), reflecting product innovation, improved production engineering and stabilized plant efficiency after February 1957.

Remarkably, by 1966, Acoustic Research loudspeaker sales had grown to approximately 32% of the entire domestic loudspeaker market (HiFi/Stereo Review annual sales reviews and Institute of High-Fidelity Manufacturers statistics), a milestone representing the highest speaker market share in U.S. high-fidelity history.

This famous New England speaker company didn’t go unnoticed, and in June 1967, Villchur and Henry Singleton, CEO of Teledyne, Inc., negotiated the sale of AR, Inc. to Teledyne, making it a subsidiary of the huge aerospace conglomerate. As a condition of this purchase agreement, Villchur insisted on a five-year employment contract for key AR management personnel in a two-tier arrangement that included the original top-management AR executives. This arrangement lasted until 1972, when the original management team of AR either retired or moved on to other pursuits, and Teledyne subsequently added management personnel to reflect a newer approach to improve sales and market-share results.

AR continued to grow and prosper in the years after the 1967 acquisition—still under the direction of the old AR management—but its once dominant position in the industry declined as many rival companies emerged (some were AR descendants), creating greater competition and innovation for market share. Most of the new speaker companies were located on the East Coast and produced similar products, including KLH, Advent, Cizek, EPI (formerly Epicure), Bose, Allison Acoustics, ADS, Boston Acoustics, dbx, Genesis, HH Scott, Fisher, NHT, Rectilinear, Snell, Dahlquist, Design Acoustics, Dynaco, RadioShack, Empire, Infinity, Marantz, Polk plus many others located in the Midwest and on the West Coast, as well as many well-known products from Europe and Asia.

The years of dramatic, “revolutionary” inventions, such as the acoustic-suspension woofer and the dome tweeter, were behind the company. Teledyne therefore recognized the need to invest significantly in Acoustic Research by updating facilities and manufacturing, adding personnel and ultimately moving the entire operation in 1973 to a large, modern factory on American Drive in Norwood, Massachusetts. This marked the beginning of many changes, as Teledyne began to modernize existing products and introduce new products and policies. During this period, AR hired many talented engineers and designers, and at one point was rumored to have more loudspeaker engineers than any other consumer-loudspeaker company.

During the early 1970s, AR speaker sales grew, but market share declined as new competitors emerged. Realizing this dilemma, AR updated existing models and developed a new series of improved bookshelf loudspeakers known as the “Advanced Development Division” (ADD) products. Importantly, Teledyne management was keenly aware that from 1967 until 1972, new-product development and marketing strategy under the legacy management team was somewhat stodgy and had evolved within a conservative, business-as-usual framework.

For example, several competitive loudspeaker manufacturers shot past AR in sales and surpassed it in market share. At half its cost, competitor Advent Corporation’s The Advent Loudspeaker (the large Advent speaker) easily outsold the AR-3a, offering nearly identical overall performance and better value. “Just as good at half the price,” rang Advent’s sales motto, often enthusiastically recited by hi-fi salesmen on the showroom floor. Technically, it may not have been as fine a product as the AR-3a, but at its bargain price, few consumers seemed to care. In fact, Advent’s speaker easily outsold other AR speakers, too, such as the AR-2ax and AR-5.

By 1972, the company’s market share dropped to 12.5%, and Teledyne began to feverishly modernize and improve existing classic products, such as the AR-2ax, AR-3a, AR-4x, AR-5, AR-6 and AR-7; and in 1975, Teledyne AR introduced a new series of up-dated and improved bookshelf speakers as part of the ADD. These included the AR-10π, AR-11, AR-12, AR-14, AR-16, AR-17, AR-18 and the others to replace the older AR models, which were soon discontinued.

This new product line included improvements in driver technology, reliability, and performance. The physical appearance was modernized with new cabinets and dark-foam grills to bring the products more in line with industry trends. In addition, the company’s new sales and marketing organization established many new dealerships with more aggressive discounts and promotions for higher sales, and these measures did indeed improve sales for the company. This effort helped the company maintain marginal growth through the last half of the 1970s; but for many reasons, it was not enough to regain lost momentum in the fickle high-fidelity industry that it once led with such a huge market-share.

The AR9 loudspeaker resulted from an all-out effort by Teledyne AR and its engineers to reestablish AR’s former engineering reputation and contribution in the field of loudspeakers. Specifically, the goal of the engineering group for the AR9 was to (1) improve low-frequency bass response, (2) increase power-handling capability, (3) improve frequency-response smoothness and (4) optimize stereo-imaging capabilities.

Engineering the AR9 (and related tower loudspeakers) likely included more labor, technology, and R&D than any loudspeaker ever designed by Acoustic Research. Nearly two dozen engineers and technicians were involved with its development, and the final product resulted from collaboration between AR’s Engineering Department and the Research Department. Extensive measurements, computer analysis and testing were involved, and Teledyne was willing to invest heavily in this new speaker series.

During this period, AR’s Engineering Department was led by Tim Holl, AR’s Director of Engineering, and the Research Department was led by Robert Berkovitz, AR’s Director of Research. These men and both groups—usually very competitive—collaborated completely in the engineering development and production implementation of the AR9, and AR was justifiably proud of this teamwork. For example, Tim Holl’s group offered the best driver designs, the opposing woofer design, and the special crossover; Robert Berkovitz’ group contributed the “acoustic blanket,” impulse-response measurements and enhancements and new computer-measurement improvements. The AR9 loudspeaker design was therefore planned and implemented equally between team leaders, Tim Holl and Robert Berkovitz. Specific team members included Alex DeKoster, Carl Loguidice and Steve Goldstein under Tim Holl; James Kates, James Raymond and C. Victor Campos under Robert Berkovitz.

Chief engineer on the AR9 was Alex DeKoster, selected by Acoustic Research in 1976 (over several applicants, including speaker designer Peter Snell, who designed fine products at EPI and founded Snell Acoustics, for whom the eponymous company was named). DeKoster was responsible for much of the crossover and development work on the AR9. In particular, he helped design the innovative impedance-optimizing circuit for the two parallel-wired 4-ohm woofers, affectionately known at the time as the “electronic automatic transmission.”

AR9 Design Details

The AR9 was a large floor-standing, vertical, 4-way, 4-ohm loudspeaker with two 12-inch acoustic-suspension woofers, an 8-inch acoustic-suspension lower-midrange speaker, a 1½-inch upper-midrange dome speaker and a ¾-inch dome tweeter. Dimensions were 52 ¾”H x 15”W x 15 13/16”D with a net weight of 130 lbs., and the tower configuration eliminated most room-placement issues found with conventional speaker designs. The crossover frequencies chosen for the speaker were 200 Hz, 1200 Hz and 7000 Hz., well within the safe-operating range of each driver. The AR9 was manufactured from 1978 until 1982, priced initially from $650/each in January 1978 to a high of $900/each at the end in 1982—remarkably modest prices at the time for such excellent performance.

By using two (AR-3a/AR-11) 12-inch, high-excursion, low-resonance (18 Hz) woofers in an enclosure of 4¼ cu. ft. (nearly three-times the size of the AR-3a enclosure), the AR9’s system low-frequency output was extended, sound-output level increased and distortion level decreased from the existing AR-3a—for many years considered a reference standard for low-distortion bass and extended low-frequency response. Using two side-mounted 12-inch woofers in the larger enclosure doubled the radiating surface and lowered system resonance, but also resulted in over-damped bass response at resonance.

Note: Roy Allison faced a similar problem with his flagship IC20; and to overcome it, he wired the two 10-inch woofers in series and reduced magnetic damping by using smaller magnets with lower flux density to achieve a higher low-frequency “Q” and improved low-frequency extension.

With the AR9, however, AR wired the two 4-ohm, 12-inch woofers in parallel, giving them the desired additional output (6 dB) but a very low impedance in the region above and below system resonance.

To counteract this, DeKoster—assisted by engineer James Kates—devised the impedance-and-Q-optimizing network for the parallel-wired woofers, the so-called “electronic automatic transmission,” and this network included an LCR impedance-matching circuit utilizing heavy-duty components (including a huge 2500 µfd. electrolytic capacitor). At resonance, the paralleled woofers are at their highest impedance and are in the safe range; therefore, when the input-signal frequency drops to the area of fundamental resonance, the choke and capacitor become shorted—and the resistor essentially bypassed—thus allowing the woofers full output; above and below resonance, however, the paralleled 4-ohm woofers drop precipitously below 2 ohms, but the tuned LCR circuit adds resistance in series with the two woofers and restores the proper impedance and damping (and raises the “Q”) for uniform bass response and acceptable load impedance.

The net result is a critically damped system with a Q of 0.5—optimum for a system with two woofers close to the floor-wall boundary and with uniform output down to below 30 Hz. The LCR circuit thus obviates excessive roll-off in output at resonance—indigenous to a system with such a low Q—by raising output 6 dB; on the other hand, the low Q provides nearly ideal woofer damping, thus preventing any response “ringing” or “hangover.”

When room gain is added, the AR9 has prodigious bass output down to below 20 Hz. The DIN rating for the AR9 has it down only 8 dB at 18 Hz. when measured in half space. As such, the AR9 proved to be an exceptionally potent low-frequency loudspeaker, but one that hardly called attention to itself. The speaker has such low harmonic distortion and smooth response, that some listeners—expecting heavier bass—felt the speaker was perhaps bass-shy, yet the true measure of a high-quality bass reproducer is no bass reproduction unless called for in the program material.

The woofer section of the AR9 crossover utilizes a full-section, 18 dB/octave filter with excellent damping to seamlessly blend the output of the two 12-inch woofers with the 8-inch lower midrange driver. This important, highly damped filter provides optimum low-frequency damping while keeping unwanted woofer output away from the lower midrange; moreover, a hallmark of the AR9 is the virtually undetectable blending of the woofers and the lower midrange unit, as though all sound emanates from one source at the system crossover of 200 Hz. The components of the AR9 crossovers present low series resistance (thus not dissipating power themselves) with large-gauge wire used in the air-core chokes in addition to heavy-duty, high-power, computer-grade capacitors, and of the three AR9 crossover boards, the woofer section alone weighs nearly 10 pounds.

The importance of the AR9 woofer-crossover characteristic is clear when contrasting it with a system composed of satellite speakers and a separate subwoofer. Any serious audiophile will tell you it is not a trivial task to seamlessly blend the output of these components; the subwoofer will eventually call attention to itself because of the lack of precision blending of the two systems, even when the subwoofer is turned down to a low output.

By mounting the AR9 woofers close to the floor on either side and utilizing the low 200 Hz crossover, AR engineers were able to deal with the room-boundary effects (“Allison” effect) on low-frequency reproduction which moved the reflected energy to approximately 500 Hz., well above the woofer’s operating range. During AR’s presentation of the AR9 to the Boston Audio Society, Tim Holl acknowledged Roy Allison’s work in this area in describing the design of the AR9, similar in approach of the Allison: One speaker’s pioneering “Room-Matched” design.

The AR9’s 8-inch lower-midrange driver operates throughout much of the important range of voice reproduction, and thus the crossover frequencies were carefully chosen to avoid any transition in the middle of its 200-1200 Hz. operating range—giving the speaker a great sense of midrange clarity and smoothness. This special driver is highly damped and operates within a sealed enclosure within the larger main-cabinet interior. The lower-midrange driver, with a sharp transition at 200 Hz, is always operating well above its mounted-resonance frequency of 175 Hz. Much of the AR9’s excellent clarity throughout the midrange can be attributed to the attention paid to the design of this driver.

In a vertical array above the 8-inch lower midrange driver is the 1½-inch upper-midrange dome driver, an adaptation of the original 1967 AR-3a 1½-inch treated-fabric, soft-dome midrange driver but devoid of the screen and fiberglass covering of the original version. This improved AR9 midrange driver, with its 1200-7000 Hz range, operates significantly above the original AR-3a midrange frequencies (575-5000 Hz), keeping it above its resonance frequency (800 Hz) and well within a safe-operating range. This fabric-dome driver has an energy-absorption dope applied to the outer portion of the diaphragm; it also has a “semi-horn” top plate that improves loading in the 3 kHz – 7 kHz range. This driver uses high-temperature adhesives and a special, high-grade Ferrotec Ferrofluid magnetic-damping fluid developed specifically for the AR9 for improved power-handling capability.

The tweeter in the AR9, covering the range above 7 kHz, is a ¾-inch, treated-fabric soft-dome unit that is an adaptation of the original AR-3a hard-dome tweeter and later AR-10π cloth tweeter, but with superior power-handling capabilities due to the higher crossover frequency and the use of high-temperature adhesives and the special Ferrofluid in the magnetic gap. Energy-absorbing dope is applied to this dome as well, and the top plate provides slight loading at the highest operating frequencies for an extended treble response. All drivers on the front face of the AR9 are vertically aligned, a common practice today to reduce horizontal lobing, making this interaction between drivers less intrusive.

Note: AR further reduced vertical interference with the next-generation AR9, the AR9Ls with its Dual-Dome “Lambda” midrange-tweeter.

Part of the research into the AR9’s performance was the study of diffraction and reflection of short-wavelength frequencies in the operating area of the upper-midrange driver and tweeter, and Associate Engineering Director James Kates performed extensive impulse testing on many loudspeaker enclosures using AR’s new Digital Equipment Corporation (DEC) PDP-11/40 16-bit minicomputer, previously specified by R&D director Robert Berkovitz. Kates and Berkovitz had patented a loudspeaker system equalization network, and the two worked together to develop the PDP-11 computer for speaker-testing research.

This research revealed problem areas—particularly with wide-dispersion drivers—that resulted in the introduction of the AR9’s “Acoustic Blanket,” a method of using a felt/foam pad on the face of the speaker’s front panel to suppress unwanted reflection from the front baffle, adjacent to the lower-midrange driver, upper-midrange driver and the tweeter, in effect to absorb unwanted refracted energy. This absorption of energy on the panel helps to reduce small dips and peaks, thus further smoothing output energy and improving response clarity.

The early use of AR’s PDP-11 computer for loudspeaker design and testing was one of the earliest applications of this type of testing in the high-fidelity industry, leading the way for digitized speaker testing. During the AR9 presentation to Boston Audio Society members, Kates showed graphs of impulse tests of AR speakers (and several other speaker brands), describing how to interpret the very short-duration energy bursts being studied while researching means of suppressing unwanted diffraction and reflections. One investigation was to determine the relevance of phase relationships over time periods of more than a few milliseconds and its audible perception by listeners. It was found that phase relationships were inconsistent and not generally audible when lasting longer than a few milliseconds; however, other tests showed that reflections and refraction of sound from within the speaker structure itself, from cabinet edges and molding, contributed to a smearing of the spectral image, often reducing clarity and intelligibility.

This differed from the efforts by some speaker designers to mount drivers in a physical alignment to adjust (“time align”) the arrival time of signals; AR found that this stepped-baffle arrangement actually made things worse by increasing early reflections and thus degrading performance. AR’s Acoustic Blanket, however, reduced the reflections and bounced energy before it could be reflected from the cabinet edges, thus smoothing system frequency response. Many tests were performed with and without the energy-absorbing pads before determining that the pads helped reduce coloration, improve localization and smooth frequency response.

Kates worked for Bob Berkovitz in R&D from 1975 until 1978 and contributed to the AR9 design both with crossover work but particularly his PDP-11 impulse testing. After AR, Kates has had a career as a scientist in audio, psychoacoustics and digital design, most recently as Scholar in Residence, Department of Speech Language and Hearing Sciences at the University of Colorado in Boulder.

The AR9 received universally excellent critical reviews from the U.S. and European high-fidelity audio press. It was extensively reviewed, but probably the best description came from the dean of audio testing and review, Julian Hirsch of Stereo Review,

 “The measurements we made on the Model AR9 … yielded the widest and flattest frequency-response curve we have yet obtained from a speaker system.”

The AR9 Inspiration of Future Audio Products?

After forty-four years, the AR9 remains one of the finest loudspeakers ever designed for home use. It was indeed a reference standard—used by many manufacturers as a benchmark for their speaker designs—and few speakers of any size or price could match its outstanding low-frequency performance, its midrange and treble clarity, power-handling capability or effortless and spacious frequency response across the entire audio spectrum.

The AR9 thus contributed to the science of sound reproduction and digital-measurement techniques, and this speaker design became a showcase for Teledyne Acoustic Research’s enormous investment in time and money during its development. The speaker was produced from 1978 to 1982 and sold well for a top-of-the-line speaker, thus helping AR achieve improved sales performance. Over the next decade, the AR9 became the foundation for new speakers.

Certainly, as speaker designs have evolved, improvements in quality, reliability and value have occurred—especially in professional digital monitors used in recording studios or sound reinforcement—with the application of digital designs, advanced materials and methods of manufacturing. Changes are occurring less frequently in this mature industry with far fewer “revolutionary” advances and improvements, a reflection of the slow and inevitable decline of home audio and the diminished interest in audio equipment for the home.

Nevertheless, the AR9 was a no-compromise design and regained much of AR’s reputation for engineering excellence. Without Teledyne AR’s eagerness to restore engineering excellence and its virtually unlimited financial and technical resources, the AR9 could not have been possible. The AR9 speaker design reached a high mark in engineering excellence and arrived just as the popularity of stereophonic sound reproduction with individual components began its slow downturn. As AR used to say, “the accurate reproduction of music is one of man’s more benevolent technological gifts to himself.”

Sidebar: The “9D” Powered Speaker

During the development of the AR9—before the final product was released—Bob Berkovitz conceived of a “powered” AR9, the “9D,” and he had Erik Edvardsen (a talented electronics engineer recruited earlier from Dolby Labs by AR President Martin Borish) design a low-power, high-peak amplifier to power each section of the 4-way AR9. AR engineers began to think in terms of a 4-way, “self-powered” AR9, and it became known as the AR9D, or simply, the “9D” for “digital.” Within the company, there was considerable interest in this self-powered AR9 product, but it never received official approval and the project was dropped. At the request of Robert Berkovitz, a quad-amped AR9 prototype was designed by prominent loudspeaker designer Ken Kantor (Ken arrived at AR around 1976 during the original development of the AR9 and assisted in the AR9 project), but beyond this effort, no further work was done on it.

The amplifier design that was to be used in the “9D” later became the NAD 3020, one of the best-selling integrated amplifiers in high-fidelity history. The concept was to allow the speaker to operate at normal levels with low or moderate power, yet accurately reproduce high peak levels by using a dedicated high-peak-current amplifier for each driver section. Before it started on its own, NAD was known as “New Acoustic Dimension by AR.” At NAD—essentially spun off by Teledyne AR—the company developed many innovative electronic designs by Erik Edvardsen.

About the Author

Tyson, from North Carolina, retired from a large electronics company, was never employed by Acoustic Research (AR). For many years, however, Tyson has been an audio hobbyist and historian with a particular interest in Acoustic Research. He co-owned a hi-fi dealership in the 1970s and through the years became acquainted with many of the engineers and officers at Acoustic Research. In 1993, Tyson placed an AR-3 loudspeaker on permanent display in The National Museum of American History of the Smithsonian Institution.