Paradigm Reference Servo-15 powered subwoofer Review

If finding the best match means trying too many combinations of adjustments, the result is anomie—Durkheim’s term for increasing stress associated with a breakdown in the subtle rules of social conduct.Why? When an aftermarket subwoofer provides multiple adjustable settings for different satellite and main speakers and listening environments, an owner’s choices greatly expand.

Émile Durkheim (1858-1917) once wrote, “The less limited one feels, the more intolerable all limitation appears” (footnote 1). Although directed at the paradoxical observation that suicide rates are higher in newly prosperous countries than in those mired in poverty, his comment applies equally well to subwoofers.

John Atkinson, measuring the Servo-15 as part of the Paradigm Reference Active/450 system for J. Gordon Holt’s review in the November 1998 issue of Stereophile Guide to Home Theater (p.98), noted that “the owner…enjoys considerable flexibility in getting the system to blend optimally in the room.”Which brings me to the Paradigm Reference Servo-15, a compact, self-powered, servo-corrected subwoofer with multiple control settings.

For this review, I tested the Servo-15 system with the Quad ESL-63, a bass-shy electrostatic speaker notoriously hard to blend with subwoofers. Would the Servo-15’s elaborate controls facilitate a better match, or just produce anomie?


I liked the ruggedness and build quality of this compact subwoofer the minute I pulled it out of its shipping carton. Its sealed enclosure is constructed from heavily braced, 1″-thick MDF, and a removable, MDF-framed grille snaps into the front baffle. Four ¼” plastic feet—replaceable with spikes—raise the Servo-15 off the floor. The rear, metal service panel contains the amplifier’s heatsink and just two controls: a bass-level rotary dial, and a three-position power-on switch (Off, Signal Sense, Always On). This panel also has a detachable AC IEC jack for the AC power cord. That the Servo-15 accepts only a single line-level, unbalanced interconnect could be a problem: its lengthy single-ended cable—necessary to place the sub in an optimal spot in the listening room—could be more prone to hum than a balanced cable would be.

The massive 15″ driver occupies most of the enclosure’s front panel. Constructed from a sturdy composite reinforced with Kevlar fiber with a multilayer surround, the driver sits in a die-cast aluminum chassis. The cone has a 1″ range of excursion. Its motor assembly employs three 6″ ceramic magnets, and has a flared center vent—like a bass-reflex port—that reduces the “chuffing” sounds of air moved by the driver. Two shorting rings around the pole-piece help improve linearity and heat dissipation.

The voice-coil consists of a 2.1″-diameter aluminum former 1.4″ long, to which is bonded the “instrumentation-grade” accelerometer. This is used by a closed-loop servo circuit to compare the subwoofer cone’s movements to the input waveform. As well as the servo circuit, the 400W amplifier is controlled by a patented limiter circuit designed to prevent excessive driver motion.

The biggest physical difference between the Servo-15 and other aftermarket subwoofers is the external location of its controls. These reside in the X-30, a separate crossover module linked to the subwoofer by a heavy 23′ RCA-to-RCA connector cable. The X-30 includes two filter networks, both with 18dB/octave slopes: an active high-pass for setting the lower-frequency extension of the satellite and main speakers, and an active low-pass for setting the upper-frequency extension of the subwoofer.

The front panel’s continuous rotary controls allow you to set the subwoofer’s level, tune the low-frequency cutoff between 35 and 150Hz, and adjust the Servo-15’s phase between 0 and 180 degrees (footnote 2). The X-30’s rear panel accepts single-ended inputs from the preamplifier, and there are two blended-mono output RCA jacks—one with adjustable phase, one without—for up to two Servo-15s. To drive the satellite/main speakers’ amplifier, the rear panel sports 50Hz, 80Hz, and 120Hz RCA output jacks. AC power comes from a small wall wart, whose short cable plugs into the X-30’s rear panel.

The ruggedness of its driver, the sturdiness of its mounting hardware, the satin feel of its cabinet—all in all, I found the Servo-15’s fit’n’finish to be first-rate. In a marketplace of lookalike, vinyl-veneered, black-cube subwoofers, the Servo-15 stands out as one of the best built. Because of this high quality, I was disappointed that it did not have balanced XLR input/output connectors, defeatable high-pass filters, low AC voltage shutoff, or a wireless remote for adjusting its output from the listening chair.


I can’t say I enjoyed installing the Paradigm Reference Servo-15 in my system—it took me more than three weeks to test different control settings and room locations. During this time I attempted—with varying degrees of success—to tune the Servo-15 to three different main/satellite loudspeakers: the Quad ESL-63, the Dynaudio 3.0, and the B&W 805 Nautilus. I used a signal generator, played all of my “special bass” CDs, and occasionally switched back to my comparison subwoofer, the Velodyne FSR-18, to recalibrate my ears. For all listening comparisons, the bass-level controls of the FSR-18 and Servo-15 were set to produce similar output levels of pink noise.

Footnote 1: émile Durkheim, Suicide, pp.246-260. The Free Press (Glencoe, IL), 1951.

Footnote 2: When JGH reviewed the Servo-15 in SGHT, he suggested that its phase control made sense when distances from the listener to the subwoofer and to the main speakers were not identical. Adjusting the phase of the subwoofer could prevent the cancellation of bass signals that reach the listener from different sources at different times.

I moved the Paradigm sub around, first into corners and then to the center of the rear wall behind the main speakers, about 3′ behind the Quad ESL-63’s grilles and 10′ from my listening chair. The review took place in my 5500-ft3 listening room: 25′ long by 13′ wide, with a 12′ semi-cathedral ceiling and a rear area that opens into a large kitchen. Some subwoofers have overloaded ungracefully when played at high volumes into this large area. As a result, it’s no surprise that, when I’ve asked a subwoofer to deliver 20-25Hz bass notes, it’s the big ones—18″ drivers with lots of amplifier power—that have done best. Still, I had high hopes for the compact Servo-15.

I matched the satellite and main speakers to the Servo-15 by adjusting the X-30’s controls from my listening chair. This required hauling out and connecting a huge, multicolored nest of cables and wires, including a yellow, industrial-grade 100′ extension cord for the X-30’s AC supply, 20′ of heavy-duty AudioQuest Topaz interconnects to connect the X-30 to the Mark Levinson No.334 power amplifier, and another 15′ run of Levinson interconnect cable to reach the Mark Levinson ML-7A preamplifier. Bass sources for this calibration included a steady 42Hz sinewave from the signal generator, or the repetitive bass drum on “Cosmo…Old Friend” from the Sneakers soundtrack CD (Columbia CK 53146). I also used Natalie Merchant’s “One Fine Day,” from the soundtrack album of the same name (Columbia CK 69716), to identify the best X-30 high-pass option. Merchant’s voice sounded too thin using the 120Hz high-pass output jacks; the 50Hz jacks produced a more natural vocal timbre and gave the strongest stand-up bass sound in my listening room.

For all tests and listening, the Servo-15’s rear-panel volume control was left at its middle position (-46dB). The X-30’s front-panel controls ended up set as follows: Phase, 80 degrees or 90 degrees; Low-Pass, 100Hz or 120Hz; Bass Volume, -64 (B&W 805 Nautilus) or -59dB (Quad ESL-63). Minor bass-level adjustments were made for different CDs.

Clean sweep
I drove the X-30/Servo-15 with 60Hz-10Hz sinewaves from the Heathkit sinewave generator to check its bass extension and output level by ear, setting the level high enough to see considerable cone excursion. As noted by other reviewers (footnote 3). the Servo-15’s output seemed to drop as the frequency descended. Sweeping the generator sinewave up from 10Hz, I sensed air pressure on my face and felt vibrations beginning at 17Hz, where the output was clean, with no sign of higher-order harmonics. Listening to the subwoofer play music by itself—using “Gnomus,” from Jean Guillou’s pipe-organ transcription of Mussorgsky’s Pictures at an Exhibition (Dorian DOR-90117)—the output was ultra-clean, solid, tight, and free from rattles, static, or other spurious sounds. This is exactly the kind of performance I’ve come to expect from the best servo-controlled subwoofers.

The sound of music
First, the not-so-good news: While the Servo-15 reached all the deepest bass notes, it did not increase the punch or bass heft of the Quads. Each time I played the ponderous synthesizer rumblings of “Assault on Ryan’s House” (from James Horner’s Patriot Games soundtrack, RCA 66051-2), the Servo-15’s output was just too precise to startle or frighten me. The well-recorded bass drum in Liberty Fanfare, on the National Symphonic Winds’ Winds of War and Peace (Wilson Audiophile WCD-8823), seemed too polite over the Servo-15 in my room. Some compression was evident in the decay of the sound that followed each bass-drum whack, which, rather mechanically, intensified and shut off with the Servo-15 rather than gradually fading, as it did with the now-discontinued Velodyne FSR-18.

I tried—with limited success—to improve the Servo-15’s cautious bass character in my large listening room by adjusting the X-30’s controls. What made soundwave-generated tones sound perfect produced deep bass with excellent pitch but not enough punch. The X-30’s phase and low-frequency cutoff controls could not increase the Quad’s bass slam when they were augmented with the Servo-15. Try as I might, no phase or bass-level setting on the X-30 completely cured the Quad/Servo-15’s tendency toward thin midbass—a suckout caused, I assume, by the Quad’s backwave canceling more of the Servo-15’s output than of the Velodyne’s. Each time I thought I’d finally gotten the Servo-15’s bass impact so it qualitatively matched the FSR-18’s, I’d go back the next day and find I hadn’t succeeded.

It tood a more expensive, 18″ servo-controlled subwoofer, the Velodyne FSR-18, to overcome the challenges imposed by the room and the Quad dipoles. In level-matched comparisons, the FSR-18—which cost $1000 more than the Servo-15—bested it by having more bass slam, greater deep-bass extension, quicker bass-transient performance, and better ability to expand the Quad’s dynamic range. This should be no surprise; the FSR-18’s physical specifications differ substantially from the Servo-15’s. The driver has 44% more surface area, its excursion is double that of the Servo-15’s, its enclosure volume is 0.8ft3 greater, its amplifier is three times as powerful, and the sub itself has 20dB greater rated peak SPL output. These differences enabled the FSR-18/Quad combination to really shake my room, reproducing bass-drum whacks with powder-keg slam, while the Servo-15/Quad system had merely “excellent” bass.

Footnote 3: D.B. Keele showed (in Audio, April 1998, p.70) that the Servo-15’s limiter “put a cap on the maximum sound pressure level” and “allowed the subwoofer to generate 133dB SPL at 40Hz but only about 98dB at 20Hz.”

The good news is that the Servo-15, when teamed up with excellent dynamic monopole speakers such as the Dynaudio Contour 3.0 and the B&W Nautilus 805, had outstanding tonal accuracy, great timbre, and clear, deep bass pitch. With these speakers, the Servo-15 was a champion at resolving bass detail and instrumental color. Drum-head timbre was evident in the bass-drum strokes that open La Fiesta Mexicana: Prelude and Aztec Dance, from Fiesta (Reference Recordings RR-38CD). Pitch changes in Joel Goodman’s very deep synthesizer chords on “Silk Road,” from Of the Marsh and the Moon (Chesky WO144), were clear and decisive. Glen Moore’s double-bass notes on “The Silence of a Candle,” from Oregon’s Beyond Words (Chesky JD130), were ultra-clean and well damped, with distinct tonal steps. Bass lines were instantly resolvable, with precise pitch definition and articulation.

Th X-30’s volume controls could be adjusted with each CD to coax the best performance from the Servo-15. Listening to David Hudson’s didgeridoo on “Rainforest Wonder” (Didgeridoo Spirit, Indigenous Australia IA2003 D), for example, I had to crank the X-30’s bass level way up to achieve the same deep, raspy, throbbing sound that I heard with the Revel Ultima Salon. On other CDs, I preferred the Contour 3.0, B&W Nautilus 805, or Revel Salon driven full-range without a subwoofer.

The Servo-15 was particularly good in the reproduction of sustained, solid bass, such as pipe-organ pedal notes. Deep chords were played with clarity, power, and excellent pitch definition. There was no sign of strain, overload, or clipping. The B&W 805/Servo-15 combination produced a bass “lock” on my listening room when organist Jean Guillou held a deep pedal note, playing Mussorsky’s “Gnomus,” causing the room to shake. The Servo-15 reproduced the 32′ pipes of the Great Kleuker-Seinmeyer Organ of the Zürich Tonhalle with sustained power, vibrating the room, but did not interfere with the instrument’s flute and trumpet ranks. It also reproduced rumbling organ-pedal notes when I played César Franck’s Pièce Héroïque on the Marcel Dupré Recital CD (Mercury Living Presence 434 311-2). Even more interesting were the sounds of the church pipe organ on Elgar’s The Dream of Gerontius, Part 1 on Stereophile‘s Test CD 2 (STPH004-2). The B&W 805 Nautilus/Servo-15 combo provided the “sense of space,” and clearly enunciated the final solid organ-pedal chords that “underpin the work’s tonal foundations,” as described in JA’s liner notes.

The Servo-15’s deep-bass extension and pitch definition was best heard on sustained deep bass synthesizer notes when combined with dynamic satellite speakers. The subterranean synthesizer chords that open Emmylou Harris’ “Deeper Wells” (Spyboy, EM-25001-2) enveloped and forcefully rolled over me, and the ultra-deep but slightly airy synthesizer representing the ghosts in the opening selection of the Casper soundtrack CD (MCA-11240) was captured perfectly. Similarly, the Servo-15 gave just the right weight to the distant and ethereal synthesizer bass that backs up “Silk Road” for the soft, rainy acoustic landscape the track depicts.


The Paradigm Servo-15 subwoofer system is recommended for medium to small listening environments and dynamic main speakers. For such applications, its $1500 price makes it an excellent value in a compact, 75-lb subwoofer. It offers servo-correction, low distortion, ultra-clean deep-bass output, and outstanding tonal accuracy. Its X-30 control module gives the Servo-15 the flexibility necessary to integrate with different types of monopole dynamic loudspeakers. With enough cable, the X-30 can be adjusted from your listening chair, allowing low-pass filter and bass-level adjustments on the fly.

But those interested in auditioning the Servo-15 should first try it with high-quality dynamic main/satellite speakers before using it with dipole panel speakers like the Quads. Despite the Servo-15’s superb engineering and excellent pitch definition, its comparative lack of bass slam, inability to expand the Quad’s dynamic range, and lack of quick bass transient performance in my large listening room were disappointing. No matter how many of the X-30’s control settings I tried, I couldn’t get more sock out of the Servo-15/Quad ESL-63 combo. This experience hasn’t helped my love/hate fascination with the Quad ESL-63, but it sure increased my understanding of Durkheim’s anomie—and of the Paradigm Servo-15.

Sidebar 1: Specifications

Description: active, front-firing, sealed-enclosure, servo-controlled, closed-loop, audio/video subwoofer system with outboard X-30 crossover/control module. Drive-unit: 15″ Kevlar-fiber reinforced composite cone with 2.1″ voice-coil, 1″ maximum excursion. Frequency response: 17-80Hz, +/-2dB. Total harmonic distortion: +/-0.3% at 50Hz, 90dB SPL. Amplifier: 400W RMS continuous output. Input impedance: 20k ohms.
Controls: Subwoofer: level, 3-way power switch (Off/Signal Sense/Always On); IEC detachable line cord socket; single-ended input connector (RCA jack). Control module front panel: Phase (0-180 degrees), Low-Pass Frequency Cutoff sonic filter (35-150Hz active low-pass filter, 18dB/octave active filter), Bass Level continuous. Control module rear panel: RCA single-ended inputs, outputs for two subwoofers, three pairs of RCA jacks for high-pass outputs (40Hz, 80Hz, 120Hz active high-pass filter, 18dB/octave), AC power jack, “wall wart” power supply.
Dimensions: 20″ (508mm) H by 18″ (457mm) W by 21.75″ (552mm) D. X-30 control module: 7.25″ (184mm) W by 1.8″ (48mm) H by 5″ (125mm) D. Weights: subwoofer, 78 lbs (35.5kg); control module, 2.7 lbs (1.2kg).
Finishes: black ash laminate; light cherry or rosenut wood veneer, add $250.
Serial number of unit reviewed: 11139.
Price: $1500. Approximate number of dealers: 350. Warranty: 3 years, home use only. Does not cover thermal damage.
Manufacturer: Paradigm, 205 Annagem Blvd., Mississauga, Ontario L5T 2V1, Canada. Tel: (905) 632-0180. Fax: (905) 632-0183. Web:

Sidebar 2: Associated Equipment

Analog source: Linn Sondek LP12 turntable with Lingo power supply, Ittok tonearm, Spectral moving-coil cartridge.
FM Tuners: Day-Sequerra FM Reference Classic, Rotel RH-10, Fanfare FM-1, Magnum Dynalab MD-108/Model 205 Sleuth RF amplifier, Myryad T-100.
Digital source: Krell MD-1 CD transport, Adcom GDA-700 D/A processor.
Preamplification: Krell KBL, Mark Levinson ML-7A with L-2 phono section, Duntech MX-10 head amplifier.
Power amplifier: Mark Levinson No.334.
Loudspeakers: Snell Type A Reference system, Revel Ultima Salon, Dynaudio Contour 3.0, Quad ESL-63, B&W 805 Nautilus; Velodyne FSR-18 subwoofer.
Cables: Interconnects: 75 ohm Silver Starlight digital coax, Madrigal CZ Gel-1 balanced, Krell Cogelco Yellow balanced, Bryston Balanced, Mark Levinson HFC (with Camac connectors) single-ended, Randall Research single-ended. Speaker cables: Mark Levinson HFC-10, PSC Pristine.
Accessories: Arcici, Sumiko Franklin & Lowell stands.
Test gear: Heathkit Model SQ 5218 Sine/Squarewave Audio Generator.
Burn-in: FM source for 12 hours, followed by “Special Burn-in Noise Track” (Stereophile‘s Test CD 3) for 12 hours.
Speaker placements: 5′ from rear wall, 3′ from side walls, 8′ feet apart, 8′ from listening chair.
Listening positions: nearfield: 8′ away or 16.5′ away; ear level for both positions, 37″ from the floor.—Larry Greenhill