Floor Loudspeaker Reviews

Few products elicited as much excitement, disappointment, and debate among the Stereophile staff as did the Snell Type B dynamic loudspeaker (footnote 1). Both Peter Mitchell and I praised the Type B for its low-frequency extension, smooth treble, high power handling, and excellent dynamics. Corey Greenberg and Robert Harley faulted the speaker's sluggish and fat bass response, which they felt precluded a recommendation in Stereophile's "Recommended Components." Kickdrum recordings in pop and rock excited the Type B's bass character, a characteristic which was not so noticeable if one only listens to classical music. This bass peak was so prominent in RH's listening room that it colored the rest of the loudspeaker's range. As a result, he found that this otherwise fine loudspeaker was not as transparent as other high-quality dynamic systems in the same price range.

Help was on the way. Within six months of RH's review, Snell Acoustics introduced a smaller version, the B minor, at the 1992 Summer CES. Although this announcement was overshadowed by Snell's Digital Signal Processing (DSP) technology, the B minor's bass appeared to have been harnessed, even if the listening was done in a less-than-optimal hotel room. Although the B minor sells for less than the B ($3599/pair vs $4799), it appeals for additional reasons: The B minor is physically smaller and 42 lbs lighter than the Type B; its front baffle is narrower; and a new titanium-diaphragm tweeter is used.

Technical description
The B minor has been described by its designer Kevin Voecks as being a "3½-way speaker" employing one woofer, two midranges, one tweeter, and the "half": a rear-firing tweeter that overlaps the front tweeter's range. The B minor resembles Snell's C/IV and E/III more than it does the B. A conventional tower, the B minor's cabinet is only 10" wide, in contrast to the B's 25"-wide, five-sided enclosure. (The B's extra width comes from the two 8" "wings" of veneered wood that flank its 13"-wide central black grille.)

The B minor uses a single 12" acoustic-suspension woofer, not the two 10" units found in the B. Unlike Snell's flagship Type A/III Improved loudspeaker, the B minor's woofer approach does not take advantage of bass reinforcement from the floor, as the woofer is mounted halfway up the side baffle. The 12" long-throw woofer has a cast basket frame and its double-thick magnet structure allows linear, high-excursion motion. It was designed to have superior power-handling characteristics.

The same midrange/tweeter/midrange configuration is used in the B and B minor: ie, the midranges are mounted above and below the tweeter (footnote 2). Snell's Kevin Voecks suggests that the B minor's narrow baffle enhances its imaging abilities. Like the B, the midranges and tweeter are slightly offset toward the inside of the cabinet, to spread the interactions with cabinet edges out in frequency. The felt surrounds for the drivers also yield a smoother off-axis response. Along with the side-mounted woofers, this asymmetry means that the B minors come in mirror-imaged pairs.

The B minor's tweeter is a brand-new titanium-dome unit replacing the 1" aluminum-dome Vifa tweeter found in the Type B. The company's literature suggests that this new tweeter has excellent on- and off-axis frequency responses and is said to exhibit "true pistonic behavior" throughout the audible range. It has increased power-handling capability due, in part, to a conjugate impedance-compensation circuit and a steep filter network. The rear-firing, ¾" dome tweeter is identical to the unit used in the Types B and C/IV.

The B minor's crossover employs high-quality, non-polarized electrolytic capacitors, Mylar capacitors, and air-core inductors. The B minor's midrange drivers cover the same frequency passband, 275Hz–2.7kHz, with 18dB/octave low-pass and 24dB/octave high-pass slopes. The front tweeter takes over at 2.7kHz. The back-mounted ¾" tweeter begins to play when musical information includes frequencies higher than 5kHz, gradually increasing in volume using a first-order slope. At those frequencies, the rear tweeter contributes to the reverberant soundfield, reinforcing the speaker's total radiated energy in the highs. Like the Type B, the B minor employs Monster Cable for internal wiring.

Below the tweeter on the speaker's back panel is a connector panel with hardware identical to the B's. There's a rear-tweeter on/off switch, a fuse holder, and a front tweeter-level control with continuous action. Instructions from Kevin Voecks suggested that the tweeter-level control should be set around 12 o'clock for the flattest frequency response. There are two pairs of five-way binding posts for bi-wiring/bi-amping. The speakers are shipped with flat metal jumpers configured in the shape of a capital "E" to allow the owner to drive the speaker from a single speaker cable. The slightly recessed input terminals did not interfere with my speaker-cable connections, which included both Sumiko OCOS and Levinson HFC-10 types.

The B minor's manufacturing process involves tuning each driver's frequency response, with grilles in place, to within 0.5dB of Snell's reference master standard. This tolerance is far better than the ±4–5dB quoted by the vendors supplying the drivers. Each speaker is trimmed to match each loudspeaker's amplitude response to a reference master. This is done by overwinding inductors, then pulling turns off one at a time; starting with smaller capacitor values and adding trimmers; and, finally, adjusting variable resistors. The tweeter-level control is also calibrated against the reference master.

The cabinet work appears to be first-rate, with fine fit'n'finish—my review samples were finished in carefully matched dark walnut veneer. Cabinet sides are made from ¾" high-density particleboard, while the front baffle features 1"-thick particleboard. During the review period Snell added more internal braces to the cabinet to increase rigidity. Like the B, the B minor rests on a particleboard plate and is supplied with carpet-piercing spikes.

Speaker location & adjustments
Detailed instructions for room setup were not included with the B minors sent for review. As with any installation, listening revealed the speakers' optimal room positions and toe-ins. It seemed evident that the side-mounted woofer grilles should face each other across the center space between the speakers. Via telephone, Kevin Voecks told me that the B minors were designed to be used out in the room, not against the back wall. (This is very different from the company's Type A/III, which was designed by the late Peter Snell to hug the wall and not intrude into the room space; it depended on boundary effects to support its deep bass.)

Listening-room dimensions are critical for any loudspeaker, and perhaps particularly so for the Type B. After RH's critical review, JA wondered if the B had been designed "by tuning its bass for maximal flat low-frequency extension under anechoic conditions." Thus, it would "tend to sound bass-heavy in a real moderate-sized room." The B's powerful bass might be better balanced in a large room. I tested the Bs in my very large listening room (5500ft³, with an effective room length of 52'), and reported the results in February 1992. Readers should consult that article (Vol.15 No.2, p.181) for specific details of my room's construction, exact dimensions, contents, listening positions, rugs, windows, bass modes, and results of Snell's room-analysis software program (footnote 3).

I concluded that the Type B's overgenerous bass could be balanced so that it was not overpowering, but only after careful setup and room positioning. The speakers had to be placed 54" from the back wall and driven in bi-wired mode by certain solid-state amplifiers (eg, Bryston 4B or Krell KSA-250). Though improved, the B's bass prominence in the kickdrum region was still noticeable. Would the B minors require the same time-consuming setup to avoid bass problems?

I began my review by placing the B minors near the room's narrow back wall, as suggested by Snell's room-analysis software. The B minors were positioned 24" between the rear of speaker enclosure and the wall, which put the loudspeakers' front panels about 45" from the rear, 36" from the side walls, and 84" apart. The speakers were toed-in slightly. My listening position was 18' away from a center line between the speakers, and the seat placed my ears about 34" off the floor (about the level of the B minor's tweeter).

I then adjusted the tweeter-level controls. As Voecks had suggested, I found that the 12 o'clock setting gave the best treble response for my taste. After this, I did the "sit down, stand up, walk around" procedure with pink noise, and found that neither on- nor off-axis treble responses differed markedly. I later tried other positions, but this initial setup proved to be the best.

Like the Type Bs, the B minors produced the best tonal balance when set up with bi-wired speaker cables. I disconnected the jumpers between the B minor's tweeter and woofer five-way binding posts, and connected the double spade lugs at the end of each parallel run of OCOS cable.

I swept the loudspeakers with a Heathkit sinewave generator before I began listening to music. When I'd done this with the Type B, I'd heard the bass increase markedly as I swept the frequency down past 42Hz, very near the crossover point between woofer and subwoofer. When I'd moved around the room, there were differences in the amplitude of this bass note, suggesting that room modes were being excited.

Not so with the B minor. No 42Hz prominence was found during the room sweeps—just a slight increase in output between 47 and 60Hz. Sweeping downward, I was able to hear the low-frequency response clearly down to 35Hz; this differed from the B, where I'd found detectable output down to 28Hz (footnote 4). So far, so good: The B minors were easy to set up, and had a much less prominent bass emphasis.

Listening
I'll cut to the quick. Yes, the B minor had a mild midbass peak in my listening room, but it was not as prominent or overpowering as that of the Type B. The good news is that this more felicitous tonal balance allowed the Snell B minor to image well, show greater midrange and treble clarity, and play with transparency.

I tested the B minor's bass performance with a variety of musical sources: rock recordings for kickdrum bass peak; male vocals, including FM announcers, for upper-bass excess; organ and synthesizer recordings for deep bass; and synthesizer recordings for bass transient speed.

First I replayed all the musical selections that had provoked the Type B's misbehavior in my listening room, and that meant rock studio recordings featuring kickdrum. The B minor did better than the B with "Behind the Veil" from Jeff Beck's Guitar Shop (Epic EK 44313). When bass and kickdrum played together, bass notes remained more distinct than they had on the B. The drum kit had generous bass slam and snap without being overblown. The kickdrum on Richard Thompson's "She Misunderstood," from Rumor & Sigh (Capitol CDP 7 95713 2), showed decent pitch definition, but did not overpower the other instruments when played over the B minors.

Next I checked the reproduction of male voice. The B minor did not color the voices of FM announcers. (The Type B had produced overly resonant and barrel-like FM voicing.) José Carreras's wonderfully light, lyrical tenor remained pure and open with the B minor, showing little of the nasality heard with the B at the beginning of the "Kyria" on Ariel Ramirez's Misa Criolla (Philips 420 955-2). Harry Connick, Jr.'s rendition of "I Don't Get Around Much Anymore," from the When Harry Met Sally... soundtrack (Columbia CK 45319) was clear without any of the added warmth and overly resonant low frequencies heard with the B.

---

Footnote 1: See PWM's CES report (Vol.14 No.9, p.51); RH's review (Vol.14 No.12, p.130); my Follow-Up (Vol.15 No.2, p.181); and "Manufacturers' Comments" with JA's replies (Vol.14 No.12, p.263 and Vol.15 No.6, p.291).

Footnote 2: Many call this a "D'Appolito" configuration. However, designer Joe D'Appolito's original concept involved the use of 18dB/octave acoustical crossover slopes to get the optimal evenness of vertical lobing from the array. Not all vertical MTM drive-unit arrays are D'Appolito arrays.—John Atkinson

Footnote 3: Snell provided me with a copy of the Room-Analysis Computer Program, CARA, and LEO. These programs analyze room-resonance mode distribution and suggest speaker/listener locations for each dimension of the listening room that minimizes bass nodes. The B minors were positioned in the room according to one of LEO's suggestions for a "better" location ("best" could not be easily estimated).

Footnote 4: These frequency-response measurements are purely qualitative and do not represent the loudspeaker's true performance. However, they do yield a sense of the room-speaker interaction, and are used to confirm certain subjective impressions.