Epos Acoustics, a small loudspeaker company owned by the larger British firm
Mordaunt-Short (which in turn is a member of the much larger TGI group of companies
that includes Tannoy, Goodmans, and KEF), was founded by designer Robin Marshall
in 1983. Epos (a Latin noun from the Greek, meaning an epic poem) currently
has two speakers in its line, the compact ES11 and the larger, 8-inch two-way
ES14, which has been the very successful, principal system of Epos since 1986.
These two systems are to be joined shortly by the ES25, a full-range floor-standing
model.
The ES11, which was added to the Epos line in 1990, is a small two-way vented sys tem utilizing a 6 1/2-inch woofer and a 1-inch dome tweeter. Both drivers are custom designed and manufactured by Epos, whose goals were to create an affordable, well-balanced design with high-end aspirations.
========
SPECS
Type: Two-way, vented-box, compact system.
Drivers: 6 1/2-in. cone woofer and 1-in. dome tweeter.
Frequency Response: 60 Hz to 20 kHz (tolerance not stated).
Sensitivity: 87 dB at 1 meter, 2.83 V rms applied.
Crossover: 6-dB/octave high-pass on tweeter only (frequency not stated).
Impedance: 8 ohms nominal.
Recommended Amplifier Power: 25 to 75 watts per channel.
Dimensions: 14 3/4 in. H x 8 3/8 in. W x 9 7/8 in. D (37.5 cm x 22 cm x 25 cm).
Weight 17.6 lbs. (8 kg) each.
Price: $895 per pair; available in black ash, walnut, or mahogany; foam grilles for earlier models, $50 per pair; speaker stands, $200 per pair.
Company Address: c/o Music Hall, 108 Station Rd., Great Neck, New York, 11023.
========
Among the ES11's novel design features, the most novel is the bass driver's frame, which is an integral part of the loudspeaker's front panel. Other features include an unusual cabinet assembly, an extremely simple crossover network (two parts!), and a bass driver that has a phase plug instead of a dust cap.
The frame of the ES11 woofer is, as I've just noted, an integral part of the front panel and is injection-molded in one piece. According to Epos, "This insures that the coupling between the bass driver and the baffle is precisely and consistently defined, unlike conventional systems where the coupling is influenced by the varying tightness of fixing screws." The molding is very stiff and inert, minimizing vibrations of the front panel. Presumably, replacing the woofer requires changing the whole front panel. The tweeter is separately mounted to the front panel. The rear panel, which contains the input terminals, crossover, and reflex port, is also an injection-molded part.
The front and rear panels are held in place by four long hex-head bolts, located near the cabinet's corners, which pass through the front panel and engage threaded inserts in the rear panel. Wood bracing is used internally for added strength. When assembled, and all four bolts tightened, the cabinet's 1-inch-thick medium-density fiberboard walls are sandwiched between the front and rear panels, forming a very strong and vibration-free structure.
The woofer's cone is thermo formed from a polymer material and shaped to optimize on-and off-axis response. The surround is a synthetic high-loss rubber that minimizes mechanical travelling waves on the cone's surface. The ES11's woofer has a phase plug protruding from the center of the cone rather than the more usual dust cap. The phase plug is stationary and is attached to the woofer's pole piece. The voice-coil and attached cone move fore and aft around the plug. The bullet-shaped phase plug is said to provide better polar response than is possible with a dust cap. The roll-off of the woofer's upper frequency has been carefully tailored so that the use of a low-pass crossover filter is not required.
The ES11 tweeter dome is fabricated from an aluminum alloy and is suspended by a polyamide material. The dome is acoustically loaded by a separate rear air chamber, while the voice-coil utilizes a magnetic fluid said to improve reliability and reduce dynamic compression.
The crossover of the ES11 is a "minimalist" design. It contains only two components, a high-quality series capacitor and a parallel resistor, which form the high-pass filter driving the tweeter. The simplicity of the crossover is said to enhance the loud speaker's ability to resolve fine detail, and the absence of a low-pass filter on the woofer allows better control of its motion by the power amplifier.
The ES1 l's input connections, which can be bi-wired, do not use the conventional double set of dual five-way binding posts with jumpers. The back panel contains only two sets of flush-mounted double-banana input holes. Single (i.e., not bi wire) connections are accommodated by the inclusion of two short banana-plug adaptor links, one for positive and the other for negative. Each link has a single banana plug on each end, one plain and the other containing a piggyback banana socket.
In either configuration, bare wire connection capability is not supported; banana plugs must be used! However, Epos does provide an extra set of double-banana plugs that can be attached to bare wires and then used to connect to the loudspeakers.
Measurements
The on-axis anechoic frequency response of the ES11 is shown in Fig. 1. Measurements were taken at 2 meters, halfway between the woofer and tweeter. With 5.66 V rms applied, the result was referenced back to 1 meter. A combination of elevated free-field and ground-plane measurements was used to derive the curve.
Fig. 1-One-meter, on-axis frequency response.
Fig. 2-On-axis phase response and group delay.
Fig. 3-Energy/time response.
Fig. 4-Horizontal off-axis frequency responses.
Fig. 5-Vertical off-axis frequency responses.
Fig. 6-Impedance.
Fig. 7-Complex impedance.
The overall curve in Fig. 1 is quite smooth, fitting a fairly tight 4.7-dB window (+1,-3.7 dB referenced to 1 kHz) from 100 Hz to 20 kHz. The bass response is down 3 dB (from the 100-Hz level) at 56 Hz and down 6 dB at 47 Hz. Be low 50 Hz, the response rolls off at 24 dB/octave, as is typical of vented boxes. There is a mild but broad peak centered at 800 Hz and a slight, downward shelf at high frequencies. Except for slight irregularities, the curve is quite smooth. The speaker's foam grille does not affect the response much at all. This system is one of few that will not be sonically compromised if listened to with the grille on. Averaged over the range from 250 Hz to 4 kHz, the sensitivity of the EP11 was 85.5 dB, 1.5 dB be low the 87-dB rating. The right and left speakers were matched within a close ±0.5 dB.
The phase and group-delay responses of the ES11, referenced to the tweeter's arrival time, are shown in Fig. 2. The phase curve is well behaved and rotates an additional 240° between 1 and 20 kHz.
The group-delay curve shows a fairly low offset of about 0.15 mS between the midrange and treble.
The deviations between 100 and 200 Hz are due to minimum-phase variations in the amplitude response and would disappear if the response were flat through this range.
The ES11's energy/time response is shown in Fig. 3. The test parameters accentuate the speaker's response between 1 and 10 kHz, which includes the crossover region. The main arrival, at 3 mS, is very compact but is followed by minor delayed responses, about 23 dB down from the main peak and extending 1.5 mS after the main arrival.
Figure 4 reveals the horizontal off-axis frequency responses; the bold curve at the rear of the graph is the on-axis response. The off-axis horizontal response is very uniform. In the primary (±15°) listening window, the response is extremely uniform, staying within ±1 dB of the on-axis curve all the way to 20 kHz.
The vertical off-axis frequency responses are displayed in Fig. 5; the bold curve in the center of the graph (front to rear) is on axis.
The aberrations in the range from 5 to 7 kHz indicate that the crossover frequency is at a high 6 kHz. Because the woofer and tweeter are separated by 5 3/8 inches (center to center), a significant 2.3 wavelengths at crossover, the vertical off-axis response in the crossover range is quite rough and narrow. The curves in Fig. 5 verify the narrowness of the vertical response at crossover and show that the response is significantly irregular at angles of only ±5°.
In Fig. 6, the ES11's impedance magnitude, a high minimum impedance of 7.2 ohms occurs at 250 Hz and a high maximum of about 29 ohms occurs at 85 Hz. The curve's maximum-to-minimum variation is about 4 to 1 (28.6 divided by 7.2). Even though this variation is fairly large, the high minimum impedance of the ES11 ensures that the speaker will not be very sensitive to cable resistance.
Cable series resistance should be limited to a maximum of about 0.11 ohm to keep cable-drop effects from causing response peaks and dips greater than typical run of about 10 feet, you can use low-inductance cable of 16 or 18 gauge.
The complex impedance, plotted from 5 Hz to 30 kHz in Fig. 7, is well behaved and exhibits no extraneous resonances. The impedance phase (not shown) reached a maximum angle of +45° (inductive) at 64 Hz and a minimum angle of-40° (capacitive) at 112 Hz. Even though these angles are fairly large, the ES11 will not be a problem for any amplifier (even for two in parallel), because the minimum impedance is quite high.
Fig. 8--Three-meter room response.
Fig. 9--Harmonic distortion for E1 (41.2 Hz.
Fig. 10--Harmonic distortion for B1 (61.7 Hz).
Fig. 11--Harmonic distortion for A2 (110 Hz.)
Fig. 12--IM distortion for A4 (440 Hz) and E1 (41.2 Hz).
Fig. 13--Maximum peak input power and sound output.
When subjected to a high-level sine-wave sweep, the cabinet of the ES 11 was mostly vibration-free. There were some minor wall resonances of the top and side panels in the 370 to 390 Hz range, and slight activity of the rear panel from 460 to 480 Hz. The linear travel of the woofer was about 0.4 inch, peak to peak, with reason able distortion; maximum travel was somewhat longer. The woofer overloaded quite gracefully. No dynamic offset was noted.
Minimum excursion occurred at 55 Hz, the frequency of the ES11's vented-box resonance. Cone displacement was reduced by about 50% after the port was closed. Vent noise at and near the box resonance was fairly low.
Figure 8 shows the three-meter room response, with both raw and sixth-octave smoothed data. The ES11 speaker was in 0.1 dB. For a the right-hand stereo position, aimed to ward the main listening position, and the test microphone was at ear height (36 inches), at the listener's spot on the sofa. The system was driven with a swept sine-wave signal of 2.83 V rms (corresponding to 1 watt into the rated 8-ohm impedance). The direct sound and 13 mS of the room's reverberation are included. If you exclude room-effect dips at 325 and 425 Hz, the averaged curve fits a tight, 7.5-dB window. Above 2.1 kHz, it fits an even tighter window of about 4 dB.
Figure 9 shows the E1 (41.2-Hz) bass harmonic distortion with input power ranging from 0.05 to 50 watts (note that 20 V rms generates 50 watts into the rated 8-ohm load). The second harmonic reaches a moderate level of 10.6%, while the third attains a very high 51%. Higher harmonics include an 8.1% fourth, a high 22% fifth, and a 3% sixth. With a 50-watt input, the ESI1 reaches a marginally usable 1-meter SPL of 90 dB at 41.2 Hz.
Because the E1 distortion was very high, a result of being significantly below the ES11's passband, the harmonic distortion of a higher frequency, B1 (61.5 Hz), was measured and is shown in Fig. 10. As before, the third harmonic predominates but only reaches a moderate 7% at full power. Other harmonics are all low, 0.6% or less.
The second harmonic, which reaches only 0.4%, is hidden behind the 61.5-Hz fundamental's bleed-through ridge at the left of the graph. With 50 watts in, the Epos reached a fairly usable 1-meter SPL of 98 dB at 61.5 Hz.
In Fig. 11, the bass harmonic distortion for A2 (110 Hz), the predominant distortion is a low 1.9% second harmonic and 2.4% third. Higher harmonics are quite low, 0.7% or less. The A4 (440-Hz) distortion (not shown) rose only to the low level of 3% second harmonic. Higher harmonics were below the noise floor of my measuring gear.
Figure 12 displays the IM distortion versus power, created by tones of 440 Hz (A4) and 41.2 Hz (E1) of equal level. The IM distortion rises to the fairly high level of nearly 19% at full power. The woofer handles both tones of this IM test, which contributes to this speaker's high level of intermodulation.
The ES11's short-term peak-power input and output capabilities are shown in Fig. 13. The peak input power was calculated by assuming that the measured peak voltage was applied across the rated 8-ohm impedance.
The peak input power rises from 10 watts at 20 Hz and, after minor undulations at 80 and 120 Hz, reaches a local maximum of about 1,500 watts at 250 Hz. After falling to 830 watts at 400 Hz (where the woofer exhibited a harsh buzzing sound), the peak in put power rises smoothly to a healthy 6,000 peak watts at frequencies above As can be seen in Fig. 13, the ES11's maximum peak output SPL with room gain rises very rapidly from an unusable 71 dB at 20 Hz to reach a very usable 107 to 110 dB between 65 and 160 Hz. After reaching a peak of 116.5 dB at 250 Hz and falling slightly to 115 dB at 400 Hz, the output rises into the healthy range of 120 to 123 dB above 600 Hz. A pair of ES11s, operating in unison in a typical listening room, can attain even higher levels in the bass range. The strong maximum output above 60 Hz suggests that these speakers can be used either by themselves or as satellite systems with a subwoofer.
Use and Listening Tests
After working with some large systems recently, it was nice to handle a pair of speakers that I could hold under each arm at the same time. The ESI is were supplied to me with sturdy metal stands, which are optional and have to be assembled. I chose to use my own metal stands, which are only about 1 inch taller and have spikes on the bottom.
The operating manual is brief but thorough and consists of four 4 1/4 x 5 1/2-inch pages. Topics include power requirements, stands, positioning, connections, breaking in, and warranty. Epos points out that the ES 11 was designed to be placed fairly close to the rear wall, within 20 to 30 cm (about 8 to 12 inches), for proper bass weight and definition. For some listening, I set the ES11s closer to the rear wall than I usually do, about a foot in front of the bookshelves which line the rear of my listening room. I did most of my listening, however, with the speakers in my customary positions, well away from the rear walls.
My review systems were sup plied in a very attractive walnut finish. Although Epos specifies that the finish is a veneer, it looked like solid wood; with the front panel removed, I could see what appeared to be unfinished walnut inside the cabinet. Construction and appearance were excellent.
When first received, one system had a slight air leak at a point on the front panel's periphery. Tightening the four long bolts that hold the front panel to the cabinet sealed the leak.
My listening equipment consisted of the Krell KRC preamp and KSA-250 amp driving the ES11s through Straight Wire Maestro cabling. My reference speakers were B & W 801 Matrix Series 3s, while Onkyo and Rotel CD players provided source material. Listening was done in the regular (not bi-wired) configuration.
First listening revealed the ES11s to have a well-balanced, smooth sound, with a touch of forwardness, and significantly less bass than the reference B & Ws. Sensitivity was essentially the same as that of the references, and the two systems produced a similar overall balance and tone (excluding the low bass).
Female vocals, such as Clair Marlo on Let It Go (Sheffield Lab CD-29) and Trisha Yearwood on The Song Remembers When (MCA MCAD-10911), were very natural.
There was no harshness, glare, or undue high-frequency emphasis. In fact, the over all high-frequency reproduction of the ES11 was quite similar to the 801's in level, smoothness, and extension.
On more dynamic material such as Bob Mintzer's jazz on One Music (dmp CD-488), the ES11s did quite well in handling high-level percussion transients and complex passages. The low-end kick and bass punch of the 801s was completely missing, however. Even though the low bass was quite attenuated, the ES11s still had enough bass to be satisfying. At very high levels on this disc, the ES11s did start sounding somewhat congested; some audible modulation of the mids could be heard when high-level bass was present.
I also did some listening using Velodyne's F1500R subwoofer as an adjunct to the ES11s. (The F1500R replaced my F1500; the "R" version includes a remote.
Having remote control of level and of bass on/off is a super addition!) The ES11s worked extremely well as satellites. With low bass added, the overall sound competed quite well with the 801s.
On the pink-noise stand-up/sit-down test, the ES11s exhibited significant tonal changes in the upper midrange when I stood up. Their spectral balance on pink noise was quite good but sounded slightly more forward than the 801s did and had significantly less bass. Smoothness, al though not quite up to the B & Ws' standards, was nevertheless quite good. On third-octave band-limited pink noise, the ES11s did not have any usable output in the 20-, 25-, and 31.5-Hz bands. Although the output was just barely usable at 40 Hz, it was quite usable at 50 Hz. At 63 Hz and above, the ESI is could generate sufficient levels of clean bass.
On relatively sedate classical music, such as Boccherini's Cello Concertos performed on period instruments (Sony Classical SK 53121), the ES11s were quite open sounding, albeit a shade forward in the presence range. Stereo focus and lateral imaging were exemplary. Coloration was quite low, maybe due in part to the rigidness of the enclosure.
In summary, the ES11s demonstrated quite good performance for their size and price. They should be seriously considered by anyone who desires small, high-performance loudspeakers that are also accurate, smooth, and good looking.
-D. B. Keele, Jr.
( Audio magazine, Aug. 1994)
Also see:
Acoustic Research 338 Speaker (Auricle, Nov. 1995)
= = = =