Manufacturer's Specifications:
System Type: Four-way, vented enclosure.
Drivers: One 10-in. woofer, two 4 1/2-in. cone midranges, two 1-in. dome tweeters.
Crossover Points: 300 Hz, 2.5 kHz, and 7 kHz.
Impedance: 5 ohms nominal, 4 ohms minimum.
Frequency Range: 26 Hz to 19.5 kHz.
Dimensions: 16 1/2 in. (41.9 cm) W x 14 1/2 in. (36.8 cm) D x 38 7/8 in. (98.7 cm) H.
Weight: 74 lbs. (33.3 kg).
Price: $1,000.00 per pair.
Company Address: 4129 North Port Washington Ave., Milwaukee, Wisc. 53212.
The "CM" of Koss Corporation's CM/1030 loudspeaker stands for "computer maximized." A 250-mm (10-inch) woofer has been combined, through computer analysis, with the optimum size, shape and configuration of enclosure to provide a solid bass response. By first selecting the variables, such as the number of bandpass ranges, the efficiency, and the cabinet size, Koss claims to have been able to specify an optimum design for the ports, drivers and crossover networks.
The CM/1030 is a four-bandpass loudspeaker system which, in addition to the woofer, uses two midrange drivers and two tweeters. The grille is easily removed to reveal three equalizer switches mounted on the front panel. These are marked "Midrange," "Treble," and "Tweeter" and allow for a +3 dB, " Normal," or -3 dB level adjustment in their respective bandpass ranges. A handy user's guide is provided immediately above the switches, in the form of a symbolic frequency-response plot with color-coded bandpass ranges (orange for midrange, green for treble and blue for tweeter.) This is an excellent visual guide and can be readily understood without need to refer to an operating manual.
The ease of adjustment and resettability of these controls could allow them to be used as minor touchup program equalizers, should that be felt necessary.
The cabinet is sturdy and heavy. Although tall enough to be ranked as a small tower system, its center of gravity is sufficiently low that the enclosure could not be readily tipped over by an inquisitive toddler.
Electrical connection is made to rear-mounted, push terminals placed in a recessed cavity along with a protective fuse. The terminals are clearly marked for polarity convention, and no difficulty should be encountered in hookup.
The entire enclosure is finished--front, sides, and rear. This allows the CM/1030 to be placed anywhere in a room without detracting from the decor.
Measurements
The electrical impedance of a loudspeaker is an indication of the load which the audio power amplifier must drive in order to provide high-quality sound. It represents not only the amount of energy that will be spent in work and heat, but the amount of energy which will be stored in the loudspeaker and will have to be controlled as it pushes back against the amplifier. One could liken this to pushing a child's swing--the job isn't over when the swing gets to its highest spot; it's going to come back when you've quit shoving. Some amplifiers do not like taking the loudspeaker's recoil in the teeth.
Fig. 1--Magnitude of impedance, all controls set to normal.
Fig. 2--Complex impedance plot, 10 Hz to 20 kHz.
Fig. 3--Expanded complex impedance plot, 75 Hz to 20 kHz.
Fig. 4--On-axis frequency response at 1 meter, driven by constant voltage
corresponding to 1 average watt into 4 ohms.
Fig. 5--On-axis phase vs. delay of 3.3608 mS and frequency response at (T)
the tweeter's air-path 1 meter, corrected for delay of 2.9909 mS. (M) the
midrange's air–path
Fig. 6--Three-meter room response.
The magnitude of impedance for the Koss CM/1030, plotted in Fig. 1, has a double peak in the bass region, which is characteristic of a vented system. The average value of impedance above 100 Hz is about 5 ohms, and it never drops below its rated 4-ohm value.
The complex impedance (resistance and reactance) plot of Fig. 2 reveals a possible amplifier problem in the lowest registers. The bass alignment for uniform sound pressure level at constant-voltage drive produces impedance resonances at 21 Hz and 41 Hz for this system. This is satisfactory for proper acoustic performance and indicates no problem in design. But the result is a large capacitive phase shift to the driving impedance, which reaches peak values at 26 and 55 Hz. The 63° lag at 55 Hz, where the impedance is 10 ohms, could cause some amplifiers to distort prematurely if driven near their maximum output levels at this frequency. I recommend using high-quality amplifiers with the CM/1030 to avoid this potential problem. The remainder of the frequency range, above 100 Hz, provides a very benign load to any amplifier. The number of impedance loops in the upper range causes some confusion in the plot of Fig. 2, which has been scaled to show the bass resonance peaks.
Figure 3 is an expanded plot of the complex impedance above 75 Hz.
The substantial impedance variations above 30 Hz suggest that not only low-resistance, but low-inductance hook up wire be used between the power amplifier and this speaker. As little as a half-ohm loop impedance could cause a change of several decibels in frequency response due to line drop.
The measured one-meter on-axis anechoic frequency response is plotted in Figs. 4 and 5. Since the Koss CM/1030 is a 4-ohm system, the drive level was adjusted to represent a constant 1-watt-average power level into 4-onms resistance throughout the frequency range. Midrange, treble and tweeter controls were set to their indicated normal positions, and the amplitude response for this setting is shown in Fig. 4. Sound pressure rises gradually at an average rate of about 2.5 dB per decade up to a cutoff of 19 kHz. Bass response, as expected, is good down to 35 Hz.
The sound pressure phase response is plotted in Fig. 5.
Since the acoustic positions of tweeter and midrange differ slightly, the corrected phase plot is shown in two parts. Both parts are corrected for absolute polarity in which the air-path delay is removed; 0° phase corresponds to a sound pressure increase which is precisely in phase with a positive-going electrical drive at the loudspeaker terminal marked with a plus sign. The midrange driver has a corrected time delay of 3.3608 mS for a one-meter air path, and the tweeter has a 2.9909-mS correction.
Acoustic crossovers occur at around 3 kHz and 330 Hz.
The tweeter phase shift is well-behaved and near the 0° polarity convention, while the midrange, although well-behaved, lies near-60°. Each driver is of minimum-phase type, although the composite response has non-minimum phase properties due to the arrival-time difference. Some mild emphasis of upper musical partials, with timbral irregularities in the range of Middle C, are implied by these plots.
The three-meter room response of Fig. 6 shows a much more uniform sound aspect than does the anechoic response. In this test, the Koss was placed in its recommended listening position and the measurement performed at a normal listener position of three meters in front of the speaker at a seated ear-level position of one meter above a carpeted floor. The first 13 mS of direct sound is captured, apodized (band-limited), and displayed using a time-delay spectrometer. The time gate of 13 mS is chosen from psychoacoustic literature and is believed to correspond to the interval of time necessary to establish timbre for transient tones. Thus, this is intended to measure how the Koss "sounds" in a normal listening environment.
Fig. 7-Horizontal polar energy response.
Fig. 8-Vertical polar energy response.
Fig. 9-Harmonic distortion for the musical tones of E, (41.2 Hz), A2 (110
Hz) and A4 (440-1z).
Fig. 10--IM distortion on when mixed in 1-to-1 262 Hz (Middle C) proportion,
produced by 41.2 Hz (E1)
Fig. 11--Energy-time response.
The Koss should sound pretty good, according to this measurement. The anechoic peaks and dips are nicely leveled, and a slightly off-axis listening position above the geometric axis actually tames the top-end bite. The upper measurement shown corresponds to a straight-on forward position, such as might be encountered when sitting directly in front of the speaker or when the speaker is rotated to point to the listening position. The lower measurement corresponds to a traditional, 30° off-axis, left-channel stereo speaker location. Since the response is more uniform in the straight-ahead position, I recommend rotating these speakers toward the normal listening position.
Uniformity of sound is also evident in the polar energy plots of Figs. 7 and 8. In these measurements, the energy from 20 Hz to 20 kHz is plotted as a function of angle relative to the front of the system. The horizontal angular energy response, Fig. 7, shows a straight-ahead projection of sound with a uniform fall-off of energy as one moves either to the left or right of center line. The vertical energy response shows some lobing due to time-delay interference between midrange and tweeter drivers. This also shows why the geometric on-axis amplitude response of the anechoic measurement is less uniform than the slightly higher position used for the three-meter test. This is an effect readily discerned if one moves up and down while listening to the CM/1030, since left-right motion produces little change in sound. The implication of these tests is that good stereo lateralization, with minimal influence on imaging by objects to the far side or back of the enclosure, can be expected.
The vertical energy response does show a substantial amount of sound being launched upward and downward.
The speaker should not, therefore, be placed on an acoustically hard surface, such as a hard wood floor, or immediately below overhead projecting objects, such as shelves. In normal listening positions, however, the measurement shows good sound quality.
Harmonic distortion for the musical tones of E1 (41.2 Hz), A2 (110 Hz) and A4 (440 Hz) is plotted in Fig. 9 Tones A2 and A4 remain very clean at all power levels, up to 100 average watts, while E1 is a bit higher in distortion than I would like to see for a speaker of this high quality. This measurement is a burst test, in which the test tone is applied for only a short duration and then removed while the acoustic distortion products are computed and plotted. The intent is to measure the acoustic distortion of brief sound surges of the type likely to be found in music. A sustained sine-wave measurement at any particular frequency, such as can be used to test an amplifier, would either fry the speaker or produce invalid data due to unnatural heat rise.
Second-harmonic distortion creates a false tone which is an octave above the test fundamental, while third-harmonic distortion creates a false tone which is a fifth in the octave above. Close, but not perfect, harmonic partials, such as found in piano, may tangle with these distortion fragments and produce unnatural timbral balance. The upper partial distortion fragments of the Koss stay acceptably low below 30 average watts, while the lower register E1 is moderately clean below 10 watts. This measurement indicates it can handle high sound levels in the middle and upper registers but shouldn't be driven too hard in the lowest registers.
Intermodulation distortion on Middle C (262 Hz) caused by the addition of E1 (41.2 Hz) at equal drive level is plotted in Fig. 10. Middle C was chosen for this test since A4 (440 Hz) is substantially carried by the midrange driver, and the intent of the test is to determine the distortion which the bass driver causes when asked to carry two important musical tones simultaneously. This, again, is a burst test, where the acoustic signal is captured and Fourier analyzed for all modulation sidebands on Middle C caused by E1.
Like the harmonic distortion, low-level IM is not as low as I would like to see for a system of this high quality. The lowest measured level was 0.315%, but the IM rises only gradually with increasing drive level. Phase measurements indicate that the IM is principally phase modulation throughout the range, corresponding to a vibrato on Middle C at a rate of E1. At 60 average watts, the modulation on Middle C is 5% peak-to-peak amplitude modulation and 13° peak-to-peak phase modulation. At the peak test level of 100 average watts, the amplitude modulation (tremolo) rose to 18%, while the phase modulation (vibrato) rose to only 15° peak-to-peak. The overall IM is quite low for the Koss.
Two other distortion measurements are made to uncover any unnatural musical dynamics which the loudspeaker might create. In the first test, the acoustic transfer gain linearity is measured from very low to very high sound levels. If a loudspeaker is perfectly linear, an increase of 1 dB in amplifier drive should produce a corresponding sound pressure level increase of precisely 1 dB. If it does not, then the tone is altered, even if it does not produce distortion fragments. In the second test called (with apologies to Bert Whyte) a crescendo test, I try to determine what happens to a solo instrumental tone when every other instrument in the orchestra lets go triple forte. If the speaker is perfect, the solo tone will not be altered in level or pitch by the presence of other non-musically-related instruments. I use a fixed tone for the solo instrument and 20-kHz bandlimited white noise for the other instruments, something like a loud cymbal crash.
The Koss survived the crescendo test quite well. There is a change of less than 0.1 dB in solo voice intensity when noise is added at a 20 dB higher rms level, even up to combined instantaneous peak levels of 400 watts for tones of A2 (110 Hz) and A4 (440 Hz). Middle C remained steady at less than 0.1 dB variation from perfection up to 200 watts peak, then dropped by 1 dB at 400 watts peak, a truly good performance.
Acoustic transfer gain, however, is not as good. Middle C (262 Hz) and A2 (110 Hz) remained perfect up to 1 average watt, but Middle C began to drop in gain above that level.
Middle C was 0.3 dB lower than its desired sound level at 10 watts and 0.8 dB lower at 60 watts average. The tone of A4 (440 Hz) was similar but dropped to 0.3 and 0.6 dB, respectively.
In summation, the Koss produces clean sound from the standpoint of intermodulation by bass notes, harmonic distortion and crescendo peaks, but tends to drop intensity level with increasing orchestral peaks. This amounts to a slight compression of dynamics.
The energy-time curve for the CM/1030 is plotted in Fig. 11. This is a test of the speaker's ability to reproduce a perfect impulse transient. The energy-time curve is the true envelope of the impulse response, plotted in decibels as a function of time. The first sound is due to the tweeter, commencing at 2.85 mS and cresting at 2.99 mS for a one-meter air path between the speaker and the microphone. A subsequent arrival at 3.08 mS is due to diffraction of the tweeter sound from the front of the enclosure. Midrange driver contribution begins at 3.25 mS and effectively sup presses the net sound to produce a low-level response after that time. Perfection would be a single hump at 2.9 mS with a -30 dB width on the order of 0.1 mS. The Koss energy-time curve is good, but not spectacular. This measurement indicates a good transient response with a moderate y small time spread imparted to crisp sounds.
Use and Listening Tests
The CM/1030s were placed near, but not against, a draped wall. This gave, to my ears, a better sound than when the speakers were placed either directly against a hard wall or away from the wall by more than a half meter.
The change was in the bass region; direct wall placement yielded dominant bass, while the opposite was noted when the speakers were well into the listening area.
My first impression was that the extreme top end was down in level, so I tried listening with the tweeter control in the +3 dB position. I finally decided that the best overall balance was obtained with the " Normal" switch positions and the speaker rotated toward the listening area. Removing the grille had no noticeable effect on the sound.
Piano and voice, both solo and chorus, are generally the most difficult sorts of program material and are he ones where a system's faults will show up most readily. On both these sources, the Koss CM/1030s performed with reason able credibility and presented a fairly accurate sound. While I was not able to say, "Ahh, this is the first real illusion of a piano," I was reasonably satisfied with the reproduction of this instrument. With vocals, both solo and massed in chorus, there was no lateral spreading, so that there was a good and stable stereo image. There were no serious problems or particular changes in timbre unless I moved widely about the listening area. However, articulation and crispness of vocals, particularly choral voices, were not fully satisfactory to my ears.
With the speakers positioned as I have mentioned, general timbral balance is good, and general orchestral balance seemed proper up through the range of 1 to 5 kHz. In addition, low bass is quite good and is properly balanced with the rest of the musical spectrum so long as the speaker is properly positioned in the room. There is some moderate spectral beaming evident so that the listening area having the best balance will not extend throughout a large room.
This system is capable of providing high levels of sound without audible distress. It's not a disco speaker and shouldn't be used at lease-breaking levels, but the CM/1030 does give a good account of itself on the loudest musical passages when using high-quality program material.
In general, I found this loudspeaker system to be quite listenable for extended periods of time, as it does not provoke listening fatigue. It does not provide the very best sound I've ever heard, but the CM/1030 is extremely good for this price range.
-Richard C. Heyser
(Source: Audio magazine, May 1983)
Also see:
Koss ESP/950 Earphones (Nov. 1992)
Klipschorn Loudspeaker (Nov. 1986)
= = = =