Spectrum 108A Speaker System Review

The Spectrum 108A is a small two-way speaker whose size and weight are consistent with placement on bookshelves or on free-standing pedestals. The walnut-grain, vinyl-covered box, which is 14-1/2 inches high, 11-1/2 inches wide, and 9 inches deep, weighs only 15 pounds.

The 108A’s nominally 8-inch woofer, operating in a vented enclosure, crosses over to a 1-1/2-inch cone tweeter at 2,300 Hz, the crossover network producing 12-dB-per-oc-tave slopes. The paper tweeter cone has a 1/2-inch-diameter aluminum center cap that is said to enhance its output at frequencies above 8,000 to 10,000 Hz. The black foam-plastic grille is permanently cemented to the front of the cabinet. In fact, we had to rip the grille off of one of our test units in order to locate the drivers, a necessity for positioning the test microphone accurately for close measurements. With the grille removed, we could see that the woofer port was in the upper left corner of the panel and that the tweeter was partially covered with a strip of aluminum screen, presumably to protect it from damage since the foam grille can easily be pushed in.

The Spectrum 108A has no user-accessible controls. Recessed into its back panel are two spring-loaded terminals, which accept the stripped ends of speaker wires, and a fuse holder—a rarity with low-priced speakers. A spare 1-ampere fuse is supplied for each speaker. Price: $195 per pair.

Lab Tests

For our tests, we mounted a pair of Spectrum 108A’s on pedestals about 16 inches off the floor and 24 inches from the wall behind them. Their smoothed room frequency response showed what appeared to be a dip in output at about 500 Hz. Since this dip did not appear in the quasi-anechoic response measurements made with our IQS FFT analyzer, it seems probable that it was a result of this particular placement. The effect was not audible as such in any of our listening tests.

The overall response variation was a moderate ± 4 dB through the middle and high ranges, with the minimum occurring at 10,000 Hz and an increasing output in the next higher octave. Low-frequency response was measured separately for the woofer and its port, using close microphone spacing in both cases. Combining their two response curves, with due allowance for the relative sizes of the woofer and the port, produced a bass-response curve with its maximum at 100 Hz. Bass response fell off at about 3 dB per octave at higher frequencies, at 6 dB per octave from 100 down to 50 Hz, and at 12 dB per octave below 50 Hz.

The FFT pulse-response measurements, at a 1-meter distance, showed an axial response that varied only ± 3 dB from 200 to 18,000 Hz. The phase response was equally good, with a group-delay variation of ±0.2 millisecond from 2,000 to 20.000 Hz. Since we had already removed the grille from one speaker, we compared the responses with and without the grille. That there was no significant difference up to our measurement limit of 30,000 Hz is a testimonial not only to the acoustic transparency of the Spectrum 108A’s grille but also to the response uniformity of the regular production-run units we tested.

The impedance of the Spectrum 108A had two maxima (this is characteristic of vented enclosures), 22 ohms at 32 Hz and 30 ohms at 100 Hz. Impedance was 8.5 ohms in the octave from 150 to 300 Hz, which would justify a nominal 8-ohm rating. After a broad rise to 16 ohms at 900 Hz, the impedance fell to its minimum of 6.5 ohms at about 4.000 Hz. The speaker’s sensitivity, measured at 1 meter, was a 91-dB sound-pressure level (SPL) with an input of 2.83 volts of random noise in a one-octave band centered at 1.000 Hz, only slightly less than the 92-dB SPL rating.

We used a drive level of 2.5 volts, corresponding to our own reference SPL of 90 dB, for bass harmonic-distortion measurements. Initially we tested the speaker with its grille in place, but we were disturbed to find an unusually large harmonic-distortion content (6 to 10 percent over the 100- to 50-Hz range) accompanied by audible buzzing from the grille. At that point we decided to remove the grille, and the measured distortion without it dropped from the previous excessive level to a very low one—about 1 percent down to 70 Hz and only 5 percent at 35 Hz. The other test unit did not produce excessive distortion with its grille in place, and we surmise that this effect in the first speaker was caused by some contact between the foam grille and the screen-wire protective cover over the tweeter. In any case, the port output is dominant at all frequencies below 100 Hz, so we plotted the port distortion also; the readings were not very different from those at the cone.

Finally, we measured the maximum short-term power the speaker could withstand at low, middle, and high frequencies without visible waveform distortion of its acoustic output. We used single-cycle tone bursts with a duty cycle of 1:128 or 1:64 for these tests in order to keep the average power to the speaker at a safely low value.

Despite the small size of the Spectrum 108A’s drivers, at each test frequency the amplifier’s output clipped or distorted before the speaker did. The corresponding clipping-power levels, with the actual impedance at each test frequency, were 200 watts into 30 ohms at 100 Hz, 400 watts into 16 ohms at 1,000 Hz, and 800 watts into 8 ohms at 10,000 Hz! While these figures might suggest that the 108A can be driven to high levels with impunity, or driven without restraint by any super-power amplifier, we would nevertheless suggest caution. Remember, we were using single-cycle bursts, a rather simple signal unlikely to be found in real musical program material.

Comments

We found that the Spectrum 108 A did not have the kind of full bass that one expects from larger speakers. In some ways this characteristic can be an advantage: the 108A did not impart any unnatural heaviness to the program, as so many other speakers, of all prices, do. Overall, the spectral balance of its sound was good, the slightly bright top end and reasonable bass output making it sound surprisingly similar to some much more expensive and better-known speakers with which we compared it. Not that we couldn’t hear differences, but the Spectrum 108A was refreshingly free of the common aberrations of low-priced speakers, such as peaky response, dull highs, excessive bass distortion, and the like.

Although we routinely listen to speakers through an amplifier capable of delivering several hundred watts per channel, we do not make a practice of pushing speakers to their limits except under test conditions. But even in our listening tests we had no qualms about driving the Spectrum 108A’s with power levels that might have been expected to reduce them to ashes, and we never even blew a fuse! We still suggest a modicum of caution, however, if you play these or any small speakers with a high-powered amplifier at unusually high volume levels.

The Spectrum 108A is a surprising little speaker, as our test results suggested. While it cannot perform sonic miracles, its frequency-response, power-handling, and distortion characteristics are certainly better than might be expected from a system of its size and price. This eminently listenable speaker is a worthy contender in its class. And though it will not outperform good speakers that are several times larger in size and a great deal higher in price, it will give them a run for their money.

Read all

Spectrum reviews

Read all