Solid State Power Amp Reviews

As is my usual practice, I ran the Classé Omega Omicron monoblock at 1/3-power into 8 ohms before performing any testing. At the end of that time, the black vertical heatsink area was way too hot to touch and the chassis temperature was around 55 degrees C. Soon after, the amplifier shut itself off and the Omega symbol on the front panel started flashing red. I waited a few minutes, unplugged the AC supply, and pressed the Power button. All was normal—the amplifier's protection circuit monitors temperature as well as overload—and I continued testing.

The Omicron was noninverting through both unbalanced and balanced inputs, the latter wired with pin 2 of the XLR jack hot. The input impedance at 1kHz was very high, well over the specified 100k ohms. The voltage gain was also high, at the specified 29dB into 8 ohms from both inputs.

The source impedance was low, at 0.13 ohm across most of the audioband, rising inconsequentially to 0.16 ohm at 20kHz. (Both figures include the series resistance of 6' of speaker cable.) As a result, any modification of the amplifier's frequency response due to the usual Ohm's Law interaction between its source impedance and the loudspeaker modulus of impedance will be small, around ±0.1dB with Stereophile's simulated speaker load (fig.1, top trace at 1kHz). This graph also shows the Classé's response into 8, 4, and 2 ohms; the bandwidth is widest into the highest impedance, with a –3dB point of 143kHz. This drops to 80kHz with a 2 ohm load, resulting in a slight (–0.3dB) depression at 20kHz. Nevertheless, the 10kHz squarewave response (fig.2) was nicely square, with short risetimes. The balanced responses are shown; the unbalanced input responses were identical.

Fig.1 Classé Omega Omicron, frequency response at 2.83V into (from top to bottom at 2kHz): simulated loudspeaker load, 8 ohms, 4 ohms, 2 ohms (0.5dB/vertical div.).

Fig.2 Classé Omega Omicron, small-signal 10kHz squarewave into 8 ohms.

The wideband, unweighted signal/noise ratio (ref. 1W into 8 ohms) was great rather than excellent, at 85dB, this improving to 94.7dB when A-weighted. However, bearing in mind both the higher-than-usual voltage gain and the enormous dynamic range (see later), these figures are actually better than might be expected. Fig.3 shows how the measured percentage of distortion and noise in the Omicron's output varies with output power into 8, 4, and 2 ohms. The amplifier may be rated at 300W but no fewer than 530W were available into 8 ohms, at our standard definition of clipping: 1% THD+N, with 1kW into 4 ohms. However, the amplifier's protection circuit cut in at 1075W into 2 ohms, protecting it from current overload, which is why the trace cuts off at that figure rather than showing the true clipping behavior.

Fig.3 Classé Omega Omicron, distortion (%) vs 1kHz continuous output power into (from bottom to top): 8 ohms, 4 ohms, 2 ohms.

The traces in fig.3 show that the distortion spuriae emerge from the amplifier's noise floor at a few watts output, so I ran the distortion vs frequency curves at a voltage equivalent to 10W into 8 ohms (fig.4). These show not only the expected rise in THD into lower impedances, but also a steady rise with increasing frequency above 200Hz, presumably due to the circuit's finite open-loop gain-bandwidth product. (As the frequency increases, less overall gain margin is available for negative feedback to do its work.) But even in the worst case, at 30kHz into 2 ohms, the distortion remains well below 0.1%.

Fig.4 Classé Omega Omicron, THD+N (%) vs frequency at 9V into (from bottom to top): 8 ohms, 4 ohms, 2 ohms.

Even at high powers into high impedances, the distortion is predominantly third-harmonic (fig.5). However, some crossover spikes start to appear into 2 ohms (fig.6). Even though the absolute level of the spuriae is low, they represent the addition of higher-order harmonics that will not be as sonically benign as the third-harmonic. But as I said, they are very low in level—in fig.7, taken at 200W into 4 ohms, all the higher harmonics lie at or below –100dB (0.001%)—which means they will most probably have no subjective effect.

Fig.5 Classé Omega Omicron, 1kHz waveform at 143W into 8 ohms (top), 0.0125% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).

Fig.6 Classé Omega Omicron, 1kHz waveform at 49W into 2 ohms (top), 0.01% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).

Fig.7 Classé Omega Omicron, spectrum of 50Hz sinewave, DC–1kHz, at 195W into 4 ohms (linear frequency scale).

One aspect of the Classé Omicron's measured behavior bothered me: Before the amplifier began to clip into lower impedances, it seemed to generate random noise. This can be seen in the clipping traces into 4 and 2 ohms (fig.3), where there is an apparent rise in THD+N before the actual "knee" that represents the onset of waveform clipping. It is also graphically illustrated by two graphs taken when I was measuring the amplifier's behavior when handling an equal mix of 19kHz and 20kHz tones at high powers. Fig.8 shows the spectrum of the Omicron's output as it drove this difficult signal at 410W peak into 8 ohms, a dB or so below the point where the waveform on my oscilloscope suffered from visible clipping. All the intermodulation products are low in level, with the 1kHz difference component lying at –96dB (0.0015%). However, when I halved the load impedance, not only did the unclipped output power drop, but as I approached the onset of visible clipping, the amplifier's noise floor rose in the low and middle treble regions (fig.9). Most unusual.

Fig.8 Classé Omega Omicron, HF intermodulation spectrum, DC–24kHz, 19+20kHz at 400W peak into 8 ohms (linear frequency scale).

Fig.9 Classé Omega Omicron, HF intermodulation spectrum, DC–24kHz, 19+20kHz at 500W peak into 4 ohms (linear frequency scale).

The Classé Omega Omicron offers exemplary measured performance into 8 ohms, but obviously starts to break a sweat into lower impedances. However, its enormous dynamic-range capability should mean that there will be no subjective shortfalls even at very high listening levels.—John Atkinson