Classé CT-M600 & CA-M600 monoblock power amplifiers Measurements
Sidebar 3: Measurements
I carried out a complete set of tests on the Classé CT-M600, repeating some of the tests on the CA-M600. (The CA-M600 basically performed identically, other than a very slightly wider small-signal bandwidth.) To perform the measurements, I used Stereophile's loan sample of the top-of-the-line Audio Precision SYS2722 system (see the January 2008 "As We See It" and www.ap.com). Before doing the testing, I ran the Classé CT-M600 at one-third its rated power for 60 minutes, which thermally is the worst case for an amplifier with a class-A/B output stage. At the end of that period, the chassis was warm, with a temperature of 90°F (32°C). Though the front-to-back fan was running fast, it was not particularly noisy, and the expelled air was warm rather than hot.
The voltage gain into 8 ohms was to specification at 29.0dB into 8 ohms for both balanced and unbalanced inputs. Both inputs preserved absolute polarity (ie, were non-inverting), the XLR jack being wired with pin 2 hot. The unbalanced input impedance at low and midrange frequencies was lower than specified, at 25.5k ohms, which will not be a liability; the balanced input impedance in the same frequency regions was 51.2k ohms. At 20kHz, however, the unbalanced input impedance had dropped to 13k ohms, the balanced to 20.5k ohms.
The CT-M600's output impedance was low, at 0.09 ohm (including 6' of speaker wire) over most of the audioband, rising very slightly at 20kHz to 0.12 ohm. As a result, the modification of the amplifier's frequency response due to the interaction between this impedance and that of our standard simulated loudspeaker was less than ±0.1dB (fig.1, green trace). The CT-M600 rolls off smoothly above the audioband, reaching 3dB at 80kHz, as specified. (The CA-M600's response was 3dB at 96kHz, almost as specified.) The amplifier's reproduction of a 10kHz squarewave thus has slightly rounded transients (fig.2), but is commendably free from any overshoot or ringing. The 1kHz squarewave response was essentially perfect (fig.3).
Fig.1 Classé CT-M600, frequency response at 2.83V into: simulated loudspeaker load (green), 8 ohms (blue), 4 ohms (magenta), 2 ohms (red). (0.25dB/vertical div.)
Fig.2 Classé CT-M600, small-signal 10kHz squarewave into 8 ohms.
Fig.3 Classé CT-M600, small-signal 1kHz squarewave into 8 ohms.
The Classé CT-M600 is among the quietest amplifiers I have measured. Its wideband, unweighted signal/noise ratio (with the input shorted and ref. 1W into 8 ohms) was a superb 79.6dB. This increased to 93.1dB when the measurement was restricted to the audioband, and 95.8dB when an A-weighting filter was switched into circuit. For reference, the respective S/N ratios for the Bryston 7B SST2 were 75.6dB, 92dB, and 95.7dBA; for the Ayre Acoustics MX-R, 79.6dB, 90dB, and 92.5dBA; for the Musical Fidelity Titan, 83.5dB, 93.4dB, and 96.1dBA; and for the MBL 9007, 90.9dB, 97.4dB, and 100.6dBA.
It is no coincidence that these are all very powerful amplifiers. And with the very high power, this low noise floor endows these amplifiers with an enormous capability for dynamic range. The Classé is no different. It clips at 700W into 8 ohms (see next paragraph), which is equivalent to 56V RMS. Its unweighted audioband S/N ratio of 92dB, ref. 1W into 8 ohms, is therefore equivalent to 117dB when referred to the power at clipping, which is within spitting distance of the specified 120dB. The CT-M600 joins that select group of amplifiers that can do justice to the demands of true high-resolution recordings.
Fig.4 plots the THD+noise percentage in the amplifier's output against power. The Classé easily exceeds its 600W into 8 ohms (27.8dBW) rating, clipping at 700W into 8 ohms (28.5dBW). (We define clipping as being when the THD+N reaches 1%.) It fell a little short into 4 ohms, however, clipping at 1100W (27.4dBW) rather than the specified 1200W (27.8dBW). This is probably because, with the large current draw, the AC wall voltage had dropped from 119 to 115V. (I don't keep the AC voltage constant during testing, feeling that allowing the wall voltage to be affected by the amplifier's demands for current more truly reflects the reality of its use.) The trace in fig.4 that shows how the amplifier performs into 2 ohms cuts off at 1200W (24.8dBW), below the level where it actually clips and equivalent to a sustained current of 24.5A. This is because the Classé's protection circuit kicked in at this point, turning the amplifier off and causing the front-panel LED to flash red. Removing, then reinstating the AC supply allowed the CT-M600 to be turned on again, apparently none the worse for wear.
Fig.4 Classé CT-M600, distortion (%) vs 1kHz continuous output power into (from bottom to top at 300W): 8, 4, 2 ohms.
The shape of the traces in fig.4 suggests that the actual distortion starts to rise from the noise floor around 10W or so. I therefore plotted the manner in which the THD+N percentage changed with frequency at 20V, to be sure of looking at true distortion. The results are shown in fig.5. The CT-M600 offers superbly low distortion into 8 ohms over almost the entire audioband, with only a slight rise apparent above 5kHz. This is obviously a circuit with a wide open-loop bandwidth. The THD does increase as the load impedance decreases. However, the only harmonic present to any significant degree is the subjectively innocuous third (fig.6). (Note that I averaged 64 readings to generate this graph, in order to let the harmonic spuriae emerge from the noise floor, footnote 1.)
Fig.5 Classé CT-M600, THD+N (%) vs frequency at 20V into: 8 ohms (blue), 4 ohms (magenta), 2 ohms (red).
Fig.6 Classé CT-M600, 1kHz waveform at 25W into 4 ohms (top), 0.0013% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Spectral analysis reveals that there is some second harmonic present at a lower level (fig.7). You can also see that with the FFT length chosen, the higher-frequency noise components are at or below 140dB in this graph. Increasing the signal level to just below visible waveform clipping on the oscilloscope raised the level of the second harmonic by 6dB and the third by 20dB, and tiny amounts of the fourth and fifth harmonics can now be seen (fig.8). But that this is still an extraordinarily linear power amplifier is also evident in the spectrum of its output while it reproduces an equal mix of 19 and 20kHz tones at a power level close to visible waveform clipping into 4 ohms (fig.9). This is a maximally stressful test for an amplifier, yet the Classé produced just 0.001% of the 1kHz difference product. Even the higher-order components at 18 and 21kHz were more than 82dB down from full scale. (Ignore the little blip in the noise floor around 14kHz in this graph, which seems to be a measurement artifact.)
Fig.7 Classé CT-M600, spectrum of 50Hz sinewave, DC10kHz, at 50W into 8 ohms (linear frequency scale).
Fig.8 Classé CT-M600, spectrum of 50Hz sinewave, DC10kHz, at 395W into 8 ohms (linear frequency scale).
Fig.9 Classé CT-M600, HF intermodulation spectrum, DC24kHz, 19+20kHz at 650W peak into 4 ohms (linear frequency scale).
The Classé CA-M600 and CT-M600 offer superb measured performance. It doesn't get any better than this.John Atkinson
Footnote 1: Each doubling of the number of data captures increases the uncorrelated noise level by 3dB but the correlated distortion by 6dB, thus dropping the noise contribution by 3dB. (For the averaging, I trigger the 'scope with the unfiltered waveform so that it starts each capture at exactly the same point in time.)John Atkinson