Classé CA-200 stereo power amplifier Measurements
My 240V–rated example of the CA-200 happily delivered a neat 270Wpc into 8 ohms (24.3dBW) for program-related duty cycles driven by the UK's nominal 240V power (the actual voltage was 242V). Singly driven flat-out into 8 ohms continuous, the just-clip level was 250W (24dBW), a nice 50W reserve on the manufacturer's specification. The small loss (0.3dB) when both channels were operated together indicates a generously specified power transformer.
Protection systems precluded continuous measurement into low impedances and with difficult input frequencies; I used music-equivalent gated tonebursts to explore the CA-200's peak power and load tolerance. It could almost double its rated power into 4 ohms, my measurement showing 460Wpc (23.6dBW). There was some loss into 2 ohms (0.6dBW), with a true peak power of 808Wpc before clipping. I hadn't yet reached the CA-200's current limit, while 1 ohm served to assess that boundary with a 1.57kWpc capability as well as an associated, symmetrical peak current of 57A. In practice, this is sufficient for all but the very worst load combinations at the full power extreme.
This middle-sized Classé is a heavy hitter after all! An excellent power bandwidth was recorded—within 0.1 to 0.2dB of the 1kHz maximum when measured at the 20Hz and 20kHz limits, respectively.
Traditional levels of negative feedback endowed this amplifier with a classic combination of low distortion and low output impedance—negligible values of 0.02 to 0.03 ohm were seen over the entire audio bandwidth. The CA-200's innate character will therefore be essentially impervious to loading or level. This amp also does not have or need an internal stabilizing inductor in series with the speaker line. Depending on the linking cable, 2;uF of strongly reactive loading was disposed of without event, with and without an 8 ohm parallel resistance, though a plain 0.1;uF capacitive load would occasionally set off the protection.
Driving 0.1µF in parallel with 8 ohms, the squarewave response was excellent over the entire 20Hz–20kHz frequency range. The lack of tilt on the 20Hz waveform confirmed the extended low-frequency response: almost to DC. A mild 10% overshoot for 2µ;F was followed by the usual ringing and was quickly damped—no problems here.
The frequency response—perfectly flat in the audio range—is simply not worth reproducing here. The bandwidth is gently tailored at the audible treble boundary, –0.4dB at 20kHz, falling slowly and evenly to –3dB at 62kHz and –12dB by 200kHz. At low frequencies it extended below 1Hz. Channel separation was good under drive, measuring 54dB at 20kHz, 89dB at 1kHz, and 76dB at 20Hz. Channel balance was excellent at 0.02dB, 1kHz.
Sensitivity was almost on the button at 1.5V unbalanced/normal, with 100mV sufficing for 1W into 8 ohms. Again, the input impedance met specification at a high 75k ohms, an easy load for any preamplifier.
The audible noise at the speakers was very low, confirming the measured results relative to full power of –109dB CCIR (1kHz) ARM, –112dBA at 1W IHF, –86dB unweighted, and 88.5dBA.
Fig.1 shows the variation of distortion with frequency at a level of 100W into 8 ohms. This is technically sound behavior. This is not the place to discuss whether the THD should begin to increase from such a low breakpoint of 70Hz or not, moving from the –107dB region to –80dB by 5kHz. A relatively low open-loop bandwidth is one implication.
Fig.1 Classé CA-200, THD+noise vs frequency at 1W into 8 ohms.
Analyzing the distortion results, at full power (200W into 8 ohms) I got –80dB for 20Hz, a very low –91dB at 1kHz, and a still fine –62dB (0.08%) at 20kHz. At 100W the 20kHz figure changed little, but the 20Hz and 1kHz results improved by a further 10dB; ie, to –100dB, 0.001% for the midband. Down at 1W, potentially in the region where one half of the output devices crosses over to the other, the distortion figures, including system noise, were still very low, with –85dB of distortion noted at low and mid frequencies and –70dB for 20kHz (these figures dominated by measurement noise).
I attempted to analyze the harmonic spectrum at a cruising signal level of 200Hz at 1W into 8 ohms. Fig.2 shows some quite negligible 50Hz supply-ripple harmonics in which the harmonic series for distortion (first harmonic, 400Hz; second, 600Hz; etc.) were all well-buried. Under narrow-band analysis the actual distortion was significantly better than 110dB. Any discussion of the value of harmonic order is irrelevant at such low levels.
Fig.2 Classé CA-200, spectrum of 200Hz sinewave, DC–2kHz, at 1W into 8 ohms (linear frequency scale). Note the very low level of distortion harmonics.
The numbers were again excellent for high-frequency intermodulation distortion, the spuriae measuring –88dB at full power and –97dB at 1W—as perfect as you could wish for.—Martin Colloms