Integrated Amp Reviews

All measurements of the Audio Refinement Integrated, except as noted, were made at a level-control setting that resulted in a 1W output into an 8 ohm load with an input of 100mV. The physical position of the level control for this condition was approximately 3:00, the voltage gain into 8 ohms 29.2dB at this setting.

The amplifier was very warm, but not hot, following its 1-hour, 1/3-power preconditioning test (conducted at 8 ohms). Its line input impedance measured 54.4k ohms. This decreased to 18k ohms at the level control's maximum position, and increased to 75.4k ohms at a 12:00 setting. All values are quite acceptable.

The AR Integrated's output impedance measured a maximum of 0.01 ohms at 1kHz and 20Hz, increasing to 0.2 ohms at 20kHz with an 8 ohm load. The output impedance at the tape outputs measured 25.9 ohms with a 25 ohm source impedance and 588.7 ohms with a 600 ohm source impedance—this indicating unbuffered tape outputs. The DC offset at the main outputs measured a negligible 0.1mV in the left channel and 2.2mV in the right. Signal/Noise Ratios (unweighted ref. 1W into 8 ohms) measured 72.7dB with a 22Hz–22kHz bandwidth, 71.1dB from 10Hz to 500kHz, and 77.8dB A-weighted. The Integrated is noninverting, a positive-going input resulting in a positive output. The gain at the maximum setting of the level control is 40.6dB, and the volume control tracking is excellent.

Fig.1 shows the frequency response of the Integrated at an output of 1W into 8 ohms; the response at 2W into 4 ohms (not shown) is virtually identical. The response into a simulated load (shown) differs a bit more, but the deviations are inconsequential. Most significant here is the high-frequency rolloff: nearly 3dB down at 20kHz. I would expect this amplifier to sound slightly soft. Fig.2 is the Integrated's output with a 10kHz squarewave input; the high-frequency rolloff of fig.1 is clearly visible in the waveform's long risetime. The 1kHz squarewave (not shown) is much better, with a slight rounding of the leading edges and no overshoot or ringing.

Fig.1 Audio Refinement Complete, frequency response into (from top to bottom): 1W into 8 ohms, 2W into 4 ohms, and 2.828V into simulated loudspeaker load (0.5dB/vertical div.).

Fig.2 Audio Refinement Complete, small-signal 10kHz squarewave into 8 ohms.

Fig.3 shows the Integrated's crosstalk, a solid though not exceptional result unlikely to have any audible consequences. The commonly seen increase at higher frequencies is typically due to interchannel capacitive coupling.

Fig.3 Audio Refinement Complete, crosstalk (from top to bottom): R–L, L–R (10dB/vertical div.).

The THD+noise percentage vs frequency results in fig.4 are good. The distortion is a bit higher than the best solid-state results we've seen, but typical of an amplifier incorporating only modest negative feedback. Fig.5 shows the THD+noise waveform of the Integrated with a 1kHz input and an output 2W into a 4 ohm load. While the third harmonic is dominant, there are clear indications of higher harmonics as well.

Fig.4 Audio Refinement Complete, THD+noise (%) vs frequency at (from top to bottom at 1kHz): 4W into 2 ohms, 2W into 4 ohms, 1W into 8 ohms, and 2.83V into simulated loudspeaker load.

Fig.5 Audio Refinement Complete, 1kHz waveform at 2W into 4 ohms (top), distortion and noise waveform with fundamental notched out (bottom, not to scale).

Fig.6 shows the Integrated's output spectrum driving a 50Hz sinewave at 40W into a 4 ohm load. While unremarkable, this is a reasonable result, with the highest-level artifact being the third harmonic (150Hz), at –52.2dB (about 0.25%). The intermodulation in the output caused by a combined 19+20kHz input at 40W into 4 ohms is shown in fig.7. The largest artifact here is at 18kHz (–48dB, or about 0.4%). The spectrum resulting from a 33.4W output into an 8 ohm load—the highest output attainable with this signal without signs of clipping—is very similar, and is not shown.

Fig.6 Audio Refinement Complete, spectrum of 50Hz sinewave, DC–1kHz, at 40W into 4 ohms (linear frequency scale).

Fig.7 Audio Refinement Complete, HF intermodulation spectrum, DC–22kHz, 19+20kHz at 40W into 4 ohms (linear frequency scale).

Fig.8 shows the Integrated's THD+noise percentage vs output power curves with one channel driven at 1kHz. The power supply, like those of most modestly powered integrated amplifiers, limits the available power into lower impedances to considerably less than the theoretical ideal of doubling with each halving of the load impedance. The discrete clipping measurements (1% THD+noise at 1kHz) for the Integrated are shown in Table 1.

Fig.8 Audio Refinement Complete, distortion (%) vs continuous output power into (from bottom to top): 8 ohms, 4 ohms, 2 ohms.

On toneburst testing using the Miller Audio Research Amplifier Profiler with just one channel driven, the Audio Refinement was capable of putting out a lot more power. Into 8 ohms with a 1kHz burst (10 cycles on, 40 cycles off), 87.9W was available into 8 ohms, increasing to 144W into 4 ohms. But, as can be seen from fig.9, the amplifier was current-limited into lower loads, 12.7W being available into 2 ohms and 44W into 1 ohm. The distortion level was much higher into these loads, even at low powers.

Fig.9 Audio Refinement Complete, distortion (%) vs burst output power into (from bottom to top): 8 ohms (black trace), 4 ohms (red), 2 ohms (blue) and 1 ohm (green).

The Audio Refinement Integrated was a respectable performer on the test bench, though nothing out of the ordinary for its modest cost—except, perhaps, for its high-frequency response, which rolls off more abruptly than is typical of modern, high-fidelity power amplifiers.—Thomas J. Norton

Table 1 Audio Refinement Integrated: Discrete Clipping levels
(1% THD+noise at 1kHz)