Audio Research VSi60 integrated amplifier Measurements
I examined the Audio Research VSi60's measured performance with 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); for some tests, I also used my vintage Audio Precision System One Dual Domain. I checked the bias of each output tube after letting the amplifier warm up for an hour and before I did any testing. All the tubes had exactly 60mV of bias, as specified, and didn't need adjustment with the top-panel potentiometers, which suggests good stability.
I had one mishap during the testing: I was measuring the amplifier's clipping power from the 4 ohm output transformer tap into 16 ohms and, just as the waveform squared, there was a loud bang. I immediately pulled out the AC lead and took the cover off. The 100 ohm wire-wound plate resistor for one of the left channel's 6550 output tubes had exploded. No fuses had blown! The other channel was okay, as was the same tube's 1 ohm cathode resistor. Audio Research sent me a replacement tube and resistor, which I installed (using lead-free solder supplied by ARC), though not to the same standard of neat soldering typical of the Minnesota company's products. The repaired left channel measured the same as the undamaged right channel, and offered the same signal/noise ratio as it had before the accident. I continued with the remaining tests, the harmonic and intermodulation spectra.
The VSi60 offered a maximum voltage gain into 8 ohms of 32.65dB from the 8 ohm tap, and 29.9dB from the 4 ohm tap. Though these gains are lower than usual for an integrated amplifier, in my opinion they're optimal for CD sources. The unity-gain setting of the volume control was between the third and fourth green LEDs on the front-panel thermometer display. The VSi60 preserved absolute polarity (ie, was non-inverting) through all the inputs and from both output taps. Its input impedance was usefully high, at 51k ohms, in the midrange and below, decreasing slightly, to 37k ohms, at 20kHz. The subwoofer output appeared to be full-range, meaning that it should be used with a subwoofer that has its own low-pass input filter.
As expected for an amplifier using a single pair of 6550 tubes per channel, the output impedance was moderately high in the bass and midrange, at 2.5 ohms from the 8 ohm tap and 1.6 ohms from the 4 ohm tap. The impedance rose at 20kHz, to 3 and 2.16 ohms, from the 8 and 4 ohm taps, respectively. This highish source impedance gave rise to ±1.4dB response variations from the 8 ohm tap into the magazine's standard simulated loudspeaker (fig.1, gray trace), while the increase in source impedance at high frequencies resulted in a premature top-octave rolloff of 2dB at 20kHz into 2 ohms (fig.1, magenta). The responses from the 4 ohm tap were similar, but with less variation in response due to different load impedances, as expected (fig.2).
Fig.1 Audio Research VSi60, 8 ohm tap, frequency response at 2.83V into: simulated loudspeaker load (gray), 16 ohms (green), 8 ohms (blue), 4 ohms (red), 2 ohms (magenta), volume control maximum. (0.25dB/vertical div.)
Fig.2 Audio Research VSi60, 4 ohm tap, frequency response at 2.83V into: simulated loudspeaker load (gray), 16 ohms (green), 8 ohms (blue), 4 ohms (red), 2 ohms (magenta), volume control maximum. (0.25dB/vertical div.)
If you look closely at the traces in fig.1, you can see a slight knee at 50kHz. This graph was taken with the volume control set to its maximum; at lower settings of the volume control, the knee became a little more pronounced (fig.3), which implies the existence of some sort of parasitic resonance, which in turn results in a couple of cycles of well-damped ringing on the tops and bottoms of a 10kHz squarewave (fig.4). The 4 ohm tap was better behaved in this respect. The amplifier's reproduction of a 1kHz squarewave was superbly square (fig.5).
Fig.3 Audio Research VSi60, 4 ohm tap, frequency response at 2.83V into 8 ohms with volume control at 12:00 (blue) and at its maximum (red). (0.25dB/vertical div.)
Fig.4 Audio Research VSi60, 8 ohm tap, small-signal 10kHz squarewave into 8 ohms.
Fig.5 Audio Research VSi60, 8 ohm tap, small-signal 1kHz squarewave into 8 ohms.
Channel separation (not shown) was excellent, at >100dB in both directions below 2kHz and still 78dB at 20kHz. The amplifier's noise floor, assessed with the input shorted but the volume control set to its maximum, was also fairly low in level. The unweighted, wideband S/N ratio (ref. 2.83V into 8 ohms) was a good 73.3 and 74.8dB for both channels from the 8 and 4 ohm taps, respectively. The audioband ratios were 5 and 3dB better, while A-weighting the measurement gave 89dB from the 8 ohm tap and 90dB from the 4 ohm tap.
Plotting the THD+noise percentage in the VSi60's output against power revealed that the distortion began to rise above the noise floor at levels above 12W from the 8 ohm tap (fig.6), and at levels above 500mW1W from the 4 ohm tap (fig.7). The linear increase above those levels before actual waveform clipping suggests that the circuit uses a low amount of loop negative feedback (Audio Research quotes just 7dB). ARC specifies the VSi60's maximum power at 1.5% THD. The amplifier exceeds that specification from the 8 ohm tap into 8 ohms, with 53Wpc available (17.25dBW) at 1% THD, though only 44W (13.4dBW) were available from the 4 ohm tap into 4 ohms at 1% THD. The amplifier did meet its 50W specification (14dBW) from the 4 ohm tap at 6% THD; however, the 0.6dB shortfall in power will be inconsequential.
Fig.6 Audio Research VSi60, 8 ohm tap, distortion (%) vs 1kHz continuous output power into (from bottom to top at 1W): 16, 8, 4 ohms.
Fig.7 Audio Research VSi60, 4 ohm tap, distortion (%) vs 1kHz continuous output power into (from bottom to top at 1W): 16, 8, 4, 2 ohms.
Provided the speaker load is equal to or greater than the nominal value of the VSi60's output transformer tap, the distortion at a level equivalent to 5W into 8 ohms remains below 0.1% in the midrange and below (fig.8, 8 ohm tap; fig.9, 4 ohm tap), though with a rise in THD in the upper audio octaves. The distortion is predominantly the subjectively innocuous low-order harmonics, even at moderate power levels (figs.10 and 11). However, fig.11 indicates that the amplifier's power supply is working hard, with the fundamental and third-order harmonics of the 60Hz AC power frequency almost as high in level as the distortion harmonics, particularly in the right channel (fig.11, red trace). The 120Hz component lies at 96dB in this graph (0.0015%), suggesting that the circuit benefits from optimal grounding.
Fig.8 Audio Research VSi60, 8 ohm tap, THD+N (%) vs frequency at 6.3V into: 16 ohms (green), 8 ohms (blue), 4 ohms (red), 2 ohms (magenta).
Fig.9 Audio Research VSi60, 4 ohm tap, THD+N (%) vs frequency at 3.9V into: 16 ohms (green), 8 ohms (blue), 4 ohms (red), 2 ohms (magenta).
Fig.10 Audio Research VSi60, 8 ohm tap, 1kHz waveform at 13.7W into 4 ohms (top), 0.175% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.11 Audio Research VSi60, 4 ohm tap, spectrum of 50Hz sinewave, DC10kHz, at 10W into 8 ohms (linear frequency scale).
These odd-order power-supply harmonics led to sidebands to either side of the two signal components in the spectrum of the VSi60's output while it reproduced an equal mix of 19 and 20kHz tones at a level close to visual clipping on the oscilloscope screen (fig.12). As anticipated from the circuit's decreasing linearity in the top audio octave, the 1kHz difference tone in this graph is moderately high in level, at 60dB (0.1%), and reaches 50dB (0.3%) at 19W peak from the 4 ohm tap into 4 ohms (not shown). The higher-order intermodulation products are all lower in level, however.
Fig.12 Audio Research VSi60, 8 ohm tap, HF intermodulation spectrum, DC24kHz, 19+20kHz at 25W peak into 8 ohms (linear frequency scale).
To a large extent, the Audio Research VSi60's measurements are much as I would have expected from a tube amplifier using a single pair of 6550 output tubes per channel, with low loop negative feedback, and offering the sizes of power supply and output transformers that are possible at its price. But it's nicely engineered for all that. Just make sure you use the appropriate output transformer tap for your loudspeakers.John Atkinson