Integrated Amp Reviews

I used my Audio Precision SYS2722 system (see the January 2008 "As We See It" and www.ap.com) to measure the Octave V 40 SE. After installing the four KT88 output tubes per the setup sheet, and before doing any measurements, I let the amplifier warm up for a morning. I then checked the bias voltage for the KT88s using the front-panel rotary switch and the LED display. All four LEDs illuminated green, which, according to the excellent owner's manual, means "Setting is OK." I took a complete set of measurements with the amplifier used alone, then repeated most of the testing with the auxiliary Black Box capacitor bank connected.

The maximum voltage gain into 8 ohms was the same with or without the Black Box outboard capacitance bank, at 38.75dB, and the V 40 SE preserved absolute polarity; ie, was non-inverting. The input impedance at low and middle frequencies was, at 61.5k ohms, usefully higher than the specified 50k ohms, and was still 51k ohms at the top of the audioband. The output impedance is specified as 2 ohms; however, the measured impedance was significantly higher than this, at 3.3 ohms at 20Hz, 4.3 ohms at 1kHz, and 3.8 ohms at 20kHz. I was surprised that these measured impedances were not lowered by the Black Box.

With these high output-impedance values, the variation in frequency response with our standard simulated loudspeaker, resulting from the Ohm's law interaction between the amplifier's output impedance and the manner in which the load impedance varies with frequency, reached ±2.1dB (fig.1, gray trace). The two channels matched very closely in the audioband into 8 ohms (left channel blue, right red), though the left channel rolled off a little earlier above 20kHz. The ultrasonic rolloff is very fast, and there is a sign of a slight peak developing between 30 and 50kHz in the right channel into lower impedances (magenta and green traces). This correlates with some critically damped overshoot—one half-cycle of overshoot and no ringing—with a 10kHz squarewave (fig.2).

Channel separation (not shown) was modest, at >65dB below 1kHz, and just 38dB in both directions at 20kHz. (The specification is 40dB.) I measure an integrated amplifier's signal/noise ratio in the worst-possible case, with its input shorted to ground but the volume control set to its maximum. The wideband, unweighted ratio under these conditions, ref. 2.83V or 1W into 8 ohms and with the Black Box connected, was 77.6dB in the left channel and 78.5dB in the right. The ratios improved to 79.9 and 84.3dB, respectively, when the measurement bandwidth was restricted to the audioband, and to 84.2 and 87dB when A-weighted. Spectral analysis of the V 40 SE's output while it drove a 1kHz tone at 1W into 8 ohms (fig.3) revealed that the left channel's slightly poorer noise performance was due to a higher level of 60Hz in that channel, though at –83dB (0.007%), this won't be audible.

The V 40 SE's maximum output power is specified at 40Wpc into 4 ohms (13dBW). Figs. 4 and 5 plot the THD+noise percentage in the amplifier's output against output power into 8 and 4 ohms, respectively. We define clipping as when the THD+N reaches 1%, and these two graphs indicate that, into both impedances, the Octave clips at 26.5W (14.2 and 11.2dBW, respectively). Adding the Black Box didn't increase the power at clipping. However, the V 40 SE reaches its specified power into 4 ohms at 1.12% THD+N, and the THD+N is still just 1.18% at 50Wpc into 4 ohms (14dBW), which is when the amplifier starts to hard clip (shown by the knee in the trace). Below 50W, the THD+N rises linearly with increasing power, this due to the low level of loop negative feedback used.

The distortion below 1W is low, especially into higher impedances. Fig.6 plots the percentage of THD+N against frequency at 2.83V into 8, 4, and 2 ohms. The distortion rises at the frequency extremes, as expected from the circuit topology, but only into 2 ohms (gray trace) does it become problematically high. Fortunately, the distortion content at low powers is predominantly the subjectively innocuous second harmonic, even into low impedances (fig.7), though it is joined at higher powers by some high-order harmonics (fig.8). High-frequency intermodulation distortion (fig.9) was what I'd expected from the Octave's THD-vs-frequency behavior, with the 1kHz difference component—resulting from an equal mix of 19 and 20kHz tones at 10Wpc into 8 ohms—reaching –50dB (0.3%). There is a peculiar rise in the power-supply–rated spuriae in the mid-treble in this graph, which was taken with the Black Box connected. Disconnecting the Black Box didn't change this behavior, nor did experimenting with all the various ways of grounding the amplifier to the Audio Precision test system.

Other than that anomaly, Octave's V 40 SE measured as I would expect from a traditional design that uses a pair of KT88 output tubes for each channel. And, like Octave's RE 290, a 100Wpc power amplifier favorably reviewed by Erick Lichte in our January 2013 issue, the V 40 SE is constructed to an impressively high standard.—John Atkinson