Sidebar 4: Measurements
I performed a complete set of measurements on one of the Buckeye Purifi 1ET9040BA amplifiers. I primarily used my Audio Precision SYS2722 system, repeating some of the testing with the magazine's higher-performance APx555 system. Because class-D amplifiers emit relatively high levels of ultrasonic noise that would drive my analyzer's input into slew-rate limiting, I inserted an Audio Precision AUX-0025 passive low-pass filter between the test load and the analyzers. This filter mitigates noise above 80kHz and eliminates noise above 200kHz that would otherwise drive the SYS2722's input circuitry into slew-rate limiting. Without the filter, there was 197mV of ultrasonic noise with a center frequency of 719kHz present in the amplifier's output. I used the Audio Precision filter for all the tests other than frequency response. As the Buckeye is a class-D design, there was no need to precondition it before the testing. Nevertheless, out of habit I operated the amplifier for 30 minutes at a moderate power into 8 ohms before starting the testing.
That aside, the Buckeye Purifi 1ET9040BA offers superb measured performance, with high power combined with vanishingly low levels of noise and distortion.—John Atkinson
Fig.1 Buckeye Eigentakt 1ET9040BA, frequency response at 2.83V into: simulated loudspeaker load (gray), 8 ohms (blue), 4 ohms (magenta), and 2 ohms (red) (1dB/vertical div.).
Fig.2 Buckeye Eigentakt 1ET9040BA, small-signal 10kHz squarewave into 8 ohms.
The Buckeye preserved absolute polarity, the balanced input being wired with pin 2 hot, the AES standard. The voltage gain at 1kHz into 8 ohms was very close to the specified figures, at 25.6dB, High setting, 20.5dB, Medium, and 15.5dB, Low. The input impedance is specified as 51k ohms. However, I measured 8.4k ohms from 20Hz to 20kHz. This was not affected by the gain setting.
The output impedance was extremely low, at 0.006 ohms at 20Hz and 1kHz, rising very slightly to 0.0077 ohms at 20kHz. The variation in the amplifier's frequency response with our standard simulated loudspeaker (fig.1, gray trace) was therefore negligible. The response into resistive loads was flat in the audioband but rolled off sharply above 30kHz, reaching –3dB at 90kHz. The Buckeye's reproduction of a 10kHz squarewave into 8 ohms featured short risetimes (fig.2), with a small, critically damped overshoot.
Fig.3 Buckeye Eigentakt 1ET9040BA, spectrum of 1kHz sinewave, DC–1kHz, at 1Wpc into 8 ohms (linear frequency scale).
The amplifier's unweighted, wideband signal/noise ratio, taken with the input shorted to ground, was an excellent 83.6dB ref. 1W into 8 ohms in both channels. This ratio improved to 101dB when the measurement bandwidth was restricted to the audioband, and to 104dB when A-weighted. Spectral analysis of the low-frequency noisefloor while the Buckeye drove a 1kHz tone at 1W into 8 ohms revealed that while power supply–related spuriae at 60Hz and its the odd-order harmonics were present, these all lay at or below –120dB (fig.3).
Fig.4 Buckeye Eigentakt 1ET9040BA, distortion (%) vs 1kHz continuous output power into 8 ohms.
Fig.5 Buckeye Eigentakt 1ET9040BA, distortion (%) vs 1kHz continuous output power into 4 ohms.
Fig.6 Buckeye Eigentakt 1ET9040BA, distortion (%) vs 1kHz continuous output power into 2 ohms.
Buckeye specifies the Purifi 1ET9040BA's maximum power as 375W into 8 ohms and 750W into 4 ohms, both equivalent to 25.75dBW, and 1200W into 2 ohms (24.8dBW), all at 0.1% THD. We define clipping as when the THD+noise reaches 1%. However, I couldn't discover the Buckeye's clipping powers with a 1kHz signal as the amplifier went into protection, with the LED on the front panel turning from blue to red, before the THD reached 1%. Into 8 ohms (fig.4), the amplifier went into protection at 290W (24.6dBW). The THD+N at this power was just 0.0009%. Into 4 ohms (fig.5), it did so at 450W (23.5dBW) with 0.01% THD+N, while into 2 ohms (fig.6), it turned off at 660W (22.2dBW). In each case, I waited a couple of minutes before turning the amplifier off, then back on, with the rear-panel switch.
The FTC's updated "Amplifier Rule" states that maximum power should also be assessed at frequencies other than 1kHz. I therefore repeated the maximum power test with a 20kHz signal. The THD+N was higher at low powers than it had been with the 1kHz signal and reached 1% at just 74W into 8 ohms (18.7dBW; not shown).
Fig.7 Buckeye Eigentakt 1ET9040BA, THD+N (%) vs frequency at 20V into: 8 ohms (blue), 4 ohms (magenta), and 2 ohms (red).
In figs.4–6, actual distortion was buried beneath the noisefloor below 50W or so. I examined how the THD+N percentage varied with frequency at 20V, which is equivalent to 50W into 8 ohms, 100W into 4 ohms, and 200W into 2 ohms (fig.7). The THD+N was extremely low in the bass, midrange, and mid-treble into all three loads, but rose sharply above 5kHz. This is why the amplifier clipped at a much lower power with a 20kHz signal than it did with a 1kHz signal.
Fig.8 Buckeye Eigentakt 1ET9040BA, 1kHz waveform at 140W into 8 ohms, 0.00025% THD+N (top); distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.9 Buckeye Eigentakt 1ET9040BA, spectrum of 50Hz sinewave, DC–1kHz, at 100W into 8 ohms (frequency scale).
Fig.10 Buckeye Eigentakt 1ET9040BA, spectrum of 50Hz sinewave, DC–10kHz, at 100W into 8 ohms (linear frequency scale; note expanded vertical scale).
As well as the Audio Precision low-pass filter, I used a 20kHz "brickwall" low-pass filter to capture the distortion waveform, as without the additional filter all I could see was noise, even at 140W into 8 ohms. With the filter, the distortion at this power lay at just 0.00025% and was a mix of low-order harmonics (fig.8). These all lay at a negligible –120dB or below with 50Hz at 100W into 8 ohms (0.0001%, fig.9). Because these levels are close to the SYS2722's resolution limit, I repeated this test with the APx555. The result is shown in fig.10. Again, the low-order harmonics all lie at or below –127dB ref.100W into 8 ohms!
Fig.11 Buckeye Eigentakt 1ET9040BA, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 50W peak into 8 ohms (linear frequency scale).
Given the amplifier's reduced linearity in the top audio octave, I examined the spectrum with an equal mix of 19kHz and 20kHz tones at 50Wpc peak into 8 ohms. Fig.9 shows that while the 1kHz difference product lay at –120dB, the levels of the higher-order intermodulation products at 18kHz and 21kHz were much higher, at –70dB (0.03%). As with the possible exception of EDM, recorded music doesn't have high energy above 10kHz, the Buckeye's reduced linearity in this region will probably not have audible consequences.
