Sidebar 3: Measurements
I measured a different sample of the Eversolo DMP-A8 from that auditioned by RvB. The serial number was 800A805CC764, and my sample was fitted with the v1.2.80 firmware.
Looking first at the line-input performance with my Audio Precision SYS2722 system, the DMP-A8 preserved absolute polarity (ie, was noninverting) with both balanced and unbalanced inputs and outputs. With the volume control set to "0dB," the gain was –0.06dB from the balanced inputs to balanced outputs and from the unbalanced inputs to the unbalanced outputs. Setting the volume control to the maximum of "+10.0dB" increased the gain by 9.9dB. The control operated in accurate 0.5dB steps.
I measured a different sample of the Eversolo DMP-A8 from that auditioned by RvB. The serial number was 800A805CC764, and my sample was fitted with the v1.2.80 firmware.
Looking first at the line-input performance with my Audio Precision SYS2722 system, the DMP-A8 preserved absolute polarity (ie, was noninverting) with both balanced and unbalanced inputs and outputs. With the volume control set to "0dB," the gain was –0.06dB from the balanced inputs to balanced outputs and from the unbalanced inputs to the unbalanced outputs. Setting the volume control to the maximum of "+10.0dB" increased the gain by 9.9dB. The control operated in accurate 0.5dB steps.
Fig.1 Eversolo DMP-A8, balanced input/output, frequency response with volume control set to "0dB" at 1V into: 100k ohms (left channel blue, right red), 600 ohms (left green, right gray) (1dB/vertical div.).
The Eversolo's unbalanced input impedance was a relatively low 5k ohms at 20Hz and 1kHz and 3.5k ohms at 20kHz. The balanced input impedance was higher, at 18k ohms from 20Hz to 20kHz. The output impedance was usefully low across the audioband, at 101 ohms, balanced, and 51 ohms, unbalanced. The line preamplifier's frequency response was flat from 10Hz to 50kHz and down by just 2.1dB at 200kHz into both 100k ohms (fig.1, blue and red traces) and 600 ohms (green, gray traces). This graph was taken from the balanced outputs and with the volume control set to "0dB"; the very close channel matching was retained at higher and lower settings of the control and from the unbalanced outputs.
Fig.2 Eversolo DMP-A8, balanced input/output, spectrum of 1kHz sinewave, DC–1kHz, at 3V into 100k ohms with volume control set to "+10dB" (left channel blue, right red) (linear frequency scale).
The DMP-A8's channel separation (not shown) was superb, at >140dB in both directions below 3kHz and a still superb 128dB at the top of the audioband. The Eversolo offered extremely low noise from all its outputs, with no power-supply–related spuriae present even with the volume control set to the maximum (fig.2). The wideband, unweighted signal/noise ratio, measured at the balanced outputs with the unbalanced input shorted to ground and the volume control set to "+10.0dB," was a superb 90.9dB ref. the specified maximum output of 4V. Restricting the measurement bandwidth to the audioband increased these ratios by 14dB, while switching an A-weighting filter into circuit further improved the ratios to 107.5dB.
Fig.3 Eversolo DMP-A8, balanced output, THD+N (%) vs 1kHz output voltage into 100k ohms.
Fig.4 Eversolo DMP-A8, balanced output, THD+N (%) vs 1kHz output voltage into 600 ohms.
Fig.3 plots the percentage of THD+noise in the Eversolo's balanced output against the output voltage into 100k ohms. Actual distortion lies beneath the noisefloor below an output of 2V, and the clipping voltage, which we define as when the THD+N reaches 1%, occurred at 17V in balanced mode, 8.5V in single-ended mode. The balanced output's clipping voltage into the current-hungry 600 ohm load was only slightly lower than that into 100k ohms, at 11.1V (fig.4).
Fig.5 Eversolo DMP-A8, balanced output, THD+N (%) vs frequency at 4V into: 100k ohms (left channel blue, right red), 600 ohms (left green, right gray).
Fig.6 Eversolo DMP-A8, balanced output, spectrum of 50Hz sinewave, DC–1kHz, at 4V into 600 ohms (left channel blue, right red; linear frequency scale).
Fig.7 Eversolo DMP-A8, balanced output, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 4V into 600 ohms (left channel blue, right red; linear frequency scale).
To ensure that the THD+N reading was not dominated by noise, I measured how the DMP-A8's distortion changed with frequency at an output level of 4V. The THD+N percentage into 100k ohms was superbly low throughout the audioband into both 100k ohms (fig.5, blue and red traces) and 600 ohms (green, gray traces). The second harmonic was the highest in level (fig.6), but at just –120dB (0.0001%), even at 4V into 600 ohms, was negligible. Intermodulation distortion with an equal mix of 19 and 20kHz tones at the same peak voltage level into 600 ohms was similarly vanishingly low in level (fig.7).—John Atkinson
