Sidebar 3: Measurements
I measured the darTZeel NHB-108 model two with my Audio Precision SYS2722 system—see the January 2008 As We See It. I performed the measurements using the balanced (XLR) and single-ended (RCA) inputs and left the internal jumpers as they had been set. To avoid triggering the amplifier's protection circuitry, I turned it off before changing the inputs during the measurements. (The amplifier took a relatively long time to stop passing signal when turned off, which implies that the power supply has a lot of capacitance.) Before starting the testing, I followed the CEA's recommendation by preconditioning the amplifier for 30 minutes at one-eighth the maximum power into 8 ohms. At the end of that time, the side-mounted heatsinks were hot, at 113.3°F (45.2°C). After an additional 10 minutes at one-third power into 8 ohms, the heatsink temperature had risen to 121.6°F (49.8°C). The NHB-108 model two needs to be well-ventilated.
The darTZeel's voltage gain into 8 ohms was a low 20.15dB from the balanced inputs, but as with the NHB-468 monoblock that Michael Fremer reviewed in November 2019, the unbalanced gain was 6dB higher. The amplifier preserved absolute polarity (ie, was noninverting) for both XLR and RCA inputs. The balanced and unbalanced input impedances are specified as >30k ohms; I measured 20k ohms from 20Hz to 20kHz for the balanced inputs, which is still usefully high, and 24k ohms for the unbalanced inputs at 20Hz and 1kHz, 16k ohms at 20kHz.
Although the NHB-108 model two's output impedance is specified as 0.33 ohms, my estimate, including the series impedance of 6' of spaced-pair cable, was slightly higher, at 0.435 ohms at 20Hz, 0.455 ohms at 1kHz, and 0.465 ohms at 20kHz. Consequently, the frequency response with our standard simulated loudspeaker varied by ±0.26dB (fig.1, gray trace). The audioband response into impedances of 2 ohms and above was flat up to 20kHz, and the output into 8 ohms (fig.1, blue and red traces) was down by just 0.5dB at 200kHz. As with the other darTZeel amplifiers that have been reviewed in Stereophile, the NHB-108 model two has a wide small-signal bandwidth; its reproduction of a 10kHz squarewave into 8 ohms (fig.2) featured very short risetimes and no overshoot or ringing.
I then looked at the amplifier's output spectrum as I reduced the level in stages down to 1Wpc into 8 ohms. As you can see from fig.10, though the higher-order harmonics progressively disappear as the power is reduced, the level of the second harmonic remains constant at around –60dB (0.1%). Repeating this series of spectral analyses at different signal frequencies gave the same result. The NHB-108 model two's distortion signature is independent of input frequency and output voltage; while the level of the distortion products increases as the output current increase, this distortion signature doesn't change.
Fig.1 darTZeel NHB-108 model two, frequency response at 2.83V into: simulated loudspeaker (gray), 8 ohms (left channel blue, right red), 4 ohms (left cyan, right magenta), 2 ohms (green) (0.5dB/vertical div.).
Fig.2 darTZeel NHB-108 model two, small-signal, 10kHz squarewave into 8 ohms.
The darTZeel's wideband, unweighted S/N ratio, ref. 2.83V into 8 ohms and measured with the unbalanced input shorted to ground, was an excellent 86.7dB in the left channel and 81.4dB in the right. These ratios improved to a superb 99.4dB and 95.5dB, respectively, when the measurement bandwidth was restricted to the audioband, and to 101.8dB and 99.1dB when A-weighted. Spectral analysis of the darTZeel's noisefloor (fig.3) revealed that the spuriae related to the power-line frequency were all very low in level. The levels of these spuriae didn't change when I experimented with the grounding between the amplifier and analyzer.
Fig.3 darTZeel NHB-108 model two, spectrum of 1kHz sinewave, DC–1kHz, at 1Wpc into 8 ohms (linear frequency scale).
The NHB-108 model two's maximum power is specified as 150W into 8 ohms (21.8dBW) and 225W into 4 ohms (20.5dBW). With "clipping" defined as when the THD+noise reaches 1%, the darTZeel with both channels driven slightly exceeded its specified powers into both impedances. It clipped at 155Wpc into 8 ohms (21.9dBW, fig.4) and at 230Wpc into 4 ohms (20.6dBW, fig.5). The NHB-108 model two's maximum power is not specified into 2 ohms; I didn't test the amplifier's maximum power into that load, as I broke the original version when I attempted to do so.
Fig.4 darTZeel NHB-108 model two, distortion (%) vs 1kHz continuous output power into 8 ohms.
Fig.5 darTZeel NHB-108 model two, distortion (%) vs 1kHz continuous output power into 4 ohms.
I measured how the darTZeel amplifier's THD+N percentage varied with frequency at a fairly high level, 20V, which is equivalent to 50W into 8 ohms and 100W into 4 ohms. Though the distortion is relatively high in level into both impedances, fig.6 reveals that the NHB-108's distortion doesn't increase with frequency, which is commendable considering that the NHB-108 model two has no global negative feedback.
Fig.6 darTZeel NHB-108 model two, THD+N (%) vs frequency at 20V into 8 ohms (left channel blue, right red), 4 ohms (left cyan, right magenta).
The waveform of the residual distortion and noise with a 1kHz signal at 50W into 8 ohms (fig.7) is smooth, which English audio engineer Ben Duncan has suggested correlates with good sound quality. This graph suggests that the subjectively benign second harmonic is dominant, which was confirmed by spectral analysis with a 50Hz tone (fig.8), though higher-order harmonics are present at lower levels. Other than a proportional increase in the levels of the harmonics when I repeated this analysis at the same voltage into 4 ohms, the spectrum was identical (fig.9).
Fig.7 darTZeel NHB-108 model two, 1kHz waveform at 50W into 8 ohms, 0.118% THD+N (top); distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.8 darTZeel NHB-108 model two, spectrum of 50Hz sinewave, DC–1kHz, at 50Wpc into 8 ohms (linear frequency scale).
Fig.9 darTZeel NHB-108 model two, spectrum of 50Hz sinewave, DC–1kHz, at 100Wpc into 4 ohms (linear frequency scale).
Fig.10 darTZeel NHB-108 model two, spectrum of 50Hz sinewave, DC–1kHz, at 1Wpc into 8 ohms (linear frequency scale).
It was only when I performed my usual high-frequency intermodulation test with an equal mix of 19kHz and 20kHz tones that the NHB-108 model two stumbled a little. With a peak signal level of 50W into 8 ohms, the second-order difference product lay at –60dB (fig.11). Although many higher-order products can be seen, almost all of these lie at or below –80dB (0.01%).
Fig.11 darTZeel NHB-108 model two, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 50Wpc peak into 8 ohms (linear frequency scale).
The darTZeel NHB-108 model two's measured performance is very similar to that of the original version of the amplifier, which John Marks and Wes Phillips favorably reviewed in September 2003 and April 2005, respectively. Yes, the distortion is relatively high in level, which will be due to the designer's decision not to use any global negative feedback. But I keep coming back to the independence of the amplifier's transfer function from the input signal. I have long felt that this behavior is related to good sound quality. I will be interested to read what JVS thought of the darTZeel's sound.—John Atkinson
