Solid State Preamp Reviews

I measured a different sample of the Naim NSC 222, serial number 519712, than that auditioned by Martin Colloms. Using my Audio Precision SYS2722 system, I performed a complete set of tests with the preamplifier connected to the NPX 300 power supply. (The NPX 300 was sitting on the floor, as far away from the preamplifier chassis as the two umbilical cables allowed.) I then repeated some of the tests with the NSC 222 powered by its internal supply.

I connected the Naim to my router with an Ethernet cable, and when I powered up the preamplifier, it asked to be paired with the remote control. After I did that, it asked which language, country, and time zone I preferred, then checked that its firmware was up to date. It was: The preamp was running v5.0.2.7936. The Roon app running on my iPad mini recognized three versions of the NSC 222: as a Roon Ready device; as a Chromecast streaming device; and as an AirPlay streaming device. I activated the Roon Ready version and used that, along with the Audio Precision's optical and coaxial S/PDIF outputs, to send test-tone files to the NSC 222.

The Naim's optical S/PDIF input accepted data sampled up to 96kHz, the coaxial S/ PDIF inputs accepted data sampled up to 192kHz. With Roon, the preamplifier accepted data sampled up to 384kHz. (I don't have any PCM files sampled at a rate higher than 384kHz.) The digital inputs preserved absolute polarity at the NSC 222's balanced, unbalanced, and headphone outputs. With the volume control set to its maximum, a 1kHz digital signal at –12dBFS resulted in an output level of 1.34V into 100k ohms from the balanced output and 3.156V from the unbalanced and headphone outputs. The balanced output impedance was a low 45 ohms at 1kHz and 20kHz, rising to 1149 ohms at 20Hz, presumably due to the presence of output coupling capacitors. The unbalanced output behaved similarly, with a source impedance of 23 ohms at 1kHz and 20kHz but 447 ohms at 20Hz. The headphone jack offered an appropriately low output impedance of 5.3 ohms, this value consistent across the audioband.

All the subsequent tests were performed at the balanced outputs. The NSC 222's impulse response with 44.1kHz data (fig.1) was a minimum-phase type, with all the ringing following the single sample at 0dBFS. With 44.1kHz-sampled white noise (fig.2, red and magenta traces), the NSC 222's filter response reached full stop-band suppression at 26kHz. An aliased image at 25kHz of a full-scale tone at 19.1kHz (blue and cyan traces) can be seen at –70dB, and the distortion harmonics of the 19.1kHz tone are all low in level, with the second harmonic the highest in level at –73dB (0.02%). The Naim's frequency response with data sampled at 44.1kHz, 96kHz, and 192kHz is shown in fig.3. The response with 192kHz data (blue and red traces) doesn't extend any higher in frequency than with data sampled at 96kHz (cyan and magenta traces), the responses at both rates reaching –6dB at 29kHz.

Channel separation was excellent, at >95dB below 1kHz in both directions, falling to 73dB at 20kHz. The reduction at the top of the audioband is presumably the result of capacitive coupling between the channels. Even with the volume control set to its maximum, the balanced low-frequency noisefloor with the NPX 300 power supply (fig.4) was very low in level, and supply-related spuriae were absent. Spuriae at 60Hz and its odd-order harmonics made an appearance with the NSC 222 powered by its own internal supply (fig.5). These will be due to magnetic interference from the power transformer, but they are still very low in level, at or below –115dB ref. 0dBFS.

Increasing the bit depth from 16 to 24 with a dithered 1kHz tone at –90dBFS lowered the noisefloor by 12dB (fig.6), meaning that the NSC 222 offers around 18 bits' of resolution. With undithered data representing a tone at exactly –90.31dBFS (not shown), the three DC voltage levels described by the data were well resolved, and the waveform was noisy but symmetrical. With undithered 24-bit data, the result was a somewhat noisy sinewave (also not shown).

At low frequencies (fig.7), the second harmonic was the highest in level, at –86dB (0.005%). The second-order intermodulation product at 1kHz with an equal mix of 19 and 20kHz tones with a peak level of –3dBFS lay at a very low –96dB (0.0015%, fig.8). The higher-order products at 18kHz and 21kHz were even lower in level and, as expected from fig.2, the aliased images of the primary tones can be seen in this graph, albeit at low levels.

The Naim's rejection of word-clock jitter with 16-bit J-Test data was superb. Even with optical S/PDIF data, all the odd-order harmonics of the LSB-level, low-frequency squarewave lay at the correct levels, indicated by the sloping green line in fig.9. Repeating this test with Roon data gave an identical result.

Turning to the single-ended analog line inputs, I looked at the behavior via the RCA jacks, as I don't have a suitable DIN connector. With the Naim's volume control set to its maximum, the voltage gain at 1kHz measured 8.15dB from the balanced and single-ended outputs and 15.6dB from the headphone output. All the outputs preserved absolute polarity. The line input impedance is specified as a high 47k ohms; I measured 53k ohms at 20Hz and 1kHz, 40.7k ohms at 20kHz.

As anticipated from the digital input measurements, the line input's frequency response rolled off above the audioband (fig.10), reaching –0.75dB at 25kHz, –3dB at 41kHz, and –6dB at 60kHz. The blue and red traces in this graph were taken into the high 100k ohms load; the increase in output impedance at low frequencies meant that the response into 600 ohms (green and gray traces) rolled off below 200Hz, reaching –3dB at 60Hz and –6dB at 32Hz. As Naim's NAP 250 power amplifier, which Martin Colloms reviewed in the November 2023 issue, offers an input impedance of 47k ohms at low frequencies, there will not be any issues with that pairing. Fig.10 was taken with the volume control set to its maximum, "100." Both the frequency response and the excellent channel matching were preserved at lower settings of the control and from the single-ended and headphone outputs. The volume control operated in approximate 0.4dB steps at settings of "80" and higher, 1dB steps at "60" and lower.

The line input's channel separation was excellent, at >100dB below 1kHz and still 73dB at 20kHz. With the NSC 222 powered by the NPX 300, the line inputs shorted to ground, and the volume control set to the maximum, the wideband, unweighted signal/noise ratio (ref. 1V, which is the NAP 250 amplifier's sensitivity for full power) measured 77.3dB in the left channel, 79.5dB in the right channel. Restricting the measurement bandwidth to 22kHz increased the ratios to an excellent 92dB and an A-weighting filter increased them further, to 95dB. Spectral analysis of the Naim's low-frequency noisefloor with it powered by the NPX 300 revealed that with the volume control set to its maximum (fig.11, blue and red traces), some low-level power supply–related spuriae were present. Other than the spurious tone at 60Hz, these disappeared with the control set to –20dB, the input level increased by the same amount (green and gray traces). These spuriae were present at slightly higher levels with the NSC 222 powered by its internal supply, but the highest in level, at 60Hz, still lay below –90dB ref. 1V.

Fig.12 shows how the THD+noise percentage in the balanced output changed as the output voltage into 100k ohms increased. The downward slope below 4V output is due to actual distortion lying below the noisefloor. The distortion starts to rise above that output level but is still very low, before actually clipping at a high 8V. The unbalanced and headphone outputs clipped at the same voltage into 100k ohms. The balanced clipping voltage into the punishing 600 ohms load was only slightly lower, at 7.3V!

Examined at 3V output, to make sure the THD+N reading was not dominated by random noise, the distortion into 100k ohms (fig.13, blue and red traces) was extremely low over most of the audioband, rising slightly in the top octave. Distortion was only slightly higher into 600 ohms (green and gray traces) but rose in the bass, presumably related to the rise in output impedance in this region. The distortion signature appears to be entirely second-harmonic (fig.14) and extremely low in level, at –106dB ref.1V into 100k ohms (0.0005%). Intermodulation distortion was vanishingly low, even into 600 ohms (fig.15).

To examine the performance of the Naim's MM-compatible phono input, I left the switch on the rear panel in the Default position and connected a wire from the Audio Precision's ground terminal to the grounding lug on the Naim's rear panel. This input preserved absolute polarity at all three output types. The input impedance measured 41k ohms at 20Hz and 1kHz, and 28.4k ohms at 20kHz. The maximum gain at 1kHz was an appropriate 49.2dB at the balanced and unbalanced outputs, 56.6dB at the headphone outputs. The phono input's RIAA correction (fig.16) was superbly accurate, with excellent channel matching. The lows rolled off slightly, reaching –2.3dB at 10Hz, and, like the line input, the highs rolled off above the audioband.

The phono input's channel separation was excellent, at >80dB in both directions below 1kHz and still 72dB at 20kHz. The wideband, unweighted S/N ratio ref. 1kHz at 5mV, assessed with the NPX 300 power supply, the inputs shorted to ground, and with the volume control set to the maximum, was a good 75dB in both channels. Restricting the measurement bandwidth to the audioband increased this ratio by 2dB, while inserting an A-weighting filter resulted in a ratio of 80.9dB. These ratios were all 6dB less with the preamplifier powered by its internal supply, but this is still a relatively low level of noise. With the external power supply, the phono input's noisefloor almost entirely comprised random noise, even with the volume control set to the maximum (fig.17). Odd-order supply-related spuriae are present with the preamplifier's internal supply (fig.18), but these all lie at or below –73dB ref. 1kHz at 5mV and will be lower in level at lower volume control settings.

The phono input's overload margins, ref. 1kHz at 5mV, were excellent, at 23dB at 20Hz and 1kHz and 25dB at 20kHz. The second was the highest-level harmonic in the phono input's distortion signature (fig.19), but at –96dB (0.0015%) is inconsequential. The levels of the intermodulation products with an equal mix of 19 and 20kHz tones were just as low (fig.20).

Naim's NSC 222 offers generally excellent measured performance via its digital inputs and its line and phono analog inputs. The rolloff above 20kHz in the frequency response is unusual but will not have audible consequences, and some earlier Naim amplification products behaved similarly (footnote 1). The only measurable difference I found when I replaced the external NPX 300 power supply with the preamplifier's internal power supply was the appearance of spuriae at 60Hz and its odd-order harmonics, which will be due to magnetic interference from the transformer, but these are still low in level.—John Atkinson