Solid State Power Amp Reviews

Wes Phillips mentioned at the start of this review that I dislike measuring amplifiers with a class-D output stage. This is because such designs can throw out enough ultrasonic noise that you can never be quite sure that what you're actually measuring is the input stage of your analyzer being driven into slew-rate limiting. (For more on this subject, see the measurements sidebar accompanying Michael Fremer's review in April of the class-D Yamaha MX-D1.) What I do, therefore, is to insert, when appropriate, a sixth-order low-pass filter set to 20kHz or 30kHz between the output of the amplifier and the analyzer.

Our sample of the D-100 had a voltage gain into 8 ohms of 26.36dB. It preserved absolute polarity (ie, was noninverting), and its input impedance measured 95k ohms over most of the audioband, dropping inconsequentially to 78k ohms at 20kHz. The amplifier's output impedance was a moderately low 0.16 ohm in the bass and midrange, rising slightly to 0.19 ohm at 20kHz. As a result, the modification of the amplifier's frequency response, due to the Ohm's Law interaction between this source impedance and that of the loudspeaker, will be very mild. With our standard simulated speaker load, this modification remained within ±0.1dB limits (fig.1, top trace at 2kHz), and the D-100's output doesn't drop by much into lower impedances. But you can see from this graph how the amplifier's ultrasonic output rolls off rapidly, presumably due to the low-pass filter necessary to minimize the HF noise produced by the switching output stage—except that CIA claim that there is no output filter!

Fig.1 Channel Islands Audio D-100, frequency response at 2.83V into (from top to bottom at 2kHz): simulated loudspeaker load, 8 ohms, 4 ohms (1dB/vertical div.).

Even so, a relatively high level of this noise is still present in the D-100's output, as can be seen in fig.2, which shows the amplifier's reproduction of a 10kHz squarewave. The tops of the waveform are overlaid by what appears to be ringing; with no signal present it is actually a pure tone at 412kHz that has an amplitude of 263mV. This is presumably the residue of the output-stage switching frequency and affects the measured signal/noise ratio. The unweighted wideband figure, ref. 1W into 8 ohms, was compromised by the ultrasonic content to just 21.9dB. Restricting the measurement bandwidth to the audioband increased the ratio to a respectable 76.3dB, while switching in an A-weighting filter further improved it to 80.5dB.

Fig.2 Channel Islands Audio D-100, small-signal 10kHz squarewave into 8 ohms.

This VHF noise was present during my measurements of the D-100's clipping power, as my active low-pass filter will not handle the high signal voltages concerned. The results of this test are shown in fig.3. The downward slope to the right of each trace up to 20W or so indicates that the THD+noise percentage is dominated by the ultrasonic noise. However, the traces above that power indicate that the amplifier is running out of headroom into each load, these ranging from 2 to 16 ohms. Taking clipping as being our usual 1% THD+N, the D-100 clips at 59W into 16 ohms (20.7dBW), 118W into 8 ohms (20.7dBW), 105.5W into 4 ohms (17.2dBW), and 55W into 2 ohms (11.4dBW). While it exceeds its specified power into 8 ohms by a useful 0.7dB, the D-100 apparently fails to meet its 4 ohm power spec by 2.2dB.

Fig.3 Channel Islands Audio D-100, distortion (%) vs 1kHz continuous output power into (from bottom to top at 1W): 16, 8, 4, and 2 ohms.

I examined the manner in which the amplifier's small-signal THD+N percentage changed with frequency using the 30kHz low-pass filter. The results are shown in fig.4, which is of necessity restricted to a 10kHz upper limit. (The distortion harmonics for frequencies higher than this increasingly fall above the filter's passband.) The bottom trace was taken into 8 ohms; it shows a distortion figure of around 0.07% in the bass through the low treble, which is respectable. The THD increases into low loads, however, with the 2 ohm figure exceeding 0.1%. (This measurement was taken after the amplifier had been running at 1/3-power for 60 minutes. When first turned on, the D-100's measured THD+N was around half this figure. The D-100 ran cold, by the way, even after the preconditioning period.)

Fig.4 Channel Islands Audio D-100, THD+N (%) vs frequency at 2.83V into (from bottom to top): 8, 4, and 2 ohms, with sixth-order low-pass filter at 30kHz in front of analyzer.

Of more subjective importance than the level of THD is its spectrum, and here again the ultrasonic switching noise raised its head. The lower trace in fig.5 shows the waveform of the distortion and noise at 1.2W into 8 ohms. The measured percentage figure of 0.085% is obviously invalid because of the switching noise visible in this graph. Repeating the measurement with the bandwidth ahead of the analyzer restricted to 20kHz reduced this figure to 0.076% and revealed that while the THD is primarily second-harmonic in nature, there are also some higher-order harmonics present (fig.6).

Fig.5 Channel Islands Audio D-100, 1kHz waveform at 1.2W into 8 ohms (top), 0.085% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).

Fig.6 Channel Islands Audio D-100, 1kHz waveform at 1.2W into 8 ohms (top), 0.076% THD+N; distortion and noise waveform with fundamental notched out, with sixth-order low-pass filter at 20kHz in front of analyzer (bottom, not to scale).

This can also be seen in fig.7, which shows the spectrum of the amplifier's output signal driving a low-frequency tone at 12W into 4 ohms, with the bandwidth ahead of the analyzer restricted to 30kHz. Increasing the power level to 66W, which necessitated removing the low-pass filter, brings up the higher-order harmonics more than it does the second and third harmonics (fig.8), which each remain around –70dB in level (0.03%). In general, the D-100 seems happier driving 8 ohms than 4 or 2 ohms.

Fig.7 Channel Islands Audio D-100, spectrum of 50Hz sinewave, DC–1kHz, at 12W into 4 ohms, with sixth-order low-pass filter at 30kHz in front of analyzer (linear frequency scale).

Fig.8 Channel Islands Audio D-100, spectrum of 50Hz sinewave, DC–1kHz, at 66W into 4 ohms (linear frequency scale).

Finally, the Channel Islands amplifier's performance on the high-level, high-frequency intermodulation test (fig.9) is okay without being outstanding. The 1kHz difference tone, resulting from an equal mix of 19kHz and 20kHz tones driven at 50W peak into 8 ohms, lies at –84dB (0.006%)—respectably low—but the higher-order products at 18kHz and 21kHz reach –70dB (0.03%), which is less good.

Fig.9 Channel Islands Audio D-100, HF intermodulation spectrum, DC–24kHz, 19+20kHz at 50W peak into 8 ohms (linear frequency scale).

Considering its relatively affordable price, the Channel Islands Audio D-100 offers good measured performance, though I would not use it with speakers whose impedance dipped below 4 ohms.—John Atkinson