Sonic Impact Model TA2024 Super T power amplifier Measurements
I preconditioned the Sonic Impact Super T amplifier with both channels running at one-third maximum output power for an hour into 8 ohms, but, as expected with an amplifier with a high-efficiency switching output stage, the chassis was as cold at the end of that period as it had been at the start. The THD+noise percentage was 0.0911% at the start and 0.0916% at the end; ie, no significant change. The maximum voltage gain into 8 ohms was a little lower than usual, at 25.1dB, and the amplifier inverted signal polarity.
The input impedance depended both on frequency and on the setting of the volume control. With the latter set to its maximum, the input impedance measured a fairly low 7.8k ohms in the bass and midrange, dropping slightly to 6.5k ohms at 20kHz. With the volume control set to 12:00, these rose to 19k ohms at low and midrange frequencies and 13k ohms at high frequencies.
The output impedance was a moderately high 0.35 ohm at 20Hz and 1kHz, rising to a very high 5 ohms at 20kHz, the latter due, I imagine, to the series low-pass filter used to prevent ultrasonic switching noise from being radiated from the speaker wires. This filter must be quite severe in its action; there were only a few millivolts of ultrasonic noise present at the amplifier's output, compared, for example, with the several hundred mV of noise present in the PS Audio GCC-100's output. The wideband, unweighted signal/noise ratio (taken with the volume control at its maximum but the input shorted) was therefore only 58.2dB ref.2.83V into 8 ohms, though this improved to a more respectable 78dB with an unweighted audioband measurement, and to 87.5dB when A-weighted.
The series output filter, however, will interact with the load impedance, and this interaction can be seen in fig.1, which shows the Super T's frequency response with the amplifier driving resistive loads ranging from 2 to 8 ohms, as well as Stereophile's standard simulated loudspeaker. While the high-frequency response is drastically curtailed into 4 ohms and below—the output into 4 ohms is down by 2dB at 20kHz and a very audible 6.5dB into 2 ohms—it actually peaks by 1.5dB at 23kHz into 8 ohms. This gives rise to a noticeable overshoot in the Sonic Impact's reproduction of a 1kHz squarewave (fig.2), though the 10kHz squarewave confirms that there is no ringing, the overshoot being well-damped (fig.3). These graphs were all taken with the volume control wide open; there were no significant changes in response at lower settings.
Fig.1 Sonic Impact Super T, frequency response at 2.83V into (from top to bottom at 2kHz): simulated loudspeaker load, 8, 4, 2 ohms (0.5dB/vertical div., right channel dashed).
Fig.2 Sonic Impact Super T, small-signal 1kHz squarewave into 8 ohms.
Fig.3 Sonic Impact Super T, small-signal 10kHz squarewave into 8 ohms.
Channel separation was excellent at almost 90dB in both directions at 1kHz, this decreasing (due to the usual capacitive coupling) to a still-good 70dB (L–R) and 62dB (R–L) at 20kHz (not shown).
As with all amplifiers that use a switching output stage, examining the Sonic Impact's distortion performance was complicated by the presence of RF energy in its output. Even at the level of a few millivolts, unless this energy is filtered, you can never be sure that you're not actually measuring the interaction between the RF content and the test gear instead of the absolute performance of the amplifier under test (again, see the PS Audio review). Therefore, to ensure that I was measuring what I thought I was measuring, I repeated the THD-related measurements using an active sixth-order low-pass filter set to 30kHz.
Fig.4 plots the percentage of THD+N in the Sonic Impact's output against output power in watts, taken without the low-pass filter. The bottom two traces show the behavior with a 1kHz sinewave into 8 and 4 ohms. The amplifier just fails to meet its specified power into 8 ohms: instead of 6W (7.8dBW), I measured 5.5W (7.4dBW) at clipping (defined as 1% THD+N). The shortfall was a little worse into 4 ohms: 9.2W (6.6dBW) instead of 11W (7.4dBW). Into 2 ohms, the amplifier didn't clip but simply stopped at 13W (fig.4, top trace). These measurements didn't surprise me. The amplifier's power supply is a 12V, 3A wall wart, and even with the Super T's high-efficiency output stage, there just aren't enough volts and amps to be delivered into the load to generate higher powers than these.
Fig.4 Sonic Impact Super T, distortion (%)vs 1kHz continuous output power into (from bottom to top at 2W): 8, 4, 2 ohms.
Interestingly, the Super T clipped at lower powers when powered by the linear 12V Monolithic supply with which Wes Phillips experimented.
Peculiarly, the Super T is more linear at low frequencies into low impedances than it is into high ones (fig.5). Above 1kHz, however, the amplifier clearly has trouble maintaining its linearity into all loads. Looking at the distortion content without the low-pass filter engaged, what appears to be subjectively innocuous third-harmonic content is overlaid with spikes that coincide with the signal's zero-crossing points (fig.6). Switching in the low-pass filter smooths these spikes over (fig.7) but doesn't eliminate them, confirming that they are characteristic of the Sonic Impact, not the result of any unwanted interaction between the ultrasonic noise in the amplifier's output and my Audio Precision test set.
Fig.5 Sonic Impact Super T, THD+N (%)vs frequency at 2.83V into (from bottom to top): 8, 4, 2 ohms (right channel dashed).
Fig.6 Sonic Impact Super T, 1kHz waveform at 2W into 4 ohms (top), 0.108% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.7 Sonic Impact Super T, 1kHz waveform at 2W into 4 ohms (top), 0.108% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale), with 30kHz sixth-order low-pass filter.
What was interesting was that these spikes are not present with the Super T passing low frequencies. Fig.8, for example, shows a spectral analysis taken with the low-pass filter of the amplifier's output while it drove a 50Hz sinewave at relatively high power into 4 ohms. The second harmonic is the highest in level, at an excellent –83dB (0.007%), with the third and fifth harmonics lying at –90dB (0.003%). (These odd-order harmonics disappear if the load impedance is increased to 8 ohms, leaving just the benign second harmonic as the primary spurious component.) But with a 1kHz sinewave at the same power into the same load, again with the 30kHz low-pass filter in front of the analyzer's input (fig.9), the third is now the highest-level harmonic, at –66dB (0.05%), and some high-order components can be seen, correlating with the spikes seen in the waveform graph.
Fig.8 Sonic Impact Super T, spectrum of 50Hz sinewave, DC–1kHz, at 6.5W into 4 ohms (linear frequency scale).
Fig.9 Sonic Impact Super T, spectrum of 1kHz sinewave, DC–10kHz, at 6.5W into 4 ohms (linear frequency scale).
Finally, the relatively poor HF linearity seen in fig.5 results in somewhat disappointing performance with the amplifier driving an equal mix of 19kHz and 20kHz tones into 4 ohms, even at a power level below visible waveform clipping (fig.10).
Fig.10 Sonic Impact Super T, HF intermodulation spectrum, DC–24kHz, 19+20kHz at 3W peak into 4 ohms (linear frequency scale).
The Sonic Impact Super T amplifier's measured performance is impacted by its use of a small external power supply and by its disappointing high-frequency linearity. But at its very affordable price, it should provide acceptable performance when matched to high-sensitivity speakers.—John Atkinson