Solid State Power Amp Reviews

Sidebar 3: Measurements

I tested the Primare A35.8 with my Audio Precision SYS2722 system. As the A35.8 has class-D output stages, all measurements other than frequency response were taken with Audio Precision's auxiliary AUX-0025 passive low-pass filter, which mitigates noise above 80kHz and eliminates noise above 200kHz. (The relatively high levels of ultrasonic noise emitted by class-D amplifiers would otherwise drive the analyzer's input into slew-rate limiting.) Without the filter, 325mV of ultrasonic noise was present at each pair of loudspeaker terminals, with a center frequency of 396kHz. Because the AUX-0025 filter is a two-channel device, I could only measure two of the Primare's eight channels at a time. Almost all the measured performance was taken from channels 6 and 8, with the bridged-mode tests performed with channels 5 and 6 linked and channels 7 and 8 linked. I don't precondition class-D amplifiers before testing, but after an hour's operation, the top panel's temperature had stabilized at a warm 102.2°F (39°C).

As with Primare's two-channel A35.2 that Herb Reichert reviewed in May 2020, the A35.8's balanced inputs offered 6dB lower voltage gain into 8 ohms than the single-ended inputs, at 19.95dB compared with 25.95dB. (A balanced input usually offers 6dB higher gain than a single-ended input.) The voltage gains in bridged mode were identical, but they were exactly 6dB higher in "bridged +6" mode. The amplifier preserved absolute polarity (ie, was noninverting) with both balanced and unbalanced input signals in both normal and bridged modes. The balanced input impedance was close to the specified 36k ohms from 20Hz to 20kHz. The unbalanced input impedance, specified as 15k ohms, was 14.7k ohms at low and middle frequencies, dropping slightly and inconsequentially to 13.25k ohms at the top of the audioband.

Including the series impedance of 6' of spaced-pair loudspeaker cable, the A35.8's output impedance in normal mode was a very low 0.078 ohm at 20Hz and 1kHz, rising slightly to 0.092 ohm at 20kHz. Bridged mode increased these impedances only slightly. The modulation of the amplifier's frequency response, due to the Ohm's law interaction between this source impedance and the impedance of our standard simulated loudspeaker, was minimal, at ±0.1dB (fig.1, gray trace). The response into an 8 ohm resistive load (fig.1, blue and red traces), taken without the Audio Precision low-pass filter, was flat almost to 20kHz with then a rolloff of 1dB at 30kHz. Unusually, the ultrasonic rolloff was disturbed by resonant peaks at 60kHz and 100kHz. The Primare's reproduction of a 10kHz squarewave (fig.2), taken with the Audio Precision low-pass filter so that the waveform was not obscured by ultrasonic noise, was clean, without any ringing on the tops and bottom and with minimal overshoot.

Separation between adjacent channels was excellent, at >100dB in both directions below 1kHz, reducing to a still good 76dB at the top of the audioband. Measured with the unbalanced inputs shorted to ground and with the auxiliary low-pass filter in circuit, the amplifier's unweighted, wideband signal/noise ratio was a good 74dB ref. 1W into 8 ohms. This ratio improved to a superb 96dB when I restricted the measurement bandwidth to the audioband, and to 98.6dB when the reading was A-weighted. Low-level spuriae at the 60Hz power supply frequency and its harmonics were absent from the Primare's low-frequency noisefloor (fig.3).

Primare specifies the A35.8's maximum power as 150W into 8 ohms and 300W into 4 ohms, both equivalent to 21.76dBW. Using our definition of clipping, which is when the output's percentage of THD+noise reaches 1%, the Primare exceeded its specified power with two channels driven into 8 ohms, clipping at 200Wpc (23dBW, fig.4). With two channels driven into 4 ohms, the A35.8 clipped at the specified 300Wpc (fig.5). Again, with two channels operating but in bridged mode, the Primare clipped at 730W into 8 ohms (28.6dBW, fig.6), just below the specified 740W into this load. In bridged mode into 4 ohms, the clipping power was 1kW (27dBW, not shown) compared with the specified 750W.

I examined how the percentage of THD+noise changed with frequency in normal mode at 20V, which is equivalent to 50W into 8 ohms, 100W into 4 ohms, and 200W into 2 ohms. The THD+N was very low in the midrange into 8, 4, and 2 ohms (fig.7), but rose in the top octave. Peculiarly, with each channel driven separately, the THD+N was lower in the midrange and below into 2 ohms (gray trace) than it was into 4 ohms (cyan and magenta traces).

The A35.8's distortion at 50W into 8 ohms was very low and predominantly the third harmonic (fig.8), though bursts of HF noise coincide with the waveform peaks, even with the Audio Precision auxiliary filter in circuit. The third harmonic lay at a very low –108dB (0.0004%) in channel 8 (fig.9, blue trace), –103dB (0.0007%) in channel 6 (red trace), with the second harmonic close to 10dB lower in level in both channels. In bridged mode at the same power into 8 ohms (fig.10), the third harmonic lay at the same –103dB but the second and fifth harmonics were about 10dB higher in level than they had been in normal mode. With an equal mix of 19 and 20kHz tones at 100W into 4 ohms (fig.11), while high-order intermodulation products lay at or below –94dB (0.002%), the second-order difference product at 1kHz lay below –110dB (0.0003%). This spectrum was taken in normal mode. The intermodulation distortion was similar in bridged mode.

The Primare A35.8 offers high power, especially in bridged mode, with very low levels of distortion and audioband noise.—John Atkinson