Details

Parent Category: Products

Category: Amplifier Measurements

Created: 15 June 2024

Link: reviewed by Roger Kanno on SoundStage! Hi-Fi on June 15, 2024

General information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The MXA-8400 was conditioned for 1 hour at 1/8th full rated power (~25W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The MXA-8400 has eight balanced inputs (XLR), and eight speaker-level outputs (Neutrik speakON). Each pair of inputs (e.g., 1&2 and 3&4, etc) can be independently configured as a single channel in bridge mode. The MXA-8400 was evaluated as a stereo amplifier using channels 1 and 2, as well as a two-channel bridged amplifier using channels 1/2 (bridged) and 3/4 (bridged). The MXA-8400 also offers a low gain mode (6Vrms sensitivity) and a more typical high gain mode (2Vrms sensitivity). Unless otherwise stated, the high gain mode was used.

Because the MXA-8400 uses a digital amplifier technology that exhibits considerable noise above 20kHz (see FFTs below), our typical input bandwidth filter setting of 10Hz-90kHz was necessarily changed to 10Hz-22.4kHz for all measurements, except for frequency response and for FFTs. In addition, THD versus frequency sweeps were limited to 6kHz to adequately capture the second and third signal harmonics with the restricted bandwidth setting.

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Lyngdorf for the MXA-8400 compared directly against our own. The published specifications are sourced from Lyngdorf’s website, either directly or from the manual available for download, or a combination thereof. With the exception of frequency response, where the Audio Precision bandwidth was extended to 500kHz, assume, unless otherwise stated, 10W into 8ohms and a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

Our primary measurements in two-channel mode revealed the following using the line-level balanced analog input (unless specified, assume a 1kHz sinewave at 440mVrms, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the MXA-8400 was able to sustain 548W into 4 ohms (~1.2% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (54.8W) for 5 seconds, for 5 continuous minutes without inducing a fault or the initiation of a protective circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the top of the MXA-8400 was only slightly warm to the touch.

Our primary measurements in bridge mode revealed the following using the line-level balanced analog input (unless specified, assume a 1kHz sinewave at 230mVrms, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

Frequency response (8-ohm loading)

In our frequency-response plots above, measured across the speaker outputs at 10W into 8 ohms, the MXA-8400 is essentially flat within the audioband. At the extremes the MXA-8400 is at 0dB at 5Hz and -3dB just past 60kHz. In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Phase response (8-ohm loading)

Above are the phase response plots from 20Hz to 20kHz for the line-level input, measured across the speaker outputs at 10W into 8 ohms. The MXA-8400 does not invert polarity and exhibits at worst, about 30 degrees (at 20kHz) of phase shift within the audioband.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8 ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 5Hz to 100kHz, in stereo mode. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Here we find that maximum deviation between no-load and a 4-ohm load is very small, at around 0.025dB. This is an indication of a very high damping factor, or low output impedance. With a real speaker, the deviations are smaller, at roughly 0.01dB.

THD ratio (unweighted) vs. frequency vs. output power (two-channel mode)

The chart above shows THD ratios at the output into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input in stereo mode. The blue and red plots are for left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange at the rated 200W. The 10W data yielded the lowest THD figures, ranging from 0.00004% from 20Hz to 200Hz, then up to 0.0005% at 6kHz. These are extraordinarily low THD ratios, nearing the limits of the APx555 analyzer. At 1W, THD ratios were more constant, from 0.0001% from 20Hz to 1kHz, then up to 0.0003% at 6kHz.  A 200W, THD ratios ranged from 0.00006% from 20Hz to 200Hz, then up to 0.0005% at 6kHz.

THD ratio (unweighted) vs. frequency vs. output power (bridge mode)

The chart above shows THD ratios at the output into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input in bridge mode. The blue and red plots are for left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange at the rated 450W.  The 1W data yielded the most constant results, ranging from 0.0002% from 20Hz to 2kHz, then down to 0.0001% at 3-4kHz. At 10W, THD ratios ranged from 0.00006% from 20Hz to 1kHz, then up to 0.0006% at 6kHz.  A 450W, THD ratios ranged from 0.00002/0.00003% from 20Hz to 100Hz, then a steady climb to 0.0004% at 6kHz.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms (two-channel mode)

The chart above shows THD ratios measured at the output of the MXA-8400 as a function of output power for the analog line level-input in two-channel mode, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). The 8-ohm data ranged from about 0.0003% at 50mW, down to 0.00007% from 10 to 50W, then up to the “knee” right around 200W. The 4-ohm data ranged from about 0.0006% at 50mW, down to 0.00007% from 5 to 100W, then up to the “knee” around 350W. The 1% THD marks were hit at 278W (8-ohm loading) and 540W (4-ohm loading).

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the MXA-8400 as a function of output power for the analog line level-input in stereo mode, for an 8-ohm load (blue/red for left/right) and a 4-ohm load (purple/green for left/right channels). THD+N values for the 8-ohm data ranged from 0.003% down to just below 0.0002% at 100-200W. The 4-ohm data yielded THD+N values 3-4dB higher.

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms (bridge mode)

The chart above shows THD ratios measured at the output of the MXA-8400 as a function of output power for the analog line-level-input in bridge mode for a 4-ohm load (blue/red for left/right channels). THD ratios were not measured into an 8-ohm load because our dummy-load configuration allows for a power handling of only 500W (but 1000W for 4 ohms). Maximum (1% THD) power into 8 ohms was measured in bridge mode with very short 1-second iterative measurements where a staggering 950W with two bridged channels driven was observed. The 4-ohm data ranged from about 0.002% at 300mW, down to 0.0001-0.0002% from 1.5 to 500W, then up to the “knee” around 700-800W. The 1% THD mark was hit at 1040W.

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only, two-channel mode)

The chart above shows THD ratios measured at the output of the MXA-8400 as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yields 50W at the output into 8 ohms (and roughly 100W into 4 ohms, and 200W into 2 ohms) for the balanced analog line-level input in two-channel mode. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. The 8 and 4-ohm THD data are nearly identical, ranging from 0.00002-0.00003% from 20Hz to 100Hz, then up to 0.0004% at 6kHz. The 2-ohm data ranged from 0.0001% to 0.003% across the audioband. It’s clear that these amplifier modules are optimized for 4- and 8-ohm loads. Nonetheless, they are stable into 2 ohms, yielding admirably low THD ratios.

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only, two-channel mode)

The chart above shows THD ratios measured at the output of the MXA-8400 as a function of frequency into two different loads (8/4 ohms) for a constant input voltage that yields 200W at the output into 8 ohms (and roughly 400W into 4 ohms) for the analog line-level input in bridged mode. THD ratios were essentially identical, ranging from 0.00004% at low frequencies, then up to 0.0004% at 4kHz.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the MXA-8400 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the balanced analog line-level input in stereo mode. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). The 8-ohm plot is fairly flat and between 0.0001% and 0.0003% from 20Hz to 6kHz. Between 1kHz and 6kHz, the THD ratios when real speakers were used as loads are identical to the dummy load.  The two-way speaker THD results were as high as 0.02% at 20Hz. Between 40Hz and 200Hz, the speaker THD results were roughly 10-20dB higher than that of the dummy load. While THD ratios remain low and below the threshold of audibility into real-world speaker loads, the NAD M23, using similar amplifier modules, performed better in this regard than the MX-8400.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the MXA-8400 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method is used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three IMD plots are within 10-15dB of one another, hovering between 0.0002% and 0.0008%. The IMD results for the real-world speaker loads can be seen both above an below the resistive dummy load results, depending on frequency.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the MXA-8400 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input in stereo mode. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three data sets are close enough to be judged as identical, hovering around the 0.001% level.

FFT spectrum – 1kHz (line-level input, two-channel mode)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input in two-channel mode. We see that the signal’s third (3kHz) and fifth (5kHz) harmonics are at -125dBrA, or 0.00006%, and -130dBrA, or 0.00003%, respectively. The remaining visible signal harmonics are below the -135dBrA, or 0.00002%, levels. These are extraordinarily low THD levels.  The power-supply-related noise peak at the fundamental (60Hz) frequency is barely seen at just above the -150dBrA, or 0.000003%, level; however, this peak is inherent to the AP’s signal generator. A rise in the noise floor can be seen above 20kHz, indicative of this type of digital amplifier technology. This is an exceptionally clean FFT result.

FFT spectrum – 1kHz (line-level input, two-channel mode, low gain)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input in two-channel mode, this time with the gain set to low. We see that the signal’s second (2kHz) and third (3kHz) harmonic, are nearing the absurdly low -140dBrA, or 0.00001%, level. As a point of comparison, below is a 1kHz FFT with the AP analyzer in loopback mode (generator internally feeds the analyzer) with the same 9Vrms signal amplitude. The only differences are that the overall noise floor, from uncorrelated thermal noise, is roughly 10dB higher with the MXA-8400 in the signal path (in the audioband), and the 2kHz signal harmonic peak is just below (instead of above) of -140dBrA, and the 3kHz peak is at -150dBrA instead of -140dBrA. In low-gain mode, in can be said that the MXA-8400 adds only a very small amount of uncorrelated noise (hiss), and from a THD perspective, is essentially perfectly transparent. In other words, the very definition of a “straight wire with gain.”

FFT spectrum – 1kHz (loopback, 9Vrms)

Shown above is the fast Fourier transform (FFT) for a 9Vrms 1kHz input sinewave stimulus, measured with the AP analyzer in loopback mode, for comparison to the above MXA-8400 FFT charts. The overall noise floor is around the -165dBrA level, and the only visible peaks are at 60Hz (-150dBrA), 120/240Hz (-155dBrA), 2kHz (-140dBrA), 3kHz (-150dBrA), and 4/6kHz (-160dBrA).

FFT spectrum – 1kHz (line-level input, bridge mode)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input in bridged mode (high-gain setting). We see that the signal’s third (3kHz) harmonic is at -130dBrA, or 0.00003%, while the second (2kHz) and fifth harmonics (5kHz) are even lower at -135dBrA, or 0.00002%. These THD levels are slightly lower than the already extraordinarily low levels in two-channel mode. The overall noise floor (from uncorrelated thermal noise) is a few dB higher in bridge mode (slightly above versus slightly below -150dBrA), however. This is an unavoidable consequence of using two amplifier modules instead of just one to drive the load.

FFT spectrum – 50Hz (line-level input, two-channel mode)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the analog line-level input in two-channel mode. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most dominant (non-signal) peaks are the second (100Hz) and third (150Hz) signal harmonics at roughly -145dBrA, or 0.000006%, and -140dBrA, or 0.00001%. THD levels are even lower at 50Hz than the already ultra-low levels at 1kHz. The power-supply-related noise peak at the fundamental (60Hz) frequency is evident at the extremely low -140dBrA, or 0.00001%, level. Another near-perfect FFT.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input, two-channel mode)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the analog line-level input in two-channel mode. The input RMS values are set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -125dBrA, or 0.00006%, and the third-order modulation products, at 17kHz and 20kHz, are at roughly -110dBrA, or 0.0003%.

Intermodulation distortion FFT (line-level input, APx 32 tone, two-channel mode)

Shown above is the FFT of the speaker-level output of the MXA-8400 with the APx 32-tone signal applied to the input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and are below the very low -140dBrA, or 0.00001%, level. This is an another ultra-clean IMD FFT.

Square-wave response (10kHz)

Above is the 10kHz squarewave response using the balanced analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the MXA-8400’s slew-rate performance. Rather, it should be seen as a qualitative representation of the MXA-8400’s mid-tier bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. In the case of the MXA-8400, however, what cn be seen in the plateaus of the squarewave in the top graph is a 500 kHz sinewave, the frequency at which the switching oscillator in the Class D amp is operating (see FFT below).

Square-wave response (10kHz)—250kHz bandwidth

Above is the 10kHz squarewave response using the analog input, at roughly 10W into 8 ohms, with a 250kHz restricted bandwidth to remove the modulation from the 500kHz oscillator. Here we find a relatively clean squarewave, with some overshoot in the corners.

FFT spectrum of 500kHz switching frequency relative to a 1kHz tone

The MXA-8400’s amplifier relies on a switching oscillator to convert the input signal to a pulse-width-modulated (PWM) squarewave (on/off) signal before sending the signal through a low-pass filter to generate an output signal. The MXA-8400 oscillator switches at a rate of about 500kHz, and this graph plots a wide bandwidth FFT spectrum of the amplifier’s output at 10W into 8 ohms as it’s fed a 1kHz sinewave. We can see that the 500kHz peak is quite evident, and at -40dBrA. There is also a peak at 1MHz (the second harmonic of the 500kHz peak), at -70dBrA. Those three peaks—the fundamental and its second/third harmonics—are direct results of the switching oscillators in the MXA-8400 amp modules. The noise around those very-high-frequency signals are in the signal, but all that noise is far above the audioband—and therefore inaudible—and so high in frequency that any loudspeaker the amplifier is driving should filter it all out anyway.

Damping factor vs. frequency (20Hz to 20kHz, two-channel mode)

The graph above is the damping factor as a function of frequency in two-channel mode. We see both channels with very high and constant damping factor, around 650 to 780 across the audioband.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 01 June 2024

Link: reviewed by Jonathan Gorse on SoundStage! Ultra on June 1, 2024

General information

All measurements were taken using an Audio Precision APx555 B Series analyzer.

The NAP 350 was conditioned for 1 hour at 1/8th full rated power (~25W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The NAP 350 is a single-channel amplifier with one balanced (XLR) input and one set of speaker level outputs. An input of 330mVrms was required to achieve the reference 10W into 8 ohms.

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Naim for the NAP 350 compared directly against our own. The published specifications are sourced from Naim’s website, either directly or from the manual available for download, or a combination thereof. Assume, unless otherwise stated, 10W into 8 ohms and a measurement input bandwidth of 10Hz to 22.4kHz.

Our primary measurements revealed the following (unless specified, assume a 1kHz sinewave at 330mVrms at the input, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the NAP 350 was able to sustain about 382W into 4 ohms (~2% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (38.2W) for 5 seconds, for 5 continuous minutes without inducing a fault or the initiation of a protective circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the sides of the NAP 350 were warm to the touch.

Frequency response (8-ohm loading)

In our frequency-response (relative to 1kHz) plot above, measured across the speaker outputs at 10W into 8 ohms, the NAP 350 exhibits a near-flat frequency response across the audioband (0/-2dB at 20Hz/20kHz). At low frequencies, the NAP 350 is -0.3dB at 5Hz. The -3dB point is just past 100kHz.

Phase response (8-ohm loading)

Above is the phase-response plot from 20Hz to 20kHz for the balanced line-level input, measured across the speaker outputs at 10W into 8 ohms. The NAP 350 does not invert polarity and exhibits, at worst, about -20 degrees of phase shift at 20kHz, and +5 degrees at 20Hz.

RMS level vs. frequency vs. load impedance (1W, single channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 10Hz to 100kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Here we find the maximum deviation between an 8-ohm load and no-load to be around 0.5dB. This is an indication of a low damping factor, or high output impedance, for a solid-sate amplifier. With a real speaker, the deviations are lower, at just over 0.3dB.

THD ratio (unweighted) vs. frequency vs. output power

The chart above shows THD ratios at the output into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input. The blue plot is at 1W output into 8 ohms, purple at 10W, and pink at 160W. The 1W data ranges from 0.002% at 20Hz, down to 0.0005% at 200-300Hz, then up to 0.01% at 20kHz. The 10W data ranges from 0.005% at 20Hz, down to 0.002% at 60-2kHz, then up to 0.01% at 20kHz. The 160W data ranges from 0.04% at 20Hz, down to 0.007% at 40-1kHz, then up to 0.1% at 20kHz.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the output of the NAP 350 as a function of output power for the analog line-level input, for an 8-ohm load (blue) and a 4-ohm load (purple). The 8-ohm data ranged from about 0.002% from 50mW to 20W, then up to 0.01% at the “knee,” at roughly 180W. The 4-ohm THD data were close, but 5-6dB higher, with a “knee” at just past 300W. The 1% THD thresholds were reached at 201W and 370W into 8 and 4 ohms.

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the NAP 350 as a function of output power for the analog line-level input, for an 8-ohm load (blue) and a 4-ohm load (purple). The 8-ohm data ranged from about 0.015% down to 0.0025% at 10-30W. The 4-ohm data ranged from about 0.02%% down to 0.004% at 10-30W.

THD ratio (unweighted) vs. frequency at 8 and 4 ohms

The chart above shows THD ratios measured at the output of the NAP 350 as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yields roughly 50W at the output into 8 ohms (blue), 100W into 4 ohms (purple), and 200W into 2 ohms (pink). The 8-ohm data ranged from 0.008% at 20Hz down to 0.003% from 30Hz to 2kHz, then a steady rise to 0.03% at 20kHz. The 4-ohm THD data were only about 5dB higher compared to the 8-ohm data from 20Hz to 2kHz, while above this frequency THD ratios were the same at 4 and 8 ohms. The 2-ohm data ranged from 0.02% at 20Hz down to 0.006% from 200Hz to 2kHz, then a steady rise to 0.04% at 20kHz.This shows that the NAP 350 is perfectly stable into 2 ohms, and it yields THD results that are comparable to those seen at 8 and 4 ohms.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (single channel only)

The chart above shows THD ratios measured at the output of the NAP 350 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). At very low freqencies, the two-way speaker yielded the highest THD ratios (0.4%). Generally, below 500Hz or so, THD ratios into the real speakers were higher than the resistive load, from 5 to 30dB. Above 1kHz, THD ratios into the real speakers were either lower (by up to 5dB) or close to the resistive-load values.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (single channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the NAP 350 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three IMD plots are close, within about 5dB, most often within a couple dB throughout the frequency sweep. IMD values ranged from 0.01 to 0.03%, which is not a particularly strong result.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (single channel only)

The chart above shows IMD ratios measured at the output of the NAP 350 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). The IMD results into real speakers were lower than those measured into the resistive load, by about 5dB. IMD values ranged from 0.03 to 0.01%, again, not a particularly strong result for a modern solid-state amplifier.

FFT spectrum – 1kHz (line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the balanced analog line-level input. We see that the signal’s second (2kHz) and third (3kHz) harmonics dominate at around -95dBrA, or 0.002%. The subsequent signal harmonics range from the -110dBrA level, or 0.0003%, down to -140dBrA, or 0.00001%. The highest noise-related peaks are at the third (180Hz) and fifth (300Hz) harmonics at a low -125dBrA, or 0.00006%.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the balanced analog line-level input. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most dominant (non-signal) peaks are the signal’s second (100Hz) and third (150Hz) harmonics at close to -100dBrA, or 0.001%. Noise-related harmonics are much lower, at the -125dBrA, or 0.00006%, and below level.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the balanced analog line-level input. The input RMS values were set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at the relatively high -85dBrA, or 0.006%, level, while the third-order modulation products, at 17kHz and 20kHz, are lower at -95dBrA, or 0.002%.

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the NAP 350 with the APx 32-tone signal applied to the input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and are below the -120dBrA, or 0.0001%, level.

Square-wave response (10kHz)

Above is the 10kHz squarewave response using the balanced analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the NAP 350’s slew-rate performance. Rather, it should be seen as a qualitative representation of the NAP 350’s relatively wide bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. Here we find clean corners with only mild softening.

Damping factor vs. frequency (20Hz to 20kHz)

The final graph above is the damping factor as a function of frequency. We find quite low damping factor values, right around 35 across the audioband. This is a poor damping factor result for a modern solid-state amplifier.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 01 May 2024

Link: reviewed by Dennis Burger on SoundStage! Access on May 1, 2024

General Information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The Marantz Model 50 was conditioned for one hour at 1/8th full rated power (~9W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The Model 50 offers four line-level analog inputs (RCA), one moving-magnet (MM) phono input (RCA), a sub output and left/right variable and fixed pre-outs plus a power amp input (all RCA), two pairs (A and B) of speaker-level outputs, and one headphone output over 1/4" TRS connector. For the purposes of these measurements, the following inputs were evaluated: analog line-level and phono.

Most measurements were made with a 2Vrms line-level analog input and a 5mVrms phono input. The signal-to-noise ratio (SNR) measurements were made with the default input signal values but with the volume set to achieve the rated output power of 70W (8 ohms). For comparison, SNR measurements were also made with the volume at maximum.

Based on the variability and non-repeatability of the left/right volume channel matching (see table below), the Model 50 volume control is digitally controlled operating in the analog domain. The Model 50 overall volume range is from -57.6dB to +41.8dB (line-level input, speaker output). It offers 0.5dB increments throughout the volume range.

Our typical input bandwidth filter setting of 10Hz–22.4kHz was used for all measurements except FFTs and THD vs. frequency sweeps, where a bandwidth of 10Hz –90kHz was used. Frequency response measurements utilized a DC to 1MHz input bandwidth.

Volume-control accuracy (measured at speaker outputs): left-right channel tracking

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Marantz for the Model 50 compared directly against our own. The published specifications are sourced from Marantz’s website, either directly or from the manual available for download, or a combination thereof. With the exception of frequency response, where the Audio Precision bandwidth was extended to 1MHz, assume, unless otherwise stated, 10W into 8 ohms and a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

Our primary measurements revealed the following using the line-level analog input and digital coaxial input (unless specified, assume a 1kHz sinewave at 2Vrms or 0dBFS, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the Model 50 was able to sustain 145W into 4 ohms (~3.5% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (14.5W) for 5 secondss, for 233 seconds of the 500-second test before inducing the fault-protection circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the top of the Model 50 was warm to the touch.

Our primary measurements revealed the following using the phono-level input, MM configuration (unless specified, assume a 1kHz sinewave at 5mVrms, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

Our primary measurements revealed the following using the analog input at the headphone output (unless specified, assume a 1kHz sinewave, 2Vrms output, 300-ohm loading, 10Hz to 22.4kHz bandwidth):

Frequency response (8-ohm loading, line-level input)

In our frequency-response plots above (relative to 1kHz), measured across the speaker outputs at 10W into 8 ohms, the Model 50 is essentially perfectly flat within the audioband (20Hz to 20kHz). At the extremes, the Model 50 is -0.1dB at 5Hz and -0.2dB at 100kHz. The Model 50 can be considered a high-bandwidth audio device. In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Frequency response (8-ohm loading, line-level input, bass and treble controls)

Above is a frequency response (relative to 1kHz) plot measured at the speaker-level outputs into 8 ohms, with the bass and treble controls set to maximum (blue/red plots) and minimum (purple/green plots). We see that for the bass and treble controls, roughly +/-11dB of gain/cut is available at 20Hz, and roughly +/-10dB of gain/cut at 20kHz.

Phase response (8-ohm loading, line-level input)

Above are the phase response plots from 20Hz to 20kHz for the line-level input, measured across the speaker outputs at 10W into 8 ohms. The Model 50 yields very little phase shift (as expected given the extended frequency response), with less than +5 degrees at 20Hz (the Model 50 is not DC coupled) and less than -5 degrees at 20kHz.

Frequency response (line-level pre and sub outputs)

Above is a frequency response plot measured at the line-level outputs into 8 ohms, where the left/right pre-outs are in purple/green and the sub-out is in blue.  The pre-outs show an extended frequency response, -0.1dB at 5Hz and -0.5dB at 70kHz. The sub-out is low-pass filtered with a -3dB point at around 150Hz.

Frequency response (8-ohm loading, MM phono input)

The chart above shows the frequency response (relative to 1 kHz) for the phono input (MM configuration) and shows very small maximum deviations of about +0.25/-0.1dB (100-200Hz/20kHz) from 20Hz to 20kHz. What is shown in this chart is the deviation from the RIAA curve, where the input signal sweep is EQ’d with an inverted RIAA curve supplied by Audio Precision (i.e., zero deviation would yield a flat line at 0dB).

Phase response (MM input)

Above is the phase-response plot from 20Hz to 20kHz for the phono input, measured across the speakers outputs at 10W into 8 ohms. For the phono input, since the RIAA equalization curve must be implemented, which ranges from +19.9dB (20Hz) to -32.6dB (90kHz), phase shift at the output is inevitable. Here we find a worst case of about -60 degrees at 200Hz and -90 degrees at 20kHz.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8 ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 5Hz to 50kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Here we see that the deviations between no load and 4 ohms are around 0.15dB. This is a reasonably strong result for an integrated class-AB amp, and an indication of a low output impedance, or high damping factor. With a real speaker load, deviations measured just below the 0.07dB level—well below the threshold of audibility.

THD ratio (unweighted) vs. frequency vs. output power

The chart above shows THD ratios at the speaker-level outputs into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input. The blue and red plots are for left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange at the rated 70W. The power was varied using the Model 50 volume control. Between 20Hz and 2kHz, all THD ratios are fairly constant and similar, between 0.004 and 0.006%. Between 2kHz and 20kHz, THD ratios were higher at higher power levels, but not by a significant margin. At 20kHz, we measured around 0.01% at 1W and 10W, and 0.03% at 70W.

THD ratio (unweighted) vs. frequency at 10W (MM input)

The graph above shows THD ratio as a function of frequency plot for the phono input measured across an 8-ohm load at 10W. The input sweep was EQ’d with an inverted RIAA curve. The THD values vary from 0.01% (20Hz) down to between 0.001 and 0.002% from 200Hz to 3kHz, then up to 0.006% at 20kHz.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the speaker-level outputs of the Model 50 as a function of output power for the analog line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). THD ratios were essentially identical into 8 and 4 ohms up to the 8-ohm “knee” at roughly 70W (the 4-ohm “knee” is at roughly 100W). These ranged from 0.003% at 50mW, down to 0.001% at 1 to 3W, then up to 0.005% at the “knees.” The 1% THD values were hit at 83W and 132W into 8 and 4 ohms, respectively.

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the Model 50 as a function of output power for the line level-input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels).  Both data sets track closely, but the 4-ohm data yielded a couple dB more noise than the 8-ohm data. THD+N ratios into 8 ohms ranged from 0.02% at 50mW, down to 0.002% at 10W, then up to just below 0.005% at the 8-ohm “knee.”  

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only)

The chart above shows THD ratios measured at the output of the Model 50 as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yielded 20W at the output into 8 ohms (and roughly 40W into 4 ohms, and 80W into 2 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. We find essentially identical THD ratios (0.005%) into all three loads up to about 1kHz. From 3kHz to 20kHz, there is a roughly 5dB increase in THD between the 2-ohm load and the 8/4-ohm loads. At 20kHz, we measured 0.015% into 8/4 ohms, and 0.025% into 2 ohms. This is a strong result, and shows that the Model 50 is stable into 2 ohms.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the Model 50 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). Generally, THD ratios into the real speakers were similar to the resistive dummy load, with the expection of the two-way speaker at 20 to 30Hz. THD ratios hovered between 0.007 and 0.01% from 40Hz to 20kHz for all three loads. At 20Hz, the THD ratio was 0.07% into the two-way speaker. This is a relatively strong result, and shows that the Model 50 will yield consistently low THD results into real-world speaker loads.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the Model 50 as a function of frequency into an 8-ohm load and two different speakers, for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find essentially identical IMD ratios into all three loads, at a relatively flat and constant 0.005%.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the Model 50 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a thee-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find essentially identical IMD ratios into all three loads, at a relatively flat 0.02% from 40Hz to 500Hz, then down to 0.0015% from 500Hz to 1kHz.

FFT spectrum – 1kHz (line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input. We see that the signal’s second (2kHz) and third (3kHz) harmonics dominate at -85dBrA, or 0.006%, and -105dBrA, or 0.0006%, while subsequent signal harmonics are below -110dBrA, or 0.0003%. On the right side of the signal peak, we see the primary (60Hz) noise-related peak and its harmonics (120, 180, 240, 300Hz, etc.) at and below the very low -120dBrA, or 0.0001%, level.

FFT spectrum – 1kHz (MM phono input)

Shown above is the FFT for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the phono input, configured for MM. We see the second (2kHz) signal harmonic dominate at -95dBrA, or 0.002%, while subsequent signal harmonics are below the -110dBrA, or 0.0003%, level. The noise related peaks are at and below the -85dBrA, or 0.006%, level.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the analog line-level input. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peaks are the second (100Hz) and third (150Hz) signal harmonics at a -90dBrA, or 0.003%, and -105dBrA, or 0.0006%. Noise related peaks can be seen below the -110dBRa, or 0.0003%, level.

FFT spectrum – 50Hz (MM phono input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the phono input configured for MM. The most predominant (non-signal) peaks are that of the power-supply fundamental (60Hz) and third (180Hz) harmonics at -90dBrA, or 0.003%, and -85dBrA, or 0.006%. The signal’s second (100Hz) harmonic is at the -95dBrA, or 0.002%, level.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the analog line-level input. The input RMS values were set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -95dBRa, or 0.002%, while the third-order modulation products, at 17kHz and 20kHz are at -100dBrA, or 0.001%.

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the Model 50 with the APx 32-tone signal applied to the analog input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and are at and below the very low -120dBrA, or 0.0001%, level.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, MM phono input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the phono input configured for MM. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -100dBRa, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are at the same level.

Squarewave response (10kHz)

Above is the 10kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the Model 50’s slew-rate performance. Rather, it should be seen as a qualitative representation of the Model 50’s extremely high bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. In this case, due to the Model 50’s very extended bandwidth, we see a near-perfect squarewave, with sharp corners and no ringing.

Damping factor vs. frequency (20Hz to 20kHz)

The final graph above is the damping factor as a function of frequency. We can see here a constant damping factor right around 200, from 20Hz to roughly 20kHz. This is a relatively strong damping factor result for an affordable class AB integrated amp.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 01 May 2024

Link: reviewed by George de Sa on SoundStage! Simplifi on May 1, 2024

General Information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The Rotel RAS-5000 was conditioned for one hour at 1/8th full rated power (~17W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The RAS-5000 offers one pair of line-level analog inputs (RCA), a sub output (RCA), left/right pre-outs (RCA), one coaxial S/PDIF input (RCA), one optical S/PDIF input (TosLink), one USB digital input, a pair of speaker level outputs, and one headphone output over a 1/4" TRS connector. Bluetooth, streaming, and HDMI (eARC) inputs are also offered. For the purposes of these measurements, the following inputs were evaluated: digital coaxial, analog line-level.

Most measurements were made with a 2Vrms line-level analog input and 0dBFS digital input. The signal-to-noise ratio (SNR) measurements were made with the default input signal values but with the volume set to achieve the rated output power of 140W (8 ohms). For comparison, on the line-level input, an SNR measurement was also made with the volume at maximum.

Based on the variability and non-repeatability of the left/right volume channel matching (see table below), the RAS-5000 volume control is digitally controlled but operating in the analog domain. The RAS-5000 overall volume range is from -69dB to +32.4dB (line-level input, speaker output). It offers 1dB increments up to about the 75% mark, and then 0.5dB steps to full volume.

Our typical input bandwidth filter setting of 10Hz–22.4kHz was used for all measurements except FFTs and THD vs. frequency sweeps where a bandwidth of 10Hz–90kHz was used. Frequency response measurements utilize a DC to 1 MHz input bandwidth.

Volume-control accuracy (measured at speaker outputs): left-right channel tracking

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Rotel for the RAS-5000 compared directly against our own. The published specifications are sourced from Rotel’s website, either directly or from the manual available for download, or a combination thereof. With the exception of frequency response, where the Audio Precision bandwidth was extended to 1MHz, assume, unless otherwise stated, 10W into 8ohms and a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

*protection circuit enabled after a few seconds

Our primary measurements revealed the following using the line-level analog input and digital coaxial input (unless specified, assume a 1kHz sinewave at 2Vrms or 0dBFS, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

*protection circuit enabled after a few seconds

For the continuous dynamic power test, the RAS-5000 was able to sustain 237W into 4 ohms (~2% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (23.7W) for 5 seconds, for 233 seconds of the 500-second test before inducing the fault protection circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the top of the RAS-5000 was warm to the touch.

Our primary measurements revealed the following using the analog input at the headphone output (unless specified, assume a 1kHz sinewave, 2Vrms output, 300-ohm loading, 10Hz to 22.4kHz bandwidth):

Frequency response (8-ohm loading, line-level input)

In our frequency-response plots above (relative to 1kHz), measured across the speaker outputs at 10W into 8 ohms, the RAS-5000 is essentially perfectly flat within the audioband (20Hz to 20kHz). There’s a +0.4dB rise in the frequency response at 200kHz, and -0.2dB at 10Hz. The RAS-5000 can be considered a high-bandwidth audio device. In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Frequency response (8-ohm loading, line-level input, bass and treble controls)

Above is a frequency-response (relative to 1kHz) plot measured at the speaker-level outputs into 8 ohms, with the bass and treble controls set to maximum (blue/red plots) and minimum (purple/green plots). We see that for the bass and treble controls, roughly +/-11dB of gain/cut is available at 20Hz, and roughly +/-9dB of gain/cut at 20kHz.

Phase response (8-ohm loading, line-level input)

Above are the phase-response plots from 20Hz to 20kHz for the line level input, measured across the speaker outputs at 10W into 8 ohms. The RAS-5000 yields very little phase shift (as expected given the extended frequency response), with +10 degrees at 20Hz (the RAS-5000 is not DC coupled), and less than -5 degrees at 20kHz.

Frequency response (line-level pre and sub outputs)

Above is a frequency response plot measured at the line-level outputs into 8 ohms, where the L/R pre-outs are in blue/red, and the sub-out is in purple.  All three plots overlap perfectly, with a ruler-flat and extended frequency response, as was seen with the speaker outputs. The sub-out is not low-pass filtered in any way, and is likely simply a summed L/R version of the pre-outs.

Frequency response vs. input type (8-ohm loading, left channel only)

The chart above shows the RAS-5000’s frequency response (relative to 1kHz) as a function of input type measured across the speaker outputs at 10W into 8 ohms. The two green traces are the same analog input data from the speaker-level frequency response graph above, but limited to 80kHz. The blue and red traces are for a 16-bit/44.1kHz dithered digital input signal from 5Hz to 22kHz using the coaxial input, and the purple and green traces are for a 24/96 dithered digital input signal from 5Hz to 48kHz, and the pink and orange traces are for a 24/192 dithered digital input signal. At low frequencies, all four plots yield the same -0.2dB at 10Hz. The -3dB points for the 16/44.1, 24/96, and 24/192 digital input data are: 21.0kHz, 35.1kHz, and 71.2kHz.

Digital linearity (16/44.1 and 24/96 data)

The chart above shows the results of a linearity test for the coaxial digital input for both 16/44.1 (blue/red) and 24/96 (purple/green) input data, measured at the line-level pre-outputs of the RAS-5000, where 0dBFS was set to yield 2Vrms. The digital input was swept with a dithered 1kHz input signal from -120dBFS to 0dBFS, and the output was analyzed by the APx555. The ideal response would be a straight flat line at 0dB. The 16/44.1 data were essentially perfect as of -100dBFS down to 0dBFS, while the 24/96 data were near perfect down to -120dBFS. We all also extended the sweep down to -140dBFS, to . . .

. . . see how well the 24/96 would perform. We can see here, only a +3/+1dB (L/R) overshoot at -140dBFS. This is an exemplary digital-linearity result.

Impulse response (24/44.1 data)

The graph above shows the impulse response for the RAS-5000, fed to the coaxial digital input, measured at the line-level pre-outputs, for a looped 24/44.1 test file that moves from digital silence to full 0dBFS (all “1”s), for one sample period then back to digital silence. We find a reconstruction filter that minimizes pre-ringing, with short post-ringing.

J-Test (coaxial)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level pre-outputs of the RAS-5000 where 0dBFS is set to 2Vrms. J-Test was developed by Julian Dunn in the 1990s. It is a test signal—specifically, a -3dBFS undithered 12kHz squarewave sampled (in this case) at 48kHz (24 bits). Since even the first odd harmonic (i.e., 36kHz) of the 12kHz squarewave is removed by the bandwidth limitation of the sampling rate, we are left with a 12kHz sinewave (the main peak). In addition, an undithered 250Hz squarewave at -144dBFS is mixed with the signal. This test file causes the 22 least significant bits to constantly toggle, which produces strong jitter spectral components at the 250Hz rate and its odd harmonics. The test file shows how susceptible the DAC and delivery interface are to jitter, which would manifest as peaks above the noise floor at 500Hz intervals (e.g,, 250Hz, 750Hz, 1250Hz, etc.). Note that the alternating peaks are in the test file itself, but at levels of -144dBrA and below.  The test file can also be used in conjunction with artificially injected sinewave jitter by the Audio Precision, to show how well the DAC rejects jitter.

Here we see an average J-Test result, with peaks flanking the 12kHz fundamental, as high as -110dBrA. This is an indication that the RAS-5000 DAC may be susceptible to jitter.

J-Test (optical)

The chart above shows the results of the J-Test test for the optical digital input measured at the line-level pre-outputs of the RAS-5000. The optical input yielded similar but slightly worse results compared to the coaxial input.

J-Test (coaxial, 10ns jitter)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the RAS-5000, with an additional 10ns of 2kHz sinewave jitter injected by the APx555. The results are very clear, as we see the sidebands at 10kHz and 14kHz (12kHz main signal +/- 2kHz jitter signal) manifest at near -100dBrA. This is further indication that the DAC in the RAS-5000 has poor jitter immunity. For this test, the optical input yielded effectively the same results.

J-Test (coaxial, 100ns jitter)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the RAS-5000, with an additional 100ns of 2kHz sinewave jitter injected by the APx555. The results are very clear, as we see the sidebands at 10kHz and 14kHz (12kHz main signal +/- 2kHz jitter signal) manifest at near -80dBrA. This is further indication that the DAC in the RAS-5000 has poor jitter immunity. For this test, the optical input yielded effectively the same results.

Wideband FFT spectrum of white noise and 19.1kHz sine-wave tone (coaxial input)

The chart above shows a fast Fourier transform (FFT) of the RAS-5000’s line-level pre-outputs with white noise at -4dBFS (blue/red) and a 19.1 kHz sinewave at -1dBFS fed to the coaxial digital input, sampled at 16/44.1. The gentle roll-off around 20kHz in the white-noise spectrum shows that the RAS-5000 does not use a brick-wall type reconstruction filter. There are very clear low-level aliased image peaks within the audio band at the -90dBrA and below level. The primary aliasing signal at 25kHz is prominent at -20dBrA, while the second and third distortion harmonics (38.2, 57.3kHz) of the 19.1kHz tone are at -80dBrA.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 5Hz to 50kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Here we that the deviations between no load and 4 ohms are around 0.05dB. This is a strong result for a class-AB amp, and an indication of a low output impedance, or high damping factor. With a real speaker load, deviations measured just below the 0.05dB level—well below the threshold of audibility.

THD ratio (unweighted) vs. frequency vs. output power

The chart above shows THD ratios at the speaker-level outputs into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input. The blue and red plots are for the left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange just under 137W (just shy of the rated 140W). The power was varied using the RAS-5000 volume control. Between 20Hz and 1kHz, all THD ratios are fairly constant and similar, between 0.001 and 0.002%. Between 1kHz and 20kHz, THD ratios were higher at higher power levels, but not by a significant margin. At 20kHz, we measured 0.002% at 1W, 0.003% at 10W, and 0.01% at 137W.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the speaker-level outputs of the RAS-5000 as a function of output power for the analog line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). Into 4 ohms, the right channel outperformed the left by more than 10dB between about 2 and 20W. The right channel THD ratios into 4 ohms ranged from 0.002% at 50mW, down to nearly 0.0002% at 5W, then up to 0.002% at just shy of 200W, where the RAS-5000 protection circuit engaged and shut down the unit. THD ratios into 8 ohms ranged from 0.001% at 50mW, then down to 0.0003-0.0005% between 1 and 10W, then up to 0.002% at the “knee” at roughly 140W, then up to the 1% THD mark at 154W. Note that we were able to achieve 230W into 4 ohms (1% THD) with the RAS-5000 in Bench Mode, but only for a few seconds before the protection circuit engaged and shut down the unit.

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the RAS-5000 as a function of output power for the line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels).  Both data sets track closely except for the left input into 4 ohms, which yielded about 5dB more THD+N from 10W to over 100W. Otherwise, THD+N ratios ranged from 0.02% at 50mW, down to 0.001% at 20 to 50W, then up to just below 0.002% at the 8-ohm “knee.”  

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only)

The chart above shows THD ratios measured at the output of the RAS-5000 as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yielded 20W at the output into 8 ohms (and roughly 40W into 4 ohms, and 80W into 2 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. We find essentially identical THD ratios (0.0015%) into all three loads up to about 200Hz. From 2kHz to 20kHz, there is a roughly 7-8dB increase in THD every time the load is halved. At 20kHz, we measured 0.003% into 8 ohms, 0.009% into 4 ohms, and 0.02% into 2 ohms. This is a strong result, and shows that the RAS-5000 is stable into 2 ohms.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the RAS-5000 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). Generally, THD ratios into the real speakers were similar to the resistive dummy load, with the expection of the two-way speaker at 20 to 30Hz, and the three-way speaker at 10 to 20kHz. THD ratios hovered between 0.001 and 0.003% from 40Hz to 6kHz for all three loads. At 20Hz, the THD ratio was 0.04% into the two-way speaker, and at 20kHz, 0.008% into the three-way speaker. This is a relatively strong result, and shows that the RAS-5000 will yield consistently low THD results into real-world speaker loads.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the RAS-5000 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find very similar IMD ratios into all three loads, between 0.001 and 0.003% from 2.5kHz to 20kHz. Another strong result.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the RAS-5000 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find very similar IMD ratios into all three loads, between 0.002 and 0.005% from 40Hz to 1kHz. Another strong result.

FFT spectrum – 1kHz (line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input. We see that the signal’s second (2kHz) and third (3kHz) harmonics dominate at a low -100dBrA, or 0.001%, while subsequent signal harmonics are near and below -120dBrA, or 0.0001%. On the right side of the signal peak, we see the primary (60Hz) noise-related peak and its harmonics (120, 180, 240, 300Hz, etc.) at the -105dBrA, or 0.0006%, and below level.

FFT spectrum – 1kHz (digital input, 16/44.1 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 16/44.1. Both the signal and noise-related harmonic peaks are very similar to the analog FFT above, but for a higher noise floor (-135dBFS) due to the 16-bit depth.

FFT spectrum – 1kHz (digital input, 24/96 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 24/96. Both the signal and noise-related harmonic peaks are very similar to the analog FFT above.

FFT spectrum – 1kHz (digital input, 16/44.1 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 16/44.1 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at the correct amplitude, and noise-related peaks below the -110dBrA, or 0.0003%, level. There are no signal related peaks above the -135dBFS noise floor.

FFT spectrum – 1kHz (digital input, 24/96 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 24/96 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at the correct amplitude, and noise-related peaks below the -110dBrA, or 0.0003%, level. Signal-related peaks are difficult to discern above the -145dBFS noise floor.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the analog line-level input. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peak is the second (100Hz) and third (150Hz) signal harmonic at a low -100dBrA, or 0.001%. Noise related peaks can be seen at -105dBRA, or 0.0006%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the analog line-level input. The input RMS values were set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -110dBRa, or 0.0003%, while the third-order modulation products, at 17kHz and 20kHz, are at -105dBrA, or 0.0006%. This is a strong IMD result.

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the RAS-5000 with the APx 32-tone signal applied to the analog input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and are at and below the very low -120dBrA, or 0.0001%, level. The low frequency peaks that rise near and above -110dBrA, are due to power-supply noise.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 16/44.1)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 16/44.1 (-1dBFS). We find that the second-order modulation product (i.e., the difference signal of 1kHz) is just below -100dBrA, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are around -95dBrA, or 0.002%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 24/96)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 24/96 (-1dBFS). We find that the second-order modulation product (i.e., the difference signal of 1kHz) is just below -100dBrA, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are around -95dBrA, or 0.002%.

Squarewave response (10kHz)

Above is the 10kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the RAS-5000’s slew-rate performance. Rather, it should be seen as a qualitative representation of the RAS-5000’s extremely high bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. In this case, due to the RAS-5000’s very extended bandwidth, we see a near-perfect squarewave, with sharp corners and no ringing.

Damping factor vs. frequency (20Hz to 20kHz)

The final graph above is the damping factor as a function of frequency. We can see here a constant damping factor right around 300, from 20Hz to roughly 15kHz, then a dip down to around 80 at 20kHz. This is a relatively strong damping factor result for an affordable class-AB amp.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 01 March 2024

Link: reviewed by Dennis Burger on SoundStage! Access on March 1, 2024

General Information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The Dayton Audio HTA200 was conditioned for one hour at 1/8th full rated power (~4W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The HTA200 offers one pair of line-level analog inputs (RCA), one pair of moving-magnet (MM) phono inputs (RCA), an RCA sub output, one digital coaxial (RCA) input, one optical (TosLink) input, one USB digital input, a pair of speaker level outputs and one headphone output over 1/4″ TRS connector. A Bluetooth input is also offered. For the purposes of these measurements, the following inputs were evaluated: digital coaxial, analog line-level, as well as phono.

Most measurements were made with a 2Vrms line-level analog input, 5mVrms MM input, and 0dBFS digital input. The signal-to-noise (SNR) measurements were made with the default input signal values but with the volume set to achieve the rated output power of 70W (4 ohms). For comparison, on the line-level input, a SNR measurement was also made with the volume at maximum.

Based on the inaccuracy and non-repeatability of the left/right volume channel matching (see table below), the HTA200 volume control is a potentiometer operating in the analog domain. The HTA200 overall volume range is from -59dB to +28.6dB (line-level input, speaker output).

Our typical input bandwidth-filter setting of 10Hz–22.4kHz was used for all measurements except FFTs and THD vs frequency sweeps where a bandwidth of 10Hz–90kHz was used. Frequency response measurements utilize a DC to 1 MHz input bandwidth.

Volume-control accuracy (measured at speaker outputs): left-right channel tracking

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Dayton Audio for the HTA200 compared directly against our own. The published specifications are sourced from Dayton’s manual provided with the review sample (note: the PDF availiable on Dayton's website shows much higher power ratings that are less realistic than those shown here). With the exception of frequency response, where the Audio Precision bandwidth was extended to 1MHz, assume, unless otherwise stated, 10W into 8 ohms and a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

Our primary measurements revealed the following using the line-level analog input and digital coaxial input (unless specified, assume a 1kHz sinewave at 2Vrms or 0dBFS, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the HTA200 was able to sustain 133W into 4 ohms (~7% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (13.3W) for five seconds, for five continuous minutes without inducing a fault or the initiation of a protective circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the top of the HTA200 (except for the tubes) was only slightly warm to the touch.

Our primary measurements revealed the following using the phono-level input, MM configuration (unless specified, assume a 1kHz 5mVrms sinewave input, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

Our primary measurements revealed the following using the analog input at the headphone output (unless specified, assume a 1kHz sinewave, 2Vrms output, 300-ohm loading, 10Hz to 22.4kHz bandwidth):

Frequency response (8-ohm loading, line-level input)

In our frequency-response plots above (relative to 1kHz), measured across the speaker outputs at 10W into 8 ohms, the HTA200 is not flat within the audioband (20Hz to 20kHz). There’s a +1dB rise in the frequency response at 40Hz and at nearly 20kHz. There’s also a -0.3dB dip at 5-7kHz. The HTA200 appears to be AC coupled (or at least purposefully high-pass filtered) with a -3dB point at 10Hz. There is extreme brickwall-type filtering at 20kHz, with a -3dB point at around 21.8kHz. There is a high probability that the analog input is digitized, because this type of brickwall filtering is easier to implement in the digital domain. Further evidence for this can be seen in the FFTs in this report. In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Frequency response (8-ohm loading, line-level input, bass and treble controls)

Above is a frequency response plot measured at the speaker-level outputs into 8 ohms, with the bass and treble controls set to maximum (blue/red plots) and minimum (purple/green plots). We see that for the bass and treble controls, roughly +7.5/-6.5dB of gain/cut is available centered at 100Hz/8kHz.

Phase response (8-ohm loading, line-level input)

Above are the phase-response plots from 20Hz to 20kHz (wrapped, or every time the phase delay exceeds 360 degrees, the plot loops back up to +180 degrees) for the line-level input, measured across the speaker outputs at 10W into 8 ohms. The HTA200 appears to invert polarity and yields an astonishing -60000 degrees of phase shift at 20kHz.

Frequency response (line-level sub out)

Above is the frequency-response plot (relative to 80Hz) measured at the line-level sub out of the HTA200. The same rise at 40Hz that was observed at the speaker-level outputs can be seen here. The -3dB point is just past 2kHz.

Frequency response vs. input type (8-ohm loading, left channel only)

The chart above shows the HTA200’s frequency response (relative to 1kHz) as a function of input type measured across the speaker outputs at 10W into 8 ohms. The pink trace is the same analog input data from the previous graph. The blue trace is for a 16-bit/44.1kHz dithered digital input signal from 5Hz to 22kHz using the coaxial input, and the purple trace is for a 24/96 dithered digital input signal from 5Hz to 48kHz, and the 24/192 dithered digital input signal is not shown because it yielded the same result as the 24/96 data. Because the HTA200 appears to digitize incoming analog signals, as well as resample all digital signals to 44.1kHz, the same brickwall-type behavior is seen right at 20kHz regardless of the input type. At low frequencies, there is slightly more extension for the analog input, with a -3dB point at 10Hz versus about 12Hz for the digital input. The digital input shows less significant deviations within the audioband, with a rise of only 0.5dB at 50Hz versus the 1dB rise at 40Hz for the analog input. The same is true at 20kHz, where the digital input is at +0.25dB compared to the +0.8dB for the analog signal.

Frequency response (8-ohm loading, MM phono input)

The chart above shows the frequency response (relative to 1kHz) for the phono input (MM configuration). What is shown is the deviation from the RIAA curve, where the input signal sweep is EQ’d with an inverted RIAA curve supplied by Audio Precision (i.e., zero deviation would yield a flat line at 0dB). We see a maximum deviation of about -5dB/+0.8dB (20Hz and 20kHz) from 20Hz to 20kHz. The dip at 5-7kHz seen with the line-level input is also seen here; however, with a channel-to-channel deviation of about 0.3dB (from 5kHz to 20kHz).

Phase response (MM input)

Above is the phase-response plot from 20Hz to 20kHz (excess, or above and beyond the true input to output phase delay as seen in the plot for the line-level input above) for the phono input (MM configuration) measured across the speaker outputs at 10W into 8 ohms. For the phono input, since the RIAA equalization curve must be implemented, which ranges from +19.9dB (20Hz) to -32.6dB (90kHz), phase shift at the output is inevitable. Here we find a worst case of about +150 degrees at 20Hz.

Digital linearity (16/44.1 and 24/96 data)

The chart above shows the results of a linearity test for the coaxial digital input for both 16/44.1 (blue/red) and 24/96 (purple/green) input data, measured at the speaker outputs of the HTA200 at 1W into 8 ohms. The digital input is swept with a dithered 1kHz input signal from -120dBFS to 0dBFS, and the output is analyzed by the APx555. The ideal response would be a straight flat line at 0dB. The 16/44.1 data were essentially perfect as of -90dBFS down to 0dBFS, while the 24/96 data were near perfect down to -100dBFS. Both overshot the mark by over 10dB at -120dBFS. This is a poor linearity-test result, although it should be pointed that the linearity test is measured at the line-level pre-out when available. In this case, there are none, and the speaker outputs will invariably be noisier.

Impulse response (24/44.1 data)

The graph above shows the impulse response for the HTA200, fed to the coaxial digital input, measured at the speaker level output at 1W into 8ohms, for a looped 24/44.1 test file that moves from digital silence to full 0dBFS (all “1”s), for one sample period then back to digital silence. We find a reconstruction filter that adheres to a typical symmetrical sinc function, although the HTA200 does invert polarity.

J-Test (coaxial)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the speaker level output of the HTA200 at 1W into 8 ohms. J-Test was developed by Julian Dunn the 1990s. It is a test signal—specifically,  a -3dBFS undithered 12kHz squarewave sampled (in this case) at 48kHz (24 bits). Since even the first odd harmonic (i.e., 36kHz) of the 12kHz squarewave is removed by the bandwidth limitation of the sampling rate, we are left with a 12kHz sinewave (the main peak). In addition, an undithered 250Hz squarewave at -144dBFS is mixed with the signal. This test file causes the 22 least-significant bits to constantly toggle, which produces strong jitter spectral components at the 250Hz rate and its odd harmonics. The test file shows how susceptible the DAC and delivery interface are to jitter, which would manifest as peaks above the noise floor at 500Hz intervals (e.g., 250Hz, 750Hz, 1250Hz, etc.). Note that the alternating peaks are in the test file itself, but at levels of -144dBrA and below.  The test file can also be used in conjunction with artificially injected sinewave jitter by the Audio Precision, to show how well the DAC rejects jitter.

It is difficult to assess the results, with the significant rise in the noise floor (-80dBrA) centered on the main peak. We see a peak at -75dBrA at 8kHz, and another at -85dBrA at 16kHz. This is a poor J-Test result, indicating that the HTA200 DAC may be susceptible to jitter. When we attempted to inject artificial jitter at a level of only 10ns, we could not capture a reliable result.

J-Test (optical)

The chart above shows the results of the J-Test test for the optical digital input measured at the speaker-level output of the HTA200 at 1W into 8 ohms. The optical input yielded essentially the same result as the coaxial input.

Wideband FFT spectrum of white noise and 19.1kHz sine-wave tone (coaxial input)

The chart above shows a fast Fourier transform (FFT) of the HTA200’s speaker level output at 1W into 8 ohms, with white noise at -4dBFS (blue/red) and a 19.1 kHz sinewave at -1dBFS fed to the coaxial digital input, sampled at 16/44.1. The steep roll-off around 20kHz in the white-noise spectrum shows the brickwall-type behavior of the HTA200’s reconstruction filter. There are low-level aliased image peaks within the audioband at the -100dBrA level. The primary aliasing signal at 25kHz cannot be seen and is completely suppressed, while the second and third distortion harmonics (38.2, 57.3kHz) of the 19.1kHz tone are at -55dBrA.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 5Hz to 50kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Here we find that between 100Hz and 8kHz, the deviations between no load and 4 ohms are around 0.15dB. This is a fair result for a class-AB amp, and an indication of a relatively low output impedance, or high damping factor. Due to the nonlinear nature of the HTA200’s frequency response, it is difficult to assess the fluctuations in response versus frequency for the real speaker load.

THD ratio (unweighted) vs. frequency vs. output power

The chart above shows THD ratios at the output into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input. The blue and red plots are for left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange just under 35W. The power was varied using the HTA200 volume control. At 1W, THD ratios are fairly constant at 0.3% from 20Hz to 20kHz. At 10W, THD ratios are fairly constant across the audioband from 1% to 0.8%, while at 35W, THD ratios are as high as 2% at 20Hz, then a constant 1.5% from 100Hz to 20kHz. The HTA200 is definitely a high-distortion amplifier, considering the class-AB transistor-based output.

THD ratio (unweighted) vs. frequency at 10W (MM input)

The chart above shows THD ratio as a function of frequency plots for the phono input measured across an 8-ohm load at 10W. The input sweep is EQ’d with an inverted RIAA curve. The THD values vary from around 0.5% to 0.8% across the audioband.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the output of the HTA200 as a function of output power for the analog line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). Both data sets track closely except for maximum power, with the 4-ohm data slightly outperforming (2-3dB) the 8-ohm data at medium power. THD ratios range from as low as 0.1% at 50mW, then a steady linear climb to the “knees” at 80W (2% into 8 ohms) and just past 100W (2% into 4 ohms). The 1% THD marks were reached at just 16.5W (8 ohms) and 32W (4 ohms).

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the HTA200 as a function of output power for the line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels).  Both data sets track closely except for maximum power, with the 4-ohm data slightly outperforming (2-3dB) the 8-ohm data at lower power. Overall, THD+N values for both loads ranged from 0.2% at 50mW, then a steady linear climb to the “knees” at 80W (2% into 8 ohms) and just past 100W (2% into 4 ohms). Because THD ratios are so high with the HTA200, it’s the THD component of THD+N that dominates in these graphs.

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only)

The chart above shows THD ratios measured at the output of the HTA200 as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yielded 10W at the output into 8 ohms (and roughly 20W into 4 ohms, and 40W into 2 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. We find essentially identical THD ratios into all three loads, from 1% to 0.8%. Ordinarily, having all three data sets plot identically would be commendable; however, in this case, THD ratios are very high for a solid-state amplifier output.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the HTA200 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find essentially identical THD ratios into all three loads, just above and below 0.3% to 0.8%. As above, ordinarily, having all three data sets plot identically would be commendable; however, in this case, THD ratios are very high for a solid-state amplifier output.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the HTA200 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find essentially identical IMD ratios into all three loads, around 0.3%. Once again, having all three data sets plot identically would be commendable; however, in this case, the IMD ratios are very high for a solid-state amplifier output.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the HTA200 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). We find essentially identical IMD ratios into all three loads, 1% from 40Hz to almost 500Hz, and then a dip to 0.02% from 500Hz to 1kHz. As mentioned above, having all three data sets plot identically would be commendable; however, in this case, the IMD ratios are very high for a solid-state amplifier output.

FFT spectrum – 1kHz (line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input. We see that the signal’s second (2kHz) harmonic dominates at a very high -40dBrA, or 1%, while subsequent signal harmonics are at and below -70dBrA, or 0.03%. On the right side of the signal peak, we see the primary (60Hz), second (120Hz), and fourth (240Hz) noise-related harmonics dominate at -85dBrA to -95dBrA, or 0.006% to 0.002%. While difficult to see in this graph, we did find two peaks at 43.1kHz and 45.1kHz, which are telltale IMD products that are a result of the 1kHz analog signal being digitized and sampled at 44.1kHz.

FFT spectrum – 1kHz (digital input, 16/44.1 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 16/44.1. We see that the signal’s second (2kHz) harmonic dominates at a very high -40dBrA, or 1%, while subsequent signal harmonics are at and below -55dBrA, or 0.2%. On the right side of the signal peak, we see the primary (60Hz), second (120Hz), and fourth (240Hz) noise-related harmonics dominate at -85dBrA to -95dBrA, or 0.006% to 0.002%.

FFT spectrum – 1kHz (digital input, 24/96 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 24/96. We see that the signal’s second (2kHz) harmonic dominates at a very high -45dBrA, or 0.6%, while subsequent signal harmonics are at and below -90dBrA, or 0.003%. On the right side of the signal peak, we see the primary (60Hz), second (120Hz), and fourth (240Hz) noise-related harmonics dominate at -85dBrA to -95dBrA, or 0.006% to 0.002%.

FFT spectrum – 1kHz (digital input, 16/44.1 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 16/44.1 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at just under the correct amplitude, and noise-related peaks at -90dBrA, or 0.003%, and below.

FFT spectrum – 1kHz (digital input, 24/96 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 24/96 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at just under the correct amplitude, and noise-related peaks at -90dBrA, or 0.003%, and below.

FFT spectrum – 1kHz (MM phono input)

Shown above is the FFT for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the MM phono input. We see that the signal’s second (2kHz) harmonic dominates at a very high -40dBrA, or 1%, while subsequent signal harmonics are at and below -70dBrA, or 0.03%. On the right side of the signal peak, we see the primary (60Hz) noise-related peak dominate at -65/-75dBrA (left/right), or 0.06/0.02%.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the analog line-level input. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peak is the second (100Hz) signal harmonic at a high -40dBrA, or 1%. Several other signal-related harmonic peaks can be seen throughout at -70dBrA, or 0.03%, and below.

FFT spectrum – 50Hz (MM phono input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the MM phono input. The X axis is zoomed in from 40 Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peak is from the second (100Hz) signal harmonic at a high -45dBrA, or 0.6%. The worst-case noise-related peak is from the fundamental (60Hz) for the left channel at -60dBrA, or 0.1%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the analog line-level input. The input RMS values are set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -45dBRa, or 0.6%, while the third-order modulation products, at 17kHz and 20kHz, are at -80dBrA, or 0.01%.

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the HTA200 with the APx 32-tone signal applied to the analog input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e. the “grass” between the test tones—are distortion products from the amplifier and are at and below the -70dBrA, or 0.03%, level.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 16/44.1)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 16/44.1 (-1dBFS). We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -50dBrA, or 0.3%, while the third-order modulation products, at 17kHz and 20kHz, are around -75dBrA, or 0.02%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 24/96)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 24/96 (-1dBFS). We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -50dBrA, or 0.3%, while the third-order modulation products, at 17kHz and 20kHz, are around -90dBrA, or 0.003%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, MM phono input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the MM phono input. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -50dBrA, or 0.3%, while the third-order modulation products, at 17kHz and 20kHz, are around -60dBrA, or 0.1%.

Squarewave response (10kHz)

Above is the 10kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the HTA200’s slew-rate performance. Rather, it should be seen as a qualitative representation of the HTA200’s extremely restricted bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. In this case, because of the brickwall-type cutoff at 20kHz, we find only the 10kHz fundamental sinewave, with all the upper harmonics filtered out.

Squarewave response (1kHz)

Above is the 1kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms, We can see significant overshoot/undershoot in the corners of the squarewave, a consequence of the HTA200’s limited bandwidth.

Damping factor vs. frequency (20Hz to 20kHz)

The final graph above is the damping factor as a function of frequency. We can see here a constant damping factor right around 100, from 20Hz to 20kHz.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 01 February 2024

Link: reviewed by Dennis Burger on SoundStage! Access on February 1, 2024

General Information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The Technics Grand Class SU-GX70 was conditioned for one hour at 1/8th full rated power (~5W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The SU-GX70 offers two line-level analog inputs (RCA), one moving-magnet (MM) phono input (RCA), one pair of preamp outputs (RCA), one digital coaxial (RCA) and two optical (TosLink) S/PDIF inputs, one USB digital inputs, two pairs of speaker-level outputs and one headphone output over 1/4″ TRS connector. For the purposes of these measurements, the following inputs were evaluated: digital coaxial, analog line-level, as well as phono.

The SU-GX70 is a sophisticated device that digitizes all incoming signals and can apply DSP for various functions. An “initialization” was performed before any measurements were made, to ensure that any room EQ DSP had been cleared. Unless otherwise stated, Pure Amplification was turned on, MQA off, and LAPC off, although comparisons between the on and off effects of these functions can be seen in this report.

Most measurements were made with a 2Vrms line-level analog input, 5mVrms MM input and 0dBFS digital input. The volume control is variable from -99dB to 0dB. The signal-to-noise (SNR) measurements were made with the default input signal values but with the volume set to achieve the rated output power of 40W (8 ohms). For comparison, on the line-level input, a SNR measurement was also made with the volume at maximum.

Based on the high accuracy and repeatability of the left/right volume channel matching (see table below), the SU-GX70 volume control operates in the digital domain. The SU-GX70 offers 1dB volume steps ranging from -99dB to -54dB, then 0.5dB steps from -53.5dB to 0dB. Overall range is -59.3dB to +39.6dB (line-level input, speaker output).

Because the SU-GX70 is a digital amplifier technology that exhibits considerable noise above 20kHz (see FFTs below), our typical input bandwidth filter setting of 10Hz-90kHz was necessarily changed to 10Hz-22.4kHz for all measurements, except for frequency response and for FFTs. In addition, THD versus frequency sweeps were limited to 6kHz to adequately capture the second and third signal harmonics with the restricted bandwidth setting.

Volume-control accuracy (measured at speaker outputs): left-right channel tracking

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Technics for the SU-GX70 compared directly against our own. The published specifications are sourced from Technics’ website, either directly or from the manual available for download, or a combination thereof. With the exception of frequency response, where the Audio Precision bandwidth was extended to 250kHz, assume, unless otherwise stated, 10W into 8 ohms and a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

Our primary measurements revealed the following using the line-level analog input and digital coaxial input (unless specified, assume a 1kHz sinewave at 2Vrms or 0dBFS, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the SU-GX70 was able to sustain 105W into 4 ohms (~6% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (10.5W) for five seconds, for five continuous minutes without inducing a fault or the initiation of a protective circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the top of the SU-GX70 was only slightly warm to the touch.

Our primary measurements revealed the following using the phono-level input, MM configuration (unless specified, assume a 1kHz 5mVrms sinewave input, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

Our primary measurements revealed the following using the analog input at the headphone output (unless specified, assume a 1kHz sinewave, 1Vrms output, 300 ohms loading, 10Hz to 22.4kHz bandwidth):

Frequency response (8-ohm loading, line-level input)

In our frequency-response plots above (relative to 1kHz), measured across the speaker outputs at 10W into 8 ohms, the SU-GX70 is nearly flat within the audioband (20Hz to 20kHz). At the extremes the SU-GX70 is -0.1dB at 20Hz and +0.5dB down at 20kHz. There’s a rise in the frequency response above 20kHz, where we see +2.2dB just past 40kHz, which is a result of the digital amplifier and its high output impedance at high frequencies. Into a 4-ohm load (see RMS level vs. frequency vs load impedance graph below), the response is essentially flat at and above 20kHz. The -3dB point was also explored and found to be at roughly 46kHz, exactly where it was measured for a 24-bit/96kHz digital input signal (see “Frequency response vs. input type chart” below). In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Frequency response (8-ohm loading, line-level input, bass and treble controls)

Above is a frequency response (relative to 1kHz) plot measured at the speaker-level outputs into 8 ohms, with the bass and treble controls set to maximum (blue/red plots) and minimum (purple/green plots). We see that for the bass and treble controls, roughly +/- 5dB of gain/cut is available at 20Hz/20kHz.

Phase response (8-ohm loading, line-level input)

Above are the phase-response plots from 20Hz to 20kHz for the line-level input, measured across the speaker outputs at 10W into 8 ohms. The SU-GX70 does not invert polarity and exhibits at worst, 20 degrees (at 20Hz) of phase shift within the audioband.

Frequency response vs. input type (8-ohm loading, left channel only)

The chart above shows the SU-GX70’s frequency response (relative to 1kHz) as a function of input type measured across the speaker outputs at 10W into 8 ohms. The green trace (overlapping the purple trace) is the same analog input data from the previous graph. The blue trace is for a 16-bit/44.1kHz dithered digital input signal from 5Hz to 22kHz using the coaxial input, and the purple trace is for a 24/96 dithered digital input signal from 5Hz to 48kHz, while the pink trace is for a 24/192 dithered digital input signal from 5Hz to 96kHz. The 16/44.1 data exhibits brickwall-type filtering, with a -3dB at 21.1kHz. The 24/96 (and analog input) and 24/192 kHz data yielded -3dB points at 46.8kHz and 92.9kHz respectively. The analog data looks nearly identical to the 24/96 digital data, which is evidence for the SU-GX70 sampling incoming analog signals at 96kHz. 

Frequency response vs. MQA (16/44.1)

The chart above shows the SU-GX70’s frequency response (relative to 1kHz) for a 16/44.1 dithered digital input signal from 5Hz to 22kHz using the coaxial input, with MQA turned on. We find no difference in the measured frequency response for 16/44.1 data input whether MQA is turned on or off.

Frequency response (8-ohm loading, MM phono input)

The chart above shows the frequency response (relative to 1kHz) for the MM phono input without (blue/red) and with (purple/green) the subsonic filter enabled. What is shown is the deviation from the RIAA curve, where the input signal sweep is EQ’d with an inverted RIAA curve supplied by Audio Precision (i.e., zero deviation would yield a flat line at 0dB). We see a maximum deviation of about +0.5/-0.2dB (150Hz and 20kHz/20Hz) from 20Hz to 20kHz. With the subsonic filter engaged, we find the -3dB point at 20Hz.

Phase response (MM input)

Above is the phase response plot from 20Hz to 20kHz for the MM phono input without (blue/red) and with (purple/green) the subsonic filter enabled, measured across the speaker outputs at 10W into 8 ohms. The SU-GX70 does not invert polarity. For the phono input, since the RIAA equalization curve must be implemented, which ranges from +19.9dB (20Hz) to -32.6dB (90kHz), phase shift at the output is inevitable. Here we find a worst case of about +80 degrees at 20Hz without the subsonic filter and +160 degrees with the filter.

Digital linearity (16/44.1 and 24/96 data)

The chart above shows the results of a linearity test for the coaxial digital input for both 16/44.1 (blue/red) and 24/96 (purple/green) input data, measured at the line-level output of the SU-GX70. The digital input is swept with a dithered 1kHz input signal from -120dBFS to 0dBFS, and the output is analyzed by the APx555. The ideal response would be a straight flat line at 0dB. Both data were essentially perfect as of -100dBFS down to 0dBFS. At -120dBFS, the 16/44.1 data were only +2dB (left) and +4dB (right) above reference, while the 24/96 data were within +1dBFS.

Impulse response (24/44.1 data)

The graph above shows the impulse response for the SU-GX70 with MQA turned off (blue) and MQA turned on (purple), fed to the coaxial digital input, measured at the line level output, for a looped 24/44.1 test file that moves from digital silence to full 0dBFS (all “1”s), for one sample period then back to digital silence. We find a reconstruction filter that adheres to a typical symmetrical sinc function. There appears to be no difference in the impulse response with MQA on or off through the coaxial input.

J-Test (coaxial, MQA off)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the SU-GX70 with MQA turned off. J-Test was developed by Julian Dunn the 1990s. It is a test signal—specifically, a -3dBFS undithered 12kHz squarewave sampled (in this case) at 48kHz (24 bits). Since even the first odd harmonic (i.e., 36kHz) of the 12kHz squarewave is removed by the bandwidth limitation of the sampling rate, we are left with a 12kHz sinewave (the main peak). In addition, an undithered 250Hz squarewave at -144dBFS is mixed with the signal. This test file causes the 22 least significant bits to constantly toggle, which produces strong jitter spectral components at the 250Hz rate and its odd harmonics. The test file shows how susceptible the DAC and delivery interface are to jitter, which would manifest as peaks above the noise floor at 500Hz intervals (e.g., 250Hz, 750Hz, 1250Hz, etc.). Note that the alternating peaks are in the test file itself, but at levels of -144dBrA and below.  The test file can also be used in conjunction with artificially injected sinewave jitter by the Audio Precision, to show how well the DAC rejects jitter.

The coaxial input exhibits low-level rises (-135dBrA) in the noise floor within the audioband at 6.5kHz and 13kHz. This is a good J-Test result, indicating that SU-GX70 DAC should yield good jitter immunity.

J-Test (optical, MQA off)

The chart above shows the results of the J-Test test for the optical digital input measured at the line-level output of the SU-GX70. The optical input yielded essentially the same result compared to the coaxial input.

J-Test (coaxial, MQA on)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the SU-GX70 with MQA turned on. The result is similar to the one with MQA turned off, only slightly improved, with the rises in the noise floor no longer visible.

J-Test with 100ns of injected jitter (coaxial, MQA off)

Both the coaxial and optical inputs were also tested for jitter immunity by injecting artificial sinewave jitter at 2kHz, which would manifest as sidebands at 10kHz and 14kHz without any jitter rejection. Jitter immunity proved exceptional, with no visible sidebands at the 100ns jitter level, and only a spurious peak at 2kHz at the -135dBrA level. The coaxial input is shown, but both performed the same.

Wideband FFT spectrum of white noise and 19.1kHz sine-wave tone (coaxial input, MQA off)

The chart above shows a fast Fourier transform (FFT) of the SU-GX70’s line-level output with white noise at -4dBFS (blue/red) and a 19.1 kHz sinewave at 0dBFS fed to the coaxial digital input, sampled at 16/44.1, with MQA turned off. The steep roll-off around 20kHz in the white-noise spectrum shows the behavior of the SU-GX70’s reconstruction filter. There are low-level aliased image peaks within the audioband at around 2kHz and 13kHz, at or near -120dBrA. The primary aliasing signal at 25kHz is at -95dBrA, while the second and third distortion harmonics (38.2, 57.3kHz) of the 19.1kHz tone are at -80 and -60dBrA.

Wideband FFT spectrum of white noise and 19.1kHz sine-wave tone (coaxial input, MQA on)

The chart above shows a fast Fourier transform (FFT) of the SU-GX70’s line-level output with white noise at -4dBFS (blue/red) and a 19.1 kHz sinewave at 0dBFS fed to the coaxial digital input, sampled at 16/44.1, with MQA turned on. The steep roll-off around 20kHz in the white-noise spectrum shows the behavior of the SU-GX70’s reconstruction filter. There are low-level aliased image peaks within the audioband at around 2kHz and 7kHz, at -120dBrA.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 5Hz to 100kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Here we find that between 20Hz and 6kHz, the deviations between no load and 4 ohms are around 0.45dB, but at high frequencies, the differences are larger, at about 1dB at 20kHz. This is a relatively poor result, and an indication of a relatively high output impedance, or low damping factor. When a real speaker is used, deviations are within around 0.4dB throughout the audioband.

RMS level vs. frequency (1W, left channel only, real speaker, LPAC on and off)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8ohms or 2.83Vrms) as a function of frequency, for the analog line-level input swept from 20Hz to 20kHz. Both plots are for the Focal Chora 806 speaker, with (purple) and without (blue) LAPC enbaled. The SU-GX70 provides a feature called Load Adaptive Phase Calibration (LAPC). This feature measures the outputs of the amplifier while the speakers are connected using test tones to establish a correction curve to deal with the amplifier’s inherently high output impedance at high frequencies. The theoretical goal is to achieve a flat frequency response for the user’s speakers when LAPC is enabled. We can see here that the purple trace is not flat, but closer to ideal compared to when LAPC is disabled. When LAPC is disabled, deviations reach about 0.35dB, while only 0.15dB with LAPC enabled.

THD ratio (unweighted) vs. frequency vs. output power

The chart above shows THD ratios at the output into 8 ohms as a function of frequency for a sinewave stimulus at the analog line-level input. The blue and red plots are for left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange just under 40W. The power was varied using the volume control. At 1W, THD ratios are fairly constant and range from 0.02% at 20Hz, down to 0.01% from 40Hz to 6kHz. At 10W, THD ratios are as high as 0.3% at 20Hz, with a steady decline to 0.01% at 6kHz. At nearly 40W, THD ratios are as high as 0.6% at 20Hz, with a steady decline to 0.02% at 6kHz.

THD ratio (unweighted) vs. frequency at 10W (MM input)

The chart above shows THD ratio as a function of frequency plots for the MM phono input measured across an 8-ohm load at 10W. The input sweep is EQ’d with an inverted RIAA curve. The THD values vary from around 0.3% at 20Hz, then a steady decline down to 0.015% at 6kHz.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the output of the SU-GX70 as a function of output power for the analog line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). Both data sets track closely except for maximum power, with the 4-ohm data slightly outperforming the 8-ohm data at lower power. THD ratios range from as low as 0.0025% at 0.5-1W, up to 0.07% (8-ohm) and 0.2% (4-ohm) at the “knees” at just below 50W and 90W, respectively. The 1% THD marks were reached at just past 50W (8 ohms) and just shy of 100W (4 ohms).

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the SU-GX70 as a function of output power for the line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). Both data sets track closely except for maximum power, with the 4-ohm data slightly outperforming the 8-ohm data at lower power. Overall, THD+N values for both loads ranged from 0.05% at 50mW, down to near 0.01% at 3-5W, then up to the “knees,” as described in the caption for the chart directly above.

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only)

The chart above shows THD ratios measured at the output of the SU-GX70 as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yields 10W at the output into 8 ohms (and roughly 20W into 4 ohms, and 40W into 2 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. We find roughly the same THD values of 0.02% from 1kHz to 6kHz for the 8- and 4-ohm data. From 20Hz to 1kHz, there is a roughly 5dB increase in THD every time the load is halved. However, even into a 2-ohm load, which the SU-GX70 is not designed to drive, THD ratios range from 0.3% at 20Hz, down to 0.03% from 1kHz to 6kHz.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the SU-GX70 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). In general, the measured THD ratios for the real speakers were close to the 8-ohm resistive load, hovering between 0.01 and 0.02% from 100Hz to 6kHz. The two-way Focal yielded the highest THD values (0.2% at 20Hz) at very low frequencies.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the SU-GX70 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All IMD results are similar, hovering from 0.03 to 0.015% across the measured frequency range.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the SU-GX70 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). Between 40Hz and 60Hz, all result are essentially identical, around -81dB. Above 60Hz, the highest IMD ratios are associate with the Paradigm speakers, rising up to -74dB from 100Hz to 250Hz. All IMD results are essentially identical, from 0.05% from 40Hz to 400Hz, then 0.025% from 500Hz to 1kHz.

FFT spectrum – 1kHz (line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input. We see that the signal’s second (2kHz) and third (3kHz) harmonics are at a relatively high -80dBrA, or 0.01%, while subsequent signal harmonics are at and below -90dBrA, or 0.003%. Since the SU-GX70 uses a switching power supply, there are no obvious peaks at 60Hz or subsequent harmonics. There are, however, several significant noise peaks (as high as -65dB, or 0.06%) that are likely a result of IMD products between the signal, it’s harmonics, and the high-frequency oscillator used in the class-D amplifier section. Of note is that the analyzer would ignore these peaks, which are actually larger in magnitude than the signal harmonics, when calculating THD. There is also a rise in the noise above 20kHz, characteristic of digital amplifiers. This is far from what is considered a clean FFT.

FFT spectrum – 1kHz (line-level input, Pure Amplification off)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input, but with Pure Amplification turned off. The FFT is similar to the FFT above, where Pure Amplification was turned on, except for low-level peaks (-120dBrA, or 0.0001%) that can be seen here at low frequencies that are not present in the first FFT.

FFT spectrum – 1kHz (digital input, 16/44.1 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 16/44.1. Signal harmonics are different to the analog input FFT above. The second (2kHz) harmonic is low at -115dBRa, or 0.0002%, while the third (3kHz) harmonic is much higher, at -75dBrA, or 0.02%. Subsequent signal harmonics are at and below -90dBrA, or 0.003%. The same IMD peaks can also be seen here, as high as -65dB, or 0.06%, flanking the main 1kHz signal peak.

FFT spectrum – 1kHz (digital input, 24/96 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 24/96. The FFT is very similar to the 16/44.1 input FFT above, but for a more predominant second (2kHz) signal harmonic at -95dBrA, or 0.002%.  

FFT spectrum – 1kHz (digital input, 16/44.1 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 16/44.1 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at the correct amplitude, and no other peaks above the noise floor at -130dBrA.

FFT spectrum – 1kHz (digital input, 24/96 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 24/96 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at the correct amplitude, and no other peaks above the noise floor at -135dBrA.

FFT spectrum – 1kHz (MM phono input)

Shown above is the FFT for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the MM phono input. We see the third (3kHz) signal harmonic dominating at around -75dBrA, or 0.02%. Other signal harmonics can be seen at -95dBrA, or 0.002%, and below. The most significant power-supply-related noise peaks can be seen at 60Hz at -85dBrA, or 0.006%. Higher-order power-supply-related peaks can also be seen at lower amplitudes. The same IMD peaks can also be seen here, as high as -65dB, or 0.06%, flanking the main 1kHz signal peak.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the analog line-level input. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peak is the third (150Hz) signal harmonic at a high -55dBrA, or 0.02%. Several other signal-related and IMD peaks can be seen throughout at -70dBrA, or 0.03%, and below.

FFT spectrum – 50Hz (MM phono input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the MM phono input. The X axis is zoomed in from 40 Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The 60Hz power supply fundamental can be seen at -90dBrA, or 0.003%. The most predominant (non-signal) peak is the third (150Hz) signal harmonic at a high -55dBrA, or 0.02%. Several other signal-related and IMD peaks can be seen throughout at -70dBrA, or 0.03%, and below.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the analog line-level input. The input RMS values are set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at nearly -80dBrA, or 0.01%, while the third-order modulation products, at 17kHz and 20kHz, are at roughly the same level.

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the SU-GX70 with the APx 32-tone signal applied to the input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and are below the -90dBrA, or 0.003%, level.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 16/44.1)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 16/44.1. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -100dBRa, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are around -80dBrA, or 0.01%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 16/44.1, MQA on)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 16/44.1, with MQA turned on. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -100dBRa, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are around -80dBrA, or 0.01%. This is essentially the same result as with the FFT with MQA turned off.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 24/96)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 24/96. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -100dBRa, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are around -80dBrA, or 0.01%. This is essentially the same result as with the 16/44.1 IMD FFT.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, MM phono input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the MM phono input. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -95dBRa, or 0.002%, while the third-order modulation products, at 17kHz and 20kHz, are just below -80dBrA, or 0.01%.

Square-wave response (10kHz)

Above is the 10kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the SU-GX70’s slew-rate performance. Rather, it should be seen as a qualitative representation of the SU-GX70’s mid-tier bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. In this case, because of the digital nature of the amplifier, we see a 400kHz switching frequency (see 1MHz FFT below) riding on top of the squarewave.

Square-wave response (10kHz, restricted 500kHz bandwidth)

Above is the same 10kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms, this time with a 250kHz input bandwidth on the analyzer to filter out the 400kHz switching frequency. We can see significant over/undershoot in the corners of the squarewave, a consequence of the SU-GX70’s mid-tier bandwidth.

FFT spectrum (1MHz bandwidth)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input, with an extended 1MHz input bandwidth. This enables us to see the high-frequency noise above 20kHz reaching almost -70dBrA at 80kHz. We also see a clear peak at 400kHz, reaching just past -20dBrA, as well as its harmonics (800kHz, 1.2MHz). These peaks, as well as the noise, are a result of the digital amplifier technology used in the SU-GX70. However, they are far above the audioband—and are therefore inaudible—and so high in frequency that any loudspeaker the amplifier is driving should filter it all out anyway.

Damping factor vs. frequency (20Hz to 20kHz)

The final graph above is the damping factor as a function of frequency. We can see here the clear trend of a higher (although still poor in absolute terms) damping factor at low frequencies—around 35 from 20Hz to 3kHz, and then a decline down to 18 at 20kHz. This is a limitation of the digital amplifier technology used in the SU-GX70, and the reason Technics has incorporated their clever Load Adaptive Phase Calibration (LAPC) feature to compensate for losses into low impedances at high frequencies.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 15 November 2023

Link: reviewed by Philip Beaudette on SoundStage! Hi-Fi on November 15, 2023

General Information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The Rotel Michi X5 Series 2 was conditioned for one hour at 1/8th full rated power (~43W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The X5 Series 2 offers a multitude of inputs, both digital and analog, line-level analog pre-amp outputs, subwoofer line-level outputs and two pairs of speaker-level outputs (for biwiring). There is also a ¼″ TRS headphone output on the front panel. For the purposes of these measurements, unless otherwise stated, the following inputs were evaluated: digital coaxial 1 (RCA), analog balanced (XLR), as well as RCA phono, configured both for moving-magnet (MM) and moving-coil (MC) inputs. Comparisons were made between unbalanced (RCA) and balanced (XLR) line-level inputs, and no differences were seen in terms of THD+N (FFTs for both can be seen in this report); however, the balanced input offers 3.84dB less gain than the unbalanced inputs. Bluetooth is also offered; however, our APx555 does not currently have a Bluetooth board.

Most measurements were made with a 2Vrms line-level, 0dBFS digital input, 5mVrms MM-level, and 0.5mVrms MC-level analog input. The signal-to-noise ratio (SNR) measurements were made with the same input signal values but with the volume set to achieve the rated output power of 350W (8 ohms). For comparison, on the line-level input, a SNR measurement was also made with the volume at maximum, but with a lower input voltage to achieve the same 350W output.

Based on the accuracy and random results of the left/right volume channel matching (see table below), the X5 Series 2 volume control is likely digitally controlled but operating in the analog domain. The X5 Series 2 offers 96 volume steps. Between steps 1 and 5, step increases are 2dB, steps 6 to 18 are 1.5dB, 19 to 66 are 1 dB, 66 to 86 are 0.5dB, and volume settings 87 to 96 are 0.25dB.

Volume-control accuracy (measured at speaker outputs): left-right channel tracking

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Rotel for the Michi X5 Series 2 compared directly against our own. The published specifications are sourced from Rotel’s website, either directly or from the manual available for download, or a combination thereof. With the exception of frequency response, where the Audio Precision bandwidth is set at its maximum (DC to 1MHz), assume, unless otherwise stated, 10W into 8 ohms and a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

Our primary measurements revealed the following using the balanced line-level analog input and digital coaxial input (unless specified, assume a 1kHz sinewave at 2Vrms or 0dBFS, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the X5 Series 2 was able to sustain 700W into 4 ohms (3.6% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (70W) for five seconds, for five continuous minutes without inducing a fault or the initiation of a protective circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the sides of the X5 Series 2 were warm to the touch, but not enough to cause discomfort.

Our primary measurements revealed the following using the phono-level input, MM configuration (unless specified, assume a 1kHz sinewave at 5mVrms, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

Our primary measurements revealed the following using the phono-level input, MC configuration (unless specified, assume a 1kHz sinewave at 0.5mVrms, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

Our primary measurements revealed the following using the balanced line-level inputs at the headphone output (unless specified, assume a 1kHz sinewave 2Vrms input, 2Vrms output, 300 ohms loading, 10Hz to 22.4kHz bandwidth):

Frequency response (8-ohm loading, line-level input)

In our measured frequency response (relative to 1kHz) plot above, the X5 Series 2 is nearly flat within the audioband (-0.2dB at 20Hz, -0.05dB at 20kHz). At the extremes the X5 Series 2 is -0.7dB at 10Hz, and -0.5dB at 100kHz. These data only half corroborate Rotel’s claim of 10Hz to 100kHz (0/-0.6dB). The X5 Series 2 can be considered a high-bandwidth audio device. In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Phase response (8-ohm loading, line-level input)

Above is the phase-response plot from 20Hz to 20kHz for the balanced line-level input, measured across the speakers outputs at 10W into 8 ohms. The X5 Series 2 does not invert polarity and exhibits, at worst, +20 degrees (at 20Hz) of phase shift within the audioband.

Frequency response (treble/bass at minimum and maximum settings, 8-ohm loading, line-level input)

Above are two frequency response plots (relative to 1kHz) for the balanced line-level input, measured at 10W (8-ohms) at the speaker outputs, with the treble/balance controls set at both minimum and maximum. They show that the X5 Series 2 will provide a maximum gain/cut of approximately 12dB at 20Hz, and a maximum gain/cut of approximately 9dB at 20kHz.

Frequency response vs. input type (8-ohm loading, left channel only)

The chart above shows the X5 Series 2’s frequency response (relative to 1kHz) as a function of input type. The green trace is the same (but limited to 80kHz) analog input data from the previous graph. The blue trace is for a 16-bit/44.1kHz dithered digital input signal from 5Hz to 22kHz using the coaxial input, the purple trace is for a 24/96 dithered digital input signal from 5Hz to 48kHz, and finally pink is 24/192 from 5Hz to 96kHz. The behavior at low frequencies is the same for all the digital sample rates, as well as the analog input: -0.25dB at 20Hz. The behavior at high frequencies for all three digital sample rates is as expected, offering filtering around 22, 48, and 96kHz (half the respective sample rate). The 44.1kHz sampled input signal exhibits typical “brick-wall”-type behavior, with a -3dB point at 21kHz. The -3dB point for the 96kHz sampled data is at 46kHz, and 68kHz for the 192kHz sampled data.

Frequency response (8-ohm loading, MM input)

What is shown above is the moving-magnet (MM) phono stage’s deviation from the RIAA curve, where the input signal sweep is EQ’d with an inverted RIAA curve supplied by Audio Precision. No deviation would yield a flat line at 0dB. So the chart above shows the frequency response (relative to 1 kHz), which displays very small maximum deviations of about -0.25/-0.04dB (20Hz/20kHz) and +0.1dB (100Hz) from 20Hz to 20kHz.

Frequency response (8-ohm loading, MC input)

The chart above shows the frequency response for the phono input (MC configuration). We see essentially the same result as with the MM configuration, with the exception of the left channel deviating by +0.08dB at 4kHz.

Phase response (MM and MC phono inputs)

Above is the phase response plot from 20Hz to 20kHz for the MM phono input (MM and MC configurations behaved identically), measured across the speakers outputs at 10W into 8 ohms. For the phono input, since the RIAA equalization curve must be implemented, which ranges from +19.9dB (20Hz) to -32.6dB (90kHz), phase shift at the output is inevitable. Here we find a worst case of about +60 degrees at 20Hz and +40 degrees at 20kHz.

Digital linearity (16/44.1 and 24/96 data)

The chart above shows the results of a linearity test for the coaxial digital input (the optical input performed identically) for both 16/44.1 (blue/red) and 24/96 (purple/green) input data, measured at the line-level pre-outs of the X5 Series 2. For this test, the digital input was swept with a dithered 1kHz input signal from -120dBFS to 0dBFS, and the output was analyzed by the APx555. The ideal response would be a straight flat line at 0dB. Both digital input types performed similarly, approaching the ideal 0dB relative level at -100dBFS, then yielding perfect results to 0dBFS. At or near -120dBFS, both sample rates overshot by 3 to 5dB.

Impulse response (24/48 data)

The graph above shows the impulse response for a looped 24/44.1 test file that moves from digital silence to full 0dBFS (all “1”s) for one sample period then back to digital silence, measured at the line level pre-outs of the X5 Series 2. We can see that the X5 Series 2 utilizes a reconstruction filter that favors no pre-ringing and significant post-ringing. Since the initial pulse/peak shows a negative voltage, it's likely that the digital input inverts polarity.

J-Test (coaxial input)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the X5 Series 2. J-Test was developed by Julian Dunn the 1990s. It is a test signal—specifically, a -3dBFS undithered 12kHz squarewave sampled (in this case) at 48kHz (24bit). Since even the first odd harmonic (i.e., 36kHz) of the 12kHz squarewave is removed by the bandwidth limitation of the sampling rate, we are left with a 12kHz sinewave (the main peak). In addition, an undithered 250Hz squarewave at -144dBFS is mixed with the signal. This test file causes the 22 least significant bits to constantly toggle, which produces strong jitter spectral components at the 250Hz rate and its odd harmonics. The test file shows how susceptible the DAC and delivery interface are to jitter, which would manifest as peaks above the noise floor at 500Hz intervals (e.g., 250Hz, 750Hz, 1250Hz, etc). Note that the alternating peaks are in the test file itself, but at levels of -144dBrA and below.  The test file can also be used in conjunction with artificially injected sinewave jitter by the Audio Precision, to show how well the DAC rejects jitter.

The coaxial SPDIF X5 Series 2 input shows obvious peaks in the audioband from -95dBrA to -140dBrA. This is a poor J-Test result and an indication that the X5 Series 2’s DAC may be susceptible to jitter.

J-Test (optical input)

The chart above shows the results of the J-Test test for the optical digital input measured at the line-level output of the X5 Series 2. The results here are similar but slightly better than the coaxial input, with the highest peaks just below -100dBrA.

J-Test (coaxial, 2kHz sinewave jitter at 10ns)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the X5 Series 2, with an additional 10ns of 2kHz sinewave jitter injected by the APx555. The results are very clear, as we see the sidebands at 10kHz and 14kHz (12kHz main signal +/-2kHz jitter signal) manifest at near -70dBrA. This is a clear indication that the DAC in the X5 Series 2 has poor jitter immunity. For this test, the optical input yielded effectively the same results.

J-Test (coaxial, 2kHz sinewave jitter at 100ns)

The chart above shows the results of the J-Test test for the coaxial digital input measured at the line-level output of the X5 s2, with an additional 100ns of 2kHz sinewave jitter injected by the APx555. The results are very clear, as we see the sidebands at 10kHz and 14kHz (12kHz main signal +/-2kHz jitter signal) manifest at near -50dBrA. This is further indication that the DAC in the X5 Series 2 has poor jitter immunity. For this test, the optical input yielded effectively the same results.

Wideband FFT spectrum of white noise and 19.1kHz sinewave tone (coaxial input)

The chart above shows a fast Fourier transform (FFT) of the X5 Series 2’s line-level output with white noise at -4dBFS (blue/red) and a 19.1 kHz sinewave at 0dBFS fed to the coaxial digital input, sampled at 16/44.1. The steep roll-off around 20kHz in the white-noise spectrum shows that the X5 Series 2 uses a brick-wall-type reconstruction filter. There are no obvious aliased images within the audioband, with the exception of a small peak at -115dBrA at 15kHz. The primary aliasing signal at 25kHz is highly suppressed at -100dBrA, while the second and third distortion harmonics (38.2, 57.3kHz) of the 19.1kHz tone are at -85 and -75dBrA.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8 ohms or 2.83Vrms) as a function of frequency, for the balanced analog line-level input swept from 5Hz to 100kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Zoomed in, we can see a maximum deviation within the audioband of only about 0.15dB (at 20kHz) from 4 ohms to no load, and much less (0.05dB) within the flatter portion of the curve, which is an indication of a high damping factor, or low output impedance. The maximum variation in RMS level when a real speaker was about the same, deviating by about 0.04dB within the flat portion of the curve (100Hz to 5kHz), with the lowest RMS level, which would correspond to the lowest impedance point for the load, exhibited around 200Hz, and the highest RMS level, which would correspond to the highest impedance point for the load, at around 3-5kHz. The more significant deviations in RMS level between loads at 10kHz and 20kHz is an indication of a dip in damping factor in this frequency range. This can be seen in our damping factor graph (see last graph in the report).

THD ratio (unweighted) vs. frequency vs. output power

The graph above shows THD ratios at the output into 8 ohms as a function of frequency for a sinewave stimulus at the balanced line-level input. The blue and red plots are for the left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange at the full rated power of 350W. The power was varied using the volume control. The 10W and 1W data exhibited effectively the same THD values, and remained commendably flat with the entire audioband, between 0.006% and 0.01%. At the full rated power of 350W, THD values were remarkably close to the 1/10W data, ranging from 0.006% to 0.01% up to 10kHz, then up to 0.2% at 20kHz. 

THD ratio (unweighted) vs. frequency at 10W (phono input)

The graph above shows THD ratio as a function of frequency plot for the phono input measured across an 8 ohms load at 10W. The MM configuration is shown in blue/red (left/right channels) and MC in purple/green (left/right channel). The input sweep was EQ’d with an inverted RIAA curve. The THD values for the MM configuration vary from around 0.02% (20Hz) down to just above and below 0.005% (1kHz to 10kHz). The MC THD values were higher, ranging from 0.3/0.2% (20Hz, left/right channel), down to 0.005% (2kHz, left channel). Between 1kHz and 3kHz, the left channel outperformed the right channel for the MC configuration by as much as 10dB.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the output of the X5 Series 2 as a function of output power for the balanced line-level input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). The 8-ohm data outperformed the 4-ohm data by about 5-6dB, and both data sets show fairly constant THD values across measured output power levels until the “knees” at just past 300W (8 ohms) and 500W (4 ohms). THD levels for the 8-ohm data are around 0.005-0.007%, and 0.01-0.015% for the 4-ohm data. The 1% THD mark for the 8-ohm data is at 400W, and 659W for the 4-ohm data.

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the X5 Series 2 as a function of output power for the balanced line-level-input for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). Overall, THD+N values for the 8-ohm load before the “knee” ranged from around 0.1% (50mW) down to about 0.007%. The 4-ohm results were similar, but 2-4 dB worse.

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only)

The chart above shows THD ratios measured at the output of the X5 Series 2 as a function of load (8/4/2 ohms) for a constant input voltage that yields 40W at the output into 8 ohms (and roughly 80W into 4 ohms, and 160W into 2 ohms) for the balanced line-level input. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. We find increasing levels of THD from 8 to 4 to 2 ohms, with about a 5dB increase between each halving of the load, but nearly 20dB difference at 20kHz between the 8- and 2-ohm data. Overall, even with a 2-ohm load at roughly 160W, THD values ranged from as low as 0.02% through most of the audioband to 0.07% at 20kHz.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the X5 Series 2 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). Generally, THD ratios into the real speakers were similar to the resistive dummy load, which hovered between 0.01 and 0.005%. THD ratios were higher (5-10dB) at 20Hz for the two-way speaker and at 20kHz for the three-way speaker, but also lower (5-10dB) than the resistive load between 800Hz and 6kHz. This is a strong result, and shows that the X5 Series 2 will yield consistently low THD results into real-world speaker loads.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the X5 Series 2 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). Generally, THD ratios into the real spakers were similar to the resistive dummy load, which hovered around 0.005%. At lower frequencies, both speakers yielded lower IMD results (0.002-0.003%), and at higher frequencies, the 3-way speaker yielded higher IMD results than the resistive load, 0.01% at 20kHz.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the X5 Series 2 as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three results are similar enough to be judged identical, hovering around 0.02%.

FFT spectrum – 1kHz (XLR line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the balanced line-level input. We see that the signal’s second harmonic, at 2kHz, is at -85dBrA or 0.005%, and around -100dBrA, or 0.001%, at the fourth (4kHz). The sixth (6kHz) and eighth (8kHz) harmonics follow at -110 and -120dBrA, or 0.0003 and 0.0001%. Below 1kHz, we see peaks from power-supply noise artifacts at 60Hz (around -100dBrA or 0.001%), and then the odd harmonics (180Hz, 300Hz, 420Hz) dominating at between -90dBrA, or 0.003%, and -100dBrA, or 0.001%.

FFT spectrum – 1kHz (RCA line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the unbalanced line-level input. We see effectively the same results as with the XLR balanced input FFT above.

FFT spectrum – 1kHz (digital input, 16/44.1 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 16/44.1. We see effectively the same results as with the analog balanced input FFT above.

FFT spectrum – 1kHz (digital input, 24/96 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the coaxial digital input, sampled at 24/96. We see essentially the same signal and noise harmonic profile within the audioband as with the 16/44.1 sampled input.

FFT spectrum – 1kHz (digital input, 16/44.1 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 16/44.1 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at the correct amplitude, along with the 60Hz power-supply peak (-110dBrA) with a multitude of subsequent harmonics at and below -95dBrA.

FFT spectrum – 1kHz (digital input, 24/96 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 24/96 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at the correct amplitude, along with the 60Hz power-supply peak (-110dBrA) with a multitude of subsequent harmonics at and below -95dBrA.

FFT spectrum – 1kHz (MM phono input)

Shown above is the FFT for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the phono input, configured for MM. We see the signal harmonic profile is similar to the line-level balanced input, with the second harmonic dominating at -85dBrA, or 0.005%. The noise-related peaks are at and below the -80dBrA level, or 0.01%.

FFT spectrum – 1kHz (MC phono input)

Shown above is the FFT for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the phono input, configured for MC. The main signal harmonic is again the second harmonic (2 kHz) at around -80dBrA or 0.01%. What dominates the FFT are the noise peaks, which is due to the very high gain required for an MC cartridge, and are as high as almost -55dBrA, or around 0.2% at 60, 180, and 300Hz.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the balanced line-level input. The X axis is zoomed in from 40 Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peaks are that of the signal’s second (100Hz) harmonic at -85dBrA, or 0.005%, and the fifth power-supply noise harmonic (300Hz) at -90dBrA, or 0.003%.

FFT spectrum – 50Hz (MM phono input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the phono input configured for MM. The most predominant (non-signal) peaks are that of the signal’s second (100Hz) harmonic at -85dBrA, or 0.005%, and the primary (60Hz), third (180Hz), and fifth (300Hz) power-supply noise harmonics at nearly the same level.

FFT spectrum – 50Hz (MC phono input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the phono input configured for MC. The most predominant (non-signal) peaks are the primary (60Hz), third (180Hz), and fifth (300Hz) power-supply noise harmonics at around -60dBrA, or 0.1%. The second (2kHz) signal harmonic is just below -80dBrA, or 0.01%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8 ohm load at 10W for the balanced line-level input. The input RMS values are set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -90dBRA, or 0.003%, while the third-order modulation products, at 17kHz and 20kHz, are lower, at -100dBrA, or 0.001%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 16/44.1)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 16/44.1. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -90dBRA, or about 0.003%, while the third-order modulation products, at 17kHz and 20kHz, are at -100dBrA, or 0.001%. We also see the main aliased peaks at 25.1kHz and 26.1kHz around -105dBrA.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 24/96)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital coaxial input at 24/96. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -90dBRA, or about 0.003%, while the third-order modulation products, at 17kHz and 20kHz, are at -100dBrA, or 0.001%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, MM phono input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the phono input configured for MM. Here we find close to the same result as with the balanced line-level analog input. The second order 1kHz peak is at -90dBrA, or 0.003%, while the third-order peaks are at -100dBrA, or 0.001%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, MC phono input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the phono input configured for MC. The second order 1kHz peak is at -80dBrA, or 0.01%, while the third-order peaks are at -100dBrA, or 0.001%. 

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the X5 Series 2 with the APx 32-tone signal applied to the input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and below the -110dBrA, or 0.0003%, level. The peaks at lower frequencies that reach the -90dBrA level are not IMD products but power-supply-related noise peaks.

Square-wave response (10kHz)

Above is the 10kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the X5 Series 2’s slew-rate performance. Rather, it should be seen as a qualitative representation its extended bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. The X5 Series 2’s reproduction of the 10kHz square wave is very clean, with only very mild softening in the edges.

Damping factor vs. frequency (20Hz to 20kHz)

The final graph above is the damping factor as a function of frequency. Both channels show a relatively steady decline in damping factors from low to high frequencies, and track very closely to one another. From 20Hz to 2kHz, damping factors ranged from 600 to just shy of 500, then a decline to 125 at 20kHz. These are strong damping factor results.

Diego Estan
Electronics Measurement Specialist

Details

Parent Category: Products

Category: Amplifier Measurements

Created: 01 November 2023

Link: reviewed by Dennis Burger on SoundStage! Access on November 1, 2023

General Information

All measurements taken using an Audio Precision APx555 B Series analyzer.

The Ampster BT II was conditioned for one hour at 1/8th full rated power (~3W into 8 ohms) before any measurements were taken. All measurements were taken with both channels driven, using a 120V/20A dedicated circuit, unless otherwise stated.

The Ampster BT II offers one analog input (RCA), one digital optical (S/PDIF), and one Bluetooth input, plus a subwoofer output (RCA) and a pair of speaker-level outputs. For the purposes of these measurements, the following inputs were evaluated: optical (S/PDIF) and the analog line-level unbalanced (RCA) input.

Based on the different results at various volume levels of the left/right channel matching (see table below), the Ampster BT II volume control is likely operating in the analog domain but is digitally controlled. The volume control offers a total range from -62dB to +28dB with step sizes ranging from 1 to 4dB.

Most measurements were made with a 1.1Vrms line-level analog input, or a 0dBFS digital input. We avoided our typical standard 2Vrms analog input level because this caused severe distortion at the input. We found that a 1.5Vrms analog input yielded 1% THD at the output of the Ampster BT II, while maintaining a modest 1W into 8 ohms. This could be considered a significant design flaw, as most modern DAC outputs are between 2 and 2.2Vrms, with some even exceeding 4Vrms for a 0dBFS digital input. This means that a typical DAC connected to the Ampster BT II’s analog input, while decoding a digital track that is recorded with little digital headroom (peaks at or approaching 0dBFS, which are common for modern music), would cause the Ampster BT II to clip regardless of volume position and load.

The signal-to-noise ratio (SNR) measurements were made with the same input signal values but with the volume set to achieve the rated output power of 25W (8 ohms). For comparison, on the line-level input, a SNR measurement was also made with the volume at maximum, but with a lower input voltage to achieve the same 25W output.

Because the Ampster BT II uses digital amplifier technology that exhibits considerable noise above 20kHz (see FFTs below), our typical input bandwidth filter setting of 10Hz–90kHz was necessarily changed to 10Hz–22.4kHz for all measurements, except for frequency response and for FFTs. In addition, THD versus frequency sweeps were limited to 6kHz to adequately capture the second and third signal harmonics with the restricted-bandwidth setting.

Volume-control accuracy (measured at speaker outputs): left-right channel tracking

Published specifications vs. our primary measurements

The table below summarizes the measurements published by Tangent for the Ampster BT II compared directly against our own. The published specifications are sourced from Tangent’s website, either directly or from the manual available for download, or a combination thereof. With the exception of frequency response, where the Audio Precision bandwidth was set at its maximum (DC to 1MHz), assume, unless otherwise stated, a measurement input bandwidth of 10Hz to 22.4kHz, and the worst-case measured result between the left and right channels.

Our primary measurements revealed the following using the analog/optical input (unless specified, assume a 1kHz sinewave at 1.1Vrms or 0dBFS, 10W output, 8-ohm loading, 10Hz to 22.4kHz bandwidth):

For the continuous dynamic power test, the Ampster BT II was able to sustain about 52W (5% THD) into 4 ohms using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (5.2W) for five seconds, for five continuous minutes without inducing a fault or the initiation of a protective circuit. This test is meant to simulate sporadic dynamic bass peaks in music and movies. During the test, the sides of the BT II were only slightly warm to the touch.

Frequency response (8-ohm loading, line-level input)

In our measured frequency-response (relative to 1kHz) chart above, the Ampster BT II is nearly flat at the low end of the audioband (-0.1dB at 20Hz), but deviates from flat at high frequencies (+2.5dB at 20kHz). The -3dB point is just past 60kHz. In the graph above and most of the graphs below, only a single trace may be visible. This is because the left channel (blue or purple trace) is performing identically to the right channel (red or green trace), and so they perfectly overlap, indicating that the two channels are ideally matched.

Phase response (8-ohm loading, line-level input)

Above is the phase response plot from 20Hz to 20kHz for the analog input. The Ampster BT II does not invert polarity and exhibits a about +10 degrees of phase shift at 20Hz, and less than -5 degrees at 20kHz.

Frequency response vs. input type (8-ohm loading, left channel only)

The chart above shows the Ampster BT II’s frequency response (left channel only) as a function of input type. The green trace is the same analog input data from the previous graph. The blue trace is for a 16bit/44.1kHz dithered digital input signal from 5Hz to 22kHz, the purple trace is for a 24/96 dithered digital input signal from 5Hz to 48kHz, and finally pink is 24/192 from 5Hz to 96kHz. The behavior at low frequencies is the same across digital input types (and the analog input)— -1.5dB at 5Hz. The behavior nearing 20kHz for the 16/44.1 digital input is a brick-wall-type attenuation, with a -3dB point at 21kHz, but +0.45dB at 17-18kHz. The -3dB point for the 24/96 data is at 47kHz, and 56kHz for the 24/192 data.

Frequency response (bass and treble controls, line-level input)

Above are two frequency-response plots (relative to 1kHz) for the analog input, measured at 10W (8-ohm loading) at the speaker outputs, with the treble and bass controls set at both minimum and maximum. They show that the Ampster BT II will provide a maximum gain/cut of approximately 5dB centered around 150Hz and 9-10kHz. Due to the Ampster BT II’s inherent rise in frequency response at high frequencies, with the treble set to maximum, we measured +8dB at 20kHz.

Frequency response (subwoofer output)

Above is the frequency response (relative to 20Hz) plot for the analog input, measured at the line-level subwoofer output. The -3dB point is near 600Hz.

Digital linearity (16/44.1 and 24/96 data)

The chart above shows the results of a linearity test for the optical digital input for both 16/44.1 (blue/red) and 24/96 (purple/green) input data, measured at the speaker outputs of the Ampster BT II. For this test, the digital input is swept with a dithered 1kHz input signal from -120dBFS to 0dBFS, and the output is analyzed by the APx555. The ideal response would be a straight flat line at 0dB. The 16/44.1 data overshot the ideal output signal amplitude by about 10dB at -120dBFS, but yielded perfectly flat results from -90dBFS to 0dBFS. The 24/96 data undershot by 10dB at -110dBFS (left) and overshot by 10dB at -120dBFS, but yielded perfectly flat results from -80dBFS to 0dBFS. Interestingly, the 16/44.1 data outperformed the 24/96 data, because that usually does not happen.

Impulse response (24/44.1 data)

The graph above shows the impulse response for a looped 24/44.1 test file that moves from digital silence to full 0dBFS (all “1”s) for one sample period then back to digital silence, measured at the speaker outputs (10W 8-ohm). We can see that the Ampster BT II utilizes a reconstruction filter with minimal pre-ringing and significant post-ringing.

J-Test (optical)

The chart above shows the results of the J-Test test for the optical digital input measured at the speaker outputs (10W into 8-ohms) of the Ampster BT II. J-Test was developed by Julian Dunn in the 1990s. It is a test signal—specifically, a -3dBFS undithered 12kHz squarewave sampled (in this case) at 48kHz (24 bits). Since even the first odd harmonic (i.e., 36kHz) of the 12kHz squarewave is removed by the bandwidth limitation of the sampling rate, we are left with a 12kHz sinewave. In addition, an undithered 250Hz squarewave at -144dBFS is mixed with the signal. This test file causes the 22 least-significant bits to constantly toggle, which produces strong jitter spectral components at the 250Hz rate and its odd harmonics. The test file shows how susceptible the DAC and delivery interface are to jitter, which would manifest as peaks above the noise floor at 500Hz intervals (e.g., 250Hz, 750Hz, 1250Hz, etc). Note that the alternating peaks are in the test file itself, but at levels of -144dBFS and below.  The test file can also be used in conjunction with artificially injected sinewave jitter by the Audio Precision, to show how well the DAC rejects jitter.

We see several and clear peaks in the audioband at -105 to 130dBFS. This is an average-to-mediocre J-Test result, and an indication that the Ampster BT II DAC may have poor jitter immunity.

J-Test with 10ns of injected jitter (optical)

The chart above shows the results of the J-Test test for the optical digital input measured at the speaker outputs (10W into 8 ohms) of the Ampster BT II, with an additional 10ns of 2kHz sinewave jitter injected by the APx555. The results are very clear, as we see the sidebands at 10kHz and 14kHz (12kHz main signal +/- 2kHz jitter signal) manifest at -70dBrA. This indicates that the DAC in the Ampster BT II has poor jitter immunity.

J-Test with 100ns of injected jitter (optical)

The chart above shows the results of the J-Test test for the optical digital input measured at the speaker outputs (10W into 8-ohms) of the Ampster BT II, with an additional 100ns of 2kHz sinewave jitter injected by the APx555. The results are very clear again, as we see the sidebands at 10kHz and 14kHz (12kHz main signal +/- 2kHz jitter signal) manifest at -50dBrA. This is yet another indication that the DAC in the Ampster BT II has poor jitter immunity.

Wideband FFT spectrum of white noise and 19.1kHz sine-wave tone (optical input)

The chart above shows a fast Fourier transform (FFT) of the Ampster BT II’s speaker outputs (10W into 8-ohms) with white noise at -4 dBFS (blue/red) and a 19.1 kHz sinewave at -1dBFS, both fed to the optical digital input, sampled at 16/44.1. The sharp roll-off above 20kHz in the white-noise spectrum shows the implementation of a brick-wall-type reconstruction filter. There are several low-level aliased image peaks in the audioband at -90dBrA and below. The main 25kHz alias peak is at -45dBrA.

RMS level vs. frequency vs. load impedance (1W, left channel only)

The chart above shows RMS level (relative to 0dBrA, which is 1W into 8 ohms or 2.83Vrms) as a function of frequency, for the analog input swept from 10Hz to 100kHz. The blue plot is into an 8-ohm load, the purple is into a 4-ohm load, the pink plot is an actual speaker (Focal Chora 806, measurements can be found here), and the cyan plot is no load connected. The chart below . . .

. . . is the same but zoomed in to highlight differences. Zoomed in, we can see a maximum deviation within the audioband of nearly 0.4dB from 4 ohms to no load, which is an indication of an average damping factor, or average output impedance. The maximum variation in RMS level when a real speaker was used is less at about 0.2dB within the flat portion of the response.

THD ratio (unweighted) vs. frequency vs. output power

The chart above shows THD ratios at the output into 8 ohms as a function of frequency (20Hz to 6kHz) for a sinewave stimulus at the analog input. The blue and red plots are for the left and right channels at 1W output into 8 ohms, purple/green at 10W, and pink/orange at 25W (rated power). The power was varied using the volume control. Between 20Hz and 200Hz, all three THD plots are relatively flat and similar, hovering around 0.04% to 0.1%. From 200Hz to 6kHz, the 1W and 10W data ranged from 0.04% to 0.2% (left, 10W at 6kHz), while the 25W THD data yielded higher results, from 0.1% to 0.5% at 6kHz.

THD ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD ratios measured at the output of the Ampster BT II as a function of output power for the analog line level-input, for an 8-ohm load (blue/red for left/right channels) and a 4-ohm load (purple/green for left/right channels). Although there are many fluctuations below the “knees” (roughly 25W into 8 ohms and about 45W into 4 ohms), both the 4-ohm and 8-ohm data are relatively close, from 0.2% down to 0.02%. It’s the right channel that generally outperformed the left channel, by as much as 10dB. The exception is the right channel into 4 ohms from 4W to 20W, where THD ratios reached 0.3%, 10-15dB higher than the other data at these power levels. The 1% THD levels were reached at 28W and 49-50W into 8 and 4 ohms.

THD+N ratio (unweighted) vs. output power at 1kHz into 4 and 8 ohms

The chart above shows THD+N ratios measured at the output of the Ampster BT II as a function of output power for the analog line-level input, for an 8-ohm load (blue/red for left/right) and a 4-ohm load (purple/green for left/right). The plots are virtually identical to the THD vs. output power plot above, which means, even at low power levels, it’s the THD ratios that dominate, and the Ampster BT II is a relatively quiet amplifier.

THD ratio (unweighted) vs. frequency at 8, 4, and 2 ohms (left channel only)

The chart above shows THD ratios measured at the output of the Ampster BT II as a function of load (8/4/2 ohms) for a constant input voltage that yields 10W at the output into 8 ohms (and roughly 20W into 4 ohms, and 40W into 2 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the 4-ohm load the purple trace, and the 2-ohm load the pink trace. We find identical THD ratios between the 8- and 4-ohm loads, hovering around 0.05%, and up to 0.15% at 6kHz. Since the Ampster BT II is not designed to drive 2-ohm loads, predictably, THD ratios were higher into 2 ohms, at 0.1% to 0.2%. Nevertheless, the Ampster BT II was stable into 2 ohms, and did not shut down due to a protection circuit.

THD ratio (unweighted) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows THD ratios measured at the output of the Ampster BT II as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three results are similar enough to be judged identical, yielding THD ratios around 0.05% from 20Hz to 6kHz. The exception is the two-way speaker at 20kHz, which typically yields higher THD results in most amps, here at 0.25%. While a strong result in this test is one where the real speaker THD ratios are very close to the THD ratios into a dummy resistive load, as seen here, in this case, since THD results are already high into a dummy load, this result should be taken with a grain of salt.

IMD ratio (CCIF) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows intermodulation distortion (IMD) ratios measured at the output of the Ampster BT II as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here the CCIF IMD method was used, where the primary frequency is swept from 20kHz (F1) down to 2.5kHz, and the secondary frequency (F2) is always 1kHz lower than the primary, with a 1:1 ratio. The CCIF IMD analysis results are the sum of the second (F1-F2 or 1kHz) and third modulation products (F1+1kHz, F2-1kHz). The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three results are similar enough to be judged identical, yielding relatively high IMD ratios from around 0.05% up to 0.1%.

IMD ratio (SMPTE) vs. frequency into 8 ohms and real speakers (left channel only)

The chart above shows IMD ratios measured at the output of the Ampster BT II as a function of frequency into an 8-ohm load and two different speakers for a constant output voltage of 2.83Vrms (1W into 8 ohms) for the analog line-level input. Here, the SMPTE IMD method was used, where the primary frequency (F1) is swept from 250Hz down to 40Hz, and the secondary frequency (F2) is held at 7kHz with a 4:1 ratio. The SMPTE IMD analysis results consider the second (F2 ± F1) through the fifth (F2 ± 4xF1) modulation products. The 8-ohm load is the blue trace, the purple plot is a two-way speaker (Focal Chora 806, measurements can be found here), and the pink plot is a three-way speaker (Paradigm Founder Series 100F, measurements can be found here). All three results are similar enough to be judged identical, yielding relatively high IMD ratios around 0.2% across the frequency sweep.

FFT spectrum – 1kHz (line-level input)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the analog line-level input. We see that the signal’s second (2kHz) and third (3kHz) harmonics dominate at -70dBrA, or 0.03%; the remaining signal harmonics are at or below -80dBrA, or 0.01%. Below 1kHz, we see power-supply noise-related peaks at the fundamental (60Hz), and second (120Hz), third (180Hz) and fourth (240Hz) harmonics, at a low -105 dBrA, or 0.0006%, and below. Other lower-level noise-related peaks can also be seen. There is a significant rise in the noise floor just above 20kHz, typical for many class-D amps. It’s clear from the FFT above that THD related peaks dominate with the Ampster BT II, while noise levels are relatively low.

FFT spectrum – 1kHz (digital input, 16/44.1 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the optical digital input, sampled at 16/44.1. The main differences compared to the analog input FFT above are the fourth (4kHz), fifth (5kHz), and sixth (6kHz) signal harmonic peaks that are higher here, at or near -70dBrA, or 0.03%.

FFT spectrum – 1kHz (digital input, 24/96 data at 0dBFS)

Shown above is the fast Fourier transform (FFT) for a 1kHz input sinewave stimulus, measured at the output across an 8-ohm load at 10W for the optical digital input, sampled at 24/96. We see essentially the same signal harmonic profile as with the 16/44.1 FFT above, but with a slightly lower noise floor due to increased bit-depth.

FFT spectrum – 1kHz (digital input, 16/44.1 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 16/44.1 input sinewave stimulus at the optical digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at just slightly below the correct amplitude, and no visible signal harmonic peaks seen above the -145dBrA noise floor. The second (120Hz) and fourth (240Hz) power-supply-related harmonic peaks are slightly below -130dBrA, or 0.00003%.

FFT spectrum – 1kHz (digital input, 24/96 data at -90dBFS)

Shown above is the FFT for a 1kHz -90dBFS dithered 24/96 input sinewave stimulus at the optical digital input, measured at the output across an 8-ohm load. We see the 1kHz primary signal peak, at just below the correct amplitude, and the signal’s third (3kHz, left channel) harmonic can be seen at a very low -140dBrA, or 0.00001%. Power-supply-related harmonic peaks are similar to what is seen in the 16/44.1 FFT above.

FFT spectrum – 50Hz (line-level input)

Shown above is the FFT for a 50Hz input sinewave stimulus measured at the output across an 8-ohm load at 10W for the analog input. The X axis is zoomed in from 40Hz to 1kHz, so that peaks from noise artifacts can be directly compared against peaks from the harmonics of the signal. The most predominant (non-signal) peak is the signal second (100Hz) and third (150Hz) harmonics at -70dBrA, or 0.03%, with other signal harmonics can be seen below -80dBrA, or 0.01%. Very small power-supply-related peaks can be seen, for example, at 60Hz at -105dBrA, or 0.0005%, and 120Hz at -110dBrA, or 0.0003%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, line-level input)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the analog input. The input RMS values are set at -6.02dBrA so that, if summed for a mean frequency of 18.5kHz, would yield 10W (0dBrA) into 8 ohms at the output. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at -80/-90dBrA (left/right), or 0.01/0.003%, while the third-order modulation products, at 17kHz and 20kHz, are higher, at around -60dBrA, or 0.1%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 16/44.1)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital optical input at 16/44.1. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at 70dBrA, or about 0.03%, while the third-order modulation products, at 17kHz and 20kHz, are at -55dBrA, or 0.2%.

Intermodulation distortion FFT (18kHz + 19kHz summed stimulus, coaxial digital input, 24/96)

Shown above is an FFT of the intermodulation distortion (IMD) products for an 18kHz + 19kHz summed sinewave stimulus tone measured at the output across an 8-ohm load at 10W for the digital optical input at 24/96. We find that the second-order modulation product (i.e., the difference signal of 1kHz) is at 70dBrA, or about 0.03%, while the third-order modulation products, at 17kHz and 20kHz, are at -55dBrA, or 0.2%.

Intermodulation distortion FFT (line-level input, APx 32 tone)

Shown above is the FFT of the speaker-level output of the Ampster BT II with the APx 32-tone signal applied to the input. The combined amplitude of the 32 tones is the 0dBrA reference, and corresponds to 10W into 8 ohms. The intermodulation products—i.e., the “grass” between the test tones—are distortion products from the amplifier and are below the -100dBrA, or 0.001%, level, below 6kHz.

Square-wave response (10kHz)

Above is the 10kHz squarewave response using the analog input, at roughly 10W into 8 ohms. Due to limitations inherent to the Audio Precision APx555 B Series analyzer, this graph should not be used to infer or extrapolate the Ampster BT II’s slew-rate performance. Rather, it should be seen as a qualitative representation of its average bandwidth. An ideal squarewave can be represented as the sum of a sinewave and an infinite series of its odd-order harmonics (e.g., 10kHz + 30kHz + 50kHz + 70kHz . . .). A limited bandwidth will show only the sum of the lower-order harmonics, which may result in noticeable undershoot and/or overshoot, and softening of the edges. Here we find considerable overshoot in the corner, which may be due to the Ampster BT II’s non-linear frequency response above 10kHz. In addition, we can see the 400kHz switching oscillator frequency used in the digital amplifier section clearly visible modulating the waveform.

Square-wave response (1kHz–250kHz bandwidth)

Above is the 1kHz squarewave response using the analog input, at roughly 10W into 8 ohms, with a 250kHz restricted bandwidth to remove the modulation from the 400kHz oscillator. Here we seen a relatively clean squarewave response, with the exception of the overshoot in the corner, which again may be due to the Ampster BT II’s non-linear frequency response above 10kHz.

FFT spectrum of 400kHz switching frequency relative to a 1kHz tone

The Ampster BT II’s class-D amplifier relies on a switching oscillator to convert the input signal to a pulse-width modulated (PWM) squarewave (on/off) signal before sending the signal through a low-pass filter to generate an output signal. The Ampster BT II oscillator switches at a rate of about 400kHz, and this graph plots a wide bandwidth FFT spectrum of the amplifier’s output at 10W into 8 ohms as it’s fed a 1kHz sinewave. We can see that the 400kHz peak is quite evident, and at -35dBrA. There is also a peak at 800kHz (the second harmonic of the 400kHz peak), at -60dBrA, and at 1.2MHz (the third harmonic) at -75dBrA. Those peaks are direct results of the switching oscillators in the Ampster BT II amp modules. The noise around those very-high-frequency signals are in the signal, but all that noise is far above the audioband—and is therefore inaudible—and so high in frequency that any loudspeaker the amplifier is driving should filter it all out anyway.

Damping factor vs. frequency (20Hz to 20kHz)

The final plot above is the damping factor as a function of frequency. Both channels track very closely, and range from about 44 from 20Hz to 2kHz, then a dip to 6.5 at 20kHz. This dip in damping factor at high frequencies is typical of inexpensive class-D amp modules.

Diego Estan
Electronics Measurement Specialist