Peachtree Audio Carina GaN Integrated Amplifier Measurements

Link: reviewed by George de Sa on SoundStage! Hi-FI on January 1, 2025

General Information

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

The Peachtree Audio Carina GaN was conditioned for 1 hour at 1/8th full rated power (~12W 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 Carina GaN offers one set of line-level and phono analog inputs (RCA, switchable), left/right pre-outs, two digital coaxial (RCA) S/PDIF inputs, one optical (TosLink) S/PDIF input, one USB digital input, one set of speaker-level output, and two headphone outputs, one over 1/4″ TRS connector and the other 4.4mm balanced TRRRS. A Bluetooth input is also offered. For the purposes of these measurements, the following inputs were evaluated: digital coaxial, analog line-level and phono, and the balanced headphone output.

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 achievable output power of 100W into 8 ohms. (NOTE: despite the 200Wpc into 8/4-ohm rating, the Carina GaN could not sustain more than approximately 100Wpc into 8/4 ohms without inducing the fault protection circuit. According to Peachtree, this is abnormal behaviour and may indicate a fault with our unit under test. The issue is being investigated by Peachtree). For comparison, on the line-level input, an SNR measurement was also made with the volume at maximum.

The Carina GaN offers four different digital filters that are applied only when using the pre-out or headphone outputs. These are: linear phase fast (default), hybrid fast, minimum phase slow, and NOS.

Based on the accuracy and repeatability of the left/right volume-control channel matching (see table below), the Carina GaN volume control is operating fully in the digital domain, meaning all incoming signals are digitized. The Carina GaN overall volume range is from -62dB to +27.9dB (line-level input, speaker output). It offers 1dB increments throughout the volume range.

Our typical input bandwidth filter setting of 10Hz-22.4kHz was used for all measurements except FFTs and where a bandwidth of 10Hz-90kHz was used. Frequency-response measurements utilize a DC to 1 MHz input bandwidth. Because the Carina GaN is a digital amplifier technology that exhibits considerable noise above 20kHz (see FFTs below), the 22.4kHz bandwidth setting was maintained for THD versus frequency sweeps as well. For these sweeps, the highest frequency was 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 Peachtree for the Carina GaN compared directly against our own. The published specifications are sourced from Peachtree’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.

* 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 Carina GaN was able to sustain 170W into 4 ohms (~1.5% THD) using an 80Hz tone for 500ms, alternating with a signal at -10dB of the peak (17W) for 5 seconds, for 206 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 Carina GaN was 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 coaxial digital input at the balanced headphone output (unless specified, assume a 1kHz sinewave, 24/96 0dBFS input/2Vrms 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 Carina GaN is essentially perfectly flat within the audioband (20Hz to 20kHz). Just beyond 20kHz, however, there is brickwall filtering, followed by a peak at 30kHz, then beyond 30kHz, brickwall filtering again. This frequency response behavior, along with the behavior of the volume control, indicates that incoming analog signals are digitized within the Carina GaN. The Carina GaN is roughly -0.4dB at 5Hz.

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 Carina GaN yields an enormous amount of absolute input-to-output phase shift, which includes the time delay inherent with the digitization and processing of analog signals. Here we see roughly 17,000 degrees of phase shift at 20kHz. Below . . .

. . . is the same plot but shown “wrapped,” where each 360 degrees of phase shift just wraps back inside the plot area.

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

The chart above shows the Carina GaN’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. The blue and red traces are for a 16-bit/44.1kHz dithered digital input signal from 5Hz to 22kHz using the coaxial input, 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, the digital signals yielded a flat or slight over-response at 5Hz, whereas the analog input is -0.4dB at 5Hz. All signals show the same brickwall-type behavior just past 20kHz.

Frequency response vs. MQA (16/44.1)

The chart above shows the Carina GaN’s frequency response (relative to 1kHz) as a function of filter type measured at the pre-outs (left channel only), from 1kHz to 30kHz for a dithered 16/44.1 input signal. Please note that these filters only affect the pre-out and headphone outputs. They have no effect on the speaker-level outputs. The blue trace is the linear phase fast filter, the purple trace hybrid fast, pink is minimum phase slow, and orange is NOS. The linear phase fast filter exhibits brickwall-type filtering just past 20kHz, the hybrid fast filter is -8dB at 20kHz, the minimum phase slow filter is -4dB at 20kHz, and the NOS filter is -30dB at 20kHz, -10dB at 10kHz, and -3dB at 8kHz.

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

The chart above shows the frequency response for the MM phono input. 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 maximum deviations within ±0.2dB or so from 20Hz to 20kHz, and worst-case channel-to-channel deviations of roughly 0.1dB. Here again we find sharp attenuating just past 20kHz, meaning the phono input is also digitized. While a very strict adherence to the RIAA curve such as this may be an indication that EQ is applied in the digital domain, Peachtree have assured us that equalization is applied in the analog domain.

Phase response (MM input)

Above is the phase response plot from 20Hz to 20kHz for the MM phono input, measured across the speaker outputs at 10W into 8 ohms. The same extreme phase shift seen here was also seen for the line-level analog input.

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 Carina GaN, where 0dBFS was set to yield 2Vrms. 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 -110dBFS 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, that below -120dBFS, the 16/44.1 data significantly over-responds (over 10dB), while the 24/96  significantly under-responds (below -10dB).

Impulse response (24/44.1 data, linear phase fast filter)

The graph above shows the impulse response for the Carina GaN, fed to the coaxial digital input, measured at the line-level pre-outputs, using the Transfer Function measurement in the APx500 software, for the linear phase fast filter. We find a reconstruction filter with symmetrical pre/post ringing. A typical sinc function response.

Impulse response (24/44.1 data, hybrid fast filter)

The graph above shows the impulse response for the Carina GaN, fed to the coaxial digital input, measured at the line-level pre-outputs, using the Transfer Function measurement in the APx500 software, for the hybrid fast filter. We find a reconstruction filter that minimizes pre-ringing, but with obvious post-ringing.

Impulse response (24/44.1 data, minimum phase slow filter)

The graph above shows the impulse response for the Carina GaN, fed to the coaxial digital input, measured at the line-level pre-outputs, using the Transfer Function measurement in the APx500 software, for the minimum phase slow filter. We find a reconstruction filter that minimizes pre-ringing but exhibits some post-ringing.

Impulse response (24/44.1 data, NOS

The graph above shows the impulse response for the Carina GaN, fed to the coaxial digital input, measured at the line-level pre-outputs, using the Transfer Function measurement in the APx500 software, for the NOS filter. We find a reconstruction filter that minimizes all ringing, emulating a true non-oversampling (NOS) DAC.

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 pre-outputs of the Carina GaN where 0dBFS is set to 2Vrms. 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.

Here we see a strong J-test result, with only very low-level peaks below the -150dBrA level. This is an indication that the Carina GaN DAC should have strong jitter immunity.

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 Carina GaN. The optical input yielded similar but slightly worse results compared to the coaxial input, with peaks flanking the 12kHz fundamental at the -130dBrA level.

J-Test with 10ns of injected jitter (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 Carina GaN, with an additional 10ns of 2kHz sinewave jitter injected by the APx555. The results are strong, with only a few extra spurious peaks at the -150dBrA level.

J-Test with 100ns of injected jitter (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 Carina GaN, with an additional 100ns of 2kHz sinewave jitter injected by the APx555. The results are strong again, this time with the tell-tale peaks at 10kHz and 14kHz, but highly suppressed at the -140dBrA level. This is further indication that the DAC in the Carina GaN has strong 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 line-level pre-outputs of the Carina GaN, with an additional 100ns of 2kHz sinewave jitter injected by the APx555. The results are strong again, but slightly worse than with the coaxial input. Here the 10/14kHz peaks are higher, at the -125dBrA level.

Wideband FFT spectrum of white noise and 19.1kHz sine-wave tone (coaxial input, linear phase fast filter)

The chart above shows a fast Fourier transform (FFT) of the Carina GaN’s line-level pre-outputs 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, using the linear phase fast filter. The steep roll-off around 20kHz in the white-noise spectrum shows the brickwall-type characteristic of this filter. There is only one clear low-level aliased image peaks within the audioband at -125dBrA at roughly 18kHz. The primary aliasing signal at 25kHz is highly suppressed at -120dBrA, while the second and third distortion harmonics (38.2, 57.3kHz) of the 19.1kHz tone are at roughly the same level.

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 see that the deviations between no load and 4 ohms are around 1dB. This is a poor result and an indication of a high output impedance, or low damping factor. With a real speaker load, deviations measured lower at roughly the 0.4dB level.

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 60W. The power was varied using the Carina GaN volume control. The 1W THD ratios were the lowest, with a constant 0.005% from 50Hz to 6kHz (0.01% at 20Hz). The 10W THD ratios were higher, from 0.2% at 20Hz, down to roughly 0.01% from 1kHz to 6kHz. At 60W, THD ratios ranged from 2% at 20Hz, down to 0.07% from 1kHz to 6kHz.

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

The chart above shows THD ratios 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 for the MM configuration vary from around 0.2% (20Hz) down to 0.01% from 1kHz to 6kHz.

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 Carina GaN 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). Note: the protection circuit would engage and shut down the unit at roughly 100W into 4 ohms, and just past 100W into 8 ohms. From 50mW to 3W, both THD plots track relatively closely, from 0.02% down to 0.005%. Above 3W, THD ratios into 4 ohms are higher, from 0.02% at 5W to a plateau of 0.2% from 30W to 100W. Into 8 ohms, THD ratios ranged from 0.01% at 5W up to 0.1% from 50W to 100W.

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

The chart above shows THD+N ratios measured at the speaker-level outputs of the Carina GaN 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). Note: the protection circuit would engage and shutdown the unit at roughly 100W into 4 ohms, and just past 100W into 8 ohms. From 50mW to 5W, both THD plots track relatively closely, from 0.2% down to 0.02%. Above 5W, THD+N ratios into 4 ohms are higher, from 0.02% at 5W to a plateau of 0.2% from 30W to 100W. Into 8 ohms, THD+N ratios ranged from 0.02% at 5W up to 0.1% from 50W to 100W.

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 Carina GaN as a function of frequency into three different loads (8/4/2 ohms) for a constant input voltage that yields 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 a roughly 20dB increase (0.01% to 0.2%) in THD from 8 to 4 ohms from 400Hz to 6kHz. From 4 to 2 ohms, there is only a roughly 5dB increase in THD, with the 2-ohm data yielding a relatively steady 0.5 to 0.3% across the measured audio band.

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 Carina GaN 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 higher than those measured across the resistive dummy load. The differences ranged from 0.15% at 20Hz for the two-way speaker versus 0.006% for the resistive load, and 0.01% at 100-150Hz into the three-way speaker versus 0.005% for the resistive load. At higher frequencies (1.5kHz to 6kHz), the THD ratios measured across all three loads are similar, at around the 0.005% level.

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 Carina GaN 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 that the IMD ratios into the three-way speaker were the highest, ranging from 0.03% to 0.05%, compared to the resistive load which was a constant 0.008%. The two-way speaker ranged from 0.02%, down to 0.005%, then up to 0.03%.

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 Carina GaN 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 0.02 and 0.03% from 40Hz to 250Hz, then down to 0.005% upt 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 third (3kHz) harmonic dominates at a significant -80dBrA, or 0.01%, while the second (2kHz), fourth (4kHz) and fifth (5kHz) harmonics are around -95dBrA, or 0.002%. These are high THD results for a modern solid-state amplifier. On the right side of the signal peak, we do not see any power-supply-related noise peaks (e.g., 60/120/180Hz); however, the overall noise floor is relatively high at -120dBrA.

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 essentially the same result as with the analog FFT above. Given that analog signals are very likely digitized inside the Carina GaN, this should come as no surprise.

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 result as with the 16/44.1 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, just below the correct amplitude, with no other peaks above the noise floor. We can also see that the left channel (blue) is quieter than the right channel by more than 5dB throughout the audioband.

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, just below the correct amplitude, with no other significant peaks above the noise floor. Once again, the left channel is quieter than the right channel.

FFT spectrum – 1kHz (MM phono 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 MM phono input. We see that the signal’s third (3kHz) harmonic dominates at a significant -80dBrA, or 0.01%, while the second (2kHz), fourth (4kHz) and fifth (5kHz) harmonics are around -95dBrA, or 0.002%. On the right side of the signal peak, we find power-supply related noise peaks at 60/120/180/240Hz etc, ranging from -90dBrA, or 0.003%, down to -110dBrA, or 0.0003%.

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 -60dBrA, or 0.1%. Other peaks can be seen at -90dBRA, or 0.003%. Again, this represents high levels of THD for a modern solid-state amplifier.

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 the third (150Hz) signal harmonic at a high -60dBrA, or 0.1%. The highest power-supply related noise harmonic can be seen at 60Hz at -90dBrA, or 0.003%.

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 -110dBRa, or 0.0003%, while the third-order modulation products, at 17kHz and 20kHz, are much higher at -85dBrA, or 0.006%.

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

Shown above is the FFT of the speaker-level output of the Carina GaN 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 -100dBrA, or 0.001%, level.

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

Shown above is an FFT of the intermodulation (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 -110dBRa, or 0.0003%, while the third-order modulation products, at 17kHz and 20kHz, are much higher at -85dBrA, or 0.006%.

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 -110dBRa, or 0.0003%, while the third-order modulation products, at 17kHz and 20kHz, are much higher at -85dBrA, or 0.006%.

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

Shown above is an FFT of the intermodulation (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 around -100dBrA, or 0.001%, while the third-order modulation products, at 17kHz and 20kHz, are higher at -85dBrA, or 0.006%.

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 Carina GaN’s slew-rate performance. Rather, it should be seen as a qualitative representation of the Carina GaN’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, we see the 400Hz switching frequency from the digital amplifier riding on top of a distorted 10kHz fundamental sinewave.

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

Above is the 1kHz squarewave response using the analog line-level input, at roughly 10W into 8 ohms, this time with a 90kHz input bandwidth on the analyzer to filter out the 400kHz switching frequency, as well as other high-frequency artifacts. We see more evidence here, in the overshoot/undershoot and soft corners of the squarewave, of the Carina GaN’s very limited bandwidth.

FFT spectrum (1MHz bandwidth)

The Carina GaN’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 Carina GaN 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 -20dBrA. There are also two peaks at 800kHz and 1.2MHz (the second and third harmonic of the 400kHz peak), at -50dBrA. In addition, there is a very significant rise in the noise floor from 25kHz to 300kHz, peaking at -70dBrA. Those peaks—the fundamental and its harmonics—are direct results of the switching oscillators in the Carina GaN 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)

The final graph above is the damping factor as a function of frequency. We can see here a constant damping factor right around 18, from 20Hz to roughly 2kHz, then a dip down to around 7 at 20kHz. This is a very poor damping factor result for a modern solid-state amplifier.

Diego Estan
Electronics Measurement Specialist