Arcam FMJ A22 integrated amplifier Measurements
The Arcam FMJ A22 is apparently very sensitive to thermal loading. Though not running particularly hot, it shut down 20 minutes into its 1/3-power, 1-hour preconditioning test. After a brief cooling-down period, it came back on without incident.
The A22 proved particularly prone to test-bench shutdown during all types of heavy loading. After it would switch off, it had to be turned off entirely before it could reset and allow itself to be turned on again. I probably did this 50 times during the amp's time on the test bench. One set of measurements, phono overload, could not be performed at all because the amp shut down well below 1% THD+noise. Aside from this tendency to shut down easily (evidently not a problem in normal use, per LB), I noted no quirks while taking these measurements.
The FMJ A22's line input impedance measured 14.3k ohms. Its maximum line voltage gain measured 38.6dB in high-gain mode (which I used for all measurements), 36.1dB in low-gain. For most of the remaining measurements I set the FMJ A22's level control to a point that produced an overall line gain of 35dB, this level chosen because it represents the maximum gain of many preamp/power-amp combinations.
The Arcam's amplifier output impedance was a bit higher than is normally found in solid-state amplifiers, but not alarmingly so. It measured a maximum of 0.16 ohms at 1kHz and 20Hz, this increasing to 0.28 ohms at 20kHz, measured using an 8 ohm load. The line-level output impedance measured 47 ohms at both the tape and preamp outputs, and the tape output impedance did not change when the source impedance was altered, indicating buffered tape outputs. DC offset at the main outputs measured 3.4mV in the left channel, 3.5mV in the right. Line signal-to-noise (unweighted at 1W into 8 ohms) measured 80.9dB from 22Hz to 22kHz, 71.5dB from 10Hz to 500kHz, and 83.3dB A-weighted. The volume-control tracking was very good, the left and right channels differing by a maximum of 0.17dB. The Arcam's line stage was noninverting, a positive-going input resulting in a positive-going output.
Fig.1 shows the FMJ A22's frequency response. The RIAA response of the phono stage is displaced here by -2dB for clarity. The 10kHz squarewave response (fig.2) has a fast risetime and only the slightest rounding of the leading edge. (Such rounding is common to most amplifiers and corresponds to the ultimate ultrasonic rolloff.) The 1kHz squarewave, not shown, is virtually perfect.
Fig.1 Arcam FMJ A22, frequency response at (from top to bottom at 6kHz): 1W into 8 ohms, 2W into 4 ohms, and 2.828V into simulated loudspeaker load; phono MC (0.5dB/vertical div.).
Fig.2 Arcam FMJ A22, small-signal 10kHz squarewave into 8 ohms.
Fig.3 shows the line and phono crosstalk. Interestingly, the phono crosstalk (taken from the tape outputs, as were all the phono measurements) is significantly better than the line-level result, but neither is high enough to cause any audible degradation. The THD+noise vs frequency results (fig.4) are also very good, with the rise at the highest frequencies typical of many amplifiers. While some designs have less of an increase than this, the most common means of reducing it is an increase in feedback—which many designers prefer to avoid. The plot of the distortion waveform (fig.5) indicates a dominant third harmonic, plus some higher-order harmonics and noise. The actual level of this distortion, however, is very low. The harmonic composition was similar into 8 and 2 ohm loads (not shown).
Fig.3 Arcam FMJ A22, channel separation: L-R line (top at 1kHz); R-L (line); MM R-L; MM L-R; MC R-L; MC L-R (10dB/vertical div.).
Fig.4 Arcam FMJ A22, THD+noise (%) vs frequency at (from top to bottom at 4kHz): 4W into 2 ohms; 2.83V into simulated loudspeaker load; 2W into 4 ohms; and 1W into 8 ohms (right channel dashed).
Fig.5 Arcam FMJ A22, 1kHz waveform at 2W into 4 ohms (top), distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.6 shows the Arcam's output spectrum in response to a 50Hz input at an output of 114W into a 4 ohm load. The only significant artifact is at 150Hz (the third harmonic): -74.9dB, or less than 0.02%. Fig.7 shows the intermodulation in the output caused by a combined 19+20kHz input at 95W into 4 ohms. The only audible artifact here is at -76.9dB (about 0.015%) at 18kHz. The intermodulation distortion into an 8 ohm load is slightly lower (79.8dB, or 0.01% at 18kHz). The FMJ A22 clips with this 19+20kHz signal above 95W into 4 ohms and 60W into 8 ohms.
Fig.6 Arcam FMJ A22, spectrum of 50Hz sinewave, DC-1kHz, at 114W into 4 ohms, (linear frequency scale).
Fig.7 Arcam FMJ A22, HF intermodulation spectrum, DC-22kHz, 19+20kHz at 95W into 4 ohms (linear frequency scale).
Fig.8 shows the THD+noise vs level curves for the FMJ A22, one channel driven, at 1kHz. The amplifier's discrete clipping measurements are shown in Table 1. All are above spec. The continuous output into 2 ohms was not measured due to the amplifier's low shutdown threshold.
Fig.8 Arcam FMJ A22, distortion (%) vs continuous output power into (from bottom to top at 2kHz) 4 ohms, 8 ohms, and 2 ohms.
Table 1 Arcam FMJ A22: Clipping
|(1% THD+noise at 1kHz)|
|Both Channels Driven||One ChannelDriven|
|Load||W (dBW)||W (dBW)|
|8||105 (20.1)||107.6 (20.3)||121 (20.8)|
|4||153.4 (18.9)||159 (19)||185.4 (19.7)|
John Atkinson measured the Arcam's power output using a low-duty-cycle toneburst signal, which will be more representative of how the amplifier behaves with a music signal. The results are shown in fig.9: the modest-looking FMJ is a powerhouse with this signal, clipping at 149 into 8 ohms and 261 W into 4 ohms (clipping defined as 1% THD+N). Into the punishing 2 ohm and 1 ohm loads, the Arcam put out 401W and 360W, respectively, but with a rise in overall distortion and an earlier onset of clipping. In fact, like most amplifiers, halving the load impedance raises the THD by 6dB or so. But even into 1 ohm, the distortion remains below -70dB (0.03%) until 350W output or so, well below the usual 0.1% point of possible audibility (horizontal magenta line). And if you relax the clipping point to 3% THD, the available output power rises to 442W, equivalent to an RMS current of 21A.
Fig.9 Arcam FMJ A22, distortion (%) vs 1kHz burst output power into 8 ohms (black trace), 4 ohms (red), 2 ohms (blue), and 1 ohm (green).
Turning to the Arcam's phono stage: For moving-magnet cartridges, the input impedance measured 47.2k ohms left, 46.2k ohms right; for moving-coils, 312.2 ohms left, 313 ohms right. Voltage gain was 35.9dB (MM) and 55dB (MC). The S/N ratio (ref. 1V) measured 90dB from 22Hz to 22kHz, 73.6dB 10Hz-500kHz, and 94.2dB A-weighted for MM cartridges. For MCs, S/N measured 80.6dB, 72.8dB, and 85.5dB, respectively, for the same conditions. The phono stage is noninverting.
The phono stage's frequency response and crosstalk were presented in figs.1 and 2. Fig.10 shows THD+noise plotted against frequency for the phono input. The moving-magnet curve was taken with an input of 76mV, the moving-coil at 6.6mV, these high levels chosen to minimize the effect of noise on the results. The performance is fine other than the rapid rise in THD for ultrasonic signals at these high signal levels. Finally, the THD+noise vs output of the phono stage is shown in fig.11. The overload margins are respectably high, but the phono stage appears to clip hard.
Fig.10 Arcam FMJ A22, phono stage THD+noise (%) vs frequency, MC (top) and MM (bottom) (right channel dashed).
Fig.11 Arcam FMJ A22, phono stage distortion (%) vs output voltage, MC (top), MM (bottom).
The Arcam FMJ A22 performed well on the test bench, with no problems aside from its sensitive protection circuitry.—Thomas J. Norton