Halcro dm88 Reference monoblock power amplifier Measurements
I preconditioned one of the Halcro dm88 monoblocks (serial no. 00136) by running it for 60 minutes at one-third the specified power into 8 ohms. With the amplifier cold, the measured THD+noise was just 0.00093%! The figure was the same at the end of the hour; while the left-hand side panel was gently warm, the right-hand panel was still cool. It wasn't possible to measure the dm88's Current Mode input; I performed complete sets of measurements via the dm88's balanced and unbalanced Voltage Mode inputs, as well as via its unbalanced Minimal Path jack.
The dm88's voltage gain was a higher-than-average 29.4dB into 8 ohms via the unbalanced Voltage Mode input; unusually, it dropped by 6dB, to 23.4dB, using the balanced jack. The Minimal Path gain was also low, at 23.1dB. Both Voltage Mode inputs preserved absolute polarity— ie, were noninverting (the XLR is wired with pin 2 "hot")—but the Minimal Path input inverted signal polarity. The unbalanced Voltage Mode input impedance was a little higher than usual for a solid-state design, ranging from 25k ohms at 20Hz to 22k ohms at 1kHz, though it dropped further, to 17.5k ohms, at 20kHz. The balanced input impedance was exactly twice the unbalanced, as expected. However, the Minimal Path input had a very low impedance of 650 ohms across the audioband. As well as explaining the slightly lower voltage gain via this input, this low input impedance will be incompatible with many preamplifiers, especially those using tubed output stages.
The dm88's output impedance was very low across almost all the audioband, at 0.1 ohm including 6' of multistrand speaker cable. It rose slightly at 20kHz, to 0.14 ohm, though the increase is insignificant. As a result, any alteration of the amplifier's frequency response driving our standard simulated loudspeaker remained within tight ±0.1dB limits (fig.1, top trace at 2kHz). The responses shown in this graph were taken via the balanced Voltage Mode jack; it is 3dB down at 122kHz into 8 ohms, with a slightly earlier ultrasonic rolloff into low impedances. The low frequencies are down 0.5dB at 20Hz.
Fig.1 Halcro dm88, balanced frequency response at 2.83V into (from top to bottom at 2kHz): simulated loudspeaker load, 8, 4, 2 ohms (0.5dB/vertical div.).
The LF response was identical via the unbalanced Voltage Mode input, though the ultrasonic response extended a little higher, to –3dB at 172kHz (not shown). This extended ultrasonic response was the same via the Minimal Path jack, but the bass is now flat to below 10Hz (fig.2). The extended HF response gives rise to an excellent 1kHz squarewave response (fig.3), and the 10kHz waveform is free from ringing or overshoot (fig.4).
Fig.2 Halcro dm88, Minimal Path frequency response at 2.83V into 8 ohms (0.5dB/vertical div.).
Fig.3 Halcro dm88, small-signal 1kHz squarewave into 8 ohms.
Fig.4 Halcro dm88, small-signal 10kHz squarewave into 8 ohms.
The dm88's noise floor was very low. Referenced to a level of 2.83V into 8 ohms (equivalent to 1W), the A-weighted signal/noise ratio with the input short-circuited was an extraordinary 105.6dB. Even the wideband, unweighted measurement was 84.7dB! This is an amplifier that can take full advantage of the high resolution promised by SACD and DVD-Audio.
As expected from its pedigree, the dm88 can put out high powers with very low distortion, at least into higher impedances. Fig.5 plots the percentage of THD+noise in the amplifier's output against output power into 8, 4, and 2 ohms. The downward slope of the traces below 10W in this graph implies that the actual distortion is below the noise floor, even with the Halcro's very low level of noise in absolute terms. Above that power level, the amplifier's behavior depends on the load impedance. The distortion remains very low into 8 ohms, the small discontinuities in the trace indicating that the measurement is at the limit of the Audio Precision System One's resolving power. Into 4 ohms, however, and even more so into 2 ohms, the distortion begins to rise at high powers, though still not to any level that might be considered subjectively significant. (It is fair to point out that Halcro feels this behavior is not typical of the dm88. I am looking into this discrepancy.) The amplifier does clip very suddenly.
Fig.5 Halcro dm88, distortion (%)vs 1kHz continuous output power into (from bottom to top at 100W): 8, 4, 2 ohms.
The measured output power at 1% THD+N was greater than specified, at 295W into 8 ohms (24.7dBW) and 525W into 4 ohms (24.2dBW). The dm88 is less happy driving 2 ohms, however, clipping at 266W (18.2dBW).
The low levels of THD at small-signal levels made it more informative to plot the dm88's THD against frequency at a higher-than-usual output voltage, where the actual distortion would be above the noise. Fig.6 shows how the dm88's THD changes into 8, 4, and 2 ohms. The traces are flat with frequency, suggesting that the amplifier has a commendably wide open-loop bandwidth. However, the distortion does increase as the load drops. Peculiarly, the THD was a little higher, at 0.0017% across the band, from the Minimal Path input (not shown). Figs.7 and 8 reveal that the primary harmonic content is the third, and that while this rises in level into low impedances, it is not joined by other harmonics. Given both the subjectively benign nature of the third harmonic and the very low level of the distortion in absolute terms, even into 2 ohms, I think it safe to say that the dm88's audible distortion signature will be nonexistent.
Fig.6 Halcro dm88, THD+N (%)vs frequency at 20V into (from bottom to top): 8, 4, 2 ohms.
Fig.7 Halcro dm88, 1kHz waveform at 90W into 8 ohms (top), 0.00093% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.8 Halcro dm88, 1kHz waveform at 196W into 4 ohms (top), 0.0038% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.9 plots the spectrum of the dm88's output while it drives 50Hz at one-third power into 8 ohms. The third harmonic lies at just –106dB (0.0005%), though the second harmonic makes an appearance at –103dB (0.0007%) with this low-frequency tone. (The higher harmonics in this graph are mathematical artifacts of the FFT process and should be ignored.) Halving the load impedance didn't affect the second-harmonic level, but the third rose to –86dB (0.005%), as expected from the waveform graphs. Intermodulation distortion was very low, even at very high powers into 4 ohms (fig.10). The amplifier didn't misbehave when I drove it into clipping with this very demanding signal. (I save this test for last, as this is when optimistically specified amplifiers tend to break.)
Fig.9 Halcro dm88, spectrum of 50Hz sinewave, DC–1kHz, at 100W into 8 ohms (linear frequency scale).
Fig.10 Halcro dm88, HF intermodulation spectrum, DC–24kHz, 19+20kHz at 453W peak into 4 ohms (linear frequency scale).
As I have come to expect from Bruce Candy designs, the dm88 offers superb measured performance.—John Atkinson