Halcro dm10 preamplifier Measurements

Sidebar 4: Measurements

It's rare to find ground-lift switches in consumer audio equipment, much though they are needed, but the Halcro dm10 is one such component. Arranging the grounding between my Audio Precision System One to get the lowest noise was therefore a relatively simple business, without me having to reach for a "cheater" AC adapter. I also used the dm10's supplied Shunyata AC cord. (Note that I was only able to measure the dm10 in conventional voltage mode. I also didn't test the headphone jack, which I will do in a Follow-up.)

I first tested the dm10's phono stage. The MM input impedance at 1kHz was a high 53k ohms with the rotary control at its rightmost position ("60k"), 9.8k ohms with it set to the leftmost position ("10k"). The MC input impedance was 220 ohms as specified. The three gain settings of the moving-magnet setting provided 32dB, 38dB, and 44dB of gain at 1kHz, exactly as specified, while switching in the moving-coil stage added another 26dB. (All figures measured at the tape output jacks and with the AP source impedance set to 25 ohms.) The phono input inverted absolute polarity, and the RIAA error (see fig.1) was very low, at less than ±0.1dB. Fig.1 also shows that the Halcro implements the IEC-recommended LF rolloff, reaching -0.5dB at 30Hz and -3dB at 15Hz. The phono-stage channel separation at 1kHz was excellent, at almost 80dB.

Fig.1 Halcro dm10 phono stage, MM, RIAA error at 10mV input at 1kHz (0.5dB/vertical div., right channel dashed).

The dm10's phono stage was very quiet; the A-weighted signal/noise ratio with the input set to MM, highest gain, measured 83.2dB (ref. 5mV input), this reducing to a still excellent 67.1dB, unweighted, wideband. Set to highest-gain MC, the S/N ratio was a still superb 72dB, A-weighted (63dB, unweighted). Because of the low noise, I was able to get an accurate reading of the phono stage's distortion—a mere 0.0007% at 1kHz for the 32dB MM setting with a 6mV input signal, with the second harmonic the highest in level at just -103dB (not shown). Correlating with that low distortion, the dm10's phono overload margins were superb. For the MM setting in the lower two gain settings, I measured 27dB at 20Hz and 25.9dB at 1kHz and 20kHz, these figures dropping to 21.1dB and 19.9dB for the highest gain mode. The corresponding MC figures were 21.3dB and 19.9dB for the two lower gain settings, and 15.2dB and 13.9dB, respectively, for the highest gain mode. The latter figures are good rather than great, but it must be remembered that this mode's very high gain of 70dB will be used only with very-low-output cartridges.

As set by PB for his auditioning, the dm10's balanced line-stage gain was 5.53dB with the volume control set to "0," 25.53dB with it set to "20." The figures for the unbalanced input to unbalanced output were 6dB lower, as expected. Both balanced and unbalanced I/O jacks preserved absolute polarity, and while the input impedance was almost exactly twice that specified, at 21k ohms unbalanced and 43k ohms balanced, the difference is in the right direction for optimum system matching. The output impedance was basically to specification, at 171 ohms unbalanced and 330 ohms balanced. The tape-output source impedance was higher than specified, at 673 ohms.

The balanced line-stage frequency response was perfectly flat from 20Hz to 20kHz (fig.2), -0.4dB at 10Hz, and 1dB down at a high 144kHz. The unbalanced response was basically the same as the balanced, and the response didn't change with the volume-control setting. One thing that should be noted in this graph is the slight (0.2dB) channel imbalance, which persisted at all volume-control settings. (The balance was set to dead center for the measurements.) The line-stage channel separation (not shown) was superb, at almost 130dB at 1kHz.

Fig.2 Halcro dm10 line stage, balanced frequency response at 1V into 100k ohms with volume control at "0" (0.5dB/vertical div., right channel dashed).

The dm10 can deliver very high output voltages with virtually zero distortion. Figs.3 and 4 show the THD+noise percentage in the preamplifier's output plotted against voltage for the balanced and unbalanced outputs, respectively. At our usual 1% THD clipping point, the dm10 could output a balanced 16V into a high 100k ohms, and more than 10V even into the demanding 600 ohm load. (The corresponding unbalanced figures were half these voltages, as expected.) Note the downward slope in these graphs: this means that at voltages less than the clipping point, the measured THD+noise is actually dominated by noise, not distortion. Therefore, plotting the THD+noise against frequency at a 10V balanced output and 7V unbalanced (fig.5) really shows only noise, at least below 10kHz. With the input shorted and the volume control at its maximum setting, the unweighted wideband S/N ratio (ref. 1V) was 84.7dB, due to the presence of a small amount of ultrasonic noise. Switching in an A-weighting filter improved the ratio to 106.5dB.

Fig.3 Halcro dm10 line stage, distortion (%) vs balanced 1kHz output voltage into (from bottom to top): 100k ohms, 600 ohms.

Fig.4 Halcro dm10 line stage, distortion (%) vs unbalanced 1kHz output voltage into (from bottom to top): 100k ohms, 600 ohms.

Fig.5 Halcro dm10 line stage, THD+N (%) vs frequency at (from bottom to top): 10V balanced into 100k ohms and 600 ohms, 7V unbalanced into 100k ohms and 600 ohms.

The dm10's actual distortion is at or below the residual level of the lowest-distortion signal generators I have on hand: the Audio Precision System One's analog generator and a PC-based, National Instruments digitally synthesized generator. A plot of the dm10's unbalanced output driving a 1kHz signal from the latter into a lowish 8k ohm load is shown in fig.6. What harmonics are visible are below -110dB (0.0003%), which is extraordinary linearity. The picture was similar when I tested the Halcro's high-frequency intermodulation. The spectral products you can see in fig.7 lie at around the -100dB level (0.001%), which is negligible.

Fig.6 Halcro dm10 line stage, spectrum of 1kHz sinewave, DC-10kHz, at 1V unbalanced into 8k ohms (linear frequency scale).

Fig.7 Halcro dm10, line-stage HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1V unbalanced into 8k ohms (linear frequency scale).

Halcro's dm10 offers measured performance that is both beyond my ability to fully characterize and beyond reproach. Wow.—John Atkinson

Company Info
US distributor: On A Higher Note
26081 Via Estelita, San Juan Capistrano, CA 92675
(949) 488-3004
Article Contents
Share | |
Site Map / Direct Links