Tube Preamp Reviews

I measured the McIntosh C12000 with the preamplifier chassis sitting on top of the controller/power supply, the two joined with the supplied umbilical cables. I used my Audio Precision SYS2722 system for the measurements, confirming some of the results with the higher-performance APx555. The C12000 has five independent outputs: one balanced pair labeled "T" for tube output circuitry; another balanced pair labeled "SS" for solid state output circuitry; an unbalanced pair labeled "Main" that can be switched between tube and solid state output circuitry; a second unbalanced pair labeled "Rec/Fixed" for sending output to a recording device; and a ¼l" headphone jack. I performed a complete set of line-input tests from the two sets of balanced outputs and from the Main output in both modes, then repeated some of the tests using the headphone output.

The Controller's firmware was V 1.02. All the outputs preserved absolute polarity (ie, were noninverting) with both balanced and unbalanced line inputs. As used by SM for his review, the input Gain Trim control was set to "0.0dB." In this condition, the maximum gain was 15.1dB from the balanced line inputs to both balanced outputs, 21dB from the unbalanced inputs to the balanced outputs, and 14.8dB from the unbalanced inputs to the unbalanced Main output in either mode. Setting the volume control to "60%" on the front-panel meters reduced the gain by 20dB.

The headphone output's source impedance was <1 ohm across the audioband, and its maximum gain depended on the setting. Set to "16–40 ohms," the gain was 9.14dB; to "40–150" it was 15.1dB, and to "150–600" it was 21.1dB.

The C12000's unbalanced input impedance was lower than the specified 25k ohms, but still usefully high at 13.8k ohms at 20Hz and 1kHz, and 12.8k ohms at 20kHz. The balanced input impedance was close to the specified 50k ohms, at 48.6k ohms at 20Hz and 1kHz, and 46.8k ohms at 20kHz. The output impedance is specified as 200 ohms, balanced, and 100 ohms, unbalanced. My estimates agreed with the specified impedances at 1kHz and 20kHz, in both SS and T modes, but the output impedance rose to 795 ohms, balanced, and 315 ohms, unbalanced, at 20Hz. As long as the power amplifier has an input impedance of 10k ohms or higher, the C12000's low frequencies won't sound lightweight.

The preamplifier's frequency response into high impedances in SS mode was flat from 10Hz to 50kHz and down by just 1.7dB at 200kHz (fig.1, blue and red traces). This graph was taken with the volume control set to "100%"—the superb channel matching was retained at lower settings of the control. The C12000's response into 600 ohms (green and gray traces) was the same as that into the high impedance over most of the band but rolled off in the bass due to the increased output impedance in this region. The response into 100k ohms in Tube mode (fig.2, blue and red traces) was still flat in the audioband but was now down by 4.4dB at 200kHz.

The McIntosh preamplifier's channel separation (not shown) was superb, at >130dB in both directions below 1kHz, and a still superb 120dB at the top of the audioband. The C12000 offered extremely low noise from all its outputs, with no power-supply–related spuriae even with the Tube output and the volume control set to the maximum (fig.3). The wideband, unweighted signal/noise ratio, measured with the unbalanced input shorted to ground but the volume control set to "100%," was a very good 78.5dB ref. 1V for the balanced SS output, 81dB for the balanced T output, and 84dB for the single-ended output in Tube mode. Restricting the measurement bandwidth to the audioband increased these ratios by 15dB, while switching an A-weighting filter into circuit further improved the ratios by 2–3dB.

Not only is the C12000 a very quiet preamplifier, it also offers very low levels of distortion and very high maximum output levels. Fig.4 plots the percentage of THD+noise in the McIntosh's balanced output in SS mode against the output voltage into 100k ohms with the volume control set to the maximum. Actual distortion lies beneath the noisefloor below an output of 5V then gradually rises above that level. We specify a preamplifier's clipping voltage as being when the THD+N reaches 1%, which occurs at 24V in fig.4; it was the same in Tube mode. The single-ended Main output clipped at 12V into 100k ohms, which is still much higher than the maximum level the preamplifier will be asked to deliver in typical use. The clipping voltage into the current-hungry 600 ohm load was 8.23V from both types of balanced output (fig.5) as well as from the Main output.

To ensure that the THD+N reading was not dominated by noise, I measured how the C12000's distortion changed with frequency at an output level of 5V. The THD+N percentage into 100k ohms was superbly low throughout the audioband (fig.6, blue and red traces) but rose in the low bass into 600 ohms (green, gray traces). This graph was taken with the Tube output; the Solid State output behaved identically.

Fig.7 shows the preamplifier's spectrum with the balanced SS output driving 2V into 100k ohms. The highest-level harmonic is the second, at just –117dB (0.00014%). This harmonic rose by 5dB in Tube mode (fig.8), but this is still extremely low in level. The level of the second harmonic in both modes rose to –100dB (0.001%) into 600 ohms and was joined by the third harmonic at –110dB (0.0003%). Tested for intermodulation distortion with an equal mix of 19 and 20kHz tones at the same peak voltage level, all the intermodulation products lay at a negligible –120dB (0.0001%, fig.9) or lower.

Turning to the preamplifier's phono inputs, I connected a wire from one of the ground terminals on the preamplifier chassis's back panel to a grounding post on the Audio Precision analyzer to minimize hum. The two phono inputs can be configured for moving coil or moving magnet cartridges, and the gain and resistive and capacitive loadings can be adjusted accordingly. The phono inputs preserved absolute polarity at all the preamplifier's outputs. With the gain set to "40," the gain at the Rec/Fixed outputs with a 1kHz signal was 40.64dB. It was 15dB higher at the single-ended Main output and 21dB higher at the balanced outputs. Each 6dB increase in the nominal gain setting increased the gain by 6dB, and at the maximum setting of "64," the gain was 64.5dB at the Rec/Fixed output, 79.6dB at the Main output, and 85.6dB at the balanced output.

The input impedance can also be set to values between 25 ohms and 47k ohms. With the capacitive loading set to 100pF, the measured input impedances were all close to the nominal values. The only exception was the 47k ohm setting, where I measured 40.5k ohms at 20Hz, 41.9k ohms at 1kHz, and 39.5k ohms at 20kHz.

I performed the subsequent tests of the McIntosh's phono inputs at the Rec/Fixed output. The C12000's RIAA equalization was superbly accurate up to 20kHz (fig.10), with a slight rise at ultrasonic frequencies and a low-frequency rolloff of 3dB at 10Hz. Channel matching was excellent. Channel separation (not shown) was close to 70dB in both directions across the audioband.

The phono input's unweighted, wideband S/N ratio, measured with the input shorted to ground, depended on the gain setting. In the worst case, with 64dB of gain, it was still a very good 71.8dB (average of both channels), ref. 1kHz at 500µV. Restricting the measurement bandwidth to 22Hz–22kHz increased the ratio to 76dB in both channels, and inserting an A-weighting filter gave a ratio of 80.7dB. Spectral analysis of the C12000 phono input's low-frequency noisefloor (fig.11) revealed low levels of random noise components and no AC supply–related spuriae.

The C12000's phono input's overload margins varied with the gain setting. At the lowest gain of 40dB, the margin, ref. 1kHz at the standard MM level of 5mV, was an excellent 26.4dB at 20Hz and 1kHz, and still 23dB at the top of the audioband. Each 6dB increase in gain reduced the overload margins by around 3dB, but even with 64dB of gain, the margins ref. the standard MC level of 500µV were a high 23dB from 20Hz to 20kHz.

The McIntosh's phono inputs offered very low distortion. Fig.12 shows the spectrum of the preamplifier's output reproducing an input signal of 1kHz at 20mV with the gain set to 46dB. The only distortion harmonics that can be seen above the noisefloor are the second, at –100dB (0.001%), and the third, at –104dB (0.0006%). The phono inputs also offered a low level of intermodulation distortion. Even with the level of an equal mix of 19kHz and 20kHz tones lying 5dB below the overload voltage, the second-order difference product lay at –64dB (0.05%) and the higher-order products were all 30dB lower in level (fig.13).

With both its line and phono inputs, the McIntosh C12000's measured performance is among the best I have encountered from a preamplifier.—John Atkinson