Mark Levinson No.334 power amplifier Measurements

Sidebar 4: Measurements

A complete set of measurements of the Mark Levinson No.334 was made in the balanced configuration. Selected readings were also performed in unbalanced mode. Unless otherwise noted, the results here refer to balanced operation.

Following its one-hour preconditioning test, the Mark Levinson's heatsinks were hot, but could still be touched without discomfort. The amplifier's voltage gain into 8 ohms measured 27dB, balanced or unbalanced. The input impedance measured 98k ohms balanced, 50.7k ohms unbalanced. The DC offset was a very low 1.5mV in both channels. The No.334 is noninverting at its unbalanced input; pin 2 of the balanced input is positive. The unweighted S/N ratio at 1W into 8 ohms measured 79.2dB over a 22Hz-22kHz bandwidth, 58.1dB balanced from 10Hz to 500kHz, and 81.6dB A-weighted. The broadband, unweighted result is unusually high for a high-performance amplifier. All possible grounding schemes were tried, with little change. The unbalanced S/N measured only a small fraction of a dB higher. Most of the noise is above 80kHz; it is unlikely that it will be audible.

The No.334's crosstalk was so low as to be irrelevant. The measured output impedance of the No.334 varied from 0.014 to 0.022 ohms—inconsequential values. (The highest reading was at 20kHz.) This amplifier's frequency response should not vary with the loudspeaker load. Fig.1 shows the No.334's frequency response. It is so flat to our normal measurement limit of 50kHz that I continued the curve to 200kHz. As you can see, it remained essentially flat to 100kHz with all of our test loads. The unbalanced result (not shown) was virtually identical. Fig.2 shows the amplifier's 10kHz squarewave response. The slight overshoot visible on the leading edge is quickly damped, and risetime is excellent. The 1kHz squarewave (not shown) is similar, with the overshoot also visible (but only just).

Fig.1 Mark Levinson No.334, frequency response at (from top to bottom): 1W into 8 ohms, and 2.828V into simulated loudspeaker load (0.5dB/vertical div., right channel dashed).

Fig.2 Mark Levinson No.334, small-signal 10kHz squarewave into 8 ohms.