Tube Preamp Reviews

Levels of intermodulation distortion were also extremely low. Fig.5 shows the spectrum of the unbalanced output driving an equal mix of 19kHz and 20kHz tones into 100k ohms. The only apparent modulation product is the difference tone at 1kHz, this lying at -80dB (0.01%). This was virtually unchanged into 10k ohms (not shown), which is about the lowest typical power-amplifier input impedance the L2 is likely to see—though it did rise to -54dB (0.2%) into the punishing 1k ohm load (fig.6). But from the balanced outputs at 10V, the intermodulation spectrum was one of the cleanest I have ever seen (fig.7).

Fig.5 Lamm L2, unbalanced HF intermodulation spectrum, DC-24kHz, 19+20kHz at 5V into 100k ohms (linear frequency scale).

Fig.6 Lamm L2, unbalanced HF intermodulation spectrum, DC-24kHz, 19+20kHz at 5V into 1k ohm (linear frequency scale).

Fig.7 Lamm L2, balanced HF intermodulation spectrum, DC-24kHz, 19+20kHz at 10V into 100k ohms (linear frequency scale).

Finally, the use of MOSFET gain devices run at very high voltages results in the Lamm preamplifier offering enormous dynamic range. Figs.8 and 9 show the level of distortion and noise plotted against output voltage from the unbalanced and balanced outputs, respectively. The minimum value in each trace shows the point where the measured THD+N percentage changes from being dominated by noise to showing the actual level of distortion. Into 100k ohm loads, this occurs at around 1.5V unbalanced and 15V balanced. The former is the approximate point where an amplifier with an unbalanced input will approach its clipping point; the latter is way above a balanced power amplifier's clipping point, meaning that what distortion is present in the L2's output may well be masked by noise at any sensible operating volume.

Fig.8 Lamm L2, unbalanced distortion (%) vs output level (V) into (from bottom to top at 1V): 1k ohm, 100k ohms.

Fig.9 Lamm L2, balanced distortion (%) vs output level (V) into (from bottom to top at 2V): 600 ohms, 100k ohms.

But note that above these output voltages—all the way to the point where the traces stop because the Audio Precision signal generator has reached its maximum output level—the L2 has not reached its actual hard clipping point. Instead, the transfer function gradually becomes more curved as the output level reaches into the tens of volts.

Power amplifiers with very low input impedances should be avoided because of the rising source impedance at low frequencies, and the restricted output voltage into low loads indicated by figs.8 and 9. These measurements reveal the Lamm L2 to be a well-engineered preamplifier with a sensibly arranged gain/distortion structure. I did wonder, however, whether the L2's responses to Jonathan's mechanical tweaks were connected to its very resonant chassis.—John Atkinson