Solid State Preamp Reviews

I measured the Lamm L2.1 Reference preamplifier with my Audio Precision SYS2722 system (see the January 2008 As We See It"). All measurements were taken at the preamplifier's Direct inputs, which bypass the input switching. The gain with the volume control set to its maximum was 8.8dB, unbalanced input to unbalanced output; and, as expected, 6dB higher from unbalanced input to balanced output. The twin volume controls operated in accurate 1dB steps, and the unity-gain setting at the unbalanced outputs was with the controls set to 3:30. Both sets of outputs preserved absolute polarity (ie, were non-inverting). The unbalanced inputs, specified as offering a usefully high input impedance of 50k ohms, met that figure at 20Hz and 1kHz. Though the input impedance dropped to 44k ohms at 20kHz, this is inconsequential.

The unbalanced output impedance, specified as typically being 130 ohms, was 131 ohms at 1kHz and 139 ohms at 20kHz, but rose to 3.3k ohms at 20Hz, presumably due to the presence of an output coupling capacitor. The balanced impedances were twice the unbalanced, again as expected. With this very high output impedance at low frequencies, it came as no surprise that, when I tested the unbalanced frequency response into the very low 600 ohms load, the bass rolled off, reaching –3dB at 180Hz (fig.1, cyan and magenta traces). But into a much higher impedance (blue and red traces) the response was flat from 50Hz to 50kHz, with the output down by just 0.9dB at 10Hz and by 0.7dB at 200kHz. As long as the L2.1 is used with power amplifiers having an input impedance of 20k ohms or higher, the low frequencies will not sound lean. This graph was taken with the volume controls set to their maximum. There was no significant difference at lower settings, and the balanced response into 100k ohms was identical to the unbalanced. Also note in this graph the excellent channel matching, which was preserved at lower volume settings.

The Lamm's channel separation was superb, at >120dB below 1kHz and still 110dB at 20kHz. Noise levels were low: the wideband, unweighted signal/noise ratio, taken with the inputs shorted to ground but the volume controls set to the maximum—the worst-case situation—measured 76.8dB ref. 1V. Restricting the bandwidth to the audioband improved the ratio to an excellent 87.8dB, with an A-weighting filter giving further improvement to 90.5dB. Spectral analysis revealed that the low-frequency noise floor was smooth, other than some very low-level spuriae at the AC supply frequency of 60Hz and its odd-order harmonics (fig.2). These stem from magnetic interference, perhaps from a supply transformer—though for these tests I did have the separate power-supply chassis on the floor, well away from the preamplifier chassis.

As I found with the original L2 Reference preamplifier, my Audio Precision test set can't deliver a voltage high enough to drive the L2.1 into clipping. Fig.3 shows that at the signal generator's maximum output the L2.1 delivers 36V into 100k ohms, but the distortion is still only just above 0.1%. At lower output voltages the percentage of THD+noise is very low, reaching just 0.005% at 1.4V. Below that level, the trace in fig.3 is dominated by noise. Reducing the load impedance to 10k ohms—which is about the lowest impedance the preamplifier will encounter in a real-world system—the picture was very similar (fig.4). Only into the punishing load of 600 ohms did the Lamm actually clip (defined as 1% THD+N), but even then, not until 3.3V (not shown).

Fig.5 shows how the THD+N percentage changes with frequency, taken at a level, 2V, where I could be sure that I was looking at THD rather than N. Commendably, the measured figure remains low across the audioband, and the distortion signature is almost pure second harmonic, this lying at –90dB (0.003%, fig.6). This graph was taken into 100k ohms; the second harmonic rose slightly into 10k ohms, to –87dB (0.0045%). Intermodulation distortion was also extraordinarily low (fig.7). This graph was taken at 1V into 100k ohms; increasing the level to 2V did not increase the level of the distortion to any significant extent—this is extraordinary linearity for a circuit that is said not to use loop negative feedback.

Like the other preamplifiers designed by Vladimir Lamm that have passed through my test lab, the L2.1 Reference is well engineered.—John Atkinson