Lamm Industries LL1 Signature line preamplifier Measurements

Sidebar 3: Measurements

I used my top-of-the-line Audio Precision SYS2722 system (see www.ap.com and the January 2008 "As We See It") to measure the Lamm LL1 Signature preamplifier. With its four chassis, the LL1 occupies a lot of real estate; I placed the two signal-handling chassis on the rack next to the test system, with the umbilical cables long enough to allow me to place the two power-supply chassis on the floor.

Although the LL1 offers XLR jacks for input and output, Lamm refers to these as "pseudo-balanced," meaning that they are actually single-ended. I measured the preamp's behavior using conventional single-ended cables fitted with RCA plugs. Once everything was connected, I experimented with grounding between the Audio Precision and the four Lamm preamplifier chassis to get the lowest level of measured noise.

The two channels matched very closely when set to maximum gain, both offering 17.34dB. The attenuation switch reduced the gain by precisely the specified 12dB. As its comprehensive operating manual warns, the LL1 inverts signal polarity. The input impedance is specified as "typically 16.5k ohms"; I measured 16.4k ohms at 20Hz, 16.3k ohms at 1kHz, and 14.7k ohms at 20kHz. At middle and high frequencies the output impedance was low, at around 280 ohms, but rose to 1530 ohms at 20Hz.

This increased output impedance resulted in rolled-off low frequencies into the demandingly low 600 ohm load (fig.1, cyan and magenta traces), but into the more typical 100k ohms (blue, red traces) the response was flat from 10Hz to 30kHz. This graph was taken with the independent volume controls set to their maximum positions—note the superb channel matching. Setting the volume controls to unity gain (approximately 1:30) slightly reduced the ultrasonic bandwidth, from –3dB at 200kHz to –3dB at 130kHz, but the superb channel matching was preserved.

814Lammfig1.jpg

Fig.1 Lamm LL1 Signature, frequency response with volume control set to maximum at 1V, into: 100k ohms (left channel blue, right red), 600 ohms (left cyan, magenta right) (1dB/vertical div.).

Crosstalk was, of course, unmeasurable. The unweighted, wideband signal/noise ratio, measured with the inputs shorted but the volume control set at its maximum, was a little disappointing, at 63.5dB left and 61.7dB right (both figures ref. 1V), but restricting the measurement bandwidth to the audioband increased these ratios to a superb 106.5 and 105dB, respectively, with A-weighting giving another 3dB of improvement. Spectral analysis of the low-frequency noise floor (fig.2) indicated that the primary noise components in the audioband were 60Hz and its odd-order harmonics, though it's fair to note that these all lie at or below –100dB ref. 1V (0.001%) so will not be audible.

814Lammfig2.jpg

Fig.2 Lamm LL1 Signature, spectrum of 1kHz sinewave, DC–1kHz, at 1V into 100k ohms (left channel blue, right red; linear frequency scale).

Lamm specifies the LL1's output as 1V; plotting the percentage of THD+noise against output voltage into 100k ohms (fig.3) reveals that the actual distortion starts to rise above the noise floor just below 1V. This indicates an excellent gain architecture. Above 1V, the increase in THD is linear in level but doesn't reach 1% (our usual definition of clipping) until 48V! This is well above the 4V or so required to drive any matching power amplifier into overload, meaning that the LL1 will never be the limiting factor in a system's behavior. What surprised me was that the LL1 gave the same maximum output, and proved even more linear, into the very low 600 ohm load (fig.4), other than at very low frequencies, here the linearity was affected by the rolled-off output into that load (fig.5, cyan and magenta traces).

814Lammfig3.jpg

Fig.3 Lamm LL1 Signature, distortion (%) vs 1kHz output voltage into 100k ohms.

814Lammfig4.jpg

Fig.4 Lamm LL1 Signature, distortion (%) vs 1kHz output voltage into 600 ohms.

814Lammfig5.jpg

Fig.5 Lamm LL1 Signature, THD+N (%) vs frequency at 1V into: 100k ohms (left channel blue, right red), 600 ohms (left cyan, right magenta).

Spectral analysis of the worst-case scenario—50Hz at 1V into 600 ohms—indicated that the distortion was almost entirely the subjectively benign second harmonic, at –61dB (fig.6), the two channels behaving identically. This harmonic dropped by 12dB when the load was increased to 100k ohms (fig.7), but remained predominant. High-order intermodulation distortion was vanishingly low in level (fig.8); though the second-order difference component at 1kHz that resulted from an equal mix of 19 and 20kHz tones was higher in level, it still lay at a very low –80dB (0.01%).

814Lammfig6.jpg

Fig.6 Lamm LL1 Signature, spectrum of 50Hz sinewave, DC–1kHz, at 1V into 600 ohms (left channel blue, right red; linear frequency scale).

814Lammfig7.jpg

Fig.7 Lamm LL1 Signature, spectrum of 50Hz sinewave, DC–1kHz, at 1V into 100k ohms (left channel blue, right red; linear frequency scale).

814Lammfig8.jpg

Fig.8 Lamm LL1 Signature, HF intermodulation spectrum, DC–30kHz, 19+20kHz at 1V peak into 100k ohms (left channel blue, right red; linear frequency scale).

Summing up the LL1's measured performance is easy: This is a superbly well-designed preamplifier that, other than its second-order harmonic signature, shows no trace of being based on tubes. And even then, that signature is not accompanied by any significant intermodulation distortion. Once again, I must tip my hat to Vladimir Lamm's engineering chops!—John Atkinson

COMPANY INFO
Lamm Industries Inc.
2513 E. 21st Street
Brooklyn, NY 11235
(718) 368-0181
ARTICLE CONTENTS

COMMENTS
otaku's picture

I guess I could listen just to mono LP's and save $21,395.

markotto's picture

I certainly hope it is "the best possible sound from a line-stage preamplifier".

Oldsport's picture

Hi John, just a thought on a possible noise source (and this may be opening a new audiophile can of worms). Do you have any compact fluorescent lights (CFLs) in your testing area? Those buggers emit RF; they may also dirty up the A/C as well, I don't know. Just hold a cheap AM radio next to one and have a listen. I won't have any in my listening room--they definitely degrade the sound. In fact, I try to have all CFLs turned off on the entire floor that my listening room is on, when I'm using it. Regards!

John Atkinson's picture
Oldsport wrote:
Do you have any compact fluorescent lights (CFLs) in your testing area? Those buggers emit RF; they may also dirty up the A/C as well

Yes I do! I don't have the Lamm available anymore but I will try replacing the CFL with a normal light bulb and see if I can change the measured noise level with a different preamp.

A subsequent discussion with Vladimmir Lamm revealed that he measures wideband signal/noise ratio with a 110kHz bandwidth whereas I use 500kHz. That would explain the differences in our measured results, especially as I have the CFL in my test lab.

John Atkinson
Editor, Stereophile

X