Audible Illusions Modulus 3A preamplifier Measurements
Line-level measurements of the Audible Illusions were taken from the CD inputs to the main outputs; the phono measurements were taken at the tape outputs.
The Modulus 3's output impedance at its line output measured 1775 ohms (L) and 1802 ohms (R) with the level controls at maximum; lower settings of the main level control changed this measurement by only a few ohms. The input impedance at the CD input was just over 48k ohms and wasn't significantly affected by the main level control's setting. The output impedance at the tape output was 248 ohms regardless of the source impedance feeding the preamp—confirming the presence of buffered tape outputs.
The DC offset at the Modulus 3's outputs measured a maximum of 29mV in the left channel and 37.6mV in the right, though the values fluctuated considerably. The preamp is inverting from line input to output, but its phono stage (phono in to tape out) is non-inverting; the polarity of all inputs will therefore be inverted at the main output. Maximum line-stage gain measured 18.4dB, phono gain 53.9dB. The line-stage S/N ratio (ref. 1V, unweighted) measured 83.3dB from 22Hz-22kHz, 71dB from 10Hz-500kHz, and 87dB, A-weighted. Phono-stage S/N measured 81dB, 72dB, and 87.6dB, under the same conditions, respectively. The latter are very good figures for a moving-coil phono stage.
The Modulus 3's line- and phono-stage frequency responses into a 100k load are shown in fig.1. Note that measurements were taken with the volume controls physically matched; despite MF's experience, I found tracking to be up to 0.5dB off under those conditions. [While the RIAA error (top curve) meets AI's specifications, I would expect the response rise in the bass to be audible as a slight warming of the sound. The rise above 10kHz might also be just audible and is due to the RIAA topology used, where the phono-stage gain doesn't continue to drop with increasing frequency but levels out at unity gain.—Ed.]
Fig.1 Audible Illusions Modulus 3A, MC phono stage RIAA error (top) and line-stage frequency response with volume control set (from top to bottom) to 9:00, 12:00, and Maximum (right channel dashed, 0.5dB/vertical div.).
The 3's crosstalk is shown in fig.2. While the separation is not as good as the best we've measured for line-stages, nevertheless these are audibly inconsequential amounts of crosstalk. The high-frequency increases are typically the result of capacitive coupling between channels.
Fig.2 Audible Illusions Modulus 3A, crosstalk (from top to bottom at 10kHz): R-L, line; R-L, phono; L-R, line; L-R, phono (10dB/vertical div.).
The manner in which THD+noise varies with frequency for the Modulus 3 is shown in fig.3. These are good results. The spikes visible in the phono result are intermittent artifacts apparently due to fluctuating noise levels; they appeared at different points when the measurements were redone several times.
Fig.3 Audible Illusions Modulus 3A, THD+noise (%) vs frequency of phono-stage at 4mV input (top) and of line-stage at 100mV (right channel dashed).