Cello Palette Preamplifier Measurements
Fig.1 shows the Palette Preamp's frequency response with the equalization bypassed, both at the 9:00 setting of the level control (top curves above 20kHz, left channel solid, right channel dotted) and at unity gain (bottom curves), the latter at a level-control setting of 2:30. In fig.2, the distortion resulting from a 500mV input is shown, at three settings of the level control (only the left channel is shown for the full-volume position).
Fig.1 Cello Palette Preamplifier, frequency response in bypass mode with the level control set to 9:00 (top traces above 20kHz) and at unity gain (right channel dashed, 0.5dB/vertical div.).
Fig.2 Cello Palette Preamplifier, THD+noise (%) vs frequency for 500mV input at three settings of the volume control (full volume is the bottom curve). (Right channel dashed.) Measurement dominated by noise.
The above results were obtained with the equalizer function switched out of the circuit. Switching the equalizer in with the equalizer knobs set to a mechanically flat position resulted in the frequency response indicated in fig.3. While the deviations are still small, listeners who hear a difference when switching the EQ in and out with the controls "flat" should be aware that there is a small frequency-response deviation induced by the EQ controls in their mechanically "flat" positions. This deviation may vary slightly depending on the mechanical installation of the controls.
Fig.3 Cello Palette Preamplifier, frequency response in EQ mode with all controls set to zero (right channel dashed, 0.5dB/vertical div.).
In fig.4, note the increase in crosstalk at full gain—still quite good—from the EQ out setting (lower curve) to the EQ in (top). The interchannel crosstalk is slightly (but not significantly) assymetrical, with the commonly seen increase at higher frequencies due (most likely) to capacitive coupling between channels.
Fig.4 Cello Palette Preamplifier, channel separation with volume control full and EQ bypassed (bottom) and engaged (top) (5dB/vertical div.).
Fig.5 shows the increase in distortion resulting from switching the equalizer in with the level control fully on; the bottom curve repeats that shown in fig.2 for full volume, EQ out.
Fig.5 Cello Palette Preamplifier, THD+noise (%) vs frequency for 500mV input with EQ engaged (bottom curve is with EQ out at full volume). (Right channel dashed.) Measurement dominated by noise.
Finally, the effect of the six EQ controls in maximum boost and cut is shown in fig.6. Note the significant overlap between adjacent controls and the reduced effect of the controls toward the center of the audio band—the latter an intelligent choice as the ear is very sensitive to small changes through the midrange.
Fig.6 Cello Palette Preamplifier, frequency response in EQ mode with all controls set to maximum (right channel dashed, 5dB/vertical div.).
The gain of the line-stage-only Palette preamp measured 5.9dB at the maximum setting of the level control. Output impedance measured approximately 163 ohms, give or take less than 2 ohms, at any measured level either with the equalizer in or out. The extremely high input impedance—one of Cello's design goals—was so high that it was difficult to measure with confidence. In all but one case, I measured between two and three megohms. In the latter—the left channel at full volume—the reading was over 1 megohm. Note that at these high impedances, a 0.05 millivolt difference in indicated voltage on our Audio Precision test set—at the 100mV input level employed—could double or halve the calculated impedance.—Thomas J. Norton