Conrad-Johnson Premier 17LS line-stage preamplifier Measurements
As is usually the case with a tube preamp that doesn't use negative feedback, the Conrad-Johnson Premier 17LS provides way too much amplification for practical use. With the volume control set to its maximum position of "99," the voltage gain measured 25.9dB. The volume control, however, is sensibly arranged to operate in large steps at the top of its range (with some numbers duplicating the level; for example, "99" and "98" are both equivalent to the maximum gain), but in smaller, fraction-of-a-dB steps in the range where the control is most likely to be used. There was no exact unity gain setting; it would have been between "66" and "67," while the -20dB setting was "37."
The 17LS inverted absolute polarity, and its input impedance varied according to the indicated level setting. At "99" it was a high 125k ohms, but dropped with decreasing level to 35k at "90," 29k at "80," and 28k at "70," and leveled off at a still reasonably high 26k from "50" down. The output impedance was below a lowish 800 ohms in the midband and above, but rose to 1.6k ohms at 20Hz, due to the presence of a finite-sized output coupling capacitor. The C-J should therefore be used with a power amplifier having an input impedance of at least 22k ohms if the bass is not to sound a little lean.
Like its input impedance, the Premier 17LS's bandwidth varied with the indicated volume-control setting. Fig.1 shows the preamp's frequency response at settings ranging from "99" to "70." The "99" response is 3dB down at 12kHz, which will be very audibly dullened. However, the preamp's gain is so high at this setting that it is virtually a certainty that the preamp will never be used in this condition. At a more reasonable "80," the 17LS's output is down just 0.5dB at 20kHz, and the preamp is basically flat within the audioband at lower volume settings. However, note the consistent 0.2dB channel imbalance in this graph, due, I assume, to the difficulty of exactly matching the tubes. Channel separation (fig.2) was disappointing, at just 48dB (L-R) and 45dB (R-L) at 10kHz, due to capacitive coupling between the channels.
Fig.1 Conrad-Johnson Premier 17LS, frequency response for 100mV input at volume-control settings of (from top to bottom): "99," "95," "90," "85," "80," "75," and "70" (1dB/vertical div., right channel dashed).
Fig.2 Conrad-Johnson Premier 17LS, crosstalk at 2V output, volume control at "99" (10dB/vertical div., R-L dashed).
The distortion doesn't change with frequency, and reaches a minimum value of 0.06% into 100k ohms. Fig.3 plots the percentage of distortion and noise in the Premier 17LS's output signal against the output voltage into three loads: 1k, 10k, and 100k ohms. Into the higher impedances, the measured figure is dominated by noise below about 1V, this revealed by the linear increase in the percentage with decreasing level. Above that point, the measured figure is dominated by true distortion, which increases with level—again, in a linear manner. The C-J preamp is not a very low-distortion device, but with amplifiers having typical input sensitivities it will introduce possibly audible levels of distortion only in the top 3dB of its range, revealing a sensible balance of its gain/noise/linearity architecture.
Fig.3 Conrad-Johnson Premier 17LS, distortion (%) vs output level (V) into (from bottom to top at 1V): 100k, 10k, and 1k ohms.