Proceed PAV audio/video preamplifier Measurements
All of the measurements here, except as noted, were made from the unbalanced outputs with an unbalanced input. The PAV's gain, at a maximum setting of the level control, measured 12dB. Its output impedance measured just over 20 ohms (40 ohms balanced), its input impedance just under 95k ohms (22k ohms balanced). The output impedance at the tape outputs was just over 10 ohms, regardless of the source impedance—indicating buffered tape outputs. The DC offset at the PAV's outputs tended to fluctuate—as it does with many amps and preamps—but it generally measured 2.5mV or less on all outputs.
With the volume control set at maximum, the PAV's gain measured 12dB in stereo and Pro Logic, 10dB in THX—a difference I found peculiar. It turns out the gain in all modes is the same up to the final few steps of the volume control. The THX re-equalization circuits have an insertion loss of 2dB. In order to keep the levels equal when switching between modes, the PAV's designers increased the net gain in THX by 2dB to make up for this loss.
The steps of the PAV's main volume control measured 0.5dB within a small fraction of a dB, as specified, over most of its range. The same was true of the individual adjustments for the center, subwoofer, and surrounds. Volume-control tracking was outstanding for the left and right channels, with slightly more deviation—up to 0.08dB!—in the center channel.
The PAV was non-inverting at all of its outputs, and pin 2 of its balanced outputs was positive. Its signal/noise ratio, measured relative to a 1V output, is shown (to the nearest decibel) in Table 1. The noise increases in both the Pro Logic and THX modes—this is typical of surround-sound processors. The balanced results, shown in parentheses, are generally slightly higher than the unbalanced—interesting.
Table 1: Proceed PAV S/N Ratio (unweighted ref. 1V)
|Left||89 (88)||79 (73)||83 (77)|
|Right||96 (88)||79 (73)||83 (77)|
Fig.1 shows the frequency response of the front left channel (without high-pass filtering) and the surround channel. (To make the results more legible, the center and right channels—virtually identical to the left—aren't shown.) The curves are displaced for clarity, and the scale is expanded to show the top-end responses. The stereo mode is extremely flat. The deep bass of both the Pro Logic and THX modes have small degrees of rolloff which are unlikely to be audible, particularly with a subwoofer supplementing the bottom-end response. The top-end rolloff in the THX mode is from the required THX re-equalization. The surround channels are flat at the bottom end (because Pro Logic programming is not supposed to have any bass in the surrounds below about 80Hz, many processors incorporate a rolloff here), but exhibit the top-end rolloff typical of Pro Logic. The added top-end contouring in the THX surround response, including the ripples above 3kHz, is THX's timbre-matching.
Fig.1 Proceed PAV, frequency response, from top to bottom, of the front left channel (without high-pass filtering) in Stereo, Pro Logic, and THX modes, and the surround channel in Stereo, Pro Logic, and THX modes (2dB/vertical div.).
Fig.2 shows the subwoofer output of the PAV overlaid with the left-channel, high-pass-filtered output. This, the setup typically used in a home THX setup, shows the sharp rolloff of the THX high- and low-pass filters. With the output levels matched at 20Hz and 1kHz, as in the curve, the crossover is actually a bit lower than the THX-specified 80Hz; in practice, this won't be particularly significant.
Fig.2 Proceed PAV, subwoofer high- and low-pass responses (5dB/vertical div.).
The crosstalk (fig.3) is, as expected, lowest in the stereo mode; the increase in crosstalk with Pro Logic is simply in the nature of the multi-channel matrix. The reduced crosstalk at high frequencies in the surround channel is largely a function of the latter's high-frequency rolloff. The THX crosstalk was almost identical to that shown for Pro Logic, though, again, with some top-octave variations due to the THX HF compensation.
Fig.3 Proceed PAV, crosstalk in Pro Logic and Stereo modes (from top to bottom): L-R, Pro Logic; L-C, Pro Logic; L-Surr, Pro Logic; R-L, Stereo; L-R, Stereo (20dB/vertical div.).
The PAV's THD+noise, taken with an input of 500mV, is shown in fig.4. Note that the surround THD, taken over a limited bandwidth of 100Hz to 7kHz, is generally as low as, or lower than, the THD of the front channels. The THD+noise of the subwoofer output isn't shown, but it's comparable to that of the front channels. The THX readings, not shown, virtually duplicate those of Pro Logic. As expected, the THD+noise in the stereo mode is by far the lowest, but the THD+noise for Pro Logic/THX is still quite respectable.
Fig.4 Proceed PAV, line-stage THD+noise vs frequency at 500mV input (from top to bottom): Pro Logic Left; Pro Logic Right; Pro Logic Center; Stereo Center; Stereo Left & Right (right channel dashed); Pro Logic Surround (jagged trace).
I also measured the THD+noise vs output voltage on the PAV in both the Stereo and Pro Logic modes, at the full setting of the level control. My primary reason for doing this was to determine if a 500mV input was suitable for the other measurements (I actually made this last reading first). With a gain of 12dB, 500mV at the input translates to an output of about 2V. As you can see in fig.5, a 2V output is well within the PAV's capabilities—both in stereo and Pro Logic. The PAV's input-level adjustment capabilities will, in any event, enable it to be configured for use with virtually any CD player, D/A converter, or laserdisc player without risk of overloading.
Fig.5 Proceed PAV, distortion (%) vs output voltage in Pro Logic (top) and Stereo (bottom) modes.
Though all of my listening to the PAV was done in its unbalanced mode, I also made a number of measurements using its balanced inputs and outputs. Its balanced gain, frequency response, and THD+noise, measured in the stereo mode, were all comparable to the unbalanced results.
All in all, the PAV performed extremely well on the test bench.—Thomas J. Norton