Hovland HP-100 preamplifier Measurements
The drop-dead gorgeous Hovland HP-100 preamp offered a maximum line-stage gain of 13.6dB—the silky-feeling switched attenuator operating in accurate 2dB steps down to -58dB, with then a step to a full mute. (There's also a separate Mute button.) The volume control didn't have an exact unity-gain setting; the nearest was the 3 o'clock position, which featured a 0.135dB insertion loss. The 12 o'clock position was equivalent to -14dB referred to the "unity gain" setting.
The Hovland's line stage didn't invert absolute polarity. Its input impedance was a usefully high 100k ohms in the midrange and bass, with just a small reduction to a still-high 86k ohms at 20kHz. Despite the claim that the HP-100 uses a cathode follower output stage, its source impedance was high at 2.4k ohms, rising to 4.3k ohms at 20Hz. The partnering power amplifier would best have an input impedance of at least 47k ohms if the bass is not to sound lean.
Line-stage frequency response (fig.1) was flat within the audioband, rolling off above 10kHz to reach a probably negligible -0.25dB at 20kHz and -3dB at 80kHz. The response did not change at different volume-control settings. The HP-100's channel separation (fig.2) was disappointing: While good at low frequencies (78dB L-R, 73dB R-L), it degraded at 6dB/octave throughout the midrange and treble to a merely adequate 45dB (L-R) and 37dB (R-L) at 20kHz. This is presumably due to capacitive coupling between channels, perhaps due to a twin-triode tube shared between channels at or ahead of the volume control. I suspect the latter as the channel separation did change with the volume-control setting.
Fig.1 Hovland HP-100, line-stage frequency response at 1V output with volume control at maximum gain into 100k ohms (right channel dashed, 0.5dB/vertical div.).
Fig.2 Hovland HP-100, line-stage channel separation with volume control at maximum, R-L dashed, L-R solid (10dB/vertical div.).
Mikey said he felt the Hovland was not the quietest preamp around. Line-stage noise was only okay at -71.5dB, unweighted ref. 1V. The S/N ratio improved to 95.1dB when A-weighted, suggesting that the noise is mainly high- or low-frequency in nature. Fig.3 shows the line stage's THD+noise percentage, plotted against frequency at 750mV output into 100k ohms. The true distortion is buried within the noise below 10kHz; there is a rise in harmonic content above that frequency, but as it reaches just 0.1% above 30kHz, it is probably subjectively inconsequential, particularly as the primary distortion component is the second, at twice the frequency (fig.4).
Fig.3 Hovland HP-100, line-stage THD+N (%) vs frequency at 750mV into 100k ohms (right channel dashed).
Fig.4 Hovland HP-100, line-stage spectrum of 50Hz sinewave, DC-1kHz, at 2V into 100k ohms (linear frequency scale).