Hovland HP-100 preamplifier Measurements part 2

This graph was taken into the kind 100k ohms load. Reducing the load to 10k ohms, a typical input impedance for many solid-state power amplifiers (fig.5), increased the second harmonic almost tenfold, to -60dB (0.1%), and added some third harmonic. Dropping the load to the admittedly punishing 600 ohms raised the second harmonic to -44dB (0.6%, not shown). It is probably inadvisable to use the Hovland with those few power amplifiers that have input impedances below 10k ohms.

Fig.5 Hovland HP-100, line-stage spectrum of 50Hz sinewave, DC-1kHz, at 2V into 10k ohms (linear frequency scale).

This is reinforced by fig.6, which shows how the HP-100's THD+N percentage varies with output level. Into 100k ohms (bottom trace), the distortion at the 2V output level—which will drive most power amplifiers to their clipping points—is at or below 0.05%, and the preamp's output stage doesn't clip (defined as 1% THD+N) until 9.5V RMS, well above any level the preamp will be asked to deliver. But into 600 ohms (top trace), the distortion is almost 10 times higher at 2V, and only 4.2V is available at clipping.

Fig.6 Hovland HP-100, line-stage THD+N (%) vs output voltage at 1kHz into 100k ohms (top above 300mV) and 600 ohms (bottom).

The phono stage featured an input impedance of 500 ohms in the midrange and treble, dropping very slightly to 470 ohms at 20Hz. The voltage gain was a very high 72.8dB, with correspondingly low overload margins: approximately 6.8dB at 1kHz, 6dB at 20kHz, and 10.4dB at 20Hz (all figures referred to a nominal MC cartridge 1kHz output of 500µV, or 0.5mV). The HP-100's phono input has obviously been optimized for cartridges of very low output. With such a model's 150µV output at 1kHz, these margins will improve by 10.5dB. Even so, the resultant 17.3dB overload margin at 20kHz is still on the low side, in my opinion, the subjective result being perhaps some slight emphasis of surface-noise ticks.

With the enormous phono-stage gain available, it was not surprising that the unweighted noise was disappointing at -46dB. A-weighting the measurement improved the S/N ratio to a good 62.5dB, but I suggest that only those with the very lowest-output MC cartridges have the internal 20dB step-up transformers fitted. MF noted some hum, and I found it hard to prevent the Hovland's step-up transformers from picking up radiated 60Hz from other components' power transformers.

Finally, fig.7 shows the phono stage's RIAA error. (The input impedance of 470-500 ohms is sufficiently higher than the Audio Precision System One's 25 ohm output impedance that this graph should be free from any interaction effects.) The ±0.2dB swayback in the Hovland's phono response will be subtly audible as extra treble resolution and a very slightly rich bass, or as a very slightly recessed midrange. The response is swiftly rolled-off above 20kHz—a sensible design decision, in my opinion.

Fig.7 Hovland HP-100, phono-stage RIAA error at 1mV input (1kHz), with volume control at maximum gain into 100k ohms (right channel dashed, 0.5dB/vertical div.).

Other than the unnecessarily high phono-stage gain which leads to the higher-than-usual noise floor and lower-than-usual overload margins, and the disappointing line-stage channel separation, the Hovland appears to be a well-crafted preamplifier. It certainly sounded fabulous in Mikey's system!—John Atkinson