Rogue Audio Magnum Ninety-Nine preamplifier Measurements
The five positions of the Rogue Magnum Ninety-Nine's front-panel gain switch correspond to maximum voltage gains of 18.7dB, 13.8dB, 8.6dB, 3.2dB, and -4.4dB. This will be very useful in optimizing the preamp's amplification and noise floor to best match the overall gain structure of its user's system. The Magnum Ninety-Nine inverted signal polarity, and its input impedance measured a usefully high 36k ohms at 1kHz, this dropping slightly at the edges of the audioband.
The Ninety-Nine's output impedance was a fairly low 466 ohms over most of the frequency spectrum, though it did double to 963 ohms at 20Hz, presumably due to a finite-sized coupling capacitor. This results in a degree of bass rolloff into very low impedances, which can be seen in fig.1. Into 600 ohms, for example, which admittedly is lower than the Rogue will see in practice, the -3dB point was 21Hz. However, at the other end of the spectrum, the lower load impedance extended the ultrasonic response slightly. This graph was taken with the gain switch set to its central, 8.6dB, position, and the Magnum Ninety-Nine's HF output in this condition was 1dB down at 20kHz, 3dB down at 43kHz. At higher gain settings, the preamp's bandwidth increased slightly, to -1dB at 30kHz. There was no change in the bandwidth with different volume-control settings, as is sometimes seen in tube circuits.
Fig.1 Rogue Magnum Ninety-Nine, central gain position, frequency response into 100k ohms (top below 1kHz) and 600 ohms (bottom below 1kHz), with volume control at maximum (0.5dB/vertical div.).
Despite its dual-mono construction, the Rogue's channel separation was somewhat disappointing, at less than 50dB in both directions at 20kHz. The shape of the traces in fig.2 suggests that the crosstalk is due to capacitive coupling, probably at the shared volume control. The wideband, unweighted signal/noise ratio in the maximum-gain mode was also not as good as is usual in high-end preamps, at 58.8dB (ref. 1V), this due to some ultrasonic noise. Restricting the measurement bandwidth to the audioband improved the S/N figure to 84.4dB, while switching in an A-weighting figure gave further improvement, to 87.7dB.
Fig.2 Rogue Magnum Ninety-Nine, channel separation (R-L dashed, 10dB/vertical div.).
Fig.3 plots the percentage of distortion and noise in the preamp's output against its output voltage into 100k, 10k, and 1k ohms. Below 1V into the higher loads, the figure is dominated by the noise mentioned above. Actual distortion influences the reading above that level, but the Magnum is capable of swinging many more low-distortion volts than it will ever be required to in practice. Even into 1k ohm (top trace), the preamp will deliver 8V at the usual 1% THD+N clipping point.
Fig.3 Rogue Magnum Ninety-Nine, distortion (%) vs output voltage into (from right to left at 3%): 100k, 10k, 1k ohms.
The distortion is uniform with frequency, with only a negligible increase visible above the audioband into higher impedances (fig.4). It does increase into 1k ohm (top trace in this graph), but is still below the level that would be regarded as audible, particularly as the spectrum of that distortion (fig.5) reveals it to be almost entirely the subjectively innocuous second harmonic. Note, however, the presence of spuriae at multiples of the 120Hz full-wave-rectified power-supply frequency in this graph. Yes, they're all close to or below -100dB (0.001%), but I couldn't eliminate this noise no matter how I arranged the grounding of the test setup. This suggests the preamp has a somewhat suboptimal internal grounding arrangement.
Fig.4 Rogue Magnum Ninety-Nine, THD+noise (%) vs frequency at 1.7V into (from top to bottom): 100k, 10k, 1k ohms.
Fig.5 Rogue Magnum Ninety-Nine, spectrum of 1kHz sinewave, DC-10kHz, at 1V into 8k ohms (linear frequency scale, 10dB/vertical div.).
This power-supply noise can also be seen in the spectrum of the Magnum Ninety-Nine's output while it reproduced an equal mix of 19kHz and 20kHz tones (fig.6). However, the actual intermodulation distortion was quite low in level, with the 1kHz difference component lying at 0.0055% (-83.5dB).
Fig.6 Rogue Magnum Ninety-Nine, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1V into 8k ohms (linear frequency scale).
While not the best-measuring tube preamp I've encountered, Rogue Audio's Magnum Ninety-Nine seems to be generally well-engineered.—John Atkinson