Tube Preamp Reviews

When I unboxed the Canor Hyperion P1 preamplifier, the four 6922 tubes, two for each channel, were installed, but the 6H30PI dual-triode tubes, labeled "1" and "2," were packed separately. I carefully installed these tubes in the sockets that corresponded to their numbers and checked that the 6922 tubes were correctly seated. I performed a complete set of tests with my Audio Precision SYS2722 system then repeated some of the testing with the magazine's APx500 analyzer.

I immediately encountered a problem in that there was no output from the right channel, from all the balanced and unbalanced inputs and outputs. I checked that all the cables I was using were okay. Then, after powering down the Hyperion P1, I rechecked that all the tubes were correctly installed. The problem persisted, so after powering down the preamplifier again, I swapped the tubes in one channel with those in the other. The right channel remained dead. The 6H30PI tubes are a very tight fit in their sockets; it is possible that either when the right channel's tube was removed before the preamplifier was shipped to me, during shipping, or when I installed the tube, the printed circuit board on which the socket was mounted had flexed and a trace had gone open circuit, though I could see no indication that this was so. However it happened, this malfunction obviously occurred after Jason's audition. I restored the tubes to their original locations and continued the testing with just the left channel.

All the inputs and outputs preserved absolute polarity, ie, were noninverting. The Hyperion P1's input impedance is specified as 30k ohms; I measured 53k ohms from 20Hz to 20kHz for the balanced inputs, 24k ohms at 20Hz and 1kHz, 21.6k ohms at 20kHz, for the single-ended inputs. The preamplifier's source impedance, specified as <150 ohms, was 274 ohms at 1kHz and 20kHz but 6.6k ohms at 20Hz for the balanced outputs; the unbalanced impedances were half those values. The rise in the source impedance at low frequencies is presumably due to the outputs being capacitor-coupled.

The volume control operated in accurate 1dB steps. At 10.9dB, the maximum gain at the balanced outputs was very close to the specified 11dB with both balanced and unbalanced input signals. The gain was the expected 6dB lower at the unbalanced output. Fig.1 plots the Hyperion P1's frequency response into 100k ohms (blue trace) and 600 ohms (red trace). The response into both impedances is flat to 200kHz, but the increase in the preamplifier's output impedance at low frequencies results in a large rolloff in the bass into 600 ohms. The Hyperion P1 must be used with power amplifiers that have an input impedance of at least 20k ohms at low frequencies if the bass is to not sound lightweight. (The Dan D'Agostino and Accuphase monoblocks that JVS used for his auditioning respectively have a balanced input impedance of 183k ohms and 39.6k ohms at 20Hz, both of which are ample.) The volume control was set to the maximum to generate the traces in fig.1; with it set to –20dB the output was down by 2dB at 200kHz.

With the right channel inoperative, it wasn't possible to measure channel separation in detail. However, feeding the right channel a 1kHz signal, the crosstalk in the left channel lay at –108dB ref. 5V. With the volume control set to the maximum and the balanced input shorted to ground, the wideband, unweighted S/N ratio (ref. 2V output) was an excellent 93.5dB, increasing to 107.2dB when the measurement bandwidth was reduced to the audioband and to 111.6dB when A-weighted. Repeating this test with the single-ended input shorted to ground reduced these ratios by 12dB, but this is still very good performance. Spectral analysis of the Canor's low-frequency noisefloor with the preamplifier outputting a 1kHz tone at 2V with the volume control set to the maximum (fig.2, blue trace) revealed that both the random noise and the supply-related spuriae in the noisefloor were very low in level. Repeating the analysis with the volume control set to –12dB and the level of the input signal increased so that the output was the same 2V (red trace) slightly raised the (relative) level of the noisefloor. The (low) levels of the spuriae at 60Hz and its odd-order harmonics were the same, which suggests these are introduced after the volume control.

Fig.3 plots the THD+noise percentage in the Canor preamp's balanced output against voltage into 100k ohms with the volume control set to the maximum. Actual distortion lies below the low noisefloor up to 2V output. It increases above that level but doesn't reach 1%, our usual definition of clipping, until an extraordinarily high 56V. The unbalanced output clipped at 26V into 100k ohms (fig.4), though the THD+N was higher than it had been in the balanced output. Both output types clipped at 5.2V into 600 ohms.

The balanced output's THD+N percentage, plotted at 4V into 100k ohms, was very low across the audioband (fig.5). I haven't shown how the THD+N varied into 600 ohms because this was affected by the rise in input impedance at low frequencies, but it lay between 0.2% and 0.3% in the midrange and treble at the same 4V. The distortion signature at the same voltage into 100k ohms was primarily the second and third harmonics, both close to a very low –96dB (0.0015%, fig.6).

Intermodulation distortion with an equal mix of 19 and 20kHz tones was very low in level (fig.7), but even with the P1's reduced linearity into 600 ohms it was still very low (fig.8).

Other than the problem with the right channel, which occurred after JVS's audition, the Canor Hyperion P1 offered superb measured performance, with a wide frequency response and very low levels of noise and distortion. This is commendable given that the tubed circuit doesn't use global feedback. However, this preamplifier does need to be used with power amplifiers that have a high input impedance.—John Atkinson