LABORATORY TEST REPORT
The reason Cyrus specifies power output into 6 ohms rather than the industrystandard (and US, EU and Australian standard) 8-ohm load is evident from the tabulated chart showing the measurements that Newport Test Labs made of the i7-XR’s power output with both channels driven, which you can see was 46 watts per channel (16.6dBw) at all three frequencies tested (20Hz, 1kHz and 20kHz). Being able to claim an output above 50 watts per channel is more impressive than claiming an output power below this. Power output increases slightly (1.4dB) when only a single channel is driven, which suggests that the power supply is unregulated.
Newport Test Labs measured the bothchannels-driven power output into 4-ohm loads as 67 watts (18.3dBw) per channel at 1kHz, but at 20Hz, it was only 58 watts (17.6dBw) per channel, which reveals a power supply limitation that is presumably overcome by using the outboard power supply that’s optionally available for this model. The ‘PRO’ on the tabulated chart showing power output at 20kHz is because the i7-XR’s onboard protection circuitry triggered whenever the amplifier was tasked with delivering a continuous high-frequency signal into a low-impedance load. The amplifier shut itself down in this situation and showed a front-panel display that reads “Error. Amp Over Current.”
Separation between the left and right stereo channels was outstandingly good at low and midrange frequencies, with Newport Test Labs measuring 104dB at 20Hz and 106dB at 1kHz and still a very respectable 77dB at 20kHz. Channel balance was also a very good 0.03dB at 1kHz, which is far better than will ever be required.
Frequency response was very flat and much extended at both ends of the audio spectrum, with Newport Test Labs measuring the wideband –3dB points of the i7-XR’s frequency response at lower than 1Hz and at 70kHz. The same result was measured at low frequencies for the 1dB downpoint, with the higher frequency dropping down to 38kHz. Linearity across the audio bandwidth (and a bit beyond) is shown in Graph 1 both when the amplifier is driving a standard noninductive 8-ohm laboratory test load (black trace) and also when it’s driving a load that simulates the load that would be presented by a typical two-way bass reflex loudspeaker (red trace).
As you can see, despite the greatly
magnified vertical scale of this graph, where the top of the graph represents +3dB and the bottom –3dB, the two traces essentially overlay each other, which is an excellent result and means that the amplifier will essentially ‘sound’ the same no matter what loudspeakers you use in conjunction with it. In terms of linearity, you can see that the response is absolutely ruler-flat from 10Hz up to 3kHz, then rolls off very, very slightly, to be 0.26dB down at 20kHz. This puts the normalised audio band response at 20Hz to 20kHz ±0.13dB.
Graph 2 shows distortion in the output when the Cyrus i7-XR is delivering 1kHz into an 8-ohm load at a level of one watt.
Although there are multiple harmonically related components visible on the graph, all except two (the third-order harmonic component at –101dB [0.00089% THD] and the fifth-order at –108dB [0.00039%]) of them are more than 110dB down, so each one contributes less than 0.00031% to total THD. The odd-order distortion components
are more prominent than the even-order components (it would have been preferable for this situation to be reversed). There also seems to be a non-harmonically related signal at a frequency a tad below 20kHz whose presence I cannot explain.
You can see on Graph 2 that the overall noise floor is very low, more than 120dB down across the audio spectrum except at low frequencies, where some power supply noise is visible at around –90dB. This is an excellent result. You should note, however, that these noise measurements are each at a specific frequency, not an overall wideband measurement of noise. You can see these measurements in the tabulated results, where Newport Test Labs measured 84dB unweighted, and 89dB A-weighted, both referred to a one-watt reference output level.
Distortion performance into a 4-ohm load is shown in Graph 3 and you can see that it’s essentially similar to that into an 8-ohm load, which is very desirable, and although the third harmonic is higher in level than into 8 ohms, almost all other harmonics are slightly lower. However, all of this is of academic interest only, since none of it would be audible, being too low in level. Overall THD was measured at 0.007% at one watt, and 0.001% at rated output.
Tested at higher power output levels
(20 watts) into 8-ohm and 4-ohm loads (Graphs 4 and 5), you can see that into 8 ohms, all distortion components are more than 110dB down (0.00031%) and, apart from the four lowest-order components, all are more than 120dB down (0.0001%). Interestingly, however, the structure of the harmonic distortion components becomes completely different when the amplifier is driving a 4-ohm load at the higher power level, with more and higher-level odd harmonics at higher frequencies. The noise floor, of course, is lower on both graphs due to the higher 0dB reference at 20 watts. Noise referred to rated output was measured as 98dB unweighted and 103dB A-weighted, both of which are excellent results.
Intermodulation distortion is shown for two power levels — one-watt (Graph 6) and 20-watts (Graph 7). The primary differences between the two are the almost-complete absence of a regenerated 1kHz difference signal on Graph 6 and its presence at around –105dB (0.00056%) on Graph 7 and the expected difference in the noise floors. Overall, these results are very good.
Newport Test Labs also tested the Cyrus i7-XR’s square wave response at different frequencies and into different loads. The 100Hz response is excellent, reflecting the i7-XR’s extended low-frequency response, and also demonstrating a complete absence of any phase errors. Also, the 1kHz square wave is almost perfect — as if it came straight from the test generator itself, rather than the amplifier.
The 10kHz square wave shows some rounding on the leading edge, which reflects the i7-XR’s 3dB high-frequency downpoint of 70kHz, but it’s very clean and stable and an excellent result. Tested at 1kHz with a highly capacitive load (the fourth square wave in the series), the Cyrus i7-XR was not only completely stable but also exhibited no ringing in its output at all, which is something I expect from valve amplifiers and has been demonstrated to correlate well with amplifiers that ‘sound good’ in subjective evaluations — so it’s a definite plus for the Cyrus i7-XR on this measurement.
Also excellent was Newport Test Labs’ measurement of a damping factor of 800 at 1kHz, which comes off a measured output impedance of just 0.01 ohms at this frequency, which means that the Cyrus i7XR will be easily able to control unwanted cone motion in loudspeakers with heavy, undamped cones.
Overall, the results measured by Newport Test Labs reveal excellent performance from an obviously well-engineered amplifier.
The 1kHz square wave is almost perfect — as if it came straight from the test generator itself, rather than the amplifier