LABORATORY TEST REPORT
Newport Test Labs measured the power output of the Hegel H590 at 1kHz as 295-watts per channel when both channels were driven into 8Ω non-inductive loads. This is a touch under the claimed 301-watt specification, but then again Hegel doesn’t specify whether its 301-watt rating was measured with one or two channels driven, or the frequency at which it was measured. And, as you can see from the tabulated results, Newport
Test Labs did measure an above-specification 303-watts per channel (both channels driven) when using a 20Hz signal, and another above-average specification 305-watts single channel driven at the same test frequency. At 20kHz, output dropped to around 280-watts per channel single channel driven and 276-watts when both channels were driven. However, even if we take the lowest measured result (276-watts), this is only 0.3dB lower than specification, a difference too small to be noticeable.
Hegel does not specify power output into 4Ω loads (though it does specify that the H590 will drive 2Ω loads), but Newport Test Labs measured it with this load anyway, and you can see that at 1kHz, the Hegel H590 delivered 552-watts per channel when a single-channel was driven, and 530-watts per channel both channels driven. Again, the maximum power output increased slightly at low frequencies, and decreased slightly at high frequencies.
Frequency response was extremely flat, and more than adequately extended, with Newport Test Labs measuring the overall frequency response as 3Hz to 210kHz –3dB. With a slightly tighter level variation window, the measured result was 7Hz to 120kHz ±0.5dB. Across the audio bandwidth Newport Test Labs measured the Hegel H590’s frequency response at 20Hz to 20kHz ±0.5dB, irrespective of whether it was driving a standard 8Ω non-inductive laboratory test resistor (the black trace on Graph 7) or a reactive load simulating the load which would be presented to it by a typical two-way loudspeaker (the red trace on Graph 7). You can see from this graph that the deviation in the response (tiny though it is) essentially occurs only due to the amplifier’s low-frequency roll-off.
Above 70Hz the response is ruler-flat to 20kHz and only around 0.05dB down at the 40kHz upper graphing limit.
Channel separation was outstanding at low frequencies (121dB), superior at midrange frequencies (112dB) but slightly less than stellar at high frequencies (77dB). That said, 77dB is far more than required to provide channel separation and stereo imaging beyond what the human ear is capable of resolving. Interchannel phase was excellent at low and mid-frequencies—0.01° at 20Hz and 0.03° at 1kHz, with the error increasing to an again-inaudible 0.99° at 20kHz. Channel balance was 0.1dB at 1kHz.
Distortion at an output of 1-watt into 8Ω was very low. Newport Test Labs’ spectrogram shows only a few distortion components, all of which are at or more than 100dB down. As you can see from Graph 1, there’s a second harmonic component at –104dB (0.0006%), a third at –100dB (0.001%), an eighth also at –100dB plus one at 15kHz at –110dB (0.0003%) and another at 20kHz at –105dB (0.00056%). Overall THD+N at 1-watt was 0.005%.
Distortion at an output of 1-watt into 4Ω was also very low though, with the exception of the 2nd harmonic at 2kHz, it was fractionally higher than it was into 8Ω loads. Significantly, the Hegel H590’s noise floor referenced to the 1-watt output was down around
–128dB except at low frequencies, suggesting that most of the noise measured by Newport Test Labs (as shown in the tabulated results) was mains-related.
At higher output levels (20-watts), the Hegel’s distortion signature was almost the same irrespective of load impedance, but the H590’s performance into 8Ω was clearly superior. Into 8Ω (Graph 3) you can see a second harmonic at –96dB (0.0015%), a third at –90dB (0.0031%) and then five higher-order harmonics that vary in level between –110dB (0.0003%) and –130dB (0.00003%). Into 4Ω the third harmonic component increases to –80dB (0.01%) and the fifth to –110dB (0.0003%) but most others are reduced or remain the same. Note that thanks to the higher 0dB reference, the overall noise floor has dropped further, to –140dB, and you can see the mains noise contribution has also dropped equivalently.
So far as overall noise levels are concerned, Newport Test Labs measured results of 72dB unweighted and 86dB A-weighted referred to 1-watt, and 97dB unweighted and 111dB A-weighted referred to rated output. These last two results are spectacularly good for an integrated amplifier with an on-board DAC.
Intermodulation distortion (CCIF) was very low for both the tested levels (1-watt and 20-watts). The sidebands immediately adjacent to the two test signals were
95dB (0.0017%) down at 1-watt and 90dB (0.0031%) down at 20-watts, while the regenerated difference signals were 94dB (0.0019%) down at 1-watt and 83dB (0.007%) down at 20-watts.
In all the square wave testing performed by Newport Test Labs the Hegel H590 returned excellent waveforms, as you can see for yourself from the oscillograms. At 100Hz, there’s the tilt you’d expect from a non-d.c. amplifier, but none of the bending that would be indicative of low-frequency phase shift. The 1kHz wave is nigh-on perfectly shaped, while the 10kHz square wave is very nearly as good as the 1kHz wave, exhibiting a very fast rise-time… rather as you’d expect from an amplifier whose frequency response is so flat and extends beyond 210kHz. Square wave performance into a highly reactive load (2μF in parallel with 8Ω) is excellent, with very little initial overshoot and very quickly damped ringing (just three cycles), suggesting that the H590 will be unconditionally stable into any loudspeaker load, and easily up to the task of driving electrostatic loudspeakers.
As suggested by the H590’s frequency responses into a typical loudspeaker load, its output impedance was measured by Newport Test Labs as 0.015Ω at 1kHz, which translates to a damping factor of more than 500, which
is in turn more than ten times more than required to ensure perfect control over any bass driver. Overall amplifier gain was measured at 32.8dB, with the amplifier requiring a voltage of 1.15V at its balanced input in order to deliver its rated output power of 301-watts.
As you can see from the tabulated figures for power consumption, the H590’s circuitry is not overly efficient, consuming nearly 100-watts when idling and close to a thousand watts when operating at full power. Interestingly, according to Newport Test Labs’ measurement, the amplifier still draws nearly 30-watts in its standby mode, which is no-where near the 0.5-watts it should be in order to conform to the Australian government’s legislation regarding standby power consumption. Given this level of stand-by consumption, I would recommend turning the amplifier off completely whenever it’s not going to be used again for more than 24 hours.
Overall, the Hegel H590 returned excellent results in all the tests performed on it by Newport Test Labs. It’s clearly a very well-designed, high-performance amplifier. Steve Holding