LaBoRatoRY test ResULts

Newport Test Labs tested the power output of the Audia Flight FL Three S using its standard test procedures, which resulted in the nine measurements that are tabulated in the accompanying test result table as well as shown visually in the bar graphs. You can see that the Audia Flight FL Three S easily met its rated output power figures (of 100-watts into 8Ω and 160-watts into 4Ω) at the 20Hz and 1kHz test frequencies, but fell around 0.7dB short at 20kHz when driving 8Ω loads and 1.2dB short at 20kHz when driving 4Ω loads. The slight limitation at this frequency was not due to amplifier clipping, but instead because the waveform distortion exceeded 3.0% THD.

Driven into 8Ω loads at 1kHz, the Audia Flight FL Three S delivered 110-watts per channel with a single channel driven, and 106-watts with both channels driven. Driven into 4Ω loads at 1kHz, the Audia Flight FL Three S delivered 187-watts per channel with a single channel driven and 170-watts with both channels driven. In practise, due to the distribution of power across the audio spectrum when any amplifier is reproducing music, rather than test signals, the Audia Flight FL Three S will always deliver high frequencies at the correct levels, despite not being able to deliver its full rated power at high frequencies with continuous test signals.

Newport Test Labs usually tests power output into 2Ω loads as well, but in the case of the FL Three S, Audia Flight has built in an output current protection circuit that kicks in slowly (four seconds) if the amplifier is asked to deliver full power into any load of less than 2.5Ω, and quickly if it detects a short circuit. This protection circuit prevented any full-power testing into 2Ω loads, since it takes more than four seconds to make a continuous power output measurement.

Channel separation was an excellent 107dB at low frequencies and still a very good 80dB at 1kHz. The result measured at 20kHz (55dB) is more than sufficient to ensure perfect channel separation and stereo imaging, but I am used to seeing higher figures measured by Newport

Test Labs for this particular test. Channel balance was excellent at 0.02dB (at 1kHz) and interchannel phase also excellent, measuring just 0.01° at 20Hz, 0.05° at 1kHz and 1.11° at 20kHz. These results are particularly good since Audia Flight says this amplifier is essentially a ‘dual mono’ design.

Distortion at one watt was fairly high for both 8 and 4 test loads, with Newport Test Labs measuring overall THD+N of 0.14% for a 1kHz test signal. The structure for the distortion into an 8 load is shown in Graph 1. You can see that both the second and third harmonics are sitting at around –60dB (0.1%) with a fourth at –90dB (0.0031%) a fifth at –71dB (0.0281%) and a sixth at –88dB (0.0039%). The six higher-order components visible on this graph are all around 100dB (0.001%) or more down.

Graph 2 shows distortion into a 4 load at an output of one watt. The second harmonic sticks on –60dB (0.1%) but the third harmonic component increases to –54dB (0.1995%). The fifth harmonic also increases (to –60dB or 0.1%), with the sixth harmonic coming up to –70dB (0.0316%). At this lower impedance, higher-order harmonics rise in level compared to into an 8 load, but with two exceptions (the 8th and 9th) all are more than 90dB down (0.0031%). Overall distortion gets very slightly higher at rated output, with Newport Test Labs measuring total THD+N at 0.15% and interestingly, there is almost no difference between the measurements into 8 and 4 loads. The second harmonic distortion is at –60dB (0.1%), the third and fourth are at around –70dB (0.0316%) and the sixth and seventh at around –77dB (0.0141%). Then, right out to the 17th harmonic, all are around –70dB (0.0316%).

Intermodulation distortion (CCIF) followed the same pattern as the harmonic distortion measurements, in that they’re all more like the distortion I expect to see from a valve amplifier than what I expect from a solid-state amplifier. The two test signals at 19kHz and 20kHz (Graph 5) result in a regenerated signal down at 1kHz that’s only 58dB down, plus the high sidebands also result in additional signals at 2kHz, 3kHz and 4kHz.

The noise of the FL Three S was low, with Newport Test Labs measuring 81dB unweighted referenced to one watt, a figure that increased to 88dB with A-weighting. Referenced to rated output, the overall figures measured were 99dB (unweighted) and 107dB (A-weighted). For an integrated amplifier to so significantly break through the 100dB barrier (clearing it by 7dB, in fact) is quite an achievement. (It also exceeded its manufacturer’s claim of 95dB by a long shot!) You can also see the distribution of the noise across the audio band from Graph 1, with the noise mostly more than 120dB down: it’s only the low-frequency, mostly mains-related noise that is above that level.

For an integrated amplifier to so significantly break through the 100dB barrier (clearing it by 7dB, in fact) is quite an achievement.

The frequency response of the Audia Flight FL Three S was very extended, extending from less than 1Hz to 510kHz (–3dB) and from less than 1Hz to 251kHz (–1dB). This makes this amplifier an extremely wide-bandwidth design. As you’d expect from this, the frequency response within the audio band into a standard 8 non-inductive laboratory test load is absolutely ruler-flat, as you can see from the black trace on Graph 6. The trace doesn’t start rolling off at low frequencies until 4Hz, and is only 0.1dB down at 5Hz, which is the graphing limit. The FL Three S’s frequency response into a load that simulates a typical two-way loudspeaker (red trace) was almost equally good, with Newport Test Labs measuring an overall response for this load as extending from 5Hz to 40kHz (±0.2dB).

As you’d expect given the enormous bandwidth and the amplifier’s linearity, the square wave responses recorded by Newport Test Labs were outstandingly good, with the 100Hz and 1kHz square waves in particular looking like they’d come straight from the signal generator, rather than via the Audia Flight FL Three S. The 10kHz square wave shows just the slightest hint of rounding on the leading edge, but is otherwise perfect. The response into a highly reactive load (1kHz cap) was also exceptionally good, with basically just a single overshoot to barely 1/8th wave height that was damped within one cycle. It appears this was a design intention that resulted from the company’s decision to use current feedback rather than voltage feedback, with the result that, in the company’s words: ‘ the feedback is closed before the output stage … the resulting outcome consists of fast, stable circuits with the utmost ability to control even the wildest reactive loads.’

Output impedance was relatively high, at 0.3 , which as a direct consequence means damping factor was relatively low, at 26 (@1kHz), again meaning the Audia Flight FL Three S will be more valve-like in its relationship with the loudspeakers it’s driving.

The Audia Flight FL Three S’s test results show all the hallmarks of an amplifier that’s been designed to sound good, rather than measure well, yet it still performed very well on Newport Test Labs’ test bench, though the results returned alternated between those that would be typical of a good valve amplifier (i.e. THD, IMD, DF) and those that would be typical of a state-of-the-art solid-state design (i.e. frequency response, bandwidth, signal-to-noise ratios, square wave performance). Steve Holding