Integrated Amplifiers

Western Electric's own spec for the 91E – 2x14W, 16W and 20W/8ohm at 3%, 5% and 10% THD, respectively – is higher than typically achieved with 300B/SETs but its 'SCS' circuit arguably departs from purist principles, influencing not only power. The 2.1-2.8ohm output impedance of the 8ohm transformer tap will modify the amp/speaker frequency response in line with the varying load of the latter even though, into a non-reactive 8ohm, the 91E offers a response that's flat to –1dB from 14Hz-21.4kHz before rolling away to 100kHz/–16dB. In practice, power output was a shade lower than specified into 8ohm at 4W/1% THD, 10W/3% and 11W/5% but bang-on target at 5W/1% THD, 14W/3% and 17W/5%. Under dynamic conditions this translates to 12.3W and 15.3W into 8 and 4ohm loads but just 3.2W and 1.5W into low 2 and 1ohm loads [see Graph 1]. The trend of THD vs. output level is clear from Graph 1 as is the increase at low and high frequency extremes – from 0.45%/1kHz to 1.0%/20Hz and 2.8%/20kHz (all re. 0dBW) – in Graph 2, below. Finally, the A-wtd S/N ratio is a creditable 83.8dB (re. 0dBW).

WE's resistor-ladder volume control operates over a ~100dB range where its 0.5dB steps are most accurate from volume '90' down to '40', but are closer to ±0.6dB from '30' to '10', and less reliable below '09'. The BT input, meanwhile, offers a 94.8dB A-wtd S/N (re. 0dBV via the line out) and linearity good to ±0.1dB over 60dB and ±0.7dB over 90dB (data is muted at 96dB/16-bits). Distortion is a minimum of 0.003-0.005% but lossy compression kicks in at just –50dB above 14kHz while the response (measured using a multitone to saturate the aptX bit pool) rolls away by 5dB from 100Hz to 1kHz and 9dB from 1kHz to 20kHz. PM