Banner
Banner
cookie-banner

Intro

TESTING AMPLIFIERS

 

rs-upl

 

 

Amplifiers reached ‘perfection’ long ago, in the late 1970s when Hitachi introduced their power MOSFET output transistors and put them into amplifiers like the HMA-7500. It measured perfectly - but the sound? Flat and boring. Amplifier sound quality and its links to measured performance tests have been controversial ever since.


Few tests correlate strongly with sound quality. Frequency response is one that does, that’s why tone controls are effective. Distortion, however, is a far more complex animal. Some smaller distortions are audible and they ‘stain’ the sound, whilst some larger distortions, notably second harmonic, are inaudible except in large quantities – and then the lightening of timbre it produces is not unpleasant.


Steady distortion patterns are less of an issue than constantly changing ones and our experience with World Audio Design amplifiers showed that feedback played a part in this; less feedback meant more distortion but less modulation of the harmonic structure. Listening suggested a compromise was needed. Recent transistor amplifiers are producing steadier distortion patterns, however, with second harmonic dominant and they have a more neutral treble character.


Distortion, however, is far from the only issue in amplifier sound quality. Component quality and quantity, including active devices, has a big impact and this is probably why 1970s wonders like the HMA-7500 sounded so mediocre. Even Quad eventually  admitted that poor component quality compromised earlier amplifiers, something they realised after hearing the results of a rebuild. Our World Audio Design valve amplifiers clearly demonstrated the benefits of using special audio components like Jensen paper-in-oil capacitors and Jelmax Black Gate electrolytic capacitors, Alps and Panasonic potentiometers. Modern amplifiers commonly use components optimised for audio and this improves sound quality, but measurement reveals little change.


Noise is not now a problem, except as hum in valve amplifiers and hiss in Moving Coil phono stages. Channel separation is rarely an issue too, but when the crosstalk signal is heavily distorted this can be an issue.


Damping factor has a direct impact on bass control but a figure of around 20 is the break point for audibility, at least with acoustically and magnetically under damped loudspeakers. See more on this under Damping Factor measurement.


Plenty of amplifiers have mediocre measured performance, typically valve Single Endeds such as those from Almarro, yet they have fine sound quality, and both the 300B and 211 valve when used in good surrounding circuits deliver excellent sonics, even if measurement would seem to suggest otherwise. The biggest limitations of valves are cost and reliability, not a measured performance inferior to solid-state amplifiers.


Amplifiers need to have a sensibly good measured performance, but extreme results such as minuscule distortion of 0.0002% mean little. Happily, most amplifier designers accept this and juggle the many parameters well to arrive at a decent final product.

 

Search

Hi-Fi World, Powered by Joomla!; Hosted by Joomla Wired.