[RadioSvit]

Hello guys. An interesting topic about distortion and as always I miss the beginning of interesting topics and am late for them. If anyone is interested, I will share my experience. In the middle... in the late 80s I worked as a technician in the metrology laboratory of a large enterprise and even then I made myself a transistor stereo amplifier according to the Kvod 405 scheme, and as a technician I measured the amount of distortion. I remember well that it achieved the promised distortion of 0.05%. I still have this amplifier somewhere in the attic. And fifteen years ago I experimented a lot with tube amplifiers because I wanted to create their small-scale production. So first I made a measuring test stand for measuring distortion and the harmonic spectrum using a computer and the SpectroLab program. I took many characteristics from triodes and pentodes in different modes and created a database. So - small-signal triodes octal triodes 6SN7 and 6SL7 have a very short spectrum of harmonics - a maximum of up to the fourth and the level of the last harmonic is already at the level of lamp noise. But the level of the second harmonic is very high (2...8%), but it is melodious and does not irritate the ear. The picture is completely different with pentodes - here the spectrum of harmonics can be very long, up to the 11th harmonic. And although the magnitude of harmonics decreases with increasing their number, their perceptibility by the ear increases. And therefore 0.05% of the 11th harmonic will be much more noticeable than 3% of the 2nd harmonic. And most importantly, even harmonics do not dissonant, and odd ones are very irritating to the ear, especially when they have a large number - for example, 9...11...13. Single-ended triode tube amplifiers produce mainly even harmonics of the 2nd and 4th order, pentodes add the 3rd, 5th, 7th and sometimes the 9th harmonic. But their magnitude quickly decreases. Therefore, the sound is harmonious and pleasant. In push-pull amplifiers, the picture is different - both triode and pentode push-pulls mutually destroy even harmonics that are pleasant to the ear, but add odd ones that are dissonant. The sound becomes more dynamic, but uncomfortable due to the high level of odd harmonics. How to fix this? Only by introducing negative feedback. But even a shallow negative feedback of 20 dB kills the dynamics and increases the number of harmonics to the 13th. A closed loop. ) Therefore, music lovers and audiophiles love triode single-ended amplifiers, which have large distortions of the 2nd harmonic, which creates an atmosphere of warm tube sound. ) With transistor amplifiers, it is even worse - due to the curvilinearity of the volt-ampere characteristics of transistors, it is necessary to introduce deeper negative feedbacks of 40...60 dB, and although this reduces the total value of the harmonic level to 0.01%, it increases their order to number 15 and more, which, as is known, are more noticeable than 0.5% of the 5th harmonic. This is where the reasons for the "transistor sound" are hidden. So, much more important for the quality of the sound is not the total value of the harmonics, but their spectrum, when even harmonics prevail over odd ones (as in single-ended amplifiers) and a short harmonic tail, the shorter it is, the more pleasant the sound.
Phew ... I refreshed my knowledge of audio in my memory.