Music Harmonics, Harmonic Distortion, and Negative Feedback – Revised April 6, 2024

Revised April 6, 2024

In the world of acoustics, there are discussions of “harmonics”, and this has application to amplifiers and other components in sound reproduction. Sounds of nature, as well as musical instruments, are distinguished from one another not simply on the basis of whether they are high or low in pitch, but their harmonic content. A sound has what is called a “fundamental” frequency. Humans can hear, by normal air conduction through the ear, frequencies between 20 Hz and 20 kHz. This is only an average, and as we age, we tend to lose the ability to hear high frequencies above about 12 kHz. Secondly, the sensitivity throughout the range is not smooth. We are most sensitive to the range of about 500 Hz to 5 kHz, i.e., the midrange. The fundamental frequency of a sound is the lowest frequency component in that sound. The sound also has harmonics, which are multiples of the fundamental frequency. For example, a sound might have a fundamental frequency of 500 Hz. The second harmonic would be twice the frequency of the fundamental, or 1 kHz, the third harmonic three times the fundamental, or 1.5 kHz, and so on.

Natural sounds (not recorded) contain many harmonics. It is the combination of harmonics that allows us to distinguish a trombone from a French horn when they are both playing the same note. Amplifiers produce an artifact called “harmonic distortion”, which means that they create harmonics that are not in the music, or exaggerate harmonics that are already there. Even-ordered harmonics (second, fourth, sixth, etc.) are pleasing (consonant) to the ear, while odd-ordered harmonics (third, fifth, seventh, etc.) are irritating (dissonant).

Think of it in terms of the sound of a barber shop quartet. If the three harmony parts are all singing in tune with each other and the soloist, all is well as far as the listener is concerned, no matter how loud the harmony parts are singing. This is consonance. However, if just one of the harmony parts sings out of tune, even quietly, the listener finds this irritating. This is dissonance.

Single-ended amplifiers, particularly the ones using tubes, produce even-ordered harmonic distortion, primarily second-ordered. Push pull amplifiers, particularly solid state amplifiers, tend to produce odd-ordered harmonic distortion. Strangely enough, balanced amplifiers tend to reduce even-ordered harmonics. They have more power, but at the cost of the consonant harmonics.

If you are purchasing a single-ended triode tube amplifier, 1% total harmonic distortion (THD), which means 1% of the sound is harmonically distorted, is actually pleasing to the ear, because it is mostly even-ordered. However, with a solid state amplifier, THD as low as 0.1% can be irritating, because it tends to be mostly odd-ordered distortion. Modern solid state amplifiers all meet this specification, but what happens at high volume (higher voltage output from the amplifier) differentiates the inexpensive amplifiers from the audiophile-grade amplifiers.

On the other hand, the bottom line in all cases, is whether the sound pleases you or not. “Specsmanship” as it is called (emphasizing the specifications on the technical data sheet supplied with a component) should not be the deciding factor. Decide with your ears, but take a look at the technical data sheet before you write the check.

Negative feedback is the process of taking a portion of the output, electrically inverting it, and then feeding it back to the input. The purpose of negative feedback is to flatten the frequency response, reduce harmonic distortion, lower the output impedance, and also to reduce the effects of “parasitic oscillation” that can occur when parts of the circuitry cause an induced current to flow in other circuits where it is not wanted.

However, too much negative feedback can sometimes be used to produce an artificially low harmonic distortion specification. Second-ordered harmonics will be reduced, but fifth order harmonics will increase, and this odd-ordered distortion is much more noticeable than the original second-ordered distortion. A tell tale sign of too much negative feedback is an excess of sibilance in the human voice (the “s” is exaggerated).