Forms of Distortion

For many years there has been confusion surrounding distortion and its measurement. This paper will not discuss in detail the causes of “static” or “dynamic” distortion, or the many misconceptions surrounding their reduction or elimination. This has been done by others extremely well (Cordell 1979, Hofer 1980). Instead, a comparison of the distortion measurement techniques available today and their relationship to some common distortion mechanisms will be presented. A new distortion measurement technique using numerous simultaneous sinewaves will also be presented and compared with the more conventional techniques under the same test conditions. Many people confuse the causes or forms of distortion with measures of distortion (see for example Frey 1979). This confusion occurs frequently in popular press articles. Harmonic distortion is often thought to be something different from intermodulation distortion. However, the device under consideration has an inherent nonlinearity which will generate harmonics, intermodulation products or both when stimulated by a signal. If the signal is a single tone, only harmonic products may be generated. Both harmonics and intermodulation components will be generated when the excitation is a multitone signal. Digital audio or other sampled data systems may shift the harmonic and intermodulation frequencies due to the folding action of the sampling process. This can, for example, result in harmonics appearing at frequencies which are not harmonic multiples of the input. Since some of the test circuits in this paper demonstrate the measurement of “transient intermodulation distortion” or “TIM” some initial comments concerning it are in order. The word transient implies that the effect is not connected with steady state excitations of the device under test. The word transient carries a mystique which is helpful in marketing audio equipment but really does not apply to the behavior under consideration. The other half of the phrase TIM is also a misnomer. The distortion generated by the common TIM mechanisms will result in harmonics of the input as Jung 1977 and others have shown. These TIM mechanisms are really the early manifestations of slew rate limiting. The distortion mechanism commonly referred to as TIM is simply the result of nonlinear operation of a stage in the amplifier at high frequencies due to the large current required to charge and discharge the compensation capacitor. The term Slewing Induced Distortion (SID) has been suggested by Jung as an alternative to TIM. This is certainly preferable and is correct if interpreted the way he intended. However some people are misled into thinking that the device under test must be slew rate limiting for the distortion to occur. A better term would be Slope Induced Distortion to emphasize the signal rate of change is responsible for exciting the nonlinearity. Also, some have mistakenly assumed that TIM and SID are different mechanisms, but they are merely different names for the same thing.