Created 2008-06-13 20:50:47

Question: When should I use an IMD measurement instead of THD+N?

IMD (Inter-Modulation Distortion) measurements provide useful information about the high frequency non-linearity in a relatively sharply bandwidth limited system that would not otherwise be revealed with THD+N (Total Harmonic Distortion + Noise) testing. Such systems include CD players, portable media players and cassette tape players. Measurements of THD+N using a fundamental that is higher than about 1/3 of the system bandwidth are subject to serious error because one or more potentially significant harmonics will simply fall outside of the audio band. Indeed, THD+N measurements with a fundamental frequency above ½ of the system bandwidth will include NO harmonics at all!

Inter-Modulation Distortion measurements are made by stimulating a device with a complex signal made up of two component sinewaves. A non-linear device will reproduce the two original sinewaves plus a number of IMD products expressed by the equation mF1 + nF2, where m and n are integers and F1 and F2 are the frequencies of the two component sinewaves.

The advantages of IMD measurements are that the two frequencies can be arranged so that the distortion products appear inside the audio band. Traditional THD+N measurements can not practically be made above 7 kHz because the majority of the harmonic products would be above the bandwidth of the device being measured, typically 22 kHz for consumer digital audio devices. In addition typical IMD techniques such as DFD (aka CCIF) are more sensitive because they use high frequency tones, usually above 10 kHz, where many devices exhibit greater sensitivity because distortion usually rises with frequency as the feedback factor falls.

The disadvantage of IMD measurements is that they do not measure noise, which excludes buzz, hum, aliasing products, and all other non-harmonic components present in the output signal of an audio device.