Amplifier Specifications

Author: N.H. Crowhurst

A good amplifier specification will give information about all the types of distortion that we have discussed in order to show how good the amplifier is. The frequency response figure will indicate how closely the amplifier adheres to the same amplification at all frequencies. Sometimes a complete response curve is given and sometimes the specification merely states that the response is within 0.5 db from 20 cycles to 20,000 cycles (or some similar figure). A difference of 0.5 db corresponds with a voltage change of almost 6%, so this means that the amplification will be within 6% of constant through this frequency range.

A distortion figure is also given and, unless otherwise specified, this indicates the amount of harmonic distortion. Unfortunately, it is not usual or convenient to specify what kind of harmonic distortion the figure given may be - entirely second harmonic, third harmonic, or it may be a composite of higher harmonics. This is an unfortunate deficiency of this method of specification.

Amplifier specs indicate harmonic distortion

A good modern amplifier might specify a maximum harmonic distortion of 1%. This means that the total of all the harmonic components produced in the amplification of a single sine wave will be less than 1% of the fundamental, and when all of these voltages are squared, added together, and the square root taken, this square root will still not be more than 1% of the fundamental voltage.

Obtaining harmonics

It is important that, to obtain the total harmonic distortion, one must obtain the square root of the sum of squares of the individual harmonics. The method of combining harmonics is the same regardless of which components are combined. The different components could be second and fifth or any other combination, or a combination of more than just two individual components.

In modern amplifiers in which the harmonic percentage is specified at maximum output, the distortion usually takes the form of clipping on the tops of the waveform, due to the beginning of grid current.

Harmonic % at maximum output (a measure of distortion caused by waveform clipping)

The sharp-peaked distortion component can be analyzed into a series of odd-numbered harmonics, third, fifth, seventh, and so on up the scale. The magnitude of any individual component is quite small compared to the combined peak, as is also the measured combined value. The audible effects as well as the visible one seen on an oscilloscope, however, can still be quite noticeable. It has the sound of a knocking at the fundamental frequency, as when the voice coil of the loudspeaker knocks against its end stops. Even a measured 0.5% distortion of this kind is quite readily audible, as well as visible on the waveform displayed by the oscilloscope.

Distortion may be specified as intermodulation (IM)

Sometimes the distortion is specified as IM (intermodulation). Unless further information is given about the frequencies used for the test, such a specification is valueless. If the first method of IM test is used, the low frequency (its actual value in cycles per second) is important, as well as the ratio between the magnitudes at the two frequencies. The low frequency may be 40 cycles, 60 cycles, or even 100 cycles. Because the handling capacity of an amplifier may be quite different at these three different frequencies, a specification of IM without stating which low frequency was used for the test conveys no real comparison of the performance of different amplifiers.

The peak-to-peak waveform which the amplifier has to handle is the low-frequency voltage plus the high-frequency voltage, because one rides atop the other. Consequently the amount of distortion produced will depend upon the ratio between the two voltages, whether this is 4:1, or as is sometimes used, 1:1. Because the results obtained will depend on the precise nature of the curvature or distortion causing them, there is no ready means of converting figures obtained by one test arrangement into figures that would be obtained using the other test arrangement. Consequently the only safe basis for making comparisons between the performance of different amplifiers is to insure that the same combination is used for both tests.