[SoundStage!]Home Audio
Equipment Measurements

How We Test Amplifiers

Our power amplifiers measurements are performed in the laboratory of Bascom H. King (BHK Labs), audio-engineering consultant and equipment reviewer. The various tests are done with an Audio Precision System Two Cascade, the premier piece of audio-measurement equipment. All measurements are performed separately from the subjective evaluation -- the body of the review.

Additional Data Section

This section contains some ancillary measurements in narrative form. Some of the measurements include amplifier gain, signal-to-noise ratio, AC line power draw at idle, input impedance at 1kHz, and output impedance at 50Hz. We measure output impedance by injecting a constant 1 amp of current with frequency into the output of the amplifier under test and measuring the resulting voltage developed accross the amplifier output. Since the current is 1 amp, the measured voltage is equal to the magnitude of the output impedance in ohms.  This value is checked against the output impedance calculated from the voltage drops measured in Chart 1 at the 1-watt output level and is generally in good agreement for the vast majority of tested amplifiers.

Measurement Summary

Power output with 1kHz test signal

This section indicates the power output of the amplifier at 1% and 10% distortion levels and 8-, 4-, and 16-ohm impedances when the amplifier is driven by a 1kHz test signal. In the case of tube amplifiers with stated 8-ohm output terminals, the data was taken for that output. Otherwise, the data is taken on the single pair of output terminals provided. For solid-state amplifiers, no data is provided for output at 16 ohms; all well-designed solid-state amps should have no output-related issues at such a high impedance.

In cases where the 1% and 10% powers are close together, this generally is indicative of a design with quite a bit of negative feedback, typical of solid-state amplifiers. Where the 1% and 10% powers are quite separated, this is typical of tube or solid-state amps with little or no negative feedback. Generally speaking with the tube amplifiers, the best load match for lowest distortion is the impedance that gives the highest power at 1% distortion.

General

The main purpose of this section is to give pertinent details that correspond directly with the charts, to help readers interpret the visual data. Salient points about the chart results and additional data measurements are included, along with other pertinent comments on the amplifier’s behavior.

Chart 1 - Frequency Response as a Function of Output Loading
  • Purpose: Gives an indication of how flat and uniform the frequency response of the amplifier is and how this response varies with output loading. The response with the dummy speaker load suggests how much variation in frequency may occur when the amplifier is actually driving a loudspeaker.

    For the sake of completeness, here is a plot of the dummy load we use. Impedance magnitude is represented in red; phase is represented in magenta. The vertical scale is in ohms (disregard each "m").

nht_dummy.jpg (46887 bytes)

What it tells you: Four measurements can be seen on this chart: frequency responses with open circuit, 8-ohm load, 4-ohm load, and dummy speaker loading at the amplifier’s output. For tube amplifiers with multiple impedance output connections, the 8-ohm output connector is used. The lower the output impedance, the less the output will change with loading and therefore the flatter the response delivered to a speaker load. On the chart, the lower the output impedance, the closer the three resistive loaded curves are to each other on the chart.

Chart 2 - Distortion as a Function of Power Output and Output Loading
  • Purpose: Shows how the amplifier’s distortion (signal components in the output not present in the input) varies with amount of output power and output loading.

    What it tells you:
    Three or four measurements are displayed: 1kHz total harmonic distortion plus noise vs power output for 16-, 8-, and 4-ohm resistive loading on the 8-ohm tap for tube power amplifiers, plus SMPTE IM distortion with 8-ohm loading on the 8-ohm tap. For solid-state amplifiers, the 16-ohm loading is omitted. Typically, solid-state amplifiers will have low distortion up to the start of clipping where the amount of distortion will abruptly rise. Tube amplifiers generally have higher amounts of distortion and merge into clipping more smoothly.
Chart 3 - Distortion as a Function of Power Output and Frequency
  • Purpose: Illustrates how amplifier distortion varies with frequency.

    What it tells you:
    Four measurements are displayed here: Total harmonic distortion vs frequency at four power levels ranging from 1W to a value at or near the rated power of the amplifier. Output loading is shown for either 4- or 8-ohms, as indicated.
Chart 4 - Damping Factor as a Function of Frequency
  • Purpose: Shows how the amplifier's damping factor varies with frequency.

    What it tells you: Damping factor is the value of the output impedance at a particular frequency divided into 8. With tube amplifiers this measurement is made on the 8-ohm output tap if available. In a similar manner to Chart 1, this parameter measures the amplifier’s ability to deliver a flat frequency to the load: The higher the damping factor -- and the lower the impedence, th etwo being inverses of each other -- the flatter the response is into a speaker load. The value of output impedance at 50Hz is given in the Additional Data section.
Chart 5 - Distortion and Noise Spectrum
  • Purpose: Plot of the harmonic distortion spectrum of a 1kHz test signal at an output power of 10W into an 8-ohm load.

    What it tells you:
    Since the frequency axis is logarithmic, it also allows a measure of the leakage of power supply line harmonic frequencies into the amplifier output. Generally, it is preferable if the harmonics decrease rapidly after the second harmonic.

 

[SoundStage!]All Contents
Copyright 2005 SoundStage!
All Rights Reserved