How Audio Analyzers Measure THD+N Using Notch Filters and FFT

Audio analyzers evaluate the signal integrity of high-fidelity electronics by measuring Total Harmonic Distortion plus Noise (THD+N). This article provides a comprehensive overview of how these specialized instruments use ultra-precise analog notch filters and Fast Fourier Transform (FFT) algorithms to isolate, digitize, and quantify microscopic levels of distortion and background noise in amplifiers, digital-to-analog converters, and preamplifiers.

Understanding Total Harmonic Distortion plus Noise (THD+N)

THD+N is the benchmark metric for evaluating the linearity and quietness of audio equipment. When a pure single-frequency sine wave (the fundamental frequency) passes through an electronic device, non-linearities in the circuitry generate unwanted multiples of that original frequency, known as harmonics. Simultaneously, the device's internal components introduce random electrical hiss and hum, referred to as the noise floor.

THD+N is expressed as a percentage or in decibels (dB). It represents the ratio of the combined sum of all unwanted harmonic distortion and broadband noise to the level of the original fundamental test signal.

The Role of the Analog Notch Filter

Measuring THD+N in modern high-fidelity equipment presents a massive dynamic range challenge. A top-tier audio device might produce distortion and noise that are 120 dB lower than the fundamental test signal. If an audio analyzer fed this combined signal directly into its internal Analog-to-Digital Converter (ADC), the massive voltage of the fundamental frequency would overwhelm the ADC, masking the microscopic distortion components inside the converter's own noise floor.

To solve this, audio analyzers use a high-precision analog notch filter before the ADC stage:

  1. Signal Generation: The analyzer sends a pure sine wave, typically at 1 kHz, into the device under test (DUT).
  2. Measurement of the Total Signal: The analyzer measures the total root-mean-square (RMS) voltage of the output signal coming from the DUT.
  3. Fundamental Suppression: The signal is routed through a sharply tuned analog notch filter centered precisely at the 1 kHz fundamental frequency. This filter attenuates the fundamental tone by 100 dB or more while leaving the rest of the frequency spectrum untouched.
  4. Gain Ranging: With the massive fundamental tone removed, only the tiny residual harmonics and noise remain. The analyzer then amplifies this residual signal so it utilizes the full dynamic range of the analyzer's internal ADC.

How Fast Fourier Transform (FFT) Enhances the Measurement

Traditionally, after the notch filter removed the fundamental frequency, an analog RMS voltmeter simply measured the total energy of whatever was left to calculate THD+N. Modern audio analyzers, however, pass this amplified residual signal into an ADC and apply a Fast Fourier Transform (FFT) algorithm.

The FFT is a mathematical process that converts the digitized time-domain signal into the frequency domain. By combining the analog notch filter with digital FFT processing, the audio analyzer achieves two critical functions:

1. Separating Harmonics from Noise

A standard THD+N reading lumps distortion and noise into a single number. It cannot tell an engineer whether a poor reading is caused by a noisy power supply or clipping transistors. The FFT breaks the residual signal into thousands of narrow frequency bins. This allows the analyzer to display the exact amplitude of the 2nd, 3rd, 4th, and higher-order harmonics separately from the broadband noise floor, allowing engineers to calculate pure THD without the noise component.

2. Digital Notch Refinement

Analog notch filters have physical limitations; their suppression bandwidth has a "skirt" that can slightly attenuate frequencies very close to the fundamental, or fail to completely remove a fundamental frequency that drifts slightly. Once the analog filter removes the bulk of the fundamental's energy to protect the ADC, the FFT acts as an infinitely sharp digital filter. The analyzer software can mathematically zero out the exact frequency bins of the fundamental tone, ensuring extreme measurement accuracy.

The Complete Hybrid Measurement Process

To achieve state-of-the-art measurements, modern audio analyzers execute a hybrid analog-digital workflow in fractions of a second:

  • Step 1: The analyzer generates an ultra-low-distortion sine wave and sends it through the high-fidelity component.
  • Step 2: The analyzer receives the output and records the RMS level of the complete signal (Fundamental + Harmonics + Noise).
  • Step 3: The analog notch filter locks onto the fundamental frequency and strips it away.
  • Step 4: A low-noise preamplifier boosts the remaining distortion and noise to an optimal level for digitization.
  • Step 5: The ADC digitizes the boosted residual signal, and the FFT algorithm maps the exact frequency spectrum.
  • Step 6: The analyzer calculates the RMS sum of the residual spectrum and divides it by the total signal level recorded in Step 2, outputting the final THD+N ratio alongside a detailed visual spectrum of the device's performance.