Understanding Spectral Leakage

Devices

Portable balancer & Vibration analyzer Balanset-1A

€1,975.00 + VAT (if applicable)

Parts

Vibration sensor

€90.00 + VAT (if applicable)

Parts

Optical Sensor (Laser Tachometer)

€124.00 + VAT (if applicable)

Devices

Balanset-4

€6,803.00 + VAT (if applicable)

Parts

Magnetic Stand Insize-60-kgf

€46.00 + VAT (if applicable)

Parts

Reflective tape

€10.00 + VAT (if applicable)

Devices

Dynamic balancer “Balanset-1A” OEM

€1,735.00 + VAT (if applicable)

Spectral leakage is a form of measurement error that arises during the Fast Fourier Transform (FFT) analysis of a signal. It is the “smearing”, or spreading, of energy from a single discrete frequency peak into the spectrum’s adjacent frequency bins. This smearing distorts both the amplitude and the apparent frequency of the true vibration component, and it can mask smaller signals or lead to an inaccurate diagnosis. Understanding it is essential to trusting any FFT result.

1. Definition: What is Spectral Leakage?

In an ideal world, a pure sinusoid at one frequency would appear in the spectrum as a single, infinitely thin line. Spectral leakage is what happens in the real world instead: the energy that should sit in one FFT bin “leaks” sideways into neighbouring bins, producing a peak with broad skirts rather than a sharp spike. The result is a spectrum that looks fuzzier and noisier than the underlying physics warrants, which matters most when you are trying to separate a small fault signal from a large nearby peak.

2. The Root Cause: Discontinuity

Spectral leakage stems from a violation of the FFT’s fundamental assumption. The algorithm assumes that the finite block of time-waveform data it analyses is one perfectly repeating cycle of a periodic signal. For that to hold, the signal’s value at the very end of the block must be identical to its value at the very beginning, so the block could be looped end-to-end seamlessly.

In practice, when measuring a real vibration signal it is almost impossible to capture a block that contains an exact integer number of cycles for every frequency component present. The result is a discontinuity: the end of the captured signal does not line up with the beginning. The FFT interprets this sudden jump as a high-frequency transient — much like an impact — and that artificial transient carries energy that was never in the original signal. It is this spurious energy that leaks out across a wide range of frequencies in the resulting spectrum.

The shorter the data block and the closer two real peaks lie together, the more damaging leakage becomes — which is why leakage, frequency resolution and block length are always discussed together.

3. The Effects of Spectral Leakage

The smearing of energy produces two main negative effects:

1. Reduced amplitude accuracy: energy that should have been concentrated in a single bin is now spread across many. The main peak therefore reads lower than its true amplitude, while the adjacent “sidelobe” bins are artificially raised. An amplitude read straight off a leaky peak can be misleading for severity assessment.
2. Reduced frequency resolution: leakage can be severe enough to completely hide smaller, nearby peaks. A faint signal from an early bearing defect, for example, can be lost entirely in the broad skirt of leakage from a large 1× unbalance peak.

Both effects work directly against the analyst’s goals: accurate amplitudes for trending and severity, and clean resolution for early fault detection.

4. The Solution: Windowing

Spectral leakage is controlled with windowing functions. A window is a mathematical weighting function multiplied with the time-waveform data before it is passed to the FFT.

The most common choice for general rotating-machinery work is the Hanning window. It has a smooth, bell-shaped profile that tapers the signal down to zero at both the start and the end of the block. This tapering forces the two ends to match, effectively removing the artificial discontinuity that caused the leakage in the first place. By presenting the FFT with a smoothly periodic signal, windowing dramatically reduces leakage — yielding sharper peaks, a lower noise floor and more sensitive analysis.

Windowing is a trade-off rather than a cure. The same tapering that suppresses leakage also slightly widens the main peak and lowers its measured amplitude, which is why instruments apply an amplitude-correction factor. Different windows trade these properties differently: a flat-top window is preferred when accurate amplitude of a single tone matters (for example during calibration), a uniform (rectangular) window suits transient capture in a bump test, while Hanning remains the everyday default.

5. Why It Matters in Practice

For the field engineer the lesson is simple: a clean spectrum is a prerequisite for a sound diagnosis. Leakage that buries a small bearing tone or understates a peak’s amplitude can send an investigation in the wrong direction. When measuring 1× amplitude and phase for a balancing job — the routine task a portable instrument such as the Balanset-1A performs in the machine’s own bearings — appropriate windowing keeps that synchronous peak sharp and its amplitude reliable, so the computed correction is based on the true vibration rather than a smeared artefact.

← Back to Main Index