Measuring vibration frequency requires capturing the motion of a vibrating object, converting that motion into an electrical signal, and then analyzing that signal to identify the specific frequencies present. The most common approach uses an accelerometer attached to the surface, paired with software that performs a mathematical operation called a Fast Fourier Transform (FFT) to reveal the individual frequencies hidden within the vibration. But depending on your situation, you might use anything from a handheld industrial analyzer to a smartphone app.
What You’re Actually Measuring
Vibration has three core properties: frequency (how many times per second the object oscillates, measured in Hertz), amplitude (how far it moves), and phase (the timing relationship between different vibration sources). Frequency is the one that tells you the most about what’s causing a vibration. A motor spinning at 3,600 RPM, for example, produces a vibration at 60 Hz. Bearing wear, misalignment, and imbalance each produce distinctive frequency signatures, which is why frequency measurement is central to diagnosing mechanical problems.
Amplitude itself can be expressed in three ways. The Peak value captures the single largest spike in the signal, useful for detecting sudden impacts. Peak-to-peak measures the full range from lowest to highest point. RMS (Root Mean Square) represents the total energy in the vibration signal, making it the most reliable metric for ongoing monitoring because it isn’t thrown off by a single transient spike the way peak values can be.
Sensors That Capture Vibration
Accelerometers
The workhorse of vibration measurement is the piezoelectric accelerometer. Inside, a small crystal generates an electrical voltage when it’s compressed or flexed by mechanical motion. The voltage output is proportional to the vibration in the direction the sensor faces, and that signal feeds into a recording device or analyzer. Piezoelectric accelerometers are compact, durable, and accurate across a wide frequency range, which is why they dominate industrial and engineering applications.
Laser Doppler Vibrometers
When you can’t physically touch the vibrating surface, a laser Doppler vibrometer (LDV) bounces a laser beam off the object and measures the frequency shift in the reflected light to determine surface velocity. LDVs can detect frequencies from near zero all the way above 1 GHz, with displacement resolution finer than one picometer. That makes them ideal for lightweight structures where mounting a sensor would change the vibration behavior, high-temperature surfaces, or liquids. They’re expensive and typically reserved for labs and specialized engineering work, but their accuracy is exceptional.
Velocity and Displacement Sensors
Velocity sensors measure how fast a surface moves rather than how quickly it accelerates, and displacement sensors (often called proximity probes) track the actual distance of movement. Each type has its strengths for specific frequency ranges. In practice, though, most modern systems use accelerometers and mathematically convert the data to velocity or displacement when needed.
How Mounting Affects Accuracy
The way you attach a sensor to a surface directly controls the highest frequency you can accurately measure. A threaded stud mount provides the most rigid connection and preserves the full frequency range of the accelerometer. Magnet mounts are convenient for quick checks on flat steel surfaces, but the mass and geometry of the magnet matter: a poor fit introduces resonance that distorts high-frequency readings.
Adhesive mounts, foam tape, and improvised methods like zip ties or hose clamps limit your usable range significantly. Testing has shown foam tape stays accurate only up to about 2,000 Hz, while hose clamps and zip ties top out around 800 Hz. Foam dots and poster strips failed to produce reliable results at all. If you need measurements above a few thousand Hertz, stud mounting is the standard choice.
Turning Raw Data Into Frequency: The FFT
A vibrating object rarely oscillates at just one frequency. What your sensor picks up is a complex waveform, the sum of many overlapping sine waves at different frequencies and amplitudes. The Fast Fourier Transform is the algorithm that pulls those individual components apart. It takes your time-based signal (amplitude over time) and converts it into a frequency spectrum (amplitude at each frequency), showing you exactly which frequencies are present and how strong each one is.
The FFT works by mathematically testing the raw waveform against sine waves of progressively higher frequencies to find matches. It requires a signal length that’s a power of two (512, 1,024, 2,048 samples, etc.) and exploits symmetries in sine waves to do this efficiently. With N total samples, you get N/2 frequency “bins,” or distinct frequency lines. The spacing between those bins, your frequency resolution, equals the sample rate divided by the number of samples. Put another way, frequency resolution is the inverse of your total recording time. Record for one second and you get 1 Hz resolution. Record for 10 seconds and you get 0.1 Hz resolution.
The highest frequency you can detect is half the sample rate, known as the Nyquist frequency. So if your system samples at 10,000 times per second, you can measure frequencies up to 5,000 Hz. Anything above that gets misrepresented in your data (a problem called aliasing), which is why most analyzers include filters to block frequencies above the Nyquist limit before sampling.
Equipment Options by Use Case
For industrial maintenance and machinery diagnostics, dedicated vibration analyzers are the standard tool. These handheld devices combine a sensor input, built-in FFT processing, and a display that shows frequency spectra and time waveforms. They often include route-based data collection features so technicians can walk through a facility, take readings at predefined points, and track trends over time.
For engineering and research, a general-purpose data acquisition system paired with accelerometers and analysis software gives you more control over sample rates, filter settings, and post-processing. This setup lets you perform detailed spectral analysis, cross-correlate multiple sensor channels, and export data for custom analysis.
For quick or informal measurements, smartphone apps can tap into the phone’s built-in MEMS accelerometer. Depending on the phone model, the hardware may support output rates from around 12 Hz up to several thousand Hz. However, the apps that access this data often limit the sampling rate. Common accelerometer logging apps cap out between 30 and 60 Hz of sampling frequency, which means you can only reliably measure vibrations up to 15 or 30 Hz. That’s sufficient for detecting low-frequency vibrations like building sway or large machine movement, but far too slow for most rotating machinery.
Basic Frequency Calculations
If you know the physical properties of a vibrating system, you can calculate its natural frequency without any sensor at all. For a simple spring-mass system, the natural frequency in Hertz equals one divided by two times pi, multiplied by the square root of the spring stiffness divided by the mass. Doubling the mass drops the frequency; increasing the stiffness raises it. This is useful for predicting resonance problems during design, though real-world systems are always more complex than a single spring and mass.
When you already have measured data, you can convert between displacement, velocity, and acceleration using the vibration frequency. For simple harmonic motion, these three quantities are mathematically linked through the angular frequency. Velocity equals displacement multiplied by angular frequency, and acceleration equals velocity multiplied by angular frequency again. Most vibration analyzers handle these conversions automatically, letting you view data in whichever unit is most useful for your situation.
Industrial Standards for Evaluation
ISO 20816-1 provides the international framework for measuring and evaluating machine vibration. It establishes general procedures for taking measurements on rotating and non-rotating parts, defines evaluation criteria based on both overall vibration magnitude and changes in vibration over time, and offers guidelines for setting operational limits. Different parts of the standard cover specific machine types, from large generators to pumps and compressors. If you’re setting up a vibration monitoring program for industrial equipment, this standard is your starting reference for deciding what vibration levels are acceptable, marginal, or cause for shutdown.