You can measure vibration at home using a smartphone accelerometer app, a dedicated USB vibration sensor, or a DIY setup with a microcontroller and accelerometer chip. The right approach depends on what you’re trying to learn: whether your washing machine is shaking too hard, whether traffic is rattling your floors, or whether nearby construction could be damaging your foundation.
What Vibration Actually Measures
Vibration can be captured as three different quantities: displacement (how far something moves), velocity (how fast it moves back and forth), and acceleration (how quickly the speed changes). Each tells you something different, and the tools you use at home will typically report acceleration, measured in units of g-force (where 1g equals the pull of gravity) or meters per second squared (m/s²).
Acceleration sensors are the least expensive and easiest to use, which is why they’re built into every smartphone and most affordable vibration sensors. From an acceleration reading, software can calculate velocity and displacement, so you’re not losing information by starting with acceleration. For most home purposes, acceleration in g or m/s² is the number you’ll work with. If you’re comparing against structural damage guidelines, you’ll want peak particle velocity (PPV), usually expressed in millimeters per second.
Using Your Smartphone
Every modern smartphone contains a small accelerometer, and several free or low-cost apps can turn it into a basic vibration meter. On iOS, apps like “Vibration Meter” or “phyphox” (developed by RWTH Aachen University) display real-time acceleration data. On Android, phyphox and “Accelerometer Analyzer” are solid choices. These apps read the phone’s built-in sensor and can show raw acceleration values, waveforms, and sometimes frequency spectra.
The biggest limitation is sensitivity. Phone accelerometers are designed to detect screen orientation and step counting, not subtle structural vibrations. They typically resolve down to about 0.01g, which is fine for a rattling dryer but may miss low-level floor vibrations from distant traffic. They also top out at relatively low sampling rates, often around 100 to 200 readings per second, which limits their ability to capture high-frequency vibrations above roughly 50 to 100 Hz.
For a quick check of whether your washing machine or HVAC unit is vibrating excessively, a phone app works well enough to give you a ballpark number you can compare against manufacturer specs or share with a repair technician.
How to Mount a Phone or Sensor Correctly
How you attach your measuring device to a surface matters more than most people expect. A loose or poorly coupled sensor will filter out vibrations, especially at higher frequencies, and give you readings that underrepresent what’s actually happening. The vibration signal needs a solid, continuous path from the vibrating surface into the sensor with no gaps, soft materials, or loose joints in between.
If you’re using a phone, place it flat against the surface with the screen facing up. On hard, smooth surfaces like a concrete floor or a metal appliance housing, the phone’s own weight provides reasonable coupling for low-frequency measurements. For better contact, use a thin layer of double-sided tape or a rubber band to hold it firmly against the surface. Avoid placing the phone on carpet, foam, or painted surfaces with flaking layers, as these absorb vibration energy before it reaches the sensor.
For dedicated sensors, the ideal is a stud mount or hard-curing two-part epoxy applied to a clean, bare surface. Paint, rust, and loose coatings act as dampers that block or weaken the vibration signal before it reaches the sensor. If you can’t permanently bond a sensor, clean a small area down to bare material and use a magnetic mount with a flat contact surface. Never mount a sensor to thin sheet metal, plastic covers, or fan shrouds. These vibrate at their own frequencies and will give you data about the cover, not the machine underneath.
Dedicated Vibration Sensors
If your phone’s accelerometer isn’t sensitive enough, standalone USB or Bluetooth vibration sensors offer better resolution and wider frequency range. Products from companies like Digiducer (a calibrated USB accelerometer) plug directly into a laptop and work with free spectrum analysis software. These sensors can detect vibrations well below what a phone can pick up and sample at rates high enough to capture frequencies into the thousands of hertz.
Expect to spend $150 to $400 for a calibrated consumer-grade vibration sensor. The advantage over a phone is that these devices come with known sensitivity specifications, so your readings are traceable to real physical units rather than approximate values from an uncalibrated phone chip.
Building a DIY Vibration Logger
For under $20 in parts, you can build a vibration data logger using an Arduino microcontroller and an MPU-6050 accelerometer module. The MPU-6050 is a combined accelerometer and gyroscope that communicates with the Arduino over a simple four-wire connection. Free libraries and example code are widely available online.
The basic setup requires an Arduino Uno (or any compatible board), an MPU-6050 breakout board, four jumper wires, and a USB cable to connect to your computer. Once wired and programmed, the system streams acceleration data in three axes to your computer’s serial monitor, where you can log it to a file for later analysis. You can also add a micro SD card module to record data without a computer attached, which is useful for overnight monitoring.
This approach is particularly good for long-term logging. If you suspect vibrations are happening at specific times, like when a neighbor’s equipment runs or during certain traffic hours, you can leave the logger running for days and review the data afterward to identify patterns. The MPU-6050 measures up to ±16g and samples at up to 1,000 readings per second, which is more than adequate for most home vibration problems.
Reading a Frequency Spectrum
Raw vibration data, a squiggly line bouncing up and down over time, is hard to interpret on its own. The tool that makes it useful is called a Fast Fourier Transform, or FFT, which converts that messy time signal into a frequency spectrum: a graph showing how much vibration energy exists at each frequency.
This matters because different problems produce vibration at different frequencies. A washing machine drum that’s off-balance produces a strong peak at its spin frequency, typically between 5 and 20 Hz during the spin cycle. If you see that single dominant peak, the fix is usually redistributing the load or leveling the machine. Multiple peaks at evenly spaced intervals (harmonics) suggest something is loose or misaligned, like a motor mount that’s lost a bolt. Scattered, irregular peaks across a wide frequency range can indicate worn bearings.
Several free apps and software tools perform FFT analysis. Phyphox does it directly on your phone. On a computer, Audacity (normally an audio editor) can analyze vibration data recorded through a USB sensor. For Arduino data, you can import logged CSV files into free tools like Python with the numpy library or even into spreadsheet software that supports FFT functions.
What Counts as Too Much Vibration
The answer depends on what you’re measuring and why. For structural concerns, the key number is peak particle velocity. A widely referenced guideline from the United States Bureau of Mines recommends keeping PPV below 50.8 mm/s to prevent cosmetic cracking in residential structures. Some standards are more conservative, recommending no more than 35 mm/s for safe exposure near blasting or heavy construction. If you’re monitoring construction vibrations reaching your home, readings consistently above 35 mm/s are worth documenting and raising with the responsible party.
For human comfort, the thresholds are much lower. Whole-body vibration exposure guidelines reference ISO 2631-1, which evaluates how vibration affects people sitting or standing in buildings. Most people begin to perceive floor vibrations at acceleration levels around 0.01 to 0.02 m/s², and vibrations become annoying well before they pose any structural risk. If your floors vibrate noticeably when a truck passes, the structure is almost certainly fine, but your comfort complaint is still valid.
For appliances, there’s no single universal standard, but context helps. A washing machine in its spin cycle can produce peak accelerations of 1 to 3g at the drum housing. What matters more is how much of that transfers to your floor. If you measure more than about 0.1 to 0.2g on the floor a meter away from the machine, the feet or mounting system probably need attention: leveling, anti-vibration pads, or tightening loose components.
Monitoring Floor Vibrations From Traffic
Low-frequency vibrations from road traffic, trains, or nearby construction are one of the most common reasons people want to measure vibration at home. These vibrations typically fall between 5 and 80 Hz, with heavy truck traffic concentrated in the 10 to 30 Hz range. Research on wood-framed residential floors has shown that these structures have natural frequencies (the frequency at which they most easily vibrate) that can overlap with traffic frequencies, amplifying the sensation.
To capture these vibrations, place your sensor or phone directly on the floor slab or on a ground-floor hard surface, away from the center of a room where floor bounce is greatest. Record for at least 30 minutes to capture passing vehicles or equipment cycles. Running an FFT on the data will reveal which frequencies dominate, and that information is useful if you’re considering mitigation like stiffening a floor, adding mass, or installing isolation pads under your building’s contact points.
If you’re building a case for a complaint to a local authority about construction or traffic vibration, timestamped data logs with PPV values are far more persuasive than subjective descriptions. Even phone-based measurements, while not legally calibrated, demonstrate a pattern that authorities can then verify with professional equipment.