Mounting resonance is a resonant peak created not by the machine but by the way an accelerometer is attached to it. The sensor mass and the stiffness of its mounting form a spring-mass system with a natural frequency, and near that frequency the sensor grossly exaggerates what it measures. This matters because it sets a hard ceiling on the frequency range you can trust from that measurement, and a poorly mounted sensor pushes the ceiling down and smears false content across the spectrum.
Accelerometer Mounting Resonance in one line: Accelerometer mounting resonance is the resonant frequency of the combined system formed by the accelerometer's mass and the stiffness of its attachment to the machine. Near this frequency the sensor's output is amplified far above the true vibration, so the usable measurement range is limited to well below it, roughly one third of the mounted resonant frequency in practice.
An accelerometer is not weightless, and its attachment to the machine is not infinitely rigid. Together the sensor's mass and the compliance of the coupling behave like a mass on a spring, and every mass-on-a-spring has a natural frequency at which it responds strongly. That is the mounted resonant frequency. Below it the sensor faithfully follows the surface it is attached to; approaching it, the sensor's response rises steeply and it reports far more vibration than is actually present.
The stiffer the mounting, the higher this resonant frequency sits, and the wider the range of frequencies the sensor can measure accurately below it. A rigid metal-to-metal coupling produces a high mounted resonance and a broad usable band. A soft, compliant, or loose coupling produces a low mounted resonance and a narrow usable band. This is why mounting method, not just the sensor itself, determines how high in frequency a measurement can be trusted.
The practical consequence is a rule of thumb widely used by vibration analysts: treat the usable upper frequency limit as roughly one third of the mounted resonant frequency. Below that fraction the resonant amplification is small enough to ignore, so amplitudes read true. This is why a sensor with a high natural frequency in isolation can still deliver a disappointing usable range once it is attached through a soft mount, because the mount, not the sensor, now governs the resonance.
When the mounting resonance sits within or near the measured frequency range, real machine vibration that happens to fall near the resonance gets amplified, producing a peak in the spectrum that is not a genuine machine fault but an artifact of the mount. An analyst who does not recognize it can chase a defect that does not exist, or worse, dismiss the whole measurement. Even broadband energy such as random noise gets amplified near the resonance, raising the apparent amplitude across a band of frequencies.
A classic symptom of a bad mount is the so-called ski-slope spectrum, where the low-frequency end of the spectrum is elevated and slopes downward as frequency increases. This is typically not machine vibration at all but a mounting or sensor artifact, often associated with a poor, loose, or compliant attachment. Its presence is a signal to distrust the low-frequency amplitudes and to check how the sensor is attached before drawing any conclusions.
The remedy is to move the mounting resonance well above the frequencies of interest by mounting more stiffly, and to be aware of where the resonance lies for the mounting method in use. Because the resonance amplifies rather than attenuates, it does not simply hide a signal; it manufactures signal, which is the more dangerous failure mode. Recognizing the artifact for what it is, and keeping the usable band below one third of the mounted resonance, keeps the analysis honest.
The mounting method chosen for a permanently installed sensor directly sets the frequency range that can be fed into a SCADA condition-monitoring system. A stud-mounted accelerometer, threaded into a flat, machined spot on the machine with a thin film of coupling grease, achieves the highest mounted resonance and can read reliably into the high-kilohertz region, which is where early bearing and gear defects show up. That makes stud mounting the standard for fixed points that stream to a historian.
Temporary or convenient mounts trade that bandwidth away. A magnet drops the mounted resonance substantially, and a handheld probe drops it dramatically, sometimes limiting a useful measurement to only a few kilohertz or less. For a walkaround overall reading this may be acceptable, but for a SCADA channel expected to detect high-frequency bearing tones it is not. If the permanent installation uses a weaker mount than the diagnosis requires, the early-warning bands simply never see the energy they are meant to catch.
This is why designing a condition-monitoring input for cloud or SCADA monitoring starts with the mounting, not the sensor datasheet. The engineer decides which fault frequencies must be visible, works out the bandwidth that implies, and then specifies a mounting stiff enough that the mounted resonance keeps that band inside the trustworthy one-third region. Getting the mount right at installation is what makes the streamed data meaningful; a soft mount quietly caps the diagnostic value of every reading that follows.
The one-third rule is a practical guideline that the highest frequency you can trust from a mounted accelerometer is roughly one third of the mounted resonant frequency. Below that fraction the resonant amplification is small enough that amplitudes read close to true, while closer to the resonance the readings are inflated. Because the mounting method sets the resonant frequency, it also sets this usable ceiling.
A ski-slope spectrum, where the low-frequency end is elevated and slopes downward, is usually an artifact of a poor sensor mount or a sensor issue rather than real machine vibration. A loose or compliant attachment lowers the mounting resonance and amplifies low-frequency and noise content, producing the slope. Its appearance is a cue to check the mounting and to treat the low-frequency amplitudes with suspicion.
A magnet couples the accelerometer to the machine less stiffly than a threaded stud, which lowers the mounted resonant frequency. Since the usable range is roughly one third of that resonance, a lower resonance means a lower usable ceiling, so a magnet-mounted sensor cannot reliably measure the high frequencies a stud-mounted one can. That is why permanent points that must detect high-frequency bearing defects are usually stud mounted rather than magnet mounted.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.