How do I know if vibration is caused by unbalance or misalignment?

Unbalance produces a dominant peak at exactly 1× RPM (shaft speed) in the radial direction, with stable phase and amplitude proportional to speed². Misalignment produces significant 2× RPM component (often equal to or larger than 1×), strong axial vibration, and 180° phase shift across the coupling.

What are bearing defect frequencies (BPFO, BPFI, BSF, FTF)?

These are characteristic frequencies generated when a rolling element passes over a defect on a specific bearing component. BPFO (Ball Pass Frequency Outer race), BPFI (Ball Pass Frequency Inner race), BSF (Ball Spin Frequency), FTF (Fundamental Train Frequency). They depend on bearing geometry and shaft speed.

What is a good vibration level for rotating machinery?

ISO 10816-1 classifies vibration severity by machine class. For typical industrial machines (Class II, 15-75 kW): Zone A (good) 11.2 mm/s.

Can the Balanset-1A be used for vibration analysis?

Yes. The Balanset-1A includes a 2-channel vibration analyser with FFT spectrum display (F1 mode), vibrometer for overall/1× comparison (F5 mode), and detailed spectrum charts (F8 mode). Both channels can be used simultaneously for radial+axial comparison.

What is the difference between time waveform and FFT spectrum?

The time waveform shows vibration amplitude over time — complex when multiple faults are present. FFT decomposes this into individual frequency components, creating a spectrum where each peak corresponds to a specific source.

Vibration Analysis — Beginner's Guide to Spectrum Diagnostics — Vibromera

Vibration Analysis — Spectrum Diagnostics Guide

Q: What is vibration analysis?

Vibration analysis is the process of measuring and interpreting mechanical oscillations of rotating machinery to diagnose faults. Using FFT (Fast Fourier Transform), the complex vibration signal is decomposed into individual frequency components. Each fault type produces a characteristic 'fingerprint' in the frequency spectrum — unbalance at 1× RPM, misalignment at 2×, looseness as multiple harmonics, bearing defects at non-synchronous frequencies (BPFO, BPFI, BSF, FTF).

Q: How often should I measure vibration?

Critical equipment: weekly or biweekly. Standard production: monthly. Auxiliary/non-critical: quarterly. After any maintenance: immediately to establish new baseline.

Q: What causes subharmonic vibration (0.5×)?

Subharmonics at 0.5× RPM typically indicate severe mechanical looseness, oil whirl in journal bearings, or cracked shafts. Half-order peaks arise from impacts occurring every other revolution.

📌 Canonical Reference Article — vibromera.eu

From FFT fundamentals to fault diagnosis: learn to read vibration spectrums, calculate bearing defect frequencies, assess severity per ISO 10816, and diagnose unbalance, misalignment, looseness, bearing and gear defects — with interactive tools and the Balanset-1A.

Interactive Diagnostic Calculators

Essential tools for vibration analysis — bearing defect frequencies, gear mesh frequency, severity assessment, and unit conversion

🔵 Bearing Defect Frequency Calculator

Quick Select — Common Bearing

Shaft Speed (RPM)

Number of Rolling Elements (n)

Ball / Roller Diameter, Bd (mm)

Pitch Diameter, Pd (mm)

Contact Angle, α (degrees)

BPFO — Outer Race

—

BPFI — Inner Race

—

BSF — Ball/Roller Spin

—

FTF — Cage (Train)

—

1× shaft = — Hz | BPFO/BPFI ratio: —

📏 ISO 10816 Severity & RPM↔Hz & GMF

Vibration Velocity (mm/s RMS)

Machine Class

A Good

B Acceptable

C Alert

D Danger

🔄 RPM ↔ Hz Converter

RPM

Frequency (Hz)

1× = 25.00 Hz | 2× = 50.00 Hz | 3× = 75.00 Hz | Period = 40.00 ms

⚙️ Gear Mesh Frequency (GMF)

Shaft RPM (driving gear)

Number of Teeth (driving gear)

Number of Teeth (driven gear) — optional, for gear ratio

Gear Mesh Frequency (GMF)

—

2× GMF

—

3× GMF

—

Gear Ratio

—

Driven Shaft Speed

—

RPM

Fault Identification at a Glance

Each mechanical fault produces a characteristic "fingerprint" in the vibration spectrum

⚖️

Unbalance

1×

Dominant peak at shaft speed. Radial. Amplitude ∝ speed². Stable phase.

🔧

Misalignment

2× (+ 1×, 3×)

Strong 2nd harmonic. High axial. 180° phase across coupling.

🔩

Looseness

1×,2×,3×…n×

"Forest" of harmonics. May show 0.5× subharmonics.

🔵

Bearing Defect

BPFO/BPFI/BSF

Non-synchronous peaks. Sidebands at 1×. Noise floor rise.

⚙️

Gear Fault

GMF ± 1×

Gear mesh frequency with sidebands at shaft speed.

📊 Complete Diagnostic Reference Table

Fault	Primary Frequency	Harmonics	Direction	Phase Behaviour	Key Distinguishing Feature
Static unbalance	1×	Low / none	Radial (H,V)	In-phase both bearings	Pure 1× sinusoid. Amplitude ∝ ω².
Dynamic unbalance	1×	Low / none	Radial (H,V)	~180° between bearings	1× dominant, bearings out of phase (couple).
Parallel misalignment	2× (≥ 1×)	1×, 3×	Radial	180° across coupling	2× often > 1×. High radial at coupling.
Angular misalignment	1×, 2×	3×	Axial dominant	180° across coupling (axial)	High axial. Axial ≥ 50% of radial.
Component looseness	1×,2×…10×+	Many (~10×)	Radial	Erratic	"Forest" of harmonics. Possible 0.5× sub.
Structural looseness	1× or 2×	Few above 2×	Vertical	Unstable	Strong vertical. Responds to bolt check.
Outer race (BPFO)	BPFO, 2×BPFO…	Multiple BPFO	Radial	N/A	Non-synchronous. No 1× sidebands.
Inner race (BPFI)	BPFI, 2×BPFI…	Multiple BPFI	Radial	Modulated at 1×	BPFI harmonics with ±1× sidebands.
Rolling element (BSF)	BSF, 2×BSF…	Multiple BSF	Radial	N/A	2×BSF often > 1×BSF. Non-synchronous.
Cage (FTF)	FTF ≈ 0.4×	2,3× FTF	Radial	N/A	Sub-synchronous (< 1×).
Gear mesh	GMF=N×1×	2,3× GMF	Radial+axial	Modulated at 1×	GMF with sidebands. N = teeth.
Electrical (motor)	2× line freq	—	Radial	Drops on power-off	100/120 Hz. Instant drop test.

Interactive FFT Spectrum Demonstration — 16 Fault Scenarios

Select a fault type to see characteristic time waveform and frequency spectrum. Compare patterns to identify the root cause.

Baseline

Unbalance

Misalignment

Looseness

Bearings

Gears

Other

Normal operating condition — low, stable vibration. Small 1× peak from residual unbalance. Clean spectrum.

Time Domain (Waveform)

Frequency Spectrum (FFT)

What is Vibration Analysis?

Quick Answer

Vibration analysis is the process of measuring and interpreting mechanical oscillations of rotating machinery to diagnose faults without disassembly. Using FFT (Fast Fourier Transform), the complex vibration signal is decomposed into individual frequency components. Each fault produces a characteristic spectral "fingerprint": unbalance at 1× RPM, misalignment at 2×, looseness as multiple harmonics, bearing defects at non-synchronous frequencies. The Balanset-1A performs both balancing and spectrum analysis in one portable instrument.

Every rotating machine vibrates. In a healthy machine, vibration is low and stable — its normal "operating signature." As defects develop, vibration changes in predictable ways. By measuring and analysing these changes, we can identify the root cause, predict failure, and schedule maintenance before catastrophic breakdown. This is the foundation of predictive maintenance.

FFT: The Core of Spectrum Analysis

A vibration sensor (accelerometer) converts mechanical oscillation into an electrical signal. Displayed over time, this is the waveform — a complex, seemingly chaotic curve when multiple faults are present. FFT (Fast Fourier Transform) decomposes this complex signal into individual sinusoidal components, each with its own frequency and amplitude.

Think of FFT as a prism splitting white light into a rainbow. The complex waveform is "white light" — FFT reveals the individual "colours" (frequencies) hidden inside. The result is the vibration spectrum — the primary diagnostic tool.

Rotational Frequency

f₁ₓ = RPM / 60 (Hz)

1× = shaft rotational frequency — the reference for all spectral analysis

Key Spectrum Parameters

Frequency (X-axis, Hz): How often oscillations occur. Directly linked to the source. 1× = shaft speed. 2× = twice shaft speed.
Amplitude (Y-axis, mm/s RMS): Vibration intensity at each frequency. Higher peaks = more energy = more serious condition.
Harmonics: Integer multiples of the fundamental: 2× (2nd), 3× (3rd), 4×, etc. Their presence and relative height carry diagnostic information.
Phase (°): Timing relationship at different measurement points. Essential for distinguishing unbalance (in-phase) from misalignment (180°).

Vibration Measurement Units: Displacement, Velocity, Acceleration

Vibration can be measured as three different physical parameters. Each emphasises different frequency ranges, making them suited to different diagnostic tasks. Understanding when to use which parameter is fundamental to effective analysis.

📏 Displacement

µm (peak-to-peak) or mil

Best range: 1–100 Hz

Measures how far the surface moves. Emphasises low frequencies — ideal for slow-speed machines, shaft orbit analysis, and proximity probes on journal bearings. 1 mil = 25.4 µm.

📈 Velocity

mm/s (RMS)

Best range: 10–1000 Hz

Measures how fast the surface moves. The standard parameter for general machinery monitoring per ISO 10816. Flat frequency response gives equal weight to most fault types. Balanset-1A measures in mm/s RMS.

💥 Acceleration

m/s² or g (RMS/peak)

Best range: 500 Hz – 20 kHz+

Measures the force of vibration. Emphasises high frequencies — ideal for early bearing defects, gear mesh, and impacts. 1 g = 9.81 m/s². Used for envelope/demodulation analysis.

When to Use Each Parameter

Parameter	Unit	Frequency Range	Best For	Standards
Displacement	µm pk-pk	1–100 Hz	Slow machines (< 600 RPM), shaft orbit, proximity probes, journal bearings	ISO 7919 (shaft vibration)
Velocity	mm/s RMS	10–1000 Hz	General machinery monitoring — unbalance, misalignment, looseness. Default parameter.	ISO 10816, ISO 20816
Acceleration	g or m/s² RMS	500 Hz – 20 kHz	Early bearing defects, gear mesh, impacts, high-speed machinery	ISO 15242 (bearing vibration)

Conversion at a Single Frequency

v = 2πf · d | a = 2πf · v = (2πf)² · d

d = displacement (m), v = velocity (m/s), a = acceleration (m/s²), f = frequency (Hz)

💡 Rule of Thumb

If you have only one sensor and one parameter to choose — choose velocity (mm/s RMS). It covers the broadest range of common faults with flat response. The Balanset-1A uses this as its native parameter. Add acceleration measurement only when you need to catch early-stage bearing or gear defects at high frequencies.

Measurement Technique with Balanset-1A

Sensor Placement

The quality of diagnosis depends entirely on measurement quality. Vibration forces are transmitted through bearings, so sensors must be mounted on bearing housings — as close to the bearing as possible, on the load-bearing structure (not covers or cooling fins).

Surface preparation: Clean, flat, free of paint flakes. Magnetic base must sit flush.
Radial horizontal (H): Perpendicular to shaft, horizontal plane. Often highest amplitude.
Radial vertical (V): Perpendicular to shaft, vertical plane.
Axial (A): Parallel to shaft. Critical for detecting misalignment.

💡 Two-Channel Diagnostic Trick

The Balanset-1A has 2 channels. For diagnostics, mount both sensors on the same bearing — one radial, one axial. This gives simultaneous radial + axial spectrums, enabling instant misalignment detection.

Balanset-1A Modes for Diagnostics

F1 — Spectrum Analyser: Full FFT display. The primary diagnostic mode.
F5 — Vibrometer: Quick assessment. Compare V1s (total RMS) vs. V1o (1×). If V1s ≈ V1o → unbalance. If V1s ≫ V1o → other faults.
F8 — Charts: Detailed spectrum + time waveform. Best for harmonic patterns and bearing frequencies.

⚠️ V1s vs. V1o — The First Diagnostic Check

Before balancing, compare V1s with V1o. If V1s ≫ V1o (e.g., 8 vs. 2 mm/s), most vibration is NOT from unbalance. Balancing won't solve it — examine the full spectrum.

Phase Analysis — The Diagnostic Differentiator

Frequency tells you what is vibrating; phase tells you how. Two faults can produce identical spectrums (both dominated by 1×) — only phase analysis distinguishes them. Phase is the angular relationship between vibration at different measurement points, measured in degrees (0°–360°).

🧭 Phase → Diagnosis Reference Table

Phase Relationship	Measurement Points	Diagnosis	Explanation
0° (in-phase)	Bearing 1 ↔ Bearing 2 (radial)	Static unbalance	Both bearings move together in sync — single heavy spot in centre of rotor. Single-plane correction.
~180° (anti-phase)	Bearing 1 ↔ Bearing 2 (radial)	Dynamic (couple) unbalance	Bearings rock in opposition — two heavy spots at different planes create a rocking couple. Two-plane correction needed.
~90°	Horizontal ↔ Vertical (same bearing)	Unbalance (any type)	Normal for unbalance — force vector rotates with shaft, producing ~90° between H and V at same point.
~180°	Across coupling (radial)	Parallel misalignment	Coupling forces push shafts apart in opposite radial directions. 180° across coupling with high 2× is the signature.
~180°	Across coupling (axial)	Angular misalignment	Shafts alternately push/pull axially. 180° axial across coupling with high 1× and 2× is definitive.
0°	Across coupling (axial)	Not misalignment	Both sides moving same axial direction — likely thermal growth, piping strain, or soft foot. Not angular misalignment.
Erratic / unstable	Any consistent points	Mechanical looseness	Phase readings jump randomly between measurements — characteristic of impacts in loose joints. Unstable phase = looseness.
Slowly drifting	Any point, over time	Resonance or thermal effects	Gradual phase shift during warmup suggests structural stiffness changing with temperature (thermal misalignment).
Consistent, non-0/180°	Bearing 1 ↔ Bearing 2	Combined static + couple unbalance	Phase between 0° and 180° indicates a mix of static and couple components — requires two-plane balancing.

💡 Phase Measurement with Balanset-1A

The Balanset-1A displays phase at 1× (the F1 value in vibrometer mode) using the tachometer as reference. To compare phase between two bearings, measure each bearing in the same direction (e.g., horizontal) with the tachometer on the same reference mark. The difference in phase readings reveals the fault type. No special software needed — just subtract the two readings.

Fault 1: Unbalance

Cause: Centre of mass displaced from rotation axis. Manufacturing tolerances, deposit buildup, erosion, broken blade, lost weight.

Spectrum: Dominant peak at exactly 1× RPM. Very low harmonics. Radial vibration. Amplitude increases with speed² (quadratic). Phase is stable and repeatable.

Static Unbalance (Single-Plane)

Pure 1× peak, sinusoidal waveform. Both bearings in-phase. Single-plane correction.

Static unbalance — dominant 1× at 25 Hz (1500 RPM). Minimal harmonics.

Dynamic Unbalance (Two-Plane / Couple)

Also 1× dominant, but bearings ~180° out of phase. Two-plane correction required.

Dynamic unbalance — 1× dominant. Spectrum similar to static but phase differs at bearings.

Action: Perform rotor balancing with the Balanset-1A. G-grade tolerance per ISO 1940-1.

Fault 2: Shaft Misalignment

Cause: Axes of coupled shafts do not coincide. Can be parallel (offset) or angular (tilted), usually both.

Parallel Misalignment (Radial)

High 1× and 2× in the radial direction. 2× often ≥ 1×. 180° phase shift across coupling.

Parallel misalignment — radial direction. Strong 1× and 2× with minor 3×.

Angular Misalignment — Radial

1× and 2× present in radial, but 2× typically dominates.

Angular misalignment — radial (R). 2× > 1×.

Angular Misalignment — Axial

Axial vibration ≥ 50% of radial. 180° phase across coupling in axial. This is the key distinguishing measurement.

Angular misalignment — axial (A). Very high 2× in axial direction.

Action: Balancing will NOT help. Stop the machine and perform shaft alignment. Re-check vibration after.

Fault 3: Mechanical Looseness

Cause: Loss of structural stiffness — loose bolts, cracks in foundation, worn bearing seats, excessive clearances.

Component Looseness

"Forest" of harmonics — 1×, 2×, 3×, 4×… up to 10×+ with decreasing amplitude. May show 0.5× subharmonics.

Component looseness — many harmonics 1× through 10×. Note 0.5× subharmonic.

Structural Looseness

1× and/or 2× dominant. Few higher harmonics. Strong vertical vibration.

Structural looseness — 1× and 2× dominate. Minimal higher harmonics.

Action: Inspect and tighten mounting bolts. Check foundation. Always check looseness before balancing.

Fault 4: Rolling Bearing Defects

Cause: Pitting, spalling, wear on raceways, rolling elements, or cage.

Bearing Defect Frequencies

BPFO = (n/2)(1 − Bd/Pd·cos α) · f_s
BPFI = (n/2)(1 + Bd/Pd·cos α) · f_s
BSF = (Pd/2Bd)(1 − (Bd/Pd·cos α)²) · f_s
FTF = ½(1 − Bd/Pd·cos α) · f_s

n = rolling elements | Bd = ball dia | Pd = pitch dia | α = contact angle | f_s = RPM/60

Outer Race Defect (BPFO)

Series of peaks at BPFO, 2×BPFO, 3×BPFO… No 1× sidebands (stationary ring). Most common bearing fault.

Outer race defect — BPFO harmonics at non-synchronous frequencies. No sidebands.

Inner Race Defect (BPFI)

BPFI harmonics with ±1× sidebands (rotating ring, load zone modulation). Sideband pattern is the key identifier.

Inner race defect — BPFI harmonics with ±1× sidebands (smaller peaks flanking main peaks).

Rolling Element Defect (BSF)

BSF harmonics. 2×BSF often dominant. Non-synchronous. Often accompanied by race damage.

Rolling element defect — BSF harmonics. Note 2×BSF is highest (two-element damage).

Cage Defect (FTF)

Sub-synchronous peaks (FTF ≈ 0.4× shaft speed). Low frequency. Often accompanies other bearing damage.

Cage defect — FTF and harmonics below 1× shaft speed (sub-synchronous).

Bearing Defect Progression (4 Stages)

Stage 1 — Subsurface: Ultrasonic zone (> 5 kHz). Not visible on standard FFT. Detectable by spike energy / enveloping.

Stage 2 — Early defect: Bearing frequencies appear (BPFO, BPFI). Low amplitude. This is where Balanset-1A begins detection.

Stage 3 — Progressed: Multiple harmonics. Sidebands develop. Noise floor rises.

Stage 4 — Advanced: Broadband noise. Bearing frequencies may disappear into noise. Replacement urgent.

Envelope (Demodulation) Analysis — Early Bearing Detection

Standard FFT spectrum analysis detects bearing defects from Stage 2 onward. But in Stage 1, bearing impacts are too weak to appear above the noise floor. Envelope analysis (also called demodulation or high-frequency detection, HFD) extends detection to much earlier stages.

How It Works

When a rolling element hits a defect, it generates a short impact pulse that excites high-frequency structural resonances (typically 5–20 kHz). These resonances "ring" briefly at each impact. Envelope analysis works in three steps:

Band-pass filter: Isolate the high-frequency resonance band (e.g., 5–15 kHz) where the impacts ring.
Rectify and envelope: Extract the amplitude modulation pattern — the "envelope" that follows the peaks of the ringing.
FFT of the envelope: Apply FFT to the envelope signal. The result shows the repetition rate of impacts — which equals the bearing defect frequencies (BPFO, BPFI, BSF, FTF).

Why Envelope Detects Earlier

In the raw spectrum, a weak impact at BPFO might produce 0.1 mm/s — invisible among machine noise of 2 mm/s. But that same impact excites a resonance at 8 kHz where there is no other vibration source. After demodulation, the BPFO repetition pattern emerges clearly from a clean background.

Related Parameters

Spike Energy (SE): Overall measurement of high-frequency impact energy. Scalar trending value. Good for "go/no-go" screening.
gSE / HFD / PeakVue: Vendor-specific names for envelope-derived parameters. All based on the same principle.
Acceleration enveloping: The Balanset-1A measures in velocity (mm/s). For full envelope analysis, a dedicated analyser with acceleration input and band-pass filtering capability is ideal. However, the Balanset-1A's FFT can still detect Stage 2+ bearing defects effectively in the standard velocity spectrum.

Envelope spectrum of inner race defect — BPFI harmonics emerge clearly from demodulated high-frequency signal. Compare with raw velocity spectrum where these may be hidden in noise.

Action: Check lubrication. Plan bearing replacement. Increase monitoring frequency.

Fault 5: Gear Defects

Cause: Worn, pitted, or broken teeth. Gear eccentricity. GMF = number of teeth × shaft RPM / 60.

Gear Eccentricity

GMF with sidebands at ±1× shaft speed. Gear's 1× may also be elevated.

Gear eccentricity — GMF at 500 Hz with ±1× sidebands. Elevated 1×.

Gear Tooth Wear / Damage

Multiple GMF harmonics with dense sidebands. Severity tracks with sideband count and amplitude.

Gear wear — GMF and 2×GMF with multiple sidebands at 1× intervals.

Action: Check gearbox oil for metallic particles. Schedule inspection. Monitor GMF sideband trend.

Electrical Faults (Motors)

Electromagnetic faults produce vibration at 2× line frequency (100 Hz on 50 Hz grids, 120 Hz on 60 Hz). Critical test: vibration disappears instantly when power is cut. Mechanical faults decay gradually.

Stator eccentricity: 2× line frequency, steady amplitude.
Rotor bar defects: Sidebands around line frequency at slip frequency intervals.
Soft foot: Vibration changes when individual motor feet are loosened.

Fault 7: Belt Drive Problems

Cause: Worn, misaligned, or improperly tensioned belts. Belt drives generate vibration at the belt pass frequency, which is typically a sub-synchronous frequency (below 1× shaft speed) since the belt is longer than the pulley circumference.

Belt Frequency

f_belt = (π · D · RPM) / (60 · L)

D = pulley diameter (m) | L = belt length (m) | RPM = pulley speed
Simplified: f_belt = pulley circumference speed / belt length

Common Belt Signatures

Belt wear / defect: Peaks at belt frequency (f_belt) and its harmonics (2×, 3×, 4× f_belt). These appear below 1× shaft speed — sub-synchronous peaks are the key indicator.
Belt misalignment: Elevated axial vibration at 1× and 2× shaft speed. Similar to shaft misalignment but restricted to the belt-driven machine.
Improper tension: High 1× vibration that changes dramatically with belt tension adjustment. Overtight belts increase bearing load; loose belts cause slapping and belt-frequency peaks.
Resonance: Belt natural frequency (belt "flutter") can be excited if belt span resonance coincides with operating speed. Visible as broad peak at belt natural frequency.

Belt drive defect — sub-synchronous peaks at belt frequency and harmonics (below 1× shaft speed at 25 Hz).

Action: Check belt condition, tension, and pulley alignment. Replace worn belts. For recurring issues, verify pulley alignment with a laser tool or straight-edge.

Fault 8: Pump Cavitation

Cause: Vapor bubbles form and collapse violently when local pressure drops below the liquid's vapor pressure — typically at the pump suction. Each bubble collapse creates a micro-impact. Thousands of collapses per second generate a characteristic broadband noise.

Spectral Signature

Broadband high-frequency energy: Unlike mechanical faults (which produce discrete peaks), cavitation generates a raised noise floor across a wide frequency range, typically above 2–5 kHz. The spectrum looks like a "hump" or elevated plateau rather than sharp peaks.
Random, non-periodic: No harmonics, no relationship to shaft speed. The noise sounds like "gravel" or "crackling" — audible even without instruments.
Low-frequency effects: Severe cavitation may also cause instability at 1× and broadband low-frequency noise from flow turbulence.

Pump cavitation — broadband high-frequency noise (raised floor above 200 Hz). No discrete peaks — contrast with bearing defects which show specific frequencies.

Action: Increase suction pressure (lower pump, open suction valve, reduce suction pipe losses). Check NPSH_available vs. NPSH_required. Reduce pump speed if possible. Cavitation causes rapid erosion damage — do not ignore.

Fault 9: Oil Whirl & Oil Whip (Journal Bearings)

Cause: Fluid-film instability in journal (sleeve) bearings. The oil film wedge forces the shaft to orbit within the bearing clearance at a sub-synchronous frequency. This is distinct from rolling element bearing defects and occurs only in plain/journal bearings.

Oil Whirl

Frequency: Approximately 0.42× to 0.48× shaft speed (often cited as ~0.43×). This is a sub-synchronous peak that tracks shaft speed — if RPM increases, the whirl frequency increases proportionally.
Spectrum: A single peak at ~0.43× that shifts with speed. Amplitude may be moderate.
Condition: Precursor to oil whip. Usually not immediately destructive but indicates instability.

Oil Whip

Frequency: Locks onto the rotor's first natural frequency (critical speed). Unlike whirl, it does NOT track shaft speed — the frequency remains constant as RPM changes.
Spectrum: Large sub-synchronous peak at the rotor's first critical speed. Amplitude can be very high — destructive.
Condition: Dangerous. Immediate action required. Can lead to bearing wipe-out and shaft damage.

Oil whirl — sub-synchronous peak at ~0.43× shaft speed (≈ 10.7 Hz for 1500 RPM). Distinct from 0.5× looseness.

⚠️ Oil Whirl vs. Looseness — How to Distinguish

Both produce sub-synchronous peaks, but: Oil whirl is at ~0.43× (not exactly 0.5×) and tracks with speed. Looseness produces peaks at exactly 0.5×, 1.5×, 2.5× and does not track with speed (stays at fixed fractions of 1×). Oil whirl only occurs in journal/sleeve bearings — if the machine has rolling element bearings, it cannot be oil whirl.

Action: For oil whirl: check bearing clearance, oil viscosity, and load. Increase bearing loading or change oil viscosity. For oil whip: reduce speed immediately below the critical threshold. Consult a rotor dynamics specialist.

ISO 10816 Vibration Severity — Complete Classification Table

ISO 10816 (superseded by ISO 20816 but still widely referenced) defines vibration severity zones for four machine classes. Vibration is measured as velocity in mm/s RMS on bearing housings. The table below shows all zone boundaries for all four classes — use it as a quick reference when evaluating measurements.

📋 ISO 10816-3 Vibration Severity Zones — All Machine Classes (mm/s RMS)

Machine Class	Zone A Good	Zone B Acceptable	Zone C Alert	Zone D Danger
Class I Small machines ≤ 15 kW (pumps, fans, compressors)	≤ 0.71	0.71 – 1.8	1.8 – 4.5	> 4.5
Class II Medium machines 15–75 kW (without special foundation)	≤ 1.8	1.8 – 4.5	4.5 – 11.2	> 11.2
Class III Large machines > 75 kW (rigid foundation)	≤ 2.8	2.8 – 7.1	7.1 – 18	> 18
Class IV Large machines > 75 kW (flexible foundation, e.g. steel frame)	≤ 4.5	4.5 – 11.2	11.2 – 28	> 28

📌 How to Use This Table

Step 1: Determine your machine class by power and foundation type.
Step 2: Measure overall vibration velocity (mm/s RMS) on each bearing housing in radial direction.
Step 3: Find the zone. Zone A = newly commissioned or excellent. Zone B = unrestricted long-term operation. Zone C = acceptable only for limited periods — schedule maintenance. Zone D = damage is occurring — stop machine as soon as possible.

Remember: trends matter more than absolute values. A machine running at 3.0 mm/s (Zone B for Class II) that was previously at 1.5 mm/s has doubled — investigate the cause even though it's still "acceptable." The Balanset-1A's vibrometer mode (F5) displays overall velocity V1s for instant zone assessment.

⚠️ ISO 10816 vs. ISO 20816

ISO 10816 was formally superseded by ISO 20816 (published 2016–2022). The zone boundaries remain similar for most machine types, but ISO 20816 adds evaluation criteria for displacement and expands machine-specific parts. In practice, ISO 10816 values remain the industry-standard reference. Both the Balanset-1A and most industrial vibration programs still use ISO 10816 zones.

From Measurement to Monitoring

Trend Analysis

A single spectrum is a snapshot. The power of vibration analysis is trend analysis — tracking changes over time.

Create a baseline: Measure new or known-good equipment. Save spectrums.
Establish intervals: Critical: weekly. Standard: monthly. Auxiliary: quarterly.
Ensure repeatability: Same points, same directions, same operating conditions.
Track changes: A 2× increase from baseline is significant even if in ISO Zone A.

Decision Algorithm

Get a quality spectrum (F8 Charts, radial + axial).
Identify the highest peak — this is the dominant problem.
Match to fault type:
- 1× dominates → Unbalance → Balance with Balanset-1A.
- 2× dominates + high axial → Misalignment → Realign shafts.
- Many harmonics → Looseness → Inspect and tighten.
- Non-synchronous peaks → Bearing → Plan replacement.
- GMF + sidebands → Gear → Check oil, inspect gearbox.
Fix the dominant fault first — secondary symptoms often disappear.

← Back to Glossary Index

Frequently Asked Questions — Vibration Analysis

▸ What is vibration analysis?

The process of measuring and interpreting mechanical oscillations to diagnose machinery faults without disassembly. FFT decomposes complex vibration into frequency components. Each fault produces a characteristic spectral "fingerprint" — unbalance at 1× RPM, misalignment at 2×, looseness as multiple harmonics, bearing defects at non-synchronous frequencies.

▸ How do I tell unbalance from misalignment?

Unbalance: dominant 1× peak, radial, stable phase, amplitude ∝ speed². Misalignment: significant 2× (often ≥ 1×), high axial vibration, 180° phase across coupling.

▸ What are bearing defect frequencies?

Frequencies generated when rolling elements pass over defects. BPFO = outer race, BPFI = inner race, BSF = ball/roller, FTF = cage. They are NOT integer multiples of shaft speed. Use the bearing frequency calculator above.

▸ What is a good vibration level?

ISO 10816 zones for Class II (15–75 kW): A < 1.8, B < 4.5, C < 11.2, D > 11.2 mm/s RMS. Trends matter more — a 2× increase from baseline is significant even in Zone A.

▸ Can Balanset-1A do vibration analysis?

Yes. 2-channel FFT spectrum analyser (F1), vibrometer (F5), detailed charts (F8). Mount both sensors radial+axial on one bearing for instant misalignment detection.

▸ Time waveform vs. FFT spectrum?

Waveform shows amplitude vs. time — complex when multiple faults overlap. FFT decomposes into individual frequencies (like a prism splitting light). Each spectrum peak = one vibration source.

▸ How often should I measure vibration?

Critical: weekly. Standard: monthly. Auxiliary: quarterly. After maintenance: immediately. Consistency is key — same points, directions, conditions.

▸ What causes 0.5× (subharmonic) vibration?

Severe mechanical looseness, oil whirl in journal bearings, or cracked shaft. Half-order peaks (0.5×, 1.5×) arise from impacts every other revolution. Serious condition — prompt attention required.

Diagnose First — Then Balance

The Balanset-1A is both a 2-channel vibration analyser and a precision field balancer. Identify the fault by spectrum, then fix it — all with one instrument.

Browse Equipment →

Vibration Analysis of Machinery – Diagnostic Spectral Signatures of Common Faults

Published by Nikolai Shelkovenko on June 11, 2025 February 7, 2026