“It just sounds wrong” — the maintenance technician’s intuition is usually correct. Reading the noise correctly turns an emergency stop into a planned intervention with weeks of warning.
Worm gearbox NVH (noise, vibration, harshness) is three separate dimensions that can each indicate different root causes. Noise is what you hear (40 to 80 dB at 1 metre, dominated by gear mesh frequency). Vibration is what you measure (0.5 to 4.5 mm/s rms in healthy units, rising as wear progresses). Harshness is the felt character of the operation — smooth at one extreme, rough or pulsating at the other. Worm gear NVH is genuinely harder to diagnose than helical or spur because sliding contact suppresses the crisp sideband signatures that make spur/helical fault detection straightforward. Reading the gear mesh frequency, its harmonics, the worm shaft sideband pattern, and the time-domain vibration character together gives a working diagnosis even when no single signal is conclusive. Most reduction work is done at design time through lead angle, contact pattern, and lubricant choice; field retrofits rely on damping and isolation when the source cannot be replaced.
“It just sounds wrong.” The maintenance technician’s observation is usually correct, but it tells you nothing about whether the cause is a worn output bearing, a chipped tooth, a misaligned shaft, or simply a normal break-in transient that will settle in a few hundred hours. Worm gearbox noise is a real symptom that demands real diagnosis, not a quick reassurance from the supplier that “all worm gears make some noise.” Reading the noise correctly turns an emergency stop into a planned intervention with weeks of warning before failure.
NVH is shorthand for three distinct dimensions that need to be measured and interpreted separately. Noise is the airborne sound emitted by the gearbox housing — what an operator hears standing nearby. Vibration is the mechanical motion of the housing surface — what an accelerometer measures in millimetres per second. Harshness is the felt character of the operation — smooth, pulsating, gritty, or rattly. The three dimensions correlate but do not match: a quiet gearbox can vibrate strongly, a rough-feeling drive can register low vibration values, and noise level alone does not predict bearing life. Diagnostic teams that conflate the three dimensions miss developing failures that the right interpretation would catch.
Every worm gearbox running at steady state emits a dominant tone at the gear mesh frequency: the worm input rotation speed multiplied by the number of worm thread starts. A single-start worm at 1,450 rpm input produces gear mesh at 24.2 Hz. A two-start worm at the same input runs gear mesh at 48.3 Hz. The same calculation applied at the wheel side gives the same result, since one worm rotation engages one wheel tooth.
Healthy gearboxes show clean gear mesh tones at the fundamental frequency and small second-harmonic and third-harmonic peaks. Developing problems show up as rising harmonics, sidebands at shaft rotation frequency, or new peaks at unrelated frequencies.
| Frequency feature | Calculation | What it indicates | Healthy / problem |
|---|---|---|---|
| Worm shaft 1× rpm | Input rpm / 60 | Worm shaft balance, runout | Low / rising = unbalance |
| Wheel shaft 1× rpm | Output rpm / 60 | Wheel runout, hub eccentricity | Low / rising = wheel eccentricity |
| Gear mesh fundamental | Worm rpm × thread starts / 60 | Tooth profile, transmission error | Always present / amplitude grows = wear |
| 2× and 3× gear mesh | 2 × GMF, 3 × GMF | Tooth deflection, mesh stiffness | Small / dominant = tooth issues |
| Sidebands at GMF ± wheel rpm | GMF + 1× wheel rpm spacing | Wheel tooth fault localised | Absent / present = pitting or chip |
| Bearing fault frequencies | Manufacturer-specific BPFO/BPFI | Bearing race or roller defect | Absent / appearing = bearing wear |
Worm gearbox vibration analysis is genuinely harder than spur or helical analysis, and pretending otherwise sets up false confidence. The sliding contact between worm and wheel suppresses the crisp, periodic impact signatures that make sideband detection clean on rolling-contact gear pairs. A worm gearbox with a moderately pitted wheel often shows a vibration spectrum that looks nearly identical to a healthy unit — only an experienced analyst comparing against the unit’s own baseline catches the small changes. The practical implication for maintenance teams: invest in baseline measurements when the gearbox is new and healthy, then track changes against that specific unit’s signature rather than against generic industry norms. The baseline costs a one-hour technician visit. Reading every later measurement against it doubles the diagnostic resolution at zero ongoing cost.
ISO 10816 and ISO 20816 establish vibration severity zones for industrial machinery using overall rms velocity in millimetres per second. The zones translate measured vibration into action: continue running, monitor more frequently, plan maintenance, shut down for repair. The thresholds below apply to industrial worm gearboxes mounted on rigid foundations, measured at the housing surface near the output bearing.
| Zone | Vibration rms (mm/s) | Condition | Action |
|---|---|---|---|
| A (Good) | ≤ 1.8 | New or recently overhauled | Routine quarterly check |
| B (Acceptable) | 1.8 to 4.5 | Normal long-term operation | Monthly check, trend monitoring |
| C (Tolerable) | 4.5 to 11.2 | Wear progressing, marginal | Plan repair within 3 months |
| D (Unacceptable) | > 11.2 | Damage advancing rapidly | Shut down for repair within days |
A worm gearbox baseline measurement on a new install typically falls in zone A or low zone B. Drift over service life into upper zone B and zone C is normal as the bronze wheel breaks in and bearings settle into wear. Sudden jumps of 50 percent or more between two consecutive measurements are usually more diagnostic than the absolute reading — they indicate a developing fault, regardless of whether the absolute value is still in the “tolerable” range.
Worm gear frequency-domain analysis catches problems that have settled into steady patterns. Time-domain character catches problems that show as transient events — chipped teeth, intermittent contact, occasional rattles.
A trained ear on a long screwdriver pressed against the worm gearbox housing detects these events with surprising sensitivity, and a smartphone vibration recording app captures them well enough for later analysis.
Smooth steady whine: healthy operation. Tone changes gradually with load and speed. No transients.
Periodic click or knock at wheel rotation rate: chipped or pitted single tooth. The sound recurs once per wheel revolution; calculate the wheel rpm and confirm the period matches.
Rough or gritty character with no distinct period: general wear progressing across many teeth, often combined with bearing wear. Less diagnostic but indicates the gearbox is past mid-life.
Modulating drone that rises and falls slowly: shaft misalignment or coupling problem rather than a gear fault. Check input shaft alignment and coupling condition before disassembling the gearbox.
Rattle that disappears under load: backlash growing past acceptable limits, or worn coupling allowing free play. Increases at light load and stops when load engages the teeth firmly.
Worm gear NVH is largely set at design time, with limited field-side options once the unit is installed. The biggest design levers are tooth-profile accuracy, contact pattern quality, lead angle, and lubricant choice. Each of these can shift the dB level by 3 to 8 dB independently, which compounds to 10 to 20 dB total improvement when applied together.
The tradeoff is cost. Ground worms (DIN 5 to DIN 6) cost 30 to 60 percent more than hobbed-only (DIN 7 to DIN 8) but produce 5 to 8 dB lower mesh-frequency noise.
Tooth profile accuracy. The dominant noise source is transmission error — the deviation of actual rotation angle from theoretical at the meshing point. Ground tooth profiles reduce transmission error roughly five-fold compared to hobbed-only profiles, lowering gear mesh excitation accordingly.
Contact pattern. A correctly assembled pair shows 60 to 80 percent contact area centred on the wheel teeth. Off-centre or undersized contact concentrates load on edges, which generates higher dynamic forces and louder operation. Verify with bluing test at first installation and at every overhaul.
Lead angle. Higher lead angles (multi-start worms) produce more rolling and less sliding contact. The result is lower friction-induced acoustic emission. The trade-off is that high-lead-angle worms cannot self-lock, so the noise advantage is only available for non-self-locking applications.
Lubricant. PAG synthetic oil reduces sliding friction by roughly 15 percent compared to mineral oil, which lowers noise generation by 2 to 4 dB at the gear mesh frequency. Browse ウォームギア減速機 options that come standard with PAG synthetic fill for noise-sensitive applications.
Once a worm gearbox is installed, the worm gear noise source itself usually cannot be changed without replacement. Field-side reduction relies on three approaches: damping the worm gearbox housing radiation, isolating the structural transmission path, and absorbing airborne noise locally.
Housing damping. Constrained-layer damping treatment applied to the worm gearbox housing exterior reduces sound radiation by 3 to 6 dB across most of the spectrum. Effective on cast iron housings; less effective on aluminium housings (which are already lower-radiating).
Vibration isolation. Replacing rigid mounting feet with elastomer or wire-rope isolators decouples the worm gearbox from the supporting frame. Reduces structure-borne noise transmitted to the building or machine frame by 6 to 15 dB. Effective when the noise reaches the operator through the structure rather than directly through the air.
Acoustic enclosure. A box-type enclosure with sound-absorbing lining around the worm gearbox reduces emitted noise by 10 to 20 dB. Adds cost, requires ventilation provisions for thermal management, and complicates maintenance access. Reserved for applications where dB targets cannot be met by other means.
A Korean food packaging OEM specified a maximum 65 dB at 1 metre from each conveyor drive in a new production line. Initial measurement on the standard 30:1 hobbed-only worm gearbox: 72 dB at 1 metre, 7 dB above target. Diagnosis: gear mesh frequency dominated the spectrum at 71 Hz with strong second harmonic. The customer evaluated three options. Option A: replace with DIN 6 ground worm and PAG synthetic oil — predicted 64 to 65 dB, 35 percent unit cost premium. Option B: keep the standard unit and add an acoustic enclosure — predicted 60 dB, 22 percent total cost premium plus complicated maintenance access. Option C: switch to a helical gearbox — predicted 60 dB, but right-angle layout incompatible with conveyor frame. Decision: option A, ground worm with PAG. Final measurement after installation: 64 dB. Acceptance achieved with margin and no maintenance complications.
A Japanese pharmaceutical OEM mounted a 1.5 kW vertical-mount worm gearbox above a cleanroom inspection station. Measured floor-level vibration at the inspection table: 0.08 mm/s rms — 40 percent above the 0.06 mm/s sensitivity threshold for the optical inspection equipment. Diagnosis: the worm gearbox itself was running quietly (1.4 mm/s housing rms, zone A), but structure-borne transmission through rigid mounting bolts to the cleanroom ceiling propagated to the inspection station. Solution: install wire-rope isolators between the gearbox foot and the mounting bracket. Cost: roughly 280 USD. Post-retrofit floor vibration: 0.03 mm/s, well within tolerance. Lesson: NVH problems often live in the structural path, not in the gearbox itself. Diagnose the path before replacing the source.
A Vietnamese textile factory had 40 looms on a single floor each driven by a 0.75 kW worm gearbox. Cumulative noise at the workstations exceeded 88 dB, above the 85 dB occupational limit. Per-loom noise from each gearbox: 78 dB at 1 metre — high but not extreme for the unit class. Diagnosis: 40 sources at 78 dB combine logarithmically to 94 dB ambient. Replacing all 40 with ground-worm units would cost over 35,000 USD — unaffordable. Alternative: apply constrained-layer damping treatment to all 40 housings. Per-unit material and labour cost: 18 USD. Per-unit noise reduction: 4 dB. Cumulative ambient noise reduction: 4 dB to 90 dB. Combined with PPE programme (mandatory ear protection), workplace exposure brought below the 85 dB action threshold. Total intervention cost: under 800 USD across the factory. Lesson: when 40 sources dominate, treating all 40 cheaply often beats replacing one or two expensively.
In general, yes — by 3 to 8 dB at comparable power and speed. The continuous sliding contact between worm and wheel produces less impulsive excitation than the rolling-and-sliding mesh of involute helical gears. The difference is most noticeable on smaller drives in quiet environments. On large industrial drives in noisy plants, the gear-type difference is often masked by other sources (motor, bearings, structure). The “quiet worm” advantage is real but modest, and not a reason on its own to choose worm over helical when other factors (efficiency, ratio, layout) favour helical.
A 5 kW industrial worm gearbox at 1,450 rpm input typically emits 65 to 75 dB at 1 metre under nominal load. Higher precision (DIN 5 ground) and synthetic lubrication can drop this to 60 to 65 dB. Lower-quality hobbed-only construction with mineral oil can run 75 to 82 dB. As a rough rule, doubling the kW rating adds about 3 dB to the noise level at the same accuracy class. Below 60 dB at 1 metre is achievable but typically requires either ground worms with PAG synthetic plus damping treatment, or housing within an acoustic enclosure.
Worm gearboxes typically have a break-in period of 50 to 200 operating hours during which the bronze wheel teeth wear in to match the steel worm contact pattern. Noise is usually 2 to 4 dB louder during break-in than at steady state. If noise stays elevated beyond 200 hours, or rises during break-in instead of falling, the contact pattern was likely incorrect at assembly — the bluing test should confirm this. Persistent above-normal noise after break-in usually indicates incorrect centre distance, off-centre contact pattern, or insufficient preload on output bearings, all of which require disassembly to correct.
Three instruments cover most worm gearbox diagnostic needs. A handheld worm gearbox vibration meter (accelerometer probe with rms readout, around 200 to 500 USD) measures overall vibration severity in mm/s rms — covers ISO 10816 zone classification. A sound level meter (around 100 to 300 USD for industrial models) reads dB at standardised distance. A smartphone with a vibration analysis app (free to 30 USD) captures time-domain recordings for later FFT analysis on a laptop. Together these three instruments support trend monitoring and basic fault diagnosis without the expense of a vibration analyser system. For deeper diagnosis, a Bruel and Kjaer or similar professional analyser system (5,000 USD plus) adds spectral kurtosis, envelope analysis, and cepstral analysis but is rarely needed for routine maintenance.
Two mechanisms produce worm gearbox load-dependent pitch changes. First, varying mesh stiffness as load deflects the gear teeth — under heavier load, the teeth deflect more, the effective mesh stiffness drops slightly, and the natural frequency of the mesh shifts. The pitch change is usually 1 to 3 percent, perceptible but not large. Second, oil-film thickness changes with load — heavier loads thin the elastohydrodynamic film, increase friction-induced acoustic emission, and add high-frequency content. Pitch changes that exceed 5 percent or that produce harsh tones at heavy load indicate developing tooth deflection problems and warrant inspection.
Steady whine at the gear mesh frequency is normal and expected on every operating worm gearbox. The frequency is determined by the worm thread starts and input rpm, and the whine is the audible signature of meshing teeth — present even on perfectly built drives. What matters is the amplitude (how loud) and the stability (whether it grows over time). A whine that holds steady across thousands of operating hours is healthy. A whine that grows by 3 dB or more between quarterly measurements indicates progressing wear and should be investigated. Pitch wandering, harmonics dominating the fundamental, or new sideband peaks appearing also warrant inspection.
Backlash that has grown beyond initial spec produces rattle at light load conditions because the teeth lose contact briefly during torque reversals. Each engagement-disengagement event adds an impulsive component to the vibration spectrum. Ratio interacts with NVH through the gear mesh frequency: higher ratios with single-start worms produce lower mesh frequencies, which are easier to damp but harder to isolate (longer wavelengths penetrate isolators). Lower ratios with multi-start worms produce higher mesh frequencies, easier to isolate but harder to damp. For NVH-sensitive applications, choose the ratio with both efficiency and acoustic considerations in mind, not efficiency alone.
Worm gearbox NVH is three independent dimensions — what you hear, what you measure, what you feel — and treating worm gear NVH as one number obscures most of the diagnostic signal. Frequency analysis at gear mesh fundamental, harmonics, and sidebands tells you what is starting to fail; severity zones tell you how urgently to act; time-domain character separates transient events from steady wear; and the structural path determines whether the noise reaches the operator at all. Reduction work is overwhelmingly more effective at design time than at field retrofit, but useful retrofit options exist when the design choice cannot be revisited. The diagnostic discipline pays back faster than most maintenance teams expect — a baseline measurement on every new install costs an hour and saves weeks of unplanned downtime down the line.
For Korean and Japanese OEM design teams developing acoustic-sensitive packaging lines, cleanroom equipment, or precision factories, our engineering desk reviews application dB targets against gear-pair selection, lubricant choice, and mounting strategy. Standard catalogue ground precision worm gear sets deliver lower transmission error and quieter operation than hobbed-only equivalents at typical 30 to 60 percent cost premium. Custom NVH-optimised configurations are available on 5 to 8 week lead times — request an NVH-targeted specification review with your dB target and operating conditions and our team will return options within one Korean working day.
Send the gearbox kW and ratio, target dB at one metre, and operating conditions. We will recommend gear pair accuracy class, lubricant, mounting strategy, and any retrofit options that fit the budget — typically within one Korean working day for standard catalogue specifications.
編集者: Cxm
Worm and Worm Wheel Pair Matching — Why Mix and Match Fails A worm and…
Worm Gear Strength Calculation — DIN 3996, ISO 14521, AGMA 6034 From application torque to…
Worm Gear Surface Finish — Why Smoothness Decides Service Life Run a fingernail across the…
Worm Gear Contact Pattern — How Bluing Tests Reveal Quality A 60 to 80 percent…
Worm Gear Module — Choosing the Right Tooth Size for Torque What module do I…
Worm Gear Center Distance — How to Calculate and Standardise One millimetre of centre distance…