An inclination sensor "drifts" on a vibrating machine when the vibration is coupled into the measurement instead of being rejected: a mounting bracket that resonates or flexes converts machine vibration into real tilt at the sensor; vibration components alias into the slow tilt reading when sampling and filtering do not separate them; the sensitive axis is misaligned so machine vibration projects onto it; temperature shifts add genuine drift; and sometimes the reading is honest — the structure actually tilts under load. The first safe check is the static-versus-dynamic comparison: read the sensor machine-stopped versus machine-running at the same nominal attitude, and on a rigid reference surface if possible. Never mount, remount or wiring-work a sensor on a running machine — stopped-and-locked-out work only.
Key Takeaways
- Vibration must be rejected, not averaged away. Averaging a vibration-corrupted reading reduces noise but biases the tilt value; the fix is coupling (mounting) and filtering designed for the vibration spectrum — with sampling that avoids aliasing.
- The mounting is part of the measurement. A bracket that flexes or resonates makes the sensor measure its own motion; stiff, short, direct-to-structure mounting is the baseline, and the sensor's own filtering (e.g., the fifth-order filtering documented in KJT Sensors' inclination products) is the second line.
- Axis alignment decides what the sensor sees. A single-axis sensor mounted a few degrees off its intended plane projects machine vibration onto the sensitive axis — apparent drift that vanishes with correct alignment.
- Static-vs-dynamic is the split. Same attitude, machine stopped vs. running: identical readings → drift is thermal or electronic; different readings → vibration coupling or real structural tilt under load.
- Don't recalibrate to hide a coupling fault. Zeroing a vibration-biased reading corrupts the static accuracy; fix the coupling, then zero if the model provides it and the zero state is real.
Symptom / Likely Cause / Safe Check Table
| Symptom | Likely cause | Safe check |
|---|---|---|
| Reading wanders only while the machine runs | Vibration coupling: bracket flex/resonance, axis projection, aliasing | Static-vs-dynamic comparison at same attitude (see worksheet) |
| Reading differs stopped vs. running by a consistent amount | Structural tilt under machine load (the reading is honest) | Compare with an independent reference or measure the structure's loaded deflection by design data |
| Slow wander over hours/days regardless of machine state | Temperature drift; electronic offset drift | Correlate with temperature logs; check model's temperature specification |
| Apparent drift after remounting | Axis misalignment; different mounting stiffness | Re-verify axis alignment marks and mounting surface; rerun static reference |
| Reading jumps at certain machine speeds | Mounting resonance excited at specific frequencies | Correlate jump speeds with machine speed; inspect bracket stiffness |
| Zero offset after installation | Mounting stress or residual zero offset within specification | Zero check on a verified level reference; apply model's zero procedure if provided |
| Noisy output at the controller, clean locally (or on serial readout) | Signal-chain noise, not the sensor | Parallel comparison local vs. controller (https://www.kjt-sensors.com/show-27188.html) |
Safety Boundary
Safe observations: static-vs-dynamic reading comparisons from the controller or local display without touching the machine; correlation with temperature and machine-state logs; axis/mounting inspection with the machine stopped.
Qualified-person / procedural work: any mounting, remounting or wiring change (machine stopped and locked out); zero/calibration procedures per the model manual; controller configuration. Work on mobile machinery adds the machine's own lockout discipline.
Inputs to Collect Before Changing Settings
- Sensor model, axes (single/dual), output (analog/serial/CAN), and its documented filtering and temperature specifications;
- Machine vibration profile: dominant frequencies (from machine speed and known sources) and direction relative to the sensitive axis;
- Mounting description: bracket geometry, material, fastening, distance from structure; axis alignment method used;
- The static-vs-dynamic worksheet results (below);
- Temperature environment and its cycles;
- Reading history: drift pattern vs. machine state and temperature;
- Required accuracy and response in the application — what the control actually needs from the tilt value.
The Static-Versus-Dynamic Diagnostic Worksheet
Four measurements, same nominal machine attitude:
- Static baseline — machine stopped, locked out, settled: record the reading and the ambient temperature.
- Rigid reference — if accessible per procedure, the sensor (or a same-model reference) on a known-rigid surface nearby, machine running: records whether vibration alone corrupts the measurement chain.
- Dynamic on-machine — machine running at operating speed: record the reading, its noise band and the mean.
- Post-stop return — machine stopped again: does the reading return to the static baseline?
Reading the result: static ≈ dynamic ≈ return → the "drift" is thermal/electronic or expectation error; static ≠ dynamic but return ≈ static → vibration coupling (mounting/axis/aliasing); static ≠ dynamic and return ≠ static → real structural change or thermal drift — investigate the structure and temperature.
Diagnostic Sequence (Least Invasive to Specialized)
- Observation: reading wanders while the machine runs. Hypothesis: vibration coupling. Check: complete the worksheet. Expected result: static and return agree; dynamic differs or noisier. Next action: proceed to steps 2–4 in order.
- Observation: dynamic deviation or noise confirmed. Hypothesis: mounting flex/resonance. Check: inspect bracket geometry and fastening (machine stopped); correlate any speed-specific jumps with machine speed. Expected result: long/flexible bracket, loose fasteners, or jump speeds matching structural resonance. Next action: stiffen/shorten the mounting path; re-run the worksheet.
- Observation: coupling persists after mounting fixes. Hypothesis: axis projection — vibration direction has a component on the sensitive axis. Check: verify axis alignment marks against the intended measurement plane; review vibration direction relative to the axis. Expected result: measurable misalignment or known co-directional vibration. Next action: realign the axis; if the application needs two planes, use a dual-axis model (selection guidance in inclination-sensor selection guide).
- Observation: noise band within the reading (not just mean shift). Hypothesis: vibration aliasing into the tilt reading; filtering/sampling insufficient for the spectrum. Check: compare machine's dominant frequencies with the sensor's documented filtering and sampling behavior; review the model's filter options. Expected result: spectrum overlaps the measurement band. Next action: configure the model's filtering options per the manual; where the spectrum defeats the installed model, reselect with the vibration discussion in inclination-sensor selection guide.
- Observation: drift independent of machine state. Hypothesis: temperature drift. Check: correlate reading with temperature logs over a full thermal cycle. Expected result: reading tracks temperature within or beyond the model's specification. Next action: within spec → compensate in the application; beyond spec → model review.
- Observation: zero offset since installation or remounting. Hypothesis: mounting stress or residual offset. Check: zero verification on a trusted level reference; compare against the model's stated offset/zero characteristics. Expected result: offset within the documented class. Next action: apply the model's zero procedure against a real zero state; do not zero on a vibration-biased reading.
- Observation: dynamic deviation is real structural tilt under load. Hypothesis: the machine actually deflects. Check: compare with design deflection data or an independent reference method. Expected result: deviation matches structural behavior. Next action: the sensor is honest — treat loaded tilt as a process measurement, not a fault.
- Observation: controller value noisy while local/serial readout is clean. Hypothesis: signal-chain noise. Check: parallel comparison local vs. controller; verify wiring, shielding and scaling (analog) or bus integrity (serial/CAN). Expected result: divergence isolated to the chain. Next action: signal-path diagnosis (intermittent-PLC-input troubleshooting guide); interface trade-offs in interface-selection guide and the integration guide (NPN/PNP integration guide).
Conditional Corrective Actions
- Mounting resonance/flex confirmed → shorten and stiffen the path; mount directly to structure where possible; re-run the worksheet after each change.
- Axis projection confirmed → realign; move to dual-axis where two planes matter.
- Aliasing/spectrum problem → configure documented filter options per the manual (KJT Sensors documents fifth-order filtering in its inclination products — manufacturer-stated); if the spectrum still defeats it, reselect via inclination-sensor selection guide.
- Thermal drift within spec → compensate in application software against temperature; document it.
- Real structural tilt → use the loaded value as the process measurement; don't chase it.
- Signal-chain noise → per intermittent-PLC-input troubleshooting guide; accessory replacements follow the equivalence discipline (cable-or-bracket replacement topic).
Replacement and Reselection Criteria
Reselect when the vibration spectrum defeats the installed model's documented filtering/sampling; when the environment or accuracy class exceeds the installed model's temperature/accuracy specification; or when the application needs axes/outputs the installed model lacks (single→dual axis; analog→serial/CAN for noise immunity over distance). All selection runs through the inclination selection guide (inclination-sensor selection guide), which handles the vibrating-machine case at selection time — this article handles the installed case.
Information to Send KJT Sensors Support
- Sensor model, axes, output and documented filtering/temperature specifications;
- The completed worksheet: static, rigid-reference, dynamic, post-stop values with temperatures;
- Machine vibration description: sources, dominant frequencies, direction relative to the sensitive axis;
- Mounting photos: bracket geometry, fastening, distance to structure, axis alignment;
- Temperature environment and cycles; reading history;
- Required accuracy and response in the application.
Limitations
- Filtering characteristics, temperature specifications, zero procedures and axis conventions are model-specific: every configuration reference points to the exact manual — none is quoted beyond manufacturer-stated documentation.
- Structural deflection verification belongs to machine design engineering; the sensor article only separates measurement faults from real tilt.
Frequently Asked Questions
Why does an inclination sensor drift on a vibrating machine?
Usually it does not drift — it couples. Vibration reaches the sensing element through a flexible or resonant mounting, projects onto a misaligned sensitive axis, or aliases into the reading when filtering and sampling do not separate it from true tilt; the reading then wanders with machine state. The static-versus-dynamic worksheet proves which is happening: identical stopped/running readings rule out coupling; deviations that vanish after stopping confirm it.
How can mounting stiffness affect inclination readings?
The sensor measures the attitude of its own mounting surface. A bracket that flexes under vibration, or resonates at machine frequencies, moves that surface independently of the machine's true attitude — the sensor faithfully reports its own oscillation as tilt. Short, stiff, directly-fastened mounting keeps the sensor's frame rigid with the structure; the sensor's own filtering (documented for KJT Sensors' inclination products) then handles what remains.
Which static and dynamic data should be compared before recalibration?
Four values at the same nominal attitude: the static reading (machine stopped, settled), a rigid-reference reading if accessible (vibration without the machine's structure in the path), the dynamic reading (machine running, with its noise band and mean), and the post-stop return. Recalibration is justified only after this set shows the drift is thermal/electronic — zeroing a vibration-coupled reading corrupts static accuracy while leaving the coupling intact.
Conclusion
Vibration-related "drift" is coupling until proven otherwise: run the static-versus-dynamic worksheet, fix the mounting and axis before touching filtering, configure documented filter options for what remains, and treat thermal drift and real structural tilt as their own families. Never zero away a coupling fault, and never work on a running machine. Selection for vibrating applications runs through E08's guidance.
Sources
- KJT Sensors, Inclination Sensors page (analog-output single/dual-axis models; verified 2026-10-10) — https://www.kjt-sensors.com/list-qjcgq.html
- KJT Sensors, Measuring Sensors category (inclination family within the measuring portfolio; verified 2026-10-10) — https://www.kjt-sensors.com/list-clcgq.html
- KJT Sensors, Service & Support — https://www.kjt-sensors.com/list-services.html (verified 2026-10-10)
- KJT Sensors internal documentation: inclination sensors (KJT-DRC-K: serial, single/dual-axis, differential A/D, fifth-order filtering; wireless inclination: IP65, −40~80 °C; outputs incl. RS-485/232/422, TTL, CAN, Modbus — manufacturer-stated) — version current at drafting
- General inclination-measurement practice (vibration aliasing, mounting stiffness, thermal effects; no third-party claims)
Send Mounting and Motion Data
Provide the completed worksheet, mounting photos, the machine vibration description and the model number via the KJT Sensors service page — measurement application engineers will verify the coupling path or confirm a reselection.
https://www.kjt-sensors.com/
KJT Sensors
About Author
