Short answer: a vibration sensor’s bandwidth defines the range of motion it can measure, but the engineering value comes from matching that bandwidth to the faults you actually need to detect. The PyXis smart 3-axis vibration sensor provides a DC–6 kHz measurement bandwidth, tri-axial vibration processing, 20 harmonics with amplitude per axis, temperature monitoring and Modbus RTU access to processed condition information.
Engineering point: “6 kHz” should not be treated as a marketing number. It is the upper measurement-band limit. Whether it is useful depends on machine speed, expected fault frequencies, mounting quality, sensor orientation and the diagnostic indicators being monitored.
What does DC–6 kHz vibration bandwidth mean?
Bandwidth describes the frequency range over which a sensing system is intended to measure vibration. A low-frequency machine movement and a high-frequency mechanical impact are different physical events. A useful monitoring system therefore needs enough frequency range for the failure modes that matter on the target machine.
For PyXis, the current engineering datasheet specifies a DC–6 kHz vibration measurement bandwidth. This creates room to monitor both slower machine-motion behaviour and substantially higher-frequency vibration content within the sensor’s specified range.
That does not mean that every event below 6 kHz is automatically identified as a particular fault. Frequency range is only one part of the measurement chain. Interpretation still depends on the machine, shaft speed, bearing and gear geometry, mounting, loading state, operating history and alarm strategy.
Why bandwidth matters for different machine faults
| Monitoring question | Why frequency range matters | What to check in the application |
|---|---|---|
| Is the rotor becoming unbalanced? | Rotational problems often create strong components related to running speed. | Machine RPM, axis orientation, load state and baseline amplitude. |
| Is alignment or looseness changing? | Mechanical faults can produce multiple running-speed components and changing harmonic patterns. | Harmonic amplitudes, trend stability and mechanical mounting condition. |
| Is a gearbox developing a problem? | Gear-mesh-related content may occur well above shaft rotational frequency. | Gear ratio, tooth count, running speed and expected mesh frequencies. |
| Are high-frequency impacts increasing? | Impulsive mechanical events can contain energy at higher frequencies than ordinary running-speed motion. | Sensor mounting, bandwidth margin, repeatability and the correct diagnostic indicator. |
The practical lesson is simple: do not choose a vibration sensor only by asking whether it measures “vibration.” Start with the machine’s mechanical frequencies and the faults you want to screen.
Three axes reduce the risk of looking in the wrong direction
Machine vibration is directional. A fault may appear more strongly in one physical direction than another depending on shaft orientation, bearing housing stiffness, mounting surface and machine structure.
PyXis measures on X, Y and Z axes and provides axis-specific vibration information. The current engineering data includes acceleration RMS, peak, peak-to-peak and crest factor; velocity RMS and peak-to-peak; and displacement peak-to-peak, together with spectral and harmonic information.
This does not remove the need for correct mounting and axis alignment. It does, however, give the commissioning engineer more information than a single-axis measurement when the dominant vibration direction is not obvious or changes with machine condition.
Why PyXis provides 20 harmonics per axis
A single overall vibration value can show that something changed, but it may not explain how the vibration pattern changed. Harmonic information helps separate a simple amplitude increase from a more structured change in the machine’s frequency content.
PyXis provides 20 harmonics with amplitude per axis. Used correctly, this can help engineers compare changes in frequency components against known machine speed and mechanical geometry.
For example, a maintenance engineer can trend the overall level while also watching whether particular harmonic components are growing. That combination is often more informative than relying on one total vibration number alone.
Bandwidth is useful only when mounting is good
Higher-frequency vibration is especially sensitive to the mechanical path between the machine and the sensor. A poor mounting surface, loose fastener, flexible bracket or unsuitable adhesive layer can alter the measured response before the signal reaches the sensing element.
For a meaningful installation:
- mount the sensor on a rigid machine surface close to the mechanical source of interest;
- avoid flexible covers and thin sheet metal where possible;
- record the X/Y/Z orientation during commissioning;
- capture a healthy-machine baseline under representative operating conditions; and
- repeat the same operating state when comparing trends.
A 6 kHz sensor mounted poorly can produce less useful diagnostic information than a lower-bandwidth sensor mounted correctly. Installation remains part of the measurement system.
Why edge processing matters on an RS-485 network
Continuous raw waveform transport is not always practical across an industrial RS-485 network, especially when many monitoring points share the same communication segment. PyXis is designed to process vibration and temperature information at the machine and expose compact condition information over Modbus RTU with automatic baud detection.
This architecture lets a PLC, SCADA system or maintenance platform read health metrics, alarms and diagnostic indicators without treating the RS-485 network as a high-volume waveform streaming link.
For brownfield plants, this can be important. It allows condition information to be added to existing control infrastructure while keeping network design, polling rate and register use manageable.
A practical PyXis bandwidth-selection checklist
- Record the machine speed range. Include minimum, normal and maximum RPM.
- List the failure modes that matter. Separate imbalance, alignment, looseness, gear, bearing and process-related concerns.
- Calculate or obtain expected mechanical frequencies. Include gear ratios, tooth counts and bearing geometry where available.
- Confirm frequency margin. Ensure the diagnostic content you need falls comfortably inside the intended measurement range.
- Choose the mounting point and axes deliberately. Do not treat mounting as an afterthought.
- Define what the PLC or SCADA will read. Decide which vibration, temperature, harmonic and alarm indicators are useful operationally.
- Baseline before alarming. Healthy operating data should be captured before warning and alarm thresholds are finalised.
When 6 kHz is not the whole answer
Some applications need specialised high-frequency techniques, detailed raw-waveform analysis, laboratory instrumentation or a different sensing technology. Other applications need far less bandwidth but place more importance on low-frequency stability, process correlation or long-term trend consistency.
The correct question is therefore not “Is 6 kHz enough?” in isolation. It is:
Does the sensor’s usable frequency range, measurement set, mounting arrangement and integration method fit the failure modes we need to detect on this specific machine?
That is the engineering question PyXis should be qualified against.
Match PyXis to your machine before installing the sensor
Send us the machine type, speed range, bearing or gearbox information, mounting location, PLC/SCADA platform and the failure modes you want to detect. We can help screen the measurement and integration requirements before deployment.
For initial product matching, use the Tachyonic product selector.
Source and technical fact-check
- PyXis 222x Series Datasheet TTN-DS-2026-002 Rev 08, issue date 22 July 2026: 3-axis vibration measurement, DC–6 kHz measurement bandwidth, 20 harmonics with amplitude per axis, Modbus RTU with automatic baud detection and edge condition processing.
- PyXis product page, checked 13 September 2026.
Engineering note: bandwidth is one element of a condition-monitoring design. Final fault-detection performance depends on machine dynamics, mounting, orientation, operating state, configuration, thresholds and the suitability of the selected diagnostic indicators. Safety-related functions require a separately designed and validated safety system.
