Excitation & completion
RTDs, thermistors, strain gauges and bridge sensors may require controlled excitation, bridge completion or other sensor-specific circuitry before a usable measurement signal exists.
MEASUREMENTPERFORMANCE / CONDITIONING / UNCERTAINTY / CALIBRATION
Industrial Sensor Measurement: Accuracy, Resolution, Uncertainty & Calibration
A sensor measurement result depends on the complete measurement chain, not a single datasheet number. Sensor behaviour, signal conditioning, acquisition hardware, calibration, environmental influences and data processing can all affect the value reported by a PLC, DAQ or instrument.
Evaluate measurement performance by separating accuracy from resolution, checking how the signal is conditioned and acquired, identifying relevant uncertainty sources, and maintaining calibration and metrological traceability where the application requires it.
MEASUREMENT SCOPE4 GUIDES
- Accuracy
- PERFORMANCE TERMS
- Condition
- SIGNAL QUALITY
- Uncertainty
- RESULT CONFIDENCE
- Calibration
- REFERENCE CHAIN
ACCURACY / SIGNAL / UNCERTAINTY / TRACEABILITY
COVERSAccuracyResolutionSignal conditioningUncertaintyCalibrationTraceability
CORE MEASUREMENT GUIDES
Measurement performance is built from several different engineering concepts.
Accuracy and resolution describe different aspects of performance; signal conditioning prepares the sensor output for acquisition; uncertainty qualifies the measurement result; calibration and traceability connect that result to stated references.
PERFORMANCE01
Sensor Accuracy vs Resolution
Distinguish accuracy, resolution, precision, repeatability, linearity and related specifications without treating them as interchangeable.
OPEN →CONDITIONING02
Sensor Signal Conditioning
Excitation, amplification, attenuation, filtering, isolation, linearisation and sensor-specific conditioning before the signal is digitised or processed.
OPEN →UNCERTAINTY03
Measurement Uncertainty
Identify uncertainty contributions, distinguish uncertainty from measurement error, and understand what an uncertainty statement says about a result.
OPEN →CALIBRATION04
Calibration & Metrological Traceability
Calibration relationships, reference standards, uncertainty contributions and the documented chain used to establish traceability of measurement results.
OPEN →MEASUREMENT PATH
Follow the value from the measurand to the reported result.
The sensor is only one part of the measurement system. A physical quantity is converted into a sensor response, that response may be conditioned, the signal is acquired or digitised, and software or controller logic converts the indication into an engineering value.
Errors or uncertainty can enter at each stage. Poor mounting can alter the measurand seen by the sensor; conditioning can add offset or noise; an input module can limit resolution; scaling can be wrong even when the hardware is functioning correctly.
MEASURANDSENSORCONDITIONACQUIRERESULT
ACCURACY & RESOLUTION
More digits do not make a measurement more accurate.
A system can resolve a very small change while still carrying offset, gain, linearity, drift or calibration errors that are much larger than one display count or ADC step.
| Characteristic | What it describes | Engineering check |
|---|---|---|
| Accuracy | Closeness of agreement between a measured value and the value of the measurand, considered with the stated reference and conditions. | Read the manufacturer's numerical error specification and its basis rather than assuming resolution or repeatability implies accuracy. |
| Resolution | The smallest change in the quantity being measured that produces a perceptible change in the corresponding indication. | Check the complete chain: sensor output, conditioning, ADC or input-module resolution, scaling and displayed increment. |
| Precision | Closeness of agreement among indications or measured values obtained by replicate measurements under specified conditions. | Use repeated measurements to assess scatter, and state the conditions under which the precision applies. |
| Repeatability | Measurement precision under repeatability conditions: the same measurement procedure, operators, measuring system, operating conditions and location over a short period of time. | Use repeatability data only for the stated repeatability conditions; it does not replace absolute error or uncertainty information. |
| Linearity | How the sensor response departs from the specified linear reference over the stated range. | Confirm the reference line, range and test conditions used for the linearity specification. |
| Drift | Change in indication or metrological properties over time when the measured quantity is not causing that change. | Include long-term stability, temperature history and recalibration interval where they can affect the decision. |
| Response / bandwidth | How quickly and over what frequency range the system can follow changes in the measurand. | Match the dynamic response to the event or control loop; steady-state accuracy alone is not enough. |
For a focused comparison of these specifications, use Sensor Accuracy vs Resolution.
SIGNAL CONDITIONING
Condition the signal for the sensor, the cable and the receiving input.
Signal conditioning is not one universal stage. The required functions depend on the sensing principle, signal level, wiring, noise environment, isolation requirements and the analogue or digital input that receives the signal.
Amplification & attenuation
Match the useful signal span to the input range without clipping. Poor range use can waste converter resolution; excessive gain can amplify offsets or interference.
Filtering & bandwidth
Limit unwanted noise and frequencies without removing the process dynamics that matter. Analogue anti-alias filtering is especially important before sampled acquisition.
Isolation & linearisation
Isolation can break unwanted conductive paths and protect measurement interfaces; linearisation and compensation convert non-linear or temperature-dependent sensor behaviour into useful values.
The detailed conditioning requirements for common sensor families are covered in Sensor Signal Conditioning.
MEASUREMENT UNCERTAINTY
Uncertainty qualifies the result; it is not another name for error.
Measurement uncertainty is a non-negative parameter that characterises the dispersion of quantity values attributed to a measurand based on the information used. A measurement result is normally interpreted together with its associated uncertainty when that uncertainty is significant for the decision.
Measurement error
Measurement error is the measured value minus a reference quantity value. It can have systematic and random components, but the exact error of an ordinary measurement is generally not completely known.
- A known correction can reduce a recognised systematic effect.
- Residual effects remain after correction.
- Error and uncertainty answer different questions.
Measurement uncertainty
Uncertainty describes the dispersion associated with the values that could reasonably be attributed to the measurand from the available information.
- Contributions may come from repeatability, resolution and calibration.
- Environment, installation and the measurement model can contribute.
- The method and coverage used to report uncertainty must be stated.
For uncertainty components, budgets, standard uncertainty and expanded uncertainty, continue to Measurement Uncertainty.
CALIBRATION & TRACEABILITY
Calibration establishes a measurement relationship; traceability belongs to the result.
Calibration compares indications with quantity values provided by measurement standards, under specified conditions, and establishes the relationship used to obtain measurement results. It should not be confused with adjustment or with a simple pass/fail verification.
Define the measurand
State what quantity is being measured, the range, relevant conditions and the measurement method. Traceability is meaningful only when the result itself is defined.
Use suitable references
Reference standards and calibration methods must cover the required quantity, range and conditions with uncertainty suitable for the intended measurement.
Record uncertainty
Each calibration in a traceability chain contributes measurement uncertainty. A certificate or sticker without the necessary measurement information does not establish the quality of every later result.
Maintain the chain
Metrological traceability relates a measurement result to a stated reference through a documented unbroken chain of calibrations, each contributing to the uncertainty.
Check fitness for purpose
Traceability alone does not make a measurement adequate. The resulting uncertainty, range, dynamics and environmental performance must still satisfy the engineering decision.
See Calibration & Metrological Traceability for the relationship between calibration, reference standards, uncertainty and traceability chains.
SYSTEM-LEVEL CHECK
Judge the measurement chain, not only the sensor specification.
A high-performance sensor cannot compensate for an unsuitable process connection, noisy wiring, incorrect excitation, poor input-range use, wrong scaling or an unmaintained calibration system. The final result is limited by the complete path from the measurand to the engineering value.
Sensors
Choose sensing principles and sensor types that match the measurand, operating range, environment and installation.
MEASURAND / PRINCIPLE / RANGESignals
Match 4–20 mA, 0–10 V, PNP/NPN, pulse/frequency and wiring interfaces to the receiving control or acquisition system.
OUTPUT / WIRING / INTERFACEReference
Check terminology, SI units, standards and sensor-selection criteria before comparing datasheets or specifying equipment.
TERMS / UNITS / STANDARDS