Understanding sensor specifications is critical for obtaining meaningful measurements. Terms like accuracy, resolution and calibration are often misunderstood, leading to incorrect expectations and poor system design.
This article explains the most important measurement concepts and how they affect real-world sensor performance.
Accuracy
Accuracy describes how close a measured value is to the true value.
- Specified as ± value (e.g. ±0.5°C)
- Includes sensor and calibration errors
Example:
- A sensor with ±0.5°C accuracy may show 24.5°C to 25.5°C when the actual temperature is 25°C
Important: Accuracy does not improve by averaging readings.
Resolution
Resolution is the smallest change a sensor can detect.
- Often determined by ADC or internal processing
Example:
- A 12-bit ADC over 0-3.3V has ~0.8mV resolution
Important: High resolution does not mean high accuracy.
Repeatability
Repeatability describes how consistent measurements are under the same conditions.
- Good repeatability = stable readings
- Poor repeatability = fluctuating results
This is often more important than absolute accuracy in many applications.
Calibration
Calibration adjusts sensor output to match a known reference.
- Factory calibration is built into many sensors
- User calibration improves accuracy
Example:
- Adjusting a temperature sensor offset using a known reference thermometer
Drift
Drift is the gradual change in sensor readings over time.
- Caused by aging, contamination or environmental factors
Example:
- Humidity sensors may drift after long-term exposure
Response Time
Response time describes how quickly a sensor reacts to changes.
- Fast sensors respond quickly to changes
- Slow sensors smooth rapid variations
This is important in dynamic environments.
Noise and Filtering
Sensor readings often include noise:
- Electrical noise
- Environmental fluctuations
Common solutions:
- Averaging multiple readings
- Low-pass filtering
- Shielding and proper wiring
Comparison Overview
| Term | Meaning | Important Note |
|---|---|---|
| Accuracy | Closeness to true value | Cannot be improved by averaging |
| Resolution | Smallest detectable change | Not equal to accuracy |
| Repeatability | Consistency of readings | Often more important |
| Calibration | Adjustment to reference | Improves accuracy |
| Drift | Change over time | Requires recalibration |
Practical Examples
- A cheap sensor may have high resolution but poor accuracy
- A high-quality sensor provides stable and repeatable readings
- Calibration can significantly improve low-cost sensors
Common Mistakes
- Confusing resolution with accuracy
- Ignoring sensor drift over time
- Expecting laboratory precision from low-cost sensors
Practical Recommendations
- Choose sensors based on required accuracy, not just resolution
- Calibrate sensors when possible
- Focus on repeatability for relative measurements
Conclusion
Understanding sensor specifications is essential for reliable measurements. Accuracy, resolution and repeatability all play different roles, and confusing them can lead to incorrect system design.
With proper calibration and realistic expectations, even low-cost sensors can provide useful and reliable data.