Introduction
PIR sensors are designed to detect human body radiation, which peaks around 9.7 μm. To achieve this selectively, they incorporate optical filters that transmit only the 8-14 μm band while blocking other wavelengths.
Why Spectral Filtering Matters
Without filtering, a PIR sensor would respond to visible light, near-IR from remote controls and sunlight, mid-IR from hot objects like engines and heaters, and far-IR beyond human emission band. This would cause constant false triggers. The optical filter is essential for selective human detection.
Filter Technologies
Silicon Windows with Dielectric Coatings
Most PIR sensors use a silicon window coated with a multilayer dielectric interference filter. Silicon is transparent in IR, and the coating creates a bandpass filter centered on 8-14 μm.
Advantages: Stable, precise cutoff, can be integrated with sensor package.
Germanium Windows
Germanium has excellent IR transmission but is more expensive. Used in high-end sensors.
Polymer Filters
Some low-cost sensors use dyed or coated polymer films, but these are less stable and precise.
Spectral Response Characteristics
A typical PIR sensor spectral response:
- Blocked: < 5 μm (visible, near-IR)
- Transmission band: 8-14 μm (human detection)
- Blocked: > 15 μm (reduces background)
Impact on Detection
Human Detection
Human body emission at 300K peaks at 9.7 μm, well within the 8-14 μm band. Filter matches human emission.
Rejection of Sunlight
Sunlight contains significant IR, but its spectrum peaks at shorter wavelengths. The filter blocks most solar IR, reducing daytime false triggers.
Rejection of Artificial Light
Incandescent lamps emit broad-spectrum IR; fluorescent and LED lamps have little IR. Filter blocks residual IR from lamps.
Filter Degradation
Over time, filters can degrade due to:
- Moisture absorption (shifts spectrum)
- UV exposure (if not protected)
- Contamination (dust, oils)
- Mechanical damage (scratches)
Testing Spectral Response
To verify a sensor’s spectral response, you need a monochromator or tunable IR source, a reference detector, and a lock-in amplifier for signal extraction. This is typically done by manufacturers, not end users.
Custom Filters for Special Applications
- High-temperature detection: Filter passing shorter wavelengths (3-5 μm) for detecting hot objects like engines
- Flame detection: Special filters for CO2 emission band (4.3 μm)
- Gas sensing: Narrowband filters for specific gas absorption lines
Conclusion
The optical filter is as important as the pyroelectric element itself. It defines what the sensor “sees” and is critical for reliable human detection. Understanding filter characteristics helps in selecting the right sensor and diagnosing performance issues.