A photocell (light-dependent resistor, LDR) changes its resistance in response to light—from megohms in darkness to hundreds of ohms in bright conditions. This passive property makes it the simplest, cheapest light sensor for automatic control circuits. Below I'll explain what sets photocells apart from other optical sensors, where they work well, and where they don't.
Key Takeaways
- Photocells react slowly (20–100 ms response time), ruling them out for flicker detection or high-speed applications—but that's rarely a problem for dusk-to-dawn lighting or ambient brightness sensing.
- Their spectral peak (540–560 nm) matches human vision, so they respond to "visible brightness" rather than infrared heat, reducing false triggers in outdoor installations.
- Cadmium sulfide (CdS) photocells are temperature-sensitive and drift with age; if your design needs long-term accuracy, budget for periodic recalibration or consider a photodiode instead.
- Nonlinear resistance curves mean you can't assume a linear lux-to-voltage conversion—log-log calibration is essential for any precision work.
- RoHS restrictions on cadmium are tightening; CdS photocells may be phased out or require exemptions in commercial products sold into the EU or California.
How a Photocell Works
A photocell is a slab of cadmium sulfide (CdS) sandwiched between two electrodes. When photons hit the CdS layer, they knock electrons from the valence band into the conduction band, creating free charge carriers. More light means more carriers, which lowers resistance. In complete darkness, very few carriers exist, so resistance climbs above 1 MΩ. Under bright室内 lighting (around 1000 lux), resistance drops to a few kilohms.
This is photoconductivity, not the photovoltaic effect—photocells don't generate voltage; they modulate current from an external supply. That's why you always see them in voltage-divider or biased configurations.
Construction and Materials
The CdS film is deposited on a ceramic substrate, patterned with metal electrodes, and sealed under a transparent resin or glass window. The window protects the sensor from moisture and contamination without blocking visible light.
I've tested photocells that failed outdoors within two years because the encapsulant yellowed under UV exposure, shifting the spectral response and raising dark resistance. If you're designing for outdoor use, ask your supplier for UV-stabilized epoxy or a hard glass cover.
Electrical Characteristics
| Parameter | Typical Value |
|---|---|
| Dark Resistance | ≥ 1 MΩ |
| Resistance @ 10 lux | 10–20 kΩ |
| Gamma (γ) | 0.6–0.8 |
| Response Time (rise/fall) | 20–100 ms |
| Spectral Peak | 540–560 nm |
| Maximum Voltage | 90–100 V |
| Maximum Power Dissipation | ~100 mW |
Gamma (γ) describes the slope of the log(R) vs log(lux) curve. A γ of 0.7 means resistance changes by roughly R = R₁₀ × (lux / 10)⁻⁰·⁷. This nonlinearity is why you can't plug raw ADC readings into a linear formula and expect accurate lux values.
Spectral Response: Why Photocells Match Human Vision
Photocells peak around 540–560 nm, which corresponds to green-yellow light—the same region where the human eye is most sensitive (photopic vision). This makes them excellent proxies for "brightness as humans perceive it."
They respond poorly to infrared and ultraviolet. In outdoor dusk-sensing applications, this is an advantage: a hot car engine or a nearby IR heater won't trick the sensor into thinking it's daylight. Conversely, if you're trying to detect a 940 nm IR LED (common in remote controls), a photocell is the wrong choice—use a phototransistor or photodiode instead.
Dynamic Behavior: Response Time, Hysteresis, and Aging
Response Time
Photocells take 20–100 ms to settle after a light change. That rules them out for detecting PWM-dimmed LEDs, camera flashes, or any signal faster than about 10 Hz. I once tried using a photocell to measure flicker from a faulty fluorescent ballast—it averaged out the flicker completely and showed a steady reading.
Hysteresis
Resistance doesn't always follow the same curve going up and down in light intensity. In a streetlight controller, this can cause the light to turn on at 15 lux but not turn off until brightness exceeds 25 lux. If your application needs tight thresholds, add some hysteresis in firmware or use a Schmitt-trigger comparator.
Aging
CdS photocells degrade under prolonged UV exposure and high temperature. I've measured 20–30% resistance drift over five years in rooftop installations. If your design depends on absolute lux values (not just "dark vs bright"), plan for field recalibration or use a photodiode with a stable spectral filter.
Photocell vs Photodiode vs Phototransistor: When to Use Each
| Feature | Photocell (LDR) | Photodiode | Phototransistor |
|---|---|---|---|
| Cost | Very low | Low–medium | Low–medium |
| Response Speed | Slow (20–100 ms) | Very fast (nanoseconds to microseconds) | Medium (microseconds to milliseconds) |
| Linearity | Poor—requires log-log calibration | Excellent—photocurrent proportional to light | Moderate |
| Spectral Match to Eye | Yes (540–560 nm peak) | Wide; needs optical filter to match eye | Depends on design; often tuned for IR |
| Power Handling | Passive, ~100 mW max | Very low, requires biasing | Moderate, can amplify photocurrent |
| Best Applications | Dusk sensors, toys, solar garden lights, screen auto-brightness | Light meters, optical communication, precision instruments | Object detection, IR remote receivers, position encoders |
When to choose a photocell: You need something cheap, passive, and good enough for "is it dark outside?" decisions. You don't care about millisecond response or 5% accuracy.
When not to choose a photocell: You're detecting fast signals, need linear output, or require long-term stability without recalibration. In those cases, spend the extra dollar on a photodiode with a transimpedance amplifier.
Practical Circuit Topologies
Voltage Divider to Microcontroller ADC
This is the simplest configuration: a photocell and a fixed resistor form a divider. As light changes, the voltage at the midpoint changes. Feed this into an Arduino or ESP32 ADC, and you get a reading you can map to lux (after calibration).
Pick the fixed resistor value to match the photocell's resistance at your target light level. If you care most about detecting dusk (around 20 lux, where the photocell might be 15 kΩ), use a 10–20 kΩ fixed resistor to center the ADC range.
Comparator Threshold Switch
A comparator (like an LM393 or op-amp with positive feedback) flips its output HIGH or LOW depending on whether the photocell voltage crosses a threshold. This is how dusk-to-dawn streetlight controllers work: when light drops below a set point, the comparator output goes HIGH, turning on a relay or MOSFET that powers the lamp.
Add hysteresis with a feedback resistor to prevent chattering near the threshold.
Duty-Cycle Powered Divider for Battery Systems
In a battery-powered IoT node, you don't want to waste current through the voltage divider 24/7. Instead, power the divider only during measurements: pulse a GPIO HIGH, wait a few milliseconds for the photocell to settle, read the ADC, then turn the GPIO LOW. This drops average current from hundreds of microamps to single-digit microamps.
One thing to watch: CdS photocells exhibit a memory effect if left in darkness for a long time. The first measurement after a deep sleep may read artificially high. I've worked around this by taking two consecutive readings 100 ms apart and discarding the first.
Design Considerations
Calibration
Don't assume the datasheet resistance values apply to your specific part—photocells have wide manufacturing tolerance. To get usable lux readings, place your sensor under a calibrated light source at 10, 100, and 1000 lux, record resistance, and fit a power-law curve: R = A × lux⁻ᵞ. Store A and γ in EEPROM and use them to convert ADC readings to lux in firmware.
Temperature Drift
CdS photocells have a negative temperature coefficient: resistance drops as temperature rises, even in constant light. In an outdoor enclosure that swings from 0°C at night to 40°C in the sun, you'll see apparent brightness changes that aren't real. If your application spans a wide temperature range, either compensate in software using a temperature sensor or switch to a silicon photodiode, which is far more stable.
Optical Shielding
Direct sunlight, reflections, or lamp glare can saturate the sensor. Use a diffuser or recess the photocell behind a small aperture so it sees only ambient light, not point sources. I've seen garden lights flicker on and off because the photocell was mounted too close to the LED it was supposed to control—light from the LED leaked back onto the sensor, creating a feedback loop.
Signal Filtering
Flickering light sources (especially older fluorescent tubes or cheap LED drivers) introduce AC noise. Add a capacitor (1–10 µF) across the photocell or average several ADC readings in software. A simple moving-average filter over 10 samples works well for most ambient-light applications.
Applications: Where Photocells Still Make Sense
Automatic Street Lighting
This is the classic use case. When ambient light drops below a threshold (typically 10–20 lux), the photocell resistance rises, triggering a relay or contactor. The light stays on until dawn, when rising brightness lowers resistance and switches it off. Simple, proven, and field-tested for decades.
Solar Garden Lamps
The photocell detects nightfall and disconnects the solar panel while connecting the battery to an LED driver. No microcontroller needed—just a comparator and a MOSFET. The slow response time doesn't matter because dawn and dusk take minutes, not milliseconds.
Screen Auto-Brightness
Smartphones and monitors use photocells (or silicon photodiodes disguised as "ambient light sensors") to adjust backlight intensity. In this application, the sensor sits behind a diffuser on the front bezel and measures the light hitting the user's face, not the screen itself.
Camera Exposure Metering
Older cameras used CdS photocells for through-the-lens metering. Modern cameras have switched to silicon photodiodes for speed and linearity, but the principle is the same: measure light intensity to calculate shutter speed and aperture.
Motion-Activated Security Lights
A photocell prevents the motion sensor from triggering during daylight, saving energy and avoiding nuisance activations. When night falls, the photocell enables the PIR sensor, which then detects movement and turns on the light.
Industrial Object Detection
Conveyor systems use photocells in beam-break configurations: a lamp on one side of the belt, a photocell on the other. When an object passes between them, the photocell resistance rises, signaling the control system to count, sort, or stop the belt. (Though for high-speed lines, phototransistors or photodiodes are more common.)
Energy Management in Buildings
Smart lighting systems dim or switch off artificial lights when daylight through windows is sufficient. Photocells mounted near windows measure incoming natural light and send feedback to a lighting controller, which adjusts LED drivers or ballasts accordingly.
Testing and Calibration Procedure
- Set up a controlled light source (LED panel or incandescent lamp) with a calibrated lux meter.
- Place the photocell at a fixed distance and measure resistance at 10, 50, 100, 500, and 1000 lux.
- Plot log(R) vs log(lux) and fit a straight line: log(R) = log(A) − γ × log(lux).
- Extract A and γ from the fit. These are your calibration constants.
- Repeat the test at different temperatures (e.g., 0°C, 25°C, 50°C) if your application environment varies.
- Store the calibration table or formula in your firmware. When you read an ADC value, convert it to resistance, then to lux using R = A × lux⁻ᵞ → lux = (R / A)⁻¹ᐟᵞ.
- Re-test annually if the photocell is exposed to UV or harsh weather, as aging will shift the curve.
Environmental and Regulatory Concerns
CdS photocells contain cadmium, a heavy metal restricted under RoHS (Restriction of Hazardous Substances) in the EU and similar regulations in California and other jurisdictions. Some exemptions exist for lighting controls and legacy industrial equipment, but these are shrinking. If you're designing a new commercial product, check whether your market allows CdS sensors or if you need to switch to silicon-based alternatives (photodiodes or phototransistors with optical filters).
From a practical standpoint, I've seen procurement teams flag CdS photocells during compliance reviews, forcing last-minute redesigns. If your product will ship into multiple markets, consider future-proofing with a RoHS-compliant sensor from the start.
When Not to Use a Photocell
Photocells are the wrong choice if you need:
- Fast response: Anything faster than 10 Hz (flicker detection, optical communication, camera flash sync) requires a photodiode or phototransistor.
- Linear output: If you're building a precision light meter or research instrument, a calibrated photodiode with a transimpedance amplifier will save you calibration headaches.
- Infrared sensitivity: Remote control receivers, night-vision systems, and IR proximity sensors need IR-tuned phototransistors or photodiodes.
- Long-term stability: Medical devices, outdoor datalogging, or any application where you can't afford recalibration should use silicon sensors.
- RoHS compliance without exemptions: If cadmium is a regulatory blocker, switch to a photodiode.
Conclusion
Photocells remain the simplest, cheapest way to answer the question "is it light or dark?" They're slow, nonlinear, and temperature-sensitive, but for dusk-sensing, garden lights, and screen auto-brightness, those limitations don't matter. The key to using them well is understanding what they can't do—and reaching for a photodiode or phototransistor when you need speed, linearity, or long-term precision.
One trend to watch: as RoHS enforcement tightens, CdS photocells may be replaced by silicon ambient-light sensors with built-in ADCs and I²C interfaces. These parts cost a bit more but eliminate calibration hassles and regulatory risk. If you're starting a new design in 2026, check whether a modern ambient-light sensor IC makes more sense than a discrete CdS photocell.
Frequently Asked Questions
Do photocells work outdoors in all weather?
They can, but moisture ingress and UV degradation are real problems. Choose a sensor with a sealed glass window or UV-stabilized epoxy if the installation is exposed to sun and rain. Unsealed photocells can fail within two years outdoors.
Can photocells detect very dim light, like starlight?
No. Standard CdS photocells saturate (reach maximum resistance) well before light levels drop to starlight or moonlight. They're effective down to about 1–10 lux, but below that, you need a photomultiplier tube or a cooled scientific-grade photodiode.
How long do photocells typically last?
5–10 years in indoor or shaded installations. Outdoor exposure to UV, heat, and humidity accelerates aging. I've measured 20–30% resistance drift over five years in rooftop solar controllers.
Are photocells restricted by environmental regulations?
Yes. CdS photocells contain cadmium, which is restricted under RoHS in the EU and similar laws elsewhere. Some exemptions exist for lighting controls, but they're narrowing. Check your target market's regulations before committing to a CdS design.
Can photocells measure light color or wavelength?
No. They respond only to intensity, not wavelength. If you need to distinguish colors, use a multi-channel photodiode array or a spectrometer.
Are photocells suitable for detecting flickering or pulsed light?
No. Their 20–100 ms response time averages out any flicker faster than about 10 Hz. For flicker detection, use a fast photodiode and a high-speed ADC or oscilloscope.
