Photodetector response testing is an essential characterization process that quantifies how effectively a photodetector converts optical signals into electrical outputs, covering key parameters such as responsivity, spectral response, quantum efficiency, rise and fall time, frequency bandwidth, linearity, and noise-equivalent power. This technical article explains the testing principles, applicable sample types, measured parameters, standard test methods, and typical industry applications, with reference to international standards including the IEC 60747 series, IEC 60904-8, ASTM E1021, and GB/T 6495.8. It is intended for engineers, R&D teams, and quality managers in optical communication, sensing, imaging, and photovoltaic fields who require reliable response data from accredited third-party laboratories.
Technical Overview
A photodetector is a device that converts incident optical radiation into an electrical signal, typically a photocurrent or an output voltage. Common photodetector families include PN and PIN photodiodes, avalanche photodiodes (APDs), phototransistors, photoresistors (photoconductive detectors), photomultiplier tubes (PMTs), silicon photomultipliers (SiPMs), CCD and CMOS image sensors, and thermal detectors such as thermopiles, pyroelectric sensors, and bolometers. Because the usefulness of any photodetector is ultimately defined by how faithfully it responds to light, response testing is the core of photodetector characterization and qualification.
Response testing evaluates both the static and dynamic behavior of the device. Static response parameters describe the magnitude and wavelength dependence of the electrical output for a given optical input, expressed primarily as spectral responsivity in A/W and external quantum efficiency (EQE) in percent. Dynamic response parameters describe how quickly the detector reacts to changing light, expressed as rise time, fall time, cutoff frequency, and the -3 dB bandwidth. Noise-related parameters such as dark current, noise-equivalent power (NEP), and specific detectivity (D*) define the weakest optical signal the detector can resolve.
The most fundamental relationships used in response testing are simple but powerful. Responsivity is defined as the photocurrent generated per unit of incident optical power, R = Iph/Popt, in amperes per watt. External quantum efficiency is derived from responsivity through the relation EQE(%) = 1240 x R / λ, where λ is the wavelength in nanometers, so a silicon photodiode with 0.55 A/W responsivity at 850 nm exhibits an EQE of approximately 80 percent. Because responsivity depends strongly on wavelength, reverse bias, temperature, and optical power level, professional testing must control and report all of these conditions.
Response testing is typically performed by accredited third-party laboratories operating under ISO/IEC 17025 quality management systems, using optical sources and reference detectors that are traceable to national metrology institutes. Testing follows recognized international and national standards such as the IEC 60747 series for discrete semiconductor optoelectronic devices, IEC 60904-8 and ASTM E1021 for the spectral responsivity of photovoltaic devices, and the corresponding GB/T standards for the Chinese market.
Testing Samples
PN and PIN Photodiodes
Silicon PIN photodiodes covering the 190 to 1100 nm range, germanium photodiodes for 800 to 1650 nm, and indium gallium arsenide (InGaAs) photodiodes covering 800 to 2600 nm are the most frequently tested samples. Samples may be submitted as bare dies, chips on wafer, TO-can packaged devices, surface-mount components, or fiber-pigtailed modules. Both unbiased (photovoltaic) and reverse-biased (photoconductive) operating modes can be characterized.
Avalanche Photodiodes (APDs) and Silicon Photomultipliers (SiPMs)
APDs operated at high reverse bias exhibit internal gain, so response testing must include the multiplication gain versus bias voltage, gain-dependent responsivity, breakdown voltage, and excess noise behavior. SiPMs and other Geiger-mode devices require specialized quenching circuits and single-photon-level light sources.
Phototransistors and Photodarlingtons
These devices are tested for collector photocurrent under defined illumination, light and dark current ratios, and switching speed. Their inherent transistor gain makes linearity testing particularly important.
Photoconductive Detectors (Photoresistors)
Cadmium sulfide (CdS), lead sulfide (PbS), and lead selenide (PbSe) photoresistors are tested for light resistance, dark resistance, resistance ratio under specified illumination, and relatively slow response times. Cooled PbS and PbSe detectors are common in infrared applications.
Photomultiplier Tubes
PMTs are characterized for anode luminous sensitivity, cathode radiant sensitivity, gain at specified supply voltage, rise time, and transit time spread, using calibrated pulsed light sources and high-voltage bias networks.
CCD and CMOS Image Sensors
Linear arrays and area image sensors are tested for per-pixel photoresponse uniformity (PRNU), quantum efficiency versus wavelength, full-well capacity, and photoresponse nonlinearity across the dynamic range.
Thermal Detectors
Thermopiles, pyroelectric detectors, and bolometers exhibit broad, flat spectral response and are typically tested for voltage or current responsivity, time constant, and noise-equivalent power in the far infrared up to 20 micrometers or beyond.
Photovoltaic Cells and Receiver Modules
Crystalline silicon, thin-film, and emerging perovskite or tandem solar cells are tested for spectral response and EQE per IEC 60904-8 and GB/T 6495.8. Optical receiver modules, fiber-coupled photodiodes, and time-of-flight sensor assemblies can be tested as complete units.
Testing Items
Typical response testing items for photodetectors include the following parameters, each reported with the corresponding test conditions such as wavelength, bias voltage, temperature, and optical power level:
- DC responsivity at a specified wavelength (A/W)
- Spectral responsivity curve and spectral response range
- External quantum efficiency (EQE) and derived internal quantum efficiency
- Photocurrent under standard or customer-defined illumination
- Dark current and dark output voltage
- Shunt resistance and series resistance
- Breakdown voltage, multiplication gain, and excess noise factor for APDs
- Light-to-dark current or resistance ratio for photoresistors and phototransistors
- Rise time and fall time (10 to 90 percent), delay time, and settling behavior
- Frequency response and -3 dB bandwidth
- Linearity, dynamic range, and saturation optical power
- Noise-equivalent power (NEP) and specific detectivity (D*)
- Spatial response uniformity across the active area
- Angular response and field-of-view characterization
- Temperature coefficient of responsivity and temperature-dependent response curves
Testing Methods
DC Responsivity Measurement
A wavelength-stabilized laser diode or filtered LED source illuminates the detector at a precisely known optical power, which is measured with a calibrated reference detector or optical power meter. A source measure unit (SMU) or electrometer applies the specified bias and measures the photocurrent. The dark current is measured with the beam blocked and subtracted, yielding responsivity as R = (Ilight - Idark)/Popt. This method follows the measuring-methods framework of the IEC 60747 series for photodiodes and phototransistors.
Spectral Response and EQE Measurement
A xenon or tungsten lamp coupled to a monochromator, or a tunable laser, generates monochromatic light that is scanned across the wavelength range of interest, typically 200 to 2600 nm in steps of 1 to 10 nm. A chopper and lock-in amplifier configuration is used when the photocurrent is weak. The reference detector substitution method, in which a calibrated standard detector alternates with the device under test, ensures spectral accuracy traceable to national metrology institutes. External quantum efficiency is calculated directly from the measured spectral responsivity using EQE(%) = 1240 x R/λ. For photovoltaic samples, this method is performed in accordance with IEC 60904-8, ASTM E1021, and GB/T 6495.8, including bias-light options for bifacial and multi-junction devices.
Time-Domain Transient Response Testing
Rise time and fall time are measured by illuminating the detector with a pulsed laser whose pulse width is significantly shorter than the expected detector response. The output is terminated into a matched impedance, typically 50 ohms, passed through a bias tee, and captured with a high-bandwidth sampling oscilloscope. The 10 to 90 percent rise time and the 90 to 10 percent fall time are extracted, and the small-signal bandwidth can be estimated from the relation BW = 0.35 divided by the rise time. Picosecond-class pulse lasers and oscilloscopes allow characterization of detectors with bandwidths beyond 10 GHz.
Frequency-Domain Response Testing
The frequency response is measured by driving a laser diode or external electro-optic modulator with a swept sinusoidal signal from a vector network analyzer. The electrical response of the photodetector is recorded as an S21 transmission trace, from which the -3 dB cutoff frequency is read directly. This method is the preferred technique for high-speed receivers in optical communication and provides amplitude and phase information across frequencies from the kilohertz range to tens of gigahertz.
Linearity and Dynamic Range Testing
Linearity is evaluated by stepping the incident optical power over many orders of magnitude using calibrated neutral-density attenuators while recording the photocurrent. The superposition, or double-light, method, in which the response to two beams is compared with the sum of the individual responses, verifies linearity independent of power-meter calibration. The results define the linear dynamic range between the noise floor and the onset of saturation or compression.
Noise, NEP, and Detectivity Measurement
The detector is placed in an electrically shielded, light-tight enclosure at the specified bias and temperature. The output noise spectral density is measured with a low-noise amplifier and FFT analyzer or lock-in technique. NEP is calculated as the noise current spectral density divided by the responsivity, expressed in watts per square root hertz. Specific detectivity D* is then reported as the square root of the detector active area divided by NEP, in units of cm·Hz^0.5/W, enabling fair comparison between detectors of different sizes.
Temperature-Dependent and Environmental Response Testing
The detector is mounted inside a climate chamber, and responsivity, dark current, rise time, and noise are recorded at multiple temperature points over ranges such as -40 to +85 degrees Celsius for industrial-grade devices. The temperature coefficient of response is derived from the slope of these curves. Environmental conditioning tests, such as damp heat and thermal cycling per IEC 60068 or applicable GB/T standards, can be combined with before-and-after response measurements to assess long-term stability.
Applications
Photodetector response testing supports product development, incoming inspection, reliability monitoring, and regulatory compliance across a wide range of industries:
- Optical fiber communication: verification of PIN and APD receiver responsivity, bandwidth, and overload behavior at 850, 1310, and 1550 nm for data center and telecom transceivers
- LiDAR and time-of-flight sensing: rise-time, gain, and NEP qualification of APDs and SiPMs for automotive, robotics, and surveying systems
- Machine vision and imaging: quantum efficiency, photoresponse uniformity, and linearity characterization of CCD and CMOS sensors and camera modules
- Spectroscopy and analytical instruments: spectral response validation of InGaAs, PbS, and PbSe detectors in spectrometers and gas analyzers
- Photovoltaics: EQE and spectral response measurement of solar cells for efficiency diagnostics, current-loss analysis, and calibration of solar simulators
- Medical technology: response verification of detectors used in pulse oximetry, dental curing light meters, phototherapy dosimetry, and PET imaging modules
- Industrial automation: qualification of photoelectric sensors, encoders, light curtains, and flame detectors
- Environmental and safety monitoring: UV index meters, haze and air quality sensors, and smoke and flame detection equipment
- Consumer electronics: ambient light sensors, proximity sensors, and wearable optical sensing modules
FAQ
What is the difference between responsivity and quantum efficiency?
Responsivity is the ratio of the electrical output current to the incident optical power, expressed in A/W, and depends on wavelength. Quantum efficiency describes the percentage of incident photons that generate electron-hole pairs. The two are directly related through EQE(%) = 1240 x R/λ, with λ in nanometers, so a measurement of one fully determines the other.
What are typical responsivity values for common detectors?
Silicon PIN photodiodes typically show 0.4 to 0.6 A/W between 600 and 950 nm, InGaAs detectors reach 0.8 to 1.0 A/W around 1550 nm, germanium detectors provide roughly 0.5 to 0.7 A/W near 1500 nm, and APDs exhibit effective responsivities multiplied by their internal gain, which can range from tens to several hundreds at typical operating voltages.
What wavelength and power ranges can be covered in testing?
Using monochromator-based systems with calibrated reference detectors, spectral response testing typically covers 200 nm to 2600 nm, and thermal detectors can be characterized further into the mid and far infrared. Optical power levels from the picowatt region, using lock-in techniques, up to hundreds of milliwatts can be accommodated, supporting both detectivity and saturation measurements.
How fast a response time can be measured?
With picosecond pulsed laser sources, bias tees, and high-bandwidth sampling oscilloscopes, rise times down to the tens of picoseconds level can be resolved, corresponding to bandwidths above 10 GHz. Frequency-domain measurements using a vector network analyzer extend characterization to tens of gigahertz for communication-grade receivers.
Why are dark current and noise measurements included in response testing?
Dark current sets the shot-noise floor that limits the smallest detectable signal, and it typically doubles for every 8 to 10 degrees Celsius of temperature rise. Combined with responsivity, noise data yield NEP and D*, which are the decisive figures of merit for weak-light applications such as LiDAR, spectroscopy, and scientific imaging.
What standards govern photodetector response testing?
Discrete photodiodes and phototransistors are measured according to the IEC 60747 series and corresponding GB/T national standards, photovoltaic spectral response follows IEC 60904-8, ASTM E1021, and GB/T 6495.8, and laboratory competence is verified under ISO/IEC 17025. Environmental conditioning referenced during response stability testing follows the IEC 60068 series or equivalent GB/T standards.
How should samples be prepared for submission?
Samples should be submitted with the device datasheet, the requested test wavelengths, bias conditions, load impedance, temperature range, and any required reference standards. Packaged devices, bare dies, wafer-level samples, and fiber-coupled modules can all be tested, provided the active area is accessible and the electrical pads or leads are compatible with probing fixtures.