What is a Photodiode Array and When Do You Need One?
A photodiode array is a monolithic silicon chip containing multiple discrete photodiodes (pixels) arranged in a linear or two-dimensional grid, sharing a single package and substrate. While a single photodiode measures total light intensity hitting its active area, an array resolves spatial light distribution. You need one when building DIY spectrometers, laser beam profilers, optical encoders, or barcode scanners where the position of the light matters just as much as the intensity.
Schematic Symbol and Pinout Description
In schematics, a photodiode array is drawn as multiple standard photodiode symbols (a diode with two inward-pointing arrows) grouped together. The exact pinout depends on whether the manufacturer wired the elements as common-cathode or common-anode.
For a standard 8-element linear common-cathode array, the pinout typically follows this pattern:
- Pin 1: Common Cathode (shared return path for all pixels)
- Pins 2 through 9: Individual Anodes (Anode 1 through Anode 8)
- Pin 10 (if present): Guard ring or shield connection (tie to ground to reduce parasitic capacitance and crosstalk)
Operating Regions and Biasing Strategies
Selecting the right bias voltage is the most critical decision when integrating a photodiode array. You are trading off between linearity/noise (zero bias) and speed/bandwidth (reverse bias). The table below maps the operating regions to typical electrical characteristics.
| Operating Mode | Bias Voltage | Typical Dark Current | Junction Capacitance | Response Speed | Best Application |
|---|---|---|---|---|---|
| Photovoltaic (Zero Bias) | 0 V | < 1 pA | High (e.g., 50-100 pF) | Slow (kHz range) | Precision spectrometry, low-light DC measurements |
| Photoconductive (Low Reverse) | -1 V to -5 V | 10 pA - 1 nA | Medium (e.g., 15-30 pF) | Medium (MHz range) | General-purpose optical sensing, pulse oximetry |
| Photoconductive (High Reverse) | -10 V to -50 V+ | 1 nA - 10 nA | Low (e.g., 2-10 pF) | Fast (100s of MHz) | Laser profiling, high-speed LiDAR, optical comms |
How to select for the job: If your light source is continuous or slowly varying (like a halogen lamp in a spectrometer) and you need maximum signal-to-noise ratio, run the array in photovoltaic mode (0V bias). The dark current drops to virtually zero, eliminating shot noise. If you are profiling a pulsed laser or reading a fast-moving barcode, you must apply a reverse bias to widen the depletion region, drop the junction capacitance, and speed up carrier transit time.
Safe Default Part Numbers for Bench and Production
When prototyping, avoid obscure surplus arrays with no datasheets. The following part numbers are reliable, well-documented defaults available from major distributors like Digi-Key and Mouser in 2026.
- Hamamatsu S4111-16Q: A 16-element linear array. Each pixel is 0.9 mm × 0.9 mm with a 1.0 mm pitch. Peak wavelength is 960 nm (silicon). Packaged in a ceramic DIP. Typical price is around $45. This is the gold standard for DIY and university-level spectrometer builds.
- Hamamatsu S3901-256Q: A 256-element linear array designed specifically for UV-VIS-NIR spectrometry. 50 μm × 500 μm pixels. Requires careful handling due to the dense pitch. Typical price is $110–$130. Hamamatsu's solid-state detector catalog provides exhaustive spectral response curves for this line.
- OSRAM SFH 7072S: If you do not need a linear spectrometer array but rather a multi-chip proximity/color sensor array, this integrated SMD part combines multiple photodiodes with different spectral filters. Priced around $2, it is ideal for consumer IoT projects.
Complete Transimpedance Amplifier (TIA) Application Circuit
Photodiodes output current, not voltage. To read an array, you must convert each pixel's current into a usable voltage using a Transimpedance Amplifier (TIA). Below is a complete, stable TIA design for a single pixel of a Hamamatsu S4111-16Q operating in zero-bias (photovoltaic) mode.
Circuit Topology and Component Values
We use the Texas Instruments OPA380, a low-noise, high-speed op-amp specifically optimized for transimpedance applications. For a multi-pixel array, you would either duplicate this circuit 16 times or use an analog multiplexer (like the CD4051) on the output side, though multiplexing high-impedance nodes directly requires careful PCB guarding.
- Photodiode Connection: Connect the pixel Anode to Analog Ground. Connect the pixel Cathode to the inverting input (Pin 2) of the OPA380.
- Non-Inverting Input: Tie Pin 3 directly to Analog Ground (for zero-bias operation).
- Feedback Resistor ($R_f$): Place a 1 MΩ metal-film resistor (0.1% tolerance) between the inverting input (Pin 2) and the output (Pin 6). This sets your transimpedance gain: 1 μA of photocurrent yields 1 V of output.
- Feedback Capacitor ($C_f$): Place a 2 pF C0G/NP0 ceramic capacitor in parallel with $R_f$. This is non-negotiable. The photodiode's junction capacitance (approx. 40 pF at 0V) introduces a pole in the feedback loop. Without $C_f$, the op-amp will oscillate at high frequencies. The 2 pF cap introduces a compensating zero to ensure phase margin.
- Power Supply: Power the OPA380 with a clean, regulated dual supply of ±5V. Place 100 nF decoupling capacitors as close to the VCC and VEE pins as physically possible.
Failure Modes and Multimeter Testing
Photodiode arrays are fragile. The most common failure mode is Electrostatic Discharge (ESD), which punches a microscopic hole through the silicon junction, creating a permanent low-resistance shunt. The second most common failure is thermal overstress from soldering too close to the glass window or applying excessive reverse bias that causes thermal runaway.
How to Test with a Digital Multimeter (DMM)
You can verify the health of an array on your workbench before soldering it to your board. Set your DMM to Diode Test Mode.
- Forward Bias Test: Place the red probe on an individual Anode pin and the black probe on the Common Cathode pin. A healthy silicon photodiode will read between 0.35 V and 0.55 V. If it reads 0.00 V or near-zero, the junction is shorted (ESD damage).
- Reverse Bias Test: Swap the probes (red on Cathode, black on Anode). The meter should read OL (Over Limit) or an open circuit. If it reads a low voltage or a resistance value, the junction is leaky or shorted.
- Crosstalk/Isolation Test: Keep the black probe on the Common Cathode. Move the red probe sequentially across all individual Anode pins. Every pixel should yield the same 0.35V–0.55V forward drop. If one pixel reads differently, it has suffered localized physical or thermal damage.
Frequently Asked Questions
How do I wire a common-cathode photodiode array to a multiplexer?
Do not multiplex the raw photodiode anodes directly into a single TIA. The parasitic capacitance of the multiplexer's off-channels will sum together at the TIA's inverting node, destroying your bandwidth and causing severe ringing. Instead, use a "TIA-per-pixel" approach where each pixel has its own dedicated op-amp, and then multiplex the voltage outputs of the op-amps using a low-impedance CMOS analog multiplexer like the 74HC4051 or an ADG series switch.
Why is my photodiode array output oscillating in my TIA circuit?
Oscillation in a TIA is almost always caused by insufficient phase margin due to the photodiode's high junction capacitance interacting with the feedback resistor. According to Analog Devices' application notes on transimpedance amplifiers, you must add a feedback capacitor ($C_f$) across $R_f$. Calculate the required $C_f$ using the formula $C_f = \sqrt{(C_d + C_{in}) / (2 \pi R_f f_{GBW})}$, where $C_d$ is the diode capacitance and $f_{GBW}$ is the op-amp's gain-bandwidth product. If you already have $C_f$, try increasing its value by 1 pF or 2 pF; you are likely under-compensating.
Can I use a silicon photodiode array for high-power laser profiling?
Silicon arrays (like the Hamamatsu S-series) will saturate and potentially suffer permanent optical damage (burn-in) if exposed to unattenuated high-power lasers (>10 mW concentrated). You must use calibrated neutral density (ND) filters to drop the optical power into the microwatt range before it hits the sensor window. For profiling high-power CO2 lasers (10.6 μm), silicon is entirely transparent and useless; you must use a pyroelectric array or an InGaAs sensor specifically rated for that wavelength.
What is the difference between a photodiode array and a CMOS linear image sensor?
A raw photodiode array is an analog, passive component. It outputs continuous photocurrent proportional to light, requiring external amplification (like the TIA circuit above) and external analog-to-digital conversion. It has no internal timing, clocking, or readout logic. A CMOS linear image sensor (like the TCD1304 or modern equivalents) integrates the photodiodes with active readout circuitry, correlated double sampling (CDS), and clock drivers on the same die. The CMOS sensor outputs a timed, multiplexed analog voltage sequence. Use a raw array for custom, ultra-low-noise analog front-ends; use a CMOS sensor when you want a plug-and-play digital-readout solution.
