Phototransistor

A phototransistor is a bipolar transistor whose base-collector junction is exposed to light, so that photons absorbed near the junction generate electron-hole pairs and the resulting photocurrent is amplified by the transistor's current gain. In practice it is a light sensor with built-in gain: a package with a clear window or lens, usually just two leads because the base is left unconnected, that turns illumination directly into a much larger collector current than a bare photodiode of similar size could supply.

The gain mechanism is the entire point, and it works by treating light as a base current. Photons with enough energy to cross silicon's band gap of about 1.1 electron volts, which means visible and near-infrared light up to roughly 1100 nanometres, free carriers in the reverse-biased base-collector depletion region. The junction's electric field sweeps them apart, and this photocurrent flows into the base exactly as an injected base current would. The transistor then multiplies it by its current gain β, so the collector current is roughly (β + 1) times the photocurrent. A photodiode delivers the photocurrent itself; a phototransistor delivers a hundred times more or so from the same light, which is why it can drive a relay-sized load or a logic input with little extra circuitry.

Light reaches the base-collector junction
through a window or lens in the package
Photons create electron-hole pairs
in and near the depletion region
Photocurrent flows into the base
acts like an injected base current
Trade-offs of the gain
Larger output
collector current is about β times the photocurrent
Slower response
large base-collector capacitance with a high-gain stage
Less linear
gain varies with current and temperature
Used in optocouplers, slotted switches and reflective sensors
where on-off detection matters more than speed or precision

The price of that gain is speed and linearity, so a phototransistor is the wrong tool where a photodiode is the right one. The large base-collector junction area that collects light also has a large capacitance, and the Miller effect of the gain stage multiplies its apparent size, so response times typically lie in the range of microseconds, compared with nanoseconds for a fast PIN photodiode. Output current also varies with temperature and does not track light intensity as cleanly. Photodiodes are therefore preferred for fiber-optic receivers, precision light meters and fast pulse detection, while the phototransistor is favoured for simple presence detection and low-speed signalling where its larger signal and low cost outweigh these limits. Designers can open up the base lead, add a resistor to it, or use a Darlington arrangement to trade gain against speed or leakage.

The optocoupler, or optoisolator, is the phototransistor's best-known application. An infrared light-emitting diode and a phototransistor sit in one package, separated by a transparent insulating barrier, so an electrical signal crosses between two circuits by light with no conducting path. This lets a microcontroller sense mains-referenced signals, break ground loops, and drive power stages while surviving voltage spikes on the far side, with isolation ratings commonly in the range of several thousand volts. The key specification is the current transfer ratio, the output collector current divided by the input LED current, expressed as a percentage and often from tens to a few hundred percent for phototransistor types, and it declines as the LED ages. Slotted optical switches and reflective sensors in printers, encoders and proximity detectors work on the same principle.

Phototransistors are often confused with solar cells, but the two are optimized for opposite jobs. A solar cell is a large-area photodiode run in the power quadrant to deliver electrical energy, as in the silicon cell announced at Bell Labs in 1954, and has no transistor gain. A phototransistor is a small-signal detector biased as an amplifier, wasting no effort on harvesting power. A further trap is the assumption that the part responds to light alone: a standard transistor in an opaque case is also slightly light-sensitive, which is why early glass-encased germanium transistors were painted black, and dark current and temperature effects remain real.

Modern designs still use phototransistors, though many applications have moved to integrated sensors. Ambient light sensors and proximity modules now typically combine a photodiode with on-chip amplifiers and converters, and fast optocouplers use photodiodes with logic-level output stages. Even so, the single phototransistor remains a robust, inexpensive part in industrial sensing, remote-control receivers and relay-style isolation, a reminder that a transistor can listen to light as easily as to a voltage.

Read phototransistor as a photodiode wired into a transistor's base: light becomes the base current, gain makes it large, and the extra capacitance that comes with the gain is what makes it slower than a bare photodiode.

Take phototransistor further

Ask the copilot about this term, or have our engineers assess it against your process.