Introduction to Phototransistors
In this series of articles, we’ll explore higher-output-current alternatives to photodiodes.
In my experience, light-detection and light-measurement applications most commonly employ photodiodes. But there’s no getting around the fact that photodiodes produce very small output currents, and this can lead to design challenges that you might, in certain situations, prefer to avoid.
This article and the following article provide some basic information on two types of light-sensitive devices that produce higher output current than photodiodes: phototransistors and photosensitive ICs. The latter term refers to devices that are essentially a photodiode and an amplifier integrated into the same package.
What Is a Phototransistor?
A photodiode can generate photocurrent because its junction is exposed to incident light. A phototransistor functions in a similar way, except that the exposed semiconductor material is the base of a bipolar junction transistor (BJT).
A phototransistor is depicted as a BJT with the base terminal removed, and the arrows imply that the base is sensitive to light. The other diagrams in this article depict only NPN phototransistors.
There are two ways to think about the behavior of a phototransistor.
First, you can mentally replace the amount of current flowing into the base of a normal transistor with the intensity of incident light. In the basic model of active-mode BJT behavior, the output current (i.e., the collector current) is the input current (i.e., the base current) multiplied by the gain parameter called beta (β). With a phototransistor, incident light is like a weak signal applied to the base, and the output current is much higher than what we would expect from a photodiode, because of the transistor’s ability to internally amplify the signal applied to the base.
Second, you can imagine that a phototransistor is a BJT with a photodiode connected to the base, such that the input signal to the transistor is the photocurrent generated by the photodiode. In this conceptualization, the BJT is like an additional semiconductor device that applies current gain to the output signal of a photodiode.
A phototransistor is conceptually equivalent to a photodiode that drives the base of a bipolar junction transistor. Note the orientation of the photodiode: photocurrent is always reverse current, and the photodiode is oriented such that photocurrent is flowing into the base.
As with photodiodes, the goal with phototransistors is to produce a usable output voltage from the light-generated current. Since phototransistors have amplification built into their semiconductor structure, we don’t need an op-amp-based transimpedance amplifier (TIA). Instead, we can use amplifier configurations that we already know from non-light-sensitive BJT applications.
The common-collector and common-emitter configurations are both viable options for converting light into voltage. I prefer the common-emitter approach, because I find it more intuitive, but you might enjoy the common-collector amp if you prefer to avoid inversion—i.e., if you want higher illuminance to produce higher output voltage.
You can use the common-collector or the common-emitter amplifier configuration to turn your phototransistor into an illuminance-to-voltage converter.
Phototransistors vs. Photodiodes
Phototransistors might seem like a major improvement over photodiodes, but they’re not as popular as you might think. The internal current amplification is in theory an important advantage, but there are many resources out there to help engineers design high-performance TIAs, and I for one prefer the TIA approach.
Furthermore, phototransistors are inferior in important ways.
- Phototransistors have less ability to maintain a linear relationship between illuminance and output current. This isn’t important if all you need is an on/off light detector that produces a digital output voltage. My photosensitive applications tend to require analog output signals, and I instinctively disparage phototransistors for their limited linearity.
- Photodiodes achieve faster response than phototransistors. The importance of wide bandwidth is dependent on the requirements of the application, and in many cases a phototransistor will be perfectly adequate. At the same time, you don’t want to design your system around a phototransistor and then be forced to overhaul the design a year later when someone wants to increase the maximum operational frequency by an order of magnitude.
- Important performance specs are more sensitive to temperature in phototransistors than they are in photodiodes. This is a non-issue if your product will always operate at room temperature. If you work with automotive or military systems, a phototransistor’s temperature-induced performance variations could cause headaches.
Phototransistors offer higher light-generated output current while imposing some performance restrictions. I favor photodiodes; nevertheless, there must be quite a few applications in which it makes sense to use a phototransistor and then eliminate the cost and complexity of a TIA.
I used to teach electronics and always started photo-transistors with taking the end off a BC107 (or any) with wire snips and connecting an oscilloscope to the legs….... firstly it showed that the principal is easy (all bipolar transistors show photo sensitivity) and the effect of light on the junction….... as the exposed transistor junction gets near the ‘scope trace, the amplitude increases and the trace moves away from the transistor junction.
Very interesting article.