Design Tips for Photodiode Amplifiers
This article covers important details related to the design of transimpedance amplifiers for photodiode-based systems.
If you’re still somewhat new to the functionality and implementation of photodiodes, I definitely recommend that you read (or at least skim!) my Introduction to Photodiodes series:
- The Nature of Light and pn Junctions
- The Physical Operation of Light-Sensitive pn Junctions
- Understanding Photovoltaic and Photoconductive Modes of Photodiode Operation
- Characteristics of Different Photodiode Technologies
- Understanding the Photodiode Equivalent Circuit
How to Amplify a Photodiode Signal
The standard method of amplifying the very-low-amplitude current generated by a photodiode is to use this current as the input to an op-amp-based transimpedance amplifier (TIA). The following diagram provides an example of a photodiode connected to a TIA; the photodiode has zero voltage bias, which means that the photodiode is operating in photovoltaic mode.
Figure 1. A photodiode connected to a transimpedance amplifier
For more information on transimpedance amplifiers, please refer to AAC’s video tutorial on this topic.
Maintaining Stability in Photodiode Circuits
In the circuit shown above, only the resistor (RF) provides gain. The purpose of the capacitor (CF) is to avoid oscillation problems by compensating for the photodiode’s internal junction capacitance, which creates a pole in the feedback network. CF compensates by creating a zero in the feedback network.
Oscillation is a very real problem with photodiode circuits. It’s true that internal frequency compensation usually protects op-amp implementations from instability, but photodiode TIAs can oscillate even when you use an internally compensated op-amp.
Incorporating a DC Offset
In some situations, you may want to use a photodiode to record the waveform produced by a specific type of short-duration optical or thermal event. You can use AC coupling to eliminate the effect of ambient radiation and thereby allow the system to detect only the transient illumination, but the falling edge of the waveform may extend below ground.
This could be problematic in a single-supply system: if the op-amp’s negative supply is grounded, the portion of the waveform that extends below 0 V will be clipped.
You can remedy this by applying a small DC voltage, call it VOFFSET, to the op-amp’s non-inverting input terminal; VOFFSET will become the output level produced by the amplifier in the absence of an input signal. The falling edge of the waveform will be able to extend below this voltage, and after the transient event, the amplified output will eventually return to VOFFSET.
Figure 2. The same photodiode connected to a transimpedance amplifier as in Figure 1, but with a DC offset.
In this example, I’m using a resistive divider to generate a suitable offset voltage. The parallel capacitor helps to suppress high-frequency noise originating from the power supply.
Your choice of offset voltage will depend on the application. You don’t want to make VOFFSET any larger than necessary: if the offset is 500 mV but your input waveforms never extend more than 200 mV below ground, you’ve lost 300 mV of signal swing that may be needed for the positive portion of the waveform.
Remember that the voltage applied to the non-inverting input terminal will, thanks to the virtual short, also appear at the inverting input terminal. This means that a positive offset voltage will cause the photodiode to have a reverse bias. The effect of reverse bias on photodiode operation is discussed in Part 3 of Introduction to Photodiodes.
Even if you’re not determined to preserve the below-ground portion of the waveform, you should consider including a small (maybe 100 mV) offset voltage if you’re designing a single-supply system, because it prevents the op-amp from saturating at the negative rail.
Saturation is not catastrophic, but op-amps (unlike comparators) are not optimized for producing output voltages at the supply rails. A saturated op-amp needs some time to come out of saturation; thus, a TIA that is saturated at the negative rail will exhibit some delay when responding to an input signal.
We’ve taken a closer look at transimpedance amplifiers for photodiodes, with the discussion touching on stability, DC offsets for waveform preservation, and DC offsets for preventing op-amp saturation. There’s much more to say on photodiode TIAs, and we’ll continue this topic in the next article.