Bias current is the small amount of current that flows into the input terminals of an operational amplifier (op-amp) just to keep its internal transistors working. It’s typically measured in nanoamps or even smaller units, and while it sounds insignificant, it can introduce real errors in sensitive circuits. The formal definition is the average of the currents flowing into the two input terminals when the output is at a specified level.
Why Op-Amps Need Bias Current
An op-amp’s input stage is built from transistors, and transistors need a small amount of current to operate. In designs using bipolar junction transistors (BJTs), the base of each transistor draws current from whatever is connected to the input pin. This isn’t a design flaw. It’s a fundamental requirement of how the transistor amplifies a signal. The current is small, but it’s always there.
Op-amps that use field-effect transistors (FETs) at their inputs draw far less bias current because FETs are voltage-controlled rather than current-controlled. The tradeoff is that FET-input bias current is much more sensitive to temperature, roughly doubling for every 10°C increase. A FET-input op-amp that draws 1 picoamp at room temperature could draw significantly more in a warm enclosure.
How Bias Current Is Calculated
Every op-amp has two inputs: the non-inverting (+) and the inverting (−). Each draws its own bias current, and these two currents are rarely identical. The published “input bias current” spec on a datasheet is the average of the two:
IB = (IN + IP) / 2
The difference between the two input currents gets its own specification called “input offset current.” For a classic 741 op-amp, the bias current is about 80 nA, and the offset current is roughly a tenth of that, around 10 nA. The offset current matters because it represents the mismatch you can’t fully cancel out with circuit design tricks.
Typical Values Across Op-Amp Types
The range of bias currents across different op-amp families is enormous. At the low end, electrometer-grade FET-input amplifiers like the AD549 draw around 60 femtoamps, which works out to roughly one electron crossing the input every three microseconds. At the high end, some high-speed bipolar op-amps draw tens of microamps. That’s a span of about ten billion to one.
As a rough guide:
- Standard bipolar op-amps (like the 741): tens to hundreds of nanoamps
- Bias-compensated bipolar op-amps: 0.5 to 10 nA
- JFET-input op-amps: picoamps at room temperature
- Electrometer-grade MOSFET/JFET op-amps: femtoamps
How Bias Current Creates Errors
Bias current becomes a problem when it flows through resistance at the op-amp’s input. Any current through a resistor creates a voltage (Ohm’s law), and that unwanted voltage gets amplified along with your actual signal. The larger the resistance connected to the input, the bigger the error.
With low source resistance, the effect is negligible. For instance, 2 nA of bias current flowing through a 10-ohm resistor produces only 0.02 microvolts of error, which is invisible in most applications. But connect that same op-amp to a sensor with a 10-megohm output impedance, and the error voltage jumps to 20 millivolts. In a precision measurement, that’s enough to ruin your data.
Why It Matters for High-Impedance Sensors
Sensors that measure pH, light intensity, acceleration, and humidity often have very high output impedances, sometimes in the gigaohm range. A pH probe, for example, presents such a high impedance that even picoamps of bias current can shift the measured voltage enough to throw off a reading by a meaningful amount. Photodiode amplifiers face the same challenge: the tiny currents generated by the photodiode can be overwhelmed by the op-amp’s own bias current if you choose the wrong amplifier.
For these applications, selecting an op-amp with the lowest possible bias current is one of the most important design decisions. FET-input or electrometer-grade amplifiers are the standard choice, though designers also need to account for how the bias current will grow as the circuit warms up during operation.
Compensating for Bias Current
The most common compensation technique is adding a resistor to the non-inverting input that matches the resistance seen by the inverting input. The idea is simple: if both inputs see the same resistance, and the bias currents are roughly equal, then the error voltages at both inputs will be nearly identical. Since the op-amp amplifies the difference between its inputs, equal errors on both sides cancel out.
This technique works well when the two input bias currents are close to each other. The cancellation is limited by the input offset current, which is the mismatch between the two. For the 741 example, resistor matching would reduce the effective error from one based on 80 nA (the full bias current) to one based on 10 nA (the offset current). That’s an eight-fold improvement with a single added component.
Some modern op-amps include internal bias current cancellation circuits that inject compensating currents at the input stage. These devices can achieve bias currents well below 10 nA even with bipolar input transistors. The downside is that internally compensated bias currents can be less predictable in direction, since the cancellation circuit may overcompensate or undercompensate slightly. This means the bias current might flow into or out of the input terminal depending on temperature and manufacturing variation.
Temperature Effects
Bias current is not a fixed number. It shifts with temperature, and the direction of that shift depends on the input transistor type. BJT-input op-amps generally see bias current decrease as temperature rises, because the transistor’s current gain increases with temperature, meaning less base current is needed. FET-input op-amps behave in the opposite direction: their bias current doubles roughly every 10°C. Some FET-input datasheets specify bias current only after five minutes of operation at 25°C, because the device needs to reach thermal equilibrium before the spec is valid.
If your circuit operates in a temperature-controlled environment, this may not matter. But for equipment exposed to outdoor conditions or packed tightly in a warm chassis, the temperature coefficient of bias current can be the dominant source of measurement drift.