A current mirror is an electronic circuit that copies a reference current into one or more other branches. It does this by using matched transistors so that a voltage set by the reference branch makes the output transistor conduct approximately the same current.
Why current mirrors are used
Many circuits need a predictable current rather than a fixed voltage. A current mirror provides that current using a reference source and transistors, often within an integrated circuit. It is commonly used to bias amplifier stages, set currents in analog signal-processing circuits, and supply several branches from one reference.
The word “mirror” describes the intended relationship between currents, not a guarantee of exact duplication. Real devices differ slightly, and the output current changes somewhat with voltage, temperature, and transistor characteristics. A mirror is therefore best understood as a practical current-setting circuit with defined limits.
The basic two-transistor circuit
A simple mirror uses two similar transistors. In a common NPN implementation, the emitter terminals connect to the lower supply rail. The reference transistor has its collector connected to its base, and the second transistor’s base connects to that same node. This shared base voltage is the key to the circuit.
A reference current flows into the diode-connected transistor. It establishes the base-emitter voltage needed for that transistor to conduct. Because the second transistor has the same base-emitter voltage and, ideally, the same characteristics, it conducts a similar collector current. The output current is drawn from the output branch through the second transistor.
For a simple circuit, the reference current is often set with a resistor connected to a supply. The resistor’s voltage drop, together with the transistor’s base-emitter voltage, determines the current approximately. The precise calculation depends on the supply voltage, resistor value, transistor behavior, and circuit connections; it is not simply the supply voltage divided by the resistor.
How the circuit establishes a copy
The diode-connected transistor converts reference current into a voltage: its base and collector are joined, so the current establishes a base-emitter voltage. That voltage is also applied across the second transistor’s base-emitter junction. If the devices are sufficiently alike and operate under suitable conditions, the second device then draws a similar collector current.
This is a feedback-like setting mechanism, though the simplest mirror is not usually described as a complete amplifier feedback loop. If the reference current rises, the shared base voltage rises until the reference transistor carries the increased current. The output transistor receives the resulting base voltage and responds by increasing its current as well.
The output is not forced to remain identical under every condition. The output transistor needs enough voltage between its collector and emitter to stay in its active operating region. If the output voltage falls too far, the transistor enters saturation and can no longer regulate current in the intended way. The minimum usable output voltage is often called the compliance voltage or voltage headroom.
Setting the reference current
The reference branch determines the scale of the mirrored current. In a basic resistor-fed NPN arrangement, current through the resistor supplies both the reference transistor and the base currents of the mirror devices. Those base currents are a source of error, especially when transistor current gain is limited or several output branches are connected.
When estimating the reference current, account for the voltage across the resistor rather than assuming the full supply appears across it. The diode-connected transistor holds the resistor’s lower end near a base-emitter voltage above the lower rail. As a result, the resistor current is approximately the supply-to-reference-node voltage difference divided by the resistance. A real design also checks transistor limits, resistor tolerance, supply variation, and temperature.
For example, if a circuit needs a small bias current for an amplifier stage, a resistor can set the reference branch and a matched transistor can provide the output current. The mirror reduces the need for a separate current-setting resistor in every branch. But if the output must remain accurate over a wide range of supply or temperature, a more carefully designed reference and mirror may be necessary.
Current gain and multiple outputs
A mirror can feed more than one output branch by connecting additional transistor bases to the shared reference node. Each output transistor then attempts to draw the reference current, assuming the devices are matched and each has adequate voltage headroom. Connecting several outputs does not make the reference current disappear: the reference transistor must also supply the combined base currents.
Mirrors can also scale current. If an output transistor has a larger effective emitter area than the reference transistor, it can produce a proportionally larger current under suitable conditions. A smaller area can produce a smaller current. Integrated circuits commonly use transistor geometry to establish such ratios, though layout, matching, and operating conditions affect accuracy.
- One-to-one mirror: the reference and output devices have similar effective sizes, so their currents are intended to be similar.
- Scaled mirror: device sizes differ to produce an intended current ratio.
- Multiple-output mirror: several output devices share a reference node to serve separate circuit branches.
Accuracy limits and common problems
The simple explanation assumes identical transistors, negligible base current, and equal collector voltages. Actual circuits depart from those assumptions. Discrete transistors may not match closely, while even matched devices can carry different currents if their collector-emitter voltages differ. This voltage-related difference is associated with the Early effect in bipolar transistors.
Other sources of error include resistor tolerance, temperature differences between devices, leakage, and insufficient voltage headroom. In an integrated circuit, placing matched transistors close together can help them experience similar temperatures and process conditions. In a circuit assembled from separate components, thermal and device mismatches are generally harder to control.
Common troubleshooting checks include:
- Confirm that the transistor type and pin connections match the schematic; package pinouts are not universal.
- Check that the reference path actually carries current and that the shared base node is connected correctly.
- Measure whether the output transistor has enough voltage across it to operate as a current source.
- Check resistor values, supply polarity, and the expected transistor operating region.
- Consider base-current loading when adding multiple outputs or using low-gain devices.
A frequent misconception is that a current mirror produces perfectly constant current. It does not. It approximates a current source over a finite range of output voltage, and its accuracy depends on the circuit and components. Another misconception is that the output current is determined by the output load alone; in the intended operating region, the reference current and mirror ratio set it, while the load voltage must stay within the circuit’s compliance range.
Common mirror designs and alternatives
The two-transistor mirror is easy to understand but has limited accuracy and output resistance. More advanced designs add transistors to improve matching behavior or reduce the effect of unequal transistor voltages. A cascode mirror, for instance, uses additional devices to keep key transistor voltages more nearly controlled. It can improve current stability as output voltage changes, but it needs more voltage headroom and uses more components.
MOSFET current mirrors use a related principle. A reference current sets the gate-source voltage of a diode-connected MOSFET, and that voltage controls another MOSFET with a shared gate connection. The current ratio depends strongly on device geometry and operating conditions. MOSFET mirrors are widespread in integrated circuits, while bipolar mirrors remain useful in other circuit contexts.
Alternatives include a resistor-fed branch, a dedicated current-source circuit, or an operational-amplifier-based current regulator. A resistor is simpler, but its current changes as the voltage across it changes. A more elaborate current source can offer better regulation, but may require more components, power, or voltage headroom. The best choice depends on the required accuracy, available supply voltage, load range, and circuit complexity.
How to choose a current mirror
Start by defining the current the circuit needs and the range of voltage available at its output. Then consider how much current variation the application can tolerate. A basic mirror may suit a bias branch where moderate variation is acceptable; a precision reference or carefully matched design is more appropriate when errors directly affect performance.
- Determine the target current and whether outputs need equal or scaled values.
- Check the minimum and maximum output voltage, including the mirror’s required headroom.
- Allow for transistor mismatch, base or gate effects, resistor tolerance, and temperature.
- Check the devices’ current, voltage, and power limits against the circuit conditions.
- Decide whether a simple mirror, a higher-performance topology, or another current source is more suitable.
For a real circuit, use the device data sheet and the circuit’s operating conditions rather than relying on a single ideal equation. Simulation can help reveal headroom and tolerance issues, but it depends on the accuracy of the device models and should not replace checking the actual circuit constraints.
Frequently Asked Questions
Does a current mirror keep the current exactly constant?
No. It keeps current approximately constant over an allowed output-voltage range. Device mismatch, temperature, transistor voltage effects, and limited headroom cause variation.
Why does a current mirror need a reference current?
The reference current establishes the voltage that controls the transistor pair. The output device uses that shared control voltage to produce a corresponding current.
Can a current mirror increase current?
Yes. A mirror can be designed for a larger output current by using an output transistor with a larger effective size than the reference device. The ratio is approximate in real circuits.
What happens if the output voltage is too low?
The output transistor may leave its normal current-regulating region and enter saturation. The current then depends more strongly on the circuit voltage and load.
Is a current mirror the same as a current source?
A mirror is one way to create a current-source-like output. It copies or scales a reference current, while other current sources may use different circuit principles.
In summary
A current mirror uses transistor behavior and a shared control voltage to reproduce a reference current in another branch. Its simplicity makes it a basic building block in analog electronics, but its output is only an approximation: matching, current ratio, operating voltage, temperature, and device limits all matter when designing or troubleshooting one.