Power regulator circuits, both linear and switching regulators, include feedback loops as mechanisms to ensure the circuit outputs at a target voltage. These feedback loops play a simple but important role in power regulator circuits: they dynamically adjust the output voltage so that the target value is always reached. This dynamic adjustment occurs within some operating limits in the particular regulator circuit.
To see how power regulators respond to transients in dynamic loads, we’ve outlined how feedback loops work in these components. PMICs and small power regulators will have these circuits built into the semiconductor die, which will limit the transient response capabilities of these circuits. The external passives (capacitors and inductors) on the output of a regulator also impact the operation of a regulator–we’ll examine all of these points in this article.
How Voltage Regulation Feedback Works
If you have ever looked at a voltage regulator, you may have noticed a feedback pin on the device pinout. This will commonly be found on a PMIC or a smaller regulator that outputs at or near logic levels. This will also be found on adjustable regulators (both LDOs and switching regulators), and these pins can be found in multiple topologies. The feedback pin plays a simple role in voltage regulation: it allows the output to be measured and adjusted by the IC.
Voltage regulation through the use of a feedback loop allows a voltage regulator to perform two very important tasks:
- Adjust the output voltage if the input voltage suddenly changes
- Adjust the output voltage if the load is dynamic
- Adjust the output voltage to respond to transients on the power rail
Given these three important tasks in power regulation, a feedback loop has to be designed to stabilize the voltage output from a power regulator circuit.
Voltage Regulation Feedback Circuits
The typical implementation of a feedback loop in a voltage regulator circuit involves a negative feedback error amplifier. The error amplifier allows a comparison between the output voltage and a reference voltage by passing the output voltage across a voltage divider. The voltage divider steps down the output to a lower value, and the stepped-down voltage is compared with the reference voltage with the error amplifier.
Example LDO feedback loop.
The reference in this case is a silicon bandgap reference that outputs at 1.2 V. When the regulator is designed, the two resistors in the voltage divider will step down the voltage so that it attempts to match the precision voltage reference value (about 1.2 V). Because the inputs to an error amplifier have high impedance, the feedback loop will consume very little current, so this allows for highly efficient regulation.
With this type of circuit, if the output voltage changes, then the stepped down voltage will deviate from the target reference value. The output from the error amp then modulates the transistor to adjust the output voltage up or down so that it is restored to the target output value.
In a switching regulator, we might have a two-stage amplifier circuit for regulation: one for sensing the output, and the other for modulating the regulator’s duty cycle. An example feedback loop circuit for a switching regulator is shown below.
Example switching regulator feedback loop.
The above example is probably the simplest case of a regulation circuit where the output is adjusted by modulating a duty cycle. Here, a ramp voltage is used with a positive feedback comparator. The design space gets much more complex than this when other features are added (sync, safety shutdown, etc.). The ultimate point of this type of regulator is to compare the error amp output to the ramp voltage; as the error amp output moves higher, the time where V(ramp) > V(error) is smaller, which increases the converter’s duty cycle, and vice versa.
High Power Requires Discrete Feedback Loops
Voltage regulators that must provide much higher voltages and currents, such as 10’s of amps at high voltages, will not use integrated circuits that include feedback for regulation. The problem is that these systems require components that are simply too large to fit on a semiconductor die. The other challenge is that these systems may be isolated with high output voltage and current, which requires a transformer to implement galvanic isolation.
Instead of using integrated feedback loops, these systems require a set of discrete components to implement feedback and voltage regulation. The output will typically be fed back to the input side with a current/voltage measurement using a precision resistor. Then this value is fed to a gate drive controller, which will control the PWM signal in the switching stage of the power regulator. The PWM adjustment here then adjusts the inductor/transformer current, and finally the output voltage.
Feedback loop in a high-power switching regulator
For high power systems, the challenge in laying out these systems is to ensure stability in the feedback loop. This generally means we want higher capacitance and lower inductance in the layout, requiring tight placement and routing of components along the current path. Simulations can be used to simulate the allowed inductance by including it as a parasitic element along the current path.
Whenever you want to simulate the effectiveness of your voltage regulator feedback loop and parasitics in your circuits, use the comprehensive set of simulation tools in PSpice from Cadence. PSpice users can access a powerful SPICE simulator as well as specialty design capabilities like model creation, graphing and analysis tools, and much more.