An RL circuit pairs a resistor (R) and an inductor (L) to manage current transients, filter specific frequencies, or shape pulse waveforms. Unlike purely resistive networks, an RL circuit's behavior changes over time and frequency due to the inductor's opposition to changes in current (reactance). The fundamental metric governing this behavior is the time constant, expressed as τ = L / R (in seconds), which dictates how quickly the circuit reaches 63.2% of its final steady-state current when a DC voltage is applied.
While RC (resistor-capacitor) circuits dominate low-frequency hobbyist filtering, RL circuits are indispensable when you need to pass high DC currents without voltage drop, limit di/dt (current slew rate) to protect switching MOSFETs, or filter high-frequency noise out of a DC power rail. Below is a decision-forward guide to designing, sizing, and testing RL topologies on the bench.
Topology and Node Behavior
The most practical configuration for makers and power designers is the Series RL Low-Pass Filter. In this topology, the inductor is placed in series with the signal path, and the resistor is tied from the output node to ground.
- Node 1 (V_in): The input signal (e.g., raw PWM or noisy DC).
- Node 2 (V_mid / V_out): The junction between the inductor and the resistor. This is where you measure your filtered output.
- Node 3 (GND): The ground reference tied to the bottom of the resistor.
In this arrangement, the inductor blocks high-frequency AC components (due to rising inductive reactance, $X_L = 2\pi fL$) while passing DC with minimal loss (limited only by the inductor's internal DC resistance, or DCR). The resistor provides a DC path to ground, establishing the filter's cutoff frequency and damping factor.
Why Choose RL Over RC? (Decision Framework)
The default reflex for filtering is often an RC circuit. However, RC filters suffer from DC voltage drop under load and poor high-current handling. Use the decision tree below to determine if an RL topology is the correct engineering choice for your application.
| Design Requirement | RC Topology | RL Topology | Verdict |
|---|---|---|---|
| Pass high DC current (>500mA) with minimal voltage drop | Poor (Resistor drops voltage proportional to current) | Excellent (Inductor DCR is typically < 1 ohm) | Choose RL |
| Limit inrush current or di/dt to protect semiconductors | Poor (Capacitors draw massive inrush current) | Excellent (Inductors inherently oppose current changes) | Choose RL |
| Filter low-frequency audio (< 1kHz) on a budget | Excellent (Cheap, small components) | Poor (Large inductors are expensive and bulky) | Choose RC |
| Block DC while passing AC (High-pass coupling) | Excellent (Capacitors block DC naturally) | Poor (Inductors pass DC easily) | Choose RC |
Design Walkthrough: Sizing an RL Low-Pass Filter
Let's design a concrete RL filter to smooth a 5kHz, 12V PWM signal from an ESP32 into a quasi-DC analog voltage to drive a 0-10V industrial input. We need a cutoff frequency ($f_c$) of roughly 500Hz (one-tenth of the PWM frequency) to adequately smooth the ripple.
1. The Cutoff Formula:
For an RL low-pass filter, $f_c = \frac{R}{2\pi L}$.
2. Pick the Inductor (L):
We need an inductor that won't saturate at our expected load current (assume 50mA max) and has a low DCR. We select the Bourns 78FR10K-RC, a 10mH radial leaded inductor. It has a DCR of roughly 1.5Ω and a saturation current well above 50mA.
3. Calculate the Resistor (R):
Rearranging the formula: $R = 2\pi \times f_c \times L$
$R = 2 \times 3.14159 \times 500\text{ Hz} \times 0.01\text{ H}$
$R = 31.41\Omega$
4. Select the Real Component:
We choose the closest standard 1% resistor value: 33Ω (e.g., Vishay MRS25000C3303). A 1/4W rating is more than sufficient here since power dissipation is $I^2R = (0.05)^2 \times 33 = 0.0825\text{ W}$.
5. Verify Time Constant (τ):
$\tau = \frac{L}{R} = \frac{0.01}{33} = 303\mu\text{s}$. This means the output will reach 63.2% of a step-change in roughly 0.3ms, and settle to 99% in about $5\tau$ (1.5ms)—fast enough for most industrial control loop updates.
Behavior Matrix: Tuning R and L Values
When prototyping, you will inevitably need to tweak the response. Here is exactly what happens to the circuit's behavior when you alter one variable while holding the other constant.
| Component Change | Cutoff Freq ($f_c$) | Time Constant ($\tau$) | DC Voltage Drop | Physical Trade-off |
|---|---|---|---|---|
| Increase L | Decreases | Increases (slower response) | Increases slightly (higher DCR) | Larger physical footprint, higher cost, risk of core saturation at high currents. |
| Decrease L | Increases | Decreases (faster response) | Decreases | Less ripple attenuation, requires physically smaller but potentially more expensive core materials. |
| Increase R | Increases | Decreases (faster response) | Massive increase under load | Defeats the purpose of the RL topology by introducing severe load-dependent voltage sag. |
| Decrease R | Decreases | Increases (slower response) | Decreases | Requires higher inductance to maintain the same cutoff, risking resonance if parasitic capacitance is high. |
Failure Modes: What Breaks at the Extremes
Understanding how a circuit fails is just as critical as knowing how it works. The failure modes of Series RL and Parallel RL topologies contrast sharply.
Series RL (Low-Pass: L in series, R to ground)
- Inductor Opens: The circuit is broken. $V_{out}$ drops to 0V. This is a safe failure mode.
- Inductor Shorts: The filtering action is entirely defeated. $V_{out}$ becomes equal to $V_{in}$. If $V_{in}$ is raw PWM, the full high-frequency square wave hits your load, potentially destroying sensitive analog inputs.
- Resistor Opens: The DC path to ground is lost. $V_{out}$ will float, relying entirely on the input impedance of whatever load is connected. High-impedance loads will read $V_{in}$; low-impedance loads will drag the voltage down unpredictably.
- Resistor Shorts: $V_{out}$ is pulled directly to ground (0V). The inductor will experience a massive current spike limited only by the source impedance and the inductor's DCR, likely blowing the upstream fuse or frying the driving microcontroller pin.
Parallel RL (R and L in parallel, driven by a current source or series source resistor)
- Inductor Opens: All current is forced through the parallel resistor. The voltage across the network spikes to $I_{source} \times R$, which can easily exceed the voltage rating of downstream components.
- Inductor Shorts: The inductor acts as a dead short across the source. In a voltage-driven system, this draws infinite current (limited only by wiring resistance), causing catastrophic thermal failure or tripping the breaker immediately.
Bench Testing: Step-by-Step Verification
Do not trust the datasheet blindly; parasitic capacitance and core nonlinearities alter real-world behavior. Here is how to verify your 500Hz RL design on the bench using an oscilloscope and a function generator.
- Wire the Circuit: Place the 10mH Bourns inductor in series on a breadboard. Connect the 33Ω Vishay resistor from the inductor's output leg to the breadboard ground rail.
- Connect the Source: Attach your function generator's output to the input leg of the inductor. Set it to output a 0V to 3.3V square wave at 5kHz (simulating the ESP32 PWM).
- Probe the Output: Connect your oscilloscope probe across the 33Ω resistor (Node 2 to GND). Ensure the probe is set to 1X or 10X consistently with your scope channel settings.
- Verify Time Constant (Step Response): Switch the function generator to a 10Hz square wave (0V to 3.3V) to allow the circuit to fully settle. Trigger the scope on the rising edge. Measure the time it takes for the waveform to reach 2.08V (63.2% of 3.3V). It should read approximately 303µs.
- Verify Cutoff Frequency (AC Sweep): Switch the function generator to a sine wave. Set the amplitude to 1V peak-to-peak. Sweep the frequency upward from 100Hz. Note the frequency where the output amplitude drops to 0.707V (-3dB). For our design, this should occur right around 500Hz. If it occurs at 450Hz, your inductor's actual tolerance is likely +10% (common for ferrite core parts).
The Default Recommendation
If you are building a generic PWM-to-DC smoothing filter or a basic di/dt limiter for a microcontroller-driven load under 100mA, stop deliberating and use this exact bill of materials:
- Inductor: 10mH Radial Leaded (Bourns 78FR10K-RC or equivalent). It offers a low 1.5Ω DCR, fits easily on a 0.1-inch breadboard, and costs under $0.50 in single quantities.
- Resistor: 33Ω, 1/4W, 1% Metal Film (Vishay MRS25 series). The metal film construction ensures low thermal noise, which is critical if this filtered signal is feeding into an ADC.
This specific pairing yields a reliable 483Hz cutoff frequency and a 303µs time constant, providing excellent 5kHz PWM rejection without introducing the severe DC voltage drop inherent to RC alternatives. For deeper theoretical modeling of the transient responses discussed here, refer to the foundational texts on RL series circuit dynamics and AC impedance calculations.
