Putting inductors in series adds their inductance values ($L_{total} = L_1 + L_2$) while summing their DC resistance (DCR) and restricting the maximum current to the rating of the weakest link. You use this topology when you need to hit a non-standard inductance value, distribute heat across multiple components, or cascade different core materials to filter both low-frequency ripple and high-frequency EMI.

This guide skips the abstract textbook theory and goes straight to the bench. We will map the node topology, analyze failure modes, and walk through a real-world buck converter output filter design using off-the-shelf components.

The Series Topology: Node Labels and Core Math

In a standard series configuration, the current has only one path through the magnetic components. Let us define the nodes for a two-inductor chain:

  • Node A (Source): The input voltage or upstream circuit node.
  • L1: The first inductor.
  • Node B (Junction): The electrical connection between L1 and L2. In a properly designed PCB, this node should have minimal parasitic capacitance to ground.
  • L2: The second inductor.
  • Node C (Load): The output node feeding the downstream circuit or capacitor bank.

Assuming the magnetic fields of L1 and L2 do not interact (zero mutual inductance), the governing equations are straightforward:

  • Total Inductance: $L_{eq} = L_1 + L_2 + L_3 ... + L_n$
  • Total DCR: $DCR_{eq} = DCR_1 + DCR_2 + DCR_3 ... + DCR_n$
  • Current Rating: $I_{max} = \min(I_{sat1}, I_{sat2}, I_{rms1}, I_{rms2})$
Bench Tip: Mutual inductance will ruin your math. If L1 and L2 are placed too close together or aligned on the same magnetic axis, their fields will couple. If the fields aid each other, $L_{eq}$ spikes; if they oppose, $L_{eq}$ drops. Always place series inductors at least one component-width apart, or rotate one by 90 degrees (orthogonal placement) to nullify coupling.

Element Change Behavior Matrix

When designing multi-stage filters, you need to know how the circuit reacts if one component drifts or is swapped. Here is how the overall topology behaves when a single element in the series chain changes.

Parameter Changed in L1 Effect on Total Inductance ($L_{eq}$) Effect on Total DCR Effect on Circuit Behavior
Inductance increases Increases linearly Usually increases (more wire turns) Lowers filter cutoff frequency; increases transient response time.
DCR increases No direct change Increases linearly Increases $I^2R$ heat loss; drops output voltage under heavy DC load.
Core saturates Drops to near zero (air core) No change $L_{eq}$ collapses to just L2; massive current spikes and high-frequency noise pass through.
Physical size increases Depends on winding Usually decreases (thicker wire) Improves thermal mass and current handling, but eats PCB real estate.

Series vs. Parallel: Why Choose Series?

Why put inductors in series instead of parallel? Parallel inductors divide the current but reduce the total inductance ($1/L_{eq} = 1/L_1 + 1/L_2$). Series inductors do the exact opposite. Here is the decision contrast.

Criteria Inductors in Series Inductors in Parallel
Primary Goal Increase total inductance or cascade filter stages. Increase current handling capacity.
Total Inductance Increases ($L_1 + L_2$). Decreases (halves if identical).
Current Handling Limited by the weakest component. Shared between components (if DCR is matched).
PCB Footprint Larger (components are daisy-chained). Larger (components are side-by-side).
Best Use Case Multi-stage LC filters, EMI chokes, hitting non-standard E-series values. High-current POL (Point of Load) regulators where single inductors overheat.

A highly practical reason to choose series is material cascading. A single powdered-iron inductor is great for low-frequency switching ripple (e.g., 500 kHz) but becomes transparent to high-frequency EMI (e.g., 50 MHz). By putting a small ferrite bead (which acts as a lossy, high-frequency inductor) in series with a large powdered-iron choke, you create a broadband filter that a single component cannot achieve.

Failure Modes: What Breaks at the Extremes?

Understanding how the topology fails is critical for safety and reliability, especially in power supplies.

The Open Circuit Failure

If the internal wire of L1 snaps (open circuit), the entire series path is broken. Current drops to zero. In a buck converter, this means the output voltage collapses to zero, and the load loses power. This is a safe failure mode for the load, though it causes a system outage.

The Short Circuit Failure

If the enamel insulation inside L1 breaks down and the windings short together, L1’s inductance drops dramatically, but its DCR also drops. The total inductance becomes roughly just $L_2$. If your downstream circuit relied on $L_{eq}$ to limit ripple current, that ripple will now spike, potentially overheating output capacitors or triggering overcurrent protection.

Contrast with Parallel Failure

In a parallel configuration, if one inductor fails open, the other inductor is forced to carry 100% of the load current. If it was only sized for 50%, it will rapidly overheat, saturate, and potentially catch fire (thermal runaway). Series configurations do not suffer from this specific cascading thermal failure; an open in series simply kills the circuit, while a short in series just degrades filtering performance.

Design Walkthrough: Sizing a 4.4µH Buck Output Filter

Let us design a real output filter for a 12V-to-5V buck converter delivering 3A. The controller datasheet calls for a 4.4µH inductor. However, 4.4µH is not a standard E-series value, and the closest standard parts (4.7µH) push the physical footprint too large for our dense PCB.

We decide to use two standard 2.2µH inductors in series.

Component Selection

We select the Coilcraft XEL4020-222. Here are the datasheet specs for one unit:

  • Inductance: 2.2 µH
  • $I_{sat}$ (Saturation Current): 6.8 A
  • $I_{rms}$ (Thermal Current): 5.5 A
  • DCR (Max): 13.6 mΩ

The Math

Placing two XEL4020-222 units in series yields:

  • $L_{eq}$: $2.2\mu H + 2.2\mu H = 4.4\mu H$ (Exact target hit).
  • $I_{max}$: $\min(6.8A, 5.5A) = 5.5A$. Our load is 3A, so we have a comfortable 83% margin.
  • $DCR_{eq}$: $13.6m\Omega + 13.6m\Omega = 27.2m\Omega$.

Thermal Verification

Will the DCR cause excessive heat? We calculate the $I^2R$ loss at full load:

$P_{loss} = I^2 \times DCR_{eq} = (3A)^2 \times 0.0272\Omega = 244.8 mW$.

Splitting this 245 mW across two physical components means each inductor dissipates only ~122 mW. This keeps the component temperature well within the safe operating area, proving that the series topology successfully distributes the thermal load while hitting a non-standard inductance target. For deeper thermal and AC loss simulation, you can model this exact setup using the Würth Elektronik REDEXPERT tool or similar SPICE environments.

Breadboard and Bench Testing Step-by-Step

Before committing to a PCB layout, verify your series inductor math on the bench. You will need an LCR meter (like the DER EE DE-5000 or Keysight U1733C) and a prototyping board.

  1. Zero the LCR Meter: Short the test leads together and run the auto-zero/offset null function to cancel out lead inductance and resistance.
  2. Measure L1 and L2 Individually: Record the exact inductance and DCR of each component at your circuit's switching frequency (e.g., set the LCR meter to 100 kHz or 1 MHz, not the default 120 Hz).
  3. Wire the Series Chain: Plug L1 and L2 into the breadboard. Use the shortest, thickest jumper wires possible between Node A, Node B, and Node C to minimize parasitic DCR.
  4. Measure $L_{eq}$: Place the LCR probes on Node A and Node C. Verify that $L_{eq} \approx L_1 + L_2$.
  5. Check for Mutual Coupling: While monitoring $L_{eq}$ on the meter, physically rotate L2 by 90 degrees, or move it two inches away from L1. If the measured $L_{eq}$ changes by more than 2-3%, your initial placement had significant mutual inductance. Adjust your physical spacing accordingly.
  6. Measure $DCR_{eq}$: Switch the LCR meter to DC resistance mode (or use a precision 4-wire multimeter) across Node A and Node C. Verify it matches $DCR_1 + DCR_2$.
Warning: Never test power inductors with a standard multimeter's continuity beeper if they are already installed in a circuit with parallel decoupling capacitors. The capacitors will charge and give false readings. Always test inductors out-of-circuit or verify the board is fully discharged.

Decision Tree: When to Series Your Inductors

Use this decision path to determine if a series topology is the right choice for your next design, terminating in a concrete component strategy.

Design Constraint If True... Concrete Action / Part Pick
Need high current (>10A) but standard L? Do NOT use series. Use parallel. Pick 2x identical low-DCR inductors in parallel (e.g., 2x 1.5µH Coilcraft XEL6060 in parallel for 0.75µH at 20A).
Need to block both low-freq ripple and high-freq EMI? Use series with mixed core materials. Pick 1x powdered iron choke + 1x ferrite bead in series (e.g., Coilcraft XEL4020-102 + Würth 74279204).
Need a non-standard L value at moderate current (<5A)? Use series with identical standard parts. Pick 2x identical standard E-series inductors in series (e.g., 2x 2.2µH Coilcraft XEL4020-222 for 4.4µH).
Need to reduce physical height profile? Use series with smaller footprint parts. Pick 2x low-profile 0805 or 1008 size inductors in series instead of one massive 2525 size inductor.

For a deeper theoretical foundation on how magnetic fields interact in these configurations, refer to the Electronics Tutorials guide on series inductors. By treating series inductors not just as a math equation, but as a tool for thermal distribution and EMI cascading, you can solve PCB layout and component sourcing problems that single-component designs simply cannot handle.