When you need a specific inductance value that isn't in your parts bin, or when you need to distribute thermal load across multiple components, combining inductors in series and parallel is the standard bench workaround. The direct answer for calculating total inductance is straightforward: series inductors add linearly ($L_{total} = L_1 + L_2$), while parallel inductors follow the reciprocal rule ($1/L_{total} = 1/L_1 + 1/L_2$), provided there is zero mutual magnetic coupling between them.
However, treating inductors exactly like resistors is a fast track to melted components and unstable switching regulators. Inductors store energy in magnetic fields, meaning physical placement, saturation current ($I_{sat}$), and DC resistance (DCR) dictate whether your network will actually work on the bench. This guide breaks down the topologies, failure modes, and a real-world buck converter design to show you exactly how to implement these networks.
The Core Topologies: Series vs. Parallel Inductors
Before wiring anything, we need to define the nodes and the physical constraints of each topology. The golden rule for both configurations is physical separation. If the magnetic field of one inductor intersects the core of another, mutual inductance ($M$) alters your total inductance, rendering standard formulas useless. Keep unshielded inductors at least one full component-diameter apart, or use shielded drum-core parts.
Series Topology
In a series configuration, the current has only one path.
Node mapping: Node A $
ightarrow$ L1 $
ightarrow$ Node B $
ightarrow$ L2 $
ightarrow$ Node C.
- Formula: $L_{total} = L_1 + L_2 + L_3...$
- Current constraint: The exact same AC and DC current flows through every inductor. Your maximum continuous current is strictly limited by the inductor with the lowest $I_{rms}$ (temperature rise) rating.
- DCR impact: Total DC resistance adds up ($DCR_{total} = DCR_1 + DCR_2$), which increases $I^2R$ conduction losses.
Parallel Topology
In a parallel configuration, the current splits across multiple paths.
Node mapping: Node A splits to both L1 and L2, which recombine at Node B.
- Formula: $\frac{1}{L_{total}} = \frac{1}{L_1} + \frac{1}{L_2} + \frac{1}{L_3}...$
- Current constraint: Current divides inversely proportional to the DCR of each branch. If the inductors are identical, current splits 50/50.
- DCR impact: Total DCR drops, following the parallel resistor formula, which significantly reduces conduction losses and thermal buildup.
Behavior & Failure Modes: What Happens When Things Break
Understanding what breaks at the extremes (open or short circuits) is critical for designing reliable power filters. A shorted capacitor blows a fuse; a shorted inductor creates a subtle, often destructive shift in filter cutoff frequencies.
| Topology | Event / Change | Effect on Total Inductance ($L_{total}$) | System-Level Consequence |
|---|---|---|---|
| Series | L1 value increases | $L_{total}$ increases linearly | Filter cutoff frequency drops; slower transient response. |
| Series | L1 fails OPEN | Drops to Zero (Infinite Impedance) | Circuit breaks entirely. Power delivery stops. |
| Series | L1 fails SHORT | $L_{total}$ drops to $L_2$ | Filter cutoff shifts higher; increased output ripple. |
| Parallel | L1 value increases | $L_{total}$ increases slightly (approaches $L_2$) | Minor shift in ripple current distribution. |
| Parallel | L1 fails OPEN | $L_{total}$ drops to $L_2$ | Remaining inductor takes 100% of the current; high risk of thermal runaway if not rated for full load. |
| Parallel | L1 fails SHORT | $L_{total}$ drops to near Zero | Massive short circuit across the nodes. High current spike; likely destroys the switching MOSFET or blows the input fuse. |
The failure-mode contrast is stark: an open failure in series kills the circuit safely, while an open failure in parallel silently overloads the surviving component. Conversely, a short failure in series degrades performance, but a short failure in parallel is catastrophic. For mission-critical power rails, series inductors offer a safer, albeit higher-loss, fail-open profile.
Design Walkthrough: Sizing Inductors for a Buck Converter Filter
Why choose a parallel inductor topology over a single, larger series inductor? The answer usually comes down to thermal management and footprint constraints. Let's walk through a real design scenario for a 5V, 4A synchronous buck converter (e.g., using a TI TPS54560).
Using the standard inductor selection formula outlined in TI's buck converter design guidelines, we calculate a target inductance of 10µH to maintain a 30% ripple current ratio. We need an inductor rated for at least 5A $I_{rms}$ (thermal) and 6A $I_{sat}$ (peak).
Option A: Single 10µH Inductor
We select a single Coilcraft XGL5030-100 (10µH, 6.8A $I_{rms}$, 8.2A $I_{sat}$).
The problem: It has a footprint of 5.0 x 5.3mm and a DCR of 18.5mΩ. At 4A continuous load, the $I^2R$ loss is $4^2 \times 0.0185 = 0.296W$. In a densely packed, low-airflow enclosure, this localized hot spot pushes the component near its 40°C temperature rise limit.
Option B: Two 20µH Inductors in Parallel
Instead, we select two Coilcraft XGL4020-200 inductors (20µH each, 4.5A $I_{rms}$, 5.5A $I_{sat}$).
The math: $\frac{1}{L_{total}} = \frac{1}{20} + \frac{1}{20} \rightarrow L_{total} = 10\mu H$.
The advantage: Each XGL4020 has a DCR of 23.6mΩ. In parallel, the effective DCR is 11.8mΩ. The new $I^2R$ loss is $4^2 \times 0.0118 = 0.188W$. More importantly, that heat is distributed across two physically separated 4.0 x 4.0mm footprints, drastically lowering the local PCB temperature. The current splits evenly (2A per branch), keeping both well under their 4.5A $I_{rms}$ limits.
For high-current, space-constrained DC-DC converters, parallel inductors win on thermal distribution and DCR reduction. For simple EMI filtering where current is low, a single or series inductor is cheaper and easier to route. For more on selecting the right core material for these topologies, refer to the Coilcraft inductor selection guide.
How to Breadboard and Test Inductor Networks Step-by-Step
Testing inductors requires more than just a standard multimeter. You need an LCR meter to verify the AC inductance and a DMM to check the DC resistance. Follow these steps to validate your network on the bench.
- Prepare the Breadboard and Spacing: Insert your inductors into the breadboard. If using unshielded radial or axial inductors, ensure they are placed at least 1 inch apart and oriented at 90-degree angles to each other. This orthogonal placement minimizes mutual inductance coupling.
- Wire the Nodes: Use short, thick jumper wires (22 AWG or thicker) for parallel networks to ensure the wire resistance doesn't skew the current sharing. For series, standard 24 AWG jumpers are fine.
- Measure DCR (DC Resistance): Set your multimeter to the lowest Ohms range. Short your probes to measure lead resistance, then subtract this from your component reading. Verify that parallel branches have matched DCR (within 5%) to ensure balanced current sharing.
- Configure the LCR Meter: Set your LCR meter to measure 'L' (Inductance) in series-equivalent mode (Ls).
Frequency Selection: Set the test frequency to match your application. Use 100kHz for switching power supply inductors, 1kHz for general audio/filtering, and 120Hz for line-frequency AC chokes.
- Measure Total Inductance: Connect the LCR meter probes across the outer nodes (Node A and Node C for series; Node A and Node B for parallel). Record the value.
- Verify Against Calculation: Compare the LCR reading to your theoretical calculation. If the measured value is more than 10% higher than expected in a parallel or series setup, your components are too close together and are magnetically coupling. Move them further apart and re-test.
Frequently Asked Questions
Can I wire inductors in series to increase current handling?
No. Wiring inductors in series does not increase current handling capacity. Because the exact same current must flow through every component in a series string, the maximum continuous current ($I_{rms}$) and peak saturation current ($I_{sat}$) of the entire network is strictly limited by the weakest (lowest rated) inductor in the chain. If you need higher current handling, you must use parallel inductors or a single, physically larger inductor.
Why does my parallel inductor calculation measure wrong on the LCR meter?
The most common cause is mutual inductance. The standard parallel formula ($1/L_{total} = 1/L_1 + 1/L_2$) assumes the magnetic fields do not interact. If you place two unshielded inductors side-by-side on a breadboard, the magnetic flux from L1 cuts through the core of L2. Depending on their physical orientation, this coupling can either add to or subtract from the total inductance. To fix this, separate the components by at least one full body-length, or rotate one inductor 90 degrees so their magnetic axes are orthogonal.
Do inductors in series and parallel rules apply to AC impedance?
Yes, but you must account for the complex nature of AC impedance ($Z$). An inductor is not a pure inductance; it has parasitic DC resistance (DCR) and parallel capacitance. The total impedance is $Z = R + j\omega L$. When calculating series or parallel networks for high-frequency AC circuits (like RF filters), you cannot just add the inductance values. You must calculate the complex impedance of each branch at your specific operating frequency and combine them using standard AC network analysis. For low-frequency power applications (under 1MHz), the simple inductance formulas are sufficiently accurate.
