When you need to tune inductance, manage current, or distribute thermal load in a power or RF circuit, combining multiple inductors is a standard bench and production technique. The direct answer on which topology to choose is simple: use series inductors when you need to increase total inductance, share voltage stress, or build multi-stage filters; use parallel inductors when you need to decrease total inductance, increase current handling capacity ($I_{sat}$), and lower DC resistance (DCR) to reduce heat.
This guide breaks down the exact node topologies, failure modes, and breadboard testing protocols for series and parallel inductors, terminating in a concrete decision framework with real-world component selections.
The Core Decision: Series vs. Parallel Topologies
To analyze these configurations, we define three nodes: Node A (input/source), Node B (junction), and Node C (output/load).
Series Topology
In a series configuration, current flows sequentially through each component. Inductor L1 connects between Node A and Node B. Inductor L2 connects between Node B and Node C.
- Inductance: $L_{total} = L_1 + L_2$
- DCR: $DCR_{total} = DCR_1 + DCR_2$
- Current Handling: Limited by the lowest individual saturation current ($I_{sat}$) in the chain.
Parallel Topology
In a parallel configuration, both inductors bridge the exact same nodes. L1 and L2 both connect directly between Node A and Node C.
- Inductance: $\frac{1}{L_{total}} = \frac{1}{L_1} + \frac{1}{L_2}$ (For two identical inductors, $L_{total} = \frac{L}{2}$)
- DCR: $\frac{1}{DCR_{total}} = \frac{1}{DCR_1} + \frac{1}{DCR_2}$
- Current Handling: $I_{sat(total)} = I_{sat1} + I_{sat2}$ (Assuming matched components and symmetrical PCB traces).
Behavior Matrix: How Parameters Shift When Components Change
Understanding how parasitic and primary parameters scale is critical for power supply compensation and thermal management. Here is how the network behaves when you add a second identical inductor to the circuit.
| Parameter | Adding in Series | Adding in Parallel |
|---|---|---|
| Total Inductance ($L$) | Doubles ($2L$) | Halves ($0.5L$) |
| DC Resistance (DCR) | Doubles (More $I^2R$ heat) | Halves (Less $I^2R$ heat) |
| Saturation Current ($I_{sat}$) | Unchanged (Limited by weakest link) | Doubles (Current splits) |
| Thermal Current ($I_{rms}$) | Unchanged | Increases by $\approx \sqrt{2}$ |
| Self-Resonant Frequency (SRF) | Decreases (More parasitic C) | Increases (Less parasitic C) |
Failure Mode Contrast: What Breaks at the Extremes?
Inductors rarely fail open unless subjected to massive overcurrent that melts the winding wire. However, understanding the extremes is vital for safety and system reliability, especially in high-power DC-DC converters.
Series Topology Failures
- Open Circuit (One element fails open): The entire circuit breaks. Current drops to zero. This is a safe failure mode—the load loses power, but no downstream components are destroyed.
- Short Circuit (One element fails short): Total inductance drops to the value of the remaining inductor. If the remaining inductor cannot handle the ripple current, it will saturate, causing a massive current spike that may destroy the upstream switching MOSFET.
Parallel Topology Failures
- Open Circuit (One branch fails open): Critical Hazard. Total inductance suddenly doubles to the remaining inductor's value. More importantly, the remaining inductor is now forced to carry 100% of the load current. If it was sized for 50% of the load, it will instantly exceed its $I_{sat}$, overheat, and likely fail short, cascading into a catastrophic board-level failure.
- Short Circuit (One branch fails short): Total inductance drops to near zero (just the DCR of the shorted coil). The power supply sees a dead short to ground, resulting in an immediate overcurrent trip or blown fuse.
Design Walkthrough: Sizing Real Inductors for a 5A Buck Converter
Let’s design the output filter for a 5A nominal (7A peak) synchronous buck converter stepping 12V down to 3.3V at 500kHz. The controller datasheet recommends a target inductance of 4.7µH.
The Problem: Finding a single 4.7µH inductor that handles 7A peak without saturating, while maintaining a low DCR to keep efficiency above 90%, usually requires a massive 12x12mm or 15x15mm footprint (e.g., Coilcraft MSS1560 series). This is too tall for our low-profile enclosure.
The Parallel Solution: We will use two identical inductors in parallel to halve the inductance, double the current handling, and halve the DCR, using a smaller footprint.
- Target Individual Inductance: $4.7\mu H \times 2 = 9.4\mu H$. We select the standard 10µH value.
- Select the Component: We choose the Bourns SRP1265A-100M.
- Value: 10µH
- $I_{sat}$: 8.5A
- $I_{rms}$: 8.0A
- DCR: 18mΩ (max)
- Footprint: 12.5 x 6.5mm (Low profile 6.5mm height)
- Calculate Parallel Network:
- $L_{total} = 10\mu H / 2 = \mathbf{5.0\mu H}$ (Well within the 4.7µH target tolerance).
- $I_{sat(total)} = 8.5A + 8.5A = \mathbf{17A}$ (Massive headroom over our 7A peak requirement; core saturation is virtually eliminated).
- $DCR_{total} = 18m\Omega / 2 = \mathbf{9m\Omega}$ (Cuts copper losses in half compared to a single equivalent large inductor).
Breadboard Testing Protocol: Step-by-Step Verification
Before committing to a PCB layout, verify your series or parallel network on a breadboard. Note that breadboards introduce 2-5pF of parasitic capacitance per contact, which can skew high-frequency measurements. Use short, thick jumper wires (22 AWG solid core).
- De-energize and Measure DCR: With the circuit unpowered, use a 4-wire Kelvin multimeter (or a standard DMM if DCR > 100mΩ) to measure the total DC resistance across Node A and Node C. Verify it matches your calculated $DCR_{total}$ ±10%.
- LCR Meter Baseline: Connect an LCR meter across the network. Set the test frequency to 100kHz (the standard reference for power inductors) and measure the total inductance. For our parallel Bourns design, you should read ~4.8µH to 5.0µH.
- Build the RL Test Jig: To test behavior under AC conditions, build a simple low-pass RL filter. Connect a function generator output to a 50Ω series resistor, then to Node A. Connect Node C to ground. Connect an oscilloscope probe across the inductor network (Node A to Ground).
- Sweep and Measure Cutoff: Set the function generator to a 1Vpp sine wave. Sweep the frequency from 10kHz to 10MHz. Identify the -3dB cutoff frequency ($f_c$) where the voltage drops to 0.707Vpp.
- Calculate Real-World L: Use the formula $L = \frac{R}{2 \pi f_c}$. If your 50Ω resistor yields a cutoff at 1.6MHz, your real-world inductance is $L = \frac{50}{2 \pi \times 1,600,000} \approx 4.97\mu H$. This confirms your parallel math holds up outside the datasheet.
Decision Tree: Picking Your Final Configuration
Stop guessing. Use this decision matrix to lock in your topology and select a concrete starting part number for your next build. You can cross-reference these families using the Coilcraft Power Inductor Finder or equivalent distributor parametric search tools.
| Your Design Constraint | Topology Pick | Concrete Part Recommendation |
|---|---|---|
| Need high inductance (>100µH) for low-current (<1A) EMI filtering or snubber networks. | SERIES (Combines two smaller, cheaper coils to reach high L without massive core size). |
Wurth Elektronik 74477420 (1mH, 0.3A $I_{sat}$). Put two in series for 2mH if needed. |
| Need low inductance (1-10µH) but massive current handling (>10A) for high-power buck/boost converters. | PARALLEL (Halves DCR for thermal management, doubles $I_{sat}$ to prevent core saturation). |
Bourns SRP1265A-100M (10µH). Use two in parallel for 5µH @ 17A peak. |
| Need to share high AC voltage stress across multiple components in a resonant tank or high-frequency inverter. | SERIES (Voltage divides across the inductors, preventing dielectric breakdown of a single coil's insulation). |
Coilcraft 2929SQ series (RF inductors). Series them to distribute RF voltage. |
| Need to filter high-frequency common-mode noise on a differential data or power line. | NEITHER (Standard series/parallel won't work. You need magnetic coupling). |
Use a dedicated Common Mode Choke (e.g., Wurth 744825 series). |
For further reading on the foundational math behind these networks, refer to the All About Circuits guide on inductors in series and parallel. Always verify your final layout with a physical prototype to account for PCB trace inductance and mutual coupling effects that theoretical calculations cannot fully predict.






