Winding your own inductors is a rite of passage for anyone building custom switch-mode power supplies, RF matching networks, or audio crossovers. While modern ferrite and powdered-iron cores make the physical winding easy, getting the exact inductance value requires a solid grasp of the underlying math. The formula of self inductance bridges the gap between the physical geometry of your coil and its electrical behavior. When applied correctly, it saves you from endless trial-and-error winding; when applied blindly, it leads to saturated cores and blown semiconductors.
The Core Formula of Self Inductance and Symbol Definitions
At its most fundamental level, self inductance (L) is defined by the ratio of total magnetic flux linkage to the current producing it. This gives us the foundational equation: L = (N · Φ) / I. However, on the workbench, you rarely know the magnetic flux (Φ) directly. Instead, you know the physical dimensions of your coil and core. By substituting the magnetic circuit equivalents, we arrive at the geometric formula of self inductance for a long solenoid, widely cited in resources like Georgia State University HyperPhysics:
L = (μ0 · μr · N2 · A) / l
| Symbol | Parameter | Standard Unit | Notes & Constants |
|---|---|---|---|
| L | Self Inductance | Henries (H) | The target electrical property. |
| μ0 | Permeability of Free Space | H/m | Constant: 4π × 10-7 (≈ 1.2566 × 10-6) H/m. |
| μr | Relative Permeability | Dimensionless | Core material multiplier (1 for air, 100-10,000 for ferrites). |
| N | Number of Turns | Dimensionless | Total loops of wire. Note that N is squared in the formula. |
| A | Cross-Sectional Area | Square meters (m2) | The area of the core or coil form the flux passes through. |
| l | Magnetic Path Length | Meters (m) | The physical length of the coil or the effective magnetic path of the core. |
Rearranged Forms and the Unit Mistakes That Break Them
On the bench, you rarely solve for L directly. Usually, you have a target inductance and a specific core in your parts bin, meaning you need to solve for the number of turns (N) or the required air gap (which alters effective μr). Here are the rearranged forms of the formula of self inductance:
- Solve for Turns (N): N = √[ (L · l) / (μ0 · μr · A) ]
- Solve for Area (A): A = (L · l) / (μ0 · μr · N2)
- Solve for Length (l): l = (μ0 · μr · N2 · A) / L
- Solve for Relative Permeability (μr): μr = (L · l) / (μ0 · N2 · A)
Conversion cheat sheet:
1 cm2 = 1 × 10-4 m2 (Not 10-2! This is a squared dimension).
1 mm2 = 1 × 10-6 m2.
1 cm = 0.01 m.
Solved Problems: Tracking Units from RF to Power
Let's run through two distinct bench scenarios, tracking every unit conversion to ensure the math holds up.
Problem 1: Air-Core RF Choke
Scenario: You are winding an air-core solenoid on a 5mm diameter ceramic form for a 14MHz RF matching network. The winding length is 2cm, and you manage to fit 18 turns of enameled wire. What is the inductance?
- Identify Knowns: N = 18, l = 0.02 m, μr = 1 (air), μ0 = 1.2566 × 10-6 H/m.
- Calculate Area: Diameter = 5mm = 0.005m. Radius (r) = 0.0025m.
A = π · r2 = π · (0.0025)2 = 1.963 × 10-5 m2. - Apply Formula: L = (1.2566 × 10-6 · 1 · 182 · 1.963 × 10-5) / 0.02
- Compute Numerator: 1.2566 × 10-6 · 324 · 1.963 × 10-5 = 7.99 × 10-9
- Divide by Length: L = 7.99 × 10-9 / 0.02 = 3.99 × 10-7 H, or roughly 399 nH.
Problem 2: Ferrite-Core Buck Converter Inductor
Scenario: You need a 47µH inductor for a 500kHz buck converter. You select a T50-2 powdered iron toroid (μr = 10, Ae = 0.197 cm2, le = 3.12 cm). How many turns do you need?
- Convert to SI: L = 47 × 10-6 H. A = 0.197 × 10-4 m2 = 1.97 × 10-5 m2. l = 0.0312 m.
- Select Rearranged Formula: N = √[ (L · l) / (μ0 · μr · A) ]
- Calculate Denominator: 1.2566 × 10-6 · 10 · 1.97 × 10-5 = 2.475 × 10-10
- Calculate Numerator: 47 × 10-6 · 0.0312 = 1.466 × 10-6
- Divide and Root: N = √(1.466 × 10-6 / 2.475 × 10-10) = √(5923) ≈ 77 turns.
Bench Scenario: When the Formula Fails in the Real World
Math on a whiteboard is pristine; magnetics on a workbench are messy. Here is a real-world narrative of how blind trust in the formula of self inductance leads to hardware failure, as detailed in advanced magnetics design guides from sources like Electronics Tutorials.
The Setup: A designer was tasked with building a 10mH DC link choke for a 48V / 15A solar MPPT charge controller to smooth the input current. They selected a large, ungapped manganese-zinc ferrite E-core with a massive relative permeability (μr = 2500), a cross-sectional area of 1.5 cm2, and a magnetic path length of 8 cm.
The Numbers: Plugging these into the rearranged formula for N yielded exactly 65 turns of thick 14 AWG wire. After winding, they connected the inductor to a benchtop LCR meter set to 1kHz / 1Vrms. The meter proudly displayed 10.2mH. The math worked.
The Outcome: The inductor was soldered into the MPPT prototype. At a light 1A load, the circuit ran perfectly. But when the sun peaked and the controller pulled 15A, the high-side MOSFET violently failed, popping its package and scorching the PCB.
What Went Wrong: The designer fell victim to core saturation. The formula of self inductance assumes that μr is a constant linear value. In reality, ferrite materials exhibit a non-linear B-H curve. At just 2A of DC bias, the magnetic flux density (B) inside the ungapped core crossed the 'knee' of the saturation curve. As the core saturated, its effective μr plummeted from 2500 down toward 1 (air). Under the 15A operating load, the actual inductance collapsed to roughly 250µH. This massive drop in inductance allowed the switching ripple current to spike well beyond the MOSFET's Safe Operating Area (SOA), causing thermal runaway.
The Fix: The designer introduced a physical 1.5mm air gap into the center leg of the E-core. While this dropped the zero-bias inductance (requiring them to increase the turns to 140 to hit 10mH again), the air gap drastically increased the core's resistance to saturation. The new gapped inductor maintained a stable 9.5mH all the way up to 20A of DC bias.
Assumptions, Limits, and Realistic Magnitudes
To use the formula of self inductance effectively, you must understand the boundaries of its accuracy. The geometric formula L = (μ0 · μr · N2 · A) / l relies on several strict assumptions:
- The 'Long Solenoid' Assumption: The formula assumes the length of the coil is significantly greater than its diameter (l ≫ √A). If you are winding a short, stubby coil, the magnetic field fringes heavily at the ends, and the actual inductance will be 10% to 30% lower than the formula predicts. (Nagaoka's coefficient is used to correct this in precision RF work).
- Uniform Flux Density: It assumes all flux lines are perfectly contained within the cross-sectional area A and travel exactly along path length l. In open-core geometries (like rod inductors), flux leaks into the surrounding air, making the formula an overestimation.
- Linear Material Response: As demonstrated in the bench scenario, the formula assumes μr does not change with current. For high-current power applications, you must use the amplitude permeability curves from the core manufacturer's datasheet rather than the initial permeability value.
Finally, having an intuitive sense of realistic magnitudes prevents decimal-place errors when programming your calculator. Here is what you should expect to see on the bench:
- Nanohenries (nH): Typical for air-core RF chokes, VHF/UHF matching networks, and high-speed digital decoupling. (e.g., 10nH to 900nH).
- Microhenries (µH): The standard domain for switch-mode power supplies (buck/boost converters), EMI filtering, and audio crossovers. (e.g., 1µH to 500µH).
- Millihenries (mH): Used in low-frequency AC line filtering, heavy DC link chokes, and low-pass audio filters. (e.g., 1mH to 100mH).
- Henries (H): Rare in modern compact electronics, but found in vintage telephone hybrids, massive power factor correction banks, and specialized laboratory ballasts. (e.g., 1H to 10H).
By respecting the unit conversions, verifying your core's saturation limits, and understanding the geometric assumptions, the formula of self inductance transforms from a textbook abstraction into a highly reliable tool for custom magnetics design.






