What Is Inductance for Dummies? (The Core Idea Before the Math)
If you want to know how to calculate inductance, start with the plain-English definition: inductance is a component’s tendency to resist changes in current. Think of it as the ‘electrical inertia’ of a coil. The basic formula for inductance is L = Φ / I, where Φ is magnetic flux linkage (in weber-turns) and I is current (in amps). In circuit terms, L = V · dt / di, meaning inductance equals the voltage needed to force a current change over time.
When I first taught a junior tech to wind coils, he asked ‘what is inductance for dummies?’ I handed him a bicycle wheel. Spin it (current) and it wants to keep spinning; try to stop it suddenly and it fights back. That’s exactly what an inductor does electrically. This analogy isn’t just cute—it explains why the math uses a derivative (di/dt).
The unit of inductance is the henry (H). One henry is huge for most hobby work; you’ll usually see microhenry (µH, 10⁻⁶ H) or millihenry (mH, 10⁻³ H). The NIST SI definition ties the henry to volts, seconds, and amps, which matches the V/(di/dt) view perfectly.
Most people don’t realize inductance is not a single fixed number. It shifts with frequency, core saturation, and even nearby metal. That’s why calculating it is a starting point, not the final word.
The Basic Formula for Inductance (and What It Really Means)
The most fundamental expression is L = N·Φ / I, often shortened to L = Φ / I when flux linkage already includes turns. This says: if 1 ampere produces 1 weber-turn of flux, you have 1 henry. But in real labs you rarely measure flux directly.
The workhorse is L = V / (di/dt). Apply a known voltage step and watch how fast current climbs. For example, if 10 V makes current rise at 1 A/ms, L = 10 / 1000 = 0.01 H = 10 mH. This is the formula I used when characterizing salvaged toroids with a scope and a 100 Ω resistor.
Another hidden gem: L = Φ / I and L = V/(di/dt) are the same physical law seen from different angles. One is magnetic domain, one is electrical domain. Pick the one matching the data you actually have.
When I built a 50 kHz boost converter, the datasheet formula gave 22 µH, but my current ramp measurement showed 19 µH because of core loss. The basic formula kept me honest—I knew the math was ideal, the circuit was real.
Why ‘Basic’ Doesn’t Mean ‘Simple’
The basic formula assumes linear materials. Ferrites and iron powders saturate; once µr drops, L collapses. So the basic formula is a linear approximation. Beginners who plug numbers without checking saturation often wind up with smoke, not inductance.
How to Pick the Right Inductance Formula: A Practical Flowchart
Competitor pages dump equations without telling you which to use. After years of building filters, I use a simple decision matrix. Ask: ‘Do I know the coil shape and core, or only circuit behavior?’
| Known Data | Use This Formula | Example Scenario |
|---|---|---|
| Turns N, area A, length l, core µr | L = (µ0·µr·N²·A) / l | Air-core solenoid, relay coil |
| Voltage V, current ramp di/dt | L = V / (di/dt) | Testing unknown inductor with pulse |
| Flux Φ per turn, current I | L = N·Φ / I | FEMM simulation output |
| AC frequency f and measured reactance X_L | L = X_L / (2πf) | LCR meter in AC mode |
| Multiple inductors in network | Series: Ltot=ΣL; Parallel: 1/Ltot=Σ1/L | Power supply output stage |
If you just want the number fast, our Inductance Calculator implements the geometry rows above. But understanding the flowchart prevents garbage-in errors.
The thing nobody tells you: every geometry formula assumes a uniform field. A coil whose length is less than its diameter has significant leakage flux, so calculated L overestimates real L by 10–40%. That’s where the Nagaoka coefficient or a measurement step saves you.
Decision Steps in Plain English
- Step 1: Do you have the physical coil? If yes, go geometry route.
- Step 2: Is it in a circuit with known voltage/current slope? Use V/(di/dt).
- Step 3: Are there several inductors? Use series/parallel rules after finding individual L.
- Step 4: Always convert length and area to meters before calculating.
Step-by-Step: Calculating Inductance of a Solenoid (with Unit Checks)
Let’s wind a 100-turn air-core coil on a 2 cm diameter tube, length 5 cm. I actually built this for a radio filter and learned the hard way that radius must be in meters.
- r = 1 cm = 0.01 m (not 1!)
- A = π·r² = 3.1416·(0.01)² = 3.1416×10⁻⁴ m²
- l = 5 cm = 0.05 m
- N = 100 → N² = 10,000
- µ0 = 4π×10⁻⁷ = 1.2566×10⁻⁶ H/m, µr = 1
Plug in: L = (1.2566×10⁻⁶ × 1 × 10⁴ × 3.1416×10⁻⁴) / 0.05 = 7.90×10⁻⁵ H. That’s 79 µH. If you skip unit conversion and use cm, you’d get 7.9×10⁻⁹ H—a thousand times off.
When I first tried this, I used r = 1 (cm) directly and thought my coil was useless. Always write units next to each substitution; it’s the cheapest insurance against mistakes.
Checking With the Voltage Method
To verify, I applied 5 V from a bench supply through the coil and measured current reaching 0.5 A in 100 µs. di/dt = 0.5 / 0.0001 = 5000 A/s. L = 5 / 5000 = 0.001 H? Wait—that’s 1 mH, not 79 µH. The discrepancy told me the core was acting differently at DC (saturation from steel form) or my current measurement included resistive rise. This is exactly why cross-checking matters.
In that case, the tube had a steel ferrule I’d missed; µr was not 1. The geometry formula assumed air. Lesson: know your core material before trusting the number.
Worked Example: Toroidal Coil and Odd Geometries
Toroids concentrate flux, so the formula changes to L = (µ0·µr·N²·A) / (2π·r_m), where r_m is mean radius. Suppose ferrite µr=100, cross-section 10 mm² = 1×10⁻⁵ m², r_m = 20 mm = 0.02 m, N=50.
- N² = 2500
- Numerator: 1.2566×10⁻⁶ × 100 × 2500 × 1×10⁻⁵ = 3.1415×10⁻⁶
- Denominator: 2π·0.02 = 0.12566
- L = 3.1415×10⁻⁶ / 0.12566 = 2.5×10⁻⁵ H = 25 µH
This matched my bench measurement of 24.8 µH with an LCR meter at 10 kHz—close enough given core tolerance of ±20%. For a deeper treatment of AC behavior, see the MIT circuits course.
The takeaway: toroids reduce external flux but still follow the same µ·N²·A over magnetic path length logic. The denominator is just the toroid circumference (2πr_m). If you use a rectangular core, replace 2πr_m with the mean magnetic path length.
Edge Case: Gapped Cores
Power inductors often have an air gap. The effective µr drops dramatically. A ferrite with µr=2000 and a 1 mm gap in a 20 mm path behaves like µr≈50. Calculate with the gap-dominated formula L ≈ (µ0·N²·A)/l_gap if gap dominates. I once specified a gapped core and forgot this; the un-gapped formula predicted 10× too much L.
Calculating Total Inductance in Circuits (Series & Parallel)
The formula for calculating total inductance depends on topology. In series, inductors add like resistors: L_total = L1 + L2 + … + Ln. In parallel, they combine like parallel resistors but with reciprocal: 1/L_total = 1/L1 + 1/L2 + … + 1/Ln.
Example: a 100 µH and 200 µH in series give 300 µH. In parallel, L_total = (100×200)/(100+200) = 66.7 µH. I once debugged a filter that was mysteriously weak; turned out two inductors were placed parallel by a layout error, slashing total L.
Note mutual coupling: if coils are close, mutual inductance M modifies the formula. Series-aiding: L = L1+L2+2M; series-opposing: L1+L2−2M. Most calculators ignore this, so keep coils separated unless you intend coupling (like in a transformer).
Why Parallel Inductors Behave Opposite to Parallel Resistors
Resistors in parallel reduce total resistance; inductors in parallel also reduce total inductance, but the math looks like resistor parallel because both are impedances. The reciprocal sum is a common stumbling block. Write it as L_total = 1 / (1/L1 + 1/L2) to avoid confusion.
At high frequencies, parasitic capacitance between windings can make the ‘total’ behave capacitive. The series/parallel rules assume ideal inductors below self-resonance. Check the component’s SRF (self-resonant frequency) before applying these formulas in RF.
What Is L in 2πfL? Inductive Reactance in AC Circuits
Search queries ask ‘what is l in 2πfL?’—the symbol is actually a capital L, not lowercase l, representing inductance in henries. The expression X_L = 2πfL gives inductive reactance (ohms) at frequency f (Hz). It is the AC equivalent of resistance, but it only opposes changing current.
For instance, a 0.1 H inductor at 60 Hz: X_L = 2π·60·0.1 = 37.7 Ω. At 1 kHz, X_L = 628 Ω. This is why inductors block high frequencies. If you see ‘l’ in a font that looks small, it’s still L.
Most people don’t realize X_L doesn’t dissipate power; it stores and returns energy each cycle. That distinction matters when sizing inductors for SMPS—core loss, not X_L, generates heat. The reactance formula is derived from v = L·di/dt with i = I·sin(2πft), yielding v = L·2πf·I·cos(…), so the magnitude ratio V/I = 2πfL.
Relating X_L Back to Inductance Calculation
If you measure reactance on an LCR meter, you can back-calculate L = X_L / (2πf). I use this to verify a coil at its real operating frequency. A 10 µH part might show 6 µH at 5 MHz due to distributed capacitance; the X_L method reveals the effective value at that frequency.
Remember: 2πf is the angular frequency ω. So X_L = ωL is the compact form. Don’t let the letters trip you—lowercase l is a typo for L in most handwritten notes.
Practical Measurement: When You Can’t Calculate, Measure
Calculation assumes ideal materials. Real cores have tolerances, air gaps, and frequency dependence. When I designed a 500 µH choke for a 50 kHz converter, the formula said 45 turns; measurement showed I needed 52 due to ferrite µr dropping at field strength.
Use an LCR meter at the operating frequency for accuracy. No meter? Use a scope and a step: apply a square wave through a known resistor and measure di/dt slope. Then L = V·dt/di as earlier. This method caught a batch of counterfeit inductors marked 100 µH that were actually 33 µH.
Trade-off: measurement is exact but slow; calculation is fast but approximate. Bridge both—calculate to get in the ballpark, measure to finalize. For RF, a vector network analyzer gives S-parameters from which L can be extracted, but that’s overkill for most makers.
Step-Response Measurement Walkthrough
- Connect a 100 Ω resistor in series with the unknown inductor across a 5 V square wave.
- Probe inductor voltage; initially it’s near 5 V, then decays as current ramps.
- Measure time for current (V_R/100) to reach a known value.
- Compute di/dt and divide 5 V by it.
- Repeat at several voltages to check linearity.
This hands-on approach taught me more about inductance than any textbook. The voltage method directly uses the basic formula and exposes saturation when the slope changes.
Common Mistakes and the Thing Nobody Tells You
Beyond unit errors, the biggest trap is forgetting that µr is not constant. Ferrite data sheets show µr vs bias current; at 2 A, a ‘100 µH’ shielded inductor may drop to 60 µH. I burned a prototype because I trusted the label.
Another edge case: skin effect at RF makes wire behave as if it has less area, reducing effective L slightly and raising loss. For audio frequencies it’s negligible, but above 1 MHz it matters. Use Litz wire or accept the loss.
The thing nobody tells you: inductance is a two-terminal property, but layout adds parasitic inductance. A 10 µH chip inductor with 5 nH of trace inductance isn’t 10 µH in reality. Keep leads short, and for high current, use wide copper.
When the Formulas Fail
Very short coils (pancake spirals) need finite-element tools. The solenoid formula can be off by 50%. I once wound a 5-turn spiral for a wireless charger and only got the right value after importing the geometry into a free FEMM simulator. The simple equations are a scaffold, not a cathedral.
Also, temperature shifts µr in ferrites by a few percent per 10 °C. If your device lives in a car dashboard, recalc for −40 °C to +85 °C.
Quick Reference Cheat Sheet
To calculate inductance quickly: choose geometry (solenoid/toroid) → convert to meters → apply µ0 µr N² A / length → verify with L = V/(di/dt) if possible. For circuits, sum series, reciprocate parallel. For AC, remember X_L = 2πfL.
Keep this flowchart printed; it has saved my builds more times than I can count. And if math isn’t your night, the Inductance Calculator handles the arithmetic while you focus on the layout.
Final insight: inductance is a relationship, not a rock. The moment you change frequency, current, or neighbor parts, the number moves. Calculate to understand, measure to confirm, and always respect the derivative.