Tutorials
What Happens to Inductor Current When the Voltage Changes?
Published · Updated
An inductor is often introduced as a coil, but its useful circuit relationship concerns changing current. A drawing of wound wire does not tell you the current rating, saturation behavior or suitable frequency. Start with the voltage across an ideal inductance, predict a short current ramp, and then separate that model from the heating and magnetic limits of a real part. The result is a calculable interval and a clear list of questions for the actual device.
Voltage changes the slope, not an arbitrary current setting

For an ideal constant inductance using the passive sign convention, voltage equals L times the rate of change of current. Expressed another way, current slope is the voltage directly across the inductance divided by L. Positive voltage in the stated reference direction produces a positive slope. Reversing that voltage reverses the slope. The current value at the start of the interval is still needed to find its value at the end.
This relationship explains why a finite current change in an ideal inductance takes a finite voltage-time interval. It does not say that current can never change. It also does not say that an inductor fixes one current independently of its circuit. The source, switch states, resistances and other components determine the voltage that appears across it. A useful explanation follows those connections instead of describing the inductor as an isolated current regulator.
A finite voltage across an ideal inductance cannot produce an instantaneous finite jump in current. In a real circuit, trying to interrupt a current path can instead produce a substantial voltage while stored energy moves through available paths. That is why switch-off behavior requires a complete circuit and a defined energy destination. This introductory lesson is not a procedure for opening a powered inductive load or probing a hazardous converter.
The inductor symbol does not establish whether the physical part is shielded, which material it uses, or how its inductance changes under bias. A schematic reference such as L3 identifies a position. The BOM and exact datasheet identify what was selected. Two visually similar molded parts can have different ratings. Start by asking what voltage-time behavior the circuit needs, then obtain evidence that the selected part supports it.
Calculate one short interval with units visible
Assume an ideal 100-microhenry inductance, zero initial current and a constant 1 volt directly across the inductance for 10 microseconds. These are hypothetical model inputs. Current change equals voltage times time divided by inductance. Using SI units gives one times ten millionths divided by one hundred millionths, or 0.1 amperes. The end current is therefore0.1 amperes for the stated zero initial condition.
The same interval with an initial current of 0.2 amperes ends at 0.3 amperes. Doubling inductance halves the current change when voltage and time stay fixed. Doubling time doubles the change. A negative applied voltage reduces the signed current in this reference frame. If the model passes through zero, its algebra continues; a real switching circuit may enter another conduction state before following that continuation.
The phrase directly across the inductance matters. If a source drives a resistor and inductor in series, the full source voltage does not remain across L while current increases. The resistor voltage grows and the inductor voltage changes. OpenStax's RL circuit analysis treats that complete first-order response. Do not use the constant-voltage interval calculator as a simulation of every source-fed RL loop.
A switching converter can often be examined through successive intervals, but identifying those states requires its actual topology and operating mode. The voltage across L in one interval can differ from the supply, output or another node voltage. State boundaries, diode or switch behavior, control and load conditions must be established before combining intervals. A short linear ramp is a building block for analysis, not a complete regulator design.
Watch current and energy without inventing a measured waveform
The original animation in this lesson illustrates the ideal interval. It is a calculated concept sequence, not an oscilloscope capture, SPICE run or saturation experiment. Its purpose is to connect voltage-time area with current change. The same values remain available in ordinary text so that a moving line is not the only way to understand the explanation or inspect its assumptions.
Ideal stored magnetic energy is one half L I squared. With100 microhenries and 0.1 amperes, that equals0.5 microjoules. At0.3 amperes, it is4.5 microjoules. Tripling current increases the ideal stored energy by a factor of nine for constant L. This is a model relationship; a real magnetic part can change inductance as conditions change and can dissipate energy through several mechanisms.
Energy at the end of one interval is not the same as power dissipated during the interval. An ideal inductor stores and returns energy. A real winding has resistance, and a real core may incur losses. A regulator also has switch, diode, capacitor and other losses. Keeping these accounting categories separate prevents a stored-energy figure from being treated as a thermal load or a converter-efficiency prediction.
Before considering a switch-off case, draw the current path for the next state and identify which element can accept the energy. A missing path cannot be solved by wishing current to become zero. The actual protective arrangement must match the circuit and source evidence. Do not copy a generic diode across every inductive component: the correct clamp, polarity and operating behavior depend on the load and switching topology.
Two current ratings answer different questions
Real inductor selection commonly requires both magnetic and thermal evidence. Saturation-current information describes how inductance changes as current biases the magnetic structure under specified conditions. RMS-current information concerns heating under a stated test method. They are different limits and must not be compared as if both were universal maximum current numbers. The definition and allowed change belong beside the number.
Coilcraft's introduction to inductor specifications (source address: https://www.coilcraft.com/en-us/resources/application-notes/getting-started-an-introduction-to-inductor-specif/) explains this distinction and the influence of frequency, temperature and current. Different series can use different percentage changes or temperature-rise conditions. A larger displayed current rating is not automatically stronger evidence for your waveform. Read the criterion before using it to filter candidates.
Winding DC resistance contributes copper loss. For a constant DC current, I squared times DCR provides a first resistive-loss estimate at the resistance condition used. A varying waveform requires the appropriate RMS treatment and may have additional AC losses. DCR itself changes with temperature. Neither a DC snapshot nor a single catalog rating establishes operating temperature in your board's copper and airflow environment.
Frequency behavior also limits the ideal model. Parasitic capacitance and winding effects mean a real inductor cannot be treated as an ideal L at every frequency. Self-resonance, impedance data and the measurement conditions matter for RF and filtering tasks. A power inductor chosen for one regulator job is not automatically an RF component. Avoid inferring the intended application from its inductance value alone.
Use the ramp explorer and label what it leaves unresolved
The interactive tool accepts ideal voltage across L, inductance in microhenries, interval in microseconds and signed initial current. It calculates current change, end current and ideal stored energy. An optional DCR input also gives a clearly labelled hypothetical constant-DC copper-loss snapshot using the end current. That snapshot is not the average loss during the ramp and is not a waveform RMS calculation.
Static reference: 1 V across100 µH for 10 µs from 0 A →0.1 A; ideal energy0.5 µJ. Interactive fields appear when JavaScript is available. The explanation and example remain readable without it.
Start with 1 volt,100 microhenries,10 microseconds and zero initial current. Check the0.1 ampere increase and 0.5 microjoule energy. Then change the initial current to 0.2 amperes and predict the end value. Finally reverse the voltage sign while holding time constant. The signed results should help you follow the chosen references rather than treating negative current as automatically impossible or faulty.
If DCR is0.1 ohm and you use the0.1 ampere end current as a hypothetical steady DC current, the snapshot is1 milliwatt. The tool labels exactly that limited interpretation. In a real source-fed winding, DCR also affects the voltage division and current trajectory, which this constant-voltage-across-L model does not calculate. This deliberate separation prevents a convenient arithmetic addition from silently becoming a more sophisticated simulation.
Zero or negative inductance is rejected, and negative elapsed time is outside the worksheet's domain. Invalid input checks protect the calculation but cannot verify an entered component. Keep the result with the assumed operating interval, the applicable inductance evidence and missing saturation or thermal information. If the model result approaches an uncertain rating, obtain the actual curve or reconsider the operating case rather than extending the graph confidently.
Carry the inductor evidence into a power-rail project
A comparison with a complete ideal RL loop helps define the ramp tool's boundary. Apply a constant voltage source to an ideal inductance in series with a resistor, starting from zero current. Current eventually approaches source voltage divided by resistance. The time constant is L divided by R. Unlike the constant-voltage-across-L example, the current slope decreases because the growing resistor voltage leaves less voltage across the inductance.
For a hypothetical 1-volt source, 100-microhenry ideal inductance and 10-ohm series resistor, the time constant is 10 microseconds and the final current is 0.1 amperes. After one time constant, the ideal RL result is approximately 0.0632 amperes. The constant-1-volt-across-L ramp would instead predict a 0.1-ampere increase over that interval. The discrepancy is expected: the two calculations apply different voltage conditions to the inductance.
At the beginning of the RL response, current is zero and the full source voltage appears across the ideal inductance. Later, part of the source voltage appears across the resistor. At ideal steady DC, the voltage across L tends toward zero while current is set by the series resistance. A real winding adds its own resistance and other behavior, so a statement such as an inductor is a short at DC must be used only within its explicit ideal-model context.
Draw the resistor voltage and inductor voltage as separate quantities when checking an unfamiliar circuit. Their sum follows the source in the stated simple loop. This makes it easier to recognize whether a measured node voltage is actually the voltage required by the equation. Measuring only one terminal relative to ground may not give the voltage across the component. The actual probing method must suit the circuit and instrument, especially where switching nodes or non-ground references are involved.
The practical review question is therefore not which formula is more advanced. It is which model matches the applied conditions and time interval. Use a constant-voltage ramp for an interval where that approximation is justified. Use the complete RL response when series resistance shapes the trajectory. Obtain a more suitable model when nonlinear magnetics, switching states or parasitic behavior dominate. Preserve the model choice alongside the numerical result so later readers can assess it.
An FPGA rail is an appropriate place to see why these distinctions matter, but the regulator documentation must drive the inductor choice. Required inductance, ripple, current peaks, operating mode and loss conditions depend on the actual regulator and load. The power-rail planning tool organizes rail requirements; it does not select an inductor from this beginner model. The XC7A200T-2FBG484C evidence page is relevant only as exact-device context for a board using that FPGA.
Before assembly, preserve the regulator recommendation, inductor code, current assumptions and board revision. Check the footprint and mechanical drawing independently of the conceptual coil illustration. The PCB bring-up checklist helps separate power-on observations from a validated rail, while the PCBA preparation page supports an agreed assembly and inspection scope. The tool directory provides other verified engineering tasks as the project develops.
For parts, request an inductor quotation with the exact MPN, quantity and required date after the electrical selection. For a board or complete project, submit the project scope through Contact, including board quantity, BOM revision, sourcing responsibility and power-test requirements. Do not substitute a nominal inductance for a complete ordering code or place private design files in a query string.
The lesson's output is intentionally bounded: one defensible current interval, one ideal energy calculation, and a list separating magnetic, thermal, frequency and manufacturing checks. You should be able to explain what changing the voltage or duration does to the slope, and why the same calculation cannot establish saturation margin or converter efficiency. That is a stronger foundation than remembering only that an inductor is a coil.
Show the original inductor concept animation

Concept animation, not a hardware recording or a validated simulation. Close this disclosure to hide it. Static steps and numerical assumptions are explained in the article.