FPGA DEVICES · ENGINEERING · SOURCING Source-linked information. Clear project requirements.

Tutorials

Why Does a Capacitor Take Time to Charge?

Published · Updated

A capacitor can sit beside a steady supply or shape a changing signal, but the same small symbol does not explain either job by itself. The quickest route to understanding is to follow voltage and current through an explicitly stated charging model. This lesson starts with an initially uncharged capacitor and a resistor, builds a transparent calculation, and then shows why a real ceramic capacitor requires more than its nominal printed value.

The changing voltage is the clue

A real assortment of capacitor packages and technologies. This photograph illustrates physical variety; it is not an all-ceramic selection or a set of interchangeable parts.
A real assortment of capacitor packages and technologies. This photograph illustrates physical variety; it is not an all-ceramic selection or a set of interchangeable parts. Photograph: FDominec / Wikimedia Commons, CC BY-SA 3.0, unmodified. The photographed parts are not selected components for the hypothetical examples.
Charging is a changing circuit stateThe capacitor voltage rises while the resistor current falls in the ideal model.Charging is a changing circuit state+−Step VsVc(t)Initially unchargedOriginal conceptual teaching figure · not a footprint / pinout / measurement
The capacitor voltage rises while the resistor current falls in the ideal model. Original conceptual illustration; not a measured result or package drawing.

Take an ideal voltage source, a series resistor and an initially uncharged ideal capacitor. Close the switch at time zero. Current initially flows through the resistor while charge accumulates on the capacitor's terminals. The voltage across the capacitor rises. As that voltage approaches the source voltage, less voltage remains across the resistor and the charging current falls. This relationship is the reason the simple charging curve bends instead of rising at one constant slope.

It helps to distinguish two statements that are sometimes mixed together. Capacitance relates stored charge to voltage in the stated ideal model. The surrounding circuit determines how rapidly charge can move. A larger resistor restricts charging current for a given voltage difference. A larger capacitor needs more charge for the same voltage change. Neither component alone establishes the time taken by every possible charging circuit.

The capacitor symbol identifies a function, not a complete procurement specification. Two parallel plates commonly represent a non-polarized capacitor. Other symbols and drawing conventions can indicate polarity. Ordinary MLCCs are not polarized like many electrolytic capacitors, but that does not mean all capacitors are interchangeable. Always establish the actual technology and the design's symbol convention before deciding how a physical part should be connected.

A reference such as C8 identifies one board position. It does not identify a capacitance or package without the rest of the design record. Unmarked ceramic parts are especially difficult to identify from appearance. The BOM, manufacturer code and approved footprint remain important even when the component looks symmetrical. The question for this lesson is deliberately narrower: given an ideal R and C, what voltage and current does the charging model predict over time?

This starting point also prevents a misleading shortcut. Calling a capacitor a short circuit at startup and an open circuit at steady DC can be useful shorthand in a particular ideal analysis, but it is not a connection instruction. Actual initial voltage, source impedance and the event being studied must be known. A capacitor already carrying charge does not begin from the same state as the uncharged example below.

Build the time constant before drawing the curve

One time constant is a scale, not a finish lineIdeal normalized RC response; point at t = τ is about 63.2%.One time constant is a scale, not a finish linet / τVc / Vs1 τ: 63.2%5 τ: 99.3%Original conceptual teaching figure · not a footprint / pinout / measurement
Ideal normalized RC response; point at t = τ is about 63.2%. Original conceptual illustration; not a measured result or package drawing.

The charging time constant is tau = RC. Resistance in ohms multiplied by capacitance in farads gives seconds. For an illustrative 10000-ohm resistor and 10-microfarad ideal capacitor, tau is 0.1 seconds, or 100 milliseconds. This number provides a scale for the response; it is not the time at which charging suddenly finishes. The ideal response approaches its final value continuously.

For an initially uncharged capacitor driven by an ideal step Vs, the model gives Vc(t) = Vs times one minus exp of negative t divided by RC. The resistor current is Vs divided by R, multiplied by that exponential term. At time zero, capacitor voltage is zero and current is Vs/R. At one time constant, voltage is approximately 63.2 percent of the final step and current approximately 36.8 percent of its initial value.

Use a 5-volt step with the stated example. The initial current is 0.5 milliamperes. At 100 milliseconds, capacitor voltage is approximately 3.161 volts and current approximately 0.184 milliamperes. At five time constants, voltage is approximately 99.3 percent of the ideal final value. These are mathematical model outputs, not oscilloscope readings or evidence that a particular capacitor meets its nominal value.

The direction of the current also needs a reference. Define the capacitor voltage polarity and the current entering its positive terminal consistently. A negative sign in a more general equation can describe a chosen reference direction rather than a malfunction. Keep that distinction when moving from a charging circuit to a discharge circuit or a signal-coupling example. The same component can participate in different state transitions without changing its identity.

OpenStax's RC circuit treatment gives the underlying ideal response. The worked values here are independent calculations with declared inputs. If you want to compare the model with a later experiment, preserve the initial condition, source waveform, resistor value, capacitor conditions and probe arrangement. Without those details, a visually similar curve is weak evidence.

Make a prediction, then use the RC explorer

Doubling R is not the same as doubling CBoth double τ; only doubling R halves the initial current for fixed Vs.Doubling R is not the same as doubling CCaseτI at t = 0BaselineRCVs / R2 × R2RCVs / (2R)2 × C2RCVs / RPredict both voltage and current before using the tool.Original conceptual teaching figure · not a footprint / pinout / measurement
Both double τ; only doubling R halves the initial current for fixed Vs. Original conceptual illustration; not a measured result or package drawing.

The browser tool accepts the step voltage, resistance in ohms, capacitance in microfarads and elapsed time in milliseconds. It reports the time constant, predicted capacitor voltage, series current and stored energy at that instant. All calculations remain local. There is no automatic quotation or data upload. A copyable summary includes the model limits so the result can be reviewed without relying on a screenshot alone.

Static reference: 5 V,10000 Ω,10 µF,t=100 ms →τ=100 ms,Vc≈3.161 V. Interactive fields appear when JavaScript is available. The explanation and example remain readable without it.

First predict what happens when resistance doubles. The time constant doubles and initial current halves. At the same absolute time, the capacitor has reached a smaller fraction of the step. Now restore resistance and double capacitance. The time constant also doubles, but the initial current is unchanged because initial resistor voltage and resistance have not changed. These two changes can stretch the voltage curve similarly while leaving different current behavior at startup.

Set time to zero and verify the explicit initial condition. Then enter one, two and five time constants, converting seconds to milliseconds if necessary. A long time relative to tau brings voltage close to the step but does not turn this ideal calculation into a leakage measurement. The display rounds numbers for readability; the mathematical model does not know the actual component tolerance, effective capacitance or instrument uncertainty.

Stored energy is one half C Vc squared in the ideal model. That output concerns energy present at the selected instant. It is not a promise about deliverable load energy, converter efficiency or safe handling. Energy dissipated in the charging path is a different accounting question. If you later use a capacitor to support a load during a supply interruption, define the allowed voltage drop, load behavior and conversion losses rather than reusing the charging time as a runtime estimate.

The tool rejects zero or negative R and C, blank values and non-finite inputs. Those are arithmetic protections. It cannot establish that an entered capacitor is suitable for a supply voltage. Treat a plausible curve as a reason to continue checking, not a release certificate. A useful exercise is to write the next missing fact beside each output: effective capacitance, source impedance, load current or temperature-dependent behavior.

A real ceramic part changes the assumptions

A real capacitor is more than CConceptual equivalent elements; values and relevance need exact evidence.A real capacitor is more than CESRESLCEffective C may depend on bias and temperature.No generic percentage loss is assumed.Original conceptual teaching figure · not a footprint / pinout / measurement
Conceptual equivalent elements; values and relevance need exact evidence. Original conceptual illustration; not a measured result or package drawing.

A nominal capacitance is measured under specified conditions. Certain ceramic dielectric classes can change effective capacitance with applied DC bias. Murata's ceramic-capacitor FAQ on DC bias (source address: https://www.murata.com/support/faqs/capacitor/ceramiccapacitor/char/0005) distinguishes this behavior from temperature-compensating types. The important design habit is to inspect the exact series and value under the relevant conditions, rather than assigning one generic percentage loss to every ceramic capacitor.

For the RC calculation, a smaller effective capacitance changes the time constant. If the effective value also varies substantially during the voltage swing, a fixed-C exponential becomes an approximation rather than an exact prediction. You need the appropriate device model or a bounded analysis to describe that situation. Substituting a guessed effective value may be useful for sensitivity exploration, but the assumption should remain visible.

Equivalent series resistance, equivalent series inductance and leakage describe additional departures from the ideal model. Their importance depends on the task. A slow charging illustration, a fast load transient and an RF signal path do not emphasize the same terms. A capacitor that appears adequate by nominal capacitance can have unsuitable impedance at a frequency of interest. Likewise, a high voltage rating alone does not determine dielectric behavior, thermal conditions or assembly suitability.

Decoupling an FPGA supply is therefore a different engineering task from this isolated RC example. Device guidance, power-distribution impedance, regulator behavior, mounting inductance and the board stack-up can all matter. Use the power-rail planner to organize rail requirements and read the XC7A200T-2FBG484C device evidence if that exact FPGA is part of the project. Neither page validates a capacitor network derived from this lesson.

Physical layout matters because current travels through connections as well as through the part. A conceptual capacitor drawing is not a footprint. Check land pattern, body dimensions, termination type and the assembler's process evidence. Avoid drawing a universal conclusion from package size alone. When a value changes during substitution, reconsider the circuit assumption that used it, even if the replacement fits the same two pads.

Keep the lesson, measurement and quotation scopes separate

Initial condition changes the equationThe same RC scale supports different charging and discharging states.Initial condition changes the equationt / τNormalized VCharge from 0Discharge from V0Original conceptual teaching figure · not a footprint / pinout / measurement
The same RC scale supports different charging and discharging states. Original conceptual illustration; not a measured result or package drawing.

Discharging gives a useful comparison with charging. Disconnect the ideal source and give an initially charged capacitor a defined resistor path. In the simple no-load model, capacitor voltage is its initial voltage multiplied by exp of negative time divided by RC. The same time constant appears, but the initial condition and final value differ. Do not insert a zero supply into the charging calculator and expect it to remember a previous charge; its stated initial condition is zero.

Suppose the 10-microfarad ideal capacitor begins at 5 volts and discharges through 10000 ohms. After 100 milliseconds, the ideal voltage is approximately 1.839 volts. The initial stored energy is 125 microjoules. These independently calculated values describe only that low-energy mathematical example. They do not tell you how long an unknown charged assembly remains hazardous or establish a discharge procedure for equipment with other storage and supply paths.

Charging and discharging also show why final voltage alone is not a complete observation. Two measurements can show the same capacitor voltage but have different current directions, depending on the surrounding circuit and history. Keep the state sequence in the record. A supply that was removed, a switch that changed position or an attached load may matter as much as the nominal component value. Reconstructing that history is often a better first troubleshooting step than replacing a capacitor.

For a beginner's notebook, use separate columns for initial condition, circuit connections, elapsed time and predicted result. Add the actual observation in another column only when it has been measured. This makes it possible to compare the model with evidence without retrospectively changing the assumptions to match the result. If the comparison fails, investigate the difference explicitly. A calculated plot should remain a calculated plot even when its shape resembles a familiar laboratory trace.

If you plan a low-voltage classroom measurement, prepare the circuit and instrument method before applying power. Confirm initial conditions and use the actual instrument instructions. A probe or meter can load the circuit, especially when resistance is large. An observed time response includes the source and measurement system, not just an isolated component. Do not work on charged or high-energy assemblies using this introductory lesson as a safety procedure.

Record the intended R and C, their identities, the applied step and the measured nodes. Keep a model curve labelled calculated and measured data labelled with the setup and date. A disagreement should lead to a specific question: was the capacitor initially charged, was the source ideal enough, did the instrument load the node, or did effective capacitance differ? Changing several assumptions at once makes the experiment harder to interpret.

For an existing board, the PCB bring-up checklist provides a broader evidence workflow. The engineering tool directory helps select a later task rather than forcing every supply question into an RC calculator. Before assembly, use the PCBA preparation guidance to keep BOM, drawing and assembly revisions aligned. None of these links establishes that a particular capacitor is in stock or approved for a new circuit.

To obtain components, send a component quotation request with the exact MPN, quantity and required delivery details. To obtain a board or project estimate, submit the project scope through Contact, including board quantity, BOM revision, sourcing split and test requirements. If the capacitor remains undecided, identify the unresolved electrical conditions instead of guessing a procurement code.

The lesson should leave you with a useful distinction: the ideal model explains a response; device evidence decides whether its assumptions fit; a measurement tests a defined setup; and a quotation describes a supply or project scope. Keeping those boundaries clear makes a simple charging example valuable well beyond the first calculation, while preventing a neat animation from being mistaken for a completed design.

Show the original capacitor concept animationOriginal calculated capacitor teaching sequence with declared simplified assumptions

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.

Back to Tutorials →

Related evidence and next tasks

Continue the task with the relevant source scope. These reading paths do not establish interchangeability or current supply.

Source addresses and citation trail

Full source addresses are retained below. Read each citation with its document revision, page, table and applicability conditions above.

  1. https://commons.wikimedia.org/wiki/File:Electronic_component_capacitors.jpg
  2. https://creativecommons.org/licenses/by-sa/3.0/
  3. https://openstax.org/books/college-physics-2e/pages/21-6-dc-circuits-containing-resistors-and-capacitors