The direct answer to how much current flows into or out of a capacitor is defined by the relationship between capacitance and the rate of voltage change over time. The fundamental current through capacitor equation is i = C(dv/dt). If the voltage across the capacitor isn't changing, the current is exactly zero. If the voltage changes instantly, the theoretical current approaches infinity—a physical impossibility that causes real-world headaches like blown fuses and pitted switch contacts.
This guide breaks down the calculus and algebra behind the formula, tracks units through solved bench problems, and dissects a real-world failure where ignoring this equation destroyed hardware.
The Core Current Through Capacitor Equation
In calculus terms, the instantaneous current i through a capacitor is the product of its capacitance C and the derivative of the voltage v with respect to time t. For linear voltage ramps (like a DC power supply turning on), we use the algebraic delta form: I = C(ΔV/Δt).
| Symbol | Parameter | Standard SI Unit | Common Bench Units |
|---|---|---|---|
| i or I | Instantaneous or Average Current | Amperes (A) | mA, μA |
| C | Capacitance | Farads (F) | μF, nF, pF |
| dv or ΔV | Change in Voltage | Volts (V) | mV |
| dt or Δt | Change in Time | Seconds (s) | ms, μs, ns |
Rearranged Forms
On the bench, you rarely solve for current alone. Here are the rearranged forms for practical design:
- Solving for Capacitance: C = i / (dv/dt) — Used when sizing a bulk filter cap to limit voltage droop under a known pulsed load.
- Solving for Voltage Change: dv = (i × dt) / C — Used to calculate voltage ripple on a DC rail given a specific load transient.
- Solving for Time: dt = (C × dv) / i — Used to determine how long a hold-up capacitor can sustain a circuit during a power dropout.
Assumptions, Limits, and Realistic Magnitudes
When the Formula Applies (and When It Fails)
The equation i = C(dv/dt) assumes an ideal capacitor. In reality, every physical component has Equivalent Series Resistance (ESR) and Equivalent Series Inductance (ESL).
If your math says dt is 0.000001 seconds (1 μs) and dv is 50V, the equation demands 50,000 Amps for a 1,000μF capacitor. In the real world, the ESR of the capacitor, the resistance of the PCB traces, and the internal resistance of the power supply will bottleneck the current. The formula gives you the theoretical demand; the parasitic elements dictate the actual delivered current. For high-frequency transients (nanosecond edges), ESL dominates, and the capacitor behaves like an inductor, rendering the simple capacitive current equation insufficient without a full RLC impedance model.
Realistic Answer Magnitudes
Knowing what a 'normal' answer looks like prevents decimal-place blunders:
- Signal Coupling (Audio/Data): Nanoamps (nA) to microamps (μA). Small capacitance (pF to nF), slow voltage changes.
- Microcontroller Decoupling: Milliamps (mA). Fast edges (ns), tiny capacitance (100nF).
- Power Supply Inrush: Tens to hundreds of Amps. Large capacitance (1,000μF+), near-instantaneous voltage steps.
Solved Problems: Tracking Units from Bench to Breadboard
The most common way to fail this equation is dropping a prefix. We will track every unit conversion explicitly.
Problem 1: Signal Edge Current
Scenario: A 3.3V microcontroller GPIO pin switches from LOW to HIGH in 4 nanoseconds. The trace has a parasitic capacitance of 15 picofarads to ground. What is the peak instantaneous current required to charge this trace?
- Identify variables: C = 15 pF, dv = 3.3 V, dt = 4 ns.
- Convert to base SI units:
15 pF = 15 × 10-12 F
4 ns = 4 × 10-9 s - Calculate dv/dt: (3.3 V) / (4 × 10-9 s) = 0.825 × 109 V/s
- Multiply by C: i = (15 × 10-12 F) × (0.825 × 109 V/s)
- Final Result: i = 0.012375 Amperes, or 12.4 mA.
Bench Takeaway: 12.4 mA is well within a standard GPIO's 20mA source limit, but if you parallel ten of these traces, you'll exceed the chip's absolute maximum package current.
Problem 2: Sizing a Hold-Up Capacitor
Scenario: An RTC (Real Time Clock) module draws a steady 2 μA. We need it to survive a 500-millisecond power dropout without the voltage dropping more than 0.4V. What capacitance is required?
- Identify variables: i = 2 μA, dt = 500 ms, dv = 0.4 V.
- Convert to base SI units:
2 μA = 2 × 10-6 A
500 ms = 0.5 s - Use rearranged formula: C = i / (dv/dt) which simplifies to C = (i × dt) / dv
- Plug in values: C = (2 × 10-6 A × 0.5 s) / 0.4 V
- Final Result: C = (1 × 10-6) / 0.4 = 2.5 × 10-6 Farads, or 2.5 μF.
Bench Takeaway: A standard 10μF ceramic capacitor (derated for DC bias) will easily cover this requirement with margin.
Real-World Scenario: The Inrush Current Fuse Blower
Formulas on a whiteboard don't spark, but misapplying them on a workbench does. Here is a teardown of a classic power-system failure.
The Setup
A hobbyist is building a 48V DC electric skateboard motor controller. To smooth out battery ripple, they place a massive 4,700 μF electrolytic capacitor bank directly across the main power terminals. They protect the circuit with a standard 15A automotive blade fuse and use a heavy-duty mechanical toggle switch to turn it on.
The Numbers
When the mechanical switch closes, the voltage across the capacitor must jump from 0V to 48V. The time it takes (dt) is limited only by the switch contact bounce and the parasitic resistance of the wiring. Let's assume the wiring and battery internal resistance total 0.05 Ω.
Using Ohm's law for the absolute peak instantaneous current (ignoring inductance for a microsecond): I_peak = 48V / 0.05Ω = 960 Amps.
Even if we use the capacitive equation and assume a highly optimistic ramp time of 1 millisecond (0.001s):
i = 0.0047 F × (48V / 0.001s) = 225 Amps.
The Outcome
The moment the toggle switch is flipped, a loud 'pop' is heard. The 15A fuse blows instantly. Worse, the contacts inside the $12 mechanical toggle switch are permanently welded together and pitted from the extreme arcing. The user is now unable to turn the board off without disconnecting the battery.
What Went Wrong
The builder treated the capacitor as an open circuit without respecting the dv/dt reality of the turn-on transient. A 15A fuse cannot withstand a 225A+ inrush spike, even for a millisecond. According to fundamental capacitor charging principles, the initial uncharged state acts as a dead short.
The Fix: The builder needed a precharge circuit. By placing a 100Ω power resistor in parallel with a relay, or using an NTC thermistor (like the Ametherm SL32 2R015) in series with the main feed, the dt is artificially stretched to roughly 0.5 seconds. This limits the inrush current to under 1A, safely charging the bank before the main low-resistance path is engaged. For deeper design strategies on managing these transients, refer to RC charging network tutorials to calculate exact time constants.
The Unit Conversion Trap: Mistakes That Break the Math
If your calculated current is off by a factor of a million, you fell into the prefix trap. The equation i = C(dv/dt) demands strict adherence to base SI units (Farads, Volts, Seconds). Here are the specific mistakes that break the math:
- The Microfarad/Millisecond Cancellation Myth: Many engineers mistakenly believe that because μ (micro) is 10-6 and m (milli) is 10-3, they can just plug in μF and ms and get a 'scaled' answer. They don't cancel out. 10-6 / 10-3 = 10-3. If you plug 1000 μF and 5 ms directly into the formula without converting, your answer will be off by a factor of 1,000.
- Ignoring the 'k' in Volts: In high-voltage DC systems (like 400V EV battery packs), failing to convert kV to V (or just writing 400 instead of 400.0) is fine, but confusing mV (millivolts) with V will skew your ripple calculations by 1,000x.
- Nanosecond Decimal Slips: 10 ns is not 10 × 10-6 s. It is 10 × 10-9 s. A single missed zero here results in calculating a 100A spike when the reality is a 100mA spike.
