The capacitor discharging graph plots the exponential decay of voltage over time as a capacitor releases its stored electrical energy through a resistive load. The governing mathematical model is V(t) = V0 × e-t/RC, where V0 is the initial voltage, R is resistance in ohms, C is capacitance in farads, and t is time in seconds. In textbook theory, it takes exactly five time constants (5τ, where τ = R × C) for the capacitor to discharge to 0.7% of its initial voltage. However, on the workbench, parasitic elements like Equivalent Series Resistance (ESR) and dielectric absorption warp this ideal curve. Understanding these deviations is critical for designing reliable timing circuits, power supply filters, and audio crossovers.
The Ideal Capacitor Discharging Graph vs. Reality
To understand real-world deviations, we must first establish the baseline. If you charge a 100μF capacitor to 10V and discharge it through a 1kΩ resistor, your time constant (τ) is 0.1 seconds (100 × 10-6 F × 1000 Ω). According to the ideal capacitor discharge equations, the voltage will drop to 3.68V at 0.1s, 1.35V at 0.2s, and effectively 0V (0.07V) at 0.5s (5τ).
When you probe this circuit with an oscilloscope, the graph rarely matches the textbook perfectly due to two main parasitics:
- Equivalent Series Resistance (ESR): Every physical capacitor has internal resistance. At the exact moment discharge begins (t=0), the initial current surge creates an instantaneous voltage drop across the ESR (V = I × ESR). On your graph, this appears as a sharp vertical step-down before the smooth exponential curve begins.
- Dielectric Absorption (DA): Also known as "battery memory" in capacitors, DA occurs when the dielectric material absorbs some charge and releases it slowly. If you short the capacitor to 0V and then remove the short, the voltage will rebound. On a discharging graph, this creates a "tail" where the voltage asymptote flattens out above true zero.
Capacitor Types and How They Alter the Discharge Curve
Selecting the right component dictates how closely your physical circuit will mimic the ideal capacitor discharging graph. Below is a comparison of common types, detailing how their physical construction impacts the RC decay curve.
| Type | Construction / Dielectric | Tolerance | Tempco (Temp. Coefficient) | Discharge Graph Behavior (ESR/DA) | Typical Use |
|---|---|---|---|---|---|
| Aluminum Electrolytic | Etched aluminum foil, liquid/polymer electrolyte | ±20% (M) | Wide variance, poor at extremes | High ESR causes large t=0 voltage step; high DA causes significant rebound tail. | Bulk power supply filtering, low-frequency coupling. |
| MLCC (X7R / X5R) | Multi-layer ceramic, barium titanate | ±10% to ±20% | ±15% over temp range | Very low ESR (ideal t=0 step); moderate DA; severe capacitance drop under DC bias warps the curve shape. | Decoupling, general-purpose bypass, mid-range filtering. |
| MLCC (C0G / NP0) | Multi-layer ceramic, paraelectric materials | ±5% (J) or better | ±30ppm/°C (near zero) | Near-zero ESR and virtually no DA. Matches the ideal mathematical graph almost perfectly. | Precision timing (555 circuits), RF resonant tanks, audio signal path. |
| Film (Polypropylene) | Metallized plastic film, wound or stacked | ±5% to ±10% | Stable, predictable | Extremely low DA and low ESR. The discharge curve is exceptionally clean with no rebound tail. | High-fidelity audio crossovers, snubber circuits, high-voltage timing. |
| Tantalum | Porous tantalum anode, manganese dioxide/polymer cathode | ±10% to ±20% | Moderate stability | Lower ESR than aluminum electrolytic, but higher than MLCC. Moderate DA. Prone to catastrophic failure if reverse-biased. | Space-constrained power rail filtering, medical/aerospace hold-up caps. |
Which type for which job? If your circuit relies on the exact shape of the capacitor discharging graph for timing (like a monostable multivibrator or an RC oscillator), you must use C0G/NP0 ceramics or Film capacitors. If you are simply smoothing a rectified DC power rail where the exact microsecond decay rate doesn't matter, Aluminum Electrolytics offer the highest capacitance-per-dollar ratio.
Decoding Physical Markings to Calculate Your RC Curve
To predict your discharging graph, you need the exact capacitance value. While large electrolytics print their values plainly (e.g., "470μF 50V"), surface-mount and small radial ceramics use the EIA 3-digit code system. Misreading these codes is the most common reason a built circuit's timing graph fails to match the simulation.
- The 3-Digit Code: The first two digits are the significant figures, and the third digit is the multiplier (number of zeros) in picofarads (pF). A marking of 104 means 10 followed by four zeros: 100,000 pF, which equals 100 nF or 0.1 μF.
- Tolerance Letters: A letter following the numeric code dictates the worst-case deviation from your calculated graph. J = ±5%, K = ±10%, M = ±20%. If you are designing a 1-second delay using an M-tolerance capacitor, your actual discharge time could range from 0.8s to 1.2s.
- Voltage Ratings: Often denoted by a letter/number combo or printed directly (e.g., 50V). Never exceed this. For MLCCs, you must also account for DC bias derating; a 10μF X7R capacitor rated for 50V might only exhibit 2μF of actual capacitance when 25V is applied, drastically accelerating your discharge curve.
Failure Modes: When the Discharge Graph Breaks Down
When a capacitor degrades, the oscilloscope trace will immediately reveal the anomaly. Here is how to correlate visual symptoms on the board with changes in the discharging graph.
- Electrolytic Drying Out: Visual Symptom: Bulging top vent, crusty brown electrolyte residue on the PCB, or a shrunken heat-shrink sleeve. Graph Effect: The total capacitance drops significantly, causing the 5τ discharge time to occur much faster than calculated. Simultaneously, ESR spikes, creating a massive vertical voltage drop at t=0.
- MLCC Flex Cracking: Visual Symptom: Hairline fracture near the PCB mounting holes or board edges where mechanical flex occurs. Often invisible to the naked eye without a magnifying loupe. Graph Effect: The crack bridges the internal layers, causing a dead short. The discharging graph becomes a flat line at 0V instantly, often tripping the power supply's overcurrent protection.
- Tantalum Thermal Runaway: Visual Symptom: Scorch marks on the PCB, cracked epoxy casing, or a distinct burnt smell. Tantalums fail short-circuit and can ignite if the power supply can deliver enough current. Graph Effect: Similar to a cracked MLCC, the voltage instantly collapses to zero, usually destroying the trace or the driving IC in the process.
Safe Substitution When the Exact Part is Missing
When you are prototyping or repairing a board and the exact BOM part is out of stock, you must substitute without ruining the circuit's timing or filtering characteristics. According to dielectric material guidelines from major manufacturers, follow these rules:
- Voltage Rating: Always substitute with an equal or higher voltage rating. However, if moving to a higher voltage MLCC, check the physical footprint; higher voltage often requires a larger package size (e.g., moving from 0805 to 1206).
- Capacitance Value in Timing Circuits: Do not substitute capacitance values in RC oscillators, 555 timer circuits, or active filters. The discharging graph is the core timing mechanism; changing C changes the frequency or pulse width directly.
- Capacitance Value in Power Filtering: You can generally substitute a higher capacitance value for bulk decoupling or power supply smoothing. A larger C extends the discharge curve, reducing ripple voltage. However, be aware that a massively increased capacitance will spike the inrush current at startup, which may blow upstream fuses or trip soft-start circuits.
- Dielectric Swaps: Never swap a C0G/NP0 ceramic for an X7R/X5R in precision timing or audio signal paths. X7R materials exhibit piezoelectric effects (microphonics) and severe capacitance shifts with temperature and applied voltage, which will introduce distortion and timing jitter into your discharge curve.
Frequently Asked Questions
Why does my capacitor discharging graph show a voltage rebound after shorting?
This is caused by dielectric absorption (DA). When a capacitor is charged, the dielectric material's molecular dipoles align. During a rapid discharge, some dipoles lag behind and do not immediately release their stored energy. Once the external short is removed, these lagging dipoles slowly relax, pushing a small amount of charge back onto the plates. This is why high-voltage capacitors can deliver a dangerous shock hours after being discharged; always use a permanent bleed resistor in high-voltage designs.
How to calculate the resistor for a specific capacitor discharging graph time?
If you need the capacitor to discharge to a specific voltage within a specific time, rearrange the standard formula: R = -t / (C × ln(V(t) / V0)). For example, to discharge a 10μF capacitor from 12V down to 2V in exactly 0.5 seconds, the math yields R = -0.5 / (0.00001 × ln(2/12)) = 27,890 Ω. You would select the nearest standard 1% resistor value, which is 28.0kΩ.
Does a capacitor discharging graph ever reach absolute zero volts?
Mathematically, the exponential decay curve is asymptotic, meaning it approaches zero but never truly reaches it. In practical, real-world terms, the voltage drops below the noise floor of your measurement equipment (usually in the microvolt range) after about 5 to 7 time constants. At that point, parasitic leakage currents through the dielectric and the PCB surface contamination dominate, effectively bleeding off the remaining fractional charge.
How does DC bias affect an MLCC capacitor discharging graph?
Class II ceramics (like X7R and X5R) suffer from DC bias derating. As the voltage across the capacitor drops during the discharge cycle, its actual capacitance increases back toward its nominal rated value. This means the time constant (τ = RC) is not static; it expands as the capacitor discharges. On an oscilloscope, this warps the discharging graph, making the tail end of the curve decay much slower than the initial steep drop, deviating heavily from the ideal straight-line logarithmic plot.
