A capacitor is used in electrical and electronic circuits to store energy in an electric field, block direct current (DC) while allowing alternating current (AC) to pass, smooth power supply ripple, and decouple high-frequency noise. At a bench level, if you are designing an ESP32 circuit and the module brownouts every time the Wi-Fi radio transmits, it is because the 330mA current spike is collapsing the local 3.3V rail. You fix this by placing a 10µF bulk capacitor and a 100nF decoupling capacitor as close to the VCC pin as possible to supply that instantaneous current demand.

While the basic physics of a capacitor is straightforward—two conductive plates separated by a dielectric—selecting the right physical component for a specific job requires understanding dielectric materials, equivalent series resistance (ESR), and temperature coefficients. Below is a practical guide to selecting, reading, and troubleshooting the physical parts on your workbench.

Which Capacitor Type for Which Job?

Not all capacitors are interchangeable. A 100nF ceramic capacitor and a 100nF film capacitor will behave vastly differently in an RF oscillator versus an audio crossover. The selection criteria depend heavily on the dielectric material, which dictates the temperature coefficient (tempco), tolerance, and parasitic ESR.

Type / Dielectric Construction Tolerance Tempco / Stability Typical Use & Selection Criteria
MLCC (X7R / X5R) Barium titanate ceramic layers ±10% to ±20% ±15% capacitance shift over temp/voltage General decoupling, bypassing. Choose when: You need high capacitance in a tiny SMD footprint for power rail noise suppression.
MLCC (C0G / NP0) Calcium zirconate ceramic ±1% to ±5% ±30ppm/°C (Highly stable) RF matching, precision oscillators, PLL filters. Choose when: Timing or resonance frequency must not drift with temperature or applied DC bias.
Aluminum Electrolytic Etched aluminum foil + liquid electrolyte -20% to +80% High drift, sensitive to cold Bulk power supply filtering, audio coupling. Choose when: You need massive capacitance (100µF to 10,000µF) at low cost to smooth low-frequency ripple.
Tantalum (MnO2) Tantalum pellet + manganese dioxide ±10% to ±20% Moderate stability, low microphonics Compact bulk storage on dense PCBs. Choose when: You need stable bulk capacitance in a low-profile SMD package, but never use on high-ripple switching rails without derating voltage by 50%.
Film (Polypropylene) Metallized polypropylene film ±1% to ±5% Excellent, very low dielectric absorption Audio crossovers, AC line snubbers, high-voltage resonant tanks. Choose when: Signal purity, high AC voltage handling, or self-healing properties are required.

Decoding Physical Markings and Capacitor Codes

Reading capacitor values is notoriously inconsistent across form factors. Manufacturers use different coding schemes depending on the physical size of the part. For a deeper look at fundamental component behavior, the All About Circuits textbook chapter on capacitors provides excellent baseline theory.

1. The 3-Digit Ceramic Code

Through-hole ceramic disc and small SMD MLCCs rarely have enough surface area for printed text. Instead, they use a three-digit multiplier code based on picofarads (pF).

  • First two digits: The significant figures.
  • Third digit: The multiplier (number of zeros to add).
  • Example: A capacitor marked 104 is 10 followed by 4 zeros = 100,000 pF. This converts to 100 nF or 0.1 µF.
  • Example: A capacitor marked 472 is 47 followed by 2 zeros = 4,700 pF = 4.7 nF.

2. Electrolytic and Film Direct Printing

Through-hole electrolytics (like the Panasonic FR or Nichicon PW series) print the value directly: 470µF 25V. The critical marking here is the negative stripe on the can, which indicates the cathode. Reversing polarity on an aluminum electrolytic will cause the internal electrolyte to boil and vent violently.

3. SMD Tantalum and Large MLCC Codes

Larger SMD parts often use a microfarad (µF) base code. A Kemet T491 tantalum marked 476 means 47 followed by 6 zeros in pF, which equals 47µF. The letter or number next to it (e.g., 16 or C) indicates the voltage rating (16V). Always verify the voltage rating against the manufacturer datasheet, as a 10V SMD ceramic and a 50V SMD ceramic of the same size and capacitance look identical but will fail catastrophically if swapped.

Visual Failure Modes and Safe Substitution Rules

Safety Warning: Never test an unknown electrolytic capacitor with a standard multimeter resistance check if it has been in-circuit. Always safely discharge large capacitors using a high-wattage bleed resistor (e.g., 1kΩ 5W) before handling, and wear safety glasses when inspecting bulging aluminum cans.

Capacitors fail in distinct ways depending on their chemistry. According to Cornell Dubilier's technical papers on electrolytic degradation, heat and ripple current are the primary killers of power supply capacitors.

Aluminum Electrolytic: Drying and Venting

  • Visual Symptoms: The rubber bung at the bottom is pushed out, leaving a crusty brown residue on the PCB. Alternatively, the scored 'X' on the top dome has popped open, or the top is visibly bulging.
  • Electrical Symptom: Equivalent Series Resistance (ESR) skyrockets. A 1000µF cap might still measure 950µF on a cheap multimeter, but its ESR has jumped from 0.05Ω to 5Ω, rendering it useless for filtering switching noise.

MLCC Ceramic: Flex Cracking

  • Visual Symptoms: Often invisible to the naked eye. Kemet technical publications detail how board flexure during PCB separation or connector insertion causes microscopic cracks in the ceramic layers.
  • Electrical Symptom: Intermittent short circuits or a dead short across the power rail. If your 3.3V rail measures 0Ω to ground and the regulator is shutting down, probe the MLCCs with an ESR meter or desolder them to find the cracked culprit.

Tantalum: Thermal Runaway

  • Visual Symptoms: A scorched, blackened PCB pad, a popped or melted epoxy case, and the distinct smell of burning plastic. Tantalum capacitors fail short and do not have a self-healing mechanism.
  • Electrical Symptom: Dead short. Often caused by a voltage spike exceeding the rated DC limit, or reverse polarity installation.

Rules for Safe Substitution

When you are out of the exact part, follow these substitution rules to avoid damaging your circuit:

  1. Voltage Rating: Always substitute with an equal or higher voltage rating. Swapping a 16V cap for a 25V cap is perfectly safe and often improves longevity. Never substitute a lower voltage rating.
  2. Capacitance (Decoupling): For power rail decoupling, you can generally substitute a higher capacitance (e.g., using 220nF instead of 100nF). However, going too high can cause inrush current issues or interfere with regulator phase margins.
  3. Capacitance (Timing/Filters): For 555 timer oscillators, RC low-pass filters, or RF matching networks, the capacitance must be exact. Substituting a 10nF for a 15nF will shift your cutoff frequency or oscillator pitch.
  4. Tempco and ESR: Never substitute an X7R ceramic for a C0G/NP0 in a crystal oscillator load circuit; the frequency will drift as the board warms up. Never substitute a standard electrolytic for a 'Low ESR' or 'Polymer' cap in a switching buck converter output; the high ripple current will overheat and destroy the standard cap within hours.

Frequently Asked Questions

Why is a capacitor used in parallel with a DC motor?

A capacitor (typically a 100nF ceramic) is placed directly across the brushes of a small DC motor to act as a snubber. As the carbon brushes commutate against the armature, they create high-frequency voltage spikes and RF noise. The capacitor provides a low-impedance path for this high-frequency noise, preventing it from radiating through the motor leads and interfering with sensitive microcontroller logic or radio receivers on the same chassis.

Why is a capacitor used after a bridge rectifier?

A bridge rectifier converts AC to pulsating DC, meaning the voltage drops to zero 120 times a second (on a 60Hz grid). A large bulk electrolytic capacitor (e.g., 2200µF) is placed immediately after the rectifier to store charge during the voltage peaks and discharge it into the load during the valleys. This 'smooths' the pulsating DC into a relatively flat DC voltage, reducing the ripple that the downstream linear or switching regulator must handle.

Why is a capacitor used in series with an audio tweeter?

In a passive speaker crossover, a non-polarized film capacitor is placed in series with the tweeter to create a first-order high-pass filter. Because a capacitor's reactance ($X_c = 1 / (2\pi fC)$) decreases as frequency increases, it blocks low-frequency bass energy (which could physically destroy the delicate tweeter voice coil) while allowing high-frequency treble to pass through unimpeded.

Why is a capacitor used on a microcontroller reset pin?

A small capacitor (often 100nF) is placed between the reset (EN/RST) pin and ground to form an RC delay circuit with the pull-up resistor. This holds the reset line low just long enough for the microcontroller's internal power rails to stabilize and the oscillator to lock before releasing the chip to begin executing code. It also acts as a debouncing filter, preventing static discharge or noise spikes from accidentally triggering a system reset.