Capacitor Code Calculator

How capacitor codes work

When to use this: identifying an unmarked-looking ceramic capacitor from a parts drawer, checking a BOM against what is actually on a board, or finding out what code to look for when you know the value you need.

Ceramic capacitors are too small to print a full value on, so they use a three-digit code: two significant figures followed by a power-of-ten multiplier. The base unit is picofarads, and that is the single most important thing to remember.

104 → 10 × 10⁴ pF = 100,000 pF = 100 nF = 0.1 µF

So 104 is not 104 of anything. It is 100 nF — the decoupling capacitor that sits next to almost every integrated circuit ever manufactured, and quite possibly the most-produced electronic component in history.

Codes of one or two digits are the value directly in picofarads, with no multiplier. A capacitor marked 22 is 22 pF, not 22 nF.

Common codes

The 8 and 9 multipliers

Capacitors below 10 pF cannot use a normal power of ten, so two exponents are redefined: a third digit of 9 means ×0.1 and 8 means ×0.01.

479 → 47 × 0.1 = 4.7 pF
229 → 22 × 0.1 = 2.2 pF
108 → 10 × 0.01 = 0.1 pF

These turn up in RF work and oscillator circuits. If a code ending in 8 or 9 decodes to an absurdly large value, this is why.

Tolerance letters

Note that B, C and D are absolute tolerances in picofarads, not percentages. On a 2 pF capacitor, ±0.25 pF is more than 10% — a percentage figure would be meaningless at those values.

Z deserves particular suspicion. A capacitor marked 105Z could be anywhere from 0.8 µF to 1.8 µF, and that is before temperature and bias effects. It is only acceptable where the exact value genuinely does not matter.

Dielectric classes matter more than tolerance

The tolerance letter tells you the spread at 25 °C with no voltage applied. What it does not tell you is how much the capacitance moves once the part is in a real circuit — and for ceramics that can dwarf the tolerance.

The DC bias problem

This catches people out constantly. Class 2 ceramics — X7R, X5R, Y5V — lose capacitance when a DC voltage is applied across them. A 10 µF X5R rated at 6.3 V can measure under 4 µF with 5 V across it, and the datasheet curve is often buried or omitted entirely.

The practical consequence: if you need 10 µF of actual capacitance at 5 V, specify a part rated well above 5 V, or use a physically larger package. A tolerance letter of K tells you nothing about this — a K-rated X5R can be 50% low in circuit and still be within specification, because the specification is measured with no bias applied.

Other markings you will meet

• Direct µF values. Electrolytics and film capacitors are large enough to print the real value — 10µF 25V, 0.1 µF.
• The n notation. European film capacitors often use 4n7 for 4.7 nF, with the unit letter standing in for the decimal point — the same RKM convention used for resistors.
• The R notation. 4R7 means 4.7 pF on a capacitor, the same way it means 4.7 Ω on a resistor. Context decides the unit.
• Voltage rating. Frequently printed alongside, sometimes coded. Never exceed it — ceramic capacitors fail short, which can be spectacular on a power rail.

Why 100 nF is everywhere

A 104 next to every IC is not superstition. Digital chips draw current in sharp spikes as their outputs switch, and the supply track between the chip and the power supply has inductance. Without a local reservoir, those spikes turn into voltage dips at the chip’s supply pin, which cause logic errors.

100 nF happens to sit in a useful spot: large enough to supply the charge for a switching event, small enough that its self-resonant frequency in a small package stays high enough to be effective at digital switching speeds. Larger capacitors are better at low frequencies but their lead inductance makes them useless at high ones — which is why boards often pair a 104 with a 10 µF bulk capacitor rather than choosing between them.