What the 555 is actually timing
The 555 does one thing: it watches the voltage on a capacitor and flips its output when that voltage crosses fixed fractions of the supply. Inside the chip an internal divider of three equal resistors creates two references, one at one third of the supply and one at two thirds. A comparator on the threshold pin trips when the capacitor climbs past two thirds, and a comparator on the trigger pin trips when it falls past one third. Between those two trip points a flip-flop holds the output high or low and switches a transistor that can short the capacitor to ground through the discharge pin.
Every 555 timing equation is therefore an RC charging problem with the endpoints already chosen for you. The chip does not care what the supply voltage is, because both references scale with it: a circuit timed at 5 V keeps the same period at 12 V. That supply independence is why the part survived from 1972 into current production, and it is why you can size the components on paper and expect the bench to agree.
Two wiring patterns account for almost all uses. In the astable arrangement the capacitor cycles endlessly between one third and two thirds of the supply, giving a square wave whose frequency you set with two resistors and a capacitor. In the monostable arrangement the capacitor sits discharged until a trigger pulls the trigger pin low, then charges once from zero to two thirds of the supply and stops, giving one output pulse of a length you choose.
Where 0.693 and 1.1 come from
Both constants are logarithms, and knowing which logarithm tells you immediately why the astable and monostable formulas differ.
A capacitor charging through a resistor toward a supply V follows v(t) = V(1 − e−t/RC). In the astable circuit the capacitor never starts from zero after the first cycle: it swings between V/3 and 2V/3. Solve the charging equation for the time to move between those two levels and the supply voltage cancels, leaving t = RC · ln 2. That is where 0.693 comes from, and the cancellation of V is exactly the supply independence described above.
During the charge phase the current reaches the capacitor through R1 and R2 in series, so the high time is tH = ln2 · (R1 + R2) · C. During the discharge phase the discharge pin pulls the junction of R1 and R2 to ground, so the capacitor empties through R2 alone: tL = ln2 · R2 · C. Add them and the period is T = ln2 · (R1 + 2R2) · C, whose reciprocal is the frequency the datasheet writes as 1.44/((R1+2R2)C). The 1.44 is simply 1/ln2 rounded.
The monostable is different because the capacitor starts fully discharged, at zero, and has to climb all the way to two thirds of the supply. Solving 2/3 = 1 − e−t/RC gives t = RC · ln 3, which is 1.0986 · RC. The datasheet rounds it to 1.1·RC, a 0.13% overstatement. This calculator uses the exact logarithms in both modes, and the direction of the difference depends on which quantity you compare. In astable mode its period runs about 0.19% shorter than a 1.44-based tool and its frequency therefore about 0.19% higher; in monostable mode its pulse runs about 0.13% shorter than a 1.1·RC tool, so the retrigger rate it reports is about 0.13% higher. Both gaps sit far inside any real capacitor tolerance, but they explain small disagreements between calculators.
Two structural consequences follow from the shape of the equations. First, the duty cycle (R1 + R2)/(R1 + 2R2) contains no capacitor term, so you tune frequency with C and duty cycle with the resistor ratio, independently. Second, because the numerator always exceeds half the denominator for any positive R1, this classic astable holds its output high for more than half of every cycle. You cannot reach 50% or below by choosing resistors; you have to change the topology. The same logic underlies the RC time constant calculator, which solves the same exponential for arbitrary start and end voltages.
Worked example: a 1 Hz blinker from 10 kΩ, 68 kΩ and 10 µF
You want a roughly one-second LED blink from a 555 in astable mode, and you have a 10 kΩ resistor, a 68 kΩ resistor and a 10 µF electrolytic on the bench. Work it through by hand.
- Convert the capacitor. 10 µF = 10 × 10−6 = 1 × 10−5 F.
- Charge-path resistance. R1 + R2 = 10,000 + 68,000 = 78,000 Ω.
- High time. tH = 0.693147 × 78,000 × 1 × 10−5 = 0.693147 × 0.78 = 0.540655 s.
- Low time. tL = 0.693147 × 68,000 × 1 × 10−5 = 0.693147 × 0.68 = 0.471340 s.
- Period. T = 0.540655 + 0.471340 = 1.011995 s.
- Frequency. f = 1 ÷ 1.011995 = 0.98815 Hz.
- Duty cycle. D = 78,000 ÷ 146,000 = 0.534247 = 53.42%.
Now run it backwards. Suppose you want exactly 1.000 Hz with the same 10 kΩ R1 and the same 10 µF capacitor. The period must be 1 s, so R1 + 2R2 = 1 ÷ (0.693147 × 1 × 10−5) = 144,269.5 Ω. Subtract R1 and halve: R2 = (144,269.5 − 10,000) ÷ 2 = 67,134.8 Ω. The nearest E24 value is 68 kΩ, which is the resistor you already fitted, and that is why the circuit lands at 0.988 Hz rather than 1.000 Hz. The 865 Ω gap between the exact R2 and the 68 kΩ you fitted costs 1.2% in frequency; the 10 µF electrolytic’s own tolerance dwarfs it.
How to read the result and choose components
Start from the capacitor, not the resistors. The capacitor sets the decade of the frequency and the resistors trim within it, so pick the capacitor first from the reference table below, then solve for the resistor. Fixing the capacitor also fixes the practical resistance range, and resistance is where the 555’s real limits live.
Keep the total timing resistance inside roughly 1 kΩ to 1 MΩ. Below about 1 kΩ, the discharge transistor has to sink a current large enough to matter against the device rating, and the supply current rises sharply. Above about 1 MΩ, the threshold pin’s own input current and the capacitor’s leakage stop being negligible next to the charging current, so the measured period runs long and drifts with temperature. Both bounds are engineering rules of thumb rather than absolute limits; the datasheet electrical tables are where you check a specific part.
Read the duty cycle as a design constraint rather than a result. For an indicator LED, anything from 20% to 80% looks fine. For driving a load through PWM, the classic astable’s floor of just above 50% is disqualifying, and you should either add a diode in parallel with R2 so the charge and discharge paths separate, or use the 555 as a fixed-frequency clock and generate the PWM elsewhere.
Finally, treat the calculated frequency as a centre value with a tolerance band around it. The 555’s internal divider is tightly matched, so the chip contributes very little error; almost all of your deviation comes from the passive components. With a 5% film resistor and a 20% electrolytic, expect the frequency to land within roughly a quarter of the nominal value either way. If you need better, use a film or C0G capacitor and a 1% resistor, or trim R2 with a multi-turn preset while you watch the output on a frequency counter. When the square wave then drives a filter or a smoothing stage, size that stage separately with the RC filter cutoff frequency calculator.
Astable frequency for common resistor and capacitor pairs
| Capacitor | R1 = 10 kΩ, R2 = 10 kΩ (66.67% duty) | R1 = 1 kΩ, R2 = 10 kΩ (52.38% duty) | R1 = 10 kΩ, R2 = 100 kΩ (52.38% duty) |
|---|---|---|---|
| 10 nF | 4,809 Hz | 6,870 Hz | 687.0 Hz |
| 100 nF | 480.9 Hz | 687.0 Hz | 68.70 Hz |
| 1 µF | 48.09 Hz | 68.70 Hz | 6.870 Hz |
| 10 µF | 4.809 Hz | 6.870 Hz | 0.6870 Hz |
| 100 µF | 0.4809 Hz | 0.6870 Hz | 0.06870 Hz |
Frequency scales exactly inversely with the capacitor, so every step down a column is a factor of ten. Use the table to choose the decade, then solve for the exact resistor with the calculator.
Mistakes that make a 555 miss its calculated timing
- Using an electrolytic where the tolerance matters. Aluminium electrolytics are commonly specified at ±20% and their capacitance falls with age and temperature. Any period you calculate from one is a nominal figure, not a specification.
- Leaving the control pin floating. Pin 5 taps the internal divider at two thirds of the supply. Without the customary 10 nF decoupling capacitor to ground it picks up noise, and that noise moves the threshold and jitters your period.
- Expecting a duty cycle at or below 50% from the standard astable. The charge path always includes R1, so the high time always exceeds the low time. A diode across R2 lets the charge current bypass it and is the usual fix.
- Forgetting that the CMOS variants are not the bipolar part. A 7555 or TLC555 runs at far lower supply current and tolerates much larger timing resistors, but its output cannot sink or source the hundreds of milliamps a bipolar NE555 can. The timing equations are identical; the drive capability is not.
- Retriggering a monostable before its pulse ends. The standard 555 one-shot ignores triggers while the output is high, so a trigger train faster than the pulse width produces fewer pulses than you expect, not longer ones.
- Omitting the supply decoupling capacitor. The output stage draws a current spike at every transition. Without a capacitor across the supply pins close to the chip, that spike modulates the supply and can cause double-triggering.
Which document these equations come from
The formulas are the ones printed in the Texas Instruments xx555 Precision Timers datasheet, which covers the NE555, SA555 and SE555 and is the current maintained descendant of the original Signetics NE555 data sheet. It gives the astable frequency as 1.44/((RA+2RB)C) and the monostable width as 1.1·R·C. This calculator evaluates the same physics with exact logarithms instead of those rounded constants. The datasheet is also where you should look up the supply-current, discharge-current and output-drive figures for the specific part you are fitting, because those differ between the bipolar and CMOS families even though the timing equations do not.
When a 555 is the wrong tool
Use a 555 when you need a simple, supply-independent oscillator or one-shot built from parts you already have, and when a few per cent of frequency error is acceptable. It is unbeatable for blinking, debouncing, generating a clean reset pulse or clocking a counter at a leisurely rate.
Reach for something else in three situations. If you need frequency accuracy better than about 1%, a crystal oscillator or a microcontroller timer is the right answer, because the 555’s accuracy is capped by the passive components, not the chip. If you need a duty cycle you can sweep from 0% to 100%, a microcontroller PWM peripheral does it in software with no extra parts. And if you need timing above a few hundred kilohertz, the bipolar 555’s propagation delays start to eat into the calculated period; a CMOS variant pushes the practical ceiling higher, but a dedicated oscillator is cleaner.
Once the timing is settled, the rest of the design usually needs current limiting and level setting. Size the indicator resistor with the LED series resistor calculator, set thresholds with the voltage divider calculator, check the supply current draw with the Ohm’s law calculator, and combine capacitors you do not have as a single value with the capacitor series and parallel calculator. If the 555 is clocking an amplifier stage, the op-amp gain calculator sizes the feedback network that follows it.
Key terms
- Astable
- A circuit with no stable state. It flips continuously between high and low output, producing a square wave whose period you set with R1, R2 and C.
- Monostable
- A circuit with one stable state. It rests low until triggered, emits a single pulse of length ln3·R·C, then returns to rest. Also called a one-shot.
- Duty cycle
- The fraction of each period the output spends high, expressed as a percentage. In the standard 555 astable it equals (R1+R2)/(R1+2R2).
- Discharge pin
- Pin 7, the collector of an internal transistor that shorts the junction of R1 and R2 to ground while the output is low, emptying the timing capacitor through R2.
- Control voltage pin
- Pin 5, which exposes the two-thirds-supply reference. Grounding it through a 10 nF capacitor keeps noise off the threshold; driving it deliberately turns the astable into a voltage-controlled oscillator.