Electrical Engineering › Section O · Pulses, nonsinusoidal signals, and system analysis › Topic O.1

Pulse waveforms, repetition rate, duty cycle, and average value

Clocks in computers, data on a cable and the drive to a dimmed LED are not sine waves but pulses. A handful of measurements, from rise time to duty cycle, describes any pulse waveform, and one formula gives its average value.

Why pulses matter

Much of modern electronics runs on pulses rather than sinusoids: the clock that paces a microprocessor, the bits on a network cable, radar timing and the switching waveforms that control motors and LED brightness. A pulse jumps from one level to another, stays a while, then returns. Describing it needs its own vocabulary: how fast the edges are, how wide the pulse is, how often it repeats, how much of each period it fills, and what dc level it averages to.

The ideal pulse

An ideal pulse has upright edges, square corners and a level top. It starts from a steady base-line voltage \(V_b\), which need not be zero, jumps instantly, and later drops back just as abruptly. For a positive-going pulse the rising edge comes first and is the leading edge, and the falling one is the trailing edge. The amplitude is the size of the jump, measured from the base line. A pulse that rises in the positive direction from its base line is positive-going, even if the base line is negative; one that moves in the negative direction is negative-going. For an ideal pulse, the pulse width is simply the time between its two edges.

Real edges: rise time, fall time and pulse width

Capacitance and inductance in any real circuit resist instant changes, so actual edges slope and their corners round off. Because the start and end of a rounded edge are hard to pin down, the rise time \(t_r\) is defined as the time to go from 10% to 90% of the amplitude, measured from the base line. The fall time \(t_f\) is the time to go from 90% back down to 10%; it need not equal \(t_r\). The pulse width \(t_p\) of a real pulse is the time from the leading edge's 50% crossing to the trailing edge's 50% crossing.

Tilt, overshoot and ringing

A pulse whose top slopes downward from \(V_1\) at the start to \(V_2\) at the end shows tilt, also called sag:

\%\,\text{tilt} = \frac{V_1 - V_2}{V} \times 100\%

where \(V = \frac{V_1 + V_2}{2}\), the average top level, is taken as the amplitude. Tilt usually means the system passing the pulse has a poor low-frequency response, for example a coupling capacitor that is too small. Overshoot is a brief rise above the final level just after the leading edge, and preshoot a small opposite swing just before it; both point to high-frequency effects. Ringing is a decaying oscillation after an edge, as inductance and capacitance exchange energy at resonance.

Worked example: reading a real pulse

An oscilloscope shows a positive-going pulse on a 0 V base line. The leading edge passes 10% of the amplitude at 2.0 µs and 90% at 2.6 µs, and crosses the 50% level at 2.3 µs. The top sags from 5.2 V to 4.8 V, and the trailing edge crosses 50% at 12.5 µs. 1. Amplitude: \(V = \frac{5.2 + 4.8}{2} = 5.0\,\text{V}\). 2. Rise time: \(t_r = 2.6 - 2.0 = 0.6\,\mu\text{s}\). 3. Pulse width, between the 50% points: \(t_p = 12.5 - 2.3 = 10.2\,\mu\text{s}\). 4. Tilt: \(\frac{5.2 - 4.8}{5.0} \times 100\% = 8\%\). 5. Check: here the 10%, 50% and 90% levels are 0.5 V, 2.5 V and 4.5 V above the base line.

Pulse trains and repetition rate

A series of pulses is a pulse train; in data signals the widths or heights vary and carry information. If the pattern repeats exactly, the train is periodic, and its period T is the time between any two matching points, such as successive leading edges at the 50% level. The pulse repetition frequency, prf, also called the pulse repetition rate, is

\text{prf} = \frac{1}{T}

measured in hertz or pulses per second. The prf depends only on the period, not on the shape or width of the pulses: a narrow spike and a broad pulse repeating every 1 ms both have a prf of 1 kHz.

Duty cycle

The duty cycle states what fraction of each period the pulse occupies:

D = \frac{t_p}{T} \times 100\%

A square wave, high for half of each period, has a 50% duty cycle. Pulse-width modulation varies the pulse width at a fixed prf to control the average delivered to a load, as the later topic Pulse-width modulation, DC motor drives, inverters, and AC motor drives explains. In calculations, D is used as a decimal: 40% becomes 0.40.

Worked example: prf and duty cycle from a scope

A scope is set to 5 µs and 2 V per division. One period of a pulse train spans 4.0 divisions, the pulse width (at 50%) 1.6 divisions, and the height 3.0 divisions above a 0 V base line. 1. Period: \(T = 4.0 \times 5\,\mu\text{s} = 20\,\mu\text{s}\). 2. Repetition rate: \(\text{prf} = \frac{1}{20\,\mu\text{s}} = 50\,\text{kHz}\). 3. Pulse width: \(t_p = 1.6 \times 5\,\mu\text{s} = 8\,\mu\text{s}\). 4. Duty cycle: \(D = \frac{8}{20} \times 100\% = 40\%\). 5. Amplitude: \(3.0 \times 2\,\text{V} = 6\,\text{V}\). Check: the pulse fills 1.6 of 4.0 divisions, which is 40% of the period.

Worked example: average of a pulse train

For an ideal periodic pulse train, the area method of the topic Average and rms values reduces to a short formula:

V_{av} = D\,V_{peak} + (1 - D)\,V_b

with D as a decimal, \(V_{peak}\) the pulse's top level and \(V_b\) the base line, both measured from 0 V. Find the average of a train with a 25% duty cycle that switches between a base line of −2 V and a top of +8 V. 1. Substitute: \(V_{av} = 0.25(8) + 0.75(-2)\). 2. Evaluate: \(2.0 - 1.5 = 0.5\,\text{V}\). 3. Check by area, taking T = 1 ms: \(\frac{(8)(0.25) + (-2)(0.75)}{1} = 0.5\,\text{V}\). The amplitude is 10 V, yet the average is only 0.5 V. The formula ignores tilt and slow edges.

Measuring the average with an oscilloscope

With a scope's input coupling set to ac, an internal capacitor blocks the dc component, so the trace settles with its average on the screen's zero line; note where that line is. Switching the coupling to dc then shifts the whole trace up or down by the average value, which can be read off in divisions. The same idea explains pulse-width control. A 12 V supply switched at a 30% duty cycle onto a 0 V base line gives an average of \(0.30 \times 12 = 3.6\,\text{V}\), and a slow load such as a motor responds mainly to that average.

Common mistakes to avoid

First, measuring rise time from 0% to 100%: it runs from 10% to 90% of the amplitude. Second, measuring pulse width at the base or the top of a real pulse instead of at the 50% points. Third, measuring levels from 0 V when the base line is not zero: percentages are taken of the amplitude above the base line. Fourth, using the duty cycle as a percentage in the average formula: 25% must enter as 0.25. Fifth, thinking the prf depends on pulse width: it depends only on the period.

Key takeaways

An ideal pulse has vertical edges and a flat top; it starts from a base line V_b, and its amplitude is measured from that base line. Rise and fall times run between the 10% and 90% levels; pulse width is taken at 50%. Tilt = (V₁ − V₂)/V × 100%, with V the average top level; overshoot, preshoot and ringing are other distortions. A periodic train has period T and prf = 1/T, whatever the pulse shape. Duty cycle D = t_p/T × 100%; a square wave has 50%. The average value is V_av = D·V_peak + (1 − D)·V_b, with D as a decimal.

Flashcards

Flashcard

Base-line voltage V_b

The steady level a pulse starts from and returns to

It need not be 0 V.

Test your understanding

Question 1 of 5

A pulse on a 0 V base line has an amplitude of 10 V. Its leading edge reaches 1 V at 3.0 µs and 9 V at 3.8 µs. What is its rise time?

Question 2 of 5

A pulse train has a period of 250 µs and a pulse width of 40 µs. What are its prf and duty cycle?

Question 3 of 5

A periodic pulse train switches between a base line of −1 V and a top of +10 V, with a 20% duty cycle. What is its average value?

Question 4 of 5

Two pulse trains both repeat every 2 ms. One has narrow 0.1 ms pulses, the other broad 1.5 ms pulses. How do their pulse repetition frequencies compare?

Question 5 of 5

A pulse's top slopes from 6.3 V to 5.7 V. Using the average top level as the amplitude, what is its percentage tilt?

← Previous topicNext topic →