RobotGuide
Labs/PWM and Current Ripple

Lab 9 / 15 · Electronics · computable

PWM and Current Ripple

Change frequency, duty cycle, and inductance, and watch the ripple — higher frequency is not automatically better.

Math: The idea of area integrationNo coding neededNo hardware needed≈ 7 min

Drive parameters

PWM frequency20 kHz
Duty cycle50 %
Phase inductance0.5 mH
Bus voltageselect

Readouts

Current ripple ΔI—
Suggested current-loop bandwidth—
Relative switching loss—
Dead-time share (illustrative)—
Verdict—

An idealized Buck approximation — dead time, switching loss, and magnetic saturation are ignored. Real selection also weighs driver voltage rating, gate-drive capability, and cooling.

What problem this lab solves

«Higher PWM frequency is always better» is the most common misconception. Higher frequency means smaller current ripple and smoother control, but switching losses rise, the dead-time share grows, and the driver heats up. This experiment turns ΔI = V·D(1−D)/(L·f) into a curve you can drag.

How to read the figure

Top is the bridge's PWM; bottom is the phase current. The current is not a square wave — it is a triangle wave rounded by the inductance. Double the frequency and ripple halves; double the inductance and ripple halves too — but inductance is bounded by volume and cost, frequency by switching loss.

Hands-on

Fix 24 V, 50% duty, 0.5 mH, and sweep frequency from 1 kHz to 40 kHz — watch ripple and the «suggested current-loop bandwidth» change. Then squeeze duty to 5% and see why ripple shrinks instead of growing.

Related topics

Next labs (in play-learn order)