Controlled Impedance Basics

Controlled impedance means designing a trace so its characteristic impedance hits a defined target — commonly 50 Ω single-ended or 100 Ω differential. Getting it right keeps high-speed signals free of reflections that corrupt data.

What sets impedance

Characteristic impedance is a function of geometry and materials, not a property you set directly:

• Trace width — wider traces lower impedance.
• Dielectric height — distance to the reference plane; more height raises impedance.
• Dielectric constant (Dk) — the laminate's electrical property; higher Dk lowers impedance.
• Copper weight — thickness and the etched trapezoid shape shift the real value.

Because impedance depends on the whole stackup, it is a stackup decision, not just a trace-width decision.

Single-ended vs differential

Single-ended traces reference a plane. Differential pairs carry equal-and-opposite signals; their impedance depends on the coupling between the two traces as well as the plane, so spacing matters as much as width.

Verification

A field solver predicts impedance, but the fab proves it. An impedance coupon built on the same panel and layers is measured with a TDR (time-domain reflectometer), giving a real, traceable number.

Specify your targets, tolerance (±10% is common, ±5% for demanding links), and reference layers in the fabrication notes, and let a solver-backed stackup do the rest. See our controlled-impedance service for the full workflow.