Controlled Impedance

Section Overview

Chapter 2 taught you to read a board's copper as features — traces, vias, and pads you can see and follow (§2.1). This chapter turns to something you cannot see at all: how a fast signal behaves as it travels along a trace, and why some traces must be built with far more care than others. At low speed a trace is just a wire with a little resistance (§2.4). But a fast-changing signal treats a trace as a transmission line — a conductor whose characteristic impedance, the ratio of voltage to current that a travelling wave sees, genuinely matters (transmission line). If that impedance is not held steady along the trace, part of the signal reflects back and the waveform is degraded, harming the board's signal integrity (reflection; signal integrity). To prevent that, designers build certain traces to a specific, deliberate impedance: a controlled impedance trace, made to a target value such as fifty ohms, which the fabricator then manufactures to that figure. What sets a trace's impedance is its geometry and its material: the width of the trace, its height above the metal layer beneath it that serves as its reference plane, and the dielectric constant of the insulating material in between — a property of the laminate that says how strongly it stores the signal's electric field (dielectric). Change any of these and the impedance changes. That is exactly why a controlled-impedance trace cannot be freely rerouted or jumpered: a repair that alters its width, its path, or the plane beneath it changes its impedance and can spoil the very signal it carries (Chapter 5). For a repairer the lesson is recognition and care: learn to spot the traces built for speed — differential pairs, runs kept tight over a solid plane, carefully length-matched tracks — and treat them as the precision structures they are (§3.5). Understand controlled impedance, and you know why some traces must never be touched casually.

Why This Matters

Modern boards are full of fast signals whose traces are engineered to a precise impedance, and mistreating one of those traces in a repair can kill a signal that measured "connected" the whole time — so knowing they exist is essential. This matters because a controlled-impedance trace is a precision part: it was built to a target value on purpose, and a repair that ignores that — a jumper, a reroute, a scrape near its plane — can degrade the signal even though the connection still passes a continuity check (§2.4; Chapter 5). This matters because the fault it causes is subtle: a mismatched impedance does not open a circuit; it reflects and distorts a fast signal, so the symptom is data errors, a link that will not train, or intermittent high-speed failure — not a dead connection (signal integrity). It matters because recognition guides repair: knowing which traces are impedance-controlled tells you which ones you must reroute exactly, keep the same length, and keep over their plane, and which ordinary traces you can treat more freely (§3.5; Chapter 5). It matters because the plane is part of the trace: a controlled-impedance trace and the reference plane beneath it are one system, so damaging the plane — or cutting the return path — harms the signal as surely as cutting the trace (return path; §3.2). And it matters because it sets expectations: some high-speed traces simply cannot be hand-repaired to their original performance, and knowing that in advance shapes what you promise and how you proceed (Chapter 5). Learn what controlled impedance is, and you stop treating every trace as an ordinary wire — because some of them are not.

Required Prerequisites

• Trace Width, Current, and Resistance — Section 2.4 put numbers on a trace's resistance and showed how its width and dimensions matter; this section adds a second reason a trace's width and geometry matter — its impedance to a fast signal. Familiarity with the trace as a conductor from Section 2.1 helps. This is a knowledge and observation section — no hot work; power only if you choose to probe, with care.

Recommended Consumables

• A few boards with obvious high-speed sections — a computer motherboard, a phone board, anything with fast memory or USB, HDMI, or Ethernet — to see real controlled-impedance traces and differential pairs
• A board's fabrication notes or design files, if you can get them — to see impedance targets and which layers are controlled (§3.5)
• A notebook — to sketch which traces look impedance-controlled and why
• Isopropyl alcohol and a brush — to clean a board so fine, closely-routed pairs are clear

Recommended Practice Hardware

• A magnifier or loupe and a bright, angled light — to see tightly-routed differential pairs and length-matching wiggles
• A backlight for thin boards — to see whether a trace runs over a solid plane on the next layer
• A multimeter for continuity, if you wish — to confirm a connection — though note continuity cannot measure impedance (Volume 3)
• No iron, hot air, or hot work is needed — this section is reading and reasoning, not procedure

Real-World Applications

Recognizing controlled-impedance traces changes how a repairer approaches the fast sections of a board. A technician repairing a high-speed data line knows a plain jumper may not restore it, because the jumper's impedance will not match, and plans a length-matched, plane-referenced repair or manages expectations (Chapter 5). Someone reading an unfamiliar board spots the tightly-coupled differential pairs running to a USB or HDMI connector and recognizes them as impedance-controlled at a glance (§3.5). A repairer working near a fast trace takes care not to scrape or cut the ground plane beneath it, knowing the plane is the trace's return path and part of its impedance (§3.2). A builder checking a design reads the fabrication notes for impedance targets and confirms the controlled layers were built as specified. And anyone who has "fixed" a high-speed link with a casual bodge wire learns why it still failed — the connection was restored but the impedance was not. The failures this skill prevents: bodging a controlled-impedance trace and killing the signal, cutting a reference plane, mistaking a subtle signal-integrity fault for a dead connection, and promising a high-speed repair you cannot make to spec (Chapter 5).

Common Challenges

• Treating a high-speed trace like an ordinary wire. A controlled-impedance trace is built to a target value — a naive jumper changes its impedance and can spoil the signal (Chapter 5).
• Forgetting the plane underneath. The reference plane is part of the trace's impedance and carries its return current — damaging it harms the signal (return path; §3.2).
• Reading a signal-integrity fault as an open. A mismatched impedance distorts a fast signal without breaking the connection — continuity can pass while the link still fails (signal integrity).

Safety Notes

Risk Level: Low. Studying impedance-controlled traces on an unpowered board is a safe reading task — the cautions are the usual ones for probing a live board.

Professional Tips Before Starting

• Look for pairs and wiggles. Tightly-coupled two-trace differential pairs and the serpentine "wiggles" that length-match a bus are the visual signatures of impedance-controlled routing (§3.5).
• Find the plane before you touch the trace. A controlled-impedance trace runs over a solid reference plane that carries its return current — know where that plane is before you work near it (§3.2).
• Remember what continuity cannot tell you. A meter confirms a connection but says nothing about impedance — a high-speed trace can read "good" and still be ruined by a mismatch (Volume 3).

Traces Built for Speed

Recap and Frame

Before defining controlled impedance, it helps to see where it fits against the trace you already know. So far a trace has been a conductor that carries a signal or power, with a width and a copper weight that set the current it can handle and the resistance it has (§2.1; §2.4). For most traces on most boards, that is the whole story: they carry a signal or a supply from one place to another, and as long as the connection is intact, they work. What this chapter adds is a second world that appears when signals get fast. A fast-changing signal — a high-speed data line, a clock, a fast bus — does not see a trace as a simple wire; it sees a transmission line, a conductor along which the signal travels as a wave, and for such a signal a new property of the trace matters: its impedance (transmission line; characteristic impedance). This section is about that property and how designers control it. It is still the same copper trace you have been reading — but now its exact width, its height above the layer beneath it, and the material around it matter in a way they never did for a slow signal. The goal here is not to turn you into a high-speed designer: it is to give a repairer the recognition and the caution to treat these traces correctly, because they are the ones a careless repair most easily ruins (Chapter 5). Start with why a fast signal cares about impedance at all.

Why Impedance Matters at Speed

The whole idea of controlled impedance rests on how a fast signal travels, which is different from how a slow one does. A slow or steady signal fills a trace almost instantly, so the trace behaves as you expect — a wire with a small resistance and little else to think about (§2.4). A fast-changing signal is different: it changes so quickly that, in the time the edge takes to travel down the trace, the far end has not yet "heard" the near end, so the signal propagates along the trace as a travelling wave. A conductor carrying a signal this way behaves as a transmission line, and a travelling wave on a transmission line sees a characteristic impedance — the ratio of the wave's voltage to its current as it moves along (transmission line; characteristic impedance). Here is why that matters. If the impedance is the same all the way along, the wave travels smoothly to its destination. But wherever the impedance changes — at a discontinuity, a stub, a bad connector, or a mismatched load — part of the wave reflects back toward the source, like an echo (reflection). Those reflections add to and subtract from the signal, distorting its shape, rounding its edges, and in the worst case causing a receiver to misread it (signal integrity). For a slow signal none of this matters; for a fast one it is the difference between a link that works and one that throws errors. So the aim for a fast trace is a steady, predictable impedance from end to end — and achieving that on purpose is what controlled impedance means.

Controlled Impedance — Building to a Target

When a trace must carry a fast signal cleanly, the designer does not leave its impedance to chance — they specify it, and this deliberate design is controlled impedance. A controlled impedance trace is one built to a specific, chosen characteristic impedance — a target value the designer picks to match the signal and the parts at each end. Common targets are familiar numbers: around fifty ohms for many single traces, and around ninety to a hundred ohms for a differential pair carrying a signal on two traces at once, as used by USB, HDMI, Ethernet, and fast memory (characteristic impedance). The designer chooses the value, works out the exact trace geometry that produces it, and marks those traces as impedance-controlled in the board's fabrication notes (§3.5). The fabricator then builds to that specification: because the impedance depends on precise dimensions, the manufacturer holds the trace width, the layer spacing, and the material to tight tolerances, and may even measure a test trace to confirm the finished board hits its target. This is the key mental shift: a controlled-impedance trace is not just a connection but a manufactured value, as much a specified part as a resistor. Its whole purpose is to give a fast signal the smooth, matched path that keeps it clean, and that purpose is designed and built in, not incidental. So when you meet one, you are looking at a trace whose exact form is doing a job — which is why the next question is what, precisely, sets that impedance.

What Sets the Impedance

A trace's characteristic impedance is set by a small number of physical factors, and knowing them tells you what a repair must preserve. The first is the trace's width: a wider trace has a lower impedance, a narrower one a higher impedance, which is why a controlled trace's width is held so precisely (trace width, §2.4). The second is its height above the metal layer beneath it — the plane it is built over: the closer the trace sits to that plane, the lower its impedance; the farther, the higher. The third is the material in between. The insulating laminate that separates the trace from the plane has a property called its dielectric constant — a measure of how strongly the material stores an electric field (dielectric); a higher dielectric constant lowers the trace's impedance, a lower one raises it, so the choice of board material is part of the impedance too (§1.4). The copper's thickness plays a smaller part as well. Put together, the impedance is a recipe: a particular width, at a particular height above the plane, through a particular material, yields a particular impedance — change any ingredient and the value moves. This is the heart of why these traces are delicate to repair: a jumper is a different width at a different height through air instead of laminate, so it has a completely different impedance, no matter how good the connection it makes (Chapter 5). To hold a trace's impedance, a repair would have to reproduce its width, its spacing to the plane, and its dielectric surroundings — which ordinary hand tools cannot easily do. Know the recipe, and you know exactly what a controlled-impedance trace will not tolerate.

The Reference Plane and the Return Path

One ingredient in that recipe deserves its own treatment, because it is easy to forget and vital to the trace: the plane underneath. A controlled-impedance trace is always built over a solid metal layer — a ground or power plane — that acts as its reference plane (ground plane). The impedance is defined between the trace and this plane, so the plane is not a bystander: it is half of the structure that gives the trace its impedance. There is a deeper reason the plane matters, and it is the return current. Every signal current that flows down a trace must return to its source, and for a fast signal that return current does not wander — it flows in the reference plane directly beneath the trace, mirroring the signal's path (return path). The signal and its return travel together, the trace and the plane forming a pair, and it is this tight coupling that makes the impedance well-defined and the signal clean. This has a sharp consequence for repair: anything that disturbs the plane under a fast trace disturbs the signal. Cutting a slot in the plane, routing the trace off the edge of the plane, or breaking the return path forces the return current to detour, which changes the impedance and injects noise — even though the signal trace itself is untouched (§3.2). So when you work near a controlled-impedance trace, you must protect its reference plane as carefully as the trace itself: the two are one system. Remember that a fast trace and its plane are a matched pair, and you will not accidentally ruin a signal by damaging the copper you were not even working on.

What It Means for Repair

All of this converges on a practical rule for the repairer: recognize controlled-impedance traces, and treat them differently from ordinary ones. First, recognize them. The visual signatures are learnable: closely-coupled differential pairs running side by side, serpentine "wiggles" that length-match one trace to its partner or one bus line to another, traces that run in clean, direct paths over a solid plane, and any trace heading to a high-speed connector like USB, HDMI, or Ethernet (§3.5). Fabrication notes, when you have them, name the controlled nets and their target impedance outright (§3.5). Second, understand what a repair must preserve. To keep a controlled trace's impedance, a repair must hold its width, its length (especially relative to a paired trace), its height over the plane, and the plane itself (Chapter 5); a plain bodge wire preserves none of these, which is why a casually-jumpered high-speed line so often still fails. Third, adjust your approach. Some controlled-impedance repairs can be done well with care — a short, matched, plane-referenced fix — but some fast traces simply cannot be hand-restored to their original performance, and recognizing which case you are in is part of the job (Chapter 5). Fourth, protect what you are not repairing. Keep heat, scrapes, and cuts away from the reference planes and the neighbouring traces of a fast section (§3.2). The throughline is respect: a controlled-impedance trace is a precision structure, and knowing that turns a signal you might have unknowingly destroyed into one you handle deliberately. Learn to see these traces for what they are, and you will repair the fast parts of a board — or decline to — with your eyes open (§3.5; Chapter 5).

Common Mistakes

• Jumpering a controlled-impedance trace with plain wire. A bodge wire has a completely different impedance — it may restore the connection but ruin the fast signal (Chapter 5).
• Cutting or scraping the reference plane. The plane carries the trace's return current and sets its impedance — damaging it harms the signal like cutting the trace (return path; §3.2).
• Assuming continuity proves a high-speed trace is good. Continuity shows a connection, not an impedance — a mismatched trace passes continuity and still fails (signal integrity).
• Ignoring length matching on a pair or bus. A differential pair and matched buses depend on equal lengths — a repair that changes one length upsets the timing (§3.5).
• Treating every trace on a fast board as ordinary. Only some are impedance-controlled — learn the signatures so you know which need special care (§3.5).

Troubleshooting Guidance

Controlled-impedance problems come down to recognizing these traces and preserving what makes them work. If a high-speed link fails but reads connected: suspect an impedance problem, not an open — a mismatch distorts the signal while continuity still passes (signal integrity). If you must repair a fast trace: keep any jumper short, matched in length to its partner, and routed over its plane, and accept that some cannot be restored to spec (Chapter 5). If you cannot tell whether a trace is controlled: look for differential pairs, length-matching wiggles, and routes to high-speed connectors, or read the fabrication notes (§3.5). If a signal degraded after nearby work: check whether the reference plane or return path was cut or scraped (§3.2). If you are unsure a repair will meet spec: say so — some high-speed traces cannot be hand-restored to their original performance (Chapter 5). If you must probe a live board: do it carefully, mindful that fast sections are densely packed and easy to bridge (Volume 2). The throughline: identify the controlled-impedance traces first, then preserve their width, length, plane, and dielectric — or recognize when you cannot.

Verification & Testing Methods

Use this as a check that you understand controlled impedance, not a hot procedure:

• [ ] I can explain why a fast signal treats a trace as a transmission line whose impedance matters, while a slow one does not (§2.4).
• [ ] I can say what a controlled impedance trace is and why a designer specifies a target value like fifty ohms.
• [ ] I can name what sets a trace's impedance — its width, its height above the reference plane, and the dielectric constant of the material (§1.4).
• [ ] I can explain that the reference plane carries the signal's return current and is part of the trace's impedance (return path; §3.2).
• [ ] I can recognize a controlled-impedance trace and say why a naive jumper or a cut plane endangers it (§3.5; Chapter 5).

Then try the practice exercises below — observation and reasoning practice; scenarios differ from the quiz.

Practice Exercises

1. Spot the fast traces (5 minutes, observation). On a high-speed board, find the closely-coupled differential pairs and length-matching wiggles, and note which connectors they run to (§3.5).
2. Find the reference plane (5 minutes, observation). For a trace you think is impedance-controlled, work out — from the stackup or a backlight — which layer holds the solid plane beneath it (§3.2).
3. Reason about a jumper (4 minutes, reasoning). Explain how a plain bodge wire would change a controlled trace's impedance, naming each factor (width, height, dielectric) it fails to preserve.
4. Continuity versus impedance (4 minutes, reasoning). Explain why a high-speed trace can pass a continuity check and still fail, and what that means for diagnosing it (signal integrity).

These core ideas — why impedance matters at speed, controlled impedance, what sets it, the reference plane and return path, and what it means for repair — are tested in the Chapter Quiz at the end of this chapter, where a score of 80% is required to continue.

Key Takeaways

• A fast signal travels along a trace as a wave and sees it as a transmission line with a characteristic impedance; if that impedance is not steady, part of the signal reflects and the waveform degrades, hurting signal integrity (§2.4).
• A controlled impedance trace is one deliberately built to a target characteristic impedance — commonly around fifty ohms single-ended or ninety to a hundred ohms for a differential pair — specified by the designer and manufactured to value by the fabricator.
• A trace's impedance is set by its width, its height above the metal layer beneath it, and the dielectric constant of the material between — wider or closer to the plane lowers it, a higher dielectric constant lowers it (§1.4; §2.4).
• The solid reference plane beneath a controlled trace is half of its impedance and carries the signal's return current directly under the trace, so damaging the plane or its return path harms the signal as surely as cutting the trace (return path; §3.2).
• Because its impedance depends on precise geometry, a controlled trace cannot be freely rerouted or jumpered: recognize the signatures — differential pairs, length-matching, plane-referenced routes — and preserve width, length, plane, and dielectric, or accept that some cannot be hand-restored (§3.5; Chapter 5).

Skills Learned

• You can now explain why impedance matters for a fast signal but not a slow one.
• You can now say what controlled impedance is and why a designer specifies a target value.
• You can now name what sets a trace's impedance and how each factor changes it.
• You can now explain the reference plane's role as the signal's return path.
• You can now recognize a controlled-impedance trace and say why a naive repair endangers it.

Glossary Additions

• controlled impedance — a trace deliberately built to a specific, chosen characteristic impedance (a target value such as around 50 ohms for a single trace, or 90 to 100 ohms for a differential pair) so that a fast signal travels along it cleanly without reflections. The designer picks the value to match the signal and the parts at each end, works out the exact trace geometry that produces it, marks the net as impedance-controlled in the fabrication notes, and the fabricator manufactures the trace to that value by holding its width, spacing, and material to tight tolerances. A controlled-impedance trace is therefore a specified, manufactured value — not just a connection — and cannot be freely rerouted or jumpered without changing its impedance.
• reference plane — the solid metal layer, usually a ground or power plane, that a controlled-impedance trace is built over and whose spacing to the trace helps set its impedance. The plane is not a bystander: the signal's return current flows in it directly beneath the trace, mirroring the signal's path, so the trace and its reference plane form one coupled system. Cutting, slotting, or routing off the edge of the reference plane forces the return current to detour, which changes the impedance and injects noise even if the signal trace itself is untouched — which is why the plane must be protected as carefully as the trace.
• dielectric constant — a property of an insulating material (such as a board's laminate) that measures how strongly it stores an electric field; it is one of the factors that sets a trace's characteristic impedance. Between a controlled-impedance trace and its reference plane lies this dielectric, and a higher dielectric constant lowers the trace's impedance while a lower one raises it, so the choice of board material is part of the impedance design. Because a jumper replaces the laminate with air — a much lower dielectric constant — its impedance differs from the trace it replaces, one reason a bodge wire cannot reproduce a controlled trace's behaviour.

Suggested Next Sections

Must read next:

• Ground Planes — you have met the reference plane as the layer that sets a trace's impedance and carries its return current; the next section examines the ground and power planes in full — what they do, why solid planes matter, and how they carry return current, shield the board, and steady its supply.

Recommended:

• Trace Width, Current, and Resistance — the trace dimensions that set resistance here set impedance too.
• Trace Anatomy and Function — the trace as a conductor, now seen as a transmission line for fast signals.