Learn/Articles/IPC class 2 vs 3

IPC class 2 vs class 3 — when does it matter?

TL;DR

  • Class 2 = dedicated service electronics (most commercial products). Class 3 = high-reliability where downtime is unacceptable (medical, aerospace, defense). 3A adds harsh-environment aerospace addenda.
  • The class isn’t just an inspection bar — it tightens annular ring, plating thickness, conductor spacing, and acceptable defects, which feeds back into your design.
  • Pick the class your end-use truly requires. Over-speccing to class 3 “to be safe” can add 15–40% to fab cost and shrink your fab pool for no reliability benefit on a commercial product.

What the classes actually are

IPC-6012 (rigid board performance) and IPC-A-600 (acceptability) define three classes:

  • Class 1 — General electronics. Throwaway / limited-life. Rarely specified for designed products.
  • Class 2 — Dedicated service. Continued performance expected; minor cosmetic imperfections allowed. The default for most commercial and industrial hardware.
  • Class 3 — High reliability. Continued performance is critical; equipment downtime cannot be tolerated. Medical life-support, aerospace, defense, automotive safety.
  • Class 3A / space & military addenda. Class 3 plus additional requirements (e.g., IPC-6012 Space & Military Avionics addendum) for the harshest environments.

What actually changes for the designer

These are the design-affecting deltas — the ones that mean you can’t just “ask the fab for class 3” at the end:

ParameterClass 2Class 3
Min annular ring (after plating)2 mil (may breakout)3 mil, no breakout
Plating in hole (min copper)20 µm25 µm
Conductor width reduction allowed20–30%≤ 20%
Acceptable via voidslimitedfar stricter
Soldermask registrationloosertighter

The annular-ring rule is the big one: “no breakout” at class 3 means you must add pad diameter and tighten drill-to-copper, which can force a larger via grid or an extra layer on a dense BGA escape. That’s a layout decision driven by the class — decide it up front.

How to decide

Ask: what happens if this board fails in the field?

  • Annoyance / warranty swap → class 2.
  • Injury, mission loss, or unacceptable downtime → class 3 (and check whether your customer’s contract or the governing standard — e.g., DO-254, ISO 13485, AS9100 — mandates it).
  • Space / launch / avionics → class 3 with the relevant addendum (3A).

If a regulator or prime contractor names the class, that decision is made for you — design to it from day one.

Class 3 is a reliability tool, not a quality flex. Use it where field failure is genuinely intolerable — and class 2 everywhere else.

The cost of over-speccing

Class 3 typically adds 15–40% to bare-board cost (tighter tolerances, more inspection, higher scrap) and reduces the number of fabs that will quote it. On a commercial IoT widget, that’s pure cost with no benefit. On an implantable device, it’s table stakes. Match the class to the consequence.

The class checklist

  • ☐ Confirm whether a standard/contract mandates the class (DO-254, ISO 13485, AS9100, MIL-PRF)
  • ☐ Decide class before stackup + via planning, not at fab release
  • ☐ For class 3: add annular-ring margin, tighten drill-to-copper, widen pads on dense escapes
  • ☐ For class 3: confirm your fab is certified to build + inspect to it
  • ☐ State the class explicitly in the fab notes and the purchase order
  • ☐ Don’t mix — a class-3 board with class-2 assembly (IPC-A-610) is a common, costly mismatch

Need a class verdict?

Every PhySignoff engagement includes an IPC class pass/fail with the rationale — so your auditor (and your VP of Engineering) get a defensible answer. Scope a project.

References

  1. IPC-6012F, Qualification and Performance Specification for Rigid Printed Boards.
  2. IPC-A-600K, Acceptability of Printed Boards.
  3. IPC-6012 Space & Military Avionics Addendum (class 3A).
  4. IPC-A-610H, Acceptability of Electronic Assemblies (for the assembly-class match).