IPC-A-610 Acceptance Classes: What 'Good' Looks Like

IPC-A-610 Acceptance Classes: What 'Good' Looks Like

Key Takeaways

• IPC-A-610 defines the visual acceptance criteria for electronic assemblies: solder joints, component placement, and board cleanliness. The current edition is Revision J, published in 2024.
• It sorts products into three reliability classes: Class 1 (general electronics), Class 2 (dedicated service electronics), and Class 3 (high-reliability, life-critical electronics).
• Class 3 is tightest: more solder fillet coverage, less tolerance for voids, and zero allowance for defects Class 2 might accept as cosmetic.
• The buyer decides the class based on the product's reliability needs, not the assembler. Specify it before production starts.
• Class 2 covers most commercial and industrial hardware. Class 3 is for products where field failure is unacceptable, like medical, aerospace, and automotive safety systems.
• IPC-A-610 also sets acceptance criteria for conformal coating and component staking, and its cable-and-harness counterpart is IPC/WHMA-A-620 — both use the same three classes.

If you're outsourcing PCB assembly, you need to know which IPC-A-610 class your board is built to and what that buys you. IPC-A-610 is the industry standard that defines what an acceptable electronic assembly looks like, splitting products into three reliability classes with progressively stricter limits on solder, placement, and finish. This guide is for engineers and procurement managers who want to specify the right class and read an inspection report without guessing. By the end, you'll know what separates Class 2 from Class 3 and where your product fits.

What IPC-A-610 Actually Covers

IPC-A-610, "Acceptability of Electronic Assemblies," is the most widely used visual inspection standard in electronics manufacturing. It defines accept/reject criteria for a finished assembly: solder joint shape, component alignment, lead protrusion, board damage, and contamination. It's a workmanship standard, not a process spec, so it tells you what good looks like, not how to get there.

The companion document, IPC J-STD-001, governs the soldering process itself. The two are used together: J-STD-001 controls how the board is built, and IPC-A-610 judges whether the result passes. Reputable assemblers certify their operators against both. For how these fit alongside ISO 9001 and UL, see our guide on what pcb assembly certifications actually mean.

The standard is maintained by IPC, the global trade association for electronics manufacturing. The current edition, Revision J, was published in March 2024 and refreshed the document with updated solder-joint and cleanliness criteria, expanded reference imagery, and clearer guidance on conformal coating and wire termination. Conditions are judged as acceptable, a process indicator, or a defect; the older "target condition" tier was dropped back in Revision H (2020).

Three states for every criterion

For each inspection point, IPC-A-610 marks a condition as acceptable or a defect for a given class. A defect fails the requirements and must be reworked, repaired, or scrapped. Some conditions are flagged as a process indicator: acceptable, but a sign the process may be drifting and worth watching.

The Three Acceptance Classes Explained

IPC-A-610 defines three product classes based on the consequences of failure. Class 1 is general electronic products, Class 2 is dedicated service electronics, and Class 3 is high-performance electronics where downtime can't be tolerated. The higher the class, the tighter the acceptance limits, because the cost of a field failure rises sharply at each step.

Here's how the three classes break down:

Class 1: General Electronic Products

Class 1 covers products where the only real requirement is that the assembly works when it ships. Cosmetic imperfections and minor solder issues are acceptable as long as the board functions. It fits low-cost, short-life consumer goods where a failure has no safety consequence. In practice, few buyers specify Class 1, because the savings over Class 2 are marginal and the reliability gap isn't worth it.

Class 2: Dedicated Service Electronics

Class 2 is the default for the large majority of commercial and industrial hardware. It expects continued performance and an extended service life, but allows brief interruptions that aren't critical. Minor cosmetic flaws pass as long as they don't affect function or reliability. If you're building a consumer product, an industrial controller, or most IoT hardware, Class 2 is almost certainly what you want.

Class 3: High-Reliability Electronics

Class 3 applies where continued operation is essential and a failure could risk lives or mission success. Medical devices, avionics, automotive safety systems, and military electronics live here. The limits are tightest: more solder fillet coverage, far less tolerance for voids and disturbed joints, and stricter rules on alignment and lead conditions. There's no room for "cosmetic, but acceptable." If it's marginal, it's a defect.

Class 2 vs Class 3: Where the Real Differences Live

The ipc class 2 vs class 3 question comes down to how much margin the standard demands in the solder joint and placement. Class 3 requires more wetting, fuller fillets, tighter alignment, and lower void allowances than Class 2 across nearly every criterion. Class 2 accepts conditions that Class 3 rejects, because Class 3 assumes the product faces thermal cycling, vibration, or a long service life where any weakness eventually shows up.

A few concrete examples of how the bar rises from Class 2 to Class 3:

• Class 2 typically accepts around 75% vertical solder fill of the barrel; Class 3 requires more, because a partially filled barrel is a reliability risk under thermal stress.
• Class 3 demands more complete wetting and fuller fillets. A joint that's "good enough" cosmetically on Class 2 can be a defect on Class 3 if the fillet is thin or the wetting angle is wrong.
• Side overhang and rotation tolerances tighten under Class 3, especially on fine-pitch and small chip parts where misalignment cuts contact area.
• Class 3 allows fewer voids in solder and tighter limits on laminate damage, measling, and ionic residue, because the board may need to survive 15 years instead of three.

The practical takeaway: Class 3 isn't "Class 2 plus a coat of paint." It's a different inspection regime, often with more X-ray on hidden joints like BGAs and more documentation. It costs more in inspection time, rework, and yield, so don't specify it unless your application genuinely needs it.

Who Decides the Class, and How to Specify It

The buyer decides the class, not the assembler. IPC-A-610 is explicit that the contract or documentation defines which class applies, and absent a stated class, the assembler defaults to Class 2. So if you don't specify, you get Class 2 criteria, which is fine for most products but a problem if your board needs Class 3.

To get this right, do three things before production starts:

1. Write "Built and inspected to IPC-A-610 Class 2" (or Class 3) into the documentation. Don't leave it implied.
1. If you require Class 3 visual acceptance, your assembler should also solder to the corresponding J-STD-001 class. The two need to agree.
1. Ask whether the line's operators and inspectors hold current IPC-A-610 certification. A standard only helps if the people applying it are trained to it.

Getting the class agreed up front is one of the most effective ways to outsource pcb assembly china without nasty surprises at incoming inspection. It removes the ambiguity that causes disputes over what "good" means.

How the Class Shows Up on the Line

On a real SMT line, the IPC-A-610 class drives how every inspection gate is calibrated. Automated optical inspection (AOI) systems are programmed with the alignment, fillet, and presence tolerances for the specified class. X-ray checks hidden joints under BGAs and QFNs against the class-specific void limits. Operators reference the standard when judging borderline joints by eye.

The class you pick sets the threshold an aoi or X-ray inspector uses to flag a joint, and it feeds back into how the whole pcb assembly process is run, from stencil design to reflow profile tuning. A Class 3 board typically sees tighter process control and more inspection coverage than the same board built to Class 2.

What a real rejection looks like: an AOI system flags a chip resistor with side overhang past the Class 3 limit, an operator confirms it, and the board goes to rework. The same overhang might have passed under Class 2. Multiply that across a panel and you see why Class 3 yields run lower and cost more. At Faradine, every board is functionally tested before it ships, on top of in-line inspection, so the IPC-A-610 verdict isn't the last word, the working-product check is.

IPC-A-610 Conformal Coating Criteria

Conformal coating gets explicit acceptance criteria in IPC-A-610 because a poor coating job can cause the very failures it was meant to prevent: trapped moisture, dendrite growth under a breached film, and masked solder defects that escape inspection. When a board is specified with coating, the inspector judges the cured film against the same class structure as solder joints.

• Coverage: the areas the assembly drawing designates as coated must be fully coated, and the film must be uniform enough to protect them.
• Film condition: bubbles, blisters, delamination, dewetting, cracks, and foreign material trapped in the coating are all judged against the class limits; what passes as a process indicator on Class 2 can be a defect on Class 3.
• Keep-out areas: connectors, test points, adjustment components, and any surfaces that must mate or conduct have to stay free of coating, per the documentation.
• Order of operations matters: solder joints are inspected to the class criteria before coating goes on, because once coated, visual verification of the joints underneath is limited to what X-ray can see.

Revision J expanded the standard's conformal coating guidance, which is one reason coating questions now route to IPC-A-610 rather than to coating-material specs alone. If your product needs coating, call it out in the build documentation with the coating material, the areas to coat, and the class — the assembler's inspection plan should name all three.

Component Staking: When Adhesive Becomes an Acceptance Criterion

Staking is bonding a component to the board with adhesive so it survives vibration, shock, and handling. It shows up on large or heavy parts — electrolytic capacitors, big connectors, transformers, tall relays — and on any assembly headed for a high-vibration environment, which is why it's most common on Class 3 work. IPC-A-610 treats the staking itself as an inspectable feature, not an afterthought.

The criteria are straightforward in principle: the adhesive must actually bond the component body to the board surface, it must not wick into solder joints or onto pads and terminations, it must not stress or damage the part, and the staked component still has to sit within the alignment tolerances of its class. Too little adhesive is a weak bond; too much, or adhesive in the wrong place, can be a defect of its own. If your product will see vibration — automotive, industrial, aerospace — ask your assembler which parts they stake and against which criteria the staking is inspected.

IPC-A-610 vs IPC/WHMA-A-620: Which Standard Applies

IPC-A-610 covers printed board assemblies — components soldered to a PCB. IPC/WHMA-A-620 is the companion standard for cable and wire harness assemblies: crimped terminals, splices, connectors on wires, and harness dressing. The two share the same three-class structure, so the Class 1/2/3 logic in this article carries over, but the criteria themselves address completely different hardware.

The distinction matters for box builds and full products: the boards inside are judged to IPC-A-610, the cabling between them to IPC/WHMA-A-620. If your assembly includes both, specify both standards — and the class for each — in your documentation, otherwise the harness work defaults to whatever the shop happens to do.

IPC-A-610 Revisions: Why the Letter After 610 Matters

Searches like "IPC-A-610G" or "IPC-A-610C" are people looking for specific revisions of the standard, and the letter matters because criteria change between editions. The current edition is Revision J, published in March 2024. The previous edition, Revision H (2020), was the one that restructured every criterion into acceptable, process indicator, or defect and dropped the old "target condition" tier. Revision J kept that structure and refreshed the criteria and reference imagery, with updated guidance on conformal coating and wire termination.

Practically: put the revision letter in your purchase documentation — "built and inspected to IPC-A-610J," not just "to IPC-A-610" — and confirm your assembler's operators are certified against the current revision. An inspection report written against a superseded revision can accept or reject joints differently than the one you thought you specified.

Frequently Asked Questions

What is IPC-A-610?

IPC-A-610 is the IPC standard titled "Acceptability of Electronic Assemblies." It defines the visual accept and reject criteria for finished electronic assemblies, covering solder joints, component placement, lead protrusion, board damage, and cleanliness. It's the most widely used workmanship standard in electronics manufacturing and is published by IPC. The current edition is Revision J, released in 2024.

What's the difference between IPC Class 2 and Class 3?

Class 2 covers dedicated service electronics where extended life is expected but brief interruptions are acceptable. Class 3 covers high-reliability electronics where failure can't be tolerated, like medical and aerospace gear. Class 3 demands fuller solder fillets, more barrel fill, tighter alignment, and lower void limits. Conditions accepted as cosmetic in Class 2 are often defects in Class 3.

Which IPC-A-610 class do I need for my product?

Most commercial, consumer, and industrial products are built to Class 2, which balances reliability and cost. Choose Class 3 only if a field failure could risk lives or critical missions, such as medical life-support, automotive safety, avionics, or military systems. Class 3 raises inspection effort and lowers yield, so don't over-specify. Class 1 is rare and fits only low-cost, short-life goods.

Who decides which IPC class applies to my board?

The buyer decides, through the contract or build documentation. IPC-A-610 states that the class must be specified by agreement, and if none is stated, the assembler defaults to Class 2. Write the class explicitly into your quote and purchase order, and confirm it matches your soldering specification under IPC J-STD-001 to avoid disputes at incoming inspection.

Is IPC-A-610 the same as IPC J-STD-001?

No. IPC-A-610 is a visual acceptance standard that judges the finished assembly, while IPC J-STD-001 is a process standard that governs how the soldering is performed. They're used together: J-STD-001 controls the build, and IPC-A-610 verifies the result. Both define the same three reliability classes, so they should be specified to matching classes for consistency.

What is the current revision of IPC-A-610?

The current edition is IPC-A-610 Revision J, published in March 2024. It refreshed criteria and reference imagery throughout, with updated guidance on conformal coating and wire termination. Each inspection point is classified as acceptable, a process indicator, or a defect for a given class; the separate "target condition" tier was dropped back in Revision H (2020). Always confirm your assembler is inspecting to the latest revision.

What are the IPC-A-610 conformal coating criteria?

IPC-A-610 judges a cured conformal coating on coverage of the designated areas, film condition (bubbles, delamination, dewetting, cracks, and trapped foreign material are all graded against the class), and correct masking of keep-out areas like connectors and test points. Class 3 tolerates fewer coating defects than Class 2. Revision J (2024) expanded the coating guidance, and solder joints are inspected to class before coating because the film hides them from later visual checks.

What is component staking under IPC-A-610?

Staking is bonding a component to the board with adhesive so heavy or tall parts survive vibration and shock. IPC-A-610 inspects the staking itself: the adhesive must bond part to board, stay out of solder joints and off pads, avoid damaging the component, and leave the part within its class alignment tolerances. It's most common on Class 3 assemblies and high-vibration products.

What's the difference between IPC-A-610 and IPC-620?

IPC-A-610 is the acceptance standard for printed board assemblies; IPC/WHMA-A-620 covers cable and wire harness assemblies — crimps, splices, and wired connectors. Both use the same Class 1/2/3 structure. Box builds that contain boards and cabling need both standards specified, each with its class.

Build to the Right Class From the Start

Specifying the correct IPC-A-610 class up front is the cleanest way to align expectations with your assembler and avoid arguments over what "good" means. Decide whether your product genuinely needs Class 3 or whether Class 2 covers it, write it into your documentation, and confirm your manufacturer's operators are certified to apply it.

If you want a partner who builds to your specified class, inspects against it in-line, and functionally tests every board before shipping, talk to the Faradine team about your project. You can see how a class-specified build runs end to end in our capabilities and solutions overviews.