Component Polarity Verification Practice on the SMT Line

A reversed diode, capacitor or connector is one of the most expensive assembly defects, because it usually passes electrical test in a way that looks acceptable and fails later in the field. Polarity is decided by conventions that differ between packages, between manufacturers and between tape orientations, so verification is a documented process rather than a matter of experience. This article covers the checks that catch orientation errors before they are built in.

Where Polarity Errors Come From

Most errors start in the data rather than on the line. A library footprint with the pin one marker in the wrong place, a component rotation entered incorrectly in the placement programme or a reel wound the opposite way from the previous lot will all produce a board that matches the programme and contradicts the schematic.

Physical errors come second. An operator loading a reel in the wrong orientation, a tray inserted upside down or a feeder threaded in reverse produce a repeatable offset that affects a whole run rather than a single board, which is why the first article check is so valuable.

Marking conventions add another layer of risk. Packages that use a dot, a chamfer, a bevelled edge or a band each mean something different, and the same package family has changed its convention between manufacturers.

Package Markings and Datasheet Conventions

The marking convention belongs to the manufacturer, not to the package outline. A dot on the body may indicate pin one or the cathode, and the datasheet is the only authority that settles which applies to a specific part.

Where the marking is ambiguous, the package drawing in the datasheet usually shows the orientation of the internal die and the position of the pin one corner. Verifying the footprint against that drawing, rather than against a similar part already in the library, is the check that prevents a systematic reversal.

Reel, Tape and Tray Orientation

Tape and reel orientation is defined by the direction of travel and by which side of the pocket the pin one feature sits on. Two suppliers can supply the same part number with opposite orientation, so the reel label and the pocket layout both have to be checked when a new lot arrives.

Trays present a similar problem, because the pocket shape is often close to symmetrical. Checking the orientation notch on the tray, and confirming it against the drawing rather than against the previous tray, catches the change before the feeder is loaded.

Splices are a common failure point. A splice made with the tape reversed feeds parts correctly but inverts them, and the resulting error appears only on boards built after the splice.

Feeder Setup and Programme Data

Feeder setup defines the angle at which the pick nozzle presents the part, and the placement programme then rotates it to the board orientation. Both numbers have to agree, and a change to one without the other produces a rotation error of 90 or 180 degrees that is obvious on a diode and invisible on a symmetric capacitor.

The pick and place programme, the feeder and the physical reel have to agree, and a check is only meaningful when all three are compared together rather than in isolation.

Setup verification compares the loaded feeder against the setup sheet and against the physical part. Where the assembly uses several similar packages, the check is most useful when it includes the actual part marking on the reel rather than the feeder position alone.

Vision and OCR Checks at Placement

Many placement machines can inspect a part before placing it, using shape recognition or optical character recognition on the marking. The technique works well where the marking is legible and the contrast is sufficient, and it fails on parts with laser marked or lightly printed codes.

Vision checks are most valuable on expensive or high risk components rather than on every passive. Configuring them selectively keeps the cycle time acceptable while covering the parts whose reversal would be costly or impossible to detect later.

AOI and X-Ray Verification

Automated optical inspection can verify orientation wherever a visual feature exists, using the body outline, the chamfer or the marking. Programming the check requires a reference image and a tolerance, and it depends on consistent lighting across the panel.

X-ray is required where the feature is hidden, including area array packages with orientation marks on the underside and connectors whose pin one is internal. X-ray inspection resolves the internal structure well enough to confirm orientation on most package types, but it is slower than optical inspection and is usually applied to selected locations.

Manual Assembly and Rework Risks

Manual assembly removes the programme as a control, leaving the drawing and the operator. Where a hand placed component is polarised, the assembly drawing should show the orientation clearly and the kit should present the parts in a way that makes the correct orientation the obvious one.

Training has a specific role here that is easy to overlook. An operator who has seen a batch of boards scrapped because of a reversed tantalum capacitor checks the next one carefully, and that lesson lasts longer than a procedure read once at induction. Sharing the failure with the line, including its cost, turns a written control into a working habit.

Rework is the other high risk step. A replacement part fitted from a loose tray, without the feeder and the programme as a reference, depends entirely on the drawing and on the operator’s knowledge of the package convention.

First Article and Changeover Checks

The first article check is the point where data, feeder setup and physical part meet. It should confirm orientation for every polarised component, not only for the ones that look uncertain, because the errors that reach production are the ones nobody thought to question.

Where a product uses many similar packages, a physical sample board with the correct orientation marked is worth more than a written description. Kept at the station, it lets the operator compare a part in seconds instead of looking up a datasheet, and it works for every shift.

Changeover is the other moment of risk. Every new reel, tray or feeder loaded during the change should be verified against the setup sheet, and the check should be recorded with the operator name so the record shows who confirmed the orientation.

Documenting Orientation on the Drawing

Each polarised part should carry a clear reference designator, and the assembly drawing should show its orientation with an unambiguous marker, and it should also state the package convention where the marking on the body is easy to misread. A note that says pin one is at the chamfered corner removes the ambiguity that a bare outline leaves, and it should appear on the same drawing package that carries the orientation for every other polarised part.

gopcb verifies polarity data as part of the fabrication and assembly review, because a footprint error found at that stage costs a drawing revision while the same error found after assembly costs a batch of scrap. Recording the convention once, with the quality record for the product, keeps the check repeatable for every later build.

FAQ

Can AOI always detect reversed components? No. It works where a visual feature exists and the lighting is consistent. Hidden features on area array packages need X-ray, and some packages with subtle markings cannot be verified optically at all.

Why do the same part numbers arrive in different orientations? Tape and reel orientation is defined by the supplier’s winding convention, and it can differ between manufacturers. Check the pocket and the label on every new lot rather than assuming continuity.

What is the most effective polarity control? A first article check that covers every polarised part, combined with feeder setup verification at changeover. Both are simple, and together they catch the errors that programming mistakes create.