PCB Teardrops: Why They Matter and How to Size Them

A teardrop is a small addition to a design that costs nothing to make and prevents a whole class of field failures. It is the fillet of copper added where a trace enters a pad or a via, and it turns a fragile right angle joint into a reinforced one. This article explains what the shape does, how large it should be, where it matters and when it should be left out.

What a Teardrop Is

A teardrop is a local widening of copper at the point where a trace meets a pad, a via annular ring or a connector land. Instead of a trace of constant width running into a larger round or rectangular feature and forming a sharp internal corner, the copper flares out gradually so the transition is smooth.

Most designs use a straight taper, where two lines run out from the trace edges at roughly 45 degrees until they touch the pad, which produces a shape that looks like a tear or a water drop. Some tools also offer a curved variant, where the fillet follows an arc, and a few use a short stub that extends into the pad at an angle. The straight version is the most common because the geometry is trivial to compute and it does not consume corner area that a dense layout may need.

The feature has no electrical function. It does not lower resistance in any meaningful way, and it does not change which net the copper belongs to. Everything it does is mechanical and process related, and the two effects are worth separating.

Why They Exist

The first reason is stress. A trace entering a pad at a right angle creates a sharp corner in the copper, and a sharp corner concentrates stress. When the board is drilled, depanelised, flexed, or thermally cycled, the copper moves relative to the laminate, and the place where it moves least and crack first is the internal corner of that joint. Adding a fillet spreads the load over a longer path and eliminates the geometric notch.

The second reason is etch yield. A thin trace meeting a large pad is a copper neck, and during etching the etchant attacks the narrow region from three sides at once. If the etch runs a little long, or the etchant is locally fresh, the trace can be eaten away at the junction, leaving what the shop floor calls a mouse bite, or in the worst case an open circuit that passes a visual check but fails continuity test. A teardrop widens the neck, so the same amount of over etch removes copper that was not needed in the first place.

The third reason is drill registration. Every drill step has positional tolerance, and if the hole lands slightly off centre in a small pad, the annular ring on one side becomes very thin or disappears entirely. Where a trace arrives at that thin side, the connection can be cut through completely. A teardrop adds copper exactly in that vulnerable direction, which is why it is standard practice on boards with small pads, small drills or a tight registration budget.

Where They Matter Most

The benefit is proportional to how fragile the connection is, so the cases where teardrops earn their space are predictable.

Thin traces on thick copper are the classic case. A 0.1 mm trace on a 2 oz layer must be etched through a lot of copper, and the over etch tolerance is tight, so the junction is the most likely place for an open. Single sided and double sided boards with coarse features fall into the same category, because the etch is usually less tightly controlled and the boards are often produced in volume at low cost.

Flexible circuits are the next. A flex circuit bends thousands of times, and the copper to laminate bond at a pad junction takes the strain. Pads lift from flex, particularly at the ends of a stiffener or where a trace leaves a connector footprint, and a fillet distributes that strain over a longer copper path. Test points and connector pads that are probed repeatedly also benefit, because probe pressure pushes the pad laterally.

Small drills and tight annular rings are the third case. Microvias and laser drilled holes in HDI stacks have very small capture pads, and the registration budget is measured in tens of microns. A teardrop adds copper on the side where a trace arrives without increasing the hole or the pad diameter, which is often the only way to keep a connection inside the available space.

Thermal cycling is the last. Boards that see repeated temperature swings, such as automotive controllers or power supplies, accumulate fatigue at every copper joint. The fillet reduces the stress concentration factor at the junction, so the joint survives more cycles before a crack starts. The same reasoning applies during assembly: reflow and any subsequent rework cycle load the pads.

Sizing Rules

Teardrops are usually defined by three numbers, and the values below are the ones most fabricators accept without comment.

• Length: the flare should extend along the trace for about 1.5 to 2 times the trace width, measured from where the trace meets the pad boundary.
• Width: the widest part of the fillet should be roughly 1.5 times the trace width, or wide enough to reach the pad edges at 45 degrees, whichever is smaller.
• Angle: 45 degrees is standard. Shallower angles use more area, steeper ones concentrate stress again.

The practical limits are clearance and pad geometry. The flare must not violate the minimum gap to a neighbouring trace, and on fine pitch parts the available space between adjacent pads is often the binding constraint. Inside a pad the fillet simply merges with the copper, so it only matters that the flare does not exceed the pad boundary on the far side and create a bulge that reduces the spacing to the next feature.

On vias the same rules apply with one refinement. A teardrop on a via is normally applied to every connected layer where the copper arrives, and most tools allow the feature to be generated per layer rather than for the whole stackup. That is useful, because an inner layer with plenty of space can carry a generous fillet while an outer layer with dense routing carries a minimal one.

Tools and Automation

Teardrops are generated automatically in every major layout package now, so there is no reason to draw them by hand. In general the tool offers a choice between straight and curved fillets, a way to apply the feature to pads, vias or both, and a filter that limits the operation to objects below a certain size so that large connector pads are left alone.

Two settings are worth checking before the feature is enabled across a design. First, whether the tool allows the fillet to extend beyond the pad outline, because in a dense area that is how clearances get violated without an obvious error. Second, whether the generated shapes are kept as separate objects or merged into the pad, because a separate object can be accidentally deleted and is easier to review.

Teardrops should also be applied after the routing is finished rather than during. Adding them early makes the layout harder to edit, and moving a trace after a fillet has been attached can leave a small stub of orphan copper that no design rule check will flag unless the tool is told to look for it.

Drawbacks and When to Leave Them Out

Teardrops are not free, and the cost is space rather than money. Each fillet adds copper beside the trace and beside the pad, which reduces the gap to neighbours. On a dense board the clearance lost this way can push a design below the minimum spacing the fabricator can hold, and the fix is either to thin the trace or to skip the fillet on the affected pads.

High speed and RF designs deserve a separate look. The fillet is a small change in trace geometry, so it introduces a slight impedance discontinuity and a small amount of extra capacitance at the junction. In practice the effect is far below the noise floor for ordinary digital signals, and most high speed designs carry teardrops without measurable penalty. Where a design is already tight against an impedance budget, it is reasonable to keep the fillet small, or to leave it off on the specific nets that are being tuned, rather than removing it everywhere.

Dense HDI areas are the other place to be selective. When the pad itself is barely larger than the via and the space between via rows is measured in tens of microns, the copper a fillet consumes may be needed for the next trace. Leaving teardrops off in that region is a deliberate trade, and it should be recorded so that a later reviewer does not enable them by default and break the routing.

Finally, there is no point adding a fillet where the trace is already wide compared with the pad, or where the pad is supported by a large copper area. In those cases the joint is not the weak point, and the added copper only reduces clearance for no benefit. A teardrop on a 1 mm wide trace entering a 3 mm pad is decoration.

FAQ

• Do teardrops add cost? No. They are copper on an existing layer and need no extra process step. In some cases they reduce cost by improving etch yield.
• Should every via have one? No. Apply them where the joint is fragile: thin traces, small annular rings, tight registration, flex, and repeated thermal or mechanical load.
• Are teardrops and fillets the same thing? Yes. The terms are used interchangeably, and the shape is also called a teardrop pad or a drop fillet.
• Can a teardrop replace a proper annular ring? No. It supplements the ring on the side where a trace arrives. It does not make a pad that is too small for the drill acceptable.
• Do they help with flex fatigue? They help at the pad junction where pad lift and trace cracking usually start. They do not change the fatigue behaviour of the trace itself in a bend region.

Summary

Teardrops reinforce the place where every trace joins a pad or a via. The fillet spreads mechanical stress, widens the copper neck so over etch cannot cut the connection, and adds material on the side of a pad where drill misregistration would otherwise thin the ring. Sizing is simple: a flare roughly 1.5 to 2 times the trace width, at about 45 degrees, merged into the pad.

The decision is about density rather than about cost. Enable them across the layout by default, review clearances afterwards, and be selective in fine pitch HDI areas and in nets under an impedance budget. On a flex circuit and on thin traces over thick copper the benefit is largest and worth the space. Then let the fabrication house confirm the minimum gap still holds, because that is the only place a teardrop can do harm.