Tombstoning: Why Chip Components Stand Up
Tombstoning is a chip component that has stood up on one end during reflow, held vertically by the surface tension of a single molten joint. The mechanism is a balance of torques at the two ends, and it is decided before the board reaches the oven by the paste volumes, the pad geometry and the timing of the two joints.
What Tombstoning Looks Like and Where It Happens
One end of the part is soldered to its pad and the other is lifted off the board, standing at an angle or fully vertical. The part may be held by a single fillet or hang by a bridge to a neighbouring feature, and the joint that did form is usually sound, which is why the defect is easy to diagnose once it is seen.
It is concentrated on small chips, particularly the smallest sizes, and on the first side of a double-sided assembly. The first side sees the second reflow pass as well, which is why a part that survived its own reflow can tombstone during the second one.
The Force Balance at the Joint
At each end of the part the molten solder pulls the termination downward and inward, and the vector that rotates the part depends on the fillet shape. If both ends are molten at the same moment and the forces are equal, the part stays flat. If one end melts first, the molten joint lifts the part while the other end is still on solid paste.
The part therefore rotates about the molten joint until either the second end contacts its pad again or the part reaches a position the first joint can hold. Anything that delays the melting of one end, or increases the pulling force at the other, shifts the balance toward tombstoning.
Pad Geometry and Asymmetry
The pads have to be symmetric in area and in thermal connection, because a difference between them changes both the melting time and the force. A pad connected to a plane heats more slowly than one on a narrow trace, and a pad with a larger area holds more solder, which increases the pull on that end.
Asymmetry also comes from the track layout, from a via placed near one pad and not the other, and from the mask opening. Reviewing a tombstoning problem means checking the two pads as a pair, including what is connected to each of them, rather than examining each pad on its own.
<img src="https://www.gopcba.com/wp-content/uploads/2026/06/Sierra-Wireless.png" alt="Chip component standing upright on one pad after reflow” />
Paste Volume and Placement Offset
Paste volume is the single most controllable factor. A difference in deposit volume between the two ends changes the height of the molten joint and the force it exerts, and it also changes how quickly that end reaches liquidus. A placement offset moves the part so that one end sits closer to the pad centre and the other overhangs, which has the same effect.
Because the tolerance on paste volume and placement is a percentage rather than an absolute value, the smallest parts are the most sensitive. A 10 percent volume difference on an 0402 pad is a small absolute quantity but a large fraction of the joint, which is why the defect appears on the smallest chips first.
Thermal Asymmetry and the Profile
The profile controls when each end melts. A fast ramp heats both ends nearly together, while a slow one allows small differences in thermal mass to matter more. A soak that equalizes the assembly before the peak helps both ends cross liquidus at nearly the same time, which is the condition that keeps the part flat.
Ambient gas also matters. In a nitrogen atmosphere the wetting forces are higher and form faster, so a small imbalance produces a larger torque than it would in air. This is one of the few cases where a better wetting environment makes a defect worse, and the countermeasure is to reduce the imbalance rather than to remove the nitrogen.
Vias, Thermal Relief and Heat Sink Effects
A via in or near one pad conducts heat away and delays melting at that end, which is a common cause on boards where a part sits close to a ground connection. Thermal relief spokes narrow the connection and reduce the effect, but the spokes themselves can be asymmetric if the pattern is drawn carelessly.
Where the plane is large, the difference can be enough that the plane-connected end lags several seconds behind the other, which is well beyond what the surface tension balance can tolerate. In that case the fix is to change the thermal connection, not the paste volume.

Component and Material Factors
The part itself contributes through its termination geometry and its mass. A part with a small termination and a light body is easier to lift, and a part whose terminations differ in finish or in wetting speed will have an inherent imbalance that no process change removes.
Solder paste chemistry plays a part through the speed and force of wetting, and a paste that wets very quickly can lift a part before the other end has melted. Where a defect appears after a paste change, the wetting behaviour should be considered alongside the volume and the profile.
Prevention in Production
The practical controls are symmetric pads, equal paste volumes, accurate placement, a profile with a soak that equalizes the assembly, and a check of the thermal connection at each pad. Verifying the deposit on both ends of the part rather than the average across the board is what catches the problem before it appears in yield.
Where the defect is chronic on one part, the footprint and the thermal connection are the places to look, because process adjustments only move the balance a little. The deposit volume record and the section of a good joint together show which of the two has drifted.
Detection and Disposition
Tombstoned parts are obvious once seen but easy to miss at speed, so the inspection program should include an orientation check for small chips rather than only a presence check. Where the defect is intermittent, an X-ray or a height measurement can reveal a part standing at a shallow angle that a top-down view does not show.
Disposition is straightforward: the part is removed and replaced, and the pads are cleaned before the new part is placed. What matters is recording the occurrence, because a single tombstone in a lot is a signal about the balance at that pad pair, and ignoring it leaves the cause in place.
Points to Confirm at First Article
The cost of verification is small compared with the cost of a field failure, and it is paid at a point where the product can still be corrected. Documentation exists so that a person who was not present can reproduce the work and reach the same conclusion.
Related reading: our fabrication notes, board quality and design release notes cover the same ground.
FAQ
What causes a chip component to tombstone? One end melts before the other, and the surface tension of the molten joint lifts the part. Paste volume, pad symmetry, placement offset and thermal connection all influence the balance.
Why does tombstoning appear on the second reflow pass? The part sees an additional thermal cycle, and the paste from the first pass is already consumed. Small differences in thermal mass are expressed again, so a marginal part can lift.
Does nitrogen make tombstoning worse? It can. Wetting is faster and stronger in nitrogen, so a small imbalance produces a larger lifting force. Reduce the imbalance rather than removing the nitrogen.



