This blog explains why tombstoning happens in SMT lines and how to cut it down with better profiles, pad design, paste control and placement for cooler boards.
Tombstoning looks small, but it can kill your SMT yield fast. One chip resistor stands up like a tiny stone, one pad is soldered, the other side hangs in the air. The board may pass AOI, then fail in test or even in the field. So here we talk simple: why it happens, which knobs you can turn, and how to stop it.
In surface-mount assembly, tombstoning means one end of a small two-pad part lifts during reflow. It happens more on 0603, 0402, 0201 parts, where the body is light and solder force is strong. For a cold-room controller or a display cabinet driver, one open resistor already can break the whole cooling system. Your nice rear wire shelving is still empty, because the refrigerator never starts.
The root is simple: solder paste on the two pads does not melt and wet in the same way. One side pulls harder, the chip rotates, and now you have a defect.
Most tombstoning comes from uneven heating. One pad hits liquidus earlier, its solder wets, and surface tension pulls that side up.
If you run a very aggressive ramp, one pad reaches target faster than the other. You see more tombstone near the edge of the panel or near big copper. A smoother preheat and soak gives both pads time to follow each other.
Common practice on many lines:
You don’t need perfect numbers. The key idea: check the thermal profile on real product. Put thermocouples close to chip arrays where you see tombstone and tune the zones or conveyor speed.
The second big driver is solder paste imbalance. If one pad gets more paste, it has more molten solder, so it pulls harder when it wets.
On the stencil, both apertures for one chip should print almost the same volume. Real life is messy: worn squeegee, dirty stencil, wrong cleaning time. Still, you can tune the design:
You can even do a small line test: normal apertures vs reduced on the hot side, then compare tombstone rate. Fast shop-floor experiment, not big academic research.
Layout also plays big role. If pads and copper are not balanced, you almost invite tombstoning.
Typical layout traps:
All these change local heating and wetting. One side heats slower, but has more copper and more paste, so once it finally melts it pulls hard.
Good habits:
This is same logic like designing stable rear wire shelving in our factory at QIAO: if the feet and cross wires are not symmetric, the rack will shaking. The PCB behaves similar under thermal stress.
Very small packages like 0402 and 0201 are more sensitive. Their mass is low, so they lift easier. Different suppliers can also have different plating and tin structure, which changes wetting speed.
You can:
Sometimes engineers blame only the oven, but the component lot is also part of the story.
Placement also drives tombstoning. If the chip is already off-center before reflow, one termination sits in a small paste island, the other in a deep “pool”.
Check these points on your pick-and-place:
When the line guys at QIAO debug a new freezer control board, they always run slow speed first, with extra board support pins. Only when tombstone rate is under control they start to push for takt time.
Even with good rules, real product is complex. So you need feedback tools.
Think of it like designing a custom wire rack system: you don’t just weld and ship. You test load, bending, vibration, then you tune the wire gauge and welding pattern. Same mindset for SMT work.
| Main factor | Typical symptom on line | Simple fix ideas |
|---|---|---|
| Uneven heating / bad profile | More tombstone near edge or near heavy copper | Smooth ramp, add soak, adjust zone temps, re-profile real product |
| Paste volume / stencil | One pad with big shiny solder, other side almost dry | Balance apertures, use thinner stencil, improve cleaning and paste control |
| Pad and copper asymmetry | Same component always tombstones in same PCB area | Make pads symmetric, add thermal relief, move or mirror vias |
| Component and material | Certain lot or very small size parts fail more | Prefer bigger size, qualify suppliers, control storage |
| Placement and handling | Chips look skewed before reflow, more defects when speed up | Fix nozzle choice and Z-height, improve board support, reduce vibration |
You maybe think: “We are in wire shelving and refrigeration parts, why we talk so much SMT?” In real life they are tied together.
Cold storage room components, commercial display cabinet controllers, small PCBs hidden near the compressor protection mesh — all of them depend on stable SMT quality. If a tiny resistor tombstones and opens, your nice cabinet lights turn off, fans stop, or temperature runs out of spec. End users don’t complain about tombstoning; they complain that their food or drink is warm.
At QIAO we see both sides: we build custom rear mesh, freezer baskets, freezer components and other metal parts, and we also talk every day with customers’ engineering teams. Many of them now ask their PCB partners directly: “Show me your tombstone rate on small passives.” It became a common pain point in new product launch.
So, if you design hardware for supermarket freezers, beverage coolers, or lab cold rooms, don’t treat tombstoning like a small cosmetic defect. It is a signal that force balance in your SMT scene is wrong. Fix the thermal, paste, pads, parts, and placement early, and your shelves, racks, and compressors will have a much more quiet life later.
