Parts of an electronics circuit
Circuits are made of a combination of many parts. Some of these parts connect to the PCB (Printed Circuit Board) with leads coming out of the component package. The package leads are soldered to copper pads on the PCB. Some leads get folded under the package to reduce the package footprint on high density PCBs with no thickness restrictions. When the circuit card is limited by thickness, the leads can be clipped. This effectively creates a copper pad on the bottom of the package that is soldered to the pad on the PCB.
To lead or not to lead
The copper leads coming out of the package serve electrical and mechanical functions. They act like springs and help protect the solder joint. When the package leads are replaced by pads then all the mechanical stresses due to CTE mismatch are in the solder joint. The use of SMT (Surface Mount Technology) packages allows for dense PCBs at the cost of reliability.
If we take the lead height to a ridiculous point, where the lead is super tall, the solder joint will not fail due to CTE mismatch between the package and the PCB. The lead shape is so elastic that it bears all the stress. However, there are no free lunches, some other failure mode like vibration fatigue will now become dominant.
Shear distance
The lead height creates a distance between the part package and the PCB. The part package could be a ceramic (resistors, capacitors...etc) it can be a plastic (overmolded leadframe, diodes...etc) or some combination. The CTE mismatch between the PCB and the component acts in shear. Shear does not have a direction (unlike tension and compression) and repetitive applications of shear stress are known to cause fatigue failure in metals. For the same CTE mismatch, a larger “shear distance” will reduce the shear stress.
The combination of materials and shapes between the bottom of the package and the top of the PCB create an effective elasticity. Taken to a ridiculous extent, if we solder two materials with microscopic distance, the solder joint will crack immediately. A zero shear distance means infinite shear stress and immediate failure. Making a tall solder joint will work. Placing a copper shim under a QFN or using copper spheres embedded in a BGA solder ball prevents the solder from collapsing during reflow. In simpler terms, it is possible in some cases to increase the shear distance by making a thicker solder joint, up to a point.
Distance to neutral point
Most electronic parts have at least two solder joints to the PCB. A symmetric package with two joints will have the same stress on each of the joints. The solder joints at the corners have the largest stress since most part packages are square or rectangular. The center of the part package is called the neutral point because it has the least stress. The package corners have the largest DNP (Distance to Neutral Point) and therefore higher stress and solder joints at the corner usually fail first.
CTE mismatch can’t be avoided
Solder joint fatigue failure is driven by the CTE mismatch between two sides of the solder. There is copper from the package and copper from the PCB on either side of the solder joint. These two copper plates move in a Shear direction back and forth in every temperature cycle.
Is it possible to eliminate the CTE mismatch and therefore eliminate this failure? No.
The Silicon CTE can’t be changed. Resistors, capacitors, inductors, diodes...etc, are made of different materials that all have different CTE and stiffness. It is not practical to match a single PCB materials to all the parts. It is possible to identify critical components and try to mitigate them.
It is all a trade-off between electrical, RF, mechanical, thermal and reliability performance. Manufacture-ability, costs and time constrains also complicate things. The math is rarely simple.
Copyright Gil Sharon May 6, 2025 . All rights reserved.