Most caps are fixed to the lamp with a cement that has to survive the heat of the soldering or welding operation that attaches the lead wires, and then survive thousands of hours at operating temperature. A cap that passes every dimensional gauge and fails in service is often a cement problem, not a dimension problem.
Solder resistance testing applies a defined thermal cycle representing the assembly operation and then checks that the cap has not rotated, shifted axially, or developed a leak at the seal. The pass criteria are mechanical: a torque applied after the cycle must not move the cap relative to the bulb, and the insulation path must still meet its resistance requirement.
Three causes dominate. Under-cured cement, which softens at a lower temperature than specified. Excess cement, which expands and cracks the shell when it heats. Contaminated surfaces, where the cement bonds to oxide rather than metal. All three are process issues that a supplier can fix quickly once they are measured, and none of them show up in a dimensional inspection.
Solder resistance is a destructive test, so it belongs in type testing and in periodic audit rather than in lot acceptance. Sample at the start of a production run, after any change of cement batch, and after any change to the soldering process.
