SMT vs Through-Hole Assembly: Process and Cost Differences

Surface-mount placement scales with component count: passives, ICs, and area-array packages reflow in one or two passes. AOI covers solder joints and polarity efficiently on SMT fields. Fine-pitch and BGA packages essentially require SMT—through-hole equivalents rarely exist at modern densities.

Double-sided SMT is routine: the first side reflows, the board flips, and surface tension holds first-side parts through the second pass—heavy components may need adhesive dots. Package miniaturization keeps pushing the process floor: 0402 passives are unremarkable, 0201 is standard on capable lines, and 01005 demands tighter printing and placement control that not every shop maintains. Ask about the smallest package your assembler runs in production, not what their equipment brochure claims.

SMT yield hinges on pad design, stencil apertures, and finish planarity. OSP age limits and ENIG planarity matter when pitch drops below 0.5 mm. The reflow profile—ramp, soak, time above liquidus—is developed per board, which is why a thermally massive ground plane next to a fine-pitch QFN is an assembly conversation, not just a layout one.

Through-hole suits connectors that see insertion force, large electrolytics, and parts that must survive mechanical stress. Wave or selective solder adds process time and often requires top-side masking on mixed boards, but the joints can be more forgiving under vibration.

Hole fill is the quality metric that matters: IPC-A-610 expects vertical barrel fill (75% minimum for Class 2, with Class 3 stricter on specific conditions), and thermally connected pins on thick boards are the usual offenders. Press-fit connectors sidestep soldering entirely—compliant pins cold-weld into plated holes—but they impose tight hole-diameter and plating tolerances that belong on the fab drawing, not discovered at assembly.

Manual or semi-auto insertion still appears on low-volume industrial builds where NRE for SMT-only connectors does not pay off.

Common sequence: SMT reflow first, then through-hole on the opposite side or selective solder on the same side. Parts that cannot survive reflow must be hand-soldered or selectively soldered after SMT—mark them clearly on the assembly drawing.

Two alternatives change the economics. Pin-in-paste (intrusive reflow) prints paste into through-holes and reflows the through-hole part with the SMT pass—eliminating the wave step entirely, if the component body is reflow-rated and the paste volume works out. Selective soldering replaces full wave for boards with only a handful of through-hole joints, trading cycle time for the elimination of wave pallets and top-side masking.

Design for the sequence: keep heavy through-hole connectors away from delicate SMT areas that need rework access, and specify whether second-side SMT requires glue or dedicated carriers.

SMT quotes reflect placement count, package mix, X-ray needs, and stencil complexity—not just board size. Through-hole adds fixture, wave, or selective programming. Mixed boards carry both.

At prototype quantities, setup dominates: stencil fabrication, feeder loading, first-article programming, and wave or selective fixtures are amortized over 10 boards, not 10,000. This is why a 25-piece mixed-technology build can cost more in process setup than in components—and why the highest-leverage cost question at NPI is “can we eliminate one process pass,” not “can we shave placement cents.”

When costing a prototype, ask whether moving a connector to SMT saves a wave setup on a quantity of 25—it often does not. At pilot volume, process steps dominate more than per-joint placement cents.