# CC-L2-01: SMT Assembly in Depth — Equipment, Process Control, and Defect Modes **Clark Courses — Level 2: Professional Practice** Hamilton, Ontario | CLARK Electronics Manufacturing --- ## 1. Course Overview This course provides a thorough, technically grounded examination of the surface-mount technology assembly process from solder paste selection through reflow and defect analysis. Students move beyond operational familiarity to develop the process knowledge required to make informed decisions about materials, equipment, and parameters. The course addresses the complete SMT line — stencil printing, solder paste inspection, pick-and-place, and reflow — treating each as an interdependent system rather than a sequence of isolated steps. By the end of the course, participants will be equipped to read process data critically, diagnose defect root causes, and contribute to continuous improvement efforts on a production floor. --- ## 2. Learning Objectives - Select solder paste alloys and flux chemistries appropriate to a given board assembly requirement, citing relevant material properties and IPC classifications. - Define and apply the area ratio and aspect ratio rules to evaluate stencil aperture designs and predict print quality outcomes. - Explain the operating principles of high-speed gantry, rotary turret, and flexible placement machines and identify the trade-offs in speed, accuracy, and part range. - Interpret a reflow thermal profile, identify the function of each zone, and diagnose common profile-related defect modes including head-in-pillow, graping, and voiding. - Apply statistical process control concepts — including Cpk, Ppk, and control chart interpretation — to paste printing data generated by an SPI system. - Construct a defect catalog entry for any of the major SMT defect types, correctly identifying the mechanism, likely root cause categories, and corrective actions. - Integrate SPI, AOI, and yield data to form a process assessment and prioritize improvement actions by impact. --- ## 3. Target Audience Process engineers, quality engineers, manufacturing engineers, and senior technicians working in SMT assembly environments who need to move from task execution to process ownership. Also appropriate for equipment application engineers, EMS account managers who need technical depth, and engineers transitioning into electronics manufacturing from adjacent disciplines. --- ## 4. Prerequisites Recommended Level 1 completions: - CC-L1-01: Introduction to PCB Assembly — How Electronics Gets Built - CC-L1-02: Soldering Fundamentals — Materials, Metallurgy, and Joints Participants should be comfortable reading a PCB assembly drawing, identifying component types on a board, and have at least observational familiarity with an SMT production line. --- ## 5. Duration and Format **Duration:** 2 days (16 hours of instruction) **Format:** Instructor-led classroom with integrated lab sessions. Approximately 60% lecture and discussion, 40% hands-on and observation exercises. Lab access to an operational SMT line, SPI system, and AOI system is required. --- ## 6. Module Outline ### Module 1: Solder Paste — Alloys, Flux Chemistry, and IPC Classification **Session 1.1: Alloy Composition and Selection** The dominant assembly alloy, SAC305 (96.5% Sn, 3% Ag, 0.5% Cu), is examined in terms of its melting point (217–220°C), mechanical properties, and cost drivers relative to SAC405. The course also covers low-silver alloys developed to reduce silver cost, and low-temperature alloys such as Bismuth-Tin (BiSn, liquidus approximately 138°C) used for temperature-sensitive substrates and bottom-side reflow in double-sided assemblies. Students learn to assess alloy selection against board thermal budget, component temperature ratings, and application environment (particularly the concern with BiSn joint brittleness under mechanical stress). **Session 1.2: Flux Chemistry — RMA, ROL0, REL0, and Water-Soluble** Flux performs four functions: oxide removal, surface tension reduction, heat transfer improvement, and protecting the cleaned surface until solidification. RMA (Rosin Mildly Activated) fluxes are well-established, reliable, and leave a rosin residue that is benign in most applications but must be removed in Class 3 or high-frequency applications. No-clean fluxes (ROL0, REL0 per IPC J-STD-004 classification) are formulated so that residues are electrically safe and do not require cleaning — the course examines the conditions under which this assumption breaks down (insufficient heat activation, residue under low standoff components, SIR test requirements). Water-soluble fluxes provide the highest activation level and produce the cleanest surfaces when properly cleaned, but introduce a mandatory post-solder cleaning step. **Session 1.3: Paste Types by IPC Metal Powder Classification (Type 3, 4, 5, 6)** Solder paste is classified by IPC J-STD-005A based on metal powder particle size. Type 3 (25–45 µm) is the production workhorse suitable for 0402 components and above. Type 4 (20–38 µm) is used for 0201 and fine-pitch QFP/QFN applications. Type 5 (10–25 µm) and Type 6 (5–15 µm) are used for 01005 components and advanced packages where aperture openings are too small for larger particles without risking clogging. Students will evaluate the cost, shelf life, and printability trade-offs at each powder type level. --- ### Module 2: Stencil Technology and Print Process **Session 2.1: Stencil Materials and Aperture Design Rules** Laser-cut stainless steel stencils are the industry standard. Nickel-plated stencils offer improved paste release on fine-pitch apertures, and nano-coated stencils (hydrophobic surface treatment) further reduce paste adhesion to the stencil wall, improving transfer efficiency on small apertures. The area ratio rule (aperture area divided by aperture wall area) must exceed 0.66 — ideally above 0.80 — to achieve reliable paste release. The aspect ratio (aperture width divided by stencil thickness) must exceed 1.5. Students will work through sample aperture calculations for a variety of component types and determine appropriate stencil thicknesses. **Session 2.2: Print Parameters — Squeegee, Pressure, Separation, and Snap-Off** Squeegee speed (typically 25–75 mm/s) must be balanced against paste viscosity: too fast produces insufficient time for paste to roll and fill the aperture; too slow can push paste under the stencil. Squeegee pressure determines wiping cleanliness — too low leaves paste smear, too high causes squeegee blade deflection and paste scooping from apertures. Separation speed (the rate at which the stencil lifts away from the board) is critical: slow separation gives paste time to release cleanly, but fast separation in tight process windows causes dog-ear or pillow defects. Students will map the parameter interactions through a DOE structure. **Session 2.3: Solder Paste Inspection — 2D vs 3D, Measurement Targets, and SPC Integration** 2D SPI measures paste area coverage and X/Y offset from nominal pad position. 3D SPI adds height measurement via structured light or confocal methods, enabling calculation of paste volume — the most predictive single metric for solder joint quality. Key measurements include volume (as a percentage of nominal), area (as a percentage of pad area), height, and X/Y offset. SPC integration involves mapping SPI data into control charts — typically X-bar/R or I/MR charts — for key parameters per stencil aperture group. Cpk targets of 1.33 or higher are typical for volume on critical components. Students will interpret representative SPI data sets and identify out-of-control conditions. --- ### Module 3: Pick-and-Place — Machines, Feeders, and Vision **Session 3.1: Machine Architectures — Gantry, Turret, and Flexible** High-speed rotary turret machines (e.g., Fuji NXT series, Panasonic NPM) achieve high placement rates (up to 100,000+ CPH) by cycling multiple heads on a rotating turret over a stationary board, but are optimized for chip components and standard ICs. Gantry (portal) machines use an X/Y moving head system suited to large, odd-form, or high-accuracy components — ICs with fine pitch, BGAs, connectors — at lower throughput. Flexible placement platforms use modular beam or multi-head gantry architectures that can be configured for high volume or high mix. Students will identify which machine type is appropriate for a given component population and production volume scenario. **Session 3.2: Feeders, Nozzles, and Component Presentation** Tape-and-reel feeders (8 mm, 12 mm, 16 mm, 24 mm, and larger) are the dominant format for standard SMT components. Tube feeders handle DIP and SOIC components. Tray (JEDEC tray) feeders handle BGAs, large ICs, and sensitive components. Vibratory bowl feeders handle unpackaged or bulk components. Nozzle selection — governed by component weight, body size, and surface — affects pick reliability and placement force. Worn or mismatched nozzles are a leading cause of pick errors, damaged components, and placement inaccuracy. Students will match nozzles to a sample component list and identify appropriate feeder types. **Session 3.3: Vision Systems — Upward Camera, Downward Camera, and Fiducial Recognition** The upward (component recognition) camera images the component after pick, before placement, to determine precise lead geometry, body position relative to nozzle center, and presence/absence of leads or balls. This data corrects the placement position in real time. Downward cameras support board fiducial recognition, which compensates for board-to-board variation in registration. Fiducial marks — etched copper circles — provide two or more reference points on the board; the machine's vision system calculates any X/Y offset and theta rotation and corrects the program accordingly. Students will trace through the placement correction math for a simple two-fiducial example. --- ### Module 4: Reflow — Profiles, Atmospheres, and Profile Defects **Session 4.1: Reflow Zone Configuration and Thermal Profiling** A convection reflow oven typically has 8–14 zones: preheat zones (ramp rate typically 1–3°C/s), soak zones (isothermal hold to activate flux and equalize board temperature), ramp-to-peak zones, and cooling zones. Thermocouple attachment (adhesive dots, Kapton tape, clamped leads) and datalogger selection are discussed, with emphasis on attaching thermocouples at critical locations: the smallest component, the largest thermal mass component, and BGA balls or corner pads. Profile parameters for SAC305 (peak 235–250°C, time above liquidus 30–90 seconds, cooling rate less than 4°C/s) differ from SnPb eutectic profiles and BiSn low-temperature profiles; students will characterize the differences and identify which parameters to adjust when changing alloys or board mass. **Session 4.2: Nitrogen Atmosphere — When and Why** Nitrogen (N2) atmosphere in the reflow oven suppresses oxide formation on the molten solder and solder pads, improving wetting and reducing dross. It is used when: processing with very low-activity no-clean flux that cannot overcome surface oxides on its own; assembling with fine-pitch components where wetting defects are unacceptable; processing OSP-finished boards where oxide growth is rapid; or processing any low-temperature BiSn assembly where reduced flux activation makes oxide suppression important. The course covers the economics of N2 use (cost per hour, N2 consumption rate, sealed vs. open oven considerations) to help students make a justified recommendation. **Session 4.3: Profile-Related Defects — Cold Joint, HIP, Graping, Voiding** Cold joints result from insufficient peak temperature or time above liquidus; the joint appears dull and grainy with incomplete intermetallic formation. Head-in-pillow (HIP) is specific to BGA assemblies: the ball and the paste deposit fail to coalesce during reflow, leaving a pillow-shaped non-wet interface — typically caused by board or component warpage separating ball from paste during the liquidus window. Graping (also called micrograping or cold solder on BGA) presents as a cluster of small, incompletely merged solder spheres, caused by insufficient flux activity, low peak temperature, or contaminated paste. Voiding in BGA joints occurs when flux volatiles are trapped under the ball during solidification — driven by flux chemistry, peak temperature, time above liquidus, and via-in-pad design. Students will diagnose profile defects from photographs and thermal profile traces. --- ### Module 5: Statistical Process Control and First-Pass Yield **Session 5.1: Cpk and Ppk for Paste Printing** Cpk (process capability index) compares process variation to the specification tolerance, using the within-subgroup standard deviation — it reflects instantaneous process capability and is sensitive to shifts and drifts. Ppk uses overall standard deviation including between-subgroup variation and reflects longer-term performance. Both are calculated as the minimum of (USL − mean)/3σ and (mean − LSL)/3σ. For solder paste volume, a Cpk of 1.33 (equivalent to 4-sigma capability) is a typical production target; Class 3 assemblies may demand 1.67. Students will calculate Cpk and Ppk from provided SPI data sets and interpret what each result implies about process condition. **Session 5.2: SPC Charts — Selection, Setup, and Interpretation** X-bar/R charts monitor the process mean and range within subgroups sampled over time and are appropriate when subgroup sizes are fixed (e.g., five boards per sample). Individuals and moving range (I/MR) charts are used when measurement frequency is low or boards are measured one-at-a-time. Students will identify the eight Western Electric rules for detecting special cause variation on a control chart, practice applying them to real SPC data, and discuss the operational response when a chart signals out-of-control. **Session 5.3: DPMO, First-Pass Yield, and the Rolled Throughput Yield Concept** Defects Per Million Opportunities (DPMO) normalizes defect counts against the number of solder opportunities on a board, enabling comparison across board designs. First-pass yield (FPY) is the fraction of boards passing all inspections without any rework. Rolled Throughput Yield (RTY) multiplies the FPY of each process step to produce a system-level yield metric that exposes hidden rework. Students will calculate DPMO and RTY for a multi-step process example and discuss how these metrics drive prioritization in a continuous improvement program. --- ### Module 6: SMT Defect Catalog — Mechanism, Cause, and Prevention **Session 6.1: Bridging, Solder Balls, and Tombstoning** Solder bridging (unwanted solder connecting adjacent pads or leads) results from excessive paste volume, smeared paste from a printing anomaly, or component placement shift. Prevention involves aperture reduction, correct squeegee parameters, and SPI monitoring. Solder balls are small, isolated solder spheres ejected during reflow; causes include paste slump, excessive moisture absorption by paste, aggressive ramp rates, and contaminated or low-quality paste. Tombstoning is rotation of a passive component (resistor, capacitor) to a near-vertical position during reflow, caused by unequal wetting forces on the two pads — driven by pad imbalance, stencil aperture misalignment, thermal asymmetry, or paste volume variation between the two ends. **Session 6.2: Non-Wet Opens, Head-in-Pillow, and Cold Joints** Non-wet opens occur when solder fails to wet to the component lead or board pad; the solder may ball up but does not form an intermetallic bond. Root causes include oxidized component leads, contaminated pads (OSP breakdown, flux residue from prior processing), insufficient flux activity, or incorrect profile. Head-in-pillow is discussed in the context of BGA component warpage — a separate mechanism from non-wet, in that the surfaces are clean but physically separated during the liquidus window. Cold joints are distinguished by their metallurgical signature: they have minimal intermetallic compound formation and poor mechanical strength, typically appearing dull on cross-section. **Session 6.3: Voiding, Insufficient Solder, Component Shift, and Placement Errors** BGA voiding is evaluated by X-ray; IPC-7095 provides guidelines for acceptable void percentage (typically less than 25% per ball void area, less than 9% for thermal management balls). Insufficient solder on QFP leads or passive components results from low paste volume at print (stencil clogging, bad aperture, or squeegee skip) or from component displacement. Component shift occurs when a placed component migrates during paste tack loss, oven conveyor vibration, or solder self-centering failure. Placement errors — wrong rotation, wrong polarity, wrong part — are addressed by tracing the failure back to feeder setup, vision system calibration, and program verification practices. --- ## 7. Key Terms and Concepts **SAC305** — A tin-silver-copper solder alloy (96.5% Sn, 3% Ag, 0.5% Cu) that is the dominant lead-free assembly alloy in commercial electronics. **BiSn** — A bismuth-tin alloy with a liquidus near 138°C used for low-temperature assembly of temperature-sensitive components or substrates. **ROL0** — An IPC J-STD-004 flux classification denoting a rosin/resin-based, no-clean flux with zero halide content and low activity level. **Area ratio** — The ratio of the stencil aperture opening area to the aperture wall area; values above 0.66 are required for reliable paste release. **Aspect ratio** — The ratio of the stencil aperture width to the stencil thickness; values above 1.5 are required to prevent paste bridging in the aperture. **Transfer efficiency** — The percentage of solder paste in a stencil aperture that successfully transfers to the PCB pad; influenced by aperture geometry, stencil material, and surface finish. **SPI (Solder Paste Inspection)** — An in-line automated measurement system that quantifies paste deposit volume, area, height, and position immediately after stencil printing. **Cpk** — A process capability index using within-subgroup variation that quantifies how well a process fits within its specification limits; higher is better, with 1.33 as a common minimum target. **Head-in-pillow (HIP)** — A BGA solder joint defect in which the component ball and the paste deposit fail to merge during reflow due to component warpage separating them at the liquidus window. **Tombstoning** — The lifting and rotation of a passive SMT component to a near-vertical stance during reflow, caused by unequal solder wetting forces on its two terminations. **Time above liquidus (TAL)** — The duration during a reflow profile that the solder is in molten state, typically 30–90 seconds for SAC305; insufficient TAL leads to incomplete wetting and cold joints. **Voiding** — Entrapment of gas bubbles in a solder joint during solidification, particularly visible in BGA joints under X-ray; excessive voiding degrades thermal and mechanical performance. **DPMO** — Defects Per Million Opportunities; a normalized defect rate metric that enables comparison of quality levels across different board designs and assembly processes. **Snap-off (separation speed)** — The controlled rate at which the stencil is separated from the board after printing; slow separation promotes clean paste release while fast separation can cause paste bridging or pad smear. **Rolled Throughput Yield (RTY)** — The product of first-pass yields across all process steps, revealing the true cumulative effect of process escapes and hidden rework on overall efficiency. --- ## 8. Instructor Notes This course requires access to an operational SMT line for lab sessions, including at minimum a stencil printer, SPI station, pick-and-place machine, and reflow oven. Instructors should arrange for the SPI and AOI systems to be available in a teach/demo mode so that participants can interact with the software and data displays without interrupting production. Instructors should bring a physical defect sample kit with cross-sections, microscope photos, and if possible preserved boards showing each defect type in Module 6. The Clark facility sample library should be consulted and supplemented if any common defect types are under-represented. For Module 5 (SPC and yield), prepare a dataset in advance — ideally real anonymized production data with some intentionally planted anomalies — so that participants practice chart interpretation on realistic, messy data rather than textbook-clean examples. The calculation exercises for Cpk should be done by hand at least once before transitioning to spreadsheet tools. The instructor should be prepared for discussion to run long in Module 6. The defect catalog section tends to generate floor-experience sharing from participants, which is valuable — budget time flexibility here and use it to connect mechanisms to the earlier process parameter content. Avoid presenting defect root causes as single-factor. The most important lesson in this section is that most SMT defects have multiple interacting contributors, and corrective action requires identifying the dominant factor in a specific process context rather than applying a generic rule. --- ## 9. Hands-On and Discussion Exercises ### Exercise 1: Stencil Aperture Design Review Participants receive a component placement list for a representative mixed-density board (including 0402 passives, a 0.5 mm pitch QFP, a BGA, and a QFN). Using the area ratio and aspect ratio rules, they calculate whether a single stencil thickness (0.127 mm) satisfies the rules for all component types, identify the conflicting requirements, and propose a step-stencil design or alternative solution. Discussion covers the practical limits of step-stencil manufacturing, aperture area reduction as a technique, and when to use a separate fine-pitch stencil. ### Exercise 2: Reflow Profile Diagnosis Participants are given three thermal profile traces (from a datalogger run through the oven on a representative board assembly) and a corresponding set of defect observations from the production run: one profile associated with tombstoning and solder balls, one associated with cold joints and HIP on a BGA, and one that is nominal. Working in small groups, participants must identify what is wrong with each profile (ramp rate, soak time, peak temperature, cooling rate), connect the profile characteristics to the observed defects, and propose specific parameter changes to move toward the nominal. Groups present their analysis and recommendations. ### Exercise 3: SPC Data Interpretation Challenge Participants receive a printed SPC run chart for paste volume on a critical BGA component across a 5-day production period. The chart contains at least three Western Electric rule violations embedded in the data. Participants must: identify all violations, classify each as a trend, run, or point anomaly, hypothesize what process events could produce each signal, and recommend which signals warrant immediate process stop vs. investigation while running. The exercise concludes with a discussion of what an appropriate SPC response protocol looks like in a production environment. --- ## 10. Assessment Suggestions **Knowledge Check Quiz (End of Day 1):** 15–20 multiple-choice and short-answer questions covering alloy selection, flux classification, area/aspect ratio rules, and print parameter effects. Used formatively to identify gaps before the second day's lab-heavy sessions. **Defect Root Cause Report:** Each participant selects two defect types from the Module 6 catalog and writes a one-page root cause analysis for each, structured as: defect description, mechanism, primary root cause categories, contributory factors, recommended corrective actions, and one process control measure to prevent recurrence. Evaluated on technical accuracy, completeness of cause analysis, and practicality of recommendations. **SPC Response Scenario:** A written scenario presents a production situation where SPC charts for paste volume are showing a trending signal on one of three BGA components. Participants write a brief (one page) process response recommendation, including: what additional data to gather, what immediate containment actions are appropriate, what root cause hypotheses to investigate, and who should be involved. Evaluated on systematic thinking and appropriate application of SPC concepts. --- ## 11. Recommended Resources - IPC J-STD-001: Requirements for Soldering Electrical and Electronic Assemblies (current revision) - IPC-A-610: Acceptability of Electronic Assemblies (current revision) — particularly the SMT defect acceptance criteria - IPC J-STD-005A: Requirements for Soldering Pastes - IPC J-STD-004B: Requirements for Soldering Fluxes - IPC-7527: Requirements for Stencil Design - IPC-7095: Design and Assembly Process Implementation for BGAs - Clyde Coombs (ed.), *Printed Circuits Handbook*, McGraw-Hill — Chapters on SMT assembly and reflow - Ray Prasad, *Surface Mount Technology: Principles and Practice*, Springer — foundational reference for SMT process engineering - SMTA (Surface Mount Technology Association): smta.org — technical papers library, particularly the SMTAI proceedings for defect mode research - Indium Corporation Technical Library (indium.com) — application notes on paste selection, profile optimization, and low-temperature soldering - Kester Solder Technical Resources — flux and paste application notes