CHIP DESIGN · COURSE

Package, PCB, SI/PI, Thermal & Laboratory Engineering

Carry chip interfaces from die pads into a safe measurable physical system. Select package technology from electrical, thermal, mechanical, test, cost, reliability, and supply constraints. Reconcile pad, bump/bond, substrate, ball, PCB, fixture, tester, and firmware maps; then design escape routing and assembly tolerances. Release a controlled PCB stackup, placement, return path, routing, fabrication, and assembly plan. Analyze transmission-line flight, reflections, termination, crosstalk, loss, jitter, and bounded eyes. Build a target-impedance PDN with real regulator, decoupling, anti-resonance, startup, and fault behavior. Connect power maps, hotspots, interfaces, airflow, stress, and mission profiles to thermal and mechanical reliability. Create ESD-safe staged laboratory procedures with calibrated instruments, known loading, uncertainty, and validated de-embedding. Finally, correlate one traceable die–package–board system without mistaking one booting board for production qualification.

Before this course: Completed Physical Design, Power & Signal Integrity; Analog & Mixed-Signal Foundations; Tapeout; PDK/Design Rules; IP/I/O Qualification; EDA Automation; and Embedded Firmware Co-Design. Requires circuits, waves, impedance, S-parameters, probability, heat transfer, mechanics, PCB basics, and safe laboratory practice. Package design kits, fabrication/assembly vendors, controlled stackups, field solvers, calibrated instruments, chambers, safety training, components, and silicon require lawful external resources and are not supplied.

COURSE FACTSStage, chapters, units, prerequisite, and outcome
Chapter 1

Package choice is a multi-domain product decision

Objective: For “Package choice is a multi-domain product decision,” which frozen inputs determine the result, what is the first independently observable claim, and which mutation proves the check is alive?

Wire bond, flip chip, wafer-level, leadframe, laminate, interposer, and other options differ in pin count, pitch, parasitics, power delivery, heat path, warpage, assembly, board cost, test access, yield, and supply. This lesson uses the route “build the smallest observable case.” Begin with a hand-checkable instance before invoking automation: name the state that enters the step, the transformation that is permitted, the observation that must change, and the evidence that would falsify the claim. Select package technology from electrical, thermal, mechanical, assembly, test, cost, volume, reliability, and supply constraints. Connect every abstraction back to the physical structure or executable evidence it represents, and state where that representation stops being reliable.

Package choice is a multi-domain product decision has a reviewable contract: The selected package satisfies every mandatory interface, current, impedance, thermal, mechanical, manufacturing, inspection, reliability, test, cost, schedule, and sourcing requirement with margin. Name the applicable scope, identities, units, conditions, exclusions, threshold, evidence source, owner, and change rule before using the result. Separate control-plane success from design evidence: a process can exit zero while consuming the wrong revision, skipping work, reusing stale output, suppressing a violation, or publishing an incomplete artifact. A low-cost wire-bond package meets pin count but bond inductance violates simultaneous-switching ground bounce. The learner must identify the first divergence and repair the dependency, not merely rerun until a dashboard becomes green.

The selected package satisfies every mandatory interface, current, impedance, thermal, mechanical, manufacturing, inspection, reliability, test, cost, schedule, and sourcing requirement with margin. This invariant is accepted only for the named candidate and declared environment; any changed input invalidates every dependent result until reconstruction proves otherwise.

Freeze the exact objects, conditions, units, and source evidence in the worked case “Score two teaching packages against 20 mandatory and weighted requirements, including one current/inductance limit.” First freeze the candidate and predict the expected observation without reading a generated summary.

Apply the stated physical or engineering model, showing each transformation and preserving values that fail, are missing, or remain outside the model. Then execute the smallest transformation, retaining raw standard output, standard error, exit status, generated files, and resource use.

Compare the derived observation with “The selected package satisfies every mandatory interface, current, impedance, thermal, mechanical, manufacturing, inspection, reliability, test, cost, schedule, and sourcing requirement with margin.” and identify the first downstream decision invalidated by the failure boundary. Finally reconcile the observation with the invariant, inject the named failure, and verify that the expected consumer refuses the corrupted or stale state.

Score two teaching packages against 20 mandatory and weighted requirements, including one current/inductance limit. Before revealing the trace, predict the exact command or state transition, expected exit and artifact status, first checker that should react, and minimum safe recovery.

  1. Freeze the exact objects, conditions, units, and source evidence in the worked case “Score two teaching packages against 20 mandatory and weighted requirements, including one current/inductance limit.”
  2. Apply the stated physical or engineering model, showing each transformation and preserving values that fail, are missing, or remain outside the model.
  3. Compare the derived observation with “The selected package satisfies every mandatory interface, current, impedance, thermal, mechanical, manufacturing, inspection, reliability, test, cost, schedule, and sourcing requirement with margin.” and identify the first downstream decision invalidated by the failure boundary.

Result: Any mandatory failure rejects the option before weighted preference; cheapest acceptable is different from cheapest listed. Accept the result only after a clean second execution reproduces the decisive artifact and a targeted mutation fails at the predicted boundary.