This compact index covers the accepted 10-module, 56-lesson curriculum in order. It binds 18 tests and 18 quantitative plot jobs to 26 accepted P3 scenario manifests. It is a review candidate, not an acceptance claim.
Chapter 1 counterfactual: the hard-switching case is an avoided-cost teaching counterfactual. It is not nominal LLC loss or measured efficiency.
Outcome: Distinguish hard-switching overlap from charge-complete ZVS at one frozen commutation edge.
Decision: Decide whether the edge has helpful current polarity and sufficient charge movement before turn-on.
Evidence: switched_model
Outcome: Classify each power-stage block and state its energy-transfer role.
Decision: Decide where a proposed observation or loss belongs in the power path.
Evidence: no_simulator_evidence
A static labelled schematic and learner annotation are the direct evidence; numerical interaction would add false precision before interval analysis.
Learner action: Annotate a static power-path schematic.
Observable: Bridge, tank, transformer, rectifier and output-filter boundaries.
Outcome: Apply the prerequisite concepts and measurement-safety constraints needed to read later evidence.
Decision: Decide whether a learner is ready to interpret a full-period result or must remediate a prerequisite.
Evidence: switched_model
Outcome: Match each engineering claim to the evidence layer capable of supporting it.
Decision: Escalate from equation to switched model, component model or bench when the current layer cannot answer the question.
Evidence: comparison_fixture
Outcome: Relate bridge states and dead time to tank excitation over a complete period.
Decision: Choose consistent event origins and polarity for interval analysis.
Evidence: switched_model
Outcome: Decompose primary current into magnetising and reflected load components by interval.
Decision: Determine which current component drives energy transfer or circulation at a chosen instant.
Evidence: switched_model
Outcome: Explain the energy path and current relationship at series resonance.
Decision: Use the resonance case as a reference without treating it as universally optimal.
Evidence: switched_model
Outcome: Identify below-resonance interval changes and capacitive-region risk.
Decision: Decide whether the operating point approaches an unsafe commutation boundary.
Evidence: switched_model
Outcome: Identify above-resonance energy transfer and circulating-current changes.
Decision: Trade reduced gain against frequency and loss consequences.
Evidence: switched_model
Outcome: Classify secondary conduction intervals from current and rectifier state.
Decision: Determine whether commutation and output charging assumptions match the operating region.
Evidence: switched_model
Outcome: Calculate the bridge fundamental using the declared amplitude convention.
Decision: Choose peak or RMS convention and carry it consistently.
Evidence: fha
Outcome: Derive the reflected equivalent AC load with the accepted ratio convention.
Decision: Confirm the load basis before calculating Qe or gain.
Evidence: fha
Outcome: Calculate fr and fp and explain which network each frequency represents.
Decision: Use the correct resonance when interpreting a region or trajectory.
Evidence: fha
Outcome: Build a normalized FHA gain curve from fn, Ln and Qe.
Decision: Choose a normalized parameter set that exposes rather than hides assumptions.
Evidence: fha
Outcome: Read gain and FHA input angle together to identify candidate operating regions.
Decision: Reject points that meet gain while violating the declared region cue.
Evidence: fha
Outcome: Attribute disagreement to model scope, data identity or hardware effects.
Decision: Choose the next evidence layer needed to resolve a claim.
Evidence: comparison_fixture
Outcome: Use FHA input angle as a bounded region cue.
Decision: Exclude candidate points on the wrong side of the region boundary before checking charge.
Evidence: fha
Outcome: Calculate charge margin and interpret current polarity for nominal marginal and failing cases.
Decision: Decide whether the declared current can move 200 nC within 250 ns.
Evidence: equation_screen
Outcome: Identify zero-current secondary commutation candidates and state the evidence needed for reverse recovery.
Decision: Decide whether a current interval supports benign commutation or requires higher-fidelity evidence.
Evidence: switched_model
Outcome: Locate peak gain and identify where required gain loses reserve or frequency margin.
Decision: Reject a requirement corner that cannot be regulated within accepted limits.
Evidence: fha
Outcome: Trace startup and declared fault movement across frequency region current output and protection states.
Decision: Choose or reject a startup and fault-response sequence based on region crossings current limits and reproducible state order.
Evidence: switched_model
Outcome: Convert each requirement corner into a traceable required-gain value.
Decision: Decide which requirement corners govern the design envelope.
Evidence: fha
Outcome: Compare bridge form and turns-ratio choices using explicit gain definitions.
Decision: Select a pair that makes the required envelope attainable without hiding a scaling change.
Evidence: fha
Outcome: Select a first-pass fr Ln and Qe set and expose the coupled trade-offs.
Decision: Choose candidates that meet gain and region requirements before component synthesis.
Evidence: fha
Outcome: Solve the component set and recompute current and capacitor stress screens.
Decision: Accept or revise a candidate based on both electrical targets and stress.
Evidence: fha
Outcome: Verify a candidate against frequency gain region overload and stress constraints together.
Decision: Accept the first-pass design or identify the controlling violated requirement.
Evidence: fha
Outcome: Run a declared corner set and identify which constraint controls each iteration.
Decision: Revise the variable that addresses the controlling failure without concealing new trade-offs.
Evidence: comparison_fixture
Outcome: Explain how Lm moves gain, circulating current and the ZVS screen.
Decision: Choose whether a measured Lm shift requires model or design revision.
Evidence: comparison_fixture
Outcome: Place measured leakage correctly in the tank model and avoid double counting.
Decision: Treat leakage as part of Lr external to Lr or an uncertain split with explicit measurement.
Evidence: comparison_fixture
Outcome: Compare integration choices across tolerance loss construction and measurability.
Decision: Choose an integration strategy and list the evidence needed to control it.
Evidence: no_simulator_evidence
The lesson is an implementation-architecture comparison whose decisive inputs are construction capability tolerance ownership and measurement access, not a numeric plant response.
Learner action: Complete a qualitative architecture trade-off table.
Observable: Construction dependencies, tolerance ownership and measurement access.
Outcome: Trace how winding construction changes parasitics and uncertainty.
Decision: Specify which construction variables and measurements must be controlled.
Evidence: no_simulator_evidence
The accepted depth is causal construction reasoning and evidence planning; a generic simulator would hide geometry and material assumptions that this lesson must expose.
Learner action: Annotate a winding and field-risk diagram.
Observable: Coupling capacitance fringing and tolerance mechanisms.
Outcome: Design bounded fixtures for Lm leakage and tank measurements with uncertainty.
Decision: Decide whether a measurement is admissible for model update.
Evidence: measured_data
Outcome: Recalculate resonances and affected screens from admissible measurements.
Decision: Accept the update revise the design or request better measurement evidence.
Evidence: measured_data
Outcome: Compare secondary conduction paths and reflected effects for both rectifier forms.
Decision: Select a rectifier form consistent with winding current voltage stress and loss evidence.
Evidence: switched_model
Outcome: Relate conduction intervals to capacitor charging ripple and load response.
Decision: Decide whether capacitor ripple current and voltage meet declared limits.
Evidence: switched_model
Outcome: Separate conduction-loss screening from reverse-recovery evidence by region.
Decision: Choose whether equation switched component or bench evidence is needed.
Evidence: switched_model
Outcome: Diagnose early and late SR timing from event-aligned current intervals.
Decision: Choose a timing window that limits body-diode conduction and reverse current.
Evidence: component_model
Outcome: Compare conduction benefit against capacitance and commutation penalties.
Decision: Choose a device candidate using operating-region and timing evidence.
Evidence: component_model
Outcome: Explain why SR disable and light-load mode transitions require coordinated timing.
Decision: Choose transition conditions that avoid reverse current and unacceptable ripple.
Evidence: switched_model
Outcome: Trace required and attainable gain as line and load move.
Decision: Decide whether frequency control stays inside the validated window and region.
Evidence: fha
Outcome: Compare what each control concept senses commands and limits.
Decision: Select a concept whose observable state and constraints match the operating problem.
Evidence: fha
Outcome: Compare loop gain and phase across declared operating points and evidence layers.
Decision: Decide whether compensation and margins require revision or measurement.
Evidence: component_model
Outcome: Compare startup sequences from zero and prebiased outputs.
Decision: Select an event sequence that controls current voltage and region crossings.
Evidence: switched_model
Outcome: Compare light-load modes and quantify simulated ripple and switching activity.
Decision: Choose transition limits while identifying which audible and EMI claims require hardware.
Evidence: switched_model
Outcome: Verify event order and bounded responses for transient and protection cases.
Decision: Accept revise or reject the protection state machine and limits.
Evidence: switched_model
Outcome: Build a source-badged primary loss budget across operating points.
Decision: Identify the primary loss term that needs design action or better evidence.
Evidence: component_model
Outcome: Separate tank and magnetics loss mechanisms by equation model or measurement source.
Decision: Choose which construction or component change addresses the dominant supported loss.
Evidence: component_model
Outcome: Attribute secondary and output losses across rectifier modes and load.
Decision: Select rectifier and capacitor changes based on the supported dominant term.
Evidence: component_model
Outcome: Identify which tolerance and temperature corner controls each margin.
Decision: Revise limits components or evidence plan for failing corners.
Evidence: comparison_fixture
Outcome: Compare layers without granting unsupported authority to any trace.
Decision: Decide which evidence gap blocks a claim and what test closes it.
Evidence: comparison_fixture
Outcome: Design a safe measurement plan with bandwidth reference and uncertainty controls.
Decision: Decide whether a proposed capture can support the intended engineering claim.
Evidence: no_simulator_evidence
The evidence is a safety-critical measurement plan and capture audit; a simulator cannot validate probe isolation loading bandwidth or operator procedure.
Learner action: Complete a probe and capture evidence checklist.
Observable: Safety reference loading bandwidth de-embedding and uncertainty requirements.
Outcome: Assemble a traceable corner fault and evidence matrix without collapsing unlike jobs.
Decision: Decide whether every critical requirement has bounded evidence or an explicit unknown.
Evidence: comparison_fixture
Outcome: Compare LLC with alternatives using requirement-linked evidence and uncertainty.
Decision: Defend architecture selection or rejection without a hidden weighted winner.
Evidence: comparison_fixture
Outcome: Identify requirement categories that strain or exclude a basic LLC architecture.
Decision: Decide whether added stages complexity or an alternative architecture deserves evaluation.
Evidence: comparison_fixture
Outcome: State the trigger that makes each named alternative worth a separate evaluation.
Decision: Select which alternative enters a follow-on study and what evidence that study must produce.
Evidence: no_simulator_evidence
This lesson defines when to commission a separate architecture study; comparing unscoped models would fabricate precision and duplicate courses not accepted at Gate 1.
Learner action: Map requirements to alternative-architecture study triggers.
Observable: Directionality range control device and implementation distinctions.
Outcome: Produce an evidence-backed architecture and design judgement with explicit unknowns.
Decision: Approve revise reject or request evidence for the proposed LLC design.
Evidence: comparison_fixture
Question: What cost is avoided when the nominal commutation current slews the switch node to zero before turn-on?
Purpose: Isolate only the sign of commutation current at the frozen nominal point and distinguish avoided counterfactual loss from nominal converter behavior.
Model/evidence level: switched_model
scenario_selector: select exactly one frozen casefrozen_parameters: inspect values without editingvSw [V], iComm [A], residualVdsAtTurnOn [V], chargeShortfall [C], linearOverlapEnergyEstimate [J], zvsAchieved [boolean]
Limitations: One static commutation edge only; Constant signed current during dead time; Lumped total charge rather than nonlinear capacitance; Linear overlap screen is not device switching-loss integration; Parasitics ringing reverse recovery and driver dynamics are excluded
Prohibited inference: The 52.94802456797902 uJ and 12.48 W values are avoided counterfactual cost, not nominal LLC loss or measured efficiency.
Question: How do bridge, resonant, magnetising and secondary currents change below, at and above series resonance?
Purpose: Connect switching intervals and secondary conduction directly to full-period waveforms, a job the one-edge P2 model cannot perform.
Model/evidence level: switched_model
frequency_case: select a validated caseplot_periods: choose one or two complete periodsbridge_state_overlay: show the categorical state and interval overlaysvSw [V], iRes [A], iMag [A], iSecReflected [A], currentDecompositionResidual [A], frequencyRegion [enum], secondaryConductionState [enum], secondaryCurrentMode [enum], secondaryCurrentZeroCrossing [event], secondaryZeroCurrentCommutationState [enum], commutationPolarity [enum], bridgeState [enum], deadTimeInterval [boolean], tankExcitationState [enum]
Limitations: GAP-01 is unresolved; Existing shared LLC plant uses a different 300 W tank and cannot be reused without revalidation; No handmade or hand-shaped waveform is allowed; A model waveform cannot prove hardware reverse recovery or safe probing; This test assesses only waveform literacy for les-01-03 and does not assess phasors resonance reflected impedance RMS loss or measurement safety
Prohibited inference: Do not call this switched-model or quantitative plot evidence unless P3 validates the full-period plant against the frozen tank; even after validation it cannot establish the les-01-03 prerequisite pass decision.
Question: How do normalized frequency, inductance ratio and effective quality factor move gain and input impedance angle?
Purpose: Build the normalized design map and expose the inductive or capacitive region cue without treating it as a ZVS proof.
Model/evidence level: fha
bridge_form: select the excitation form and apply its explicit Fourier factoramplitude_convention: select the displayed fundamental convention without changing physical excitationnormalized_frequency: sweep fninductance_ratio: change Lneffective_quality_factor: change Qeload_fraction: change reflected loadbridgeForm [enum], amplitudeConvention [enum], bridgeFundamentalFactor [dimensionless], bridgeFundamentalPeak [V], bridgeFundamentalRms [V], Rac [ohm], Qe [dimensionless], fr [Hz], fp [Hz], frNormalized [dimensionless], fpNormalized [dimensionless], gain [dimensionless], inputImpedanceAngle [deg], regionCue [enum]
Limitations: FHA replaces switched waveforms with fundamental equivalents; Most credible near series resonance; Cannot establish transient behavior switching ripple ZVS or hardware loss
Prohibited inference: Positive FHA angle is an inductive-region cue only; it does not prove that charge moves within dead time.
Question: Which claims can FHA, switched simulation, component models and imported measurements each support?
Purpose: Put aligned outputs beside their evidence badges and limitations so disagreement causes evidence escalation rather than a winner claim.
Model/evidence level: comparison_fixture
evidence_layers: show or hide available layersimported_dataset: select only provenance-bound external dataalignedGainOrWaveform [declared_per_trace], difference [declared], evidenceLevel [enum], evidenceBadge [enum], sourceIdentity [sha256], modelIdentity [string], modelVersion [string], limitation [text], availabilityState [enum]
Limitations: No fixed comparison dataset is accepted; Difference magnitude is case-specific; Imported measurements do not validate unmeasured operating points
Prohibited inference: Do not promote the rejected fixed reference gains or a policy fixture as switched-model or measured truth.
Question: Is the signed commutation current sufficient to move the declared charge before turn-on?
Purpose: Separate the charge boundary from the FHA region cue and show nominal, marginal and wrong-polarity cases.
Model/evidence level: equation_screen
signed_commutation_current_A: select magnitude and polarity including the frozen P2 nominal and its wrong-polarity counterpartdead_time_and_charge: inspect frozen P2 valueschargeAvailable [C], chargeMargin [C], polarityHelpful [boolean], zvsEligibilityState [enum], residualVdsAtTurnOn [V]
Limitations: Equation screen assumes constant current; Signed classification is defined for the declared high-to-low edge only; Qoss is a lumped total charge; Nonlinear capacitance parasitics and timing skew are excluded
Prohibited inference: Passing this screen does not prove device ZVS or hardware margin.
Question: Can the selected bridge and turns ratio meet line, load and overload requirements without leaving the declared frequency and inductive-region limits?
Purpose: Turn requirements into a gain envelope and test reserve, overload and frequency-window boundaries together.
Model/evidence level: fha
requirement_corner: execute one required corner from the complete declared setinput_voltage_case: expose the selected corner input-voltage dimensionoutput_voltage_case: expose the selected corner output-voltage dimensionload_case: expose the selected corner load dimensionhold_up_case: expose whether the selected corner is a hold-up conditionregulation_case: expose the selected corner regulation dimensionbridge_form: compare only with explicit gain-definition transformturns_ratio: evaluate candidate ratio without changing P2 Chapter 1 defaultrequiredGain [dimensionless], attainablePeakGain [dimensionless], gainReserve [percent], solvedFrequency [Hz], impedanceAngle [deg], windowPass [boolean], bridgeForm [enum], turnsRatio [dimensionless], gainDefinitionTransform [string], bridgeRatioSelectionPass [boolean], requirementCornerId [enum], inputVoltageCase [enum], outputVoltageCase [enum], loadCase [enum], holdUpCase [enum], regulationCase [enum], cornerExecutionState [enum], requirementSourceIdentity [string]
Limitations: FHA design screen only; Diode drop loss allowance and temperature must be reinstated beyond Chapter 1; Wider line and load trajectory is not frozen by P2
Prohibited inference: Do not reuse the rejected operating trajectory or silently treat the Chapter 1 supersessions as a settled Module 4 5 or 8 policy.
Question: Which regions and stresses does the converter cross while startup is sequenced from an unenergized or pre-biased output?
Purpose: Show time-ordered startup movement across operating regions without conflating it with steady-state protection policy.
Model/evidence level: switched_model
prebias_case: select initial output conditionsoft_start_profile: change only a validated sequence parameterstartup_case_limits: bind output current and frequency limits to each initial condition before executioninitialConditionId [enum], switchingFrequency [Hz], switchingFrequencyLimit [Hz], switchingFrequencyPass [boolean], normalizedFrequency [dimensionless], inputImpedanceAngle [deg], regionState [enum], regionCrossing [event], resonantCurrent [A], resonantCurrentLimit [A], resonantCurrentPass [boolean], outputVoltage [V], outputVoltageLimit [V], outputVoltagePass [boolean], startupLimitSourceIdentity [string], startupCasePass [boolean]
Limitations: Startup capability is not established; FHA angle overlay is not switched truth; Device stress and protection thresholds require component data
Prohibited inference: This scenario cannot certify hardware startup safety or establish a controller policy before P3 implements and validates the required model.
Question: Which Lr Cr and Lm values realize the selected resonant frequency Ln and Qe, and what first-pass stress follows?
Purpose: Convert normalized choices into physical components and immediately challenge them with frequency gain and stress screens.
Model/evidence level: fha
resonant_frequency: choose design targetinductance_ratio: choose Lneffective_quality_factor: choose Qecomponent_rounding: compare solved and candidate partsselectedFr [Hz], Ln [dimensionless], Qe [dimensionless], Lr [H], Cr [F], Lm [H], recomputedFr [Hz], inputImpedanceAngle [deg], gainReserve [percent], rmsStress [declared], peakStress [declared], componentRecomputePass [boolean], gainPass [boolean], inductiveAnglePass [boolean], stressPass [boolean], candidatePass [boolean], supportStatus [enum]
Limitations: First-pass FHA and equation stress only; No thermal magnetic saturation insulation or layout certification; Component rounding and tolerance require the separate iteration test
Prohibited inference: Do not present unsupported device or magnetic stress as switched-model or measured evidence.
Question: Which component temperature and controller corners move gain ZVS loss and frequency margins enough to require redesign?
Purpose: Propagate declared corners through all supported metrics and preserve unsupported cells as evidence gaps rather than false numbers.
Model/evidence level: comparison_fixture
tolerance_source_values: enter or select component limitstemperature_case: select temperature cornercandidate_design: compare one traceable iterationrequired_corner_set: execute the complete declared corner table in stable ordercornerId [string], cornerOrder [ordinal], cornerSourceIdentity [string], cornerAssumptionIdentity [string], resonance [Hz], resonanceSupportState [enum], gainMargin [percent], gainMarginSupportState [enum], zvsMargin [declared], zvsMarginSupportState [enum], loss [W], lossSupportState [enum], controllerLimit [boolean], controllerLimitSupportState [enum], evidenceGap [enum], controllingConstraint [enum], iterationState [enum], reviewDecision [enum]
Limitations: Corner set depends on source-backed component data; Correlation and manufacturing distribution are not inferred; A completed matrix is not product qualification
Prohibited inference: The rejected resonance-only corner table cannot support current loss or ZVS conclusions, and no unsourced tolerance may be called a device specification.
Question: How do measured leakage magnetising inductance and capacitance move resonance gain and the ZVS screen?
Purpose: Replace nominal magnetics with provenance-bound measurements and show exactly which predictions move.
Model/evidence level: measured_data
measurement_dataset: select or import a low-energy measurement setfixture_type: select the measurement purposefixture_connection: declare the exact energized and terminated nodesinstrument_identity: enter instrument and calibration identitymeasurement_frequency_Hz: enter the measurement frequencyexcitation_amplitude: enter amplitude and unituncertainty_percent: enter bounded measurement uncertaintyleakage_allocation: declare where each provenance-bound leakage contribution enters the series modelfixtureMetadataComplete [boolean], measurementAdmissionState [enum], admissionFailureReasons [list], measuredLr [H], measuredLm [H], measuredCr [F], leakageAllocationId [enum], measuredLeakageContribution [H], externalLrContribution [H], totalModeledSeriesInductance [H], declaredSeriesInductanceTarget [H], allocationEqualityResidual [H], doubleCountDetected [boolean], allocationPass [boolean], leakageAllocationDecision [enum], fr [Hz], fp [Hz], gainShift [percent], circulatingCurrent [A], circulatingCurrentShift [A], zvsScreenShift [declared]
Limitations: The browser does not perform a physical measurement; Mapping leakage into Lr depends on the declared measurement connection; Capacitance coupling fringing and nonlinear effects may require higher-fidelity models
Prohibited inference: Entered numbers without measurement provenance are comparison inputs, not measured evidence.
Question: How do rectifier topology and operating region shape secondary conduction output-capacitor current and diode stress?
Purpose: Prevent a primary-only view by tying secondary path states to output ripple and load response.
Model/evidence level: switched_model
rectifier_topology: compare only if both models pass P3 capability validationload_case: select a source-declared steady or stepped load with explicit pre-step and post-step operating-point identityoperating_region: select a validated frequency-region casediode_parameter_availability: select only a provenance-bound parameter layeroutput_ripple_voltage_limit: enter the accepted peak-to-peak ripple limitoutput_capacitor_rms_current_limit: enter the accepted capacitor RMS-current limitrectifierTopology [enum], operatingRegion [enum], diodeParameterAvailability [enum], reverseRecoveryEvidenceState [enum], evidenceBadge [enum], loadCaseId [string], loadStepSourceIdentity [string], preStepOperatingPointId [string], postStepOperatingPointId [string], loadCommand [fraction], loadStepEventOrigin [s], secondaryCurrent [A], diodeCurrent [A], diodeVoltage [V], outputCapacitorCurrent [A], outputCapacitorRmsCurrent [A], outputCapacitorRmsCurrentLimit [A], outputVoltage [V], outputRipplePeakToPeak [V], outputRippleVoltageLimit [V], outputLimitPass [boolean], outputResponsePeakDeviation [V], outputResponseDeviationLimit [V], recoverySettlingTime [s], recoverySettlingTimeLimit [s], loadResponsePass [boolean], loadResponseDecision [enum], diodeLossScreen [W]
Limitations: Bridge variant capability is unestablished; Ideal or lumped diodes cannot prove reverse recovery; Thermal and layout effects are excluded
Prohibited inference: Do not infer hardware reverse-recovery loss from ideal diode state changes.
Question: When do timing delay body-diode conduction reverse current and MOSFET capacitance erase the conduction benefit of synchronous rectification?
Purpose: Compare timing and device trade-offs on a validated secondary waveform rather than a fixed policy table.
Model/evidence level: component_model
sr_timing: sweep turn-on and turn-off delaydevice_candidate: compare Rds on and capacitance datarectifier_mode: compare from common statedeviceCandidateId [string], deviceSourceIdentity [string], deviceModelIdentity [string], deviceModelVersion [string], deviceTemperature [degC], rdsOn [ohm], capacitanceCharge [C], candidateOutputCapacitance [F], capacitanceReflectionModelIdentity [string], reflectedCapacitanceToTank [F], commonTankOperatingPointId [string], commonSecondaryWaveformId [string], secondaryCurrent [A], tankCurrentBaselineRms [A], tankCurrentCandidateRms [A], tankCurrentBurdenIncrease [A], tankCurrentBurdenLimit [A], tankBurdenPass [boolean], channelConductionTime [s], bodyDiodeTime [s], reverseCurrent [A], conductionLoss [W], bodyDiodeLoss [W], capacitanceLoss [W], timingRisk [enum], deviceTimingDecision [enum]
Limitations: Capability requires a validated secondary and SR model; Device capacitance may be voltage dependent; Gate-loop parasitics and thermal feedback require higher evidence
Prohibited inference: The rejected 8 A fixture and low watch high policy labels are not controller truth.
Question: How does each supported control action move the operating point on the gain surface across line and load?
Purpose: Map control commands to operating-point movement while keeping conceptual mode comparison separate from unproven controller performance.
Model/evidence level: fha
line_and_load: choose operating requirementcontrol_concept: select only an implemented concept or view an explanatory disabled statecomparison_dimension: inspect the same dimension for every declared conceptsensedVariable [enum], commandedVariable [enum], limitingConstraint [enum], conceptComparisonState [enum], commandedFrequency [Hz], operatingPoint [coordinates], gain [dimensionless], impedanceAngle [deg], conceptAvailability [enum]
Limitations: FHA trajectory does not establish dynamic stability; Current or charge and hybrid models may not be feasible in P3; Controller implementation is not bound to one IC
Prohibited inference: The rejected line and load frequency policy is not a validated controller trajectory.
Question: How does plant and loop response change across operating points, and which result is model-based versus imported measurement?
Purpose: Compare loop response at multiple operating points with explicit model or measurement provenance.
Model/evidence level: component_model
operating_point: select line and loadevidence_source: select source with visible badgemagnitude [dB], phase [deg], crossoverFrequency [Hz], bandwidth [Hz], phaseMargin [deg], sourceIdentity [sha256], evidenceBadge [enum], availabilityState [enum], operatingPointLoopPass [boolean]
Limitations: Dynamic model and perturbation method are not yet established; Measured response requires external safe bench work; A single nominal response cannot establish whole-range stability
Prohibited inference: Model-based Bode response is not a measured loop response.
Question: What happens to current output ripple frequency limit and mode state as load falls?
Purpose: Show the physical and control transition at light load, including output ripple and mode hysteresis, without claiming audible perception.
Model/evidence level: switched_model
load_fraction: ramp down or uplight_load_strategy: select only validated continuous SR-disabled burst or skip modesmode_transition_limits: bind entry exit reverse-current and ripple limits before mode evaluationswitchingFrequency [Hz], burstState [enum], srEnable [boolean], outputRipple [V], outputRippleLimit [V], resonantCurrent [A], reverseCurrent [A], reverseCurrentLimit [A], reverseCurrentPass [boolean], eventRate [Hz], entryLoadThreshold [fraction], exitLoadThreshold [fraction], hysteresisState [enum], transitionPass [boolean], modeLimitSourceIdentity [string]
Limitations: Mode policy requires P3 source and implementation; Model event rate cannot prove acoustic output or human audibility; Thermal and EMI consequences are outside this test
Prohibited inference: Do not claim a universal light-load threshold or audible result from the model.
Question: Does the declared controller and protection sequence limit current and recover safely across a traceable set of disturbances?
Purpose: Exercise dynamic protection actions and restart state rather than replaying a declared policy table.
Model/evidence level: switched_model
disturbance: apply one accepted scenariopolicy_parameters: inspect or change within validated boundscase_acceptance_limits: bind output-voltage resonant-current and switching-frequency limits to every required disturbancedisturbanceId [enum], outputVoltage [V], outputVoltageLimit [V], outputVoltagePass [boolean], resonantCurrent [A], resonantCurrentLimit [A], resonantCurrentPass [boolean], switchingFrequency [Hz], switchingFrequencyLimit [Hz], switchingFrequencyPass [boolean], boundSourceIdentity [string], numericBoundsPass [boolean], caseAcceptanceDecision [enum], normalizedFrequency [dimensionless], regionCue [enum], regionCrossing [event], currentLimitState [enum], protectionState [enum], restartState [enum]
Limitations: Controller and protection capability is unestablished; Simulated protection cannot certify hardware fault energy; Parasitics device SOA and thermal behavior require separate evidence
Prohibited inference: The rejected six-point frequency and protection table is not a controller design or proof.
Question: Which primary tank magnetic secondary and output-capacitor losses dominate at each validated operating point?
Purpose: Separate loss mechanisms and expose missing component evidence instead of hiding it in one efficiency number.
Model/evidence level: component_model
operating_point: select line and loadcomponent_candidate: select source-backed device magnetic capacitor and rectifier datamagnetics_loss_term: inspect one explicit magnetic loss mechanism without collapsing the other termssecondary_loss_term: inspect secondary conduction or switching independentlylossTermId [enum], lossTermSourceIdentity [string], lossTermModelIdentity [string], lossTermModelVersion [string], lossTermTemperature [degC], lossTermEvidenceBadge [enum], lossTermAvailability [enum], lossTermCompatibilityState [enum], primaryConductionLoss [W], turnOffLoss [W], gateDriveLoss [W], residualHardSwitchLoss [W], primaryLossSubtotal [W], resonantCapacitorLoss [W], copperLoss [W], coreLoss [W], fringingLoss [W], proximityLoss [W], magneticsLossSubtotal [W], secondaryConductionLoss [W], secondarySwitchingLoss [W], secondaryLossSubtotal [W], outputCapacitorLoss [W], unsupportedLoss [None]
Limitations: Loss models depend on component data and temperature assumptions; No thermal solution or measured efficiency is established; Unsupported mechanisms cannot be replaced by zero
Prohibited inference: A partial modeled loss sum is not measured efficiency or production thermal proof.
Question: Does the available evidence support LLC, redesign, comparison with another topology or a stop decision?
Purpose: Make the learner defend a decision from requirement and evidence records without an automatic weighted winner.
Model/evidence level: comparison_fixture
criterion: execute every required criterion for each declared candidatearchitecture_candidate: add or select a traceable architecture candidatecriticality: mark whether the criterion can gate the decisionevidence_records: attach evidence and limitationsdecision: record a justified outcomecriterionId [enum], architectureCandidateId [string], criticality [enum], criterionStatus [enum], evidenceRecordIds [list], evidenceLevel [enum], sourceIdentity [string], limitation [text], criticalGateState [enum], conclusionId [string], conclusionTraceabilityState [enum], openUncertainty [text], decision [enum], decisionRationale [text]
Limitations: This is a decision fixture not a multi-topology physics comparison; Cost density bidirectionality and paralleling inputs require external evidence; Saved rationale does not certify a product
Prohibited inference: The rejected 14 17 24 27 rank and CLLC winner are not technical truth and must never be displayed as such.
Evidence level: switched_model
Expected reading: Wrong-sign current leaves 390 V at turn-on; nominal negative current reaches the zero clamp at 58.925711111872716 ns.
Rendered SHA-256: ffa496b7dd2cf4bbd5d31442543931ae67be1856f798155728a4c901a01122d5
Limitations: Static one-edge comparison; Constant current; Lumped charge; Not a full-period waveform; Not measured evidence
Evidence level: switched_model
Expected reading: Learner can identify current polarity primary-secondary interval correspondence and each dead-time interval between complementary bridge states with the concurrent tank-excitation state without treating this plot as evidence of phasor resonance reflected-impedance RMS-loss or safe-measurement readiness.
Rendered SHA-256: f77f9ed2fda08ead7e6c833e54b874384892eae80faf222060dea2ea245d00b8
Limitations: Blocked by GAP-01; Existing shared 300 W plant cannot be inherited; No handmade waveform; Not hardware reverse-recovery evidence
Evidence level: switched_model
Expected reading: Current decomposition resonance-region classification DCM or CCM state and any zero-current secondary commutation remain explicit for each validated complete-period case under GAP-01.
Rendered SHA-256: 9def562b8bafa31d68fb279c5e89cd1403cf42cca9adc0fa658e184ea1f91153
Limitations: Blocked by GAP-01; Existing shared 300 W plant cannot be inherited; No handmade waveform; Not hardware reverse-recovery evidence
Evidence level: fha
Expected reading: Half-bridge and full-bridge fundamentals use the explicit factor-of-two transform while Rac Qe and exact fr and fp placements remain visible beside gain and region cues.
Rendered SHA-256: 728288cdd9672eb26d7e1d6d05a44389310d419ffd3adb98f5d228373c094a16
Limitations: FHA only; Most credible near series resonance; Cannot prove switching commutation transient behavior or hardware loss
Evidence level: comparison_fixture
Expected reading: Disagreement triggers evidence escalation every layer retains source model version badge limitation and availability and no trace gains authority from visual proximity.
Rendered SHA-256: 2a32f22559f48cf0e6457fa3e58b9bdaf3a8d1cc399246ebb3b3f1b86e87a5c9
Limitations: No accepted fixed reference dataset; Difference is case-specific; Bench data cannot validate unmeasured points
Evidence level: equation_screen
Expected reading: Minus 3.394104138973014 A is helpful_sufficient, minus 0.8 A is helpful_marginal, and equal positive magnitudes are visibly wrong_polarity despite identical unsigned margin.
Rendered SHA-256: caa7ba4af8963efe3eb9cc3664682e2b8eb427f5946ef68bba69fbd876cd6fe4
Limitations: Algebraic constant-current screen; Lumped Qoss; Not device or hardware proof
Evidence level: fha
Expected reading: Full-load P2 points are reproducible bridge form turns ratio and transform stay bound to every result all seven requirement corners execute or remain visibly unavailable and the 167.081 kHz conflict remains explicit.
Rendered SHA-256: c935c28676410789afd3c59ade6b1fb5dd724735a961899ade8f53babc277509
Limitations: FHA design screen; P2 Chapter 1 does not settle the wider Module 4 5 and 8 trajectory
Evidence level: switched_model
Expected reading: Both no-prebias and declared-prebias cases retain ordered region states and crossings with visible output current and frequency limits and a fail-closed per-case decision.
Rendered SHA-256: 690f763e04808e4726cafe8ee72beb10a019b1a20b3109ff0cb9642492a394c1
Limitations: Plant and soft-start policy are P3 requirements; Not hardware startup proof
Evidence level: fha
Expected reading: A valid first-pass design preserves fr Ln Qe and physical components and passes recomputation gain inductive-angle and supported stress gates together; one formula output is not a design decision.
Rendered SHA-256: d6f7c861883598b3c086068764929d527c99c693ab1c96c6ef3ba66c84f7af53
Limitations: First-pass analytical and FHA screens; Component temperature and construction require later evidence
Evidence level: fha
Expected reading: Every source-bound corner preserves metric support states and computes its controlling constraint before a shared-identity before or after candidate decision is accepted or rejected.
Rendered SHA-256: 28d34066ee1fa3c4fd5270f5b2f9d3b776f29e7ee51219b13cb1bd3f484740e8
Limitations: Coverage only matches the declared corner set; Monte Carlo claims require an explicit distribution and seed
Evidence level: comparison_fixture
Expected reading: Only an admitted source-backed allocation that maps each leakage contribution once and makes total modeled series inductance equal the declared target can update resonance gain current and ZVS screens; double counting is visibly rejected.
Rendered SHA-256: 01f115fd7e85fc888578c1ceff8a642585d0de6e938a543cbc6deaa264734c46
Limitations: P2 freezes no measured magnetics; Imported values are unavailable until provenance is supplied; Detailed magnetic design is out of scope
Evidence level: switched_model
Expected reading: Rectifier topology and source-declared load-step identity stay bound to pre-step and post-step operating points and event origin; output response settling ripple and capacitor-current limits jointly gate a fail-closed load-response decision.
Rendered SHA-256: ac32a6e8ec154ac51a39b5f3d97d0648e63bf5acc53562f2df443a532900f218
Limitations: Requires validated secondary and output plant; Device reverse recovery needs component or hardware evidence
Evidence level: switched_model
Expected reading: Device identity Rds on and capacitance stay bound to one common timing and tank case; reflected capacitance and baseline-versus-candidate tank-current burden remain visible so low resistance cannot hide tank body-diode capacitance or reverse-current rejection.
Rendered SHA-256: 8af93e7e95a65031ba65c7f5f89486072f7224ef786cee35601fa2aee486e459
Limitations: Timing screen depends on the declared secondary model; Driver propagation and PCB effects require component or bench evidence
Evidence level: fha
Expected reading: Each concept exposes what it senses commands and limits before its trajectory is compared; the P2 high-line light-load 167.081 kHz finding still prevents inheritance of a 135 kHz endpoint.
Rendered SHA-256: 468ffd214177dd1a13511cfc0ac4c3f3b7dfb342e7738cce6869916dd73d404b
Limitations: FHA trajectory; Wider operating policy is unresolved; Not loop-stability evidence
Evidence level: comparison_fixture
Expected reading: Crossover bandwidth and phase margin move with operating point while source badge and availability keep model and measured decisions distinct and absent bench data remains absent.
Rendered SHA-256: 84b1b1b55cc5a78da335fa1415bb5cda1a7e9455b0aeb2283b6cb3e6bfd1ff57
Limitations: No measured loop data is frozen; Measurement fixture and injection limits are P3 or later obligations
Evidence level: switched_model
Expected reading: Each supported mode preserves entry and exit thresholds and hysteresis order while SR-disabled acceptance is tied to reverse-current and ripple bounds without claiming audible or EMI compliance.
Rendered SHA-256: f71662595d9ef55775b93cdc023a60f72fae058df0ac6958a69d0b56cd9d75e5
Limitations: Audible and EMI behavior require physical evidence; Controller-specific policies are not frozen
Evidence level: switched_model
Expected reading: Every disturbance reports source-bound voltage current and frequency comparisons beside exact region protection and restart order and fails closed when a required bound is unavailable.
Rendered SHA-256: f619963214cb262d26f4e82a969037b65cbcea4fe7857fb03bec3bd7ec2ab0a5
Limitations: Controller and thermal behavior are abstracted; Simulation cannot establish safe hardware fault energy
Evidence level: comparison_fixture
Expected reading: Every separated loss term retains its source model version temperature badge availability and compatibility tuple and subtotals accept only compatible available terms.
Rendered SHA-256: 2795a51619f39d335d1425006a9a93d77f82fc8ce9c3f04488a58b7617356a5b
Limitations: No complete measured loss map is frozen; Temperature and parasitics require provenance-bound models or hardware
