AI summary by TwinMind
Lecture 11: Magnetics, Part 3
MIT OpenCourseWare
50m 54s
Course Introduction and Assessment Guidelines
- Textbook reference: Principles of Power Electronics 2nd edition, chapters 18 and 19-19.4.1
- New assessment format introduced: short mini-quizzes to check understanding
- Assessment guidelines:
- No collaboration allowed (unlike homeworks)
- Questions directed to instructor via email preferred over Piazza
- Core staff can only answer exam-type clarification questions
- Assessment numbers align with homework material (assessment 1 covers homework 1 material)
- Designed to be short, not 12-hour problem-solving sessions
- Will appear on Canvas or Gradescope
Transformer Circuit Model Review
- Two-winding transformer with N1 and N2 turns on high permeability core
- Magnetic circuit model incorporates:
- Main flux linking both windings (yellow component)
- Leakage flux linking winding 1 only
- Leakage flux linking winding 2 only
- Physical circuit model components:
- Magnetizing inductance (energy stored in core): finite μ creates H field, energy density b·h
- Leakage inductances: L_leakage1 and L_leakage2 represent energy stored in space by leakage flux
- Ideal transformer with physical turns ratio N1:N2
- Energy storage relationships:
- ½L_leakage1·i1² = energy in leakage flux 1
- ½L_leakage2·i2² = energy in leakage flux 2
Inductance Matrix vs Circuit Model Parameters
- Inductance matrix has 3 independent parameters: L11, L22, LM (symmetric due to reciprocity)
- Physical circuit model has 3 parameters when using physical turns ratio
- Non-physical circuit model can have 4 parameters:
- Arbitrary turns ratio selection
- Three inductance values adjusted to match terminal behavior
- Loses physical meaning but maintains mathematical accuracy
T-Model with 1:1 Transformer
- Alternative circuit representation using 1:1 ideal transformer
- Parameters: La, Lb, Lc with relationships:
- L11 - LM = La
- LM = Lb
- L22 - LM = Lc
- Advantages: easier voltage scaling analysis, direct inductance matrix mapping
- Trade-off: parameters may not represent physical quantities
Parameter Constraints for Real Devices
- Single inductor: L > 0 (energy conservation requirement)
- Proof: positive voltage step → positive current ramp → positive energy input
- Negative inductance would violate conservation of energy
- Two-port inductance matrix must be positive semi-definite
- Constraint: |LM| < √(L11·L22)
- Coupling coefficient: K = LM/√(L11·L22)
- Perfect coupling: K = 1 (no leakage flux)
Three-Winding Series Transformer
- Physical structure: three windings (N1, N2, N3) on single core path
- Magnetic circuit: three MMF sources (N1i1, N2i2, N3i3) with core reluctance
- Ideal relationships (infinite permeability):
- Voltage scaling: V1/N1 = V2/N2 = V3/N3
- Current constraint: N1i1 + N2i2 + N3i3 = 0
- Energy conservation: instantaneous power balance across all ports
- Schematic representation: multiple windings on single core with dot convention
- Non-ideal model: 3×3 inductance matrix with 6 independent parameters
Three-Winding Parallel Transformer
- Physical structure: three separate core legs with individual windings
- Magnetic circuit: three parallel reluctance paths with flux constraint
- Flux relationship: Φ1 + Φ2 + Φ3 = 0 (magnetic KCL)
- Ideal relationships:
- Voltage constraint: V1/N1 + V2/N2 + V3/N3 = 0
- Current scaling: N1i1 = N2i2 = N3i3
- Applications: current matching and splitting
- Structural dual of series transformer (voltage/current relationships swapped)
- No standard dot convention due to multiple flux paths
N-Winding Transformer Parameters
- General N-winding transformer: N(N+1)/2 independent parameters
- Parameter count progression: 1-winding (1), 2-winding (3), 3-winding (6), 4-winding (10)
- Inductance matrix remains positive semi-definite constraint
- Measurement complexity increases significantly with winding count
Extended Cantilever Model
- Purpose: circuit model with exact parameter count for terminal behavior matching
- Systematic approach for N-winding transformer modeling
- Construction method:
- Start with single inductance L11 for first port
- Add ideal transformer (1:N2) and inductance L12 for second port
- For each additional port: add ideal transformer (1:Nn) and leakage inductances to all existing nodes
- Parameter relationships enable direct measurement and inductance matrix mapping
- Advantages:
- Always provides correct parameter count
- Direct measurement procedures available
- Mathematical equivalence to inductance matrix
- Non-physical but terminal-accurate representation
Model Applications and Limitations
- Physical models: represent actual energy storage locations
- Non-physical models: mathematical convenience, may have negative inductance values
- Selection criteria: physical understanding vs. computational convenience
- Measurement-based parameter extraction possible for both approaches
