---
id: opt-07
title: "Part 2 Review - Optimization & Power Transfer"
section: "Optimization & Simulation"
difficulty: "intermediate"
estimated_time: 60
prerequisites: ["opt-01", "opt-02", "opt-03", "opt-04", "opt-05", "opt-06"]
objectives:
  - Synthesize concepts from all optimization lessons
  - Apply multiple techniques to comprehensive design problems
  - Troubleshoot common optimization errors
  - Build complete optimization workflow
tags: ["review", "comprehensive", "integration", "design"]
---

# Part 2 Review - Optimization & Power Transfer

This lesson integrates all concepts from Part 2, providing comprehensive exercises that require applying multiple techniques together.

## Part 2 Summary: Key Concepts

### Lesson 1: The Two Critical Resistances

**R_opt_phase:**
```
R_opt_phase = 1 / [ω√(C_mut(C_mut + C_sh))]
```
- Minimizes impedance phase angle magnitude
- Achieves φ_Z,min = -atan(2√[r(1+r)])
- Makes impedance "most resistive" possible

**R_opt_power:**
```
R_opt_power = 1 / [ω(C_mut + C_sh)]
```
- Maximizes real power transfer to load
- Always smaller than R_opt_phase
- Typical ratio: R_opt_power ≈ 0.5-0.7 × R_opt_phase

**Topological constraint:**
```
If r = C_mut/C_sh > 0.207:
  Cannot achieve φ_Z = -45° (inherently capacitive)

Most Tesla coils: r = 0.5 to 2.0 → φ_Z,min = -60° to -80°
```

### Lesson 2: The Hungry Streamer

**Self-optimization mechanism:**
1. Power → Joule heating
2. Temperature → Ionization (exp(-E_i/kT))
3. Ionization → Conductivity (σ ∝ n_e)
4. Conductivity → Resistance (R = L/σA)
5. Resistance → Circuit power
6. **Feedback stabilizes at R ≈ R_opt_power**

**Time scales:**
- Thermal response: 0.1-1 ms (thin channels)
- Ionization response: μs to ms
- Can track kHz modulation, not RF cycles

**Physical limits:**
- R_min ≈ 1-10 kΩ (maximum conductivity)
- R_max ≈ 100 kΩ to 100 MΩ (minimum conductivity)
- Source limitations prevent optimization if insufficient power

### Lesson 3-4: Thévenin Equivalent

**Extraction:**
```
Z_th: Drive OFF, apply 1V test, measure I_test
      Z_th = 1V / I_test = R_th + jX_th

V_th: Drive ON, no load, measure V_topload
```

**Using the equivalent:**
```
I = V_th / (Z_th + Z_load)
V_load = V_th × Z_load / (Z_th + Z_load)
P_load = 0.5 × |I|² × Re{Z_load}
P_load = 0.5 × |V_th|² × Re{Z_load} / |Z_th + Z_load|²
```

**Maximum power (conjugate match):**
```
Z_load = Z_th* → P_max = |V_th|² / (8 R_th)

Usually unachievable due to topological constraints!
```

### Lesson 5: Direct Measurement

**Alternative to Thévenin:**
- Keep full coupled model
- Measure power in spark directly
- Sweep R, find maximum
- Slower but handles nonlinearity

**Best practice:**
- Use Thévenin for exploration
- Validate with direct measurement

### Lesson 6: Frequency Tracking

**Critical principle:**
```
For each R value, retune to loaded pole frequency!
```

**Why:**
- Loading changes C_sh → shifts resonance
- Typical shift: 10-30 kHz for medium sparks
- Fixed-frequency comparison measures detuning, not matching

**Loaded frequency:**
```
f_loaded = f₀ × √(C_total,0 / C_total,loaded)

C_total,loaded = C_total,0 + C_sh
```

**DRSSTC modes:**
- Fixed frequency: Simple, but detunes with loading
- PLL tracking: Optimal, adapts in real-time
- Programmed sweep: Compromise

## Comprehensive Design Exercise

**Scenario:** You're optimizing a medium DRSSTC for a 3-foot spark target.

**Given System Parameters:**
- Operating frequency: f ≈ 190 kHz (to be refined)
- Topload: C_topload = 30 pF (measured)
- Target spark: 3 feet
- FEMM analysis gives: C_mut = 9 pF for 3-foot spark
- Secondary stray capacitance: C_stray = 5 pF
- Thévenin measurement (unloaded): Z_th = 105 - j2100 Ω at 200 kHz, V_th = 320 kV

**Your tasks:** Work through the complete optimization workflow.

---

### Task 1: Estimate Spark Capacitance

Using the 2 pF/foot rule:

**Question 1a:** What is C_sh for a 3-foot spark?

**Question 1b:** What is the total secondary capacitance (unloaded)?

**Question 1c:** What is the total capacitance with the 3-foot spark?

---

### Task 2: Calculate Loaded Frequency

**Question 2a:** If unloaded resonance is f₀ = 200 kHz, calculate the loaded resonance frequency with the 3-foot spark.

**Question 2b:** What is the frequency shift Δf?

**Question 2c:** If you operated at fixed f = 200 kHz (unloaded resonance), how detuned would you be? Express as a percentage of the original frequency.

---

### Task 3: Determine Optimal Resistances

**Question 3a:** Calculate R_opt_power at the loaded frequency (use result from Task 2).

**Question 3b:** Calculate R_opt_phase at the loaded frequency.

**Question 3c:** What is the ratio R_opt_power / R_opt_phase?

**Question 3d:** Calculate the capacitance ratio r = C_mut / C_sh.

**Question 3e:** Calculate the minimum achievable phase angle φ_Z,min. Can this system achieve -45°?

---

### Task 4: Build Lumped Spark Model

**Question 4a:** Draw the lumped spark circuit showing R, C_mut, and C_sh. Label all component values, using R = R_opt_power from Task 3a.

**Question 4b:** Calculate the spark admittance Y_spark at the loaded frequency. Express in rectangular form (G + jB).

**Question 4c:** Convert Y_spark to impedance Z_spark. Express in both polar and rectangular forms.

**Question 4d:** Verify that the phase angle matches expectations from the topological constraint.

---

### Task 5: Predict Performance with Thévenin

Now use the Thévenin equivalent to predict performance. Adjust Z_th for the loaded frequency:

**Note:** Z_th changes with frequency. For this exercise, assume:
- Z_th ≈ 108 - j2050 Ω at f_loaded (slightly adjusted from 200 kHz value)
- V_th ≈ 320 kV (approximately constant near resonance)

**Question 5a:** Calculate the total impedance Z_total = Z_th + Z_spark.

**Question 5b:** Calculate the current through the spark.

**Question 5c:** Calculate the voltage across the spark.

**Question 5d:** Calculate the real power dissipated in the spark.

**Question 5e:** What percentage of V_th appears across the spark? Why is this ratio so high?

---

### Task 6: Compare to Theoretical Maximum

**Question 6a:** What load impedance would give conjugate match?

**Question 6b:** Calculate P_max (maximum theoretical power with conjugate match).

**Question 6c:** What percentage of P_max is actually delivered to the spark (from Task 5d)?

**Question 6d:** Explain physically why the actual power is so much less than P_max. Why can't we achieve conjugate match?

---

### Task 7: Frequency Tracking Impact

Suppose you made a mistake and measured power at fixed f = 200 kHz instead of the loaded frequency.

**Question 7a:** Estimate the voltage penalty factor. Assume Q_loaded ≈ 40 and use:
```
Voltage_ratio ≈ 1 / √[1 + (2Q × Δf/f)²]
```

**Question 7b:** How much would the measured power differ from the correctly tracked measurement?

**Question 7c:** If you compared three different spark resistances at fixed f = 200 kHz, would you correctly identify R_opt_power? Why or why not?

---

### Task 8: Self-Optimization Analysis

**Question 8a:** Suppose the spark initially forms with R = 150 kΩ (cold streamer). Describe qualitatively what happens over the next 5-10 ms as the plasma heats up. Include R, T, σ, and P in your description.

**Question 8b:** Why does the plasma naturally evolve toward R ≈ R_opt_power?

**Question 8c:** If the calculated R_opt_power = 55 kΩ but physical limits give R_min = 80 kΩ, what would happen? Would the plasma reach R_opt_power?

**Question 8d:** In burst mode with 50 μs pulses, would you expect the plasma to reach R_opt_power? Explain using thermal time constants.

---

### Task 9: Alternative Measurement Validation

You decide to validate your Thévenin results with direct power measurement.

**Question 9a:** Describe the simulation setup for direct measurement. What components are included? What is varied?

**Question 9b:** You sweep R from 20 kΩ to 120 kΩ. For each R value, should you:
- (A) Measure at fixed f = 200 kHz?
- (B) Sweep frequency to find loaded pole, then measure?

Explain your choice.

**Question 9c:** The direct measurement gives P_max at R = 58 kΩ, while your calculation gave R_opt_power = 55 kΩ. Is this agreement acceptable? What might explain the small difference?

---

### Task 10: Design Recommendations

Based on your analysis, provide design recommendations:

**Question 10a:** What operating frequency should the DRSSTC use when driving this spark?

**Question 10b:** Should the drive use fixed frequency or frequency tracking? Justify your recommendation.

**Question 10c:** If using fixed frequency, what single frequency would you choose to balance unloaded and loaded operation?

**Question 10d:** What power level should the primary tank be designed to deliver (approximately)?

**Question 10e:** If you wanted a 4-foot spark instead, qualitatively describe how C_sh, f_loaded, R_opt_power, and delivered power would change.

---

## Troubleshooting Common Errors

### Error 1: "My calculated R_opt doesn't match simulation!"

**Possible causes:**
- Forgot to account for loaded frequency (used unloaded f₀)
- Used wrong capacitance values (forgot C_stray or miscounted C_sh)
- Simulation measured at wrong port (I_base instead of I_spark)
- Simulation didn't converge properly

**How to check:**
- Verify C_total = C_topload + C_stray + C_sh
- Verify ω = 2πf_loaded (not f₀!)
- Plot power vs R to visually confirm peak location
- Check simulation settings and convergence

### Error 2: "Power is much lower than expected!"

**Possible causes:**
- Operating at wrong frequency (detuned)
- High losses in simulation (R_th too large)
- Incorrect power measurement (forgot factor of 0.5, or using wrong current)
- Displacement currents included in measurement

**How to check:**
- Verify frequency matches loaded pole
- Check Z_th extraction (is R_th reasonable? 10-100 Ω typical)
- Verify power formula: P = 0.5 × I² × R for peak phasors
- Measure current through R specifically, not total base current

### Error 3: "Phase angle doesn't match theory!"

**Possible causes:**
- Using unloaded frequency instead of loaded
- Incorrect capacitance ratio calculation
- Measurement includes other components (not just spark)
- Non-ideal behavior (resistance not purely in parallel with C_mut)

**How to check:**
- Recalculate r = C_mut/C_sh carefully
- Verify φ_Z,min = -atan(2√[r(1+r)])
- Check measurement port (topload to ground, not base)
- Consider more complex model if simple lumped model doesn't fit

### Error 4: "Conjugate match power is impossibly high!"

**This is normal!** For Tesla coils:
- Z_th has low R_th (10-100 Ω)
- P_max = V_th²/(8R_th) can be tens or hundreds of MW
- Sparks cannot achieve conjugate match (topological constraints)
- Actual power is typically 0.01% to 1% of P_max

**Not an error** - just shows extreme impedance mismatch is fundamental to Tesla coil operation.

## Key Formulas Reference

### Optimal Resistances
```
R_opt_power = 1 / [ω(C_mut + C_sh)]
R_opt_phase = 1 / [ω√(C_mut(C_mut + C_sh))]
φ_Z,min = -atan(2√[r(1+r)]) where r = C_mut/C_sh
```

### Thévenin Equivalent
```
Z_th = 1V / I_test  (drive OFF, 1V test source)
V_th = V_topload    (drive ON, no load)
P_load = 0.5 × |V_th|² × Re{Z_load} / |Z_th + Z_load|²
P_max = |V_th|² / (8 R_th)
```

### Frequency Tracking
```
C_total,loaded = C_total,0 + C_sh
f_loaded = f₀ √(C_total,0 / C_total,loaded)
C_sh ≈ 2 pF/foot for typical sparks
```

### Lumped Model
```
Y_spark = [(G + jωC_mut) × jωC_sh] / [G + jω(C_mut + C_sh)]
where G = 1/R
```

### Power Measurement
```
P = 0.5 × |I|² × Re{Z}  (peak phasors)
P = 0.5 × Re{V × I*}     (complex power)
```

## Practice Problems - Solutions in Appendix

### Problem Set A: Quick Calculations

**A1.** Calculate R_opt_power for f = 180 kHz, C_mut = 7 pF, C_sh = 9 pF.

**A2.** A spark has r = 1.5. Calculate φ_Z,min. Can it achieve -45°?

**A3.** Z_th = 92 - j1950 Ω, V_th = 290 kV. Calculate P_max.

**A4.** Unloaded f₀ = 205 kHz, C₀ = 32 pF. A 3.5-foot spark appears. Calculate f_loaded.

**A5.** At f = 190 kHz with Q = 60, you're detuned by Δf = +8 kHz. Estimate the voltage penalty.

### Problem Set B: Integration Problems

**B1.** Complete Thévenin analysis:
- Z_th = 115 - j2300 Ω, V_th = 340 kV
- Spark: C_mut = 8 pF, C_sh = 5 pF, R = 65 kΩ, f = 188 kHz
- Find: Current, voltage, power, compare to R_opt_power

**B2.** Optimization with tracking:
- f₀ = 198 kHz unloaded, C₀ = 28 pF
- Test R = 40k, 60k, 80k with C_sh = 6 pF, C_mut = 9 pF
- Calculate f_loaded for each R
- Which R is closest to R_opt_power?

**B3.** Self-optimization timeline:
- R_opt_power = 70 kΩ, spark forms at R = 200 kΩ
- Sketch R(t), P(t), T(t) vs time from t = 0 to 15 ms
- Label key phases: initial, runaway, approach, equilibrium

### Problem Set C: Design Challenges

**C1.** Design matching for 4-foot target:
- Given: f = 185 kHz, C_topload = 35 pF, C_stray = 6 pF
- Determine: C_sh, C_total, f_loaded, R_opt_power, R_opt_phase
- Build lumped model and calculate Z_spark

**C2.** Frequency tracking implementation:
- Coil operates 170-210 kHz range
- Sparks vary from 2 to 5 feet
- Calculate frequency range needed
- Recommend: fixed frequency, PLL, or sweep?

**C3.** Troubleshooting:
- Simulation shows maximum power at R = 45 kΩ
- Analytical R_opt_power = 62 kΩ
- What could explain the discrepancy? List 3 possible causes and how to verify each.

---

## Transition to Part 3

You now have a complete toolkit for optimization and power transfer analysis:
- Understanding the two critical resistances
- Physical self-optimization mechanism
- Thévenin equivalent extraction and use
- Direct measurement validation
- Frequency tracking principles

**Part 3** builds on this foundation to explore:
- Spark growth physics and field requirements
- FEMM modeling for capacitance extraction
- Energy budgets and growth rates
- Voltage vs power limits
- Complete growth simulations

The optimization techniques from Part 2 combine with the growth physics of Part 3 to enable **full spark length prediction**.

---

## Key Takeaways

- **Two optimizations:** R_opt_power (max power) and R_opt_phase (min phase) are different
- **Self-optimization:** Plasma naturally seeks R ≈ R_opt_power via thermal feedback
- **Thévenin method:** Extract once, predict any load instantly
- **Direct measurement:** Slower but handles nonlinearity, good for validation
- **Frequency tracking is critical:** Must retune for each load to avoid detuning errors
- **Topological constraints:** Most Tesla coils cannot achieve -45°, inherently capacitive
- **Conjugate match unachievable:** Sparks operate far from theoretical maximum power
- **Complete workflow:** Capacitance → frequency → R_opt → lumped model → power prediction

## Practice

{exercise:opt-ex-07}

Work through the Comprehensive Design Exercise (Tasks 1-10) completely. Show all calculations and reasoning. Compare your results with the solutions appendix.

---
**Next Section:** [Part 3: Spark Growth Physics and FEMM Modeling](../../03-spark-physics/01-electric-fields.md)